identifier
stringlengths 11
32
| pdf_url
stringlengths 17
4.62k
⌀ | lang
stringclasses 120
values | error
stringclasses 1
value | title
stringlengths 2
500
⌀ | source_name
stringlengths 1
435
⌀ | publication_year
float64 1.9k
2.02k
| license
stringclasses 3
values | word_count
int64 0
1.64M
| text
stringlengths 1
9.75M
|
|---|---|---|---|---|---|---|---|---|---|
https://openalex.org/W1980440118
|
https://zenodo.org/records/2288905/files/article.pdf
|
English
| null |
Should ringworm be notifiable?
|
Public health
| 1,910
|
public-domain
| 4,465
|
SHOULD
RINGWORM
BE
NOTIFIABLE ? He does not say how frequently." In the Administrative County of Southamp-
ton in x9o 9 there were I44 cases of Ringworm
recorded out of I5,752 children examined in
35 o schools, being an average of x in IO 9
children examined, and i in 32z children in
average attendance. It is possible that the
true average lies somewhere between these
figures, because the former includes a number
of cases specially selected as having had the
disease, and presented for examination to see
if they were really free from it, and found on
inspection not to be so. If Ringworm were as prevalent throughout
the country as it is in London, then the
number of children constantly absent from
school through it would be about 24,i6o. Taking Sir George Newman's figures how-
ever, his averages, reckoned on the basis of
average attendance, would give a total of
21,6o 7 cases in the Counties, and I5,369 in the
town areas, or a total of 36,976 for England
and Wales. And this number mxy be taken
to be the number of cases of Ringworm con-
stantly present attending school. In this country the disease is very common. Reference to the Annual Report for 19o 9 of
Sir George Newman, the Chief Medical Officer
of the Board of Education, to whom we all
offer our hearty congratulations on his recently
bestowed honour, will show us how general it
is in this country. Thus, of over fifteen
County Areas in which the statistics are
available, on an average I
in r3o children
examined at routine examinations are found to
be affected, the proportions varying from I in
63 in Flintshire and Worcestershire to I in
287 in Monmouth. If we assume that each case is, or ought to
be, absent from school for about six months,
this would represent a loss, roughly, of £35,ooo
in Government grants alone. SHOULD
RINGWORM
BE
NOTIFIABLE ? whilst Staffordshire has its highest number
in the Urban Districts. In the Boroughs the proportion of cases is
lower than in the Counties, and is approxi-
mately i in I65 on an average, the highest
proportion being found in Northampton, x in
54, and the lowest I in 2,55 ° in Bootle. (Only
2,55o children
were
examined
at
Bootle
however). I T I T is hardly necessary for me to state why I
have brought this subject before you for
consideration, considering its widely spread
nature, and the great resultant loss not only to
the individual in the loss of education and all
that this subsequently involves by reason of an
enforced absence from school, but also to
Education Authorities by reason of diminished
Government grants,
and
possibly
also to
teachers by reason of a big drop in salary that
may ensue owing to a single constant absentee
from school. Sir George Newman remarks that " these
figures must be looked upon as approximate
only. There
is considerable variation
in
different areas in the amount of care bestowed
upon the examination of the head, and a
careful examination of each child, especially
of each girl, is required if all cases of Ring-
worm are to be detected. As to the prevalence of the disease on the
Continent we are told on the authority of
Dr. Steven in his " Medical Supervision in
Schools" that "in the German Schools Ring-
worm never occurs, according to Dr. Neufert."
Morris in his "Ringworm in the light of
recent research" says " it would also appear
to be rare in Germany. KrSsing states that
Ringworm of the scalp is very seldom met
with in Breslau; Unna says it occurs in
Hamburg, though apparently it is by no
means common there." He further says " the
small spored fungus is accountable for from
60 to 65 per cent of all cases of Ringworm met
with in France. The fungus is not, however,
met with in all parts of France, for Dubreuilhqu
and Fr~che failed to find it in Bordeaux. Mibelli has never met with it in Italy. Nor had
Ducroy of Piss or Reale of Naples ever seen it
in Italy. Fergnani has met with it in Spain. * A Paper read before the Southern Branch of the Society of Medical
Officers of Health, January 6th, ~9I L PUBLIC HEALTH. PUBLIC HEALTH. 234 MARCH, SHOULD
RINGWORM
BE
NOTIFIABLE ? By J. P. WALKER, M.D., D.P.H.,
Assistant County Medical Officer of Health, Hampshire. SHOULD
RINGWORM
BE
NOTIFIABLE ? I may remark here that when reading of this
subject we are seldom assisted by a differentia-
tion in the statistics of cases of Ringworm of
the scalp from that of the body, so that when
we read in the Report above referred to that in
Sheffield it has been found that children are
absent on account of Ringworm just over five
weeks on an average, we are unable to tell how
many of these cases were of the scalp, the only
variety that causes any difficulty in its treat-
ment. A similar period was found in the Where results are given separately for
Urban and Rural Districts the results vary
somewhat, Flintshire having a higher pro-
portion among the rural
school children,
* A Paper read before the Southern Branch of the Society of Medical
Officers of Health, January 6th, ~9I L PUBLIC HEALTH. 191.1. '2.35 earlier these cases are taken in hand, the less
difficult becomes their subsequent treatment. Borough of Leicester, and also at
King's
Norton. On the other hand in the case of
the Boroughs of Hornsey and Finchley where
no child was allowed to return to school until
apparently cured, the average duration of
absence was 2o and 18 weeks respectively. In the Ringworm Schools of the Metro-
politan Asylums Board, the average stay of the
first hundred children under ordinary medical
treatment was 19 months, whilst the average
stay of the first hundred children treated by
X-rays was four-and-half months from the time
the treatment was actually begun, yet they
were not dismissed until the re-growth of hair
appeared. p
y
The Chief Medical Officer of the Board of
Education further remarks that the " Medical
Certificates do not seem free from error," as
to the length of duration of the disease. In
the London County Council's Report for 19o 9 ,
quoted by Dr. Steven, in only 4 cases out of
151 was the distinctive organism not found, in
cases where medical certificates of freedom
from disease had been given. As he rightly
says "this seems to be a grave reflection upon
the members of the medical profession." In the treatment of this condition we all
recognise the necessity for difference in the
cases of town and country school children. Thus in Bradford it had been found that five
attendances per child are necessary for X-ray
treatment. SHOULD
RINGWORM
BE
NOTIFIABLE ? Were one to adopt a series of
Centres for treatment in this County, as is
done in the Eye Clinic, an expense of at least
,£2 for conveyances would be requisite for some
of the cases, apart from other miscellaneous
expenses, and some form of portable X-ray
apparatus would be required. Dr. Meredith
Richards,
the
school
medical officer for Croydon, also bears this
out. He says "we are much hampered by the
large number of children admitted to school on
medical
certificates
as convalescent from
Ringworm, but who are subsequently found to
be in an infectious condition." I notice that the British Medical Journal of
December 3Ist, I9IO, states that the average
cost to the London County Councit of 468
cases treated at the Charing Cross, London,
and St. George's Hospitals was ,£I IS. Iod. each. In Finchley the cost of treatment by
X-rays appears to have been £21 lOS. 6d. for
22 cases, or an average roughly of £i per case
in each instance. Assuming then each case to
cost £I for treatment by X-rays, travelling ex-
penses in the case of country school children
would in some instances bring the total expense
to about £3. It is as a well known text book on the
Diseases of Children remarks " Ringworm of
the scalp is one of the most troublesome local
diseases with which the practitioner has to
deal, and one which is apt to bring unmerited
discredit on account of the many months or
even years that the disease sometimes lasts." Cases of Ringworm are often diagnosed by
the intelligent and observant teacher,
and
promptly excluded from school; others again
are not discovered until the arrival of the
school medical inspector. In each case a
letter is usually sent to the parents informing
them of the reason of the child's exclusion
from school, with a recommendation to them
to obtain medical advice for the condition. In
some cases the exclusion is for a definite time,
say for one month, whilst in others the case
has to obtain a clean Bill of Health, either
from the school medical officer or from a
private practitioner, ere being allowed to rejoin
his fellows in school. In London, though
admitted to school on a private practitioner's
certificate, he has still to pass
the school
medical officer before being allowed to mix
in a class with the others. SHOULD
RINGWORM
BE
NOTIFIABLE ? When one considers that
each parent believes that he sends his child
to school in a clean condition, and when medi-
cal inspection reveals the fact that fully one
child in every three is attending school in a
verminous condition, and when one further
considers that even after repeated notices of
this verminous condition have been sent to the
parents, accompanied in some cases by threats
of possible prosecution and fine, notices stating
too how the verminous condition can be easily
removed by inexpensive materials at hand in
nearly every household, at any rate in the coun-
try, yet a very considerable per-centage is found
to be still verminous, one can hardly be sur-
prised that the treatment of Ringworm by
drugs at considerable expense to the parents of
the scholars falls much below the point desired. We must not forget the fact that it may be
to the teacher's immediate interest to keep up
his school attendance at any cost. Teachers
do at times resent our exclusion of children,
especially when their salaries are nearing a
border limit. In many districts, head teach-
ers are paid by scale, according to the number
of children in average attendance at the school. The maximum salary may for one grade be
£22o, but a drop of even one below the hard
and fast line of reckoning may reduce that
man's salary to the grade below, viz. £17o, or
a drop of £5o per annum. Cases similar to
this have occurred. Small wonder then that
a teacher at times fails to be enthusiastic at a
medical inspector's action in excluding one or
more children, especially when his average at-
tendance approaches the danger zone; nor can
we be surprised at his looking at these cases
with non-seeing eyes when he chances to meet
with them in the first instance. True, Sir
George Newman in the Report above cited, in
reference to this question, and its effect on
teachers, says " All Local Education Author-
ities should make it clear that no penalty of
any kind will attach to a lowering of attendance
brought about by exclusions recommended on
medical grounds."
This view has not at all
times been taken by Educational Authorities,
who are occasionally anxious to save expense
whenever facts will warrant it, and quite ignor-
ing the justice or otherwise of their action in
so doing. SHOULD
RINGWORM
BE
NOTIFIABLE ? Personally I have not lost faith in the treat-
ment of Ringworm of the scalp by drugs when
properly carried out. I believe that if the same
personal medical attention were given to the
drug treatment as is given to the X-ray treat-
ment, that almost equally good results might
be obtained. As one who has had a very con-
siderable experience of both general and hospi-
tal practice, I know what different results can
be obtained when personally devoting indi-
vidual attention to the treatment of troublesome
(what are termed uninteresting) medical cases,
and I believe that equally satisfactory results
can be obtained in the treatment of Ringworm. In the Bradford School Clinic out of 29o
cases treated by X-rays, 6"97 weeks was the
average duration of each child's absence from
school, but during the last six months of I9O 9
the average duration of treatment was four
weeks. It is common knowledge that the Dr. Forbes in an article in Public Health in
September 19o 9 records I33 cases of Ringworm
cured at the Brighton Clinic. The total ex-
pense of treating these, together with 223 other
contagious skin cases was £2 I3S. 5 d. out of C
I PUBLIC
ItEdZTtt. 236 MARCH~ MARCH~ which parents paid £2 7s. 5d. leaving a deficit
of 6s. of 134 cases of Ringworm excluded by myself
I found that 29 cases had returned to school
without being cured. I have very little faith, however, in the people
who now administer, or should administer, the
drugs. As a rule treatment by drugs is entrust-
ed to the average parent with but little direction
and less supervision. SHOULD
RINGWORM
BE
NOTIFIABLE ? Nor apparently is it clear from the
above statement whether it covers exclusions
for Ringworm or other troubles by the teacher
himself acting on his own judgment. It may
be that only exclusions ordered by the school
medical officer himself will count in this con-
nection, and it is improbable that records of
exclusions by the teacher on his own account
for such a reason are recorded at Head Offices
in his favour. It is of course possible that they
maybe so recorded, but the teacher has no
opportunity of knowing this, or of pleading his
case personally when in trouble, as he never
meets his paymasters face to face. On the other hand, competition for prizes
d d f
d
tt
d
ill l
d
t p
The bogey of expense is ever present before
the average parent,hence he limits treatment
to the lowest possible minimum. Instead of
going to the Doctor, he often visits the cheap
prescribing chemist, whose custom it is to sell
something which injures neither the child's
skin, nor yet the offending mycelium and spore. Should the medical man be consulted, the visits
to him are often confined to chance visits to
the town, much valuable time being lost ere he
is consulted, and after being once consulted,
and having once supplied an ointment, it is not
at all uncommon for him to lose sight of the
case for months, or for him to have no further
opportunity of seeing how the case is progress-
ing. The parent uses the ointment sparingly,
does not remove the old material when apply-
ing the new, and when the first supply is ex-
hausted, recourse is then often had to the
various advertised remedies, all of them styled
specifics, and most of them as useless as they
are expensive. Many cases return now to school without
having received any treatment whatever, not
even an application from the homely inkpot. The artful parent at times carefully plastering
down the hair to cover the bare place on the
child's head, it may pass a superficial glance
bestowed on it by the teacher, and it becomes
a matter of " out of sight, out of mind," until
possibly the succeeding visit some three or four
"months later of the school medical officer. SHOULD
RINGWORM
BE
NOTIFIABLE ? Out On the other hand, competition for prizes
awarded for good attendance will lead some to
attend school when not fit to do so, and
the Chief Medical Officer recommends that a
school medical officer's certificate of exclusion
should be no bar in a competition of this kind. Here nothing is said as to exclusionby ateacher, PUBLIC HEALTH. 1911. 237 a point that may have been accidentally or pur-
posely omitted. He would admit cases to school provided there
is systematic treatment, and
he
suggests
the wearing of a linen skull cap at school,
and making the Education Authority respon-
sible for the
quarantine, but not
for the
treatment, though the teacher has to
be
satisfied that some approved medicament is
being applied daily to the child's head. There is moreover the loss of Government
grant to be considered, a matter of I6S. to I7S. for infants, and els. to a2s. for elder child-
ren. At
Bitterne School
in this county
Dr. Lyster in his Annual Report for I9o 9 re-
cords that "33 children were excluded for 4,o47
attendances." This would be equivalent to a loss
of over £8 for the one school. I do not wish to recommend any particular
line of treatment, but I should like to mention a
point emphasised by Sabouraud, and that is
the importance of applying Tincture of Iodine,
I in 5, or t in IO, to the whole scalp every
night owing to
its indefinitely inoculating
healthy scalp areas, and further the wisdom of
applying some mild parasiticide for some con-
siderable time even after the disease appears to
be entirely eradicated. The Lancet stated last year that Dr. James
Kerr estimated the loss of Government Grant
to the London County Council by reason of
absences from school through this disease to
amount to £5,654 per annum. This deficiency has to be made up out
of the rates, so the expense has to be made
good by the ratepayer. Hence, the ratepayer
should not grumble if he is asked to meet
the expense incurred in its treatment. SHOULD
RINGWORM
BE
NOTIFIABLE ? To overcome the difficulties of treatment,
and for the prompt and adequate prevention of
the spread of the disease, and as a protection
to other children, whether in school or other-
wise, I think it highly desirable that cases of
Ringa orm should be brought early to the
cognisance of the district medical officer of
health, and that on him should be placed the
responsibility not only for the prevention of its
spread, but also of its treatment. For this
purpose there should be : -- With a view to stamping out the disease
some authorities place great reliance on ex-
clusion from school. This has been tried for
many years and found wanting; even where
medical inspection has been in vogue for many
years there is little or no evidence, so far as
I have been able to ascertain, of any diminu-
tion in the frequency Of the disease. Exclusion
may mean that you are playing into the hands
of the careless parent who is often only too
glad to have some one at home to mind the
baby. It offers no safeguard whatever to the
children the child may play with out of school
hours. It does not prevent him attending
Sunday schools or entertainments, or from
hanging up his infective hat on common pegs,
nor does it ensure him getting any treatment
whatever. I. Compulsory notification of all cases by
teachers, school nurses, and medical
practitioners. 2. Prompt systematic home visitation with
personal examination of the patient
by the local medical officer of health
in all cases, with
disinfection, or
replacement, of articles of clothing. Animals about the house should also
receive attention. In the City of New York it is not the
custom to exclude from school children suffer-
ing from Ringworm. In Bristol the plan of
allowing cases of Ringworm to attend school
was tried for a time, but Dr. Green,
the
school medical officer, tells me the plan has
been given up recently as it has been found
extremely difficult to ehsure the carrying out
of the necessary conditions that will permit of
this being done, viz., sifting apart, and wearing
close-fitting caps with a paper lining. 3. SHOULD
RINGWORM
BE
NOTIFIABLE ? p
The Education Authority would benefit by
an increase in its Government grant, and event-
ually the ratepayer also, for this seems to me
the most promising method of combating this
widespread disorder. I do not think that considering the low
infectiveness of this disease total exclusion from
school until quite cured is always requisite. Out of 81 schools affected by it in 19o8 , I9C 9,
and I9IO, only 22 show fresh cases, apart from
other members of the same family; moreover,
out of 8 schools affected in 19o8 in this County,
7 remained free from Ringworm in 191o. I believe the greatest failing at the present
time lies in these cases not being followed up
to the home with all speed as soon as they are
discovered. I will not advocate any method
of cure as I believe most of the methods ad-
vocated will be satisfactory if properly carried
out, for no matter how able the physician, if
the treatment is not thorough, is not continuous,
and not founded on the results of microscopical
investigation, we cannot expect satisfactory
results. Of 31 schools affected in 19o9, 16 had be-
come quite free in 191o. District medical
officers of health in this County are ten times
more numerous than school medical officers,
and as they are constantly moving about their
districts, it is no trouble to them to see these
cases when on their rounds, whilst it means
considerable expense, and perhaps the loss of
a half-day for a school medical officer to ex-
amine one special country scholar. We must not forget that though but rarely
does this disease affect a child's general health
yet it does however seriously affect his educa-
tion, and all that that implies. Therefore
it is desirable he should be returned
to
school as speedily as possible, as soon as that
can safely be done without undue risk to other
children. I believe this may quite safely
be done with but little absence from school if
adequate precautions are taken in the home,
and also in the school, and if treatment is
begun early. SHOULD
RINGWORM
BE
NOTIFIABLE ? Where no treatment has been adopted,
or where unsatisfactory treatment is
manifest, either lay or medical, then
it should be within the province of
the local medical officer of health
to offer absolutely free treatment, just
as the public vaccinator now offers
his services. He should be empowered
to deal with the case by segregation if
necessary just as in the case of a
Scarlet Fever patient. 3. Where no treatment has been adopted,
or where unsatisfactory treatment is
manifest, either lay or medical, then
it should be within the province of
the local medical officer of health
to offer absolutely free treatment, just
as the public vaccinator now offers
his services. He should be empowered
to deal with the case by segregation if
necessary just as in the case of a
Scarlet Fever patient. Another plan spoken of by the senior school
medical inspector for Staffordshire is that of
permitting attendance of the sufferers from the
disease, providing the case is systematically
under treatment, the arguments in its favour
being that Ringworm is a disease of slight
infectiveness, and being easily preventable. 4. He should satisfy himself by systematic
microscopical examination that every
case is fit to leave off all parasitic
medicaments before the case passes
from under his care. C 2 PUBLIC HEALTH. 238 MAte.OH, With suitable parasiticides employed, and
other due precautions, the fear of a spread of
the disease would be eliminated, the child
might return early to school, and thus lose but
little of his education. 5. The expense of any notification or treat-
ment thus required should fail upon
the Local Authority, just as in the
case of an infectious disease. 5. The expense of any notification or treat-
ment thus required should fail upon
the Local Authority, just as in the
case of an infectious disease. 6. The abolition of any possible effect on a
teacher's salary, or on a school grant
by reason of any exclusion from
school on account of a child's suffering
from this complaint. 6. The abolition of any possible effect on a
teacher's salary, or on a school grant
by reason of any exclusion from
school on account of a child's suffering
from this complaint. Finally the medical man would be assured of
payment for his service in every case, and
would be able to give it until the cure was com-
plete. SHOULD
RINGWORM
BE
NOTIFIABLE ? I do not consider that we are justified in
asking head teachers to extract hairs for us, so
that the specimens can be submitted to micro-
scopical examination ; this is altogether outside
their province, and the hairs to be selected for
epilation for such a purpose are matters rather
for the skilled practitioner. This plan has been
suggested, but it does not equal the method o[
having the work done by the medical officer
of health, as the case would then be under
direct medical control from first to last, and
the school nurse, or district nurse, acting under
his immediate direction, and in the first instance
in his presence, would show the responsible
parent or guardian how to carry out the neces-
sary treatment in cases where drug treatment
was being applied. The Chief Medical Officer of the Board of
Education is a firm advocate of total exclusion. He remarks however that "what appears to be
required, is that each Authority should adopt
a consistent line of action calculated to lead to
the banishment of the disease from their area."
Whilst Ringworm is prevalent not only in the
large animals, but also in the small domesti-
cated animals, the entire stamping out of this
disease in the immediate future seems beyond
our reasonable expectation. I do not think the
policy of exclusion as at present adopted likely
to lead to the desired result. The most con-
sistent and promising line of action seems to
me the adoption of the course I have here
outlined. g
pp
There would be every inducement to the
teacher to send the case early for treatment as
soon as the first case was noticed. The parent
for his part would not be able to keep a child
at home to mind the baby and avoid treatment. The child would be treated straight away,
and there would be no thought of a Doctor's
bill to worry the
parents, so they would
send the case early for treatment, and it would
be kept under continuous treatment until
quite well, the cure consequently being more
rapid. MICE AND MEAT.--A. Schellhorn reports in the
Centbl flit Bakt. that 5o per cent. of mice fed
with healthy germ-free meat died in from 3 to 5 days.
|
https://openalex.org/W2910188194
|
https://periodicos.ufpa.br/index.php/nra/article/download/6495/5217
|
Portuguese
| null |
REFLEXÕES SOBRE MARIO VARGAS LLOSA E A CONSTRUÇÃO DISCURSIVA DA AMAZÔNIA PERUANA EM EL HABLADOR
|
Nova Revista Amazônica
| 2,018
|
cc-by-sa
| 6,470
|
39 39 DOSSIÊ AMAZÔNIA DOSSIÊ AMAZÔNIA RESUMO A Amazônia tem uma dimensão discursiva e sua representatividade se constitui a partir das criações
simbólico-literárias que por meio dos romances contribuem para a configuração imagética dessa
região. Destaca-se neste estudo a narrativa etnográfica de Mario Vargas Llosa em sua obra El
Hablador (1987). Nela é descrito um grupo indígena da Amazônia peruana a partir de duas visões, a
do indígena e a de um jovem escritor do Peru que se encontra em Florença. Esta reflexão objetiva a
análise das duas perspectivas apresentadas por Vargas Llosa com o intuito de ler a obra como
enunciação sobre a Selva Amazônica peruana relacionada ao homem local, ao espaço geográfico, às
tradições culturais e à natureza. El Hablador pode ser considerada uma narrativa de cunho etnográfico
pois revela uma representação do indígena construída a partir dos procedimentos etnográficos de
observação, recolecção e registro de dados. Palavras-chave: Amazônia. Relato etnográfico. Mario Vargas Llosa REFLEXÕES SOBRE MARIO VARGAS LLOSA E A CONSTRUÇÃO DISCURSIVA
DA AMAZÔNIA PERUANA EM EL HABLADOR Ximena Antonia Díaz Merino1 Ximena Antonia Díaz Merino1 ϭPossui Estágio Pós-doutoral em História pela Universidade do Estado do Rio de Janeiro/UERJ (2017). Doutora
e Mestre em Letras Neolatinas opção Literaturas Hispânicas pela Universidade Federal do Rio de Janeiro/UFRJ.
Atualmente é Professora Adjunta de Culturas e Literaturas Hispânicas do Departamento de Letras do Instituto
Multidisciplinar da Universidade Federal Rural do Rio de Janeiro/UFRRJ. E-mail: ximenadm2@gmail.com j
p
p
Multidisciplinar da Universidade Federal Rural do Rio de Janeiro/UFRRJ. E-mail: ximenadm2@gmail.c ϭPossui Estágio Pós-doutoral em História pela Universidade do Estado do Rio de Janeiro/UERJ (2017). Doutora e Mestre em Letras Neolatinas opção Literaturas Hispânicas pela Universidade Federal do Rio de Janeir
Atualmente é Professora Adjunta de Culturas e Literaturas Hispânicas do Departamento de Letras do ϭPossui Estágio Pós-doutoral em História pela Universidade do Estado do Rio de Janeiro/UERJ (2017). Doutora
e Mestre em Letras Neolatinas opção Literaturas Hispânicas pela Universidade Federal do Rio de Janeiro/UFRJ.
Atualmente é Professora Adjunta de Culturas e Literaturas Hispânicas do Departamento de Letras do Instituto ABSTRACT Amazonia is a discursive construction and its representativeness is generated from the symbolic-
literary creations through the novels that contributed to the imaginary recreation of the region. This
work highlights the ethnographic narrative from Mario Vargas Llosa in his book El Hablador (1987). This Narrative describes an indigenous group from Peruvian Amazonia from two perspectives,
indigenous community and from a young Peruvian writer, in Florence. This reflection aims to analyse
the two optics, presented by Vargas Llosa, as an enunciation about the Peruvian Amazonian Forest
assimilating it to local community, geographic area, cultural traditions and to the nature itself. El
Hablador can be considered an ethnographic narrative as it reveals a representation of indigenous man
clearly constructed by ethnographic observation and by data research and collection;
Key words: Amazonia; Ethnographic representation; Mario Vargas Llosa. Desde mis frustrados intentos a
comienzos de los años sesenta de escribir
una historia sobre los habladores
machiguengas, el tema había seguido
siempre rondándome. Volvía, cada cierto
tiempo, como un viejo amor nunca
apagado del todo, cuyas brasas se
encienden de pronto en una llamarada. Había seguido tomando notas y
garabateando borradores que
invariablemente rompía. (VARGAS
LLOSA, 1995, p. 61) O estudo etnográfico possibilita o resgate e preservação das culturas das sociedades
em processo de extinção, pois envolve diversas maneiras de pensar e de escrever desde a
perspectiva da observação participativa, além de permitir a inserção do observador na cultura NOVA REVISTA AMAZÔNICA - VOLUME VI - NÚMERO ESPECIAL - DEZEMBRO 2018- ISSN: 2318-1346 40 objeto enquanto olha para essa cultura. A etnografia envolve um conhecimento com um duplo
fim: por um lado, serve de base para as ciências humanas, por outro, garante o registro da
memória das culturas ágrafas. Este segundo aspeto constitui para James Clifford uma das
pedras angulares da antropologia do século XX: a construção retórica das culturas em
processo de extinção. Levando em consideração que as sociedades indígenas em geral têm
enfrentado o implacável progresso urbano e a unificação nacional dos territórios em que estão
inseridas e que essas imposições têm resultado no aniquilamento de suas de línguas,
cosmologias e valores culturais, o estudo etnográfico se apresenta como alternativa para sua
preservação. Os relatos específicos contidos nas etnografias jamais podem ser limitados a um
projeto de descrição científica, na medida em que a tarefa principal do trabalho é
tornar o comportamento de um modo de vida diferente humanamente compreensível
[...] O que se vê num relato etnográfico coerente, a construção figurada do outro,
está conectado numa dupla estrutura contínua, com a qual se entende [...] O
comportamento estranho é retratado como significativo dentro de uma rede comum ABSTRACT Ao longo da história, o etnógrafo preocupado em não ser confundido com o
missionário, o viajante ou o funcionário do governo colonial, buscou construir em torno de si
mesmo a aura de uma experiência voltada para um conhecimento que se pretende objetivo. Para isso, uma estratégia textual decisiva orientava sobre a necessidade do apagamento dos
indícios que pudessem macular a pureza do encontro entre o pesquisador e os nativos: nada
era dito sobre os preparativos da expedição, sobre a eventual posição de força do governo
colonial propiciando a estada do etnógrafo, ou sobre as interferências decisivas dos
informantes nativos. E ficava de fora, principalmente, o intenso processo subjetivo, pleno de
ambivalências, vivenciado pelo etnógrafo. O caráter negociado, polifônico, do conhecimento
produzido em campo dava lugar no texto a um monólogo autoral com um mínimo de fissuras. Pressionado pelas transformações decorrentes da descolonização e da emergência dos
movimentos das minorias, esse modelo começou a viver seu ocaso a partir dos anos 60 do
século XX. Nesse novo cenário, marcado por um intenso translado de povos, de experiências
e saberes, produziu-se uma fabulosa multiplicação de vozes, e o monólogo que caracterizou a
etnografia até então, passou a soar como um anacronismo. Apoiando-nos novamente em James Clifford, o etnógrafo é um criador que a partir de
seus apontamentos constrói ficções de novos mundos com a finalidade de representar o que
está observando, motivo pelo qual as figuras do etnógrafo e do novelista podem ser tidas
como equiparáveis, já que os dois compõem, por meio de símbolos e de uma técnica
compartilhada, o imaginário do Outro. Argumenta Clifford: DOSSIÊ AMAZÔNIA 41 de símbolos – uma base comum de atividade compreensível válida para o
observador, e para o observado e por implicação para todos os grupos humanos. (CLIFFORD, 2002, p. 67-68). de símbolos – uma base comum de atividade compreensível válida para o
observador, e para o observado e por implicação para todos os grupos humanos. (CLIFFORD, 2002, p. 67-68). O crítico cubano Roberto González Echevarría defende que, a partir da segunda
década do século XX, o discurso antropológico, enquanto relato científico ocidental
dominante, influenciou as formas e o conteúdo da narrativa latino-americana. Para González
Echevarría, o propósito da etnografia como técnica da antropologia transferiu-se para a
literatura como mecanismo para criar um conhecimento da natureza, do primitivo e desse
Outro desconhecido no ocidentalizado. – Que esas culturas [indígenas] deben ser respetadas –dijo [Mascarita], suavemente,
como si, por fin, comenzara a serenarse. Y la única manera de respetarlas es no
acercarse a ellas. No tocarlas. Nuestra cultura es demasiado fuerte, demasiado
agresiva. Lo que toca, lo devora. Hay que dejarlas en paz. ¿No han demostrado de
sobra que tienen derecho a seguir siendo lo que son? – Eres un indigenista
cuadriculado, Mascarita – le tomé el pelo – (VARGAS LLOSA, 1995, p.39). ABSTRACT Por outro lado, um anônimo contador de histórias, um hablador, memória viva dos
indígenas machiguengas da Amazônia peruana, narra poeticamente a sua própria existência e
a história e mitos de seu povo: Después, los hombres de la tierra echaron a andar, derecho hacia el sol que caía. Antes, permanecían quietos ellos también. El sol, su ojo del cielo, estaba fijo. Desvelado, siempre abierto, mirándonos, entibiaba el mundo. Su luz, aunque
fuertísima, Tasurinchi la podía resistir. No había daño, no había viento, no había
lluvia. Las mujeres parían niños puros. Si Tasurinchi quería comer, hundía la mano
en el río y sacaba, coleteando, un sábalo; o, disparando la flecha sin apuntar, daba
unos pasos por el monte y pronto se tropezaba con una pavita, una perdiz o un
trompetero flechados. Nunca faltaba qué comer. No había guerra. Los ríos
desbordaban de peces y los bosques de animales. Los mashcos no existían. Los
hombres de la tierra eran fuertes, sabios, serenos y unidos. Estaban quietos y sin
rabia. Antes que después (VARGAS LLOSA, 1995, p.16). Como é possível observar no excerto anterior, esse hablador concentra em sua voz a
memória viva de sua comunidade e aparece como uma figura que causa grande atração nos
membros da tribo, que ficam como que hipnotizados durante horas sem perceber a passagem
do tempo. A figura do hablador surge na narrativa a partir da conversa entre o jovem peruano
e os linguistas Schneil por ocasião da primeira viagem à selva amazônica em 1958. Os
linguistas descrevem os habladores como: […] ese personaje raro, que no parece curandero ni sacerdote – dijo, de pronto, la
señora Schneil. […] –Tal vez, conversador. O, más bien, hablador – dijo, al fin. Y
pronunció de nuevo el ruido: bronco, sibilante, larguísimo. – Sí – sonrió él –. Creo
que es lo más aproximado. Hablador. Nunca habían visto a ninguno. Por su
puntillosa discreción – su temor a irritarlos– nunca habían pedido a sus huéspedes
una explicación detallada sobre las funciones que cumplía entre los machiguengas,
ni que les precisaran si se trataba de uno o de muchos, o, incluso, aunque tendían a
descartar esta hipótesis, si, en vez de seres concretos y contemporáneos, se trataba
de alguien fabuloso, como Kientibakori, patrón de los demonios y creador de todo lo
ponzoñoso e incomestible (VARGAS LLOSA, 1995, p.16). No decorrer da narrativa, Mascarita vai se afastando de seu amigo até desaparecer. ABSTRACT Em linha com o que se afirmou anteriormente, Mario Vargas Llosa, em El Hablador
(1974), desafia a possibilidade antropológica científica de interpretar uma cultura, bem como
a subjetividade que implica a relação do objeto estudado e o sujeito observador. Cabe destacar
que nessa narrativa Vargas Llosa se utiliza das experiências adquiridas em suas viagens à
Amazônia. O escritor peruano labora a figura do indígena peruano a partir das técnicas
etnográficas de observação, recolecção e registro de dados. A obra se constitui num informe
antropológico, num diário de campo, que procura descobrir a origem dos sistemas sociais das
culturas latino-americanas primitivas e sua relação com o mundo ocidental. Da mesma
maneira que na etnografia, em El Hablador encontramos dois tipos de observadores que se
configuram nos narradores da cultura machiguenga: um narrador direto participante, um
hablador, narrador ambulante da tribo machiguenga, e um narrador anônimo auto ficcional,
um jovem escritor peruano, duas vozes que se alternam para relatar os dois lados de uma
mesma história. Por um lado, o narrador anônimo auto ficcional, que se identificaria com o próprio
novelista, evoca as lembranças de um amigo de juventude que era chamado pelo apelido de
Mascarita que estava fascinado pela cultura amazônica, como se ilustra no seguinte
fragmento de El Hablador: Na citação Saúl Zuratas, o Mascarita, revela uma forte preocupação pela preservação
da cultura indígena. O apelido Mascarita apresenta dentro da obra um duplo sentido, faz
referência à mancha escura que cobria a metade de seu rosto e à dupla personalidade que
oculta e que somente é revelada ao final do texto. De acordo com a narrativa, Mascarita tinha NOVA REVISTA AMAZÔNICA - VOLUME VI - NÚMERO ESPECIAL - DEZEMBRO 2018- ISSN: 2318-1346 42 deixado a Faculdade de Direito para dedicar-se ao curso de Etnologia. Como consequência de
suas constantes viagens à Amazônia, Saúl Zuratas diz: “Lo que se está haciendo en la
Amazonía es un crimen […] ¿Adónde se pueden seguir yendo? Los empujan de sus tierras
desde hace siglos, los echan cada vez más adentro, más adentro. Lo extraordinario es que, a
pesar de tantas calamidades, no hayan desaparecido. Ahí están siempre, resistiendo”
(VARGAS LLOSA, 1995, p. 9). ABSTRACT O
jovem escritor indaga sobre o paradeiro de Saúl, mas não consegue informação. Ele dedica-se
também ao estudo da tribo amazônica que tanto obcecava seu amigo, desde Florença, esse
narrador auto ficcional constrói o perfil do indígena a partir de informes e testemunhos de DOSSIÊ AMAZÔNIA 43 outras pessoas, ou seja, desde um locus enunciativo distanciado do objeto de estudo: primeiro
desde a Itália e depois a partir de uma fotografia que retrata um grupo indígena Machiguenga. Interage com científicos, antropólogos, etnólogos e linguistas que lhe proporcionam
informações para seu estudo. Sobre sua primeira visita à Amazônia no ano de 1958 revela: […] el recuerdo más memorable y recurrente de aquel viaje sería lo que les oí
contar, en Yarinacocha, a una pareja de lingüistas: los esposos Schneil. Al principio,
me pareció que era la primera vez que oía nombrar a aquella tribu. Pero, de pronto,
me di cuenta que era la misma sobre la que había oído tantas historias a Saúl,
aquella con la que entró en contacto desde su primer viaje a Quillabamba: los
machiguengas […] Fue ella [esposa Schneil] la que me habló de la cosmogonía
fluvial del machiguenga, donde la Vía Láctea era el río Meshiareni por el que
bajaban los innumerables dioses y diosecillos de su panteón a la tierra y por el que
subían al paraíso las almas de sus muertos. Les pregunté si tenían fotografías de las
familias con las que habían vivido. Me dijeron que no. Pero me mostraron muchos
objetos machiguengas (VARGAS LLOSA, 1995, p. 35). No fragmento anterior, verifica-se que o jovem escritor precisa recorrer aos
antropólogos ou aos linguistas que tinham interagido com o grupo indígena, e a partir dessas
informações formula o seu discurso. Para o hablador imerso na tribo, esse participa das
dinâmicas do grupo, configurando um observador participante que apresenta o modo de vida
dos machiguenga a partir das informações coletadas nos diálogos estabelecidos com os
membros da tribo, ou seja, de dentro, como ilustrado no seguinte fragmento da novela, um
diálogo entre o hablador e um membro da tribo: «Entonces, es verdad, te robaste a una yaminahua», le comenté a Tasurinchi. Dice
que no se la robó. La cambió por una sachavaca, un saco de maíz y otro de yuca,
más bien. «Los yaminahuas deberían alegrarse, eso que les di vale más que ella», me
aseguró. «Entonces, es verdad, te robaste a una yaminahua», le comenté a Tasurinchi. Dice
que no se la robó. La cambió por una sachavaca, un saco de maíz y otro de yuca,
más bien. «Los yaminahuas deberían alegrarse, eso que les di vale más que ella», me
aseguró. Le preguntó a la yaminahua en mi delante: «¿No es así?» Y ella asintió:
«Sí, lo es», diciendo. También eso le entendí. (VARGAS LLOSA, 1995, p.43-44). «Entonces, es verdad, te robaste a una yaminahua», le comenté a Tasurinchi. Dice
que no se la robó. La cambió por una sachavaca, un saco de maíz y otro de yuca,
más bien. «Los yaminahuas deberían alegrarse, eso que les di vale más que ella», me
aseguró. Le preguntó a la yaminahua en mi delante: «¿No es así?» Y ella asintió:
«Sí, lo es», diciendo. También eso le entendí. (VARGAS LLOSA, 1995, p.43-44). […] el recuerdo más memorable y recurrente de aquel viaje sería lo que les oí
contar, en Yarinacocha, a una pareja de lingüistas: los esposos Schneil. Al principio,
me pareció que era la primera vez que oía nombrar a aquella tribu. Pero, de pronto,
me di cuenta que era la misma sobre la que había oído tantas historias a Saúl,
aquella con la que entró en contacto desde su primer viaje a Quillabamba: los
machiguengas […] Fue ella [esposa Schneil] la que me habló de la cosmogonía
fluvial del machiguenga, donde la Vía Láctea era el río Meshiareni por el que
bajaban los innumerables dioses y diosecillos de su panteón a la tierra y por el que
subían al paraíso las almas de sus muertos. Les pregunté si tenían fotografías de las
familias con las que habían vivido. Me dijeron que no. Pero me mostraron muchos
objetos machiguengas (VARGAS LLOSA, 1995, p. 35). […] ancestros: los cazadores, los recolectores, los flecheros, los nómadas, los
irracionales, los mágicos, los animistas. También eso era el Perú y sólo entonces
tomaba yo cabal conciencia de ello: un mundo todavía sin domar, la Edad de Piedra,
las culturas mágico–religiosas, la poligamia, la reducción de cabezas (en una
localidad shapra, de Moronacocha, el cacique Tariri nos explicó, a través de un
intérprete, la complicada técnica de relleno y conocimientos que exigía la
operación), es decir, el despuntar de la historia humana (VARGAS LLOSA, 1995,
p.29). […] ancestros: los cazadores, los recolectores, los flecheros, los nómadas, los
irracionales, los mágicos, los animistas. También eso era el Perú y sólo entonces
tomaba yo cabal conciencia de ello: un mundo todavía sin domar, la Edad de Piedra,
las culturas mágico–religiosas, la poligamia, la reducción de cabezas (en una
localidad shapra, de Moronacocha, el cacique Tariri nos explicó, a través de un
intérprete, la complicada técnica de relleno y conocimientos que exigía la
operación), es decir, el despuntar de la historia humana (VARGAS LLOSA, 1995,
p.29). ABSTRACT Le preguntó a la yaminahua en mi delante: «¿No es así?» Y ella asintió:
«Sí, lo es», diciendo. También eso le entendí. (VARGAS LLOSA, 1995, p.43-44). Na citação anterior se observa que o discurso do hablador se configura após sua
inserção no interior da tribo, com base na experiência humana e no simbolismo das interações
sociais como mecanismo de interpretação de uma cultura; já o discurso que apresenta o
narrador não participante, surge de fontes científicas. Dessa maneira, os dois tipos de
observador-narrador presentes na obra revelam formas diferentes de coletar informações para
construir o conhecimento do indígena. Por outras palavras , duas vozes, que da mesma forma
que na etnografia, constroem o conhecimento recorrendo a modos diferentes de coleta e
apresentação dos dados. Misha Kokotovic (2001, p. 455), um comentador de Vargas Llosa, considera que o
discurso ocidental do jovem jornalista é insuficiente para representar a realidade dos
indígenas machiguengas e acrescenta que a figura do hablador constitui a solução para narrar […] A veces, para ver hasta dónde podía llevarlo «el tema», yo lo provocaba. ¿Qué
proponía, a fin de cuentas? ¿Que, para no alterar los modos de vida y las creencias
de unas tribus que vivían, muchas de ellas, en la Edad de Piedra, se abstuviera el
resto del Perú de explotar la Amazonía? […] No teníamos alternativa. Si el precio
del desarrollo y la industrialización, para los dieciséis millones de peruanos, era que
esos pocos millares de calatos tuvieran que cortarse el pelo, lavarse los tatuajes y
volverse mestizos – o, para usar la más odiada palabra del etnólogo: aculturarse –,
pues, qué remedio (VARGAS LLOSA, 1995, p.10). ϮÉ uma serpente que pode chegar a um tamanho adulto de 2m a 4m. Existe no Brasil, onde é a segunda maior
cobra e pode ser encontrada em diversos locais, como na Mata Atlântica, restingas, mangues, no Cerrado, na
Caatinga e na Floresta Amazônica. –¿En serio o te parece que la poligamia, el animismo, la reducción de cabezas y la
hechicería con cocimientos de tabaco representan una forma superior de cultura,
Mascarita?[…] Estaban en movimiento desde tiempos remotos y era probable que
jamás hubieran vivido de manera gregaria, en colectividades. El hecho de haber sido
desplazados, cada cierto tiempo, por tribus más aguerridas, y por los blancos –en los
períodos de las «fiebres»: la del caucho, la del oro, la del palo de rosa, la de la
colonización agrícola– hacia regiones cada vez más insalubres y estériles, donde era
imposible la supervivencia para grupos numerosos, había acentuado su
fragmentación y desarrollado en ellos un individualismo casi anárquico. No existía
un solo poblado machiguenga. No tenían caciques y no parecían conocer otra
autoridad que la de cada padre en su propia familia (VARGAS LLOSA, 1995, p. 11,
33). DOSSIÊ AMAZÔNIA DOSSIÊ AMAZÔNIA 45 O relato transcrito desconstrói o valor do indígena exaltado anteriormente pelo
hablador por meio do mito e da oralidade, e os define como uma linha divisória que estanca
os processos de uma nação. Segundo a perspectiva do narrador não participante, devido ao
espírito da tribo dos indígenas amazônicos o Peru não explora a totalidade de seus recursos
naturais e portanto não se integra como provedor principal de matérias primas a uma
economia global e questiona: “¿Deberían dieciséis millones de peruanos renunciar a los
recursos naturales de tres cuartas partes de su territorio para que los sesenta u ochenta mil
indígenas amazónicos siguieran flechándose tranquilamente entre ellos, reduciendo cabezas
y adorando al boa constrictor2 [...]” (VARGAS LLOSA, 1995, p.10). Pode-se afirmar, então, que do berço da civilização ocidental, o jovem escritor propõe
que o desenvolvimento e a industrialização do Peru requerem a renúncia ao componente
indígena. Para o jovem peruano, o espírito da tribo, os modos de vida primitiva embasada na
cosmogonia mítica irracional e na organização social não coletiva não são elementos de
admiração; pelo contrário, são obstáculos que impedem alcançar a modernidade e a integração
homogénea dos indivíduos a um mercado global. Mediante a desconstrução do indígena como
cultura ideal, o narrador não participante sugere que a interpretação indigenista das tribos
amazônicas que relata o hablador [observador-participante] é uma invenção. A partir do questionamento do jovem escritor a Mascarita e da exposição das práticas
e da estrutura social indígena, o narrador não participante enfatiza que se trata de formas de
vida inferiores ao modo de vida ocidental, na medida em que os membros da tribo são seres
não sociais e entidades humanas que não conformam a coletividade: –¿En serio o te parece que la poligamia, el animismo, la reducción de cabezas y la
hechicería con cocimientos de tabaco representan una forma superior de cultura,
Mascarita?[…] Estaban en movimiento desde tiempos remotos y era probable que
jamás hubieran vivido de manera gregaria, en colectividades. El hecho de haber sido
desplazados, cada cierto tiempo, por tribus más aguerridas, y por los blancos –en los
períodos de las «fiebres»: la del caucho, la del oro, la del palo de rosa, la de la
colonización agrícola– hacia regiones cada vez más insalubres y estériles, donde era
imposible la supervivencia para grupos numerosos, había acentuado su
fragmentación y desarrollado en ellos un individualismo casi anárquico. No existía
un solo poblado machiguenga. NOVA REVISTA AMAZÔNICA - VOLUME VI - NÚMERO ESPECIAL - DEZEMBRO 2018- ISSN: 2318-1346
4 44 a intimidade da tribo de uma forma mais autêntica, posto que reproduz uma cultura oral
passível de ser percebida num relato que transcreve exatamente o ouvido. Do mesmo modo
que o etnógrafo-participante, a narração do hablador imita, mediante a transcrição, aquilo que
escuta. Portanto, os dois tipos de narrador em El hablador contrapõem o discurso científico
ocidental e a oralidade como formas de representação do indígena. O observador não
participante constrói um relato antropológico interpretativo, que planteia contradições
epistemológicas e conflitos com as fontes de informação. O relato se aproxima ao mundo
indígena interpondo o véu das disciplinas científicas, às que recorre constantemente para
interpretar a cultura machiguenga, enquanto que o narrador participante dá voz ao indígena,
de forma que a cultura nativa americana é narrada como uma experiência vivida que aproxima
o leitor à visão de mundo do nativo. A partir do diálogo que se estabelece entre o jovem e Mascarita, no fragmento citado a
seguir, pode-se afirmar que o mundo indígena constitui para o jovem peruano um obstáculo
que impede o progresso da nação: […] A veces, para ver hasta dónde podía llevarlo «el tema», yo lo provocaba. ¿Qué
proponía, a fin de cuentas? ¿Que, para no alterar los modos de vida y las creencias
de unas tribus que vivían, muchas de ellas, en la Edad de Piedra, se abstuviera el
resto del Perú de explotar la Amazonía? […] No teníamos alternativa. Si el precio
del desarrollo y la industrialización, para los dieciséis millones de peruanos, era que
esos pocos millares de calatos tuvieran que cortarse el pelo, lavarse los tatuajes y
volverse mestizos – o, para usar la más odiada palabra del etnólogo: aculturarse –,
pues, qué remedio (VARGAS LLOSA, 1995, p.10). O narrador-externo estabelece sua posição apresentando o povo machiguenga como
minoria primitiva, selvagem e atrasada, como um problema para o desenvolvimento do Peru. O crítico peruano Antonio Cornejo Polar (1996, p. 25) destaca, com relação ao imaginário
indígena presente em El Hablador, que o narrador não participante [jovem escritor] diminui a
maneira de viver do indígena peruano frente ao modo de vida do Peru ocidentalizado. Posição
revelada na narrativa ao se fazer a descrição de uma viagem realizada por ele a um povoado
machiguenga, em que se refere aos indígenas como aqueles: NOVA REVISTA AMAZÔNICA - VOLUME VI - NÚMERO ESPECIAL - DEZEMBRO 2018- ISSN: 2318-1346 46 indigenismo lhe atribui às culturas primitivas americanas. Neste sentido, através de sua
narrativa Vargas Llosa emite juízos de valor sobre a cultura indígena da Amazônia peruana e
assevera que a figura do hablador é uma representação regional mediada pela paixão do amor
à terra e ao primitivo; enquanto que o indígena do escritor é uma projeção atravessada pelos
sistemas ocidentais de representação. Dessa forma, pode-se afirmar que a narração científica ocidental desqualifica a
narração indígena do hablador, já que esta última se constrói a partir de uma mitologia
embasada em fatos sociais não verificáveis, e desde os sentimentos subjetivos do observador
participante. Da mesma maneira que o discurso do jovem escritor questiona a subjetividade da
aproximação regionalista indigenista que apresenta o hablador, a novela debate a
subjetividade implícita na aproximação progressista ocidental do indígena. O narrador, observador indireto, inventa uma realidade indígena mediante um sistema
ocidental de significados. Ele concede vida às fotografias dos indígenas periféricos desde um
discurso dominante. Por exemplo, quando o jovem descreve a silhueta de um homem numa
fotografia, o faz animando uma imagem muda; lhe dá significado e contexto à fotografia a
partir de suas percepções: Vine a Firenze para olvidarme por un tiempo del Perú y de los peruanos y he aquí
que el malhadado país me salió al encuentro esta mañana de la manera más
inesperada […] Al primer golpe de vista se advertía que aquella comunidad de
hombres y mujeres sentados en círculo, a la manera amazónica – parecida a la
oriental: las piernas en cruz, flexionadas horizontalmente, el tronco muy erguido–, y
bañados por una luz que comenzaba a ceder, de crepúsculo tornándose noche, estaba
hipnóticamente concentrada. Su inmovilidad era absoluta. Todas las caras se
orientaban, como los radios de una circunferencia, hacia el punto central, una silueta
masculina que, de pie en el corazón de la ronda de machiguengas imantados por ella,
hablaba, moviendo los brazos. (VARGAS LLOSA, 1995, p. 3-4). Vine a Firenze para olvidarme por un tiempo del Perú y de los peruanos y he aquí
que el malhadado país me salió al encuentro esta mañana de la manera más
inesperada […] Al primer golpe de vista se advertía que aquella comunidad de
hombres y mujeres sentados en círculo, a la manera amazónica – parecida a la
oriental: las piernas en cruz, flexionadas horizontalmente, el tronco muy erguido–, y
bañados por una luz que comenzaba a ceder, de crepúsculo tornándose noche, estaba
hipnóticamente concentrada. Su inmovilidad era absoluta. Todas las caras se
orientaban, como los radios de una circunferencia, hacia el punto central, una silueta
masculina que, de pie en el corazón de la ronda de machiguengas imantados por ella,
hablaba, moviendo los brazos. (VARGAS LLOSA, 1995, p. 3-4). DOSSIÊ AMAZÔNIA No tenían caciques y no parecían conocer otra
autoridad que la de cada padre en su propia familia (VARGAS LLOSA, 1995, p. 11,
33). O indígena apresentado pelo jovem escritor é um ser selvagem, atrasado, violento e
irracional; características que desmitificam o caráter harmônico, coletivo e edénico que o La primera vez había sido de pura casualidad, hacía de eso lo menos diez años.
Estaban en cuclillas en medio círculo, viejos y niños, hombres y mujeres, en torno a
un hombre que peroraba, sentado y con las piernas cruzadas, encarándolos. Era un
hablador […] No, nunca había visto antes a ese hablador. Bastante viejo, a primera
vista, aunque, usted sabe, aquí en la selva se envejece rápido. Viejo, entre los
machiguengas, puede significar treinta años. Era un hombre bajo, fortachón, muy
expresivo. Yo, usted, cualquiera que hable y hable esa cantidad de horas, quedaría
ronco y extenuado. Pe- ro él, no. Hablaba y hablaba, con mucha energía. (VARGAS
LLOSA, 1995, p.61, 69) NOVA REVISTA AMAZÔNICA - VOLUME VI - NÚMERO ESPECIAL - DEZEMBRO 2018- ISSN: 2318-1346 Vine a Firenze para olvidarme por un tiempo del Perú y de los peruanos y he aquí
que el malhadado país me salió al encuentro esta mañana de la manera más
inesperada […] Al primer golpe de vista se advertía que aquella comunidad de
hombres y mujeres sentados en círculo, a la manera amazónica – parecida a la
oriental: las piernas en cruz, flexionadas horizontalmente, el tronco muy erguido–, y
bañados por una luz que comenzaba a ceder, de crepúsculo tornándose noche, estaba
hipnóticamente concentrada. Su inmovilidad era absoluta. Todas las caras se
orientaban, como los radios de una circunferencia, hacia el punto central, una silueta
masculina que, de pie en el corazón de la ronda de machiguengas imantados por ella,
hablaba, moviendo los brazos. (VARGAS LLOSA, 1995, p. 3-4). Assim, o discurso do narrador ocidental, apoiado na ciência e no uso de instrumentos
que buscam reproduzir a realidade, está mediado pela subjetividade do sistema ocidental de
significação. Um sistema que entende o Outro como um ser inferior que se opõe a seu caráter
de sujeito observador desde um sistema de códigos inconsistentes com a realidade indígena. Justifica o carácter dominante do poder de ocidente e a posição de dependência das minorias
sociais indígenas ante uma força globalizada e progressista. A novela lança mão das técnicas
narrativas da etnografia como veículo de juízo ao discurso ocidental dominante da
antropologia e suas escritas. A este respeito, importa lembrar o que o antropólogo Maurice
Leenhardt sustentou sobre o trabalho de campo, a reciprocidade e a elaboração do texto
etnográfico: “Quando um europeu vive dois ou três anos entre os selvagens, ele está DOSSIÊ AMAZÔNIA 47 totalmente convencido de que sabe tudo sobre eles; quando fica dez anos, ou quase, entre
eles, se for um homem observador, ele vai achar que sabe muito pouco e aí ele está
começando a aprender.” (CODRINGTON, 1972, p. VII). Passados mais de vinte anos da desaparição de Mascarita, o jovem escritor acredita
que seu amigo se transmutou em um hablador, tornando-se no porta-voz dos direitos e das
tradições dos machigengas. Cabe destacar que, após essa primeira visita à selva peruana e do
primeiro encontro com os Scheneil, o jovem escritor passou a ter um grande interesse pela
cultura machiguenga, em especial pela figura desse narrador oral que provavelmente ainda
existia. ¿Por qué los etnólogos modernos jamás nombraban a los habladores? Era una
pregunta que me hacía cada vez que llegaba a mis manos alguno de esos estudios o
trabajos de campo y descubría que tampoco esta vez se mencionaba ni siquiera de
paso a aquellos ambulantes contadores de cuentos que a mí me parecían el rasgo
más delicado y precioso de aquel pequeño pueblo […] ¿Por qué había sido incapaz,
en el curso de todos aquellos años, de escribir mi relato sobre los habladores? [...]
¿Cuántas veces, en estos veintitrés años, había pensado en los machiguengas?
¿Cuántas veces había tratado de adivinarlos, de escribirlos, cuántos proyectos había
hecho para viajar a sus tierras? Por culpa de ellos, todos los personajes o
instituciones que pudieran parecerse o de alguna manera asociarse en el mundo con
el hablador machiguenga habían ejercido una instantánea fascinación sobre mí.
(VARGAS LLOSA, 1995, p.16,64) NOVA REVISTA AMAZÔNICA - VOLUME VI - NÚMERO ESPECIAL - DEZEMBRO 2018- ISSN: 2318-1346 NOVA REVISTA AMAZÔNICA - VOLUME VI - NÚMERO ESPECIAL - DEZEMBRO 2018- ISSN: 2318-134 48 Sobre a experiência com o segundo hablador, Edwin Schenell relatou: “ –Tenía un
gran lunar […] Y unos pelos más colorados que los míos. Un tipo raro. Lo que los
machiguengas llaman un serigórompi. Quiere decir un excéntrico, alguien distinto de lo
normal. Por esos pelos color zanahoria le decimos el albino, el gringo, entre nosotros […]’
(VARGAS LLOSA, 1995, p.72). Da revelação anterior surge a interrogação: qual dos narradores apresenta de forma
mais fidedigna a cultura machiguenga? De acordo com as reflexões anteriores, se poderia
afirmar que é o hablador. Contudo, no final da novela o jovem escritor revela que o hablador
é produto da ficção, ou seja, a obra sugere que não existe uma representação real do outro. Ambas narrativas, participante e não participante, são uma ficção. Nesse sentido, a obra
oferece a partir das duas aproximações etnográficas duas perspectivas do indígena, uma visão
progressista e uma visão preservacionista; as quais são desconstruídas para sugerir que todo
método de representação, sem importar a distância que afasta o observador do objeto, está
mediado por subjetividades que condenam a ideia do “Outro” a ser uma ficção. Seria um
simulacro de uma representação etnográfica dos machiguengas. A personagem de Saúl Zuratas, o Mascarita, é o símbolo da ficção do científico
ocidental nas representações do indígena. As perspectivas indigenistas e progressistas
ocidentais do nativo americano são invenções, como no caso do jovem escritor que constrói o
Outro, o hablador machiguenga, desde seu próprio sistema de conhecimento. Vargas Llosa
apresenta um forte argumento a favor da assimilação, aculturação, e desaparição das culturas
indígenas. Kokotovic sugere que a estrutura dos relatos na novela, desde dois observadores
etnográficos ocidentais, privilegia a posição do narrador escritor, pois para entender o
hablador são necessárias as informações proporcionadas pelo observador não participante. Nesse sentido, a novela parece sugerir que, se a representação do indígena é uma invenção
desde qualquer um de seus enfoques, existe uma estrutura de poder que subordina o mundo
machiguenga à voz do narrador ocidental. Tal como ocorre na narrativa novelística e nos
cadernos etnográficos, a voz do hablador machiguenga não é mais que um reflexo construído
pelo escritor. Uma interpretação subjetiva, dentro dos agitados intentos por obter
objetividade, por apropriar-se e representar esse Outro concebido como marginal dentro da
cultura ocidental. NOVA REVISTA AMAZÔNICA - VOLUME VI - NÚMERO ESPECIAL - DEZEMBRO 2018- ISSN: 2318-1346 Durante os anos que se passaram após essa viagem frequentemente se perguntava: ¿Por qué los etnólogos modernos jamás nombraban a los habladores? Era una
pregunta que me hacía cada vez que llegaba a mis manos alguno de esos estudios o
trabajos de campo y descubría que tampoco esta vez se mencionaba ni siquiera de
paso a aquellos ambulantes contadores de cuentos que a mí me parecían el rasgo
más delicado y precioso de aquel pequeño pueblo […] ¿Por qué había sido incapaz,
en el curso de todos aquellos años, de escribir mi relato sobre los habladores? [...]
¿Cuántas veces, en estos veintitrés años, había pensado en los machiguengas? ¿Cuántas veces había tratado de adivinarlos, de escribirlos, cuántos proyectos había
hecho para viajar a sus tierras? Por culpa de ellos, todos los personajes o
instituciones que pudieran parecerse o de alguna manera asociarse en el mundo con
el hablador machiguenga habían ejercido una instantánea fascinación sobre mí. (VARGAS LLOSA, 1995, p.16,64) O interesse pelos habladores é revelado aos Scheneil em 1981, após vinte e três anos
de sua primeira visita à região dos machiguengas: […] Les dije que, por una razón difícil de explicar, la existencia de esos habladores,
saber lo que hacían y la función que ello tenía en la vida de su pueblo, había sido en
esos veintitrés años un gran estímulo para mi propio trabajo, una fuente de
inspiración y un ejemplo que me hubiera gustado emular. (VARGAS LLOSA, 1995,
p.68-69) De acordo com os Scheneil, os machiguengas não escondiam nada, contavam tudo
sobre suas crenças e tradições, mas os habladores era o único tema que evitavam. Por fim,
Edwin Schneil revela que em vinte e cinco anos de estada na selva teve a oportunidade de
ouvir dois habladores: NOVA REVISTA AMAZÔNICA - VOLUME VI - NÚMERO ESPECIAL - DEZEMBRO 2018- ISSN: 2318-1346 Para concluir citam-se as palavras do antropólogo e etnógrafo francês Pìerre Clastres
(2011, p. 64): “[...] De fato, comparado à vigorosa abundância do que é a vida de uma
sociedade primitiva, o discurso do cientista parece antes o tartamudeio hesitante de um gago e
vesgo ainda por cima. Portanto, um livro um pouco amargo por nos deixar na certeza de que DOSSIÊ AMAZÔNIA 49 nos deslocamos na superfície das significações, que deslizam um pouco mais adiante a cada
passo dado para aproximar-se delas. Mas aí já não se trata de etnologia. Sendo as coisas o que
elas são, a linguagem da ciência parece permanecer – por destino, talvez – discurso sobre os
selvagens e não dos selvagens. Como eles, não podemos conquistar a liberdade de ser ao
mesmo tempo um e outro, de estar simultaneamente aqui e lá, sem perder tudo e não ter mais
lugar onde ficar. A cada um se recusa assim a astúcia de um saber que, ao tornar-se absoluto,
se aboliria no silêncio. Vargas Llosa questiona em sua novela a impossibilidade de
interpretar uma cultura, consciente dessa dificuldade demonstra que ancorados na ciência só
se consegue construir, nas palavras de Clastres citadas acima, um “discurso sobre os
selvagens e não dos selvagens”. Vargas Llosa conhece o poder da palavra e o poder da voz e
isso se constata em sua obsessão pelo hablador, pelo qual mantém um interesse constante. Dessa forma, o discurso do narrador ocidental, embasado na ciência e no uso de instrumentos
que buscam reproduzir a realidade, está mediado pela subjetividade do sistema ocidental de
significação. Um sistema que, como se indicou já, entende o indígena como um objeto
inferior que se opõe ao seu caráter de sujeito observador, desde um sistema de códigos
inconsistentes com a realidade indígena. VARGAS LLOSA, Mario. El Hablador. 4ª ed. Barcelona: Editorial Seix Barral, 1995 REFERÊNCIAS CLASTRES, Pierre. Arqueologia da violência-pesquisas de antropologia política. Trad. Paulo Neves. São Paulo: Cosac Naify, 2011. CLIFFORD, James, George E. Marcus y José L. Moreno-Ruiz. Retóricas de la
antropología. Madrid: Júcar, 1991. CLIFFORD, James. A experiência etnográfica: antropologia e literatura no século XX. Organizado por José Reginaldo Santos Gonçalves. 2ª ed. Rio de Janeiro: Ed. UFRJ, 2002. CODRINGTON, R. H. (1891). The Melanesians. New York: Dover, 1972. CORNEJO POLAR, Antonio. Una heterogeneidad no dialéctica: sujeto y discurso migrantes
en el Perú moderno. Revista Iberoamericana. 62,1996. GONZÁLEZ ECHEVARRÍA, R. Mito y archivo: una teoría de la narrativa
latinoamericana. México: Fondo de Cultura Económica, 2000. KOKOTOVIC, Misha. Artículos - Mario Vargas Llosa Writes Of(f) the Native: Modernity
and Cultural Heterogeneity in Peru. Revista Canadiense de Estudios Hispánicos. MA25.3,
2001. VARGAS LLOSA, Mario. El Hablador. 4ª ed. Barcelona: Editorial Seix Barral, 1995.
|
https://openalex.org/W2143194240
|
https://zenodo.org/records/2469758/files/article.pdf
|
German
| null |
Ueber eine neue Hydroverbindung des Indigos und deren Verwendung zur quantitativen Bestimmung
|
Angewandte Chemie
| 1,901
|
public-domain
| 1,349
|
Engler und Albrecht: Filtration von Petroleum. Engler und Albrecht: Filtration von Petroleum. ngZ2Eftc2mie. gleichfalls eine sehr starke capillare Steig-
ltraft besitzt. rhaftlichen, unter viillig gleichen ausseren Be-
ingungen stattgebabten Ursprung nicht ge-
acht werden kann. In der Hauptsache stimmen die Result:Lte
unserer Vcrsuche mit den BeobRchtungen
iiberein,
welche Fr. Goppelsr6der
bei
seinen eingehenden Experimentaluntersuc*hun-
gen uber Capillaritiitswirkungen geliister Stoffe
und derrn Anwendung auf
d i e Cal)ill:Lrnna-
lyse gevlacht hat').'
Insbrsondercl
erkliirt
sich ciarriit aucli das zuletzt erw5 hntc rasclir
Aufsteigen des im rohen Anilin cnthaltenen
gefarbten Kiirpers. Auch im Ubrigcln ent-
halt die interessante Arbeit einc Keilie wicli-
tiger Ergebnisse iiber die verschiedcne Ca-
pillaritit einzelner Bestandtheile dt.5 rolien
Petroleums , be s ondcrs auch d er P yri dinh as en,
und verdient iiberhaupt fur Studirn auf dem
Gebiete der Capillaritatswirkungen
\ ollste
Reachtung. Dass dagcgen Veriindernngen in dem sll-
emeinen Ch:traktcr des Erdiils bei der Fil-
ration durch Erdschichten vor sich gehen,
3 t unseres Eraclitcns eine vollberechtigte,
wlesscn schon srit Jahren gemachte An-
ahme, und wir stimmen hierin vollstandig
lit Day6) und niit Zaloziecki')
tiberein,
rinnern dabei auch an einen schoii TTor Jahren
in hiesigen Laboratoriuni angestellten Ver-
uch, bri wcllchem Filtration m u dunkelem
olten Petroleum durch getrockrieten Thon-
clilamm aus den Hohrliicliern von Monte-
hino in Italien, welche ein viillig ivasser-
dares Petroleum liefern, cine vollstandige
htfarbung des Rohiils ergab. Versuche mit anderrn Filtrirrnateri:tlicn,
wie z. B. Infusorienerdr, Gemengc ron Sand
mit Infusorienerde etc., ergahen keine dcut-
liche Trennung des Petroleums in leiclite und
scliwere Theile; nur ein Gcmiscli voii Sand
und Thon zeigtr gcringe Differenzen in der
Dichte der austretendcn 6le. Wir glaitben,
dass der Erzielung von brawhbarca Iiesul-
tatcn specie11 der Umstand erschwereiicl im
Wege stand, dass diese Materialien uns nicht
in annahernd SO gleichmassigrr Mxhlung und
Kiirnung zu Gcbotc standen, wic d:is Alu-
miniummagnesiuinhydrosilicat . Ueber eine neue Hydroverbindung
des Indigos und deren Verwendung ziir
quantitativen Bestimmung. \'on Wilhelm Vaubel. Bei Gelegenheit einer umfassenden Unter-
suehung des lndigoblaus und Indigroths, uber
lie demnachst eine Mittheilung erfolgen soll,
machte ich die Bcohachtung, dass Indigo in
alkoholisch-slkslischer Losung sich unter dem
Einfluss von Zinkstaub zu einer tiefrothen
Flussigkeit aufliist. Es ist einerlei, ob man
hierbei Kali- oder Natronlauge oder Ammo-
niak anwendet. Die gleiehe Verbindung wird
rrhalten, wenn man Zinli durrh Eisen, Ferro-
sulfat brz. Ferroammoniumsulfat, Aluminium,
Zinn oder Palladium ersetzt; dagegen nicht
mit Silber, Quecksilber, Kupfer und Platin. Die Bcrsuche sollen im hiesigeii Labora-
torium fortgesetzt werden. 5) ,,Capillai analyse'', Verhandlungen der Natur
forschenden Gesellschaft Basel, 13aiid XIV, be
Eniil Birlrimaser 1900. G, Petroleum Beview 1900, Supplement zu den
h
9
S 12
23
August- und Srptemberheften. 7, Naphta, 9. Jallrgang, S. 12 u. 23. G, Petroleum Beview 1900, Supplement zu den
7, Naphta, 9. Jallrgang, S. 12 u. 23.
August- und Srptemberheften. Engler und Albrecht: Filtration von Petroleum. Eine eingehendere
Untersuchung nach der Anwendung von mit
Zinkhydroxyd geslttigter Lijsung sowie auch
cler durch Reduction mit Eisen erhaltenen
Verbindung,
die iibrigens
dieselbe Farb-
nuance wie alle anderen Reductionsflussig-
keiten zeigte, ergab, dass hier nur eine Hydro-
verbindung oder hiichstens eine Alkaliver-
bindung vorliegt. Ich glaube vorerst an-
nehmen zu dilrfen, dass die betreffende Ver-
bindung folgende Zusammensetzung hat : ') Auch
Indigroth
licfert
eine
gefiirbte
Zwischenverbindung. Die Farbe ist jedoch
mehr braunlich, und ist deshalb die Bildung
dieser Verbindung leichter zu ubersehen. Auch
Indigroth
licfert
eine
gefiirbte
Zwischenverbindung. Die Farbe ist jedoch
mehr braunlich, und ist deshalb die Bildung
dieser Verbindung leichter zu ubersehen. Eine griissere Reihe von Beobachtungen
zeigtc, class anscheinend folgende VerhLltnissc
am besten sind fur das Zustandekonirnen der
Reaction : Auf 1 g Indigo nimmt man 0,8-1,2
g
Zinkstaub, 20 ccm ICalilange (15 Proc.) oder
entsprechend Natronlauge, 50 ccni Alltohol
(98 Proc.). Auf 1 g Indigo nimmt man 0,8-1,2
g
Zinkstaub, 20 ccm ICalilange (15 Proc.) oder
entsprechend Natronlauge, 50 ccni Alltohol
(98 Proc.). (
Ein Uberschuss von Alkali stiirt das Zu-
standekommen der Reaction erheblich, wah-
rend eine geringere hlenge Alkali (6-12
ccni. N.-Kali1auge)keine oder nur Husserst schwachc
Reaction hervorbringt. und als D ihy clr oin digb 1 au bezeichnet wer-
den kann. Ich stehe nicht an, diese Erscheinung
der Bildung einer gefirbten Hydrovrrbin-
dung als zur colorimetrischen Bestirnmung
ties Indigos geeignet
zu empfe'nlen. Wo-
fern man nur ininier die betreffenden Ver-
haltnisse in vollstandig glcichartiger Weise
einhalt und
TOP
oder nach dem Eintreten
noch entsprechend wit Alkohol verdiinnt, um
eine vergleichende Beobachtung zu erniiig-
lichen, wird man anscheinend eine quanti-
tativc Bestimmung mit nur geringer XuHe
nnd in ganz kurzer Zeit ausfuhren kiinnen. Der Reactionsverlauf ist also folgender
bei der Reduction des Indigblaus: Dih d
i di bl Dihydroindigblau. Dihydroindigblau. Jedenfalls ist ein Versuch anzurathen,
und inm kann ausserdem auch durch Bestim-
inung der Schnelliglieit, mit der die Reaction
eintritt , rinen Schlass aaf die Verkiipbar-
lteit cles Indigos zielien. Bei dem ,,synthe-
tischen" Indigo fLllt ja auch die Riicksielit-
nahmc auf etwa vorhandenes Indigroth weg. Dih ydroindigblau. Indigweiss. Der Badisclien Anilin- und Sodafabrik,
der Farbenfabrik von Leop. Cassella sowie
den Farbwerken vorm. Meister Lucius und
Briining, die mich durch Uberlassung von
Material und Litteratnr bereitwilligst unter-
stiitztcn, sage ich auch :tn dicser Stelle meinen
verb in dlic h s t en D an k . Engler und Albrecht: Filtration von Petroleum. g
Der aus den Versuclien von D a y virl-
fach abgelriteten Annahrne , dass die vcr-
schiedenen nstiirlichen Xrdiile durch Hlinliclie
Capillarfiltration durcli porBw Erdschiclitcn
sich differenzirt haben sollen , kiiriricn wir
nicht vollstindig bcipfljchten. Es ist ja nohl
miiglich, dass in eiiizelnen Fallen von eincm
nrspriinglichen Bildungsort NUS cine Trennung
und Wanderung leichter Theile nach einer
mchr oder wcniger entfernten Lager
stattgefundcn hat und dass auch, falls die
Communication nachtraglich unt erbrochen oder
erlieblich erschwert wurdc, cine locale dimerndc
Absonderung leichter von scliweren 6lcii ein-
getreten ist. Sofrrn jrdoch die urspriinglicht
Communication erhalten blieb . 111 usste anel
entgegen der ursprunglich durcli Capillarital
bewirkten Separation ein dnrcli Diffusion
wenn auch nur ganz alliiilhlich, lierbeige
fiihrter Wiederaiisgleich stattfndcn. Es is1
ausserdcin die cliemische Natur dtr Er&l(
verschiedener nahegelegener FiindstBt,ten qua
litativ oft SO verschiedcn, dass an giwicin Dic weitere Untersucliiing wurde in der
Weise ausgefiihrt, dass ich 5 g Indigblau init
30 ccm Natroiilaugeliisung (ca. 10-12
Proc.)
sowic mit 100 ccm Alkohol und 2 g Zinkstaub
in eineni gut verschlossenen Cylinder schiittelte. Allmkhlich, etwa nach 2-5
Minuten, begann
sich die Njschung zu fgrben, und nahm die
Intensitat immer mehr zu. Nach 12 stundigem Stehen bez. ijfterem
Schutteln des Gemisches wurde rasch von dem
unveriinderten Indigo sowie demiiberschussigen
Zinkstaub abfiltrirt. Bei Beriihrung mit der
atrnospharischen Luft fiirbt sich die tiefrothe
Fliissigkeit unter Riickbildung des Indigblaus
wiedcr blau. Der ganze Vorgang bildet eine
wunderb are F ark, enerscheinung . Gicbt man melir Zinkstaub hinzu, so tritt
oollstiindige Entfarbung cin und Bildung von Vaubel: Neue Hydroverbindung des Indigos. XIV. Jahrgmg. Aefi 36. 3. September 1901.l 893 Indigweiss. Lisst man nun etwas Luft hin-
zutreten, so geht erst wieder die Bildung
der tiefrothen Zwischenverbindung vor sich,
und dann tritt die blaue Farbe des Indigos
nuf. und H e r zog2) beobachtete Wasserstoffsuper-
oxyd (oder
eine andere Superoxydverbin-
dung?) in erheblicher Ilfenge. Die Bestiin-
mung des Verhaltnisses ist hierbei schwierig,
da die Liisung bei der Filtration schon eine
rasche Uniwandlung in Indighlaa erleidet. Doch scheint dasselbe einigermaasscn cnt-
sprerhend zu sein. Die erste Untersnchung liess vermuthen,
dass es sich hier um eine Zinknatriumver-
bindung handelte, indem aach bei der Fil-
tration
der
rothen Liisung nachher ent-
sprechende hlengen Zinkhydroxyd bei dem
Indigo gefunden wurden. I ) Die weitere Untersuchung hat ergeben, dass
dem Indigblau und Indigroth die doppelte Molecular-
forinel zukommt. 2) W. M a n c h o t uiid J. Herzog.
Liobig's
Ann. 316, 318, 1901. Engler und Albrecht: Filtration von Petroleum. Das verniuthlichc Dihydroindigblau tritt
auch mitunter
in wasscriger Liisung bei
der Bildnng des Jndigweiss bez. der Riick-
bildung des Indigblaus aus dew Indigweiss
auf. Wahrscheinlich ist aber das Dihydro-
indigblau in aikalisch ivBsseriger Tliisung nur
in ganz geringem Maasse liislich, weshal6
die Erscheinung hauptsiichlich nur in al-
ltoholischer Liisung zu beobachten ist. Daimilaclt, Technisclie Hocliscliiile. g
Auch bei der Muckbildung des Indigblaiis
aus
dem
Dihydroindigblau
findet
sich
nachher in cler LBsung clas von Manchot
|
https://openalex.org/W4378451025
|
https://zenodo.org/records/7974075/files/Role%20of%20Gender.pdf
|
English
| null |
Role of gender and political connectedness after extreme events in coastal Bangladesh
|
Zenodo (CERN European Organization for Nuclear Research)
| 2,022
|
cc-by
| 9,319
|
© The Author(s). 2022 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this article are included in the article's Creative Commons
licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons
licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the
data made available in this article, unless otherwise stated in a credit line to the data. * Correspondence: bishawjit.mallick@tu-dresden.de;
bishawjit.mallick@colorado.edu
3Institute of Behavioral Science, University of Colorado Boulder, Boulder, USA
Full list of author information is available at the end of the article RESEARCH Open Access Abstract This study focuses on how gender and political engagement contribute to disaster management in Bangladesh,
specifically in disaster relief and rehabilitation. It is based on empirical evidence from southwest coastal Bangladesh,
particularly in Dacope Upazila, and applies a mixed-method approach. Results show that gender has a significant
influence on the acquisition of post-disaster relief and rehabilitation supports. Females, including those widowed
and divorced, get more relief than males. Nevertheless, older women are not given priority for aftermath disaster
relief. Moreover, gender plays a vital role in the household’s overall economic strength and is crucial for disaster
resilience. Furthermore, respondents’ families who were, in some way, closely connected with local social and
political leaders, obtained more relief and aids. The results indicate that socio-political connectedness, irrespective of
gender, still dominates local decision-making processes in disaster recovery. Alongside women’s empowerment,
regular monitoring and evaluation of relief and rehabilitation programs must improve, to reduce the traditional
barriers to effective disaster management (arising from (dis)connectedness to local social power) in the face of
climate change. Keywords: Socio-economic and demographic factors, Gender, Local political engagement, Disaster relief, Coastal
region Bangladesh Keywords: Socio-economic and demographic factors, Gender, Local political engagement, Disaster relief, Coastal
region, Bangladesh Introduction surges and cyclones. Studies claim that Bangladesh is the
country
most
affected
by
cyclone-related
damages,
losses, and fatalities globally. For instance, almost 40% of
cyclones
with
death
tolls
above
5000
occur
in
Bangladesh, and even the two deadliest cyclones hit the
country in 1970 with casualties of 300,000, and in 1991
with 140,000 (Bianchi & Malki-Epshtein, 2021). In Bangladesh, overall disaster management is challen-
ging. Being the most vulnerable country to tropical cyc-
lone hazards globally (IPCC, 2014) and one of the most
vulnerable countries to climate change (Clement et al.,
2021), disaster-affected people in Bangladesh face many
environmental and social security problems. Due to its
unique geographical location, high population density,
and low climate resilience accelerate the vulnerability to
climate change (ADW & UNU-EHS, 2015). According
to the IPCC (2021) the coastal zone will face ongoing
sea-level rise throughout the 21st century, intensifying
the sudden onset of extreme events for instance, tidal However, Bangladesh has become a role model for
climate-related disaster management by employing a
holistic approach to disaster response and recovery in its
long experience of dealing with such disasters. Nonethe-
less, the current disaster management in Bangladesh is
based on top-down approaches, and thus the initiatives
taken into consideration in all disaster vulnerable areas
are not people-centered (Alam et al., 2011). In general,
grassroots-level disaster management approaches mostly
evade gender priorities (Masaba et al., 2017). In addition, Role of gender and political connectedness
after extreme events in coastal Bangladesh Zakia Sultana1, Bishawjit Mallick2,3*
, Bangkim Biswas4, Sadhon Chandra Swarnokar5, Dipika Biswas6,
Partha Pratim Brahma1, Abdullah Bin Kaizer1, Tapati Roy7 and Md. Yahya Tamim1 Climate Action Climate Action Climate Action Sultana et al. Climate Action (2022) 1:7
https://doi.org/10.1007/s44168-022-00008-3 Sometimes, relief and rehabilita-
tion activities are undertaken by non-government orga-
nizations also favor their beneficiaries, particularly their
micro-credit recipients and the local political leaders
(Islam & Walkerden 2017). Thus, the exploitable struc-
tures disrupt to achieve sustainable development and
threaten future disaster management by making the
disaster-affected people the most vulnerable (Akinci,
2004; Ozcelik et al., 2008). Furthermore, gender intersects with other social attri-
butes; for instance, women in rural areas, particularly in
developing countries, mainly depend on natural re-
sources for their livelihood, making them especially vul-
nerable
to
climate
change
issues
(Terry,
2009). Hemachandra et al. (2018) explored the role of women
in disaster risk governance and the constraints that limit
these roles. They upheld that women can play a pioneer-
ing role in disaster risk governance for effective manage-
ment that initiates resilience among women within a
community (Roy 2020a). Gender-based discrimination
persists in allocating disaster relief and rehabilitation
support in the aftermath of a disaster, as does discrimin-
ation based on countless other dimensions, including re-
ligion, tradition, and culture (Enarson & Meyreles, 2004;
Farrior, 2009; Roy 2020b). Women with fewer re-
sources are the most vulnerable and most likely to
be considered natural disaster victims (Lebni et al.,
2020). Gender-based discrimination and violence in-
crease after a disaster (Aryanti & Muhlis, 2020). Hu-
man rights organizations have long been working on
gender and human rights, but still, gender-based dis-
crimination is visible everywhere, including in educa-
tion,
housing,
the
workplace,
healthcare,
disaster
relief, and more (Clar, 2019; Farrior, 2009). Takasaki
(2012) highlighted that women-headed households
experience more risk and vulnerabilities as they ad-
just to disaster devastations. In such circumstances,
gender-sensitive emergency spaces are required for
women to defend themselves against post-disaster
violence and risks (Aryanti & Muhlis, 2020). Thus,
the inclusion of gender dimensions in disaster man-
agement requires active participation and empower-
ment of women in local politics. A structural framework was employed by Bhavnani
(2006) to investigate the conflicts of disasters and found
that illegal resource acquisition was significantly con-
nected to the aftermath disaster relief distribution. In
their study on cyclone Aila of 2009 in Bangladesh, Mah-
mud and Prowse (2012) observed that the post-disaster
relief and rehabilitation programs had been manipulated
by locally politicized favoritism, nepotism, and corrup-
tion. Page 2 of 12 Sultana et al. Climate Action (2022) 1:7 Sultana et al. Climate Action (2022) 1:7 Sultana et al. Climate Action Interestingly, research shows a lack of women’s in-
volvement in disaster risk reduction-related interven-
tions and programs (Masaba et al., 2017; Islam &
Walkerden, 2017); thus, women’s opinions are not con-
sidered
seriously
in the
household
decision-making
process. The ability to receive post-disaster relief has a
significant relationship to the disaster-affected people’s
socio-economic conditions, i.e., gender, age, education,
occupation, income, property damages, etc. (Hapeman,
2012; Khan & Nahar, 2014). Although the inadequate fi-
nancial resources, lack of human capacity, political inter-
ference,
misuse
of
resources,
lack
of
cooperation
between the local community, and lack of security and
supporting regulations are primarily responsible for
weak disaster risk management (Masaba et al. 2017). the operation of the disaster relief and rehabilitation
programs is significantly connected to various factors
such as political connectedness, bureaucratic entangle-
ment, political collaboration at the national and local
level (Baytiyeh, 2017; Cohen & Werker, 2008; Hapeman,
2012; Islam et al., 2014; Kovach, 2013; Macrae, 2014;
Sökefeld, 2012). In Bangladesh, grassroots level politics
play a crucial role in post-disaster relief and rehabilita-
tion programs (Bhavnani, 2006; Hayat & Amaratunga,
2017; Heide, 2004; Oliver & Reeves, 2015; Pelling & Dill,
2008). A study conducted by Sökefeld (2012) found that
disaster’s complexity with political action and power re-
lations creates a critical situation in total disaster man-
agement. Thus, the influence of local politics on disaster
management increases the people’s vulnerability of that
area (Gotham, 2012). Moreover, some other factors,
namely economic status, religion, caste system, ethnicity,
and other identity-based biases, play a vital role in disas-
ter relief and rehabilitation distribution (Ciampi et al.,
2011; Murthy, 2009). Aftermath relief, shelter, livelihood assistance like live-
stock, fishing boats, and services supports are provided
by the local governments (Islam & Walkerden, 2017;
Chipangura et al., 2017); however, many victims con-
tinue facing struggles to overcome post-disaster impacts
due to social and gender discrimination as well as polit-
ical instability (Buhaug et al., 2008; Enia, 2009; Gold-
stone et al., 2010; Kovach, 2013). Besides, many non-
victims of the disaster also in-migrate to the affected
area and list themselves as beneficiaries for emergency
relief, thus influencing the sufferings of the real victims
(Mallick & Vogt, 2012). Thus, people often move to the cities searching for
work and food security, especially those who do not get
enough relief and rehabilitation support (Care, 2008). Mallick and Vogt (2012) have similar findings identified
in their study that demonstrated mismanagement in re-
lief and rehabilitation after cyclone Aila, which influ-
enced many disaster victims to migrate to a new place. Thus, a holistic approach was required to manage the
disaster's impact from a rescue operation to relief inter-
vention to ensure effective disaster risk reduction by in-
cluding every social group in the intervention (Mendoza, Page 3 of 12 Sultana et al. Climate Action (2022) 1:7 Page 3 of 12 Sultana et al. Climate Action 2014). Parvin et al. (2019) proposed a comprehensive
supply chain management of disaster relief based on the
response of personnel (employees of GOs, NGOs, and
foreign donor organizations) involved in relief and re-
habilitation intervention. inhuman conditions after the cyclone Aila of 2009. Sev-
eral studies show that the Gabura Union of Shyamnagor
Upazila in Satkhira district and the Sutarkhali union of
Dacope Upazila in Khulna district were submerged more
than 6 months after Aila, and the livelihood conditions
were severely disrupted in those villages (Mallick et al.,
2011; Auerbach et al., 2015). Considering the accessibil-
ity (remoteness), proximity to rivers, and rurality in na-
ture,
and
after
all,
the
long-standing
severity
and
inhuman conditions after cyclone Aila compared to
other villages in nearby Upazilas, we selected the villages
in Sutarkhali Union from Dacope Upazila. Although several studies in Bangladesh have discussed
gender and the problem of relief distribution, there is
hardly any study available that deals with gender in-
volvement in local politics and its diversity in the local
level disaster management. And at present, how gender
and political affiliation contribute to local-level disaster
management in Bangladesh has become an emerging
issue. Therefore, this study focuses on the roles of gen-
der in disaster management to investigate how to pro-
gress
disaster
management
and
climate
change
adaptation. The following research questions are raised: p
p
Four villages were selected based on these criteria:
Gunari, Sutarkhali, Nalian, and Kalabogi of Sutarkhali
Union, under the Dacope Upazila of Khulna district (Fig. 1). Rivers surround the Sutarkhali Union, and the most
significant river is the Shibsa River which flows on the
west side of the Sutarkhali Union. Materials and methods
Study area The southern part of Bangladesh stands on the Bay of
Bengal and represents one-third of the country’s land
surface; and is also highly susceptible to environmental
hazards like a cyclone, tidal surges, flooding, and ero-
sion. About 28% of the country’s population lives in this
coastal region, and during the last 100 years, 508 cy-
clones originated in the Bay of Bengal, and 17% of these
cyclones hit Bangladesh. On average, the country faces
one high-speed tropical cyclone every 3 years. For in-
stance, the communities were affected by cyclone Sidr of
2007 and cyclone Aila of 2009 but did not experience
similar damages and losses. After these two consecutive
cyclones in this southwestern region of the country,
massive interventions program have been implemented. Such programs include relief and rehabilitation supports,
primary healthcare, housing, and disaster management
skill training. How those programs were implemented
and to what extent gender played a role in the distribu-
tion and management of those programs are the key
questions for this research, and therefore, we focused on
selecting
the
communities
which
have
longer-term The total amount of
land in this union is about 12,092 acres. The world's lar-
gest mangrove forest, the Sundarbans, is located close to
the study area, where one can access the Sundarbans by
crossing the surrounding rivers. The total population is
about 30,060 (male 15,205, female 14,855). About 73%
of the people are Muslim there. The study area’s average
literacy rate is about 49.5% (male 56.3%, female 42.6%)
(BBS, 2011). How do socio-economic and demographic factors
affect disaster management, and how does it differ
among the respondents? g
How much does gender- and individual-level en-
gagement in local politics contribute to disaster
management? These questions are presented based on empirical evi-
dence from southwest coastal Bangladesh, particularly in
Dacope Upazila of Khulna district. The following section
describes the methodology, and section three presents
the results and discussion. Section four concludes with
the future outlook. Razzaque et al. (2019) found that the study villages are
the most vulnerable to natural and climate-induced di-
sasters, such as tropical cyclones, floods, tidal surges, sal-
inity intrusion, and riverbank erosion. The height of
river water was about 13 feet above the land surface dur-
ing Cyclone Aila, which hit coastal Bangladesh in 2009
and damaged many embankments in several places. Many people in these areas lost their lives and were se-
verely affected then (Mallick & Vogt, 2014). The mar-
ginal people mainly depend on natural resources for
their bread and butter, and thus, they have become the
most sufferer in the aftermath of disaster (Biswas & Mal-
lick, 2020). In addition, gender discrimination was an
issue everywhere during the aftermath of relief and re-
habilitation
programs
(Razzaque
et
al. 2019). This
disaster-prone region has political instability, which ag-
gravates its vulnerabilities in post-disaster recovery inter-
ventions (Sultana et al., 2021). Therefore, this region has
been selected as the study area due to the existing re-
search gap on gender and the politics of aftermath disas-
ter management from a sociological view. Data collection A mixed-method approach was applied in the data col-
lection process. Quantitative data were obtained by con-
ducting household surveys where all sorts of households
were considered, i.e., single-headed, nuclear, and ex-
tended families. In total, two hundred thirty respondents Page 4 of 12 (2022) 1:7 Sultana et al. Climate Action (2022) 1:7 limate Action (2022) 1:7 Sultana et al. Climate Action Fig. 1 Study area were interviewed from four villages of Sutarkhali Union;
among them, 130 respondents were female, and the
remaining 100 respondents were male. Based on random
sampling, we selected 40 respondents from Nalian, 65
from Sutarkhali, 75 from Gunari, and 50 from Kalabogi. The final survey was conducted in March 2019. Before
that, a reconnaissance survey was done in December
2018. This initial visit to the study area was used to de-
termine the suitability of the proposed study. As the
study required informative, in-depth, and detailed data, a
'face to face' mode was selected for administering the
survey to the respondents. The survey questionnaire was
semi-structured to thoroughly clarify the quantitative
findings on respondents' socio-economic condition, pol-
itical involvement, and gender influences and experi-
ences during the relief and rehabilitation interventions
in the aftermath of cyclone Aila. were interviewed from four villages of Sutarkhali Union;
among them, 130 respondents were female, and the
remaining 100 respondents were male. Based on random
sampling, we selected 40 respondents from Nalian, 65
from Sutarkhali, 75 from Gunari, and 50 from Kalabogi. The final survey was conducted in March 2019. Before
that, a reconnaissance survey was done in December
2018. This initial visit to the study area was used to de-
termine the suitability of the proposed study. As the
study required informative, in-depth, and detailed data, a
'face to face' mode was selected for administering the
survey to the respondents. The survey questionnaire was
semi-structured to thoroughly clarify the quantitative
findings on respondents' socio-economic condition, pol-
itical involvement, and gender influences and experi-
ences during the relief and rehabilitation interventions
in the aftermath of cyclone Aila. Respondents were selected from those who knew and
were involved in the aftermath rehabilitation program. Likewise, four key informant interviews (KII) or in-depth
interviews were conducted purposively of respondents
who know disaster management well. The results from
the FGDs and KIIs have been summarized at the end. Data collection Research ethics principles, reliability, and validity were
rigorously maintained to keep a sound methodology for
this study. Data characteristics and analysis The list of essential variables used in the study and their
characteristics are shown in Table 1. A logistic regression model, comprised of socio-
demographic, economic, and political features, has been
employed to determine how individuals get relief sup-
port after a disaster. Here, the receiving relief has been
regarded as a dependent variable, dichotomous in man-
ner. The following Eq. 1 indicates the logistic regression
model. Furthermore, two unique and informative qualitative
methods were used to supplement the quantitative data
findings. One focus group discussion (FGD) was con-
ducted in each village, consisting of ten respondents. Page 5 of 12 Sultana et al. Climate Action (2022) 1:7 Sultana et al. Climate Action Table 1 Variable characteristics
Demographic variables
Variable name
Values (scale and range)
Age (in years)
C: 25 to 78
Gender
N: 1 = Male; 2 = Female
Religion
N: 1 = Muslim; 2 = Hindu; 3 =
Christian
Education level
O: 1= Illiterate; 2 = Primary; 3 =
Secondary
Marital status
N: 1 = Married; 2 = Widow; 3 =
Divorced
Type of family
N: 1 = Single headed; 2 = Nuclear; 3 =
Extended
Family size (in number)
C: 1 to 8
Socio-economic and socio-political variables
Occupation
N: 1 = Farming; 2 = Business; 3 = Day
laborer; 4 = Service
Monthly income (BDT)
C: 1000 to 13,000
Monthly expenditure (BDT)
C: 3000 to 40,000
Monthly savings (BDT)
C: 500 to 4000
Earning members of the family
(in number)
C: 1 to 2
Receiving relief
O: 0 = No; 1 = Yes
Source of relief support
N: 1 = GO; 2 = NGO; 3 = Both
Satisfaction with local
government initiative
O: 0 = No; 1 = Yes
Engagement in politics
O: 0 = No; 1 = Yes
Political leaders having any
nepotism tendency
O: 0 = No; 1 = Yes
Receiving help from political
leaders
O: 0 = No; 1 = Yes
C continuous, O ordinal, N nominal
Source: Field Survey, 2019 Primary, and Illiterate) and incorporated in the model. Secondly, as socio-economic indicators such as the occu-
pation of the respondents (X7), earning members (X8),
and total monthly income of the household (X9) have
been comprised in the model. Finally, a binary socio-
political variable, whether engaged in politics or not
(X10), has been included in the regression analysis. Profile of the respondents The demographic, socio-economic, and socio-political
variables are incorporated in Table 2. Among the
three age groups, around half of the respondents
(50%)
belong
to
the
age
group
of
25–40
years,
whereas the age group of 61-78 contains only about
3.5%. Approximately three-quarters of the respondents
(about 68%) are married, and divorced respondents
cover
9.6%. More
than
half
of
the
respondents
(56.5%) are female. Around 50% of the households
are Muslims, whereas Hindus and Christians follow
with 36.1% and 13.9%. The average household size is
about 4. The highest level of education of the respon-
dents is secondary, but only about 5% have this. Most
households are headed by one or two earning mem-
bers, and their monthly average income level is below
5000 BDT. The primary occupation for maintaining
their livelihood is day-labor (70%), besides which they
do farming (about 28%) and business (about 2%). Li ¼ ln
Pi
1−Pi
¼ β0 þ
X
m
i¼1
βiXji þ εi
ð1 ð1Þ where, where, j ¼ 1; 2; 3……:M; i ¼ 1; 2; 3……::n j ¼ 1; 2; 3……:M; i ¼ 1; 2; 3……::n Pi indicates the likelihood of getting relief. Yi is the re-
lief status (if people get relief, Y = 1 and if do not Y = 0). In Eq. 1,
lnðpi
1 −piÞ and β0 indicate the log of the odds
ratio and intercept respectively. Xji is a set of predictor
variables, and εi is the error term. The regression ana-
lysis has encompassed a total of ten demographic, eco-
nomic, and political characteristics. Firstly, in socio-
demographic variables, we included the age (X1), gender
(X2), marital status (X3), religion (X4), and size of the
household (X5). Along with the religion, to understand
the relief access of the people in various education levels,
we also classified the education variable (X4) (Secondary, The key finding from Table 2 is that the average
monthly expenditure is more than the respondents’
monthly average income. This indicates that the respon-
dents might take loans to bear their daily expenses. The
findings also demonstrate that about 55% of people have
access to relief, which indicates the rest do not receive
relief after a disaster. Only 5.7% of the total respondents
are engaged in local politics. Since women in the village
usually do not participate in politics, this 5.7% political
engagement likely belongs to the male respondents. Data characteristics and analysis g
y
The variance inflation factor (VIF=1.42) and Pearson
correlation tests have been employed to detect the mag-
nitude of collinearity and cross-correlation among the
regressors. The highly connected variables have been de-
tached (i.e., proxy variables) from the model to fit the
model better. Furthermore, before picking the final
model, we examined the heteroscedasticity problem of
our data and used robust standard error. To analyze data
in our study, we have used several statistical tools such
as STATA14 and SPSS22 (Statistical Package for the So-
cial Sciences). To prepare the map and indicate the geo-
graphical location of the study villages, we also used
ArcMap 10.5. Profile of the respondents Among the male respondents, those engaged in local
politics receive relief in the post-disaster intervention. Page 6 of 12 Sultana et al. Climate Action (2022) 1:7 Sultana et al. Climate Action Table 2 Demographic, socio-economic, and socio-political profile of the respondents
Variables
Mean
Std. De
Age
25–40
.5
.501
41–60
.465
.5
61–78
.035
.184
Gender
.565
.497
Marital status
Married
.678
.468
Widow
.226
.419
Divorced
.096
.295
Religion
Muslim
.5
.501
Hindu
.361
.481
Christian
.139
.347
Education
Illiterate
.457
.499
Primary
.491
.501
Secondary
.052
.223
Number of family members
3.913 [1~8]
1.171
Occupation
Farmer
.278
.449
Business
.022
.146
Day-labor
.7
.459
Number of earning members
1.513 [1~2]
.501
Monthly income of the family
4776.087 [1000~13,000]
2372.25
Monthly expenditure of the family
5830.435 [3000~40,000]
3037.16
Relief status
.552
.498
Engage in politics
.057
.231
Source: Field Survey, 2019 Std. Dev. Source: Field Survey, 2019 Post-disaster situation in general cyclone shelter, so very often our girls and women face
sexual harassment when they go outside at midnight to
use the common”. When asked how this problem should
be addressed, she said they must be rehabilitated from
this small island as early as possible. Besides, the govern-
ment must construct a sustainable and women-friendly
cyclone shelter. Affected communities require financial, institutional,
and infrastructural support after a disaster. The sources
of
relief
support
are
the
government
and
non-
government organizations. People usually receive relief
from one or the other but sometimes get support from
both. Most of the respondents who live in the study area
have a good relationship with their neighbors. However,
people are not satisfied with the local government initia-
tives since they do not receive their required facilities
from the local government. At least fifteen thousand
people in Kalabogi village face the extreme threat of cli-
matic disaster in the upcoming monsoon season due to
the lack of adequate cyclone shelters. Describing them
as floating hyacinth, Sefali Begum (42 years old, a house-
wife in Sutarkhali village) said lack of space was one of
the significant problems in the cyclone shelter. She also
disclosed, “There is no separate toilet for women in the Furthermore, Ruhul Amin, a fisherman, 35 years old,
said he faced some difficulty during cyclone Aila, when
he took his pregnant wife to a cyclone shelter four miles
away from his house. “It is hard to describe the ordeal I
had faced due to the lack of a cyclone shelter nearby. We have been denied every basic right, from drinking
water to medical care to education to shelter. We are
not given anything for years,” he said. Many people have
migrated in response to such livelihood difficulties, in-
cluding after losing their households to natural calam-
ities. Social organization is an integral element of Page 7 of 12 Page 7 of 12 Sultana et al. Climate Action (2022) 1:7 Sultana et al. Climate Action reducing poor people’s suffering (Kazal et al., 2017). However, people in this area are reluctant to participate
in social organizations like associations or clubs. They
pointed out that these organizations work for the people
engaged in politics, and therefore, engagement in local
politics usually creates opportunities for getting an enor-
mous amount of post-disaster relief and rehabilitation
support. with relief status. Post-disaster situation in general This significant correlation between
the widowed and divorced women with overall relief sta-
tus indicates that they receive more relief than married
women. In the case of other variables, it is found that
earning members of a family (0.26), monthly income (−
0.27), monthly expenditure (0.36), engagement in politics
(0.17), etc. also have a positive and negative relationship
with another set of variables that might have multiple
direct or indirect impacts on relief status. The result as-
serts that the variables' overall correlation coefficient re-
jects the null hypothesis and indicates that other factors
govern relief accessibility. How do the socio-economic and demographic factors
affect post-disaster relief activities, and how does it differ
among the respondents? The pairwise Pearson correlation coefficient is incorpo-
rated in Table 3 to indicate the degree of correlation
among the variables at correspondingly at 1%, 5%, and
10% level of significance. The correlation coefficient im-
plies that age and family are negatively correlated with
relief status, whereas gender (0.44) is moderately corre-
lated with relief status at a 1% significance level. Under
marital status, widowed
(0.28) and divorced (0.23)
women exhibit a positive correlation with relief access,
but married women (−0.40) show a negative association Extent of the influence on disaster management by
gender and individual-level engagement in local politics
A binary logistic regression model has been employed,
where the socio-economic variables and the socio-
political status of the household head have been incor-
porated, presented in Table 4. Source: Field Survey, 2019
*** p<0.01, ** p<0.05, * p<0.1 Extent of the influence on disaster management by
gender and individual-level engagement in local politics A binary logistic regression model has been employed,
where the socio-economic variables and the socio-
political status of the household head have been incor-
porated, presented in Table 4. A binary logistic regression model has been delineated
in Table 4 where demographic, socio-economic, and
socio-political
attributes
have
been
incorporated
to Table 3 Pairwise correlation of the selected variables
Variables
(1)
(2)
(3)
(4)
(5a)
(5b)
(5c)
(6a)
(6b)
(6c)
(7)
(8)
(9)
(10)
(1) Relief status
1.00
(2) Age
−
0.14***
1.00
(3) Gender
0.44***
−0.13
1.00
(4) Family members
−
0.26***
0.24*** −
0.19***
1.00
(5) Marital Status
(5a) Married
−
0.40***
0.19*** −
0.55***
0.31***
1.00
(5b) Widow
0.28***
−0.11
0.45***
−
0.20***
−
0.78***
1.00
(5c) Divorced
0.23***
−
0.15**
0.23***
−
0.20***
−
0.47***
−0.18** 1.00
(6) Religion
(6a) Muslim
−0.11
0.01
−0.11
−0.10
0.15**
−0.04*
−0.18** 1.00
(6b) Hindu
0.04
0.06
0.00
0.05
0.07
−0.06
−0.03
−
0.75***
1.00
(6c) Christian
0.11
−0.11
0.15**
0.08
−
0.31***
0.14**
0.30***
−
0.40***
−
0.30***
1.00
(7) Earning
members
0.08
0.26*** 0.16**
0.05
−
0.37***
0.26***
0.23***
−0.09
0.01
0.12
1.00
(8) Monthly income
−
0.27***
−0.03
−
0.22***
0.27***
0.54***
−
0.42***
−
0.26***
−0.01
0.06
−
0.06
−
0.41***
1.00
(9) Monthly
expenditure
−0.12
0.36*** −0.01
0.51***
0.26***
−
0.20***
−0.13*
−0.13*
0.12*
0.02
0.05
0.43*** 1.00
(10) Engage in
politics
0.11
0.00
−0.17** −0.14** 0.17**
−0.13*
−0.08
0.09
−0.03
−
0.10
−0.03
0.13*
−
0.02
1.00
Source: Field Survey, 2019 Page 8 of 12 Sultana et al. Climate Action (2022) 1:7 Sultana et al. Climate Action Table 4 Predictors of access to relief intervention
Dependent variable = relief status
(yes = 1 and no = 0)
Odds ratio
Robust Std. Err. Extent of the influence on disaster management by
gender and individual-level engagement in local politics t value
P value
Sig
Age (Z1): (25–40)R
41–60
2.063
.777
1.92
.055
*
61–78
.704
.522
−0.47
.636
Gender (Z2)
5.13
2.106
3.98
0
***
Marital Status (Z3): (Married)R
Divorced
2.117
1.253
1.27
.205
Widow
6.309
5.483
2.12
.034
**
Religion (Z4): (Muslim)R
Hindu
1.91
.679
1.82
.069
*
Christian
1.855
1.158
0.99
.323
Number of family members (Z5)
.853
.129
−1.05
.293
Education (Z6): (Illiterate)R
Primary
1.041
.376
0.11
.911
Secondary
28.056
26.138
3.58
0
***
Occupation (Z7): (Farmer) R
Business
13.2
18.175
1.87
.061
*
Day labor
.958
.417
−0.10
.921
Earning members of the family (Z8)
.424
.18
−2.02
.043
**
Monthly income of the family (Z9)
1
0
−2.95
.003
***
Engagement in politics (Z10)
5.23
5.181
1.67
.095
*
Constant
4.535
4.623
1.48
.138
Number of observations
230
Pseudo r-squared
0.30
Chi-square
81.08
Note: *** p<0.01, ** p<0.05, * p<0.1 and R indicates the reference categories
Source: Field Survey 2019 financial crisis after being hit by cyclone Aila that de-
tached us from the main village of Kalabogi. Moreover,
we cannot pay the money to the local agent of the water
supply company. Actually, we do not have any alterna-
tive jobs, only catching fish and crabs from the river. It
is difficult to buy water or pay a bribe for water with this
earnings.” Such realities add an extra burden to the life
of these poor people who struggle daily to ensure their
meals, forcing them to work even harder to avail them-
selves of drinking water after a devastating disaster. determine the relief support status under different refer-
ence categories. The results exhibit that several variables
are significantly associated with the outcome variables. The logistic regression model's odds ratio suggests that
the person whose age ranges from 41 to 60 years re-
ceives more relief than people of other age groups. The
demographic variable gender has a significant positive
relationship with the outcome variable, demonstrating
that females have greater access to the relief program
than males at 1%. Similarly, widowed and divorced
women have more access to relief opportunities than
married women. Again, the regression model shows that families with
fewer than two dependents (particularly girls elderly)
earn comparatively more than the families with more
dependents. Extent of the influence on disaster management by
gender and individual-level engagement in local politics Although religion doesn't significantly affect
relief acquisition, education plays a vital role here as
people who have secondary education in this region re-
ceive more relief than primary level education or
illiterate people. Surprisingly, a businessman receives
more relief than a farmer or day laborer, whose liveli-
hoods are more likely to face the harsher realities of Interestingly, there was also evidence of a lack of sup-
port for drinking water. For instance, Karimon Bibi (25
years old housewife in Kalabogi) claimed that after Aila
water was supplied through a local agent in their village,
they could not pay the required fees for water supply. Therefore, her husband spent almost 8 h collecting
drinking water from Chalna (Upazila town) to their vil-
lage Kalabogi. She mentioned, “We faced an extreme Page 9 of 12 Page 9 of 12 Sultana et al. Climate Action (2022) 1:7 Sultana et al. Climate Action disaster. Similar observations were found during the dis-
cussion with the farmers and fishers groups. For in-
stance, Md. Abdul Halim (45 years, a fisher from Gunari
village) described, “I went to a NGO listing for providing
building (house), but they did not put me in their list be-
cause I could not pay the money they asked for registra-
tion. But our local leader has received two houses from
them.” This reflects the way of manipulating the relief
and rehabilitation activities. We also asked if gender
plays any role in distributing such houses during the
group discussion. Some participants claimed that the
NGO/GO usually writes the wife’s name on the paper,
but getting the house money was more important than
the gender. 2014), also reported in our study’s findings. Involvement
in politics plays a role in getting relief. However, our
study did not explain if a female involved in local politics
has higher chances of getting more relief than a male in
politics. Again, the political elites in Bangladesh use disaster
events to strengthen their resilience to future disasters
in the affected areas (Mallick & Vogt, 2012), which was
also reported in our study. In such cases, the use of
wife’s name in getting stable and durable housing facil-
ities is very regular, reflecting the nepotism and mani-
festation of the dependency on the external resources
for the vulnerable groups. Extent of the influence on disaster management by
gender and individual-level engagement in local politics g
p
In rural society, political nepotism is to some extent
accepted and overlooked (Mallick, 2012), and is also a
barrier to effective disaster management at the grass-
roots level. The primary forms of nepotism and favorit-
ism influence the mismanagement of the relief and
rehabilitation process and illegal resource acquisition
connected to the relief distribution (Khan & Rahman,
2007; Azad & Khan, 2015; Bhavnani, 2006). When a nat-
ural disaster strikes, it simply actualizes potential in-
stability already present in the governing regime and is
determined by the limitations of the adaptive capacity of
political institutions (Goldstone et al., 2010). The vulner-
ability or adaptive capacity of the political institutions of
a governing regime determines the likelihood of civil
conflict in the aftermath of a natural disaster (Buhaug
et al., 2008; Enia, 2009). Thus, resource allocation filters
through unequal distribution, extortion, nepotism, law-
lessness, and abuse of political power, which are deeply
embedded within the political system in Bangladesh
(TIB, 2020), which extends the inequitable distribution
of resources exacerbating existing inequalities and re-
source shortages for various communities. So it is very
much essential to know the overall disaster management
knowledge and practices in the affected communities
and how it differs among different socio-economic
groups. In response to strengthening the community’s
resilience to disaster management, after cyclone Aila
2009, different NGOs and the government have started a
series of training and campaign programs to increase
disaster preparedness (Sadik et al., 2018). Again, house-
holds’
affiliation
with
either
government
or
non-
government is the key to securing more external sup-
ports during disaster management. Another significant
factor, which is not considered in our study, is the trust
in the government’s responsibility. In their study, Ack-
erly et al. (2015) claims that those who did not show any
trust in the responsibility of the ruling government get
more external support, confirmed the local politician
wants to win the next election. Therefore, they provided
more assistance to those who did not trust their activity. Future studies need to revisit such dynamics of political On the contrary, the regression model shows that the
number of family members and the family's monthly in-
come is negatively and significantly associated with ac-
cess to relief. It depicts the differences in the selection
process of the recipients based on the scale and level of
investment in relief and rehabilitation programs. Extent of the influence on disaster management by
gender and individual-level engagement in local politics As de-
scribed above, the wealthier people received quality
housing than the poorer, whereas the poorer received
more relief goods than the wealthier. Thus, the poorer
can solve their immediate food problems but hardly re-
ceive any stable housing options, which manifests their
dependency on external support during every extreme
event. Finally, the logistic model predicts that people engaged
in politics have more significant opportunities for receiv-
ing relief. However, there is no significant evidence that
gender differences in politics play any role in getting
extra relief. It was because of the remoteness of study
villages
which
also
shows
the
cultural
barriers
to
women’s openness in society. The following section pre-
sents more details of these socio-cultural and political
aspects in post-disaster society based on the group dis-
cussion and key informant's interviews. Funding Open Access funding enabled and organized by Projekt DEAL. This
publication is not intended to reflect the opinions of these institutions or
individuals. Any errors remain the responsibility of the authors. B. Mallick has
received funding from the European Union’s Horizon 2020 research and
innovation program under the Marie Skłodowska-Curie grant agreement No
846129. Authors’ contributions ZS and BM conceptualized the paper and wrote the first draft in
collaboration with BB and SCS. The methodology was designed by BM and
ZS. While DB, PPB, ABK, TR and MYT collected and cleaned data under the
supervision of ZS, the formal analysis was conducted by BB, ZS and BM. The
statistical modeling was performed by BB and ZS, while MYT, DB, ABK, and
PPB were engaged in the visualization of the results. The review and editing
was made by BM, ZS and BB. The authors read and approved the final
manuscript. Availability of data and materials
Th d
d d
i
d/ The datasets generated during and/or analyzed during the current study are
available from the corresponding author on reasonable request. Ethics approval and consent to participate In this study, the steps/procedures for data collection, treatment, and
analysis have been provided. Equally, the free prior informed consent of all
participants was sought before data collection. Prior informed consent of all
participants was sought before data collection. Discussion This study explains whether gender plays any role in
getting post-disaster relief and rehabilitation support. Most people are illiterate or have a primary education
level and lack proper knowledge of disaster management
and climate change adaptation. A significant relationship
has been found between gender and relief support. Women receive more relief than men, but widows and
divorced couples are prioritized, and married women re-
ceive minimal relief. Likewise, older people do not re-
ceive much relief in the aftermath of a disaster. Families
having fewer dependent members earn more and be-
come more resilient than families with many dependent
members. Actually, pre-disaster political trajectories are the most
significant influence on post-disaster outcomes (Mallick, Page 10 of 12 Page 10 of 12 Page 10 of 12 Sultana et al. Climate Action (2022) 1:7 Sultana et al. Climate Action Sultana et al. Climate Action (2022) 1:7 engagement and opposition’s trust in the ruling party’s
role in post disaster relief and rehabilitation activites. engagement and opposition’s trust in the ruling party’s
role in post disaster relief and rehabilitation activites. Competing interests The authors declare that they have no competing interests. Consent for publication
f
d
f Prior informed consent for publication of all participants was sought before
data collection. Conclusion The significant findings of this study delineate that
the aftermath disaster relief and rehabilitation inter-
vention are closely connected to the demographic,
socio-economic,
and
socio-political
factors;
mainly,
the acquisition of disaster relief is mostly gender and
political connectedness biased. The root cause of nat-
ural disasters in the study area is its unique geo-
graphical location: the surrounding rivers. Among the
surrounded rivers, the Shibsa River is the largest and
significantly influences natural disasters. During tidal
surges, excess water from the river often enters the
area and causes floods every year. The affected people
lose their valuable assets and often lead miserable life
having no or very few livelihood options. Therefore,
they require aftermath disaster relief and rehabilita-
tion support from both the GOs and NGOs for their
life improvement and livelihood opportunities, espe-
cially to adapt to the adverse effects of the disasters
and reduce the future vulnerability of upcoming di-
sasters. However, in reality, they suffer much in the
aftermath of the disaster, not having the expected
support from the local government and other organi-
zations. As a result, the overall disaster management
interventions at the grassroots level are not satisfac-
tory to the local people. Nepotism tendency, corrup-
tion, illiteracy, lack of awareness, and responsibility
are critical factors influencing the instability of the
disaster management activities there. The male re-
spondents who are engaged in politics receive after-
math disaster relief support. Thus, it is observed that
social supremacy still dominates the local decision-
making processes. Author details
1 1Department of Environmental Science and Disaster Management,
Bangabandhu Sheikh Mujibur Rahman Science and Technology University,
Gopalganj 8100, Bangladesh. 2Chair of Environmental Development and Risk
Management, Faculty of Environmental Sciences, Technische Universität
Dresden, 01062 Dresden, Germany. 3Institute of Behavioral Science, University
of Colorado Boulder, Boulder, USA. 4Coastal Research Centre (CRC), Khulna
9250, Bangladesh. 5Environmental Science Discipline, Khulna University,
Khulna 9208, Bangladesh. 6Department of Sociology, University of Dhaka,
Dhaka 1000, Bangladesh. 7Department of Agronomy, Khulna Agricultural
University, Khulna 9202, Bangladesh. In a nutshell, people’s socio-political characteristics
and engagement in local-level politics, irrespective of
gender, influence disaster politics and management is-
sues when a natural disaster strikes (Mallick & Vogt,
2012; Roy, 2020b). Therefore, measuring the impact of
social and political factors through research is vital for
future disaster management and climate change adapta-
tion planning. Alongside women’s empowerment, regu-
lar monitoring and evaluation of relief and rehabilitation
programs should reduce the traditional barriers to ad-
equate disaster management and relief arising from
(dis)connectedness to local social and political power. Moreover, eradicating corruption, nepotism, and bur-
eaucratic entanglement should be considered in the dis-
aster management policy agenda. Received: 25 October 2021 Accepted: 23 March 2022 Received: 25 October 2021 Accepted: 23 March 2022 References
k l Ackerly BA, Anam MM, Gilligan JM (2015) Environment, political economies and
livelihood change. In: Etzold B, Mallick B (eds) Environment, migration and
adaptation: Evidence and politics of climate change in Bangladesh. AH
Development Publishing House (AHDPH), Bangladesh, pp 27–40 ADW, & UNU-EHS. (2015). ADW(Alliance DevelopmentWorks) and UNU-EHS
(United Nations University - Institute for Environment and Human Security),
World Risk Report 2015. Available online: http://www.worldriskreport.org/
(accessed on 08 Feb 2021). ADW, & UNU-EHS. (2015). ADW(Alliance DevelopmentWorks) and UNU-EHS
(United Nations University - Institute for Environment and Human Security),
World Risk Report 2015. Available online: http://www.worldriskreport.org/
(accessed on 08 Feb 2021). Akinci F (2004) The aftermath of disaster in urban areas: An evaluation of the
1999 earthquake in Turkey. Cities 21(6):527–536. https://doi.org/10.1016/j. cities.2004.08.010 Alam K, Shamsuddoha M, Tanner T, Sultana M, Huq MJ, Kabir SS (2011) The
Political Economy of Climate Resilient Development Planning in Bangladesh. IDS B ll 42(3) 52 61 h
//d i
/10 1111/j 1759 5436 2011 00222 Alam K, Shamsuddoha M, Tanner T, Sultana M, Huq MJ, Kabir SS (2011) The
Political Economy of Climate Resilient Development Planning in Bangladesh
IDS Bull 42(3):52–61 https://doi org/10 1111/j 1759-5436 2011 00222 x Alam K, Shamsuddoha M, Tanner T, Sultana M, Huq MJ, Kabir SS (2011) The
Political Economy of Climate Resilient Development Planning in Bangladesh. IDS Bull 42(3):52–61. https://doi.org/10.1111/j.1759-5436.2011.00222.x
Aryanti T, Muhlis A (2020) Disaster, gender, and space: Spatial vulnerability in
post-disaster shelters. IOP Confer Series 447(1). https://doi.org/10.1088/1
755-1315/447/1/012012 Aryanti T, Muhlis A (2020) Disaster, gender, and space: Spatial vulnerability in
post-disaster shelters. IOP Confer Series 447(1). https://doi.org/10.1088/1
755-1315/447/1/012012 y
org/10.13189/sa.2015.030203 g
g
Islam ARMT, Tasnuva A, Islam T, Haque R (2014) Management Approach to Islam ARMT, Tasnuva A, Islam T, Haque R (2014) Management Approach to
Disaster Scenario in Bangladesh: An overview. Int J Sci Res Publ 4(3):1–7 Baytiyeh H (2017) Socio-cultural characteristics: the missing factor in disaster risk
reduction strategy in sectarian divided societies. Int J Disaster Risk Reduct 21:
63–69. https://doi.org/10.1016/j.ijdrr.2016.11.012 Islam R, Walkerden G (2017) Social networks and challenges in government
disaster policies: A case study from Bangladesh. Int J Disaster Risk Reduction
22:325–334. https://doi.org/10.1016/j.ijdrr.2017.02.011 BBS. (2011). Bangladesh Bureau of Statistics Populatio BBS. (2011). Bangladesh Bureau of Statistics Population and housing census. Bhavnani R (2006) The author would like to thank Doug Bond, Joe Bond, Jennifer
Leaning, and Gary King for comments on this project. Harvard University,
Cambridge Kazal MMH, Rahman S, Hossain MZ (2017) Poverty Profiles and Coping Strategies
of the Haor (Ox-bow lake) Households in Bangladesh. J Poverty Alleviation
Int Dev 8(1):167–191 Bianchi E, Malki-Epshtein L (2021) Evaluating the risk to Bangladeshi coastal
infrastructure from tropical cyclones under climate change. Int J Disaster Risk
Reduction 57(4):102147. https://doi.org/10.1016/j.ijdrr.2021.102147 Khan MMH, Nahar N (2014) Natural disasters: socio-economic impacts in
Bangladesh. Banglavision 13(1):58–67 Khan MMH, Nahar N (2014) Natural disasters: socio-economic impacts in
Bangladesh. Banglavision 13(1):58–67 Khan MR, Rahman MA (2007) Partnership approach to disaster management in
Bangladesh: a critical policy assessment. Nat Hazards 41:359–378. https://doi. org/10.1007/s11069-006-9040 Khan MR, Rahman MA (2007) Partnership approach to disaster management in
Bangladesh: a critical policy assessment. Nat Hazards 41:359–378. https://doi. org/10.1007/s11069-006-9040 Biswas B, Mallick B (2020) Livelihood diversification as key to long-term non-
migration: evidence from coastal Bangladesh. Environ Dev Sustain. https://
doi.org/10.1007/s10668-020-01005-4 Kovach T (2013) The Politics of Disaster Response: Disaster Diplomacy and the
Responsibility to Protect after Cyclone Nargis Kovach T (2013) The Politics of Disaster Response: Disaster Diplomacy and the
Responsibility to Protect after Cyclone Nargis Buhaug H, Gleditsch NP, Theisen OM (2008) Implications of climate change for
armed conflict. In social dimensions of climate change. https://doi.org/http://
siteresources.worldbank.org/INTRANETSOCIALDEVELOPMENT/Resources/
SDCCWorkingPaper_Conflict.pdf. Accessed 08 July 2021 Lebni JY, Khorami F, Ebadi Fard Azar F, Khosravi B, Safari H, Ziapour A (2020)
Experiences of rural women with damages resulting from an earthquake in
Iran: A qualitative study. BMC Public Health 20(1):1–13. https://doi.org/10.11
86/s12889-020-08752-z Care. (2008). Care (Cooperative for Assistance and Relief Everywhere). Chipangura P, Van Niekerk D, Van Der Waldt G (2017) Disaster risk problem
framing: Insights from societal perceptions in Zimbabwe. report_116.pdf. Accessed 14 July 2021 https://doi.org/10.1093/
jhuman/hup002 Mendoza MJ (2014) The Policies, Practices and Politics of Dealing with Disasters:
An Analysis of How Disaster Risk Reduction Policies and Actions are Framed
in the Philippines Mendoza MJ (2014) The Policies, Practices and Politics of Dealing with Disasters:
An Analysis of How Disaster Risk Reduction Policies and Actions are Framed
in the Philippines Goldstone JA, Bates RH, Epstein DL, Gurr TR, Lustik MB, Marshall MG, Ulfelder J,
Woodward M (2010) A Global Model for Forecasting Political Instability. Am J
Polit Sci 54(1):190–208. https://doi.org/10.1111/j.1540-5907.2009.00426.x Murthy RK (2009) Review of Sexual and Reproductive Health and Rights in the
Context of Disasters in Asia. In: Asian Pacific Resource and Research Centre
for Women (ARROW). https://doi.org/10.13140/RG.2.1.3925.5201 Murthy RK (2009) Review of Sexual and Reproductive Health and Rights in the
Context of Disasters in Asia. In: Asian Pacific Resource and Research Centre
for Women (ARROW). https://doi.org/10.13140/RG.2.1.3925.5201 Gotham KF (2012) Disaster, Inc.: Privatization and post-Katrina rebuilding in New
Orleans. Perspect Politics 10(3):633–646. https://doi.org/10.1017/S1537592712
00165X Oliver AJ, Reeves A (2015) The politics of disaster – Nicaragua. In: Scott R, Kosslyn
S (eds) Emerging Trends in the Social and Behavioral Sciences. Wiley, pp 1–8. https://doi.org/10.1111/j.1467-7717.1985.tb00949.x Oliver AJ, Reeves A (2015) The politics of disaster – Nicaragua. In: Scott R, Kosslyn
S (eds) Emerging Trends in the Social and Behavioral Sciences. Wiley, pp 1–8. https://doi.org/10.1111/j.1467-7717.1985.tb00949.x Oliver AJ, Reeves A (2015) The politics of disaster – Nicaragua. In: Scott R, Kosslyn
S (eds) Emerging Trends in the Social and Behavioral Sciences. Wiley, pp 1–8. https://doi.org/10.1111/j.1467-7717.1985.tb00949.x Hapeman K (2012) The effects of politics on natural sisasters: lessons learned
from Bangladesh. https://www.du.edu/korbel/crric/media/documents/katie_
hapeman1.pdf. Accessed 24 January 2021 Ozcelik H, Langton N, Aldrich H (2008) Doing well and doing good: The
relationship between leadership practices that facilitate a positive emotional
climate and organizational performance. J Manag Psychol 23(2):186–203. https://doi.org/10.1108/02683940810850817 Ozcelik H, Langton N, Aldrich H (2008) Doing well and doing good: The
relationship between leadership practices that facilitate a positive emotional
climate and organizational performance. J Manag Psychol 23(2):186–203. https://doi.org/10.1108/02683940810850817 Hayat E, Amaratunga D (2017) The Role of Local Government in Post-Disaster
Road Reconstruction: Assessment of Factors Affecting Local Government
Road Maintenance Capacity. In: Djalante R, Garschagen M, Thomalla F, R. (eds) Shaw (Eds.), Disaster Risk Reduction in Indonesia. Springer, pp 255–279. https://doi.org/10.1007/978-3-319-54466-3 Parvin GA, Sakamoto M, Shaw R, Nakagawa H, Sadik MS (2019) Evacuation
scenarios of cyclone Aila in Bangladesh: Investigating the factors influencing
evacuation decision and destination. y
org/10.13189/sa.2015.030203 Int J Disaster Risk
Reduction 22:317–324. https://doi.org/10.1016/j.ijdrr.2017.02.012 Macrae D (2014) Managing a political crisis after a disaster: how concern
assessment can address the political aspects involved in framing a
solution. J Risk Res 19(8):1036–1042. https://doi.org/10.1080/13669877.2
014.910693 Ciampi MC, Gell F, Lasap L, Turvill E (2011). Gender and Disaster Risk Reduction: A
training pack. http://reliefweb.int/sites/reliefweb.int/files/resources/Full_
report_116.pdf. Accessed 14 July 2021 Mahmud T, Prowse M (2012) Corruption in cyclone preparedness and relief
efforts in coastal Bangladesh: Lessons for climate adaptation? Global Environ
Change 22(4):933–943. https://doi.org/10.1016/j.gloenvcha.2012.07.003 report_116.pdf. Accessed 14 July 2021 Clar C (2019) How demographic developments determine the management of
hydrometeorological hazard risks in rural communities: The linkages between
demographic and natural hazards research. Wires Water 6(6):1–20. https://doi. org/10.1002/wat2.1378 Mallick B (2012) Social disparities in disaster-induced migration process. Available
at SSRN: https://ssrn.com/abstract=2173166. Accessed 15 Oct 2021 Mallick B (2014) Cyclone shelters and their locational suitability: an empirical
analysis from coastal Bangladesh. Disasters 38(3):654–671. https://doi.org/1
0.1111/disa.12062 Clement V, Rigaud KK, de Sherbinin AJB, Adamo S, Schewe J, Sadiq N, Shabahat
E (2021) Groundswell Part 2 : Acting on Internal Climate Migration. World
Bank, Washington, DC © World Bank. https://openknowledge.worldbank.org/
handle/10986/36248; License: CC BY 3.0 IGO. Accessed 25 February 2022 Mallick B, Rahaman KR, Vogt J (2011) Coastal livelihood and physical
infrastructure in Bangladesh after cyclone Aila. Migr Adapt Strateg Global
Change 16(6):629–648. https://doi.org/10.1007/s11027-011-9285-y Cohen C, Werker ED (2008) The Political Economy of “Natural” Disasters. J Conflict
Resolut 52(6):795–819. https://doi.org/10.1177/0022002708322157 Mallick B, Vogt J (2012) Cyclone, coastal society and migration: empirical
evidence from Bangladesh. Int Dev Plan Rev 34(3):217–240 Enarson E, Meyreles L (2004) International Perspectives on Gender and Disaster:
Differences and Possibilities. Int J Sociol Soc Policy 24:1–26. https://doi.org/1
0.1108/01443330410791064 Mallick B, Vogt J (2014) Population displacement after cyclone and its
consequences: Empirical evidence from coastal Bangladesh. Nat Hazards 73:
191–212. https://doi.org/10.1007/s11069-013-0803-y Mallick B, Vogt J (2014) Population displacement after cyclone and its
consequences: Empirical evidence from coastal Bangladesh. Nat Hazards 73:
191–212. https://doi.org/10.1007/s11069-013-0803-y Enia J (2009) Shaking the Foundations of Violent Civil Conflict: Institutions,
Disasters, and the Political Economies of State-Rebel Interaction Enia J (2009) Shaking the Foundations of Violent Civil Conflict: Institutions,
Disasters, and the Political Economies of State-Rebel Interaction
Farrior S (2009) Human rights advocacy on gender issues: Challenges and
opportunities. J Human Rights Pract 1(1):83–100. https://doi.org/10.1093/
jhuman/hup002 Masaba S, Mungai DN, Isabirye M, Nsubuga H (2017) Implementation of landslide
disaster risk reduction policy in Uganda. Int J Disaster Risk Reduction 24:326–
331. https://doi.org/10.1016/j.ijdrr.2017.01.019 Masaba S, Mungai DN, Isabirye M, Nsubuga H (2017) Implementation of landslide
disaster risk reduction policy in Uganda. Int J Disaster Risk Reduction 24:326–
331. https://doi.org/10.1016/j.ijdrr.2017.01.019 Farrior S (2009) Human rights advocacy on gender issues: Challenges and
opportunities. J Human Rights Pract 1(1):83–100. Code availability
Not applicable. Code availability
Not applicable. Page 11 of 12 Page 11 of 12 Page 11 of 12 Sultana et al. Climate Action Sultana et al. Climate Action (2022) 1:7 Sultana et al. Climate Action (2022) 1:7 (2022) 1:7 Auerbach LW, Goodbred SL Jr, Mondal DR, Wilson CA, Ahmed KR, Roy K, Steckler
MS, Small C, Gilligan JM, Ackerly BA (2015) Nature Climate Change Cambridge, United Kingdom and New York, NY, USA. https://doi.org/10.101
7/cbo9780511976988.002 IPCC. (2021). AR6 Climate Change 2021: The Physical Science Basis in the Sixth
Assessment Report of the Intergovernmental Panel on Climate Change. Islam ARMT, Tasnuva A, Islam T, Haque R (2014) Management Approach to
Disaster Scenario in Bangladesh: An overview. Int J Sci Res Publ 4(3):1–7
Islam R, Walkerden G (2017) Social networks and challenges in government
disaster policies: A case study from Bangladesh. Int J Disaster Risk Reduction
22:325–334. https://doi.org/10.1016/j.ijdrr.2017.02.011 y
Azad MAK, Khan MM (2015) Post Disasters Social Pathology in Bangladesh: A
Case Study on AILA Affected Areas. Sociol Anthropol 3(2):85–94. https://doi. org/10.13189/sa.2015.030203 Azad MAK, Khan MM (2015) Post Disasters Social Pathology in Bangladesh: A IPCC. (2021). AR6 Climate Change 2021: The Physical Science Basis in the Sixth
Assessment Report of the Intergovernmental Panel on Climate Change. Case Study on AILA Affected Areas. Sociol Anthropol 3(2):85–94. https://doi. org/10.13189/sa.2015.030203 report_116.pdf. Accessed 14 July 2021 Prog Disaster Sci. 2(100032):1–12. https://doi.org/10.1016/j.pdisas.2019.100032 Parvin GA, Sakamoto M, Shaw R, Nakagawa H, Sadik MS (2019) Evacuation
scenarios of cyclone Aila in Bangladesh: Investigating the factors influencing
evacuation decision and destination. Prog Disaster Sci. 2(100032):1–12. https://doi.org/10.1016/j.pdisas.2019.100032 Heide E (2004) Common misconceptions about disasters: Panic, the “disaster
syndrome,” and looting. In: O’Leary M (ed) The first 72 hours: a community
approach to disaster preparedness. iUniverse Publishing, Lincoln, pp 340–380 Pelling M, Dill K (2008) Disaster Politics : from Social Control to Human Security. In: Environment, Politics and Development Working Paper Series Pelling M, Dill K (2008) Disaster Politics : from Social Control to Human Security. In: Environment, Politics and Development Working Paper Series Pelling M, Dill K (2008) Disaster Politics : from Social Control to Human Security. In: Environment, Politics and Development Working Paper Series Hemachandra K, Amaratunga D, Haigh R (2018) Role of women in disaster risk
governance. Procedia Eng 212(February):1187–1194. https://doi.org/10.1016/j. proeng.2018.01.153 Razzaque MA, Alamgir M, Rahman MM (2019) Climate Change Vulnerability in
Dacope Upazila, Bangladesh. J Sci Res Rep 21(4):1–12. https://doi.org/10. 9734/jsrr/2018/45471 Razzaque MA, Alamgir M, Rahman MM (2019) Climate Change Vulnerability in
Dacope Upazila, Bangladesh. J Sci Res Rep 21(4):1–12. https://doi.org/10. 9734/jsrr/2018/45471 IPCC (2014) Summary for policymakers. In: Climate Change 2014: Impacts,
Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects. Contribution of Working Group II to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change. Cambridge University Press, Roy S (2020a) Post-cyclone Aila and Mobility Rights of the Shora Muslim Women
of the Bangladesh Sundarbans Forest. In: Leal FW, Nagy G, Borga M, Chávez
MP, Magnuszewski A (eds) Climate Change, Hazards and Adaptation Options. Roy S (2020a) Post-cyclone Aila and Mobility Rights of the Shora Muslim Women
of the Bangladesh Sundarbans Forest. In: Leal FW, Nagy G, Borga M, Chávez
MP, Magnuszewski A (eds) Climate Change, Hazards and Adaptation Options. Page 12 of 12 Sultana et al. Climate Action (2022) 1:7 Sultana et al. Climate Action (2022) 1:7 Sultana et al. Climate Action Climate Change Management. Springer, Cham. https://doi.org/10.1007/978-3-
030-37425-9_32 Climate Change Management. Springer, Cham. https://doi.org/10.1007/978-3-
030-37425-9_32 Roy S (2020b) Transforming Gendered Lives and Livelihoods in Post-disaster
Settings in the Bangladesh Sundarbans Forest. In: Leal FW, Nagy G, Borga M,
Chávez MP, Magnuszewski A (eds) Climate Change, Hazards and Adaptation
Options. Climate Change Management. Springer, Cham. report_116.pdf. Accessed 14 July 2021 https://doi.org/10.1
007/978-3-030-37425-9_24 Sadik MS, Nakagawa H, Rahman R, Shaw R, Kawaike K, Fujita K (2018) A Study on
Cyclone Aila Recovery in Koyra, Bangladesh: Evaluating the Inclusiveness of
Recovery with Respect to Predisaster Vulnerability Reduction. Int J Disaster
Risk Sci 9(1):28–43. https://doi.org/10.1007/s13753-018-0166-9 Risk Sci 9(1):28–43. https://doi.org/10.1007/s13753-018-0166-9 Sökefeld M (2012) Exploring the Link between Natural Disasters and Politics: Case
Studies of Pakistan and Peru. Scrutiny 6:1–32. https://doi.org/10.5282/ubm/
epub.22296 Sultana Z, Mondal P, Roy T, Biswas B, Mallick B (2021) Socio-political Distancing
Amid Disaster: Empirical Evidence from Bangladesh. In: Alam GMM, Erdiaw-
Kwasie MO, Nagy GJ, Leal FW (eds) Climate Vulnerability and Resilience in
the Global South. Climate Change Management. Springer, Cham. https://doi. org/10.1007/978-3-030-77259-8_18 Takasaki Y (2012) Natural disasters, gender and handicrafts. J Dev Stud 48(1):115–
132. https://doi.org/10.1080/00220388.2011.598509 Takasaki Y (2012) Natural disasters, gender and handicrafts. J Dev Stud 48(1):115–
132. https://doi.org/10.1080/00220388.2011.598509
Terry G (2009) No climate justice without gender justice: An overview of the
issues. Gender Dev 17(1):5–18. https://doi.org/10.1080/13552070802696839
TIB. (2020). Annual Report - Transparency International Bangladesh (TIB). Terry G (2009) No climate justice without gender justice: An overview of the
issues. Gender Dev 17(1):5–18. https://doi.org/10.1080/13552070802696839
TIB. (2020). Annual Report - Transparency International Bangladesh (TIB). Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
|
https://openalex.org/W2982916409
|
https://aacr.figshare.com/articles/journal_contribution/Supplementary_Figure_S3_from_Tumor-Associated_Fibroblasts_Promote_HER2-Targeted_Therapy_Resistance_through_FGFR2_Activation/22473386/1/files/39924878.pdf
|
Aragonese
| null |
Tumor-Associated Fibroblasts Promote HER2-Targeted Therapy Resistance through FGFR2 Activation
|
Clinical cancer research
| 2,020
|
cc-by
| 1,414
|
A
B
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
SK-BR-3
BT-474
MDA-MB-453
***
***
***
***
***
***
Control
Trastuzumab resistant
Lapatinib resistant
IC50Tz res=20,1µg/ml
IC50Tz res=14,2µg/ml
IC50wt=5,6µg/ml
IC50wt=3,7µg/ml
IC50wt=3,8µg/ml
IC50wt=2,3µM
IC50wt=2,4µM
IC50wt=0,1µM
Tz-resistant
Parental
Lap-resistant
-
+ -
-
+ -
-
-
-
+ -
-
-
+
Trastuzumab
Lapatinib
SK-BR-3
BT-474
MDA-MB-453
Lap-resistant
Parental
Tz-resistant
BT-474
MDA-MB-453
SK-BR-3
C A
B
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
SK-BR-3
BT-474
MDA-MB-453
***
***
***
***
***
***
Control
Trastuzumab resistant
Lapatinib resistant
IC50Tz res=20,1µg/ml
IC50Tz res=14,2µg/ml
IC50wt=5,6µg/ml
IC50wt=3,7µg/ml
IC50wt=3,8µg/ml
IC50wt=2,3µM
IC50wt=2,4µM
IC50wt=0,1µM
Tz-resistant
Parental
Lap-resistant
-
+ -
-
+ -
-
-
-
+ -
-
-
+
Trastuzumab
Lapatinib
SK-BR-3
BT-474
MDA-MB-453
Lap-resistant
Parental
Tz-resistant
BT-474
MDA-MB-453
SK-BR-3
C A
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
SK-BR-3
BT-474
MDA-MB-453
***
***
***
***
***
***
Control
Trastuzumab resistant
Lapatinib resistant
IC50Tz res=20,1µg/ml
IC50Tz res=14,2µg/ml
IC50wt=5,6µg/ml
IC50wt=3,7µg/ml
IC50wt=3,8µg/ml
IC50wt=2,3µM
IC50wt=2,4µM
IC50wt=0,1µM A
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
SK-BR-3
BT-474
MDA-MB-453
***
***
***
***
***
***
Control
Trastuzumab resistant
Lapatinib resistant
IC50Tz res=20,1µg/ml
IC50Tz res=14,2µg/ml
IC50wt=5,6µg/ml
IC50wt=3,7µg/ml
IC50wt=3,8µg/ml
IC50wt=2,3µM
IC50wt=2,4µM
IC50wt=0,1µM 0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
BT-474
***
IC50Tz res=14,2µg/ml
IC50wt=3,7µg/ml 0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
MDA-MB-453
***
IC50Tz res=20,1µg/ml
IC50wt=5,6µg/ml 0
5 0
1 0 0
1 5 0
2 0 0
2 5 0
6 0
8 0
1 0 0
1 2 0
T ra s tu z u m a b (g /m l)
% s u rv iv a l
SK-BR-3
***
IC50wt=3,8µg/ml A MDA-MB-453 T ra s tu z u m a b (g /m l) T ra s tu z u m a b (g /m l) 6
0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
***
***
Control
Trastuzumab resistant
Lapatinib resistant
IC50wt=2,3µM 0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
***
IC50wt=0,1µM 0
2
4
6
0
5 0
1 0 0
1 5 0
L a p a tin ib (M )
% s u rv iv a l
***
IC50wt=2,4µM Control
Trastuzumab resistant
Lapatinib resistant L a p a tin ib (M ) L a p a tin ib (M ) B B
Tz-resistant
Parental
Lap-resistant
-
+ -
-
+ -
-
-
-
+ -
-
-
+
Trastuzuma
Lapatinib
SK-BR-3
BT-474
MDA-MB-453
C Tz-resistant
Parental
Lap-resistant
-
+ -
-
+ -
-
-
-
+ -
-
-
+
Trastuzumab
Lapatinib
SK-BR-3
BT-474
MDA-MB-453
Lap-resistant
Parental
Tz-resistant
BT-474
MDA-MB-453
SK-BR-3
C Tz-resistant
Parental
Lap-resistant
-
+ -
-
+ -
-
-
-
+ -
-
-
+
Trastuzumab
Lapatinib
SK-BR-3
BT-474
MDA-MB-453
Lap-resistant
Parental
Tz-resistant
BT-474
MDA-MB-453
SK-BR-3
C Trastuzumab
Lapatinib Lap-resistant
Parental
Tz-resistant
BT-474
MDA-MB-453
SK-BR-3
C C Lap-resistant Tz-resistant Fernandez-Nogueira et al., Supplementary Figure S3
|
https://openalex.org/W1584838080
|
https://revistes.uab.cat/ciencies/article/download/n23-sedo-fontnova-barrau-etal/378
|
Catalan; Valencian
| null |
La Motxilla Bioclimàtica
|
Ciències
| 2,012
|
cc-by
| 2,064
|
El per què climàtica on es treballa tècnicament la relació entre
l’edifici, la despesa energètica i la qualitat ambien-
tal interior assolida. La Motxilla Bioclimàtica també
apropa aquests coneixements més tecnològics a
les aules de secundària. El medi ambient i la salut son les àrees de medi
natural més rellevants socialment i significatives
per a l’educació primària i secundària. La funcionalitat dels continguts científics en
qualsevol àmbit de l'educació ambiental sempre es
fa ben evident quan assolim un grau d’experimen-
tació important. En el cas que ens ocupa, els tre-
balls sobre estalvi i eficiència energètica es trans-
formen en treball científic sempre que la mesura, el
tractament de dades, les hipòtesis, etc, entren en
joc. La Motxilla Bioclimàtica es planteja com un tre-
ball experimental on fer, pensar i comunicar és la
seva raó de ser (Pujol, 2003). L'activitat científica
porta a plantejar i comprendre qüestions de l'im-
pacte de la nostra activitat sobre l'ambient i sobre
la salut. Però el diàleg entre la educació ambiental, l’e-
ducació per la salut i l’educació científica sovint no
és planer. Potser sigui per influències històriques i
culturals però creiem que també per dificultats me-
todològiques (Zeyer i Dillon, 2012). En aquest sentit cal una nova mirada que aporti
didàctiques adreçades a la integració d’aquests
àmbits des de la perspectiva actual de la complexi-
tat. Presentem en aquest treball la Motxilla Biocli-
màtica, una proposta didàctica que ha madurat
amb l'objectiu d’educar en ciències, medi ambient i
salut partint d’un context quotidià. El projecte educatiu la Motxilla Bioclimàtica és
impulsat i coordinat des del Camp d’Aprenentatge
de Juneda. El lligam entre la despesa energètica al centre i
la salut del seus ocupants és fàcil d'intuir però difícil
de quantificar. La despesa energètica de l'edifici és
una realitat força abstracta per als alumnes. Calia
apropar el balanç d’energia al propi cos i a l’aula
per relacionar-los amb les condicions ambientals i
la gestió de l’aula i la pròpia salut: un repte. En
aquest context la Motxilla Bioclimàtica és original
en proposar un treball experimental entre profes-
sors i alumnes de l’aula i les seves condicions bio-
climàtiques amb l'objectiu, en primer terme, de co-
nèixer la seva qualitat a nivell energètic i ambiental. I en segon terme, l'impacte sobre el medi ambient i
la salut dels alumnes. Experiències didàctiques i treballs pràctics Experiències didàctiques i treballs pràctics Ciències 23 (2012) Paraules clau: educació ambiental, balanç d’energia, mesura, qualitat ambient interior, bioclimatisme Paraules clau: educació ambiental, balanç d’energia, mesura, qualitat ambient interior, bioclimatisme La Motxilla Bioclimàtica Àngel Sedó Camp d'Aprenentatge Juneda (Dept. Ensenyament)
Núria Fontova Associació Lo Secanet
Jérôme Barrau Universitat de Lleida
Manel Ibáñez Universitat de Lleida Àngel Sedó Camp d'Aprenentatge Juneda (Dept. Ensenyament)
Núria Fontova Associació Lo Secanet
Jérôme Barrau Universitat de Lleida
Manel Ibáñez Universitat de Lleida Presentem la Motxilla Bioclimàtica, una proposta que té per objectiu educar en ciències,
medi ambient i salut partint d’un context quotidià. La Motxilla Bioclimàtica facilita a l'equip
de professors i alumnes un conjunt d'eines per avaluar la qualitat ambiental i bioclimàtica
de les aules. 1. Els índexs de qualitat de l’ambient interior La primera condició a complir a l’aula per fer
l’activitat és que el nostre cos pugui satisfer el ba-
lanç tèrmic. Això vol dir que els mecanismes fisio-
lògics de la termo-regulació siguin capaços de por-
tar l’organisme a un estat d’equilibri tèrmic entre el
guany d’energia tèrmica (d’origen ambiental i me-
tabòlic) i la seva eliminació. Però l’equilibri tèrmic
en si mateix està lluny de proporcionar sensació de
confort. L’organisme pot aconseguir satisfer el ba-
lanç tèrmic en un ampli ventall de combinacions de
situacions ambientals i taxes d’activitat però només
una estreta franja d’elles condueix a situacions que
les persones reconeixem com a confortables. En el món occidental actual, el manteniment de
les condicions ambientals òptimes a l’interior dels
edificis s’assoleix en gran mesura augmentant el
consum energètic. Valors rècord en el consum d’e-
nergia són noticia gairebé cada nou hivern i estiu. La societat no renuncia a viure en òptimes condici-
ons en els ambients interiors, que permeten gaudir
de salut, benestar i confort. Ben al contrari, un mal
ambient interior interfereix en el rendiment escolar,
causa desconfort i irritació, així como problemes de
salut en els escolars i el professorat (Bakó-Biró i
al., 2007). Aquesta qualitat ambiental també s’ha d’assolir
en les nostres aules. Però ho podem fer malbara-
tant energia o bé aprofitant els recursos naturals
amb una gestió adequada dels espais i una bona
arquitectura bioclimàtica de l’edifici. L'impacte d'una
i altra forma sobre el medi ambient és ben diferent. L’ésser humà emet un elevat nombre de subs-
tàncies: diòxid de carboni, compostos orgànics vo-
làtils, aerosols que poden contenir microorganis-
mes viables o no, vapor d’aigua, etc. De tots ells el
més important des del punt de vista quantitatiu és
el diòxid de carboni. Això el fa idoni per al càlcul del
cabal de ventilació necessari per assolir un ambient
de qualitat respiratòria acceptable. Quan els nivells
de diòxid de carboni superen certs valors, moltes
persones comencen a experimentar incomoditat,
mal de cap, cansament i problemes respiratoris,
depenent de la concentració i de la durada de l’ex-
posició. Uns símptomes que s’agreugen en el cas
dels infants. Aquest plantejament ens porta a investigar la
qualitat ambiental i la qualitat bioclimàtica de l’aula. La qualitat ambiental dóna una mesura de les con-
dicions de benestar i salut. La qualitat bioclimàtica
orienta sobre la sostenibilitat energètica de com
assolim aquestes condicions. El per què El professorat té a l’abast propostes d’educació
ambiental per auditar el consum energètic del seu
centre amb l’objectiu de conscienciar els alumnes
sobre la limitació dels recursos fòssils, la seguretat
energètica o el canvi climàtic. Però molt sovint l'ac-
tivitat queda deslligada dels continguts científics
que la fan possible i gairebé sempre oblida les in-
terrelacions que hi ha entre la despesa energètica
al centre i la salut del seus ocupants. Per altra ban-
da, la relació entre les característiques constructi-
ves d'un edifici i la seva salubritat és ben coneguda
pels arquitectes. Hi ha escoles d'arquitectura bio- 2 Experiències didàctiques i treballs pràctics Ciències 23 (2012) 1. Els índexs de qualitat de l’ambient interior La Motxilla Bioclimàtica planteja una primera
etapa d’experimentació i treball científic: Són bones
les condicions de treball? Quin efectes tenen sobre
el nostre rendiment i la salut? Quin impacte energè-
tic té assolir aquest ambient interior? I ofereix d’una
banda les eines conceptuals i procedimentals per
orientar la tasca i, d’altra banda, els aparells de
mesura per portar-la a terme. Un cop avaluada l'au-
la, una segona etapa d’acció consisteix en una pro-
posta d'educació ambiental i educació per la salut:
Què podem fer per millorar la qualitat ambiental in-
terior? Com podem reduir l’impacte energètic del
nostre centre? La qualitat de la il·luminació als centres educa-
tius és un factor molt important per determinar la
velocitat, precisió i confort necessaris per desenvo-
lupar les tasques visuals. La il·luminació ha de sa-
tisfer les necessitats psicològiques i emotives dels
alumnes, així como influir positivament en el seu
comportament i en la capacitat de concentració en
la feina. 2. Els índexs de qualitat bioclimàtica Les condicions de l’aula són dinàmiques: en tot
moment tenim fluxos de matèria i energia que mo-
difiquen les condicions ambientals interiors. 4. La Calculadora Bioclimàtica Els índexs de qualitat ambiental és poden ava-
luar puntualment i mantenen tot els seu sentit però
els índexs de qualitat bioclimàtica es basen en un
balanç energètic diari a l’aula. Avaluar-los requereix La Calculadora Bioclimàtica és l’eina informàtica
que facilitat tota la feina de tractament de dades. Tot i que es faciliten els conceptes i relacions per
avaluar els índexs proposats a partir de les mesu-
res, es pot emprar aquesta eina si els usuaris volen
alleugerir la feina de càlcul, no pas la d’anàlisi dels
resultats. Mesures a l’interior de l’aula
Magnitud
Sensor
Activitat
Ocupació
Taxa metabòlica
Ambient
Temperatura
Sensor Wöhler
CDL 210
Humitat relativa
Concentració CO2
Luminància
Luxòmetre PCE-174
Calefacció
Temperatura
entrada
Termòmetre
PCE-T390
Temperatura
sortida
Cabal
Cabalímetre PCE
Il·luminació
Potència elèctrica
Comptador Efenergy2
Tancaments
opacs
Temperatura
superfície
Termòmetre infraroig
PCE-888
Finestres
Irradiació
Piranòmetre SPM1
Superfície
captació
Distanciòmetre
PCE-LDM50
Mesures a l’exterior de l’aula
Ambient
Temperatura
Estació meteorològica
PCE-FWS 20
Humitat relativa
Concentració CO2
Sensor Wöhler
CDL 210
Irradiació
Piranòmetre
SPM1
Taula 1. Equips i paràmetres de mesura 5. Espai d'intercanvi Un cop calculats els diferents índexs bioclimà-
tics de l'aula és el moment d’avaluar els resultats i
passar a l'acció. Com a usuaris de l’espai ens po-
dem plantejar quins canvis es proposen. Apareixeran, de ben segur, els canvis en l'equi-
pament, que seran més complicats, però es poden
analitzar quins elements arquitectònics ajudarien a
obtenir millores a l'aula. Un cop haguem decidit,
amb tots els companys i companyes, quines són
les accions que ens interessa dur a terme, redacta-
rem un informe on s’expliqui l’avaluació realitzada. També s’hi adjuntaran les millores i canvis que es
proposen fer, i els beneficis que aportaran. Aquest informe presentat a la comunitat educa-
tiva del centre es pot compartir a la web de la Mot-
xilla Bioclimàtica. Experiències didàctiques i treballs pràctics Ciències 23 (2012) El balanç d’energia diari plantejat dependrà ne-
cessàriament de les condicions meteorològiques
dels dies estudiats. Per fer els resultats compara-
bles entre localitats i èpoques de l’any s’introdu-
eixen quatre índexs de qualitat bioclimàtica de l’au-
la. Aquests índexs integren tant el comportament
arquitectònic com la gestió que en fan els usuaris. dades representatives de tot un dia. Els aparells de
mesura a disposició a la Motxilla Bioclimàtica me-
suren i emmagatzemen les mesures de forma con-
tínua i segons l'interval que desitgem. Les memòri-
es dels diferents aparells permeten emmagatzemar
dades d’uns quants dies i posteriorment descarre-
gar-les a l’ordinador. La taula 1 detalla les 17 varia-
bles que mesurarem, acompanyades de l’instru-
ment que emprarem en cada cas. Les eines Per portar a terme la nostra investigació sobre
la qualitat ambiental i bioclimàtica de l’aula, la Mot-
xilla Bioclimàtica facilita a l'equip de professors i
alumnes un conjunt d'eines. Unes eines –ofertes a
través de la plana web als centres interessats que
en fan ús de forma parcial o total– que es poden
dividir en 5 grans grups: Per això, un objectiu de la nostra investigació és
determinar aquests fluxos per avaluar la sostenibili-
tat energètica de l’aula. Tots aquest fluxos es com-
binen en el balanç d’energia tèrmica de l’aula, una
relació ben intuïtiva que ens diu que l’energia tèr-
mica entra, surt o s’emmagatzema en un edifici. Aquests fluxos depenen principalment del disseny
arquitectònic de l’edifici, de l’aïllament tèrmic, la
massa tèrmica, les finestres, la ventilació controla-
da per portes i finestres, la infiltració no controlada i
l’ús de l’aula. Això implica que es poden modificar i
que la gestió eficient de l’aula és a les nostres
mans. 1) els índexs de qualitat de l’ambient interior
2) els índex de qualitat bioclimàtica
3) els aparells de mesura
4) la Calculadora Bioclimàtica, i
5) l’espai d'intercanvi 3 Experiències didàctiques i treballs pràctics Experiències didàctiques i treballs pràctics Ciències 23 (2012) Conclusions La Motxilla Bioclimàtica ha completat aquest
curs la seva etapa de prova. S’ha assajat en alguns
centres de secundària de les comarques de Lleida
amb una bona resposta per part de professors i
alumnes. Com a proposta didàctica ha demostrat
que permet treballar nombrosos conceptes del cur-
rículum de secundària aplicant eines i procedi-
ments científics. La Motxilla Bioclimàtica participa en l’objectiu
global d’educar en ciències, medi ambient i salut
partint d’un context quotidià i oferint-se com una ei-
na més al servei de la comunitat educativa. 4 Experiències didàctiques i treballs pràctics Bakó-Biró, Z., Kochhar N., Clements-Croome D.,
Awbi, H. i Williams M. (2007). Ventilation Rates
in Schools and Learning Performance. Procee-
dings of Clima 2007 WellBeing Indoors. Helsin-
ki. Referències Pujol, R.M. (2003). La didáctica de las ciencies en
la educación primaria. Ed. Síntesis. Madrid. Pp. 349. Pujol, R.M. (2003). La didáctica de las ciencies en
la educación primaria. Ed. Síntesis. Madrid. Pp. 349. Bakó-Biró, Z., Kochhar N., Clements-Croome D.,
Awbi, H. i Williams M. (2007). Ventilation Rates
in Schools and Learning Performance. Procee-
dings of Clima 2007 WellBeing Indoors. Helsin-
ki. Zeyer A., Dillon J. (2012). Science, Environment,
Health. Towards a reconceptualization of three
critical and inter-linked areas of education, In-
ternational Journal of Science Education, 34, 2,
327-328. 5
|
https://openalex.org/W2083114631
|
https://gupea.ub.gu.se/bitstream/2077/4693/1/gupea_2077_4693_1.pdf
|
English
| null |
Effects of VDT and paper presentation on consumption and production of information: Psychological and physiological factors
|
Computers in human behavior
| 2,005
|
public-domain
| 17,607
|
Department of Psychology, Gothenburg University, Sweden, 2007 1 Translated by J. I. Beare. Originally published in Ross, W. D. (Ed.) (1930). The works of Aristotle (vol.
3). Oxford: Clarendon Press. E. Wästlund: Experimental Studies of Human-Computer Interaction Abstract: Experimental Studies of Human-Computer Interaction: Working
memory and mental workload in complex cognition Abstract: Experimental Studies of Human-Computer Interaction: Working
memory and mental workload in complex cognition Complex cognition is readily described as cognitive tasks requiring the
coordination of multiple steps of processing or tasks exceeding short term memory
capacity. Similarly, mental workload may be described as the use and temporary
expenditure of a finite amount of information processing capacity. In the current study,
consisting of eight experiments, the mental workload of complex cognition was
manipulated through variations in the mode of presentation (Study I) with the
information being presented either printed on paper or displayed on a computer screen
as well as through variations in page layout (Study II) with the information being
presented, either using a page layout designed to fit the computer screen or on a long
page of scroll type. In Study III, the short-term memory demands of the complex
cognitive tasks themselves were explored. The aim of the experiments was to
investigate the effect of onscreen vs. paper presented materials on complex problem
solving (Study I), the effect of page layout of onscreen presented materials on mental
workload (Study II), and the configuration of short-term memory demands of complex
problem solving (Study III). The principal findings of the three studies may be summarized by the following
points: • Both Consumption- and Production of information is more effective
when information is presented on paper rather than displayed on a
computer screen (Study I). • Consumption of information generates less mental workload when the
page layout is adapted to fit the computer screen (Study II: Experiments
1 & 2). • Problem solving processes, including both Consumption and Production
of information, may be described both in terms of their reliance on either
ST-WM or LT-WM (Study III: Experiments 1, 2 & 3) and in terms of
their reliance on specific ‘slave-systems’ of the tripartite model (Study
III: Experiments 1 & 3). Taken together, Studies I and II show that the presentation of information on
screen, versus in printed form, exerts detrimental effects on human information
processing and that some of those effects may be attributed to differences in the
navigational properties of the two media. In addition, Study II demonstrated that an
adaptation of the page layout of the presented material so that it fitted its intended
media, mental workload may be alleviated. Working memory and mental workload in complex cognition Working memory and mental workload in complex cognition Erik Wästlund Department of Psychology, Gothenburg University, Sweden, 2007 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction © Erik Wästlund
Printed in Sweden
Department of Psychology
Gothenburg 2007
ISSN 1101-718X ISRN GU/PSYK/AVH--183—SE
ISBN 978-91-628-7066-9 © Erik Wästlund
Printed in Sweden
Department of Psychology
Gothenburg 2007
ISSN 1101-718X ISRN GU/PSYK/AVH--183—SE
ISBN 978-91-628-7066-9 2 2 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction For when one actually remembers,
this impression is what he contemplates,
and this is what he perceives. For when one actually remembers, ~ Aristotle1 3 Erik Wästlund, Dept. of Psychology, Karlstad University, SE-651 88, Karlstad, Sweden.
Tel: 0046 54 700 2528, Fax: 0046 54 83 91 65. Email: erik.wastlund@kau.se Tel: 0046 54 700 2528, Fax: 0046 54 83 91 65. Email: erik.wastlund@kau.se Preface This thesis consists of this summary and the following three papers, which will be
referred to using roman numerals: I:
Wastlund, E., Reinikka, H., Norlander, T., & Archer, T. (2005). Effects of VDT
and paper presentation on consumption and production of information:
Psychological and physiological factors. Computers in Human Behavior, 21, 377-
394. II:
Wastlund, E., Norlander, T., & Archer, T. (Submitted). Effect of Page Layout on
Mental Workload: A Dual-Task Experiment. II:
Wastlund, E., Norlander, T., & Archer, T. (Submitted). Effect of Page Layout on
Mental Workload: A Dual-Task Experiment. III: Wastlund, E. & Archer, T. (Submitted). Working Memory Loads derived from
Computer-based Primary- and Secondary-Tasks E. Wästlund: Experimental Studies of Human-Computer Interaction Finally, the results of Study III showed
that, in order to understand the memory demands of complex cognition, it is necessary
to include elements of both the ST- and LT-WM paradigm of Ericsson & Kintsch and
the tripartite model of Baddeley & Hitch. Keywords: Information processing, Problem-solving, Working memory, ST-WM, LT-
WM, Dual-task, Mental workload, Reading, Page layout, VDT, Convergent; Divergent. 4 4 E. Wästlund: Experimental Studies of Human-Computer Interaction 1 Introduction The science of psychology and psychological measurement may be applied to
practically all aspects of human behavior. Given the influx of computers in so many
aspects of our day-to-day lives, it is no surprise that psychological tools, most notably
those of cognitive psychology, have been used to understand the human aspect of
human-computer interaction [HCI]. The aim of HCI studies is most often one of
usability, i.e. to evaluate a user interface in order to facilitate human information
processing (Study 1). This aim, however, may be expanded into encompassing general
psychological mechanisms and thus generate knowledge, not only relating to a user
interface in particular, but also about cognition in general (Study II). Additionally, the
tools and methods of HCI may be used without any direct interest in the user interface,
but somewhat exclusively to investigate the psychology of complex cognition (Study
III). Although the scope of HCI today remains both wide and multi-faceted, in order to
fully comprehend the effects of computers on cognition, one needs to start with the most
basic aspect of computing, i.e. reading a text on a computer screen and comparing this
with its predecessor - the printed page. Acknowledgements This paper has only one author, but there would have been no paper without the
help of my main advisor Trevor Archer and many others. To all those who have helped
me (you know who you are); thank you. In particular, I wish to extend my sincere
gratitude to colleagues and friends for their encouragement and support, to students and
anonymous participants for lending me their time and lastly, and most importantly, to
Carola and Hedda for filling my life with the stuff that really matters… December 2006
Erik Wästlund 5 5 E. Wästlund: Experimental Studies of Human-Computer Interaction Contents
1
INTRODUCTION ................................................................................................................ 7
1.1
PROBLEM STATEMENTS ..................................................................................................... 7
1.2
HUMAN INFORMATION PROCESSING .............................................................................. 10
1.2.1
Consumption and Production of Information ................................................................. 10
1.2.2
Working Memory ........................................................................................................... 12
1.2.3
Complex Cognition......................................................................................................... 14
1.2.4
Mental Workload ............................................................................................................ 15
1.3
NOTES ON METHOD .......................................................................................................... 17
1.3.1
Dual-task experiments .................................................................................................... 17
1.3.2
Computerized experiments ............................................................................................. 18
2
SUMMARY OF STUDIES................................................................................................. 19
2.1
INTRODUCTION................................................................................................................. 19
2.2 STUDY I: EFFECTS OF VDT AND PAPER PRESENTATION ON CONSUMPTION AND
PRODUCTION OF INFORMATION....................................................................................... 19
2.2.1
Experiment 1................................................................................................................... 19
2.2.2
Experiment 2................................................................................................................... 20
2.3
STUDY II: THE EFFECT OF PAGE LAYOUT ON MENTAL WORKLOAD .......................... 21
2.3.1
Experiment 1................................................................................................................... 21
2.3.2
Experiment 2................................................................................................................... 23
2.4 STUDY III: WORKING MEMORY LOADS DERIVED FROM COMPUTER-BASED PRIMARY-
AND SECONDARY-TASKS.................................................................................................. 24
2.4.1
Experiment 1................................................................................................................... 24
2.4.2
Experiment 2................................................................................................................... 26
2.4.3
Experiment 3................................................................................................................... 27
2.4.4
Experiment 4................................................................................................................... 28
3
DISCUSSION...................................................................................................................... 29
3.1
INTRODUCTION TO CONCLUSIONS................................................................................... 29
3.2
FINAL CONCLUSION.......................................................................................................... 33
4
REFERENCES.................................................................................................................... 36
5
APPENDIX.......................................................................................................................... 40
STUDY I....................................................................................................................................... 40
STUDY II ..................................................................................................................................... 40
STUDY III.................................................................................................................................... 40
Figure 1: A simple general model of Human Information Processing........................... 10
Figure 2: The working memory model proposed by Baddeley & Hitch........................ 13
Figure 3: The relationship between amount of work and mental workload................... 16
Figure 4: An elaborated model of Human Information Processing................................ 34
Table 1: Summary of input and output variables. …………………………………….. 9 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction avoid the confounding variables of page layout, e.g. line length, fonts, and kerning i.e. letter spacing, the materials in Study I were presented by means of a “pdf file”, either
printed out on paper or displayed on a computer screen. As information processing may
be divided into consumption and production of information, the investigation consists of
two separate experiments, each investigating one type of information processing. Among those that advocate that presentation on paper is superior, two
explanations can be found in the literature, focusing on either (i) the quality of the
equipment or (ii) the individual’s handling of the equipment. The former refers to the
inferior quality of material presented on-screen, due to which the acquisition of
information proceeds more slowly compared to paper presentation (e.g. Belmore, 1985;
Gray, 1991). Under circumstances of a strictly limited schedule to complete the
assignment, it is not surprising that conditions leading to greater time consumption in
task completion will induce performance deterioration in a test of maximal
performance. Nevertheless, the degree of consensus regarding the requirement for
longer on-screen reading time, as opposed to printed material, is not quite convincing
(e.g. Rice, 1994). The other school of thought focuses on the additional activity of
handling the computer. The original or basic assignment of consumption of information
is thus combined with the task of coping with computer operating requirements (Waern,
1989), resulting in a dual-task situation that may explain the observed performance
deterioration (Jolicoeur & Dell, 1999; Koch & Prinz, 2002). Thus, the detrimental
effects of on-screen presentation may be explained either in terms of reading speed or
cognitive load. The latter may originate from additional cognitive resources being
allocated to the reading process in order to compensate for the inferior presentation
quality, or may be due to the dictates of the dual-task of reading and handling the
computer. In Study II, the focus was shifted from comparing modes of presentation to
measuring mental workload associated with different types of page layout. Whereas one
simply turns the page to move forward or backward through a printed document,
navigation through an on-screen document entails using either a pointing device, for
scrolling, or the arrow keys on the keyboard for paging (Piolat, Roussey, & Thunin,
1997). 2 A4 (210*297 mm.) is the international standard size for printing paper. US Letter (215.9*278.4 mm.),
by comparison, is slightly wider and slightly shorter. 1.1 Problem statements As computer usage increases, so too does the need for research into the effects of
this development. Since computers are used mainly for information processing, in the
professional setting at least, one critical important question is; how does computer usage
influence human information processing? One obvious starting point for such a question
is the comparison of information processing on a computer with its predecessor, i.e. the
written paper (Study I). As early as 1985, Belmore warned that text presentation on video display
terminals [VDTs] was of a lower quality than text presentation on paper, leading to a
poorer understanding of the material (Belmore, 1985). Furthermore, in a comparison
between information searches on either paper or VDTs, the latter was shown to lead to a
slower reading speed (Gray, 1991). Nine years after Belmores’ study, Rice (1994) found
no difference between reading from VDTs and paper with regard to text remembrance,
implying that the mode of text presentation was not critical.Additionally, he found no
difference regarding the level of understanding between text presented on paper and on
VDTs. Nevertheless, reviews concerned with comparisons between text reading on
VDTs and on paper maintained that paper text is superior (e.g. Ziefle, 1998). Explanations for the observed detrimental effects range from the quality of the
equipment used in the studies, most notably the technical characteristics of cathode-ray-
tube [CRT] monitors (Noyes & Garland, 2003) and line length limitations (Kolers,
Duchnicky, & Ferguson, 1981), to the suggestion that reading from computer screens
reduces working memory capacity (Mayes, Sims, & Koonce, 2001). Although a few studies investigating this question are in existence, their results
appeared mixed in their message. Also, points regarding the importance, and lack, of the
matching of materials between the two modes of presentation have been raised. To 7 7 E. Wästlund: Experimental Studies of Human-Computer Interaction effects of page movement. The problem is accentuated on the World Wide Web where
long texts are often presented without any page-resembling background graphics at all,
displaying the text in a hi-tech version of ancient text scrolls. effects of page movement. The problem is accentuated on the World Wide Web where
long texts are often presented without any page-resembling background graphics at all,
displaying the text in a hi-tech version of ancient text scrolls. The purpose of Study III was to examine the mnemonic demands of complex
problem solving. This shift from ‘procedure-to-process’ is motivated by an observation
in Study I and a lack of certain predicted effects in Study II. In comparing the results
of the two experiments in Study I, participants seem to have rated the mental workload
of the two experiments differently. Although cross-experimental comparisons are
problematic at best, the result is at least noteworthy. However, as the tasks in the two
experiments of Study I differ in many respects, in Study III a set of tasks not quite so
dissimilar were chosen for comparisons. Additionally, as the results of Study II showed
no effect on the primary tasks of the secondary tasks, an additional aim of Study III
was to evaluate the soundness of the novel secondary tasks used in Study II. The eight experiments in Studies I, II and III may also be summarized in terms
of the variations in information input, process demands their measurable outputs. Input
was varied in terms of the mode of presentation (paper vs. screen) in Study I, in terms
of page layout (screen vs. scroll) in Study II, and in terms of task configuration in
Study III. In addition, the two experiments in Paper 1 differed in terms of primary
assignment (consumption of information in Experiment 1 and production of information
in experiment 2) while the two experiments in Study II differed in terms of secondary
task working memory load (phonological loop in Experiment I and visuo-spatial
sketchpad in experiment 2). The effect of these variations in input was then measured in
terms of the output of the cognitive processes. Thus, consumption of information was
measured in terms of correct responses while production of information was measured
in terms of the number of responses (fluency). E. Wästlund: Experimental Studies of Human-Computer Interaction It is this difference in the navigational method that forms the basis of the
proposition that the handling of the computer in conjecture with the original assignment,
e.g. reading results in a dual-task situation, in turn resulting in higher mental workload. However, in addition to the activity of navigating itself, there is also the question of
evaluating the effect of the procedure. Using printed material, this is simple: turn a page
– end of story. On-screen documents, on the other hand, present no such clear cues
regarding action completion. Viewing a document on-screen gives the impression that
the document is something tangible, however a document like this is really just a text
superimposed on a page resembling background graphic, and thus the dimensions of the
page are, in theory, arbitrary. In practice, the page size of on-screen documents
complies with printing standards such as A4 or US letter2. As most computer screens
are unable to satisfactorily display a readable version of such a format, navigation has to
be performed in steps, displaying different parts of the page. Thus, navigating through
an on-screen document, in addition to the activity in question, also entails assessing the In Study II, the focus was shifted from comparing modes of presentation to
measuring mental workload associated with different types of page layout. Whereas one
simply turns the page to move forward or backward through a printed document,
navigation through an on-screen document entails using either a pointing device, for
scrolling, or the arrow keys on the keyboard for paging (Piolat, Roussey, & Thunin,
1997). It is this difference in the navigational method that forms the basis of the
proposition that the handling of the computer in conjecture with the original assignment,
e.g. reading results in a dual-task situation, in turn resulting in higher mental workload. However, in addition to the activity of navigating itself, there is also the question of
evaluating the effect of the procedure. Using printed material, this is simple: turn a page
d
f
O
d
h
h
h
d
h
l 8 8 Figure 1: A simple general model of Human Information Processing. In terms of the described model, Study I and Study II both examine the effects of
variations in input whereas Study III is focused on the cognitive processes taking place
between in- and output. More explicitly, Study I compares input presented on either a
computer screen or in printed form. In Study II, all input is presented onscreen but
instead varied in terms of graphical layout which lead to different interactive demands
on the user. In Study III, focus is shifted from input to the memory demands of the
cognitive processes which underlie all human information processing. E. Wästlund: Experimental Studies of Human-Computer Interaction Mental workload was measured in two
ways, by means of subjective rating scales (Studies I and II), and by means of reaction
times (Study II only). The input and output variables are summarized in Table 1. Input
Process
Output
Study: Exp
Comparisons
Memory
Load
Assignment
Primary
Task
Secondary
Task
Subjective
Ratings
I:1
P – S
Na
Cons. Cr
Na
MWL
I:2
P – S
Na
Prod. Fl
Na
MWL
II:1
Sn – Sl
Phono
Cons. Cr
RT
MWL
II:2
Sn – Sl
Visuo
Cons. Cr
RT
MWL
III:1
RC – Syl
Phono
Cons. Cr
RT
---
III:2
RC – Syl
Visuo
Cons. Cr
RT
---
III:3
He – Sni
Phono
Prod. Fl
RT
---
III:4
He – Sni
Visuo
Prod. Fl
RT
---
Table 1: Summary of input and output variables. P-S = Paper – Screen, Sn – Sl = Screen – Scroll, He –
Sni = Headlines – Snippets, Na = Not applicable, Phono = Phonological loop, Visuo = Visuo-spatial
sketchpad, Prod = Production of information, Cons = Consumption of information, Cr = Correct
responses, Fl = Fluency, RT = Reaction Time, MWL = Mental workload. Table 1: Summary of input and output variables. P-S = Paper – Screen, Sn – Sl = Screen – Scroll, He –
Sni = Headlines – Snippets, Na = Not applicable, Phono = Phonological loop, Visuo = Visuo-spatial
sketchpad, Prod = Production of information, Cons = Consumption of information, Cr = Correct
responses, Fl = Fluency, RT = Reaction Time, MWL = Mental workload. 9 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction 1.2 Human Information Processing In order to understand the process of working on a computer, one has to begin
with basic Human Information Processing [HIP]. The most basic model of HIP consists
of some type of informational input received by an individual who processes this
information by means of cognitive operations and then delivers some type of output. Cognition
Input
Output Figure 1: A simple general model of Human Information Processing. E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction Amabile (1996) argues that a product is creative to the extent that suitable observers,
independently of each other, judge the product to be creative. Suitable observers must
possess expertise within the enterprise toward which the product is aimed. Thus,
creativity may be seen as the quality of a product that is judged creative by competent
judges. Although this may sound like a straight forward process, it is important to bear
in mind that sets of judges may not agree. Convergent production results in the synthesis of a correct solution to a problem,
the right and unique answer (Guilford, 1967). Convergent production is based upon
logic and deduction, two processes that are also studied within the context of reasoning
(Markman, 2001). Although searching within divergent problem-solving is much
broader than for convergent, the latter offers clearer criteria for goal-achievement
(Guilford, 1967). Consumption of information presumes that individuals draw a specific
conclusion regarding certain materials whereas production of information provides
scope for any number of solutions. Consequently, the consumption and production of
information may be described and measured in terms of convergent and divergent
production. The most prominent example of consumption of information, regardless of media,
is of course reading, and thus consumption of information may be measured in terms of
reading comprehension. It is, however, important that questions be constructed in such a
manner that they do not solely tap memory but that they also require both an
understanding of and deduction from the material. Production of information, on the
other hand, is best described by the writing of a text. That said, it should be remembered
that all writing, except possibly experimental Dadaistic poetry, also contains elements
of consumption of information since the written text has to be evaluated if the plot is to
evolve in a coherent manner. In the same way, assigning a title to a text document or
assigning a subject line to an e-mail message entails the consumption of information,
i.e. the contents of the text, in order to produce a legible a subject or title. Measuring
production of information can thus be done in the form of writing assignments. In order
to construct meaningful tests, of both consumption and production of information, it is,
as always when constructing tests, important that the assignments bear resemblance to
tasks that are familiar to the testee in question. 1.2.1 Consumption and Production of Information From an HCI perspective, information may be seen as arising from, and returning
to, the computer. From a psychological perspective, input can come from any sensory
modality or from previously memorized information. In the same manner, output can be
any type of action, both psychological and physiological. The processes in-between
input and output, may, to a large extent, be divided into two categories: Consumption of
information and Production of information. Both types of work (performance) are based
upon cognition, briefly described as the process through which information is encoded,
organized, stored, remembered, and applied/recalled (Martinsen & Kaufmann, 1999). Both the consumption and production of information entail the formation of new
cognitive structures and novel combinations of ideas or thoughts. Ghiselin (1963)
identifies both these consequences of cognitive processes as psychological aspects of
creativity. The concept of creativity is complex and multi-faceted yet a fundamental
distinction has been made between input-creativity, the ability or process of interpreting
information and forming abstract connections between concepts, and output-creativity,
the ability or process of generating new or cultivated material (Partridge & Rowe,
1994). The input versus output-creativity distinction may be traced to Guilford (1967)
who separates two forms of creative production based on the possible variations in the
number of correct solutions. Divergent production is the production of one or more
solutions, or attempts at explanations, which can be seen as samples from an infinite
number of possible solutions. Since the only limitation arises from the individual’s
imagination, or evaluation and subsequent discarding of ideas, the possibilities are, in
fact, limitless. The absence of restrictions leads to difficulties in the qualitative evaluation of
divergent production, a problem apparent in all discussions regarding art and music. 10 E. Wästlund: Experimental Studies of Human-Computer Interaction Thus, it is pointless to ask your average
participant to write a novel or read and understand a postgraduate text on quantum
physics. This short description of the measurement of consumption and production of
information points to two important methodological issues; one of assessment and one
of generalizability. In order to successfully assess the ability of an individual, with some
degree of precision, the task should include some items that the participant will pass and
some that he or she will fail. When testing a group of participants, this principle must
apply to all participants so optimally that all participants should pass some items and no
participants should pass all of them. The question of generalizability pertains to
ecological validity, i.e. if the behaviors observed in the studies reflect behaviors outside
the controlled environment of the data collection situation. Ecological validity can be
enhanced by using tasks that mimic tasks which the participants are accustomed to, and
by presenting the tasks in a familiar setting. Regarding the simple HIP model, what can be said about the production and
consumption of information while working on a computer? First, the input arrow can be 11 E. Wästlund: Experimental Studies of Human-Computer Interaction seen as symbolizing information presented on the computer screen. This seems obvious
in relation to reading but, as shown, it is also important in relation to writing. The output
arrow, on the other hand, may be interpreted as symbolizing information entered into
the computer, either in the form of a novel text or by answering questions in a multiple-
choice test. However, as consumption of information only rarely leads to some type of
test, but instead is assumed to result in information being stored in memory and thus the
output arrow could equally be seen as symbolizing the formation of memories. In the
same manner, while production of information may be based on information presented
on-screen, it can equally draw on information stored in memory. Last, but not least,
both processes entail manipulation of information, either in order to organize
meaningful memories or in order to produce novel ideas. Although the classification of production processes makes it possible to
discriminate between types of tasks, we need to look into the study of memory in order
to understand how cognitive tasks are actually performed. E. Wästlund: Experimental Studies of Human-Computer Interaction As previously stated, the temporal qualities of STM have been known since James
(1890) described the introspective studies of Ebbinghause. However, it was the findings
of Miller (1956), i.e. that STM could contain 7 ± 2 pieces of information, and of the
Petersons (1959), i.e. that STM-stored information was rapidly forgotten if people were
distracted or otherwise preoccupied, that led to the popularization of the topic during the
1960s as well as to the inclusion of information processing as a part of STM. Today, information processing, both in terms of on-line processing and temporary
storage, is generally described and explored in terms of working memory (Duff &
Logie, 2001). Most often, working memory is viewed as a multiple resource system
containing several specialized subsystems. This notion is based on the work of
Baddeley & Hitch (1974) who, over thirty years ago, proposed a model containing
separate temporary storage systems for phonological information (the phonological
loop) and visuo-spatial information (the visuo-spatial sketch pad). Activity in these
storage systems is directed by the central executive which is not only seen as
responsible for directing attention but also as responsible for on-line processing
(Baddeley, 1996). This model has been applied successfully to a variety of cognitive
activities such as language acquisition, syllogistic reasoning, mental arithmetic, and
complex perceptual-motor skills (Duff & Logie, 2001). Figure 2: The working memory model proposed by Baddeley & Hitch (1974). Central executive
Visuo-spatial
sketchpad
Phonological
loop Figure 2: The working memory model proposed by Baddeley & Hitch (1974). Despite, or perhaps because of, the success of the model, several alternative
working memory models have emerged (see Miyake & Shah (1999) for a review). A
specific point of divergence pertains to how the very narrow span of the working
memory can account for the large processing demands made by complex problem
solving. One plausible solution was the notion proposed by Ericsson & Kintsch (1995)
who argued that in order to explain the working memory capacity during skilled
performance, it is necessary to divide the working memory entity into a short-term
working memory [ST-WM] and a long-term working memory [LT-WM]. During
skilled cognitive activities, the ST-WM is supported by information processing
strategies in the LT-WM and thus the ST-WM’s capacity for the activity may surpass
the usual constraints attributed to the working memory (Ericsson & Lehmann, 1996). 1.2.2 Working Memory The history of memory begins with Simonides of Ceos (556 – 468 B.C.) who is
attributed with the invention of visual memory aids (Yates, 1966). The aim of the
technique advocated by Simonides was to memorize poems for later recitation. Although there are no records of Simonides’ thoughts on the matter, the notion that
memory is a function of separate subsystems was introduced by William James (1890),
who distinguished primary from secondary memory, terms equivalent to the
contemporary concepts of short-term memory [STM] and long-term memory [LTM]. The main difference between STM and LTM lies in their temporal qualities. Speaking
with James, whose main focus was attention, the difference was one of past versus
present. Today, the difference is still viewed as temporal but the STM is also seen as
exerting much responsibility for the processing of information. In its turn, the LTM is seen as consisting of several subsystems based on domains
of memory. The lack of consensus regarding the classification of memory domains has
resulted in the production of several LTM models (e.g. Squire, 1992; Tulving, 1999),
which, despite their differences, display more commonalities. One basic distinction is
made between the explicit or declarative and the implicit or non-declarative memory
classes. Implicit / explicit refers to the degree of consciousness of access whereas
declarative/non-declarative refers to the possibilities of communicating memory content
to others. The explicit / declarative memories are divided into two classes based on
personal involvement. Memories that are based on personal experience are labeled
episodic memories whereas memories of facts are, in the term of Tulving (1999),
labeled semantic memories. An important difference between the two concepts lies in
the experiencing of the two. Remembering an auto-biographical experience, involves to
some extent a ‘time-lapse’ back to the original incident, whereas remembering a fact
does not involve any experience of the situation when the encoding took place. The
implicit / non-declarative memories may, according to the view of Tulving, also be
divided into two subclasses, procedural memories, which are memories for perceptual
motor skills, and priming which is an association between stimuli and actions or
thought. 12 1.2.3 Complex Cognition Although there appears to be a paucity of agreement on what constitutes complex
cognition (Kintsch, Healy, Hegarty, Pennington, & Salthouse, 1999), tasks exceeding
typical working memory limitations (e.g. reading comprehension) and those requiring
the coordination of multiple steps of processing are readily described as complex (e.g. syllogistic reasoning). The end-point of reading comprehension is to generate long-term
memory structures containing integrated representations of the text. In short, the process
leading up to this end-point may be described as the phonological loop of working
memory being fed with single words, which are combined into phrases, sentences and
so on. These phrases and sentences are then integrated with previously stored
information, stemming from both the text and the long-term memory (Van Merriënboer
& Sweller, 2005). Individual differences in reading comprehension have been explained
as differences in working memory capacities (e.g. Daneman & Carpenter, 1980;
Daneman & Merikle, 1996) but also in terms of more sophisticated and complex
comprehension strategies loaded into the LT-WM (Ericsson & Kintsch, 1995; Kintsch,
Patel, & Ericsson, 1999). In contrast, syllogistic reasoning is based on dual statements
(e.g. Peter is taller than Paul and Paul is shorter than Mary) and the object is to
determine which conclusions may be drawn from these statements. This objective is
usually achieved using a third statement (e.g. Peter is taller than Mary) and participants
are asked to indicate weather or not the final statement is true, false or impossible to
deduce. Solving syllogistic tasks hardly leads to any long term memories of the
statements or deductions. The key role of working memory in the solution of syllogisms
has been demonstrated, by for instance, Gilhooly, Logie, Wetherick, & Wynn (1993)
who concluded that there is a slight involvement of the phonological loop and a major
involvement of the central executive. Reading comprehension and syllogistic reasoning are thus similar insofar as they
may both be shown to rely on working memory for processing, but differ in terms of
LT-WM and long-term memory involvement. Additionally, they are both tasks wherein
the object is to process information in order to form a valid conclusion, to which there is
no alternative, which in Guilford’s (1967) terms appears to constitute convergent
production. Divergent production, on the other hand, involves tasks which are
characterised by the lack of a single solution (Guilford, 1967). E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction activated for both short time retention and as the central processing structure in the
service of complex cognition. Despite the similarities of the two models and their
proven merits, research into working memory seldom includes both of the models. In,
Paper III, a more or less eclectic view of working memory will be adopted which
combines the two aforementioned models. Thus, working memory will be seen as being
comprised of the long and short term systems of Ericsson and Kintsch in which the
‘slave-systems’ of Baddeley and Hitch reside. The reason for combining and including
both models, in Paper III, was to try to investigate complex cognition from a broader
perspective than strict adherence to a specific model would allow. E. Wästlund: Experimental Studies of Human-Computer Interaction Although the models appear quite different at first glance, the ST-WM of Ericsson
and Kintsch may be seen as containing the same representations of information as the
‘slave-systems’ of Baddeley and Hitch and whereas the latter postulates the
manipulation of information as stemming from the central executive, Ericsson and
Kintsch attribute this function to information processing schemata activated in the LT-
WM. Irrespective of model, at the heart of the matter lies a short term memory system 13 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction vocalised in the phonological loop. In terms of working memory demands, it has been
shown that narrative texts imply more imagery and thus rely more heavily on the visuo-
spatial sketchpad than do expository texts which place greater demands on the
phonological loop (for a review see Olive, 2003). Additionally, as the composition of a
long text entails the elaboration of information from a given starting point, it can be
assumed that LT-WM processes are needed in order to form coherent texts longer than a
single sentence. However, text production does not necessary entail elaboration. During
‘note-taking’, for instance, the object is condensation rather than elaboration. Note-
taking refers to the production of short paragraphs of text containing, for instance, a
shopping list or the minutes of a work meeting (Piolat, Olive, & Kellogg, 2005). It has
been stipulated that the production of notes places high demands on the central
executive functions as it concurrently involves comprehension, selection, and
production processes. Additionally, it has been shown that ‘note-taking’ requires less
executive mobilization than writing and that when note-taking is based on a text, the
length of that text is positively correlated with cognitive effort (Piolat et al., 2005). One
possible explanation for these findings could be that the amount of information
concurrently being processed in the LT-WM is what defines the cognitive demands. y
g p
g
Independent of memory demands, information processing in a human-computer
interaction setting, not only entails task completion but also the activity of controlling
and inputting data into the computer, an activity which, in itself, draws on information
processing capabilities. Basically, this handling consists of using either the keyboard or
a pointing device such as a computer mouse or a touchpad. The type of memory process
in use during a particular activity is determined by the user’s experience of the activity
in question and the configuration of the activity itself. Given that an activity entails few
or no options and that the user is well-experienced in the activity, as is the case with a
professional typist at work, motor-control (i.e. implicit memory) will be based more or
less solely on procedural memory (i.e. ‘how?’) and will entail only small or no amounts
of conscious processing. E. Wästlund: Experimental Studies of Human-Computer Interaction On the other hand, if an activity is new to the user and includes
a series of options, as is often the case with new releases of operating systems, both
motor-control and the choices will demand conscious processing (i.e. the involvement
of explicit memory) and, thus, working memory resources. Additionally, there are
activities during which motor-control might be automated but still reliant on conscious
processing for the determination of task completion, e.g. scrolling through a long page
of text. 1.2.3 Complex Cognition In a similar fashion, just as convergent production relies on various cognitive
processes, so too does divergent production. The writing of a long text involves a large
number of cognitive processes. Information has to be retrieved from the long-term
memory, reorganised via collaboration between LT-WM and ST-WM and finally be 14 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction degrees of dual-task interference (Pashler, 1998). Today, the ‘multiple resources’ notion
predominates (Xie & Salvendy, 2000). In addition, within the framework of the
Cognitive Load Theory, the phrase cognitive load is used with the focus not being on
attention per se but on working memory (Paas, Renkl, & Sweller, 2003). degrees of dual-task interference (Pashler, 1998). Today, the ‘multiple resources’ notion
predominates (Xie & Salvendy, 2000). In addition, within the framework of the
Cognitive Load Theory, the phrase cognitive load is used with the focus not being on
attention per se but on working memory (Paas, Renkl, & Sweller, 2003). To complicate the situation further, the depletion of information processing
capacity may stem from both under and overload of stimuli (see Figure 3, for a
graphical representation of the relationship between stimuli and workload) Optimally, the amount of stimuli should fall within the range of ideal stimulation, i.e. the amount of stimuli should be adapted to an individual’s processing capacity. Such a
situation leads to a low mental workload. As the amount of stimuli increases,
unsurprisingly, so too does the mental workload. The same effect occurs, however,
when the amount of stimuli decreases. Figure 3: The relationship between amount of work and mental workload. Adapted from (Kumashiro,
1995). decreasing stimuli increasing
w
o
r
k
l
o
a
d
range of ideal
stimulation Figure 3: The relationship between amount of work and mental workload. Adapted from (Kumashiro,
1995). Although decreases and increases in stimuli both lead to the same result, i.e. an
increase in mental workload, they vary in the degree to which they draw upon the
available resources. Whereas underload usually stems from stimuli that place high
demands on attention but low demands on processing, e.g. monitoring for low
occurrences of stimuli, overload can be the result of high demands on processing
capacity in addition to high demands on attention. Examples of both under and overload
can be found in the world of aviation (Wickens, Mavor, & McGee, 1997). Monitoring a
radar screen for enemy aircrafts that may or may not appear, provides an example of a
situation that places high demands on attention but low demands on processing
capacity, thereby resulting in a high workload. 1.2.4 Mental Workload The concept of workload is more readily understood when used in a physiological
context. Within the psychological tradition, however, the concept is rather ill-defined
and applied in a variety of ways. In the context of Kahnemans’ (1973) limited-capacity-
model of attention, the concept of “mental workload” has been used while the term
“cognitive workload” is associated with Wickens’ (1984) multiple-resource-model of
attention. Both models postulate a limited amount of attention that can be utilized for
information processing and that, when demands exceed the available resources,
processing will be impaired. The particular difference of interest between the two
models lies in their views of attention, as based on single or on multiple resources. Support for the latter is derived from dual-task research showing that tasks similar in
terms of input or output modality, or in terms of choice or judgment, lead to higher 15 15 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction workload and cognitive load as the latter two are to a greater extent associated with
specific theories. Although Cognitive Load Theory places the source of attention within
the working memory, as opposed to outside the working memory as is the case with
both the attention-based models, the focus of Cognitive Load Theory is not in line with
the current investigation. As the aim of the study is to investigate the effects of
continuous information processing, as opposed to vigilance tasks, the depletion of
processing capacity is assumed to stem from overload rather than underload. The measurement of mental workload can be divided into three major categories
(Wierwille & Eggemeier, 1993): (i) subjective measures, e.g. the NASA task load index
[TLX] (Hart & Staveland, 1988) and The Subjective Workload Assessment Technique
[SWAT] (Reid & Nygren, 1988) where respondents rate their subjective experience of a
task on a series of scales; (ii) measures of physiological correlates, e.g. heart rate
variability (e.g. S. Miyake, 2001) or eye activity (e.g. Van Orden, Limbert, Makeig, &
Jung, 2001); (iii) and a wide variety of performance measures like error rates (e.g. van
der Linden, Frese, & Meijman, 2003) or task completion time (e.g. Waters & Caplan,
2004). A special case of performance measures is found within the dual-task-paradigm
where participants are subjected to two concurrent tasks. Previous research has shown
that reaction time [RT] data offers a suitable secondary task measurement of mental
workload (Verwey & Veltman, 1996; Wickens, 1984). E. Wästlund: Experimental Studies of Human-Computer Interaction Monitoring an air traffic control radar
screen at an international airport, providing pilots with in-flight instructions, places an
equal amount of strain on attention as well as high demands on processing capacity
(Wickens et al., 1997). In the present study, the term mental workload will be used in its most generic
sense, i.e. referring to the use and temporary depletion of a finite amount of information
processing capacity. The term mental workload is chosen in preference to cognitive 16 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction the reaction times to the secondary task ought to increase accordingly. The dual-task
experiment may also be designed in reverse, i.e. if the mental workload of the
secondary task is increased, performance on the primary task will decrease. The
influence of the secondary task on the primary task has raised some concern regarding
the techniques’ usefulness (Paas, Tuovinen, Tabbers, & Van Gerven, 2003). However,
if both primary and secondary task workloads are varied, the cognitive nature of the
tasks themselves may be analyzed. In accordance with the Additive-Factor Method
(Sternberg, 2001), interactions between the two tasks indicate that they are based on the
same cognitive resource, whereas if the reaction times are additive, the tasks will utilize
different mental resources. Although the secondary tasks described above have been shown to be effective,
they are rather impractical since participants need to be constantly monitored by the
researcher as well as in solitude in order not to disturb each other. Thus, in Studies II &
III, alternative secondary tasks are used. 1.3.1 Dual-task experiments A special case of performance measures is found within the dual-task-paradigm
where participants are subjected to two concurrent tasks. Depending on the set-up of the
experiment, different types of inferences may be made from the results. Within working
memory research an often used technique is to investigate the memory demands of a
certain primary task by testing the effects of various continuous secondary tasks
previously shown to draw on a certain working memory resource. The result of two
tasks taping the same resources is a performance deterioration on the performance of the
primary- and/or secondary task. Frequently used secondary tasks include the random
generation task whereby the participant generates random sequences of numbers or
letters (the central executive), the spatial tapping task whereby the participant is
instructed to tap sequentially the four corners of a square using a finger (the visuo-
spatial sketchpad), and articulatory suppression whereby the participants are instructed
to repeat syllables, words, or phrases aloud (the phonological loop) (A. Miyake & Shah,
1999). An alternative to continuous secondary task is the application of probing tasks,
whereby the participant is instructed to engage in the primary task, with intermittent
interruptions by a secondary task probe. Previous research has shown that reaction time
[RT] data to probing secondary tasks is a suitable measurement of mental workload
(Verwey & Veltman, 1996; Wickens, 1984). The rationale behind using RTs as a
measure of mental workload is that the amount of capacity demanded by the primary
task will determine the reaction time to the occurrence of the secondary task (Brunken,
Plass, & Leutner, 2003). Thus, by increasing the mental workload of the primary task, 17 E. Wästlund: Experimental Studies of Human-Computer Interaction The ELMA CyberLab3 web application, utilized in Study II and Study III, is a
combination of server and client-side applications. When the participant clicks on the
“start experiment” link, a small program is downloaded and run on his or her computer. The program then contacts the server-side database for group allocation and, based on
this information, the participant is then shown a specific sequence of instructions and
assignments. Throughout the experiment, ELMA CyberLab collects input from the user,
in terms of both responses to questions/tasks and reaction times to various stimuli. After
the experiment, the collected data is sent back to the server for storage in the database. The experimenter may at any time download the collected data from the database and
import it into suitable statistical software for analysis. Design A 2*2 factorial design was used with Type of Presentation (Paper or Computer)
and Gender as between-group variables. Dependent variables were consumption of
information, measured in terms of correct responses on a reading comprehension test
and ratings of subjective experience. 3For the technically-minded: ELMA CyberLab consists of a series of Macromedia Flash files that
communicate with a MySQL database via php scripts. ELMA CyberLab has been designed by the auth 2.1 Introduction In the following three papers, a total of 374 participants took part in eight
experiments. Study I describes two experiments, each with 72 participants, concerning
the effects of mode of presentation (paper vs. computer screen). Study II describes two
experiments, with final samples consisting of 42 and 40 participants, respectively,
concerning the effect of page layout. Study III describes four experiments regarding the
memory demands of complex cognition with final samples consisting of 40, 32, 44, 32
participants, respectively. 1.3.2 Computerized experiments As noted above, computers and computer applications are being used
increasingly, this also being the case within the scientific community. Within the
various disciplines of the behavioural sciences, computers are not only applied as tools
for text processing and computing data, but also as a means of conducting experiments. The benefits of using computers in experimentation include controlling information, i.e. treating all subjects equally and the possibility of precise timing, both in terms of
presenting materials as well as for the timing of responses. In addition, by utilizing web
technology, computerized experiments allow the simultaneous testing of a theoretically
infinite group of participants. It is important to note that utilizing web technology does
not imply that an experiment will be made public via the Internet and thus open to
unrestricted participation. Through distributed web experimenting, researchers conduct
web-based experiments using participants recruited and assisted by local collaborators
(Reips, 2002). At the outset of the experiment, the participants or collaborators log on to
the same password-protected Internet-based experiment, thus access to the material is
strictly regulated. Such an experiment can easily be transformed into an ‘open-to-all’
web experiment by removing the password protection and advertising the existence of
the project on relevant websites. Taking experiments into the public domain via the
Internet poses special problems, e.g. self-selection and diminished commitment to
fulfilling the experiment. However, there are also benefits, e.g. the possibility to reach
large numbers of participants of varying demographic backgrounds. Web-based data collection utilities can be divided into server-side and client-side
utilities (Reips, 2002). Server-side utilities include a web server for the storage and
distribution of the experimental material as well as database applications for data
collection and the distribution of dynamic content. Conducting an experiment solely on
the server side, e.g. by means of a webpage, may lead to experimental errors in time
measurement, due to lags in Internet traffic and visual discrepancies between various
computer platforms. The solution to this is using client-side utilities, i.e. materials that
are downloaded from the server and run on the participants’ computers. 18 General aim The following two experiments were designed to measure the effects of mode of
presentation (paper vs. computer screen) on consumption (Experiment I) and production
of information (Experiment II). In addition, the participants rated their subjective
experience of the assignments. b) Dependent variables: i. The Higher Education Entrance Examination READ test. The READ test
is designed to measure Swedish language reading comprehension. The
test contains five texts (averaging 1,000 words) each followed by four
multiple choice questions. Participants where given thirty minutes to
complete this test. ii. PostSTH – Stress –Tiredness – Hunger. Three questions measuring how
tired, stressed and hungry subjects felt. Subjects were required to mark,
with a cross, the point on a VAS scale that they considered most
appropriate. Procedure The experiment was conducted over a two-day period. On the first day, the text
assignments were distributed via computer screen and on the second day, via paper. Since the participants were allowed to choose which day and time they would
participate, they also, unwittingly, chose which assignment they would take part in. Upon arrival, the participants were instructed to turn off their cellphones and “take a
seat”. In front of them, they found an envelope containing the test materials and
instructions relevant to all tests. The only difference between days one and two was
that, during day one, the text assignments were presented on-screen by means of a pdf
document, with the same document being presented in printed form during day two. E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction ii. TRI – Technology Readiness Index. The TRI test (Parasuraman, 2000) is
a multiple item scale design to measure the individual’s propensity to
embrace new technology. iii. SE – Stress and Energy. The SE instrument allows individuals to rate
their subjective experience of energy and stress (Kjellberg & Iwanowski,
1989). iv. PANAS – Positive- and negative affect scale. The PANAS instrument
(Watson, Clark, & Tellegen, 1988) measures the individual’s degree of
positive and negative affectivity through ratings using twenty adjectives. v. PreSTH – Stress –Tiredness – Hunger. Three questions measuring how
tired, stressed, and hungry subjects felt. Subjects were required to mark,
with a cross, the point on a VAS scale that they considered most
appropriate. b) Dependent variables: Instruments a) Pre-experimental questionnaires to ensure group homogeneity: i. Demographics. A simple demographics survey containing questions
regarding gender, age, university credits, and participants’ computer
usage. 19 Results The results of the data analysis showed that Type of Presentation had significant
effects on both the READ test, where participants in the Paper assignment scored
higher, and on the Post STH test, where participants in the Computer assignment
reported higher ratings on the stress and tired scales and lower ratings on the hunger
scale. Procedure (See Experiment 1). (See Experiment 1). Results The results of the data analysis showed that Type of Presentation had a significant
effect on the Headlines test whereby participants in the Paper assignment produced
more headlines but that there were no differences in the quality of those headlines. Neither were there any significant effects on the Post STH test although there was a
trend indicating that the participants in the Computer assignment reported higher ratings
of stress. Instruments Instruments
a) Pre-experimental questionnaires to ensure group homogeneity: i. Demographics. (See Experiment 1). i. Demographics. (See Experiment 1). ii. TRI – Technology Readiness Index. (See Experiment 1). iii
SE
Stress and Energy (See Experiment 1) g p
(
p
)
ii. TRI – Technology Readiness Index. (See Experiment 1). ii. TRI – Technology Readiness Index. (See Experiment 1). iii. SE – Stress and Energy. (See Experiment 1). gy
p
iii. SE – Stress and Energy. (See Experiment 1). gy
p
iv. PANAS – Positive- and negative affect scale. (See Experiment 1). v. PreSTH – Stress –Tiredness – Hunger. (See Experiment 1). b) Dependent variables: Dependent variables: i. The verbal creativity test “Headlines” measures divergent production
(Ekvall, 1969). The test assesses the ability to transform one unit of
information into another, e.g. deriving from a complete newspaper article
the essential text that might constitute a headline. The test requires the
testee to write down as many headlines as possible for short newspaper
articles; in the test’s original design, this involved four newspaper
articles where the time allocated to each article was three minutes. In the
present study, this was modified to encompass 10 articles with a total
time limitation of thirty minutes to equal the assignment time of
Experiment 1. In addition to the quantitative judgements, the number of
headlines, the test also included a qualitative assessment carried out by
two journalists who subjectively rated every headline on a scale of 1 to 5. ii
PostSTH – Stress –Tiredness – Hunger (See Experiment 1) i. The verbal creativity test “Headlines” measures divergent production
(Ekvall, 1969). The test assesses the ability to transform one unit of
information into another, e.g. deriving from a complete newspaper article
the essential text that might constitute a headline. The test requires the
testee to write down as many headlines as possible for short newspaper
articles; in the test’s original design, this involved four newspaper
articles where the time allocated to each article was three minutes. In the
present study, this was modified to encompass 10 articles with a total
time limitation of thirty minutes to equal the assignment time of
Experiment 1. In addition to the quantitative judgements, the number of
headlines, the test also included a qualitative assessment carried out by
two journalists who subjectively rated every headline on a scale of 1 to 5. ii. PostSTH – Stress –Tiredness – Hunger. (See Experiment 1). E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction information, measured in terms of the number and quality of the responses during the
“Headlined test” and ratings of subjective experience. 2.3.1 Experiment 1
Design 2.2.2 Experiment 2
Design g
A 2*2 factorial design with Type of Presentation (Paper or Computer) and Gender
as between group variables was used. Dependent variables were production of 20 Instruments a) Pre-experimental instruments to ensure group homogeneity: a) Pre-experimental instruments to ensure group homogeneity: i. Demo. A simple demographics survey containing questions about
gender, age, university credits, number of times the participant had taken
the Higher Education Entrance Examination, average hours of computer
use per week, vision and vision correction. ii. PreRightNow. The first instance of the RightNow test (see dependent
variables for full description). b) Independent variables: i. Layout. To investigate the effects of page layout, the primary task text
assignments were created in two versions (Screen and Scroll). In the
Screen version, the text was divided up so that each page was the same
size as the computer screen. In the Scroll version, the text was not
divided up at all, i.e. it was presented as one long document. ii. PopType. To investigate the effects of pre-loading the phonological loop,
two types of pop ups (PopUp and PopMem) were used. In the PopUp
assignment, participants had to respond to pop ups by simply clicking on
a button labelled ‘close’. In the PopMem assignment, the participants
were first presented with a two digit target number. The following pop
up showed a number, which may or may not have been the same number. The participant was then instructed to indicate whether or not this was
the target number by clicking on a ‘yes’, ‘no’, or ‘don’t remember’
button. Having done so, a new target number was then presented and the
participant was allowed to continue reading the text. c) Dependent variables: i. Primary task. The primary tasks are text assignments forming part of the
Higher Education Entrance Examination – READ. The test is designed to
measure Swedish language reading comprehension. The original test
consists of 10 pages containing five different texts, averaging 1,000
words, which are each followed by four multiple-choice questions. In
this experiment, two of these texts were used. As the scoring is a simple
summation of the correct answers, the respondents can receive a
maximum of 4 points in each of the two assignments. p
g
ii. Secondary task. During the course of the text assignments, the
participants are interrupted by 10 pop ups covering the screen, with the
measurement of interest being the mean reaction time to these. iii. RightNow. This test is designed to measure the participants’ current state
of stress, general tiredness, and optical fatigue. E. Wästlund: Experimental Studies of Human-Computer Interaction A 2*2 mixed factorial design was used with Layout (Screen / Scroll) as within
group variable and PopType (PopUp / PopMem) as between group variable was used. To avoid possible order effects, Layout was balanced for order and text order, i.e. there
were four subgroups in both PopType groups. General aim The following two dual-task experiments were designed to measure the effects of
the page layout of computer-presented text documents and the effects of working
memory load (phonological loop in Experiment 1 and visuo-spatial sketchpad in
Experiment 2) on reading comprehension and mental workload. 2.3.1 Experiment 1
Design 2.3.1 Experiment 1
Design 21 21 E. Wästlund: Experimental Studies of Human-Computer Interaction Design Design
(See Experiment 1). (See Experiment 1). d) Post-experimental evaluation: i. Performance. This test consisted of two questions, one positively and one
negatively worded, which allowed the participants to rate, on a five point
Likert scale, their own performance during the two assignments. The
participants were asked to rate their performance during both the first
and second text assignments. Procedure The experiment was conducted in a computer room containing 16 computers at
Karlstad University. When the participants arrived, the only verbal instructions they
were given were to “take a seat”, to turn off their cellphones, and to look for all the
information they might require on screen. Once all the participants had done this, they
were then instructed to click the ‘start experiment’ button on their computer screens. When the participants started the experiment, the ELMA CyberLab randomly assigned
them to one of the two PopType groups and to one of the four Layout subgroups. Results The results showed that there were no significant effects on reading
comprehension but that the scroll page layout had led to a significantly higher mental
workload. Instruments The questions were
phrased as thus: How state in question are you feeling right now? The
participants were asked to respond to the questions using a Visual
Analogue Scale [VAS] (range 0 – 100), with two extremities and a
midpoint. The RightNow test was presented on three occasions during 22 E. Wästlund: Experimental Studies of Human-Computer Interaction the experiment: i.e. before the first text assignment, and both before and
after the second text assignment. It is thus possible to calculate the effect
of Layout (Screen / Scroll) on the three states by subtracting the value
reported post-assignment from the value reported pre-assignment. d) Post-experimental evaluation: d) Post-experimental evaluation: d) Post-experimental evaluation: i. Performance. (See Experiment 1). Procedure (See Experiment 1). Results The results replicated those of Experiment 1, showing that there were no
significant effects on reading comprehension, but that the scroll page layout led to a
significantly higher mental workload. c) Dependent variables: c) Dependent variables: i. Primary task. The primary tasks were the same as in Experiment 1,
except for the inclusion of two additional questions to each text, making
the possible total 6 points for each text. p
p
ii. Secondary task. (See Experiment 1). iii. RightNow. (See Experiment 1). General aim The following four dual-task experiments were designed to examine the memory
demands of complex cognition in regards to both consumption and production of
information. In Experiments 1 and 2, Reading comprehension is contrasted with
Syllogistic reasoning. Additionally, the role of the phonological loop is investigated in
Experiment 1 and the role of the visuo-spatial sketchpad is investigated in Experiment
2. In Experiments 3 and 4, text production is contrasted with text condensation. Also,
the role of the phonological loop is investigated in Experiment 3 and the role of the
visuo-spatial sketchpad is investigated in Experiment 4. b) Primary tasks: Instruments a) Pre-experimental instruments to ensure group homogeneity: a) Pre-experimental instruments to ensure group homogeneity: i. Demo. (See Experiment 1). ii. PreRightNow. (See Experiment 1). b) Independent variables: ii. PreRightNow. (See Experiment 1). i. Layout. (See Experiment 1). y
p
ii. PopType. To investigate the effects of visuo-spatial load, two types of
pop ups (PopSolE / PopSolD) were used. Both PopTypes are designed
like a simple game of solitaire with a stack of four cards, one from each
suit, face down located at the top of the screen. Below the cards, there
are four foundations marked with the four suits. Upon clicking on the
back of the top card, this turns over and the object is to drag this card to
the corresponding foundation. The difference between the two PopTypes
lies in the degree of difficulty in determining which foundation the card
should be moved to. In the easy PopType (PopSolE), the cards in the
stack were all aces and lay in the same order as the foundations, when
viewed from left to right. In the difficult PopType (PopSolD), both value
and suit were randomized. ii. PopType. To investigate the effects of visuo-spatial load, two types of
pop ups (PopSolE / PopSolD) were used. Both PopTypes are designed
like a simple game of solitaire with a stack of four cards, one from each
suit, face down located at the top of the screen. Below the cards, there
are four foundations marked with the four suits. Upon clicking on the
back of the top card, this turns over and the object is to drag this card to
the corresponding foundation. The difference between the two PopTypes
lies in the degree of difficulty in determining which foundation the card
should be moved to. In the easy PopType (PopSolE), the cards in the
stack were all aces and lay in the same order as the foundations, when
viewed from left to right. In the difficult PopType (PopSolD), both value
and suit were randomized. 23 E. Wästlund: Experimental Studies of Human-Computer Interaction Design g
A 2*2 mixed factorial design was used with the primary tasks as within group
variable and the secondary tasks as between group variable. To avoid possible order
effects, the primary tasks were balanced for order, i.e. there were two subgroups in both
secondary task groups. E. Wästlund: Experimental Studies of Human-Computer Interaction The objective of both the primary tasks (Reading Comprehension and Syllogistic
reasoning) was to measure information processing using multiple choice questions. The difference between the two primary tasks lies in the amount of information that
is necessary to process in order to answer the questions. i. The primary task Reading Comprehension forms part of the Higher
Education Entrance Examination – READ (a Swedish equivalent of the
SAT test). This test is designed to assess Swedish-language reading
comprehension. The original test consists of 10 pages containing five
different texts, averaging 1000 words, with Reading Comprehension
being measured via four multiple-choice questions per text. In the current
experiment, one such text is used. As the scoring is a simple summation
of the correct answers, the respondents could receive up to 4 points The
questions are devised in order to be as varied as possible and are
distributed in such a way that the whole text is used: it is thus necessary
to form an overall understanding of the material in order to answer the
questions. The text assignments are constructed in order to cover three
cognitive levels; remembering, understanding, and deduction, where
understanding and deduction predominate. g
p
ii. The primary task Syllogistic Reasoning (Holmquist, 1974) used in this
experiment contains forty syllogisms all of which include two premises
(e.g. A is smaller than B and B is smaller than C) and a question (e.g. Is
A smaller than C?). The respondent’s task was to answer “Yes”, “No” or
“Impossible to know” using multiple-choice questions. Half of the
syllogisms contain premises based on letters while the other half were
based on names (e.g. Peter is taller than Paul. Paul is taller than Mary). ii. g
p
ii. The primary task Syllogistic Reasoning (Holmquist, 1974) used in this
experiment contains forty syllogisms all of which include two premises
(e.g. A is smaller than B and B is smaller than C) and a question (e.g. Is
A smaller than C?). The respondent’s task was to answer “Yes”, “No” or
“Impossible to know” using multiple-choice questions. Half of the
syllogisms contain premises based on letters while the other half were
based on names (e.g. Peter is taller than Paul. Paul is taller than Mary). y
Instruments Pre-experimental instruments to ensure group homogeneity: a) Pre-experimental instruments to ensure group homogeneity: i. Demo. A simple demographics survey to enable testing of group
homogeneity and consisting of questions regarding gender, age,
university credits, number of times the participant had sat the Higher
Education Entrance Examination, average hours of computer use per
week. b) Primary tasks: b) Primary tasks: 24 E. Wästlund: Experimental Studies of Human-Computer Interaction Instruments a) Pre-experimental instruments to ensure group homogeneity: a) Pre-experimental instruments to ensure group homogeneity: i. Demo. (See Experiment 1). i. Demo. (See Experiment 1). b) Primary tasks: b) Primary tasks: i. Reading comprehension (See Experiment 1). ii. Syllogistic reasoning (See Experiment 1) c) Secondary tasks: Design (See Experiment 1). d) Post-experimental evaluation: d) Post-experimental evaluation: i. Performance. This test consisted of two questions, one positively and one
negatively worded, which allowed the participants to rate, on a five point
Likert scale, their own performance during the two assignments. The
participants were asked to rate their performance during both the first
and second text assignments. Procedure The experiment was conducted in a computer room containing 16 computers at
Karlstad University. When the participants arrived, the only verbal instructions they
were given were to “take a seat”, to turn off their cellphones, and to look for all the
information they might require on screen. Once all the participants had done this, they
were then instructed to click the ‘start experiment’ button on their computer screens. Upon initializing the experiment, ELMA CyberLab randomly allocated each participant
to one of the two secondary task groups and to one of the two primary task subgroups. c) Secondary tasks: c) Secondary tasks:
To investigate the effects of increased memory load on the phonological loop, two
types of pop-ups (PopUp and PopMem) were used. In both assignments, responses
were performed by placing the mouse pointer over the relevant button and clicking
the left mouse button. During the course of the each primary task, the participants
were interrupted by 10 pop ups. PopUp incidence was based on a list of ten
intermissions varying between 30 and 76 seconds. The order of the intermissions was
randomized before each trial making it seem like the pop-ups appeared at random,
from the participants’ point of view. The dependent variable of the secondary task
was the mean reaction time of the responses to PopUp occurrences. In order to
minimize error variance, the trimmed mean was used, i.e. the highest and lowest RT
value of each participant was excluded, thus, the individual mean RT was based on 8
of the 10 responses. i. In the PopUp assignment, the participants were instructed to respond to
pop-ups by simply clicking on a button labelled ‘close’. i. In the PopUp assignment, the participants were instructed to respond to
pop-ups by simply clicking on a button labelled ‘close’. i. In the PopUp assignment, the participants were instructed to respond to
pop-ups by simply clicking on a button labelled ‘close’. ii. In the PopMem assignment, the participants were first presented with a
two-digit target number. The next pop-up showed a number which was
either the same or different. The participant was then instructed to
indicate whether or not this was the target number or not by clicking on a
‘yes’, ‘no’, or ‘don’t remember’ button. Having done so, a new target 25 E. Wästlund: Experimental Studies of Human-Computer Interaction number was then displayed and the participant was allowed to continue
reading the text. Results The results showed that preloading the phonological loop had a detrimental effect
on Syllogistic reasoning and that mean RTs were shorter during the Syllogistic
reasoning task than during Reading comprehension. i.
Performance. (See Experiment 1). Procedure
(See Experiment 1). Procedure
(See Experiment 1). c) Secondary tasks: In order to investigate the effects of visuo-spatial load, two types of pop-ups
(PopUp / PopUpV) were applied. In order to magnify the load on the sketchpad, the
pop-ups occurred in sequences of four with half-second intervals between the click
to close the pop-up and its next appearance, thus resulting in a total number of forty
secondary task responses. However, in comparisons of secondary tasks the reaction
time measurement used is from the appearance to closure of the first in the
sequence. q
i. The secondary task PopUp was identical to the one used in Experiment
1. ii. PopUpV was also identical in all respects, except for the position on the
screen where it was displayed. Instead of constantly popping up in the
centre of the screen, the position of the PopUpV was randomized. Thus, 26 E. Wästlund: Experimental Studies of Human-Computer Interaction in order to click on the close button and continue, the participants were
required to identify the exact position on the screen to were the pointer
should be moved and then execute the movement. d) Post-experimental evaluation: Design (See Experiment 1). Results The results showed no effect of visuo-spatial load on any of the two primary
tasks. The results of the secondary task analysis, however, replicated the results of
Experiment 1 were mean RTs were lower during Syllogistic reasoning than during
Reading comprehension. Reading comprehension. Instruments a) Pre-experimental instruments to ensure group homogeneity: a) Pre-experimental instruments to ensure group homogeneity: i. Demo. (See Experiment 1). ii. LetterRain. The object of this test is to measure the participants’
keyboard skills in order to ensure group homogeneity. During this one
minute test, letters rain down from the top of the screen (with a
maximum of four letters on the screen simultaneously) and the
respondents task is to press the keys corresponding to the letters. When
the correct key is pressed, the letter disappears and a new letter appears
at the top of the screen. The dependent variable was the number of
correct keystrokes. d) Post-experimental evaluation: i. Performance. (See Experiment 1). c) Secondary tasks: i. PopUp. (See Experiment 1). ii. PopMem. (See Experiment 1). Procedure Procedure
(See Experiment 1). Procedure
(See Experiment 1). Design (See Experiment 1). Instruments a) Pre-experimental instruments to ensure group homogeneity: i. Demo. (See Experiment 1). ii. LetterRain. b) Primary tasks: b) Primary tasks: i. Snippets (See Experiment 3) ii. Headlines (See Experiment 3) c) Secondary tasks: d) Post-experimental evaluation: i. Performance. (See Experiment 1). Results The results showed no effect of preloading the phonological loop. The results of
the secondary task analysis, however, showed that mean RTs were shorter during then
Headlines test than during the Snippets test. b) Primary tasks: The primary tasks Snippets and Headlines are both verbal creativity tests
measuring divergent production. The main difference between the two tasks lies in
the amount of information to be processed and produced for task completion. During both primary tasks, each encompassing a duration of ten minutes, the
participants can switch between the three stimuli at any point. The dependant
variable was character fluency. The primary tasks Snippets and Headlines are both verbal creativity tests
measuring divergent production. The main difference between the two tasks lies in
the amount of information to be processed and produced for task completion. During both primary tasks, each encompassing a duration of ten minutes, the
participants can switch between the three stimuli at any point. The dependant
variable was character fluency. i. The primary task, Snippets, assesses the ability to transform and
elaborate information. Participants are presented with three headlines and
the task is to write a newspaper snippet that might accompany the given
headline. The test is a variation of the ESL composition task (Sandlund,
Linnarud, & Norlander, 2001), where the participants are given four
random words as a starting point. ii. The primary task, Headlines (Ekvall, 1969), assesses the ability to
transform one unit of information into another, in this case to condense, 27 E. Wästlund: Experimental Studies of Human-Computer Interaction from a newspaper snippet, the essential text to provide a headline. During the task, the participants are presented with three snippets and the
object of the test is to construct as many headlines as possible for the
snippets. c) Secondary tasks: Results The results showed that a visuo-spatial load had a grater detrimental effect on the
Snippets task. d) Post-experimental evaluation: d) Post-experimental evaluation: i. Performance. (See Experiment 1). i. Performance. (See Experiment 1). Procedure
(See Experiment 1). Procedure
(See Experiment 1). c) Secondary tasks: In order to investigate the effects of visuo-spatial load, two types of pop-ups
(PopUpCN / PopUpVN) were applied. The object of the secondary tasks was the
same as in Experiment 2 i.e. to investigate the effects of visuo-spatial load. However, as the response execution of the primary tasks differs between Experiment
2 (mouse only) and Experiment 4 (keyboard only). The PopUps were adapted so
that responses were executed by means of pressing keys instead of clicking on the
muse. All pop-ups presented the participant with a number (0-9), upon which he or
she had to press the corresponding key in order to continue with the primary task. i. During the low-load task (PopUpCN) the PopUp occurred at the same
position on the screen every time and the number was always the same
(0). ii. During the high-load task (PopUpVN), the pop-ups occurred at random
positions and displayed a random one-digit number. 28 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction maximal performance may be influenced rests on the assumption that there is a limit to
the processing capacity of the working memory. The main finding of Study I, Experiment 2, indicated that production of
information, as assessed by testing creative production, resulted in a higher level of
fluency when the material was presented on paper, albeit without affecting the quality of
the headlines. Essentially, this finding confirms that of Study I, Experiment 1, and may
be explained in the same manner. Nevertheless, methodological discrepancies between
the two studies require some consideration. In addition to the assignments’ different
forms, i.e. consumption versus production of information, there was a discrepancy in
the time dimension. The consumption of information assignment was arranged so that
the participants were allowed 30 minutes in which to complete the assignment, whereas
the production of information assignment was arranged so that the participants were
required to spend 30 minutes on the task, the former necessitating completion and the
latter requiring only compliance. In addition, the amount of text in the production of
information assignment was less than 10% of the amount of text in the consumption of
information assignment. It is important to bear in mind that the performance
deterioration in the divergent problem-solving task only affected fluency and not
quality. The lack of differences in assessed quality may be an effect of the assignment
configuration, with regard to either the shorter period of time that participants were
actually handling the computers, or to the assignment’s divergent nature. Divergent
production demands flexibility in the thought processes (Lubart, 2001). It is thus
possible that the dual-task nature of the original assignment, combined with and
interrupted by handling the computer, facilitates flexible thinking by ensuring that
individuals move their focus of thought between different avenues. In sum then, despite the methodological discrepancies observed, both Study I,
Experiment 1, and Study I, Experiment 2, showed that on-screen presentation impaired
performance. The finding that the participants in Study I, Experiment 1 (production of
information), performed and responded in a similar fashion to the participants in Study
I, Experiment 2 (consumption of information), in spite of the considerably smaller
amount of text in the assignment, may be interpreted as an indication that cognitive load
on a limited-capacity working memory, as opposed to reading speed, is a central
component of the detrimental effects of text presented on computer screens. 3.1 Introduction to conclusions The main findings of the three studies can be summarized in the following points • Consumption of information is more effective when the information is
presented on paper than when displayed on a computer screen (Study I:
Experiment 1). • Equally, production of information is more effective when the
information is presented on paper than when displayed on a computer
screen (Study I: Experiment 2). • Consumption of information generates less mental workload when page
layout is adapted to fit the computer screen (Study II: Experiments 1 &
2). • Problem solving processes, including both Consumption and Production
of information, may be described both in terms of their reliance on either
ST- or LT-WM (Study III: Experiments 1, 2 & 3) and in terms of their
reliance on specific ‘slave-systems’ of the tripartite model (Study III:
Experiments 1 & 3). Since the basic aim of any experiment is to draw conclusions regarding the
independent variable on the basis of measurements of the dependent variable, it is
important to understand the configuration of the latter. The primary tasks, consumption
and production of information, are both tests of maximal performance, i.e. they measure
an individual’s capacity to perform a given assignment. The object of the experiments,
therefore, is to influence this capacity, to such extent that it is measurable, through
variations in the mode of presentation (Study I) as well as the page layout (Study II) or
through the use of secondary tasks (Studies II & III). The main finding of Study I Experiment 1 was that consumption of information,
measured via a reading comprehension test, is more difficult when the assignment text
is presented on a computer screen than when it is presented in printed form. In other
words, presentation of the material on a computer screen reduces an individual’s
capacity for consumption of information. The notion that the capacity for a test of 29 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction any measurable extent. However, it is important to remember that the reading
comprehension task in Study I, Experiment 1, consisted of a total of 20 questions and
that the participants worked on their assignment for 30 minutes. In the experiments in
Study II, the timeframes and the number of questions were considerably smaller (10
minutes and 4 questions in Experiment I, and 7 minutes and 6 questions in Experiment
II). Thus, in comparison, the manipulation was weaker and the measurements less
precise in Study II. In Studies II and III, a secondary task was added to the primary tasks of
consumption and production of information. If the secondary task, which is also a task
of maximal performance, utilizes the same memory resource as the primary task this
consideration ought to lead to impairment of at least one of the two tasks. However, in
order to generate measurable detrimental effects the two tasks have to, not only draw on
the same resources, but also exceed available resources. Thus, as in Study II, when the
primary task is manipulated, but not to the extent that it becomes impaired, the effects
of manipulation on mental workload can be measured through the secondary task
(Brunken, Steinbacher, Plass, & Leutner, 2002; Marcus, Cooper, & Sweller, 1996). In
Study III on the other hand, the different primary tasks are contrasted with each other
rather than slightly manipulated making the interpretation of RTs more complex as the
differences might just as well stem from differences in the tasks memory demands as
their inherent mental workload. In terms of detrimental effects on the primary task by the secondary tasks the
result of Study III, Experiment 3, show that the secondary task utilizing resources of
the visuo-spatial sketchpad has detrimental effects on text elaboration and Study III,
Experiment 1, show that syllogistic reasoning is impaired by pre-loading the
phonological loop. No effects on the primary tasks of snippets (Study III) and reading
comprehension (Studies II and III) were, however, observed. The latter results are in
line with previous research showing that neither brief interruptions (Glanzer, Dorfman,
& Kaplan, 1981) nor a three-digit working memory load (Baddeley & Hitch, 1974),
while reading, has any significant effect on comprehending the material. E. Wästlund: Experimental Studies of Human-Computer Interaction This notion
corroborates that of Mayes et al. (2001), who showed that cognitive workload plays an
important role for performance in computer-aided environments. Since the documents used in Study I were based on the A4 format, the question
is, then, whether or not the layout aspect of navigation can be identified as a source of
the demonstrated detrimental effects of on-screen presentation. Based on this
observation, the aim of Study II was to investigate the effects of page layout. It is again
important to point out that the nature of the task of consumption of information, as
measured through a test of reading comprehension, is one of measuring maximal
performance and that the aim of the experiment is to manipulate and measure the ability
to complete the task. Thus, in order for the manipulation to have a measurable effect, it
needs to be of such magnitude that the mental workload is not just increased but
increased to such an extent that performance is impaired. The results of both
experiments in Study II indicate that performance was not affected by page layout to 30 E. Wästlund: Experimental Studies of Human-Computer Interaction PopUps occur quite frequently when using a computer and thus offer the clear potential
for a promising avenue of development. PopUps occur quite frequently when using a computer and thus offer the clear potential
for a promising avenue of development. In terms of RTs, the main finding of Study II, Experiments 1 and 2, indicated that
consumption of information generates less mental workload, as measured through
reaction times to the secondary task, when page layout is adapted in order to fit the
computer screen. The rationale used in this interpretation of the reaction time results is
that the level of mental workload determines the cognitive resources available for
reacting to the occurrence of a secondary task stimulus (Brunken et al., 2002; Marcus et
al., 1996; Verwey & Veltman, 1996; Wickens, 1984), in this case the “PopUps”. There
are two plausible explanations for the observed workload differences: Firstly,
manipulating an on-screen text document via scrolling necessitates a shift of focus from
the text to the action of controlling the page movement. The main difference between
the screen and scroll layouts is the amount of visual cues facilitating control of the page
movement. In the absence of visual cues, the reader is obliged to match the movement
of the page with lines of text, which in turn has to be matched against the memory of the
text, in order to determine whether or not the text has been moved a satisfactory
distance or whether more scrolling is required. In terms of visual cues, the A4 format
falls between the screen and scroll layouts as it is possible to use the top (but not the
bottom) of the page as a meaningful navigational cue. Secondly, the absence of visual
cues impedes the encoding of information in a two-dimensional space, thereby
rendering the decoding of meaning more cumbersome (Piolat et al., 1997). It is
noteworthy that, despite the fact that no differences were shown in terms of the primary
task, a significant effect was found with regard to the secondary task. On the one hand,
these findings support the notion of using dual-task experiments to investigate mental
workload. However, on the other hand, they also raise further questions regarding the
paper vs. computer issue that pertains to possible Type II errors in studies showing that
no differences exist. E. Wästlund: Experimental Studies of Human-Computer Interaction The result of
Study III, Experiment 1, however, appears to contradict a previous finding (Gilhooly et
al., 1993) which asserted that the phonological loop only played a minor role in
syllogistic reasoning. On closer inspection of the methodology and model used by
Gilhooly et al. (1993), the findings do not seem so disparate, but rather warrant a
reinterpretation of the their interpretations. Taken together, these findings from Studies II and III, i.e. that the additional
processing demands made by a secondary task on the ‘slave-systems’ of the Baddeley
and Hitch model only affect problem solving tasks relying on the specific ‘slave-
system’, confirms rather than elaborates the present notions of working memory
processes. Nevertheless, the type of methodology introduced, i.e. web-distributed
experimentation, offers a novel approach in this area of research. The advantages of the
method not only provide the possibility of testing groups of individuals simultaneously,
but also the opportunity of creating secondary tasks with a higher ecological validity. Although, concurrent articulatory suppression has been proven to be an effective
secondary task, it hardly mimics any real-world phenomena, whereas different types of 31 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction The use of a secondary task is not the only way of measuring mental workload. A
more frequently used method consists of various forms of subjective rating scales. Ratings of subjective experiences usually take one of two forms: either the task is
appraised or the respondent rates his or her current state. The two methods can be
exemplified using the questions: “how cumbersome was the task” and “how tired did
the task make you”. Although the former can generally be said to be more precise, there
is a greater risk of task appraisal than of rating the actual task. State rating, on the other
hand, is less afflicted by task appraisal. However, on the other hand, it has less precision
as it is not a rating of the task itself but of the affects of the task on the individual. In
both Study I and Study II, the participants’ subjective ratings of their current states
were included. The results of Study I, Experiment 1 (consumption of information), demonstrated
that participants rated their subjective states of stress and tiredness higher in the group
where the material was presented on-screen than when the material was presented in
printed form. In Study I, Experiment 2 (production of information), no significant
differences were observed but there was a trend indicating that the participants in the
on-screen group reported higher levels of experienced stress. In the two experiments in
Study I, rating was performed after task completion and the results were compared
between the groups. As Study II involved both repeated measurement and counter
balancing of the task order, ratings were carried out on three occasions, before and after
the first task and then again after the second task. The difference between before and
after a given task was then used as a measure of the task’s effect on the individual. The
results of the two experiments in Study II show no significant differences: neither
between the type of layout nor between the type of secondary task. It is again important
to remember the differences in respect of the assignment times in Studies I and II. The
absence of any observed differences in Study II may be interpreted as being either non
existent or too small, due to the short assignment time, and thus difficult for participants
to evaluate and report (Mayes et al., 2001). E. Wästlund: Experimental Studies of Human-Computer Interaction The observed increased stress and tiredness levels in Study I, Experiment 1, may
be explained in the same way as the observed performance deterioration. Either the
focus is on the quality of on-screen presented material, which necessitates a greater
mobilization of both perceptual and executive cognitive resources in order to
compensate for the deficiency of the information presentation, or one may focus on the
dual-task of handling the equipment and the performance of the original assignment,
which necessitates a greater mobilisation of cognitive resources to compensate for the
added mental workload. Furthermore, in the light of the results of Study II, at least
some of the added workload may be attributed to the differences in page display and
their consequences for document manipulation. Irrespective of which particular mode of
explanation is embraced, the heart of the matter pertains to an increase in mental
workload. E. Wästlund: Experimental Studies of Human-Computer Interaction In terms of RTs the main finding from Study III, Experiments 1, 2 and 3, was
that problem solving processes that draw more heavily on LT-WM memory elicit longer
reaction times than do processes which are processed in the ST-WM. This interpretation
is based on the finding of Cantor & Engle (1993), who not only showed that long term
memory structure activation is an integral part of working memory performance, but
also that the degree of activation may be measured through the use of reaction times. It
is noteworthy that this finding was observed both when comparing reading
comprehension with syllogistic reasoning (Experiments 1 and 2) and when comparing
the Snippets and Headlines tasks (Experiment 3). It is equally noteworthy that the effect
was shown independent of secondary task ‘slave-system’ load (phonological loop in
Experiments 1 and 3 and visuo-spatial sketchpad in Experiment 2). Given the rather low
power of the experiments, the fact that the effect was observed in three out of four
experiments, despite there inherent differences, indicates that this is a reliably stable
effect. As the primary tasks contrasted in Study III, have been shown to utilize
different memory resources through the use of secondary tasks, it is more feasible that
the observed differences in mean RTs are more a result memory demands than
attributable to mental workload. 32 3.2 Final conclusion In brief, Study I indicated that the presentation of information on-screen vis-à-vis
in printed form produced a detrimental effect on human information processing and
Study II showed that some of that detrimental effect may be attributed to the 33 E. Wästlund: Experimental Studies of Human-Computer Interaction differences in the navigational properties of the two media. In addition, the result from
Study II demonstrates that, by adapting the page layout of the presented material to fit
its intended medium, mental workload may be alleviated. Taken together, both Study I
and Study II show that the presentation of information affects information processing
and thus highlighting the importance of further human computer interaction research in
this area. The results of Study III showed that, in order to understand the memory
demands of complex cognition, it is important to include elements of both the ST- and
LT-WM paradigm of Ericsson & Kintsch (1995) and the tripartite model of Baddeley &
Hitch (1974). From a theoretical point of view, the results show that, in order to understand the
memory demands of complex cognition, it is important to include elements of both the
ST- and LT-WM paradigm of Ericsson & Kintsch and the tripartite model of Baddeley
& Hitch. This notion is based on the findings that reading comprehension and text
production are similar in the sense that they both depend on LT-WM demands but that
they differ in that, of the two, only text production is susceptible to an additional load
on the visuo-spatial sketchpad. In the same fashion, syllogistic reasoning and the
condensation of a short text are similar in that they both rely on ST-WM processes,
whereas only syllogistic Reasoning is susceptible to an additional load on the
phonological loop. Thus, future research investigating complex cognition and human
information processing ought to expand current views on the structure of memory
resources (see figure 4). Figure 4: A proposed model for complex cognition that includes both the LT-/ST-WM and the
tripartite working memory models [modified from Baddeley and Hitch (1974) and Ericsson and
Kintsch (1995)]. The explicit processes of the visuo-spatial sketchpad and phonological loop are
placed within the ST-WM whereas the implicit processes of the central executive are placed
within the LT-WM. 3.2 Final conclusion LTM
LT-WM
ST-WM
Visuo-Spatial
sketchpad
Phonological
loop
Input
Output
Central Executive Figure 4: A proposed model for complex cognition that includes both the LT-/ST-WM and the
tripartite working memory models [modified from Baddeley and Hitch (1974) and Ericsson and
Kintsch (1995)]. The explicit processes of the visuo-spatial sketchpad and phonological loop are
placed within the ST-WM whereas the implicit processes of the central executive are placed
within the LT-WM. 34 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction From a methodological point of view, the most important aspect concerns the
measurement of information processing power and mental workload using dual-task
experiments. As the measurement of processing ability is, in essence, a measure of
maximum capacity, the manipulation has to add sufficient mental workload to the
original task so that task demands exceed the available processing resources. However,
using secondary task reaction time data, it is possible to measure the available
resources. Thus, when the primary task is manipulated, but not to the extent that it
exceeds available resources, the effects of the manipulation on mental workload may be
measured through the secondary task. A final methodological note ought to be made
regarding the appropriateness of conducting dual-task studies of human information
processing using distributed web experimenting. Not only is it possible to design
experiments of high ecological validity, as tasks and graphics may be designed to
resemble those usually encountered by participants, the possibility of group testing and
easy data acquisition also makes it a very appealing alternative. From a practical point of view, the results demonstrate two aspects of complex
cognition. Firstly, comparing the effects of the mode of presentation can give us clues
regarding what influences mental workload. It is, however, pointless to do this in order
to advocate the abolishment of digitally presented material. Nevertheless, such findings
are useful starting points in research concerning how to minimize mental workload
during human-computer interaction. Secondly, and more importantly, in order to
optimize the information processing ability during consumption of information of
onscreen presented material, it is time to abandon the high-tech recycling of the ancient
scroll and to break free of the traditions of printing, i.e. displaying written material in
the standard A4 format and instead adapt the page layout to fit its intended medium. 35 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction 4 References Amabile, T. M. (1996). Creativity in Context. Oxford: Westview press. Baddeley, A. D. (1996). Exploring the Central Executive. Quarterly Journal of
Experimental Psychology: Section A, 49(1), 5-28. Baddeley, A. D., & Hitch, G. J. (1974). Working memory. In G. A. Bower (Ed.), Recent
advances in learning and motivation (Vol. 8, pp. 47–90). New York: Academic
Press. Belmore, S. M. (1985). Reading computer-presented text. Bulletin of the Psychonomic
Society, 23(12-14). Brunken, R., Plass, J. L., & Leutner, D. (2003). Direct Measurement of Cognitive Load
in Multimedia Learning., Educational Psychologist (Vol. 38, pp. 53-61):
Lawrence Erlbaum Associates. Brunken, R., Steinbacher, S., Plass, J. L., & Leutner, D. (2002). Assessment of
Cognitive Load in Multimedia Learning Using Dual-Task Methodology. Experimental Psychology, 49(2), 109-119. Cantor, J., & Engle, R. W. (1993). Working-Memory Capacity as Long-Term Memory
Activation: An Individual-Differences Approach. Journal of Experimental
Psychology: Learning, Memory and Cognition, 19(5), 1101-1114. Daneman, M., & Carpenter, P. A. (1980). Individual differences in working memory
and reading. Journal of Verbal Learning and Verbal Behavior, 19, 450–466. Daneman, M., & Merikle, P. M. (1996). Working memory and language
comprehension: A meta-analysis. Psychonomic Bulletin & Review, 3(4), 422-
433. Duff, S. C., & Logie, R. H. (2001). Processing and storage in working memory span.,
Quarterly Journal of Experimental Psychology: Section A (Vol. 54, pp. 31-48):
Psychology Press (T&F). Ekvall, G. (1969). Rubriker: Användarmanual [Headlines: user manual]. Malmö:
Psykologiförlaget [Psychology Publishers]. Ericsson, K. A., & Kintsch, W. (1995). Long-term working memory. Psychological
Review, 102, 211-245. Ericsson, K. A., & Lehmann, A. C. (1996). Expert and exceptional performance:
Evidence and maximal adaptation to task constraints, Annual Review of
Psychology (Vol. 47, pp. 273): Annual Reviews Inc. Ghiselin, B. (1963). Ultimate Criteria for Two Levels of Creativity. In F. Barron (Ed.),
Scientific Creativity. New York: Whiley and Son. Gilhooly, K. J., Logie, R. H., Wetherick, N. E., & Wynn, V. (1993). Working memory
and strategies in syllogistic-reasoning tasks. Memory & Cognition, 21(1), 115-
124. Glanzer, D., Dorfman, D., & Kaplan, B. (1981). Short-term strage in the processing of
text. Journal of Verbal Learning and Verbal Behavior, 20, 656-670. Gray, S. H. (1991). Information search with dynamic text vs. paper text: an emperical
comparison. International Journal of Man-machine Studies, 35(575-586). Gray, S. H. (1991). Information search with dynamic text vs. paper text: an emperical
comparison. International Journal of Man-machine Studies, 35(575-586). Guilford, J. P. (1967). The nature of human intelligence. New York: McGraw-Hill. E. Wästlund: Experimental Studies of Human-Computer Interaction Hart, S. G., & Staveland, L. (1988). Development of the NASA task load index (TLX):
Results of empirical and theoretical research. In N. Meshkati (Ed.), Human
mental workload (pp. 39–183). Amsterdam: North-Holland. pp
Holmquist, R. (1974). Testbeskrivning Syllogismer I och II [Manual to the Syllogisme
I and II Test]. Malmö: Meddelande fron Metodoteket. James, W. (1890). The Principles of Psychology. New York: Dover. Jolicoeur, P., & Dell, A. R. (1999). Attentional and structural constraints on visual
encoding. Psychological research/ Psychologische forschung(62), 154-164. g
y
g
y
g
f
g(
)
Kahneman, D. (1973). Attention and Effort. Englewood Cliffs, NJ: Prentice-Hall. Kintsch, W., Healy, A., Hegarty, M., Pennington, B., & Salthouse, T. (1999). Models of
working memory: Eight questions and some general answers. In P. Shah (Ed.),
Models of working memory: Mechanisms of active maintenance and executive
control. New York: Cambridge University Press. g
y
Kintsch, W., Patel, V. P., & Ericsson, K. A. (1999). The role of long-term working
memory in text comprehension. Psychologia, 42, 186-198. Kjellberg, A., & Iwanowski, S. (1989). Stress/energiformuläret: Utveckling av en metod
för skattning av sinnesstämning i arbetet. [The Stress/arousal questionnaire:
Development of a method for assessment of mood at work]
(Undesrsökninsrapport [Research report] No. 1989:26). Solna: Sweden:
Arbetsmiljöinstitutet [National Institute for Workplace Protection]. Koch, I., & Prinz, W. (2002). Process interference and code overlap in dual-task
performance. Journal-of-Experimental-Psychology: Human Perception and
Performance(28), 192-201. Kolers, P. A., Duchnicky, R. L., & Ferguson, D. C. (1981). Eye movement
measurement of CRT displays. Human Factors(23), 517-527. Kumashiro, M. (1995). Practical measurement of psychophysiological functions for
determining workloads. In E. N. Corlett (Ed.), Evaluation of human work: A
practical ergonomics methodology (2nd ed.). Philadelphia, PA: Taylor and
Francis. Lubart, T. I. (2001). Models of the creative process: Past, present and future. Creativity
Research Journal, 13(3-4), 295-308. Marcus, N., Cooper, M., & Sweller, J. (1996). Understanding Instructions. Journal of
Educational Psychology, 88(1), 49-63´. y
gy
Markman, A. B. G., Dedre. (2001). THINKING. Annual Review of Psychology, 52(1),
223. Martinsen, Ø., & Kaufmann, G. (1999). Cognitive Style and Creativity. In S. R. Pritzker
(Ed.), Encyclopedia of creativity. San Diego: Academic Press. Mayes, D. K., Sims, V. K., & Koonce, J. M. (2001). Comprehension and workload
differences for VDT and paper-based reading. International Journal of
Industrial Ergonomics, 28(6), 367-378. Miller, G. A. (1956). The Magical Number Seven, Plus or Minus Two: Some Limits on
Our Capacity for Processing Information. 4 References comparison. International Journal of Man-machine Studies, 35(575-586). Guilford, J. P. (1967). The nature of human intelligence. New York: McGraw-Hill. Guilford, J. P. (1967). The nature of human intelligence. New York: McGraw-Hill. 36 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction Noyes, J. M., & Garland, K. J. (2003). VDT versus paper-based text: reply to Mayes,
Sims and Koonce. International Journal of Industrial Ergonomics, 31(6), 411-
423. Olive, T. (2003). Working Memory in Writing:Empirical Evidence From the Dual-Tas
Technique. European Psychologist. Olive, T. (2003). Working Memory in Writing:Empirical Evidence From the Dual-Task
Technique. European Psychologist. Paas, F., Renkl, A., & Sweller, J. (2003). Cognitive Load Theory and Instructional Paas, F., Renkl, A., & Sweller, J. (2003). Cognitive Load Theory and Instructional
Design: Recent Developments. Educational Psychologist, 38(1), 1-4. Paas, F., Tuovinen, J. E., Tabbers, H., & Van Gerven, P. W. M. (2003). Cognitive Load
Measurement as a Means to Advance Cognitive Load Theory., Educational
Psychologist (Vol. 38, pp. 63-71): Lawrence Erlbaum Associates. Parasuraman, A. (2000). Technology Readiness Index [TRI] A Multiple-Item Scale to
Measure Readiness To Embrace New Technologies. Journal of Searvice
Research, 4(2), 307-320. Partridge, D., & Rowe, J. (1994). Computers and Creativity. Wiltshire, England:
Intellect Books. Pashler, H. E. (1998). The Psychology of Attention. Cambridge, Mass.: MIT. Peterson, L. R., & Peterson, M. J. (1959). Short-term retention of individual verbal
items. Journal of Experimental Psychology, 58, 193-198. Piolat, A., Olive, T., & Kellogg, R. T. (2005). Cognitive effort during note taking. Applied Cognitive Psychology, 19(3), 291-312. Piolat, A., Roussey, J.-Y., & Thunin, O. (1997). Effects of screen presentation on text
reading and revising. International Journal of Human-Computer Studies, 47(4),
565-589. Reid, G. B., & Nygren, T. E. (1988). The Subjective Workload Assessment Technique:
A Scaling Procedure for Measuring Mental Workload. In N. Meshkati (Ed.),
Human Mental Workload (pp. 185-218). Amsterdam: North-Holland. Reips, U.-D. (2002). Standards for Internet-based experimenting. Experimental
Psychology, 49(4), 243-256. Rice, G. (1994). Examining constructs in reading comprehension using two presentation
modes: paper vs. computer. Journal of Educational Computing Research, 11,
153-178. Sandlund, E., Linnarud, M., & Norlander, T. (2001). Effects of Stress versus Flotation-
REST Relaxation on Creativity and Literacy Skills in Advanced English as a
Second Language (ESL) Composition. RASK:, 15, 95-113. Squire, L. R. (1992). Memory and the hippocampus: A synthesis from findings with
rats, monkeys, and humans., Psychological Review (Vol. 99, pp. 195). Sternberg, S. (2001). Separate modifiability, mental modules, and the use of pure and
composite measures to reveal them. Acta Psychologica, 106(1-2), 147-246. Tulving, E. (1999). Study of Memory: processes and systems. In M. Jelicic (Ed.),
Memory: systems, process, or function? E. Wästlund: Experimental Studies of Human-Computer Interaction The Psychological Review, 63, 81-97. Miyake, A., & Shah, P. (1999). Models of working memory: Mechanisms of active
maintenance and executive control. New York: Cambridge University Press. g
y
Miyake, S. (2001). Multivariate workload evaluation combining physiological and
subjective measures. International Journal of Psychophysiology, 40(3), 233-238. 37 E. Wästlund: Experimental Studies of Human-Computer Interaction Oxford: Oxford University Press. y
y
p
f
y
Waern, Y. (1989). Cognitive aspects of computer supported tasks. Chicester: Wiley. van der Linden, D., Frese, M., & Meijman, T. F. (2003). Mental fatigue and the control
of cognitive processes: effects on perseveration and planning. Acta
Psychologica, 113(1), 45-65. Van Merriënboer, J. J. G., & Sweller, J. (2005). Cognitive Load Theory and Complex
Learning: Recent Developments and Future Directions. Educational Psychology
Review, 17(2), 147-177. 38 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction Van Orden, K. F., Limbert, W., Makeig, S., & Jung, T. P. (2001). Eye activity correlates
of workload during a visuospatial memory task. Human Factors, 43(1), 111-
121. Waters, G. S., & Caplan, D. (2004). Verbal working memory and on-line syntactic
processing: Evidence from self-paced listening, Quarterly Journal of
Experimental Psychology: Section A (Vol. 57, pp. 129-163): Psychology Press
(T&F). Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief
measures of positive and negative affect: The PANAS scales. Journal of
Personality and Social Psychology, 54, 1063-1070. Verwey, W. B., & Veltman, H. A. (1996). Detecting Short Periods of Elevated
Workload: A Comparison of Nine Workload Assessment Techniques. Journal of
Experimental Psychology: Applied, 2(3), 270-285. Wickens, C. D. (1984). Processing resources in attention. New York: Academic. Wickens, C. D., Mavor, A. S., & McGee, J. P. (Eds.). (1997). FLIGHT TO THE
FUTURE: HUMAN FACTORS IN AIR TRAFFIC CONTROL. Washington,
D.C.: NATIONAL ACADEMY PRESS. Wierwille, W. W., & Eggemeier, F. T. (1993). Recommendations for mental workload
measurement in a test and evaluation environment. Human Factors, 35, 263–
281. Xie, B., & Salvendy, G. (2000). Review and reappraisal of modelling and predicting
mental workload in single- and multi-task environments., Work & Stress (Vol. 14, pp. 74-99): Taylor & Francis Ltd. Yates, F. M. (1966). The Art of Memory. London: Pimlico. Ziefle, M. (1998). Effects of display resolution on visual performance. Human Factors,
40(4), 554-568. 39 E. Wästlund: Experimental Studies of Human-Computer Interaction E. Wästlund: Experimental Studies of Human-Computer Interaction Study I Wastlund, E., Reinikka, H., Norlander, T., & Archer, T. (2005). Effects of VDT and
paper presentation on consumption and production of information: Psychological
and physiological factors. Computers in Human Behavior, 21, 377-394. Wastlund, E., Reinikka, H., Norlander, T., & Archer, T. (2005). Effects of VDT and
paper presentation on consumption and production of information: Psychological
and physiological factors. Computers in Human Behavior, 21, 377-394. Study II Wastlund, E., Norlander, T., & Archer, T. (Submitted). The Effect of Page Layout on
Mental Workload: A Dual-Task Experiment. Wastlund, E., Norlander, T., & Archer, T. (Submitted). The Effect of Page Layout on
Mental Workload: A Dual-Task Experiment. Study III Wastlund, E. & Archer, T. (Submitted). Working Memory Loads derived from
Computer-based Primary- and Secondary-Tasks 40
|
https://openalex.org/W4280521365
|
https://epub.uni-bayreuth.de/6643/1/s41598-022-11572-3.pdf
|
English
| null |
Preprocess dependence of optical properties of ensembles and single siphonaxanthin-containing major antenna from the marine green alga Codium fragile
|
Scientific reports
| 2,022
|
cc-by
| 12,516
|
Tatas Hardo Panintingjati Brotosudarmo1,2, Bernd Wittmann1, Soichiro Seki3,
Ritsuko Fujii3,4 & Jürgen Köhler1,5,6* Tatas Hardo Panintingjati Brotosudarmo1,2, Bernd Wittmann1, Soichiro Seki3,
Ritsuko Fujii3,4 & Jürgen Köhler1,5,6* The siphonaxanthin-siphonein-Chl-a/b-protein (SCP) is the light-harvesting complex of the marine
alga Codium fragile. Its structure resembles that of the major light-harvesting complexes of higher
plants, LHC II, yet it features a reversed Chl a:Chl b ratio and it accommodates other variants of
carotenoids. We have recorded the fluorescence emission spectra and fluorescence lifetimes from
ensembles and single SCP complexes for three different scenarios of handling the samples. While
the data obtained from ensembles of SCP complexes yield equivalent results, those obtained
from single SCP complexes featured significant differences as a function of the sample history. We
ascribe this discrepancy to the different excitation intensities that have been used for ensemble
and single complex spectroscopy, and conclude that the SCP complexes undergo an aging process
during storage. This process is manifested as a lowering of energetic barriers within the protein,
enabling thermal activation of conformational changes at room temperature. This in turn leads to the
preferential population of a red-shifted state that features a significant decrease of the fluorescence
lifetime. Photosynthesis is exploited by plants, bacteria, and algae to convert solar energy into biochemical energy. This is
achieved by the absorption of sunlight in specialised light-harvesting complexes (LHCs), and subsequent ultrafast
and efficient transport of the excitation energy to a photochemical reaction centre (RC) that acts as transducer
driving a transmembrane charge separation1. Since the prerequisites that have to be fulfilled for electron transfer
are quite demanding, the structure of the RCs is mainly conserved across the various organisms that perform
photosynthesis. In contrast, the rules for energy transfer are rather tolerant resulting in a rich variety for the
structures of the light harvesting systems allowing for perfect adaption to the special conditions that apply to the
habitat of the organism2. In contrast to the LHCs from higher plants the antennae systems from marine algae
have only recently attracted more attention3–7. Codium fragile is a marine alga that occurs in open coasts and
tidal pools, but it can also be found under water in a depth of up to about 20 m. Its LHC is the siphonaxanthin-
siphonein-Chl-a/b-protein (SCP) that is optimized for using blue-green light, which is the dominating spectral
range under water. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports Scientific Reports | (2022) 12:8461 Preprocess dependence of optical
properties of ensembles and single
siphonaxanthin‑containing major
antenna from the marine green
alga Codium fragile
OPEN Tatas Hardo Panintingjati Brotosudarmo1,2, Bernd Wittmann1, Soichiro Seki3,
Ritsuko Fujii3,4 & Jürgen Köhler1,5,6* Tatas Hardo Panintingjati Brotosudarmo1,2, Bernd Wittmann1, Soichiro Seki3,
Ritsuko Fujii3,4 & Jürgen Köhler1,5,6* For SCP the Chl a
molecules in position 602, and either those in positions 610 or 612 are exchanged to Chl b as indicated by the
cyan colour. The Chl molecules are numbered according to the scheme given in9 (1RWT). The figure has been
produced using PyMOL ver. 2.5.0. (Sx + Sn):Neo:Chl a:Chl b = 2.5:1:6:8 was obtained by HPLC analysis for the pigment structure from resonance
Raman spectroscopy4. A similar Chl a:Chl b ratio was determined earlier also for the LHCs of PS II from the
siphonous green algae Bryopsis maxima and Bryopsis corticulans6,7. In summary, in SCP the two molecules of
Lut in LHC II are replaced by Sx/Sn, Vio is not present, and the Chl a to Chl b ratio is reversed with respect to
that in LHC II. Both the exchange of the carotenoids and the larger amount of Chl b molecules enhance the
light-harvesting efficiency of the complex in the blue-green spectral region around 450 nm, which reflects the
adaption of the species to the natural habitat. Since the ligands in the Chl binding pockets are the same in LHC
II and SCP7 the reversed Chl a to Chl b ratio hints for some structural flexibility of the SCP complex at these
positions. Based on recent cryoEM data10 the Chl a to Chl b exchange sites are assigned to the position 602, and
either to positions 610 or 612, see Fig. 1, according to the numbering scheme used in9. Usually, the details in the optical spectra from a macroscopic ensemble of proteins are washed out due to
ensemble averaging. In contrast, studying the complexes individually provides information about the distribu-
tions of parameters rather than only about their moments, which allows to identify subpopulations that would
be obscured otherwise. In the past, single-molecule spectroscopy has been applied extensively to the antennae
complexes of purple bacteria11–20 and to the light-harvesting systems associated with plant photosystems I and
II21–29 for obtaining information about the energetics and dynamics both at cryogenic temperatures as well as
under ambient conditions, recently excellently reviewed in30. At low temperatures thermal motions of the nuclei
are frozen out and the electronic transition energies that serve as input for structure-based modelling can be
detected with enhanced spectral resolution. Tatas Hardo Panintingjati Brotosudarmo1,2, Bernd Wittmann1, Soichiro Seki3,
Ritsuko Fujii3,4 & Jürgen Köhler1,5,6* The structure of this complex is still elusive but given the high sequence identity with the
major LHC II complex from green plants7 the idea is that the structure of SCP resembles that of LHC II, which
is known with atomic resolution from x-ray crystallography8,9. LHC II is a heterotrimer and each monomer
binds 8 Chl a, 6 Chl b, and 4 molecules of carotenoid (Car), namely two non-equivalent Luteins (Lut), one 9’-cis
neoxanthin (Neo), and one Violaxanthin (Vio) or Zeaxanthin (Zea)9. For SCP it is known that is forms as well a
trimeric complex that accommodates Chl a, Chl b, and that the two Lut molecules are replaced by siphonaxan-
thin (Sx) and its esterified variant siphonein (Sn), respectively, and 9’-cis neoxanthin (Neo). Recently, a ratio of 1Spectroscopy of Soft Matter, University of Bayreuth, 95440 Bayreuth, Germany. 2Department of Food
Technology, Universitas Ciputra, Citraland CBD Boulevard, Surabaya 60219, Indonesia. 3Graduate School of
Science, Osaka City University, 3‑3‑138 Sugimoto, Sumiyoshi‑ku, Osaka 558‑8585, Japan. 4Research Center for
Artificial Photosynthesis, Osaka Metropolitan University, 3‑3‑138 Sugimoto, Sumiyoshi‑ku, Osaka 558‑8585,
Japan. 5Bavarian Polymer Institute, University of Bayreuth, 95440 Bayreuth, Germany. 6Bayreuther Institut für
Makromolekülforschung (BIMF), 95440 Bayreuth, Germany. *email: juergen.koehler@uni-bayreuth.de | https://doi.org/10.1038/s41598-022-11572-3 Scientific Reports | (2022) 12:8461 www.nature.com/scientificreports/ Figure 1. High-resolution structure of a monomer of LHC II from spinach [PDB:1RWT]. (A, C) Side view of
the arrangement of the Chl molecules with (A) and without (C) the protein backbone. (B, D) Top view of the
arrangement of the Chl molecules with (B) and without (D) the protein backbone. Chl a molecules are shown
in green, Chl b molecules are show in blue. For clarity the carotenoids have been omitted. For SCP the Chl a
molecules in position 602, and either those in positions 610 or 612 are exchanged to Chl b as indicated by the
cyan colour. The Chl molecules are numbered according to the scheme given in9 (1RWT). The figure has been
produced using PyMOL ver. 2.5.0. Figure 1. High-resolution structure of a monomer of LHC II from spinach [PDB:1RWT]. (A, C) Side view of
the arrangement of the Chl molecules with (A) and without (C) the protein backbone. (B, D) Top view of the
arrangement of the Chl molecules with (B) and without (D) the protein backbone. Chl a molecules are shown
in green, Chl b molecules are show in blue. For clarity the carotenoids have been omitted. Results Using
τAv =
i Aiτi
i Ai , yields 3.6 ns for the amplitude averaged lifetime.f i Ai
In order to diminish bleaching effects, the experiments on single SCP complexes were conducted under
oxygen-free conditions in a vessel that was first flushed with gaseous Argon to remove residual air and then kept
under vacuum. In contrast to the experiments reported in29 that were conducted using an electrokinetic trap
ensuring a well-defined orientation of the pigment-protein complexes, in our experiments the three-dimensional
spatial orientation of the SCP complexes is not controlled. Hence, for each individual complex the projection of
the absorbing transition-dipole moment on the electric field vector of the excitation light is different leaving the
effective excitation intensity unknown. Because of this ambiguity, the emitted intensity is not a reliable parameter
for comparing different individual SCP complexes with each other. Instead, we recorded for each individual SCP
complex synchronously the fluorescence lifetime and the emission spectrum as a function of time. For doing so,
the spectra were read out consecutively every 5.1 s, and the lifetimes were read out consecutively every 500 ms. From these experiments the spectral peak positions of the emission spectra were extracted by fitting a Gaussian
to the main peak (see SI). For having the same statistics for the complexes that were studied under different
preparation conditions the observation time for each individual complex was restricted to 20 s. If not stated
otherwise the accuracy for the lifetimes is limited by the instrument response which amounts to 250 ps, and the
accuracy for the spectral peak positions amounts to 0.5 nm (11 cm−1).h y
p
p
p
(
)
The data extracted will be displayed in a two-dimensional "lifetimes versus spectral peak positions" diagram,
where the horizontal axis corresponds to the spectral peak position and the vertical axis to the fluorescence life-
time, see Fig. 2E. For obtaining these diagrams, periods of both constant peak position and constant lifetime con-
tribute one data point as pointed out in Fig. 2D. The example in Fig. 2D shows on the left-hand side the lifetimes
and spectral peak positions as a function of time that have been extracted from an experiment on a single SCP
complex. All fluorescence decays were compatible with monoexponentials. (For an explanation that addresses
the discrepancy between the monoexponential decays of the single complexes and the non-monoexponential
decays observed for the ensembles, see SI). Results Samples of unaggregated trimeric SCP complexes dissolved in buffer solution (concentration of 1.45 × 10–7 M)
have been prepared from fresh samples defrosted on ice under ambient conditions (protocol 1). The correspond-
ing room-temperature absorption spectrum is shown in Fig. 2A by the black line. It features strong peaks in the
blue spectral region at 439 nm and 473 nm, and two weaker bands in the red spectral region peaking at 652 nm
and 672 nm. The band at 473 nm can be attributed to both the carotenoids and the Chl b Soret band, whereas the
band at 439 nm is assigned to the absorptions of the Chl a/b Soret bands4,32. The maxima in the red spectral region
correspond to the absorptions of the Qy transitions of Chl b and Chl a, respectively. In order to avoid selective
excitation of specific Chl a/b sites in the SCP complexes we have chosen 561 nm as the excitation wavelength,
because this is outside the 650 nm spectral range of the excitonically coupled Chl molecules. For the experiments
on single SCP complexes these were immobilised in PVA, and the experiments were carried out at an excita-
tion intensity of 525 W/cm2 for obtaining a reasonable signal-to-noise ratio. For better comparison the same
excitation intensity has been used also for all experiments on ensembles of SCP complexes immobilised in PVA. y
p
p
In Fig. 2B the ensemble emission spectra from a solution of SCP complexes (full black line) and from SCP
complexes immobilized in a PVA matrix (dashed black line) are compared. Despite the large difference in the
excitation intensity used for recording the two spectra, i.e. 30 μW/cm2 for the SCP complexes in solution versus
525 W/cm2 for those in PVA, both spectra are identical in shape and feature a pronounced peak at 680.0 nm
(corresponding to 14,705 cm−1; FWHM 383 cm−1) accompanied by a low energy wing that extends from 700 to
760 nm. The comparison of the two spectra testifies that the spectral profile of the emission spectrum is neither
affected by the immobilization of the SCP complexes in PVA nor by the difference in excitation intensity. The
observed transient of an ensemble of SCP complexes embedded in PVA, Fig. 2C, is consistent with a biexponen-
tial decay featuring lifetime components of 4 ns, and 1.2 ns with amplitudes of 85%, and 15%, respectively. www.nature.com/scientificreports/ distinct, well-defined state in contrast to an ensemble of proteins due to the lack of synchronization31. To the best
of our knowledge results from optical spectroscopy on single SCP complexes have not been published to date. Here we report about spectroscopy on ensembles and single trimeric SCP complexes at room temperature for
three different protocols for handling the samples. The first protocol (protocol 1) refers to fresh samples prepared
on ice under ambient conditions, the second protocol (protocol 2) refers to SCP complexes from the same batch
that have been stored for 7 months at − 80 °C in the dark, and that are prepared under the same conditions as
before. The last protocol (protocol 3) refers to experiments where the samples from the 7 months old SCP com-
plexes have been prepared in a cold room at 5 °C. While the absorption and emission spectra from ensembles
of SCP complexes in solution featured a high degree of similarity for the three protocols, the single complex
approach revealed strong differences in the spectral details as a function of the sample preparation history. Tatas Hardo Panintingjati Brotosudarmo1,2, Bernd Wittmann1, Soichiro Seki3,
Ritsuko Fujii3,4 & Jürgen Köhler1,5,6* On the other hand, experiments under ambient conditions are closer
to the natural environment and provide valuable information about the conformational dynamics of the pigment-
protein complexes. In particular, a single protein that undergoes conformational transitions is at any time in a Scientific Reports | (2022) 12:8461 | https://doi.org/10.1038/s41598-022-11572-3 www.nature.com/scientificreports/ Results The excitation wavelength is 561 nm and
the excitation intensity amounts to 525 W/cm2. (D) Fluorescence lifetimes and spectral peak positions of a
single SCP complex. The fluorescence lifetime (scale on the right) was read out every 500 ms and is given by
the open dots. The drop out after 4 s presumably results from unresolved blinking. The bin time for recording
the spectra was 5.1 s (full line) and the observed spectral peak position of the accumulated spectrum is given
by the black dot (scale on the left). The sequence of emission spectra is shown on the right-hand side from top
to bottom. For better comparison, the broken vertical line refers to the spectral peak position of the ensemble
emission spectrum. (E) Two-dimensional representation of the fluorescence lifetime vs. spectral peak position
of individual SCP complexes. Due to fluctuations of both lifetime and/or spectral peak position a single complex
can contribute more than one data point to the diagram. The number of individual complexes that contribute
to the pattern is given in the top left corner. The histograms correspond to the distributions of the spectral
peak position (top) and fluorescence lifetime (right). For ease of comparison an ensemble emission spectrum
has been overlaid in the top histogram. The statistical parameters of the two histograms (mean ± sdev) are
(14,723 ± 109) cm−1 for the peak positions, and (3.4 ± 0.8) ns for the lifetimes, respectively. Figure 2. Spectroscopy on ensembles and single SCP complexes. The samples for spectroscopy have been
prepared under ambient conditions (protocol 1). (A) Room temperature absorption spectrum of an ensemble
of SCP complexes dissolved in bulk buffer solution at a concentration of 1.45 × 10–7 M. (B) Peak normalised
emission spectra from the bulk buffer solution for an excitation intensity of 30 μW/cm2 (full line) and from
an ensemble of SCP complexes embedded in a thin film of PVA (dashed line) for an excitation intensity of
525 W/cm2. The excitation wavelength for recording the emission spectra was 561 nm as indicated by the
green arrow in (A). (C) Fluorescence lifetime of an ensemble of SCP complexes immobilized in PVA in semi-
logarithmic representation. The full line corresponds to a biexponential fit (see residuals at the bottom) with
decay components (amplitudes) of 4.0 ns (85%) and 1.2 ns (15%). The excitation wavelength is 561 nm and
the excitation intensity amounts to 525 W/cm2. Results (D) Fluorescence lifetimes and spectral peak positions of a
single SCP complex. The fluorescence lifetime (scale on the right) was read out every 500 ms and is given by
the open dots. The drop out after 4 s presumably results from unresolved blinking. The bin time for recording
the spectra was 5.1 s (full line) and the observed spectral peak position of the accumulated spectrum is given
by the black dot (scale on the left). The sequence of emission spectra is shown on the right-hand side from top
to bottom. For better comparison, the broken vertical line refers to the spectral peak position of the ensemble
emission spectrum. (E) Two-dimensional representation of the fluorescence lifetime vs. spectral peak position
of individual SCP complexes. Due to fluctuations of both lifetime and/or spectral peak position a single complex
can contribute more than one data point to the diagram. The number of individual complexes that contribute
to the pattern is given in the top left corner. The histograms correspond to the distributions of the spectral
peak position (top) and fluorescence lifetime (right). For ease of comparison an ensemble emission spectrum
has been overlaid in the top histogram. The statistical parameters of the two histograms (mean ± sdev) are
(14,723 ± 109) cm−1 for the peak positions, and (3.4 ± 0.8) ns for the lifetimes, respectively. standard deviation of 109 cm−1. The average peak position is in close agreement with the 14,705 cm−1 that was
found for the peak of the ensemble spectrum. The distribution of the fluorescence lifetimes is relatively broad,
and the statistical parameters amount to 3.4 ns ± 0.8 ns (mean ± sdev), also in good agreement with the average
lifetime of 3.6 ns obtained for a large ensemble of SCP complexes. standard deviation of 109 cm−1. The average peak position is in close agreement with the 14,705 cm−1 that was
found for the peak of the ensemble spectrum. The distribution of the fluorescence lifetimes is relatively broad,
and the statistical parameters amount to 3.4 ns ± 0.8 ns (mean ± sdev), also in good agreement with the average
lifetime of 3.6 ns obtained for a large ensemble of SCP complexes. g
Similar experiments have been conducted on SCP complexes from the same batch that has been stored for
7 months at − 80 °C in the dark, and that were prepared under the same conditions as before (protocol 2). Results For the first 4 s the fluorescence lifetime amounts to about 4.5 ns
before it drops to about 3.4 ns for the remaining observation time. The sharp minimum at about 4 s is attributed
to blinking processes that are faster than the 5.1 s bin time for reading out the spectra. For this particular SCP
complex the first value is slightly longer than the 3.6 ns that have been found for the average ensemble lifetime. On the right-hand side of Fig. 2D the emission spectra that have been accumulated during the four time intervals
are shown. For all spectra the spectral peak positions of 14,684 cm−1 are close to the ensemble value and show
only little variation as a function of time. For this particular complex we find in total one peak position and
two lifetimes, and it therefore contributes two data points to the aforementioned two-dimensional lifetime vs. spectral peak position diagram in Fig. 2E. The two-dimensional diagram with contributions from 31 individual
complexes is shown in Fig. 2E, together with the distributions of the spectral peak positions and the fluorescence
lifetimes on top of and next to the pattern. The spectral peak positions observed for the individual complexes
show only little variation and the corresponding distribution is characterized by a mean of 14,723 cm−1 and a https://doi.org/10.1038/s41598-022-11572-3 Scientific Reports | (2022) 12:8461 | www.nature.com/scientificreports/ t
d
d d
i ti
f 109
−1 Th
k
iti
i i
l
t
ith th 14 705
−1 th t
Figure 2. Spectroscopy on ensembles and single SCP complexes. The samples for spectroscopy have been
prepared under ambient conditions (protocol 1). (A) Room temperature absorption spectrum of an ensemble
of SCP complexes dissolved in bulk buffer solution at a concentration of 1.45 × 10–7 M. (B) Peak normalised
emission spectra from the bulk buffer solution for an excitation intensity of 30 μW/cm2 (full line) and from
an ensemble of SCP complexes embedded in a thin film of PVA (dashed line) for an excitation intensity of
525 W/cm2. The excitation wavelength for recording the emission spectra was 561 nm as indicated by the
green arrow in (A). (C) Fluorescence lifetime of an ensemble of SCP complexes immobilized in PVA in semi-
logarithmic representation. The full line corresponds to a biexponential fit (see residuals at the bottom) with
decay components (amplitudes) of 4.0 ns (85%) and 1.2 ns (15%). Results The sequence of emission spectra is shown on the right-hand side from top
to bottom. For better comparison, the broken vertical line refers to the spectral peak position of the ensemble
emission spectrum. (E) Two-dimensional representation of the fluorescence lifetime vs. spectral peak position
of individual SCP complexes. Due to fluctuations of both lifetime and/or spectral peak position a single complex
can contribute more than one data point to the diagram. The number of individual complexes that contribute
to the pattern is given in the top left corner. The histograms correspond to the distributions of the spectral peak
position (top) and fluorescence lifetime (right). For ease of comparison an ensemble emission spectrum from
the fresh sample embedded in PVA has been overlaid in the top histogram. The statistical parameters of the
two histograms (mean ± sdev) are (14,319 ± 178) cm−1 for the peak positions, and (2.0 ± 0.5) ns for the lifetimes,
respectively. Figure 3. Spectroscopy on ensembles and single SCP complexes that have been stored for 7 months at − 80 °C
in the dark. The samples for spectroscopy have been prepared under ambient conditions (protocol 2). (A)
Room temperature absorption spectrum of an ensemble of SCP that has been stored for 7 months at − 80 °C in
the dark dissolved in bulk buffer solution at a concentration of 1.45 × 10–7 M (black line). For comparison the
corresponding absorption spectrum from the fresh sample (cf. Fig. 2A) has been underlaid as filled spectrum. Both spectra have been peak normalised at 476 nm. (B) Emission spectrum from SCP complexes embedded
in PVA for an excitation intensity of 525 W/cm2. The filled area shows the emission spectrum from a sample of
fresh SCP complexes embedded in PVA. Both spectra have been peak normalised for better comparison. The
excitation wavelength for recording the emission spectrum was 561 nm as indicated by the green arrow in (A). (C) Fluorescence lifetime of an ensemble of the 7 months old SCP complexes immobilized in PVA in semi-
logarithmic representation. The full line corresponds to a biexponential fit (see residuals at the bottom) with
decay components (amplitudes) of 4.3 ns (33%) and 1.2 ns (67%). The excitation wavelength is 561 nm, and the
excitation intensity amounts to 525 W/cm2. (D) Fluorescence lifetimes and spectral peak positions of a single
SCP complex. Results The
room-temperature absorption spectrum from a solution of these complexes is shown in Fig. 3A by the black
line. For reference the absorption spectrum obtained from the fresh sample is underlaid as a filled spectrum. Apart from very small changes below 400 nm the absorption spectra obtained for the different protocols are
identical. In Fig. 3B the full line corresponds to the emission spectrum from an ensemble of these SCP complexes https://doi.org/10.1038/s41598-022-11572-3 Scientific Reports | (2022) 12:8461 | www.nature.com/scientificreports/ cientificreports/
embedded in PVA, which is compared with the emission spectrum obtained from the fresh sample that is again i
Figure 3. Spectroscopy on ensembles and single SCP complexes that have been stored for 7 months at − 80 °C
in the dark. The samples for spectroscopy have been prepared under ambient conditions (protocol 2). (A)
Room temperature absorption spectrum of an ensemble of SCP that has been stored for 7 months at − 80 °C in
the dark dissolved in bulk buffer solution at a concentration of 1.45 × 10–7 M (black line). For comparison the
corresponding absorption spectrum from the fresh sample (cf. Fig. 2A) has been underlaid as filled spectrum. Both spectra have been peak normalised at 476 nm. (B) Emission spectrum from SCP complexes embedded
in PVA for an excitation intensity of 525 W/cm2. The filled area shows the emission spectrum from a sample of
fresh SCP complexes embedded in PVA. Both spectra have been peak normalised for better comparison. The
excitation wavelength for recording the emission spectrum was 561 nm as indicated by the green arrow in (A). (C) Fluorescence lifetime of an ensemble of the 7 months old SCP complexes immobilized in PVA in semi-
logarithmic representation. The full line corresponds to a biexponential fit (see residuals at the bottom) with
decay components (amplitudes) of 4.3 ns (33%) and 1.2 ns (67%). The excitation wavelength is 561 nm, and the
excitation intensity amounts to 525 W/cm2. (D) Fluorescence lifetimes and spectral peak positions of a single
SCP complex. The fluorescence lifetime (scale on the right) was read out every 500 ms and is given by the open
dots. The dropouts after 2.5 s and 14 s presumably result from unresolved blinking. The bin time for recording
the spectra was 5.1 s (full line) and the observed spectral peak position of the accumulated spectrum is given
by the black dot (scale on the left). www.nature.com/scientificreports/ The room temperature
absorption spectrum from a solution, Fig. 4A, resembles closely the one obtained from the fresh sample, also
for the spectral range below 400 nm Fig. 4A. Interestingly, this holds true as well for the emission spectrum
that features a peak position at 679.6 nm (14,715 cm−1) and a width (FWHM) of 395 cm−1 Fig. 4B, and which
reproduces the spectral profile observed for the fresh sample. Similarly for the fluorescence decay with lifetime
components (amplitudes) of 4.0 ns (79%), and 1.3 ns (21%). The relative weight of the two decay components is
nearly restored to the values observed for the fresh SCP complexes. Accordingly, the averaged lifetime amounts
to 3.4 ns. The fluorescence lifetime from the single complex shown in Fig. 4D amounts to 3.2 ns for the first
9 s and then drops to 2.0 ns for the remaining duration of the experiment. During this time the spectral peak
position does not show significant variations and is centred around 14,735 cm−1 corresponding to a slight blue
shift with respect to the ensemble value. As before such a complex contributes 2 data points to the respective
two-dimensional diagram shown in Fig. 4E.h g
g
The statistical parameters for the full dataset with contributions from 26 individual SCP complexes are
14,537 cm−1 ± 181 cm−1 for the peak positions and 2.8 ns ± 0.9 ns for the fluorescence lifetimes. However, closer
inspection of the two-dimensional representation of the data suggests that this diagram represents a superposi-
tion of the two patterns observed before for the fresh SCP complexes and for those that have been stored for
7 months in the cold, cf. Figs. 2 and 3. This is corroborated by the separate evaluation of the statistics of the two
clusters of data points that are indicated by the blue and red-shaded areas in Fig. 4E. For convenience these will
be referred to as the blue and the red cluster in the following. For the data in the blue cluster, the distribution
of the peak position is characterised by 14,715 cm−1 ± 60 cm−1 and that for the lifetimes by 3.4 ns ± 0.9 ns. These
numbers are both in very close agreement with the parameters found for the fresh sample. The corresponding
parameters for the red cluster amount to 14,382 cm−1 ± 77 cm−1 for the peak positions and 2.1 ns ± 0.4 ns for the
lifetimes. www.nature.com/scientificreports/ decay times are about the same as those observed before, their relative weight has shifted towards the short-lived
component which is reflected in an averaged lifetime of 2.2 ns. A tentative explanation for this observation is to
consider two prevailing conformations of the complexes, each associated with one of the lifetime components. Variations of the population ratio of these conformations will directly impact on the corresponding amplitudes
of the two decay components, yet without affecting the time constants.hi f
An example for a single complex is shown in Fig. 3C. The setup of the figure is similar to the setup of Fig. 2C. For this complex the fluorescence lifetime amounts to 1.6 ns during the whole observation time, which is even
shorter than the average lifetime of 2.2 ns found for an ensemble under the same conditions. The spectral peak
position shows a significant shift from about 14,430 cm−1 during the first 10 s to about 14,120 cm−1 during the
last 20 s, which both are red shifted with respect to the ensemble value. Hence, for this complex we find two peak
positions and one lifetime, and this complex contributes two data points to the corresponding two-dimensional
lifetime vs. peak position diagram shown in Fig. 3E. For the SCP complexes that have been prepared accord-
ing to protocol 2 the "peak position versus lifetime" diagram, and the resulting histograms with contributions
from 31 individual complexes are shown in Fig. 3E. The distribution of the spectral peak positions does not
show a pronounced maximum but is red shifted and clearly broader than the histogram obtained from the fresh
complexes. This is also reflected by the statistical parameters of 14,319 cm−1 ± 178 cm−1. Interestingly, now the
histogram for the fluorescence lifetimes is relatively narrow and shifted towards shorter decay times. The statisti-
cal parameters for this distribution amount to 2.0 ns ± 0.5 ns, which is in the range of the ensemble average for
this preparation protocol. Finally, experiments have been conducted on the 7 months old SCP complexes where all preparation steps
including spin coating as well as flushing and evacuation of the sample chamber took place in a cold room
at 5 °C (protocol 3). The experiments were carried out immediately after the preparation of the sample. The
experimental results are summarized in Fig. 4 which has a similar layout as Figs. 2 and 3. Results The fluorescence lifetime (scale on the right) was read out every 500 ms and is given by the open
dots. The dropouts after 2.5 s and 14 s presumably result from unresolved blinking. The bin time for recording
the spectra was 5.1 s (full line) and the observed spectral peak position of the accumulated spectrum is given
by the black dot (scale on the left). The sequence of emission spectra is shown on the right-hand side from top
to bottom. For better comparison, the broken vertical line refers to the spectral peak position of the ensemble
emission spectrum. (E) Two-dimensional representation of the fluorescence lifetime vs. spectral peak position
of individual SCP complexes. Due to fluctuations of both lifetime and/or spectral peak position a single complex
can contribute more than one data point to the diagram. The number of individual complexes that contribute
to the pattern is given in the top left corner. The histograms correspond to the distributions of the spectral peak
position (top) and fluorescence lifetime (right). For ease of comparison an ensemble emission spectrum from
the fresh sample embedded in PVA has been overlaid in the top histogram. The statistical parameters of the
two histograms (mean ± sdev) are (14,319 ± 178) cm−1 for the peak positions, and (2.0 ± 0.5) ns for the lifetimes,
respectively. embedded in PVA, which is compared with the emission spectrum obtained from the fresh sample that is again
underlaid as a filled spectrum. In contrast, to the absorption spectrum the ensemble emission spectrum from
the 7 months old sample is red shifted with respect to the initial spectrum and peaks at 686.6 nm (14,564 cm−1). Moreover, its width (FWHM) is increased to 440 cm−1, and the relative intensity of the low-energy shoulder in
the 710 nm to 760 nm spectral range has risen significantly. The decay of the fluorescence from the ensemble is
again consistent with a biexponential with lifetimes (amplitudes) of 4.3 ns (33%), and 1.2 ns (67%). While the https://doi.org/10.1038/s41598-022-11572-3 Scientific Reports | (2022) 12:8461 | www.nature.com/scientificreports/ www.nature.com/scientificreports/ The number of individual complexes that contribute to the pattern is given in the top left corner. The histograms correspond to the distributions of the spectral peak position (top) and fluorescence lifetime
(right). For ease of comparison an emission spectrum from a sample of fresh SCP complexes embedded in
PVA has been overlaid in the top histogram. The statistical parameters of the two histograms (mean ± sdev) are
(14,537 ± 181) cm−1 for the peak positions, and (2.8 ± 0.9) ns for the lifetimes, respectively. Figure 4. Spectroscopy on ensembles and single SCP complexes that have been stored for 7 months at − 80 °C
in the dark. The samples for spectroscopy have been prepared in a cold room at 5 °C (protocol 3). (A) Room
temperature absorption spectrum of an ensemble of SCP dissolved in bulk buffer solution at a concentration
of 1.45 × 10–7 M (black line). For comparison the corresponding absorption spectrum from the fresh sample
(cf. Fig. 2A) has been underlaid as filled spectrum. Both spectra have been peak normalised at 476 nm. (B)
Emission spectrum from SCP complexes embedded in PVA for an excitation intensity of 525 W/cm2. The filled
area shows the emission spectrum from a sample of fresh SCP complexes embedded in PVA. Both spectra have
been peak normalised for better comparison. The excitation wavelength for recording the emission spectrum
was 561 nm as indicated by the green arrow in (A). (C) Fluorescence lifetime of an ensemble of SCP complexes
immobilized in PVA in semi-logarithmic representation. The full line corresponds to a biexponential fit (see
residuals at the bottom) with decay components (amplitudes) of 4.0 ns (79%) and 1.3 ns (21%). The excitation
wavelength is 561 nm, and the excitation intensity amounts to 525 W/cm2. (D) Fluorescence lifetimes and
spectral peak positions of a single SCP complex. The fluorescence lifetime (scale on the right) was read out
every 500 ms and is given by the open dots. The bin time for recording the spectra was 5.1 s (full line) and
the observed spectral peak position of the accumulated spectrum is given by the black dot (scale on the left). The sequence of emission spectra is shown on the right-hand side from top to bottom. For better comparison,
the broken vertical line refers to the spectral peak position of the ensemble emission spectrum. (E) Two-
dimensional representation of the fluorescence lifetime vs. www.nature.com/scientificreports/ The mean for the lifetimes is close to the value found for SCP complexes that were prepared according
to protocol 2, whereas the mean for the spectral peak positions is more shifted to the red. However, the most
intriguing observation in the experiments on these SCP complexes was that the classification whether a result
felt into the blue or red cluster correlated with the duration of the total experiment. Starting the experiment on
a sample immediately after the preparation in the cold room yielded spectral peak positions and fluorescence
lifetimes that belonged to the blue cluster, whereas after about 120 min duration of the experiments the observed
combination of spectral peak positions and fluorescence lifetimes felt into the red cluster. Jumps of individual
complexes between the two clusters were not observed. The experimental results for the three protocols are
summarized in Table 1.h The experiments on single SCP complexes were carried out using an excitation intensity that exceeds by far
the (spectrally integrated) intensity of less than 100 μW/cm2 from the sun on a bright day in the natural habitat
of the algae. In order to test whether the observed changes of the spectral signatures correlate with the excitation
intensity it would be desirable to repeat the experiments using a lower excitation intensity. Due to signal-to-noise
restrictions this, however, is only possible for ensembles of SCP complexes. In Fig. 5 the ensemble emission spec-
tra for the various protocols are compared as a function of the excitation intensity. At high excitation intensity,
Fig. 5A, the spectrum recorded according to protocol 3 initially reproduces the "protocol 1 like" spectrum (dashed
black line), yet changes into the "protocol 2 like" spectrum (red dashed line) after a waiting time (without illu-
mination) of about 2 h. These observations are consistent with the results obtained on single SCP complexes for
the different protocols. At lower excitation intensities, Fig. 5B, the emission spectra for all protocols are identical,
and moreover reproduce the emission spectrum obtained for protocol 1 at high excitation intensity. This suggests
that the formation of the red-shifted quenched state correlates with the excitation intensity. https://doi.org/10.1038/s41598-022-11572-3 Scientific Reports | (2022) 12:8461 | www.nature.com/scientificreports/ Discussion
We have detected the fluorescence lifetimes and the emission spectra from ensembles and single SCP complexes
Figure 4. Spectroscopy on ensembles and single SCP complexes that have been stored for 7 months at − 80 °C
in the dark. www.nature.com/scientificreports/ The samples for spectroscopy have been prepared in a cold room at 5 °C (protocol 3). (A) Room
temperature absorption spectrum of an ensemble of SCP dissolved in bulk buffer solution at a concentration
of 1.45 × 10–7 M (black line). For comparison the corresponding absorption spectrum from the fresh sample
(cf. Fig. 2A) has been underlaid as filled spectrum. Both spectra have been peak normalised at 476 nm. (B)
Emission spectrum from SCP complexes embedded in PVA for an excitation intensity of 525 W/cm2. The filled
area shows the emission spectrum from a sample of fresh SCP complexes embedded in PVA. Both spectra have
been peak normalised for better comparison. The excitation wavelength for recording the emission spectrum
was 561 nm as indicated by the green arrow in (A). (C) Fluorescence lifetime of an ensemble of SCP complexes
immobilized in PVA in semi-logarithmic representation. The full line corresponds to a biexponential fit (see
residuals at the bottom) with decay components (amplitudes) of 4.0 ns (79%) and 1.3 ns (21%). The excitation
wavelength is 561 nm, and the excitation intensity amounts to 525 W/cm2. (D) Fluorescence lifetimes and
spectral peak positions of a single SCP complex. The fluorescence lifetime (scale on the right) was read out
every 500 ms and is given by the open dots. The bin time for recording the spectra was 5.1 s (full line) and
the observed spectral peak position of the accumulated spectrum is given by the black dot (scale on the left). The sequence of emission spectra is shown on the right-hand side from top to bottom. For better comparison,
the broken vertical line refers to the spectral peak position of the ensemble emission spectrum. (E) Two-
dimensional representation of the fluorescence lifetime vs. spectral peak position of individual SCP complexes. Due to fluctuations of both lifetime and/or spectral peak position a single complex can contribute more than
one data point to the diagram. For the distinction between the data points in the red and blue shaded areas
see text. The number of individual complexes that contribute to the pattern is given in the top left corner. The histograms correspond to the distributions of the spectral peak position (top) and fluorescence lifetime
(right). For ease of comparison an emission spectrum from a sample of fresh SCP complexes embedded in
PVA has been overlaid in the top histogram. www.nature.com/scientificreports/ spectral peak position of individual SCP complexes. Due to fluctuations of both lifetime and/or spectral peak position a single complex can contribute more than
one data point to the diagram. For the distinction between the data points in the red and blue shaded areas
see text. The number of individual complexes that contribute to the pattern is given in the top left corner. The histograms correspond to the distributions of the spectral peak position (top) and fluorescence lifetime
(right). For ease of comparison an emission spectrum from a sample of fresh SCP complexes embedded in
PVA has been overlaid in the top histogram. The statistical parameters of the two histograms (mean ± sdev) are
(14,537 ± 181) cm−1 for the peak positions, and (2.8 ± 0.9) ns for the lifetimes, respectively. www.nature.com/scientificreports/ The statistical parameters of the two histograms (mean ± sdev) are
(14,537 ± 181) cm−1 for the peak positions, and (2.8 ± 0.9) ns for the lifetimes, respectively. Figure 4. Spectroscopy on ensembles and single SCP complexes that have been stored for 7 months at − 80 °C
in the dark. The samples for spectroscopy have been prepared in a cold room at 5 °C (protocol 3). (A) Room
temperature absorption spectrum of an ensemble of SCP dissolved in bulk buffer solution at a concentration
of 1.45 × 10–7 M (black line). For comparison the corresponding absorption spectrum from the fresh sample
(cf. Fig. 2A) has been underlaid as filled spectrum. Both spectra have been peak normalised at 476 nm. (B)
Emission spectrum from SCP complexes embedded in PVA for an excitation intensity of 525 W/cm2. The filled
area shows the emission spectrum from a sample of fresh SCP complexes embedded in PVA. Both spectra have
been peak normalised for better comparison. The excitation wavelength for recording the emission spectrum
was 561 nm as indicated by the green arrow in (A). (C) Fluorescence lifetime of an ensemble of SCP complexes
immobilized in PVA in semi-logarithmic representation. The full line corresponds to a biexponential fit (see
residuals at the bottom) with decay components (amplitudes) of 4.0 ns (79%) and 1.3 ns (21%). The excitation
wavelength is 561 nm, and the excitation intensity amounts to 525 W/cm2. (D) Fluorescence lifetimes and
spectral peak positions of a single SCP complex. The fluorescence lifetime (scale on the right) was read out
every 500 ms and is given by the open dots. The bin time for recording the spectra was 5.1 s (full line) and
the observed spectral peak position of the accumulated spectrum is given by the black dot (scale on the left). The sequence of emission spectra is shown on the right-hand side from top to bottom. For better comparison,
the broken vertical line refers to the spectral peak position of the ensemble emission spectrum. (E) Two-
dimensional representation of the fluorescence lifetime vs. spectral peak position of individual SCP complexes. Due to fluctuations of both lifetime and/or spectral peak position a single complex can contribute more than
one data point to the diagram. For the distinction between the data points in the red and blue shaded areas
see text. www.nature.com/scientificreports/ www.nature.com/scientificreports/ www.nature.com/scientificreports/ Table 1. Summary of the spectroscopic results from SCP complexes as a function of the sample preparation
conditions. For the ensemble data the spectral peak positions and the widths (FWHM) of the emission spectra
are provided. For the single complex data, the means and standard deviations of the respective distributions are
given. The excitation intensity at 561 nm was always 525 W/cm2. Ensembles
Protocol 1
Protocol 2
Protocol 3
Spectral peak position (cm−1)
14,705 cm−1
14,564 cm−1
14,715 cm−1
Spectral peak position (nm)
680.0
686.6
679.6
FWHM (cm−1)
383
440
395
Lifetimes (ns) (amplitudes)
4.0 (85%)
1.2 (15%)
4.3 (33%)
1.2 (67%)
4.0 (79%)
1.3 (21%)
Averaged lifetime (ns)
3.6
2.2
3.4
Single complexes
Protocol 3 blue cluster
Protocol 3 red cluster
Spectral peak position mean ± sdev (cm−1)
14,723 ± 109
14,319 ± 178
14,537 ± 180
14,715 ± 60
14,382 ± 77
Mean of spectral peak position (nm)
679.2
698.4
687.9
679.6
695.3
Lifetime mean ± sdev (ns)
3.4 ± 0.8
2.0 ± 0.5
2.8 ± 0.9
3.4 ± 0.9
2.1 ± 0.4 Table 1. Summary of the spectroscopic results from SCP complexes as a function of the sample preparation
conditions. For the ensemble data the spectral peak positions and the widths (FWHM) of the emission spectra
are provided. For the single complex data, the means and standard deviations of the respective distributions are
given. The excitation intensity at 561 nm was always 525 W/cm2. Table 1. Summary of the spectroscopic results from SCP complexes as a function of the sample preparation
conditions. For the ensemble data the spectral peak positions and the widths (FWHM) of the emission spectra
are provided. For the single complex data, the means and standard deviations of the respective distributions are
given. The excitation intensity at 561 nm was always 525 W/cm2. Figure 5. Emission spectra of ensembles of SCP complexes (A) embedded in PVA for excitation intensities
of 525 W/cm2, and (B) dissolved in bulk buffer solution at a concentration of 1.45 × 10–7 M, and excitation
intensities of 30 µW/cm2. Key: Dashed black lines—emission spectra for protocol 3; dashed red lines—emission
spectra for protocol 3 after a waiting time of 2 h. For ease of comparison, the emission spectra for protocols 1
and 2 are shown by the underlaid filled areas, and the full black lines, respectively. Figure 5. Discussion We have detected the fluorescence lifetimes and the emission spectra from ensembles and single SCP complexes
using three different protocols for handling the samples. Taking the data from the fresh sample (protocol 1) as a
reference, protocol 2 yields a red shifted ensemble emission spectrum featuring a pronounced low energy shoul-
der, and a shortening of the average fluorescence lifetime, whereas these parameters are nearly unchanged for
applying protocol 3. Since the results obtained from the fresh samples are reproduced for protocol 3, we conclude
that the SCP complexes are still intact immediately after taking them out of the low temperature storage. One https://doi.org/10.1038/s41598-022-11572-3 Scientific Reports | (2022) 12:8461 | www.nature.com/scientificreports/ Moreover, for protocol 3 the single SCP
complexes undergo a change in the course of time from "protocol 1 like" results during the first two hours after
sample preparation to "protocol 2 like" results later.l For LHC II from green plants fluorescence lifetimes between 1.2–1.4 ns and 3.4–3.8 ns with relative contribu-
tions between 10–20% for the fast decay time and 80–90% for the slower one, respectively, have been observed
before40. The fast component is in very good agreement with the fast component found for SCP, whereas the
other component for LHC II is slightly faster with respect to the corresponding lifetime component of SCP. The
relative weight of the two contributions, however, closely resembles the partition found for fresh samples of SCP,
see results for protocol 1. A reduction of the fluorescence lifetime from about 4 ns to 2 ns has been reported for
LHC II complexes in thylakoid membranes41 which, however, was not accompanied by a red shift of the emission.t p
y
p
yt
For single LHC II trimers a reversible switching of the emission between a red-shifted state above 700 nm and
a state at 683 nm has been observed before26,28,34. Unfortunately, information about the fluorescence lifetimes of
these states is not available. The red-shifted emission was associated with a mixed exciton-charge transfer (CT)
state that involves the sites Chl a603, and Chl b609 in close proximity to the lutein 2 molecule, and the appear-
ance/disappearance of this band was ascribed to a conformational change within the protein that affects the
mutual interactions of these chromophores28,34. Owing to the high degree of homology of the protein structures
of SCP and LHC II we hypothesize that in SCP the red-shifted quenched state can be associated as well with
structural changes within the protein scaffold that induce the formation of the mixed exciton-charge transfer
(CT) state. Given the relatively low fluorescence quantum yield of SCP of 5–10% a large fraction of the average
absorbed energy is dissipated by radiationless decay providing a sufficient amount of energy for inducing con-
formational fluctuations of the protein backbone. This is consistent with the occurrence of the red-emitting state
only for higher excitation intensities, see Fig. 5. www.nature.com/scientificreports/ Emission spectra of ensembles of SCP complexes (A) embedded in PVA for excitation intensities
of 525 W/cm2, and (B) dissolved in bulk buffer solution at a concentration of 1.45 × 10–7 M, and excitation
intensities of 30 µW/cm2. Key: Dashed black lines—emission spectra for protocol 3; dashed red lines—emission
spectra for protocol 3 after a waiting time of 2 h. For ease of comparison, the emission spectra for protocols 1
and 2 are shown by the underlaid filled areas, and the full black lines, respectively. might argue that the relatively high excitation intensity of 525 W/cm2 causes some damage to the SCP complexes. However, we note that at 561 nm this intensity produces about the same number of excited states per unit time
as an intensity of 250 W/cm2 at 640 nm, which is a common intensity used for spectroscopy of single LHC com-
plexes from green plants26,27,33,34. Since such a red-shifted emission has been observed before for LHC II from
green plants34, and because we do not observe any spectral features that are indicative of free Chlorophyll, namely
a fluorescence lifetime of about 5–8 ns35–37 and an emission band around 675 nm38, we exclude that the emission
from the red-shifted quenched state is caused by a deterioration of the samples. Moreover, changes in the PVA
matrix causing this shift are excluded as well. Firstly, for all three sample preparation scenarios the PVA was added
during the last step of the procedure. Hence, any presumed influence of the PVA on the outcome of the experi-
ments should be identical for all three scenarios. Secondly, in previous work on single light-harvesting complexes
from purple bacteria it was shown that the experimental results obtained from complexes that were embedded
in PVA and the results from complexes that were reconstituted into a phospholipid bilayer were equivalent16,39.l For all three protocols the fluorescence transients recorded from ensembles of SCP complexes are consistent
with biexponentials with decay times that are identical within experimental accuracy, namely around 4 ns and
1.2 ns. The decrease of the averaged lifetime from 3.6 ns to 2.2 ns for protocol 2 results solely from the variation of
the relative contributions of the two lifetime components to the total decay. In contrast, the experiments on single
SCP complexes reveal that all fluorescence decays of single complexes were compatible with monoexponentials. https://doi.org/10.1038/s41598-022-11572-3 Scientific Reports | (2022) 12:8461 | www.nature.com/scientificreports/ Figure 6. www.nature.com/scientificreports/ Schematic sketch of the highest tier of the protein energy landscape of the SCP protein structure. For the fresh samples (blue line) the states emitting around 680 nm and around 690–700 nm are separated by a
high energy barrier. For the 7 months old samples (red line) the barrier height is lowered such that the available
thermal energy at room temperature is sufficient to overcome this threshold. Figure 6. Schematic sketch of the highest tier of the protein energy landscape of the SCP protein structure. For the fresh samples (blue line) the states emitting around 680 nm and around 690–700 nm are separated by
high energy barrier. For the 7 months old samples (red line) the barrier height is lowered such that the availa
thermal energy at room temperature is sufficient to overcome this threshold. Figure 6. Schematic sketch of the highest tier of the protein energy landscape of the SCP protein structure. For the fresh samples (blue line) the states emitting around 680 nm and around 690–700 nm are separated by a
high energy barrier. For the 7 months old samples (red line) the barrier height is lowered such that the available
thermal energy at room temperature is sufficient to overcome this threshold. Figure 6. Schematic sketch of the highest tier of the protein energy landscape of the SCP protein structure. For the fresh samples (blue line) the states emitting around 680 nm and around 690–700 nm are separated by a
high energy barrier. For the 7 months old samples (red line) the barrier height is lowered such that the available
thermal energy at room temperature is sufficient to overcome this threshold. However, for a single complex the decay time may fluctuate as a function of time, and for the micro ensemble
of individual complexes studied the decay times feature a distribution. Moreover, for protocol 3 the single SCP
complexes undergo a change in the course of time from "protocol 1 like" results during the first two hours after
sample preparation to "protocol 2 like" results later.l However, for a single complex the decay time may fluctuate as a function of time, and for the micro ensemble
of individual complexes studied the decay times feature a distribution. Materials and methods
Cultivation of algae.
Cod Subsequently a small droplet (25 μL) of the solution was spin-coated in the dark onto a glass
substrate (for 10 s at 500 rpm and 60 s at 2000 rpm) forming a thin polymer film of about 100 nm thickness
with embedded SCP complexes. Prior to the spin coating the glass substrate was three times carefully cleaned
with acetone and dried with a strong flow of nitrogen gas. The substrate served as window of a vacuum chamber
with the sample positioned on the inside which allows to work under oxygen-free conditions. Immediately after
mounting the window the vacuum chamber was flushed 3 times with gaseous Argon to remove residual air and
then evacuated to 10–3 mbar. Protocol 2 Protocol 2 refers to SCP samples that have been stored at − 80 °C 7–8 months, and that were pre-
pared for the spectroscopic experiments following the same procedures as detailed above for protocol 1. g
Protocol 3 Protocol 3 refers to SCP samples that have been stored at − 80 °C 7–8 months, but where the whole
preparation of the samples for the spectroscopic experiments including spin coating, and mounting of the
samples was carried out in a cold room (5 °C) in the dark. After mounting the sample chamber was evacuated,
moved to the optical setup, flushed with gaseous Argon and then evacuated again. Optical experiments. For recording the UV/VIS spectra the dissolved SCP complexes at a concentration
of 1.45 × 10–7 M were filled into quartz-glass cuvettes (Hellma QS) that was mounted in commercial spectrome-
ters (absorption: Perkin Elmer Lambda75, emission: Varian Cary Eclipse spectrometer). The emission spectrum
was recorded for an excitation wavelength of 561 nm and an intensity that corresponded to 30 µW/cm2. The
single-complex experiments were conducted on a commercial single-molecule optical microscope (MicroTime
200, PicoQuant). The SCP complexes were excited with a laser diode at 561 nm (LHD-D-TA-560B, PicoQuant)
operated in pulsed mode with a repetition rate of 20 MHz and an intensity of 525 W/cm2. The output from the
laser was coupled into an optical fibre and entered an inverted confocal optical microscope where it was focused
with a water immersion objective (60 × UPlanS APO UIS2, NA = 1.2, Olympus) onto the sample. Materials and methods
Cultivation of algae.
Cod Cultivation of algae. Codium fragile (KU-065, KU-MACC, Kobe, Japan) was cultivated in floating form
in Provasoli’s Enriched Sea water (PES) medium as described in42 using artificial sea water (Marine Art SF-1,
Tomita Pharmaceutical Co., Ltd, Tokushima, Japan) at 21 °C, and illuminated with a white light LED (AS-010,
Fujikura, Osaka, Japan) at 0.5 mW/cm2 with 12 h/12 h light cycles. Protein preparation/purification. Thylakoid membranes were prepared from lyophilized algae accord-
ing to45. SCP was purified from the thylakoid membranes as detailed in4 with slight modifications: After solu-
bilization, the samples were loaded onto the layer of sucrose density gradient as described in46. The major band
was collected and purified by anion exchange chromatography, see4. This step was repeated, and the material was
further purified using another sucrose density gradient followed by gel filtration (Sephacryl S-200HR, Merck
KGaA, Darmstadt, Germany) with 20 mM Tris–HCl (pH 8.2) containing 0.03% n-dodecyl β-D-maltoside
(β-DDM). Purified SCP was flash frozen with liquid nitrogen, stored at − 80 °C, and transported at liquid nitro-
gen temperature using Dry Shipper (CX100, Taylor-Wharton, Baytown, Tx, USA). Sample handling. The SCP samples with an optical density of 0.2 at 670 nm were defrosted on an ice
bucket and diluted in 25 mM Tris–HCl buffer pH 8.2 containing 0.03% n-dodecyl β-D-maltoside (β-DDM) to
give a 1.45 × 10–7 M solution of SCP complexes. For removing oxygen, the buffer was degassed by ultrasonica-
tion for 30 min. The solution was immediately divided into aliquots of 10 μL, and either used immediately, or
snap-frozen in liquid nitrogen and stored at − 80 °C for later use. All handlings took place in the dark while the
samples were kept cold in an ice bucket. Protocol 1 Protocol 1 refers to fresh samples. To prepare the sample, the aliquoted SCP solution (1.45 × 10–7 M)
was defrosted on an ice bucket. For the experiments on ensembles of SCP complexes embedded in a thin film
of polyvinyl alcohol (PVA) about 1 µL from the aliquots was further diluted with the buffer to a volume of 200
µL and 2% of PVA (w/v; mw 124.000–186.000) were added. For the experiments on the single SCP complexes
the sample was further diluted with the buffer in three steps to 0.07 pM, and in the last dilution step 2% (w/v)
PVA was added. www.nature.com/scientificreports/ However, for using high excitation intensities both ensembles
and single SCP complexes prepared according to protocol 3 feature a transition from "protocol 1 like" spectra
to "protocol 2 like" spectra after a waiting time of about two hours, which presumably reflects the time that is
required for warming up the sample chamber to room temperature. From this we infer that for the 7 months
old samples the formation of the red-shifted state requires next to a high excitation intensity in addition some
thermal activation, and that the corresponding barrier is in the order of the available thermal energy at room
temperature, i.e., about 200 cm−1. Accordingly, the height of this barrier will be significantly higher than this
for the fresh sample (protocol 1), because for this scenario the formation of the red-emitting states was not
observed despite using high excitation intensities and handling the samples at room temperature prior to the
optical experiments. This hypothesis is sketched in Fig. 6.hl h
The change of the barrier height during storage of the samples could reflect a conformational change that
occurs in a part of the protein scaffold such as the formation or the loss of a hydrogen bond, or an isomerization
of one of the chromophores or one of the residues in the protein scaffold. As has been shown in42,43 for LHC
II and a similar light-harvesting structure very little changes in the liganding of the chromophores will have
drastic effects for the photophysical properties of the complexes such as spectral position of the emission and/
or fluorescence lifetimes. Whether these observations are related to the processes that regulate the photoprotec-
tion of the photosynthetic machinery44 under strong illumination conditions will need some further attention. Scientific Reports | (2022) 12:8461 | https://doi.org/10.1038/s41598-022-11572-3 www.nature.com/scientificreports/ Finally, we note that it is not uncommon to store isolated and purified photosynthetic antennae proteins
for some time span at − 80 °C in the dark, and to verify the integrity of the samples by taking UV/VIS spectra. This is usually carried out using low excitation intensities in the order of some 10–100 µW/cm2. However, the
results presented here for the light-harvesting complexes of the marine alga Codium fragile show that this might
be problematic and that (at least for this particular light-harvesting complex) ensemble UV/VIS spectroscopy
might not be sufficient to verify the intactness of the samples. www.nature.com/scientificreports/ transients were deconvoluted with the instrument response function using commercial software (SymPhoTime
64, Picoquant). All experiments have been conducted at room temperature. transients were deconvoluted with the instrument response function using commercial software (SymPhoTime
64, Picoquant). All experiments have been conducted at room temperature. Fluorescence quantum yield measurement. The fluorescence quantum yields were measured and
determined according to the procedure described in47. The measurements on SCP were carried out at room
temperature using a laser diode (Thorlabs, HL6358MG) operating at 639 nm as excitation light source. For
detection a CCD-camera (ANDOR IDUS 420) that was connected to a spectrograph (MS125), which in turn
was fibre-coupled to an integrating sphere was used. Received: 1 December 2021; Accepted: 26 April 2022 Received: 1 December 2021; Accepted: 26 April 2022 References H. P. et al. Single-molecule spectroscopy reveals that individual low-light LH2 complexes from Rhodopseu-
domonas palustris 2.1.6. have a heterogeneous polypeptide composition. Biophys. J. 97, 1491–1500 (2009). 16. Richter, M. F., Baier, J., Cogdell, R. J., Köhler, J. & Oellerich, S. Single-molecule spectroscopic characterization of light-harvesting
2 complexes reconstituted into model membranes. Biophys. J. 93, 183–191 (2007). p y
17. Landsiedel, R. et al. Testing metal-oxide nanomaterials for human safety. Adv. Mater. 22, 2601–2627 (2010). ll
l
f
l
h
l h h
f ll
d b
l
l
l 18. Gall, A. et al. Conformational switching in a light-harvesting protein as followed by single-molecule spectroscopy. Biophys. J. 2713–2720 (2015).h 19. Löhner, A. et al. The origin of the split B800 absorption peak in the LH2 complexes from Allochromatium vinosum. Photosynth. Res. 123, 23–31 (2015).il 0. Tubasum, S. et al. Protein configuration landscape fluctuations revealed by exciton transition polarizations in single light harvesting
complexes. J. Phys. Chem. B 120, 724–732 (2016). p
y
1. Jelezko, F., Tietz, C., Gerken, U., Wrachtrup, J. & Bittl, R. Single-molecule spectroscopy on photosystem I pigment−protein com-
plexes. J. Phys. Chem. B 104, 8093–8096 (2000). p
y
2. Brecht, M., Radics, V., Nieder, J. B., Studier, H. & Bittl, R. Red antenna states of photosystem I from Synechocystis PCC 6803
Biochemistry-US 47, 5536–5543 (2008). y
3. Brecht, M., Hussels, M., Schlodder, E. & Karapetyan, N. V. Red antenna states of Photosystem I trimers from Arthrospira platensis
revealed by single-molecule spectroscopy. Biochim. Biophys. Acta 1817, 445–452 (2012). 4. Brecht, M., Radics, V., Nieder, J. B. & Bittl, R. Protein dynamics-induced variation of excitation energy transfer pathways. Proc
Natl. Acad. Sci. USA 106, 11857–11861 (2009). 5. Tietz, C. et al. Single molecule spectroscopy on the light-harvesting complex II of higher plants. Biophys. J. 81, 556–562 (2001). 6 K ü
T P J N
d
hki
V I Ili
i
C &
G
d ll
R Fl
t
l d
i
f i
l LHCII t i 25. Tietz, C. et al. Single molecule spectroscopy on the light-harvesting complex II of higher plants. Biophys. J. 81, 556–562 (2001). 26. Krüger, T. P. J., Novoderezhkin, V. I., Ilioaia, C. & van Grondelle, R. Fluorescence spectral dynamics of single LHCII trimers. Biophys. J. 98, 3093–3101 (2010). 6. Krüger, T. P. J., Novoderezhkin, V. I., Ilioaia, C. & van Grondelle, R. Fluorescence spectral dynamics of single LHCII trimers
Biophys. J. References g
g
y
2. Blankenship, R. E. (ed.) Molecular Mechanisms of Photosynthesis (Wiley, 2014). 2. Blankenship, R. E. (ed.) Molecular Mechanisms of Photosynthesis (Wiley, 2014). . Akhtar, P. et al. Spectral tuning of light-harvesting complex II in 3. Akhtar, P. et al. Spectral tuning of light-harvesting complex II in the siphonous alga Bryopsis corticulans and its effect on energy
transfer dynamics. Biochim. Biophys. Acta 1861, 148191 (2020). transfer dynamics. Biochim. Biophys. Acta 1861, 148191 (2020) y
y
4. Streckaite, S. et al. Pigment structure in the light-harvesting protein of the siphonous green alga Codium fragile. Biochim. Biophys
Acta 1862, 148384 (2021). 5. Nguyen, H. L. et al. An exciton dynamics model of bryopsis corticulans light-harvesting complex II. J. Phys. Chem. B 125, 1134–
1143 (2021). 6. Chen, G. et al. Characterization of chlorophyll-protein complexes isolated from a Siphonous green alga, Bryopsis corticulans
Photosynth. Res. 96, 75–81 (2008). y
7. Wang, W. et al. Spectral and functional studies on siphonaxanthin-type light-harvesting complex of photosystem II from Bryopsis
corticulans. Photosynth. Res. 117, 267–279 (2013).h y
(
)
8. Kühlbrandt, W. & Wang, D. N. Three-dimensional structure of plant light-harvesting complex determined by electron crystal
lography. Nature 350, 130–134 (1991). g
p y
(
)
9. Liu, Z. et al. Crystal structure of spinach major light-harvesting complex at 2.72 A resolution. Nature 428, 287–292 (2004 10. Seki et al. Unveiling the blue-green light utilization by an ocea g
g
g
y
g
g
11. Bopp, M. A., Sytnik, A., Howard, T. D., Cogdell, R. J. & Hochstrasser, R. M. The dynamics of structural deformations of immobi
single light-harvesting complexes. Proc. Natl. Acad. Sci. USA 96, 11271–11276 (1999). g
g
g
p
2. Hofmann, C., Aartsma, T. J., Michel, H. & Köhler, J. Direct observation of tiers in the energy landscape of a chromoprotein: a
single-molecule study. Proc. Natl. Acad. Sci. USA 100, 15534–15538 (2003). 12. Hofmann, C., Aartsma, T. J., Michel, H. & Köhler, J. Direct observation of tiers in the energ
single-molecule study. Proc. Natl. Acad. Sci. USA 100, 15534–15538 (2003). 13. Ketelaars, M. et al. Spectroscopy on Individual light-harvesting 1 complexes of Rhodopseudomonas acidophila. Biophys. J. 83,
1701–1715 (2002).h 14. Cogdell, R. J., Gall, A. & Köhler, J. The architecture and function of the light-harvesting apparatus of purple bacteria: from single
molecules to in vivo membranes. Q. Rev. Biophys. 39, 227–324 (2006). Q
p y
,
(
)
5. Brotosudarmo, T. Materials and methods
Cultivation of algae.
Cod In order to find
a single complex, the focal spot on the sample could be scanned over a range of 250 µm × 250 µm using a galvo
scanner (FLIMbee unit, PoicoQuant). The emission from the sample passed a dichroic mirror (ZT488/561 kpc,
AHF/Croma) and suitable optical filters (LP H 560 LPXR, AHF) and was then split in a beamsplitter cube in a
ratio of 50/50. One part of the signal was directed towards a spectrometer (Shamrock SR-163, grating 600 lines/
mm blazed at 500 nm) via another fibre, and the emission spectrum was detected using an electron-multiplying
charge-coupled device (EMCCD Newton 970, Andor). The spectral resolution was 0.5 nm corresponding to
11 cm−1 at 680 nm. The other half of the signal was focused onto a single-photon counting avalanche diode
(SPCM-AQRH-14-TR, Excelitas) and used for time tagged time-resolved (TTTR) fluorescence data collection
employing time correlated single photon counting (TCSPC TimeHarp 260 PICO Dual, PicoQuant). The tran-
sients were measured with a temporal resolution of 250 ps. In order to avoid pile-up effects in the photon count-
ing statistics when measuring ensembles, we ensured that the detection count rate was less than 1% of the laser
repetition rate by placing an OD 3 in front of the detector. For obtaining the fluorescence lifetimes the measured https://doi.org/10.1038/s41598-022-11572-3 Scientific Reports | (2022) 12:8461 | www.nature.com/scientificreports/ www.nature.com/scientificreports/ An improved experimental determination of external photoluminescence quantum
efficiency. Adv. Mater. 9, 230–232 (1997). Acknowledgements We thank Werner Reichstein for assistance with the MT 200 microscope and Dr. Frank-Julian Kahle for assis-
tance with the fluorescence quantum yield measurements. T.H.P.B. thanks the Humboldt Foundation for a
Georg-Forster Fellowship. Financial support from the Humboldt Foundation and the State of Bavaria within the
initiatives “Solar Technologies go Hybrid" and the Elite Network "Biological Physics" is gratefully acknowledged. Financial support from the grant-in-aid of the Sasakura Enviro-Science Foundation (to S.S.) and the OCU Stra-
tegic Research Grant for Basic Researches (to R.F.) are also acknowledged. Author contributions Conceptualization, J.K. and R.F.; methodology, J.K.; software, B.W.; vali-dation, T.H.P.B., B.W., J.K.; formal analy-
sis, T.H.P.B.; investigation, T.H.P.B., S.S.; resources, J.K.; data curation, J.K. and R.F.; writing—original draft
preparation, J.K.; writing—review and editing, T.H.P.B., S.S., R.F., J.K.; visualization, T.H.P.B.; supervision, J.K.;
project administration, J.K.; funding acquisition, J.K. All authors have read and agreed to the published version
of the manuscript. References 98, 3093–3101 (2010). p y
27. Krüger, T. P. J., Ilioaia, C., Valkunas, L. & van Grondelle, R. Fluorescence intermittency from the main plant light-harvesting
complex: sensitivity to the local environment. J. Phys. Chem. B 115, 5083–5095 (2011). y
y
28. Krüger, T. P. J., Wientjes, E., Croce, R. & van Grondelle, R. Conformational switching explains the intrinsic multifunctional
plant light-harvesting complexes. Proc. Natl. Acad. Sci. USA 108, 13516–13521 (2011). hl
h
l
l
l
l
d
fi
f
h d
d
h d
f
h
h g
g
29. Schlau-Cohen, G. S. et al. Single-Molecule Identification of Quenched and Unquenched States of LHCII. J. Phys. Chem. Lett. 6,
860–867 (2015).l 0. Kondo, T., Chen, W. J. & Schlau-Cohen, G. S. Single-molecule fluorescence spectroscopy of photosynthetic systems. Chem. Rev
117, 860–898 (2017). 1. Köhler, J. & Cogdell, R. J. Studying conformational changes of proteins via single-molecule spectroscopy: Cryogenic temperatures
versus room temperature. Adv. Bot. Res. 91, 1–31 (2019).fi p
(
)
32. Ilioaia, C., Johnson, M. P., Horton, P. & Ruban, A. V. Induction of efficient energy dissipation in the isolated light-harvesting
complex of Photosystem II in the absence of protein aggregation. J. Biol. Chem. 283, 29505–29512 (2008).h p
y
p
gg
g
3. Ramanan, C. et al. The role of exciton delocalization in the major photosynthetic light-harvesting antenna of plants. Biophys. J
108, 1047–1056 (2015). (
)
34. Krüger, T. P. J., Ilioaia, C., Johnson, M. P., Ruban, A. V. & van Grondelle, R. Disentangling the low-energy states of the major light-
harvesting complex of plants and their role in photoprotection. Biochim. Biophys. Acta 1837, 1027–1038 (2014). https://doi.org/10.1038/s41598-022-11572-3 Scientific Reports | (2022) 12:8461 | www.nature.com/scientificreports/ In: Watanabe A, Hattori A (end) Cultures and collections of
algae. Proceedings of the US-Japan conference held at Hakone. Jpn. Soc. Plant Pysiol. 1966, 63–75 (1966). p
y q
p y
42. L. Provasoli. Media and prospects for the cultivation of marine algae. In: Watanabe A, Hattori A (end) Cultures and collections of
algae. Proceedings of the US-Japan conference held at Hakone. Jpn. Soc. Plant Pysiol. 1966, 63–75 (1966). g
g
J p
Jp
y
,
(
)
43. Wientjes, E., Roest, G. & Croce, R. From red to blue to far-red in Lhca4: how does the protein modulate the spectral properties of
the pigments?. Biochim. Biophys. Acta 1817, 711–717 (2012). g
g
J p
Jp
y
,
(
)
43. Wientjes, E., Roest, G. & Croce, R. From red to blue to far-red in Lhca4: how does the protein modulate the spectral properties of
the pigments?. Biochim. Biophys. Acta 1817, 711–717 (2012). 44. Pascal, A. A. et al. Molecular basis of photoprotection and control of photosynthetic light-harvesting. Nature 436, 134–137 (2005). 45. Anderson, J. M. Chlorophyll-protein complexes of a Codium species, including a light-harvesting siphonaxanthin-Chlorophylla
ab-protein complex, an evolutionary relic of some Chlorophyta. Biochim. Biophys. Acta 724, 370–380 (1983). 4. Pascal, A. A. et al. Molecular basis of photoprotection and control of photosynthetic light-harvesting. Nature 436, 134–137 (2005)
5. Anderson, J. M. Chlorophyll-protein complexes of a Codium species, including a light-harvesting siphonaxanthin-Chlorophylla
ab-protein complex, an evolutionary relic of some Chlorophyta. Biochim. Biophys. Acta 724, 370–380 (1983). 44. Pascal, A. A. et al. Molecular basis of photoprotection and control of photosynthetic light-harvesting. Nature 436, 134–137 (2005). 45. Anderson, J. M. Chlorophyll-protein complexes of a Codium species, including a light-harvesting siphonaxanthin-Chlorophylla
ab-protein complex, an evolutionary relic of some Chlorophyta. Biochim. Biophys. Acta 724, 370–380 (1983). p
p
y
p y
p y
6. Fujii, R., Yamano, N., Hashimoto, H., Misawa, N. & Ifuku, K. Photoprotection vs. photoinhibition of photosystem II in transplas-
tomic lettuce (Lactuca sativa) dominantly accumulating astaxanthin. Plant Cell Physiol. 57, 1518–1529 (2016). 7. de Mello, J. C., Wittmann, H. F. & Friend, R. H. An improved experimental determination of external photoluminescence quantum
efficiency. Adv. Mater. 9, 230–232 (1997). 47. de Mello, J. C., Wittmann, H. F. & Friend, R. H. An improved experimental determination of external photoluminescence quantum
efficiency. Adv. Mater. 9, 230–232 (1997). 47. de Mello, J. C., Wittmann, H. F. & Friend, R. H. Additional informationh Additional information
Supplementary Information The online version contains supplementary material available at https://doi.org/
10.1038/s41598-022-11572-3. Correspondence and requests for materials should be addressed to J.K. Reprints and permissions information is available at www.nature.com/reprints. © The Author(s) 2022 www.nature.com/scientificreports/ www.nature.com/scientificreports/ 5. Nordlund, T. M. & Knox, W. H. Lifetime of fluorescence from light-harvesting chlorophyll a/b proteins. Excitation intensity
dependence. Biophys. J. 36, 193–201 (1981).l p
p y
36. Tredwell, C. J. & Searle, G. F. Picosecond fluorescence from photosynthetic systems in vivo. Ciba F. Symp. 14, 257–281 (197 l
p
y
y
) Picosecond events and their measurement (Academic Press Inc, 1 l
37. Seibert, M. (ed.) Picosecond events and their measurement (Ac 38. van Oort, B., van Hoek, A., Ruban, A. V. & van Amerongen, H. Equilibrium between quenched and nonquenched conformations of
the major plant light-harvesting complex studied with high-pressure time-resolved fluorescence. J. Phys. Chem. B 111, 7631–7637
(2007). (
)
39. Böhm, P. S., Kunz, R., Southall, J., Cogdell, R. J. & Köhler, J. Does the reconstitution of RC-LH1 complexes from Rhodopseu-
domonas acidophila strain 10050 into a phospholipid bilayer yield the optimum environment for optical spectroscopy?. J. Phys. Chem. B 117, 15004–15013 (2013).l (
)
40. Moya, I., Silvestri, M., Vallon, O., Cinque, G. & Bassi, R. Time-resolved fluorescence analysis of the photosystem II antenna pro
in detergent micelles and liposomes. Biochemistry 40, 12552–12561 (2001). 40. Moya, I., Silvestri, M., Vallon, O., Cinque, G. & Bassi, R. Time-resolved fluorescence analysis of the photosystem II antenna proteins
in detergent micelles and liposomes. Biochemistry 40, 12552–12561 (2001). 41. Belgio, E., Johnson, M. P., Jurić, S. & Ruban, A. V. Higher plant photosystem II light-harvesting antenna, not the reaction center,
determines the excited-state lifetime-both the maximum and the nonphotochemically quenched. Biophys. J. 102, 2761–2771 (2012). 42. L. Provasoli. Media and prospects for the cultivation of marine algae. In: Watanabe A, Hattori A (end) Cultures and collections of
algae. Proceedings of the US-Japan conference held at Hakone. Jpn. Soc. Plant Pysiol. 1966, 63–75 (1966). 41. Belgio, E., Johnson, M. P., Jurić, S. & Ruban, A. V. Higher plant photosystem II light-harvesting antenna, not the reaction center,
determines the excited-state lifetime-both the maximum and the nonphotochemically quenched. Biophys. J. 102, 2761–2771 (2012). 42. L. Provasoli. Media and prospects for the cultivation of marine algae. In: Watanabe A, Hattori A (end) Cultures and collections of
algae. Proceedings of the US-Japan conference held at Hakone. Jpn. Soc. Plant Pysiol. 1966, 63–75 (1966). determines the excited-state lifetime-both the maximum and the nonphotochemically quenched. Biophys. J. 102, 2761–2771 (2012). 42. L. Provasoli. Media and prospects for the cultivation of marine algae. Funding Open Access funding enabled and organized by Projekt DEAL. Competing interests
The authors declare no competing interests. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2022 https://doi.org/10.1038/s41598-022-11572-3 Scientific Reports | (2022) 12:8461 |
|
https://openalex.org/W2569360452
|
https://europepmc.org/articles/pmc5223340?pdf=render
|
English
| null |
Sijunzi Decoction attenuates 2, 4, 6-trinitrobenzene sulfonic acid (TNBS)-induced colitis in rats and ameliorates TNBS-induced claudin-2 damage via NF-κB pathway in Caco2 cells
|
BMC complementary and alternative medicine
| 2,017
|
cc-by
| 8,662
|
© The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35
DOI 10.1186/s12906-016-1549-3 Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35
DOI 10.1186/s12906-016-1549-3 Sijunzi Decoction attenuates 2, 4,
6-trinitrobenzene sulfonic acid (TNBS)-
induced colitis in rats and ameliorates
TNBS-induced claudin-2 damage via
NF-κB pathway in Caco2 cells Yue Lu1,2†, HanJie Lin1,2†, JinWei Zhang1,2, JianAn Wei1,2, Jing Sun1,2 and Ling Han1,2* Yue Lu1,2†, HanJie Lin1,2†, JinWei Zhang1,2, JianAn Wei1,2, Jing Sun1,2 and Ling Han1,2* Yue Lu1,2†, HanJie Lin1,2†, JinWei Zhang1,2, JianAn Wei1,2, Jing Sun1,2 and Ling Han1,2* * Correspondence: linghan36@163.com
†Equal contributors
1The Second Clinical College, GuangzhouUniversity of Chinese Medicine,
Guangzhou 510000, China
2Guangdong Provincial Hospital of Chinese Medicine, Guangzhou 510000,
China Abstract Background: SijunziDecoction (SJZD) is a traditional Chinese medicine prescription used to treat the diseases of
gastrointestinal tract since ancient times. The objective of this study was to investigate the protective effects of
SJZD on TNBS-induced colitis in rats and TNBS-damaged Caco2 cells. Methods: The rat colitis model was induced by 2, 4, 6-trinitrobenzene sulfonic acid (TNBS). SJZD (2.8 5.6, 11.2 g/kg)
or salazosulfapyridine (SASP) (0.4 g/kg) was administrated orally in rats for 7 days. DAI, pathological scores and the
expression of claudin-2 were evaluated. Then we explored the effect and mechanism of SijunziDecoction Serum
(SJZDS) onTNBS-damaged Caco2 cells to figure out intestinal barrier protective effect and mechanism of SJZD. Results: SJZD significantly ameliorated the severity of TNBS-induced colitis and downregulated the level of claudin-2
in colonic tissues. SJZDS promoted proliferation and inhibited apoptosis ofTNBS-damaged Caco2 cells. In Caco2 cell
monolayers, we provided mechanistic evidence that SJZDS-induced increased TEER and decreased permeability after
TNBS damage, which were mediated through claudin-2 and NF-κB pathway, including the upregulation of claudin-2,
decreased activity of NF-κB p65, reduced level of NF-κB p65 and MLCK. Conclusions: Our results indicated that SJZD possesses protective effect of intestinal barrier towards TNBS-induced
colitis in rats and TNBS-damaged Caco2 cells in vitro. SJZDis a potential protective agent of intestinal barrier that
deserves further investigation. Keywords: Claudin-2, Inflammatory bowel disease, NF-κB pathway, SijunziDecoction, Tight junction Preparation of Sijunzi Decoction (SJZD) p
j
(
)
The preparation of concentrated water decoction of SJZD
was as follows: The herbs (composed of Ginseng Radix et
Rhizoma, Atractylodes Macrocephalae Rhizoma, Poria,
and Glycyrrhizae Radix et Rhizoma Praeparate cum
Melle, the ratio of four herbs was 3: 3: 3: 2 and the total
quantity was 216 g) were placed in a container, and soaked
in cold water of about 7 times the amount of herbs for
2 h; the herbs were boiled for 30 min and then filtered. Next, the herbal drugs were added with water of about 5
times the amount of the drugs, then decocted and boiled
for 30 min and filtered again. At last, the two filtrates were
mixed together. Three drug concentrations (1.12 0.56 and
0.28 g/ml) were prepared and stored at 4 °C. Sijunzi Decoction (SJZD),also known as Four Gentleman
Decoction, is a classical prescription for curing spleen
deficiency in traditional Chinese medicine, consisting
of Ginseng Radix et Rhizoma, or Codonopsispilosula,
Atractylodes Macrocephalae Rhizoma, Poria, and Gly-
cyrrhizae
Radix
et
Rhizoma
Praeparatecum
Melle. Modern pharmacological experiments have proved that
saponin, flavonoid, and polysaccharide are the most active
ingredients in SJZD [13]. SJZD has been used for years in
China to regulate the gastrointestinal function and en-
hance the immunity [14]. However, molecular mecha-
nisms by which SJZD suppressed inflammation bowel
disease were unclear. In the present study, we aimed to in-
vestigate the therapeutic efficacy of SJZD against IBD. Also, we analyzed the expression of tight junction related
protein to investigate the mucosal barrier protective
mechanism of SJZD in vitro. Preparation of Sijunzi Decoction-serum (SJZDS)
Twenty male SD rats (6–8 weeks old, 180–200 g) were
purchased from the Experimental Animal Center of
Guangdong Provincial Hospital of Chinese Medicine. Rats were divided randomly into two groups and were
given SJZD or equivalent volume of saline. Rats were ad-
ministered orally with SJZD at 0.5 g/kg twice a day for
4 days. Then, 1.5 h after the last administration, blood of
the rats was collected from the abdominal aorta under
ether anesthesia. Blood from the same group was mixed
and centrifuged to obtain serum. The serum was heat-
inactived at 56 °C and sterilized by membrane filtration
before use. The serum of rats administered with vehicle
alone was also prepared as control. Induction of colitis Colitis was induced using the TNBS, as described previ-
ously. The rats that had been fasted for 24 h with free
access to water were anesthetized with 10% chloral hy-
drate, and a polyethylene catheter (2 mm in outer diam-
eter) was inserted rectally (6–8 cm from the anus). Then, 100 mg/kg TNBS dissolved in 0.25 ml of 50%
ethanol was administered through the catheter. After Animals Female Wistar rats (6–8 weeks old, 180–200 g) were
obtained from the Experimental Animal Center of
Guangdong Province (Guangzhou, China). All rats were
fed on a standard diet, had free access to water and were
housed under standard laboratory conditions. All experi-
ments were approved by the Animal Welfare and Ethics
Branch
of
the
Biomedical
Ethics
Committee
of
Guangzhou University of Chinese Medicine. DMEM medium, fetal bovine serum (FBS) and NEAA
were purchased from GIBCOLaboratories (Grand Island,
NY, USA). 2, 4, 6-trinitrobenzene sulfonic acid (TNBS),
MTT were purchased from Sigma-Aldrich (St. Louis,
Mo, USA). Salazosulfapyridine (SASP) was purchased
from Tongda Pharmaceutical Company Ltd. (Datong,
Shanxi, China). The anti-claudin 2 antibody, anti-myosin
light chain kinase antibody and NF-κB p50/p65 tran-
scription factor assay kit were all obtained from Abcam
(Cambridge, MA, USA). Trizol and cDNA synthesis kit
were obtained from Invitrogen (Carlsbad, CA, USA). The RT-PCR primers were synthesized by Invitrogen
(Shanghai, China). Background T cells [3]. The ongoing activation of the mucosal im-
mune system are considered to be the major patho-
logical agents of IBDs [4]. Inflammatory bowel diseases (IBDs) are chronic recur-
rent inflammatory diseases of the gastrointestinal tract,
mainly including Crohn’s disease (CD) and ulcerative
colitis (UC) [1, 2]. The development of IBDs is linked to
characteristic changes in the colon, such as epithelial cell
necrosis and ulceration, local infiltration by immune
cells, including a significant number of neutrophils and The epithelium of the mucosa plays an essential role
in maintaining balance of intestinal ecosystem [5]. The
tight junction complexes play a decisive role in the
maintenance of barrier integrity [6]. Tightjunctions are
multi-protein complexes, which composed of transmem-
brane proteins, peripheral membrane (scaffolding) pro-
teins and regulatory molecules. The claudin protein
family is the most important transmembrane proteins,
which can impact the tight junction permeability. The
transmembrane proteins also include occludin, which * Correspondence: linghan36@163.com
†Equal contributors
1The Second Clinical College, GuangzhouUniversity of Chinese Medicine,
Guangzhou 510000, China
2Guangdong Provincial Hospital of Chinese Medicine, Guangzhou 510000,
China * Correspondence: linghan36@163.com
†Equal contributors
1The Second Clinical College, GuangzhouUniversity of Chinese Medicine,
Guangzhou 510000, China
2Guangdong Provincial Hospital of Chinese Medicine, Guangzhou 510000,
China Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Page 2 of 13 interacts directly with claudins and actin. Zonula oc-
cludens 1 (ZO1) and ZO2 are peripheral membrane pro-
teins which play a crucial role on tight junction assembly
and maintenance [7]. and Glycyrrhizae Radix et Rhizoma Praeparate cum
Melle, and these four drugs were in same ratio as
described in Pharmacopoeia. Dried herbs were pur-
chased from Kangmei Pharmaceutical Company Ltd. (Guangzhou, Guangdong, China). Currently, conventional drugs used for treatment of IBD
are including immunosuppressants, corticosteroids, disease
improving agents and biological agents, such as TNF anti-
bodies. However, most of these medicines only temporarily
alleviate the symptoms, and are associated with mild to
serious side effects, leading to their limited clinical applica-
tions [8]. Thus traditional Chinese Medicine Formulas
gaining more and more attention in the treatment of IBD,
due to its effective and safe [9]. Some clinical and experi-
mental studies have shown that Chinese formulas for treat-
ment of IBD were effective, and most of their mechanisms
were related with anti-inflammatory, anti-oxidant and re-
store the function of intestinal barrier [10–12]. Electrical resistance measurements Transepithelial electrical resistance (TEER) of Caco2 cells
was monitored using an EVOM TEER meter (Millipore). TEER increased until day 7 when a steady state of higher
than 200 Ω cm2 was reached, indicating the formation of
complete tight junction, an intact monolayer and maxi-
mized integrity of barrier function [15]. TNBS stimulation
of Caco2 cells lasted for 24 h until addition of SJZDS. After removal of TNBS, the Caco2 cell monolayers was
washed with PBS, numerical readings of TEER were re-
corded at 24 h after the co-incubation of Caco2 cells and
SJZDS. Permeability study All transport studies were conducted at 37 °C, in a
medium of DMEM + 10% FBS + 1%NEAA. Prior to the
transport study, cell monolayers were washed with PBS,
phenolsulfonphthalein at final concentrations of 20 mg/L
in distilled water was added to the apical compartment. The basolateral compartment contained only water. After
4 h, 150 μl of samples were removed from the basolateral
compartment into 2 ml EP tubes which possess 1.5 ml
NaOH (20 μmol/ml) for neutralization. The absorbance at
570 nm was measured. Plant materials Sijunzi Decoction (SJZD) was composed of Ginseng Radix
et Rhizoma, Atractylodes Macrocephalae Rhizoma, Poria, Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Page 3 of 13 DMEM supplemented with 10% FBS and 1% NEAA in a
humidified 5% CO2 atmosphere at 37°C. DMEM supplemented with 10% FBS and 1% NEAA in a
humidified 5% CO2 atmosphere at 37°C. DMEM supplemented with 10% FBS and 1% NEAA in a
humidified 5% CO2 atmosphere at 37°C. administration of TNBS, rats were held in headfirst
position for 3 min to prevent the solution leaking
out. Then, the animals were placed in separate cages
with free access to food and water. Control rats re-
ceived 0.9% saline. Stool consistency and occult blood
were recorded daily. Rats were sacrificed after 7 days
treatment. The colonic tissues were quickly removed
and rinsed with ice-cold PBS. Histological assessment of colitis Formalin-fixed colon sections were embedded in paraffin
and sectioned (7 μm), then stained by hematoxylin and
eosin (H&E). Histopathologic severity was scored using
three different parameters as previously described: severity
of inflammation (0, none; 1, mild; 2, moderate; 3, severe);
extent of inflammation (0, none; 1, mucosa; 2, mucosa
and submucosa; 3, transmural);crypt damage (0, none; 1,
basal one-third damaged; 2, basal two-thirds damaged; 3,
crypt lost but surface epithelium present; 4, crypt and sur-
face epithelium lost). All parameters were scored by two
independent observers in a blinded manner. SJZD administration Cell viability was determined by the MTT assay. Cells
were plated at a density of 5000 per well in 96-well plates
in DMEM + 10% FBS + 1% NEAA and then treated with
200 μg/ml TNBS. After exposure to TNBS for 24 h, SJZDS
or control serum (cs) was added for 24 h, then10 μl MTT
solution was added for 4 h. The cells were lysed with
0.04 N HCL in isopropyl alcohol, and the absorbance was
read at 570 nm. Cell viability was calculated as follows: cell
viability (%) = absorbance of test group/absorbance of con-
trolled cell group × 100%. After a 1-week adaptive period, rats were randomly divi-
dedinto four groups (n = 8) as follows. (1) The control
group and TNBS group received 2 ml saline via the rec-
tum; (2) The TNBS + low dose of SJZD group received
SJZD at concentration of 2.8 g/kg;(3) The TNBS +
medium dose of SJZD group received SJZD at concen-
tration of 5.6 g/kg; (4) The TNBS + high dose of SJZD
group received SJZD at concentration of 11.2 g/kg;(5)
The TNBS + SASP group received SASP at concentra-
tion of 0.4 g/kg; Drug administration to animals began
24 h after induction of colitis and was treated for 7 days. TNBS damage model of cells For TNBS damage model of cells, the Caco2 cells were
treated with 200 μg/ml TNBS for 24 h. The Caco2 cells
incubated with DMEM media were used as control. Hoechst 33342 staining
l
f
h For analysis of chromatin condensation or nuclear frag-
mentation, SJZDS-treated cells after TNBS damage were
harvested and fixed in 3% paraformaldehyde (PFA) for
20 min, incubated in 0.1% Triton-X100-PBS for 30 min,
then stained in 0.25 μg/ml Hoechst 33342 for 15 min. The cells were observed using the IX70 fluorescence
microscope (Olympus, Japan). Immunochemistry The sections were dewaxed by heating at 55 °C for
30 min, washed twice for 15 min, rehydrated in ethanol
for 15 min, proceeded in water bath at 95 °C for 5 min
for antigen unmasking, and treated with 3% hydrogen
peroxide for 30 min. The sections were then incubated
at 4 °C overnight with anti- claudin-2 antibody diluted
1:500. Next, the sections were washed with PBS and in-
cubated secondary antibody at 37 °C for 30 min. 3’3-di-
aminobenzidine tetrachloride was added to observe the
positive expression. Negative control sections were incu-
bated with PBS instead of primary antibody. The expression
of claudin-2 was quantified by image-pro plus software. Statistical analysis Results were expressed as the mean ± SEM. Statistical
analysis was carried out by the Student’s t-test, P values
less than 0.05 were considered significant. At least three
independent experiments were performed. SJZD decreased histological changes in rats with TNBS-
induced colitis Caco2 cells were seeded in the wells of 6-well plates and
cultured overnight. TNBS damaged cells were treated
with SJZDS (15%) for 0, 24, 36, 48 h. After the treatment
of SJZDS, Caco2 cells were fixed in 3% paraformalde-
hyde (PFA) for 30 min, incubated overnight at 4 °C with
anti-claudin 2 antibody, followed by incubation for 1 h
with anti-rabbit IgG-FITC at 1:200 dilution. Finally, an
immunofluorescence assay was carried out according to
the protocol by a fluorescence microscope (Olympus,
BX51, Japan). From Fig. 1b and c,histological analysis of the rat colonic
tissues revealed a significant reduction in colon inflamma-
tion and epithelial cells disruption after the administration
of SIZD, compared to the TNBS group. Numerous neu-
trophils and granulocytes, erosion of mucosal layers were
present in the colon of the TNBS group, and a signifi-
cantly reduced influx of inflammatory cells and intact
architecture of the crypts were observed in the colon of
the SJZD and SASP treated rats. SJZD ameliorated clinical parameters in rats with TNBS-
induced colitis To determine whether oral administration of SJZD could
ameliorate the intestinal damage in colitis rats, we induced
colitis by administration of TNBS and then treated the
rats with SJZD or SASP (positive control) for 7 days. From
Fig. 1a, the TNBS group had a sharp increase of the DAI
(3.83) from start to day 5, and symptoms were maintained
during the experimental period. Compared to the TNBS
group, medium and high dose of SJZD significantly de-
creased the disease severity of TNBS-induced colitis. SASP also reduced the DAImarkedly compared with the
TNBS group. Western blot analysis y
The SJZDS-treated Caco2 cells were harvested at 24 h
after the addition of TNBS or vehicle to analyze the ex-
pression of ZO-1, ZO-2, ZO-3, claudin-1, claudin-2,
claudin-3, claudin-4, occludin, JAM, E-cadherin and
GAPDH mRNAs. Total RNA was isolated using Trizol re-
agent. RNA purity and concentration were assayed with a
NanoDrop 2000 devise (Thermo Scientific, Wilmington,
DE, U.S.A.). The mRNAs were then reverse transcribed
directly into cDNA using a RT-PCR kit according to the
manufacturer’s instructions. PCR amplification conditions
were as follows: initial denaturation at 95 °C for 15 s
followed by 35 cycles of denaturation at 95 °C for 5 s and
annealing at 61 °C for 15 s. The sequences of the respect-
ive sense and antisense primers were as follows (from
5’to 3’): CAACATACAGTGACGCTTCACA and CAC
TATTGACGTTTCCCCACTC for ZO-1; ATGGAAGA
GCTGATATGGGAACA and TGCTGAACTGCAAAC
GAATGAA
for
ZO-2;
GCTTTGGCATTGCGATC
TCTG and GATGTGGTCGCCTGTCTGTAG for ZO-3;
AGGAATTAACTGCATACGTTTTGG and TAGCCAC
AGAAAGCATCGGG for claudin-1; GCCTCTGGATGG
AATGTGCC and GCTACCGCCACTCTGTCTTTG for
claudin-2; AACACCATTATCCGGGACTTCT and GCG
GAGTAGACGACCTTGG for claudin-3; TGGGGCTAC
AGGTAATGGG and GGTCTGCGAGGTGACAATGTT
for claudin-4; ACAAGCGGTTTTATCCAGAGTC and
GTCATCCACAGGCGAAGTTAAT for occludin; TGT
TTCAGTTCTGTGTCATGGT and
TGCAGACAAGG
TGTTTTCCAG for JAM; CGAGAGCTACACGTTCA
CGG and GGGTGTCGAGGGAAAAATAGG for E-
cadherin; GGAGCGAGATCCCTCCAAAAT and GGC
TGTTGTCATACTTCTCATGG for GAPDH. Relative
mRNA
quantities
were
determined
by
using
the
2ΔΔCt method with data normalized to the GAPDH
housekeeping gene. y
The cells were plated in 6-well plates at density of 1 × 106
per well, TNBS damaged cells were treated with SJZDS
(15%) for different times. After the indicated times, cells
were collected and lysed in lysis buffer [50 mM Tris
(pH 7.4), 150 mM NaCL, 1% Triton X-100, 1% sodium
deoxycholate, 0.1% sodium dodecylsulfate (SDS), 1 mmol/
l phenylmethylsulphonyl fluoride, and protease inhibitors]. The protein concentration was measured by BCA assay
kit, equal amounts of protein were separated by 12.5%
SDS-PAGE, then transferred to nitrocellulose membranes. The membranes were blocked and incubated overnight at
4 °C with primary antibodies, followed by incubation for
1 h with the secondary antibodies. Finally, protein ex-
pression was detected using the Bio-rad Imaging System
(Bio-rad Biosciences, USA). Cell culture The human colon adenocarcinoma cell line Caco2, pur-
chased from the Cell Culture Unit of Shanghai Scicence
Academy
(Shanghai,
China),
Cells
were
grown
in Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Page 4 of 13 Page 4 of 13 SJZD upregulated the level of claudin-2 in colon of TNBS-
induced colitis rats The cellular extracts were prepared, the activity of NF-
κB p65 transcription factor was measured using NF-κB
p50/p65 transcription factor assay kitaccording to the
manufacturer’s instructions. Since claudins are the most important transmembrane
proteins which can impact the permeability of tight Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35
Page 5 of 13 Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Page 5 of 13 Fig. 1 SJZD has a protective effect against TNBS-induced colitis. a The disease activity index (DAI) were monitored. b Representative histological
photograph of colon sections. c Microscopic score of sections (*P < 0.05, **P < 0.01 and ***P < 0.001 vs. control group; #P < 0.05, ##P < 0.01 and
###P < 0.001 vs. TNBS-induced colitis group; n = 8) Fig. 1 SJZD has a protective effect against TNBS-induced colitis. a The disease activity index (DAI) were monitored. b Representative histological
photograph of colon sections. c Microscopic score of sections (*P < 0.05, **P < 0.01 and ***P < 0.001 vs. control group; #P < 0.05, ##P < 0.01 and
###P < 0.001 vs. TNBS-induced colitis group; n = 8) Fig. 1 SJZD has a protective effect against TNBS-induced colitis. a The disease activity index (DAI) were monitored. b Representative histological
photograph of colon sections. c Microscopic score of sections (*P < 0.05, **P < 0.01 and ***P < 0.001 vs. control group; #P < 0.05, ##P < 0.01 and
###P < 0.001 vs. TNBS-induced colitis group; n = 8) SJZDS inhibited cell apoptosis of TNBS-damaged Caco2 cells
To determine whether the growth protection of SJZDS on
TNBS-damaged Caco2 cells was related to the inhibition
of cell apoptosis, several apoptotic parameters were mea-
sured. Fluorescence microscopic examination indicated
that Caco2 cells displayed markednucleus condensation
after TNBS treatment, as shown in Fig. 3b. After treat-
ment of 15% SJZDSfor 24 h, the amount of shriveled nu-
cleus decreased. After treatment of TNBS-induced Caco2
cells with 15% SJZDSfor 24 h, the percentage of cell apop-
tosis decreased from 69.6 ± 1.6% to 44.2 ± 3.8% (Fig. 3c, d). junction, we investigated the expression of claudin-2 in
colon tissue by immunochemistry. From Fig. 2, com-
pared to control group, the expression of claudin-2 was
downregulated in TNBS-induced colitis rats. The level
of claudin-2 was upregulated after the treatment of SJZD
and SASP, compared to the TNBS group. SJZD upregulated the level of claudin-2 in colon of TNBS-
induced colitis rats SJZDS promotes the growth of TNBS-damagedCaco2 cells
The protective effect of SJZDS on TNBS-damagedCaco2
cells was initially determined by the MTT viability assay. From Fig. 3a, it was noted that SJZDSpromoted the pro-
liferation of TNBS-damagedCaco2 cells significantly in a
dose-dependent manner, when the concentration of
SJZDS achieved15 and 20%, the growth promotion was
moresignificant. SJZDS protectedbarrier function of Caco2 cells monolayer
via regulating claudin-2 These results suggested that the protein of claudin-2
might be involved in the protective mechanism of bar-
rier function induced by SJZDS. There is increasing evidence that the change of tight
junction can influence the intestinal barrier function,
which is usually related to an increased permeability and
a decrease in TEER [16]. To investigate the underlying
protective mechanism of barrier function in Caco2 cells,
we analyzed the tight junction protein following SJZDS
treatment. From Fig. 5, TNBS can induce the downregu-
lation of different tight junction proteins, and SJZDS
promotes the expression of claudin-2, as measured by SJZDS enhanced the TEER of Caco-2 cells and reduced the
permeability SJZDS enhanced the TEER of Caco-2 cells and reduced the
permeability To investigate whether SJZDS can protect the barrier
function of Caco2 cells monolayer, we analyzed the Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Page 6 of 13 Fig. 2 Effect of SJZD on tight junction protein claudin 2 of TNBS-induced colitis in rats by immunochemistry. a Representative immunochemical
photograph of colon sections, the original magnification was 400×. b Quantification of integrated optical density (IOD) in claudin-2. Data represented
the mean ± SEM of at least three independent experiments (*P < 0.05 vs. control group; #P < 0.05, ##P < 0.01 and ###P < 0.001 vs. TNBS-induced colitis
group; n = 8) Fig. 2 Effect of SJZD on tight junction protein claudin 2 of TNBS-induced colitis in rats by immunochemistry. a Representative immunochemical
photograph of colon sections, the original magnification was 400×. b Quantification of integrated optical density (IOD) in claudin-2. Data represented
the mean ± SEM of at least three independent experiments (*P < 0.05 vs. control group; #P < 0.05, ##P < 0.01 and ###P < 0.001 vs. TNBS-induced colitis
group; n = 8) TEER and permeability of Caco2 cells following SJZDS
treatment. From Fig. 4a, SJZDSenhanced the TEER of
Caco2 cells monolayer at 24 h, as measured by EVOM
TEER meter. Besides, SJZDS can reduced the permeabi-
lityof Caco2 cells (Fig. 4b). TEER and permeability of Caco2 cells following SJZDS
treatment. From Fig. 4a, SJZDSenhanced the TEER of
Caco2 cells monolayer at 24 h, as measured by EVOM
TEER meter. Besides, SJZDS can reduced the permeabi-
lityof Caco2 cells (Fig. 4b). RT-PCR, but SJZDS has no significant effect on the ex-
pression of other tight junction proteins. Furthermore, we quantify the protein expression of
claudin-2 by western blot and immunofluorescence. As
shown in Fig. 6, the level of claudin-2 was decreased sig-
nificantly after TNBS damage at 24 h, SJZDS treatment
upregulates the expression of claudin-2 for 24 and 36 h,
correspondingly,
the
results
of
immunofluorescence
were similar (Fig. 7). SJZDS protectedbarrier function of Caco2 cells monolayer
via regulating claudin-2 SJZDS upregulated the level of claudin-2 to protect
barrier function via inhibiting NF-κB signaling activation Research reveals that claudin-2 has an important role in
inflammation of colon,there is a relationship between
claudin-2 and MLCK, which followed the activation of Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Page 7 of 13 Fig. 3 SJZD Serum promoted proliferation and inhibited cell death of TNBS-damaged Caco2 cells. a Caco-2 cells were treated with various
concentrations of 5–20% Control Serum (CS) or SJZD Serum (SJZDS) for 24 h, the effect of SJZDS on cell viability was measured using the MTT assay. b
TNBS-damaged Caco2 cells were incubated in the medium with 15% CS or 15% SJZDS for 24 h, then stained with Hoechst 33342 and observed using
a fluorescence microscope. Apoptotic cells wereindicated by the arrows (×200magnification). c TNBS-damaged Caco2 cells were treated with 15% CS
or 15% SJZDS for 24 h. The induction of apoptosis was determined byAnnexin V-FITC/PI staining assay. d Quantification of the number of apoptotic
cells. Data represented the mean ± SEM of at least three independent experiments (***P < 0.001 vs. control Caco2 cells group; ##P < 0.01
vs. TNBS-damaged Caco2 cells group) Fig. 3 SJZD Serum promoted proliferation and inhibited cell death of TNBS-damaged Caco2 cells. a Caco-2 cells were treated with various
concentrations of 5–20% Control Serum (CS) or SJZD Serum (SJZDS) for 24 h, the effect of SJZDS on cell viability was measured using the MTT assay. b
TNBS-damaged Caco2 cells were incubated in the medium with 15% CS or 15% SJZDS for 24 h, then stained with Hoechst 33342 and observed using
a fluorescence microscope. Apoptotic cells wereindicated by the arrows (×200magnification). c TNBS-damaged Caco2 cells were treated with 15% CS
or 15% SJZDS for 24 h. The induction of apoptosis was determined byAnnexin V-FITC/PI staining assay. d Quantification of the number of apoptotic
cells. Data represented the mean ± SEM of at least three independent experiments (***P < 0.001 vs. control Caco2 cells group; ##P < 0.01
vs. TNBS-damaged Caco2 cells group) NF-κBpathway [17]. To investigate whether SJZDS has
effect on NF-κB signaling in TNBS-damaged Caco2
cells, we measured the activity of NF-κB p65transcrip-
tion factor and protein expression level of NF-κB p65
and MLCK. prime the abnormal immune responses [18]. The integ-
rity of intestinal barrier can stabilize the entire intestinal
ecosystem. And tight junction (TJ) complexes have a de-
cisive effect on the maintenance of barrier integrity [6]. SJZDS upregulated the level of claudin-2 to protect
barrier function via inhibiting NF-κB signaling activation The tight junction (TJ) is a multiprotein complex has
roles in selectively regulating paracellular transport of
ions and small molecules, preventing endotoxins and
microorganisms passing through. Tight junction pro-
teins include transmembrane proteins (such as occludin
and claudins), peripheral membraneproteins (such as
ZO-1, ZO-2, ZO-3) and regulatory molecules. [19–21]. Defective TJs of the intestinal epithelium have been
shown to be an important pathogenic factor of inflam-
matory bowel diseases [22–25]. Restore the TJ function
would be possible as a new target for treatment of IBD. The cellular subfractions were prepared after 15%
SJZDS treatment for 3, 6 and 9 h. As shown in Fig. 8a,
compared to control group, the activity of NF-κB p65
was increased significantly after TNBS damage for 24 h,
after SJZDS treatment for 3 and 6 h, the activity of NF-
κB p65 was markedly decreased. As shown in Fig. 8b and c, the western blot analysis
revealed that SJZDS decreased the level of NF-κB p65
protein in TNBS-damaged Caco2 cells after treatment
for 6 and 9 h. Moreover, as shown in Fig. 8d and e, treat-
ment with SJZDS resulted downregulation in the level of
MLCK at different times. SijunziDecoction (SJZD) is one of the most famous
Traditional Chinese herbal formula and had been used
in clinical for treatment of gastrointestinal disorders over
2000 years. It could effectively attenuate nausea,vomitin-
gand diarrhea to restore the homeostasis of the digestive
tract in patients [26]. Ten active components in SJZD
had been verified, including ginsenoside Rg1, Re, Rb 1,
liquiritin, liquiritigenin, glycyrrhizic acid, atractylenolide Discussion As an important physical barrier, the intestinal epithe-
lium acts as a guard to protect intestinal tract against
bacteria, pathogens and other antigens invaded into the
intestinal mucosa and contacted with immune system to Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Page 8 of 13 Fig. 4 Effect of SJZDS on barrier function of Caco2 cells monolayers. a SJZDS significantly increased the TEER of Caco2 cells monolayers. b SJZDS
markedly attenuated the phenolsulfonphthalein flux across the epithelia monolayers. Data represented the mean ± SEM of at least three independent
experiments (*P < 0.05, **P < 0.01 vs. control Caco2 cells group; #P < 0.05, ##P < 0.01 vs. TNBS-damaged Caco2 cells group) Fig. 4 Effect of SJZDS on barrier function of Caco2 cells monolayers. a SJZDS significantly increased the TEER of Caco2 cells monolayers. b SJZDS
markedly attenuated the phenolsulfonphthalein flux across the epithelia monolayers. Data represented the mean ± SEM of at least three independent
experiments (*P < 0.05, **P < 0.01 vs. control Caco2 cells group; #P < 0.05, ##P < 0.01 vs. TNBS-damaged Caco2 cells group) I, atractylenolide II, atractylenolide III and pachymic
acid [14]. Many of these active ingredients had been
reported could reduce the symptoms of IBD. For ex-
ample, GinsenosideRg1, could restrict inflammation
progress via improving hypercoagulability and micro-
circulation, including the prolonged prothrombin time
(PT), activated partial thrombin time (APTT) and
thrombin time (TT), downregulation of thromboxane
B2 (TXB2) to alleviate the symptoms of UC [27]. The
butenolidederivatives
from
Atractylodesmacrocephala
showed potent inhibitory effects on NO production in
lipopolysaccharide-induced RAW 264.7 cells and ex-
hibited moderate cytotoxicity against HL-60 cell line
[28]. Our research group has found Carboxytmethyl-
pachymaran (CMP), extracted from Poriacocos could
ameliorate TNF-α-induced damage of intestinal epi-
thelial barrier in Caco-2 monolayers by suppression
MLCK-MLC phosphorylationsignaling pathway
[29]. However, the research on pharmacological effects and
mechanisms
of
SJZD
for
treatment
IBD
is
still
limited. TNBS has long been used as an effective inducer to
make mimic experimentally colitis model to investigate
the
pathophysiological
mechanism
of
inflammatory
bowel diseases and to determine the mechanism and ef-
ficacy of drugs. TNBS-ethanol administration could in-
duce the inflammation and structure disrupt of mucosa,
submucosa and transmural in colon, characterized by
Th1-driven inflammation [30], and has been regarded as
a similarmodel for CD [31, 32]. Discussion In this study, we provided the first evidence that SJZD
plays an important role in the protection of TNBS-
induced colitis. Our results showed that DAI scores and
microscopic scores of colon pathology were significantly
reduced in TNBS-induced colitis by SJZD. Moreover,
the most important finding was that SJZD could upregu-
late the level of claudin 2 by immunochemistry in
TNBS-induced colitis. These findings suggest that SJZD
effectively reduces inflammation and tissue damage of
colon in rats. In the 1980s, Japanese scholars first proposed the
serum contained Chinese medicine as an important
method of pharmacological study. There are many
componentsin Chinese medicine, which may be chan-
ged in the period of absorption and metabolism in
vivo [33, 34]. The componentsafter absorption and
metabolism, would be transported to the target organ
or tissue to make effect [35]. Therefore, we can use Thus, in the present experiments, we explored the ef-
fects of SJZD on TNBS-induced colitis rat model, then
we investigated TEER, permeability, expression of TJs
and NF-κB signaling way of TNBS-damaged Caco2 cells
monolayer, further to figure out intestinal barrier pro-
tective effect and mechanism of SJZD. Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Page 9 of 13 Fig. 5 Expression of tight junction proteins in Caco2 cells monolayers. Total RNA was isolated from Caco2 cells, RT-PCR was used to investigated
the level of various tight junction proteins. Data represented the mean ± SEM of at least three independent experiments (*P < 0.05 vs. control
Caco2 cells group; ##P < 0.01 vs. TNBS-damaged Caco2 cells group) Fig. 5 Expression of tight junction proteins in Caco2 cells monolayers. Total RNA was isolated from Caco2 cells, RT-PCR was used to investigated
the level of various tight junction proteins. Data represented the mean ± SEM of at least three independent experiments (*P < 0.05 vs. control
Caco2 cells group; ##P < 0.01 vs. TNBS-damaged Caco2 cells group) serum of SJZD (SJZDS) to further explore the effect
and mechanism of SJZD in vitro. SJZDS-treated Caco2 cells, TEER increased from 324 ±
1.8 to 341 ± 0.88 Ω cm2. TNBS induced drop in Caco2
TEER correlated directly with the increase in paracellu-
lar permeability. SJZDS could decrease paracellular per-
meability of Caco2 cell monolayers for 24 h. Discussion Human Caco2 intestinal epithelial cells are polarized
epithelial cells which could form apical junction com-
plexes, resulting in high electrical resistance, are useful
for studying effects of therapies on permeability. To explain the relationship between the TJ proteins and
barrier function, we analyzed a series of TJ proteins by
RT-PCR, the results showed TNBS could downregulate
the mRNA level of ZO-1, claudin-1, claudin-2 in Caco2
cells, after SJZDS treatment, the mRNA level of claudin-2
was upregulated. So we next focused the target of SJZDS
on claudin-2, the treatment of TNBS-damaged Caco2 cells
with SJZDS resulted in the increase in the expression of
claudin-2 by western blot and immunofluorence. Then, we investigated the effect and mechanism of
SJZDS on intestinal barrier protection in TNBS damaged
Caco2 cells. Our results showed TNBS could inhibit
proliferation and induce apoptosis of Caco2 cells, lead
morphological changes by Hoechst 33342. SJZDS was
able to antagonize this effect, which could promote pro-
liferation and reduce apoptosis of cells. Restoring physical barrier function is crucial in the
treatment of intestinal inflammatory disorders. After
24 h TNBS damage on Caco2 cell monolayers, TEER de-
creased from 355.2 ± 6.9 to 324 ± 1.8 Ω cm2, then in NF- κB, an important nuclear transcriptional factor, its
signaling
pathway
has
been
involved
in
numerous
physiological
processes,
including
regulation
of Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Page 10 of 13 Fig. 6 SJZDS modified the expression of claudin 2. a The figure showed a representative western blot of claudin 2 in TNBS-damaged Caco2 cells
treated by SJZDS for 24 h. b Quantification of the amounts of claudin 2relative to GAPDH for 24 h. c The figure showed a representative western
blot of claudin 2 in TNBS-damaged Caco2 cells treated by SJZDS for 36 h. d Quantification of the amounts of claudin 2relative to GAPDH for 36 h. e
The figure showed a representative western blot of claudin 2 in TNBS-damaged Caco2 cells treated by SJZDS for 48 h. f Quantification of the amounts
of claudin 2relative to GAPDH for 48 h. Data represented the mean ± SEM of at least three independent experiments (*P < 0.05, **P < 0.01 vs. control
Caco2 cells group; ##P < 0.01 vs. TNBS-damaged Caco2 cells group)
Fig. 7 Effect of SJZDS on tight junction protein claudin 2 of TNBS-damaged Caco2 cells by Immunofluorescence. Discussion Results were reported from
three independent experiments, original magnification was 400× Fig. 6 SJZDS modified the expression of claudin 2. a The figure showed a representative western blot of claudin 2 in TNBS-damaged Caco2 cells
treated by SJZDS for 24 h. b Quantification of the amounts of claudin 2relative to GAPDH for 24 h. c The figure showed a representative western
blot of claudin 2 in TNBS-damaged Caco2 cells treated by SJZDS for 36 h. d Quantification of the amounts of claudin 2relative to GAPDH for 36 h. e
The figure showed a representative western blot of claudin 2 in TNBS-damaged Caco2 cells treated by SJZDS for 48 h. f Quantification of the amounts
of claudin 2relative to GAPDH for 48 h. Data represented the mean ± SEM of at least three independent experiments (*P < 0.05, **P < 0.01 vs. control
Caco2 cells group; ##P < 0.01 vs. TNBS-damaged Caco2 cells group) Fig. 6 SJZDS modified the expression of claudin 2. a The figure showed a representative western blot of claudin 2 in TNBS-damaged Caco2 cells
treated by SJZDS for 24 h. b Quantification of the amounts of claudin 2relative to GAPDH for 24 h. c The figure showed a representative western
blot of claudin 2 in TNBS-damaged Caco2 cells treated by SJZDS for 36 h. d Quantification of the amounts of claudin 2relative to GAPDH for 36 h. e
The figure showed a representative western blot of claudin 2 in TNBS-damaged Caco2 cells treated by SJZDS for 48 h. f Quantification of the amounts
of claudin 2relative to GAPDH for 48 h. Data represented the mean ± SEM of at least three independent experiments (*P < 0.05, **P < 0.01 vs. control
Caco2 cells group; ##P < 0.01 vs. TNBS-damaged Caco2 cells group) Fig. 7 Effect of SJZDS on tight junction protein claudin 2 of TNBS-damaged Caco2 cells by Immunofluorescence. Results were reported from
three independent experiments, original magnification was 400× Fig. 7 Effect of SJZDS on tight junction protein claudin 2 of TNBS-damaged Caco2 cells by Immunofluorescence. Results were reported from
three independent experiments, original magnification was 400× Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Page 11 of 13 Fig. 8 Effect of SJZDS on NF-κB signaling pathway. a SJZDS upregulated the activity of NF-κB transcription factor p65. Discussion b The figure
showed a representative western blot of NF-κB transcription factor p65 in TNBS-damaged Caco2 cells treated by SJZDS. c Quantification
of the amounts of p65relative to PCNA. d The figure showed a representative western blot of MLCK in TNBS-damaged Caco2 cells treated
by SJZDS. e Quantification of the amounts of MLCKrelative to GAPDH. Data represented the mean ± SEM of at least three independent experiments
(*P < 0.05, **P < 0.01 vs. control Caco2 cells group; #P < 0.05, ##P < 0.01 vs. TNBS-damaged Caco2 cells group) Fig. 8 Effect of SJZDS on NF-κB signaling pathway. a SJZDS upregulated the activity of NF-κB transcription factor p65. b The figure
showed a representative western blot of NF-κB transcription factor p65 in TNBS-damaged Caco2 cells treated by SJZDS. c Quantification
of the amounts of p65relative to PCNA. d The figure showed a representative western blot of MLCK in TNBS-damaged Caco2 cells treated
by SJZDS. e Quantification of the amounts of MLCKrelative to GAPDH. Data represented the mean ± SEM of at least three independent experiments
(*P < 0.05, **P < 0.01 vs. control Caco2 cells group; #P < 0.05, ##P < 0.01 vs. TNBS-damaged Caco2 cells group) the level of MLCK. The results showed that the ex-
pression of MLCK was increased in TNBS-damaged
Caco2 cells, SJZDS decreased the level of MLCK at
different times. inflammation response and innate immunity, and there-
fore may be participation in IBD in multiple ways [36]. The increased expression of NF-κB in the acute phase of
colitis had been proved and is consistent with its role in
the onset of experimental colitis [37]. SJZDS could sig-
nificantly inhibit NF- κB p65 activity and expression in
nuclear after treatment for 3 h and 6 h, make it impos-
sible to block downstream cytokine and inflammatory
factors expression. References 26. Yang L, Yang J, Cai Z. Chemical investigation on Sijunzi decoction and its
two major herbs Panax ginseng, and Glycyrrhizauralensis, by LC/MS/MS. J
Pharm Biomed Anal. 2006;41(5):1642–7. 1. Podolsky DK. Inflammatory bowel disease (1). N Engl J Med. 1991;325(13):
928–37. 1. Podolsky DK. Inflammatory bowel disease (1). N Engl J Med. 1991;325(13):
928–37. 2. Yamamoto-Furusho JK, Podolsky DK. Innate immunity in inflammatory 27. Hao WW, Wen HZ, Ma GT, Tang ZP, He XY, Li J, Li NN, Liu YT. Ginsenoside-
Rg1regulates blood coagulation in DSS-induced colitis mice. Chin J Integr
Trad Med Dig. 2013;21(5):238–42. 27. Hao WW, Wen HZ, Ma GT, Tang ZP, He XY, Li J, Li NN, Liu YT. Ginsenoside-
Rg1regulates blood coagulation in DSS-induced colitis mice. Chin J Integr
Trad Med Dig. 2013;21(5):238–42. y
y
y
bowel disease. N Engl J Med. 1991;325(14):131–6. bowel disease. N Engl J Med. 1991;325(14):131–6. 3. Fiocchi C. Inflammatory bowel disease: etiology and pathogenesis. World J
Gastroenterol. 1998;115(1):182–205. 28. Guo F, Li Z, Xu X, Wang K, Shao M, Zhao F, Wang H, Hua H, Pei Y, Bai J. Butenolide derivatives from the plant endophytic fungus Aspergillusterreus. Fitoterapia. 2016;113:44–50. 28. Guo F, Li Z, Xu X, Wang K, Shao M, Zhao F, Wang H, Hua H, Pei Y, Bai J. Butenolide derivatives from the plant endophytic fungus Aspergillusterreus
Fitoterapia. 2016;113:44–50. 4. Podolsky DK. Inflammatory bowel disease. N Engl J Med. 2002;347(6):417–29. 4. Podolsky DK. Inflammatory bowel disease. N Engl J Med. 2002;347(6):417–29. 5. Bouma G, Strober W. The immunological and genetic basis of inflammatory
bowel disease. Nat Rev Immunol. 2003;3(7):521–33. 29. Zhang J, Lu Y, Wei J, Li L, Han L. Protective effect of
carboxytmethylpachymaran on TNF-α-induced damage in Caco-2 cell
monolayers. Int J BiolMacromol. 2016;93(Pt A):506–11. 29. Zhang J, Lu Y, Wei J, Li L, Han L. Protective effect of
carboxytmethylpachymaran on TNF-α-induced damage in Caco-2 cell
monolayers. Int J BiolMacromol. 2016;93(Pt A):506–11. 6. Mitic LL, Van Itallie CM, Anderson JM. Molecular physiology and
pathophysiology of tight junctions I. Tight junction structure and function:
lessons from mutant animals and proteins. Am J Physiol Gastrointest Liver
Physiol. 2000;279(2):G250–4. 6. Mitic LL, Van Itallie CM, Anderson JM. Molecular physiology and
pathophysiology of tight junctions I. Tight junction structure and function:
lessons from mutant animals and proteins. Am J Physiol Gastrointest Liver
Physiol. 2000;279(2):G250–4. 30. Neurath M, Fuss I, Strober W. TNBS-colitis. Int Rev Immunol. 2000;19(19):51–62 31. Jurjus AR, Khoury NN, Reimund JM. Consent for publication
Not applicable. 22. Hermiston ML, Gordon JI. Inflammatory bowel disease and adenomas in
mice expressing a dominant negative N-cadherin. Science. 1995;270(5239):
1203–7. 23. Costantini TW, Peterson CY, Kroll L, Loomis WH, Eliceiri BP, Baird A, Bansal V. Coimbra RRole of p38 MAPK in burn-induced intestinal barrier breakdown. J
Surg Res. 2009;156(156):64–9. Abbreviations DOR-β-arrestin1-Bcl2 signal transduction pathway in a rat model CD: Crohn’s disease; H&E: Hematoxylin and eosin; IBD: Inflammatory bowel
disease; SASP: Salazosulfapyridine; SJZD: SijunziDecoction; CD: Crohn’s disease; H&E: Hematoxylin and eosin; IBD: Inflammatory bowel
disease; SASP: Salazosulfapyridine; SJZD: SijunziDecoction;
SJZDS: SijunziDecoction serum; TEER: Transepithelial electrical resistance;
TJ: Tight junction.; TNBS: 6-trinitrobenzene sulfonic acid; UC: Ulcerative coliti
ZO1: Zonulaoccludens 1 ulcerative colitis. J Ethnopharmacol. 2014;154(1):88–97. 13. Ji YF, Wang RJ, Li XB. Research progress on chemical constituents and
pharmacological effects of Sijunzi decoction. ChinTrad Herb Dr. 2016;47(5):
837–43. SJZDS: SijunziDecoction serum; TEER: Transepithelial electrical resistance;
TJ Ti ht j
ti
TNBS 6 t i it
b
lf
i
id UC Ul
ti
liti j
;
p
;
TJ: Tight junction.; TNBS: 6-trinitrobenzene sulfonic acid; UC: Ulcerative colitis
ZO1: Zonulaoccludens 1 14. An K, Jin-Rui G, Zhen Z, Xiao-Long W. Simultaneous Quantification of Ten
Active Components in Traditional Chinese Formula Sijunzi Decoction Using
a UPLC-PDA Method. J Anal Methods Chem. 2014;2014(11):570359. Ethics approval and consent to participate All the procedures were conducted in strict accordance with the PR
Chinalegislation on the use and care of laboratory animals, and with the
guidelines established by the Institute for Experimental Animals of Guangzhou
University of TCM, and were approved by the university ethical committee for
animalexperiments. 24. Yasuda T, Takeyama Y, Ueda T, Shinzeki M, Sawa H, Nakajima T, Kuroda Y. Breakdown of intestinal mucosa via accelerated apoptosis increases
intestinal permeability in experimental severe acute pancreatitis. J Surg Res. 2006;135(135):18–26. University of TCM, and were approved by the university ethical committee for
animalexperiments. Received: 16 March 2016 Accepted: 19 December 2016 25. Seo GS, Jiang WY, Park PH, Sohn DH, Cheon JH, Lee SH. Hirsutenone
reduces deterioration of tight junction proteins through EGFR/Akt and
ERK1/2 pathway both converging to HO-1 induction. Biochem Pharmacol. 2014;90(2):115–25. Received: 16 March 2016 Accepted: 19 December 2016 Conclusions In this study, our results demonstrate that SJZD ame-
liorates
the
severity
of
TNBS-induced
colitis
and
downregulates the level of claudin-2 in colonic tis-
sues. In vitro data showsSJZDS promotes proliferation
and inhibits apoptosis ofTNBS-damaged Caco2 cells. In Caco2 cell monolayers, SJZDS increases theTEER
andreduces permeability after TNBS damage, we finds
that the barrier protective effect of SJZDS is mediated
through claudin-2 and NF-κB pathway, including up-
regulation of claudin-2, decreased activity of NF-κB
p65, reduced expression of NF-κB p65 and MLCK. Taken together, these findings provide evidence that,
SJZD, is a potential protective agent ofintestinal bar-
rier function. Further investigation into the mechan-
ism of SJZD on intestinal barrier as well as in vivo
research is required. p
TJs are closely link to various intracellular signaling
molecules and are regulated by multiple signal trans-
duction pathways. Myosin light chain kinase (MLCK)
is classically known to be required for the contraction
of actomyosin via the phosphorylation of myosin light
chain (MLC) [38] and has also been reported to be
expressed in the human intestinal tissue with IBD
[39]. It is also essential to regulation the permeability
of epithelial barrier and affects production of inflam-
matory cytokine, such as TNFα, in the inflamed intes-
tinal tissues. Inhibition of MLCK can attenuate TNFα
induced nuclear translocation of p65 and phosphoryl-
ation of IκB, which indicates MLCK play a role in in-
ducing activation of NF-κB [40]. We next investigated Page 12 of 13 Page 12 of 13 Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 DOR-β-arrestin1-Bcl2 signal transduction pathway in a rat model of
ulcerative colitis. J Ethnopharmacol. 2014;154(1):88–97. Funding
Th
k This work was supported by grants from the Natural Science Foundation of
China (Grant No.81202635)and Foundation of Guangdong Provincial Bureau
of Chinese Medicine (20151244). 17. Nishida M, Yoshida M, Nishiumi S, Furuse M, Azuma T. Claudin-2 regulates
colorectal inflammation via myosin light chain kinase-dependent signaling. Dig Dis Sci. 2013;58(6):1546–59. Availability of data and materials Data are all contained within the paper. 18. Sierro F, Dubois B, Coste A, Kaiserlian D, Kraehenbuhl JP, Sirard JC. Flagellin
stimulation of intestinal epithelial cells triggers CCL20-mediated migration
of dendritic cells. Proc Natl Acad Sci USA. 2001;98(24):13722–7. Acknowledgments 15. Zhao SY, Wang CJ. Influence of Sijunzi decoction’s on T lymphocyte subsets
of gastric mucosa in rats with spleen-deficiency syndrome. Chin J Integr
Trad Med Dig. 2014;22(4):204–6. We thank all of our colleagues in our team at the Guangdong Provincial
Hospital of Chinese Medicine for their excellent assistance of this study. 16. Klingberg TD, Pedersen MH, Cencic A, Budde BB. Application of
measurements of transepithelial electrical resistance of intestinal epithelial
cell monolayers to evaluate probiotic activity. Appl Environ Microbiol. 2005;
71(11):7528–30. Authors’ contributions Conception of idea and research design: LH. Conduct of research and
experimentation: YL, HJL, JWZ. Data analyses: YL, HJL. Drafting of manuscript:
YL. Review and approval of final manuscript: LH, JAW, JS. All authors read
and approved the final manuscript. 19. Garrote JA, Gómez-González E, Bernardo D, Arranz E, Chirdo F. Celiac
disease pathogenesis: the proinflammatory cytokine network. J Pediatr
Gastroenterol Nutr. 2008;47 Suppl 1:S27–32. 20. Ivanov AI, Parkos CA, Asma N. Cytoskeletal regulation of epithelial barrier
function during inflammation. Am J Pathol. 2010;177(2):512–24. Competing interests
The authors declare that they have no competing interests. Competing interests 21. Hartsock A, Nelson WJ. Adherens and tight junctions: structure, function
and connections to the actin cytoskeleton. Biochimica EtBiophysica Acta
Biomembranes. 2008;1778(3):660–9. Competing interests
The authors declare that they have no competing interests. References Animal models of inflammatory bowel
disease.J PharmacolToxicol. Methods. 2004;50(2):81–92. 7. Turner JR. Intestinal mucosal barrier function in health and disease. Nat Rev
Immunol. 2009;9(11):799–809. 7. Turner JR. Intestinal mucosal barrier function in health and disease. Nat Rev
Immunol. 2009;9(11):799–809. 8. Pithadia AB, Jain S. Treatment of inflammatory bowel disease (IBD). Pharmacol Rep. 2011;63(3):629–42. 32. Brenna Ø, Furnes MW, Drozdov I, van BeelenGranlund A, Flatberg A, Sandvik
AK, Zwiggelaar RT, Mårvik R, Nordrum IS, Kidd M, Gustafsson BI. Relevance
of TNBS-colitis in rats: a methodological study with endoscopic, Histologic
and transcriptomic [corrected] characterization and correlation to IBD. PLoS
One. 2013;8(1):e54543. 9. Triantafyllidi A, Xanthos T, Papalois A, Triantafillidis JK. Herbal and plant
therapy in patients with inflammatory bowel disease. Ann Gastroenterol. 2015;28(2):210–20. 9. Triantafyllidi A, Xanthos T, Papalois A, Triantafillidis JK. Herbal and plant
therapy in patients with inflammatory bowel disease. Ann Gastroenterol. 2015;28(2):210–20. 33. Dou ZH, Luo L, Chen JY, AN LP, YANG AH. Analysis of lignans in serum
containing drug of Schisandrachinensis by UPLC-MS/MS. Chin J Clin
Pharmacol. 2013;29(3):215–8. 10. Wang X, Fan F, Cao Q. Modified Pulsatilla decoction attenuates oxazolone-
induced colitis in mice through suppression of inflammation and epithelial
barrier disruption. Mol Med Rep. 2016;14(2):1173–9. 34. Wang X, Sun W, Sun H, Lv H, Wu Z, Wang P, Liu L, Cao H. Analysis of the
constituents in the rat plasma after oral administration of Yin Chen Hao
Tang by UPLC/Q-TOF-MS/MS. J Pharm Biomed Anal. 2008;46(3):477–90. 11. Xu BL, Zhang GJ, Ji YB. Active components alignment of Gegenqinlian,
decoction protects ulcerative colitis by attenuating inflammatory and
oxidative stress. J Ethnopharmacol. 2015;13(162):253–60. 35. Wang P, Liang Y, Zhou N, Chen B, Yi L, Yu Y, Yi Z. Screening and analysis of
the multiple absorbed bioactive components and metabolites of 12. Fan H, Liu XX, Zhang LJ, Hu H, Tang Q, Duan XY, Zhong M, Shou ZX. Intervention effects of QRZSLXF, a Chinese medicinal herb recipe, on the Page 13 of 13 Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 dangguibuxue decoction by the metabolic fingerprinting technique and
liquid chromatography/diode-array detection mass spectrometry. Rapid
Commun Mass Spectrum. 2007;21(2):99. dangguibuxue decoction by the metabolic fingerprinting technique and
liquid chromatography/diode-array detection mass spectrometry. Rapid
Commun Mass Spectrum. 2007;21(2):99. 36. Anselmi L, Huynh J, Duraffourd C, Jaramillo I, Vegezzi G, Saccani F, Boschetti
E, Brecha NC, De Giorgio R, Sternini C. 40.
Wadgaonkar R, Linz-McGillem L, Zaiman AL, Garcia JG. Endothelial cell
myosin light chain kinase (MLCK) regulates TNFalpha-induced NFkappaB
activity. J Cell Biochem. 2005;94(2):351–64. 39.
Blair SA, Kane SV, Clayburgh DR, Turner JR. Epithelial myosin light chain
kinase expression and activity are upregulated in inflammatory bowel
disease. Lab Invest. 2006;86(2):191–201. 37.
Funakoshi T, Yamashita K, Ichikawa N, Fukai M, Suzuki T, Goto R, Oura T,
Kobayashi N, Katsurada T, Ichihara S, Ozaki M, Umezawa K, Todo S. A novel
NF-κB inhibitor, dehydroxymethylepoxyquinomicin, ameliorates
inflammatory colonic injury in mice. J Crohns Colitis. 2012;6(2):215–25. 38.
Liu X, Xu J, Mei Q, Han L, Huang J. Myosin light chain kinase inhibitor
inhibits dextran sulfate sodium-induced colitis in mice. Dig Dis Sci. 2013;
58(1):107–14. Lu et al. BMC Complementary and Alternative Medicine (2017) 17:35 References Activation of μ opioid receptors
modulates inflammation in acute experimental colitis. Neurogastroenterol
Motil. 2015;27(4):509–23. 37. Funakoshi T, Yamashita K, Ichikawa N, Fukai M, Suzuki T, Goto R, Oura T,
Kobayashi N, Katsurada T, Ichihara S, Ozaki M, Umezawa K, Todo S. A novel
NF-κB inhibitor, dehydroxymethylepoxyquinomicin, ameliorates
inflammatory colonic injury in mice. J Crohns Colitis. 2012;6(2):215–25. 38. Liu X, Xu J, Mei Q, Han L, Huang J. Myosin light chain kinase inhibitor
inhibits dextran sulfate sodium-induced colitis in mice. Dig Dis Sci. 2013;
58(1):107–14. 39. Blair SA, Kane SV, Clayburgh DR, Turner JR. Epithelial myosin light chain
kinase expression and activity are upregulated in inflammatory bowel
disease. Lab Invest. 2006;86(2):191–201. 40. Wadgaonkar R, Linz-McGillem L, Zaiman AL, Garcia JG. Endothelial cell
myosin light chain kinase (MLCK) regulates TNFalpha-induced NFkappaB
activity. J Cell Biochem. 2005;94(2):351–64. • We accept pre-submission inquiries
• Our selector tool helps you to find the most relevant journal
• We provide round the clock customer support
• Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research
Submit your manuscript at
www.biomedcentral.com/submit
Submit your next manuscript to BioMed Central
and we will help you at every step: • We accept pre-submission inquiries
• Our selector tool helps you to find the most relevant journal
• We provide round the clock customer support
• Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research
Submit your manuscript at
www.biomedcentral.com/submit
Submit your next manuscript to BioMed Central
and we will help you at every step: • We accept pre-submission inquiries
• Our selector tool helps you to find the most relevant journal
• We provide round the clock customer support
• Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research
Submit your manuscript at
www.biomedcentral.com/submit
Submit your next manuscript to BioMed Central
and we will help you at every step: Submit your next manuscript to BioMed Central
and we will help you at every step:
|
https://openalex.org/W3190165006
|
http://edoc.mdc-berlin.de/20801/1/20801oa.pdf
|
English
| null |
Analysis of correlation and ionization from pair distributions in many-electron systems
|
The European physical journal plus
| 2,021
|
cc-by
| 9,279
|
Eur. Phys. J. Plus (2021) 136:763
https://doi.org/10.1140/epjp/s13360-021-01747-8 Eur. Phys. J. Plus (2021) 136:763
https://doi.org/10.1140/epjp/s13360-021-01747-8 Regular Article Analysis of correlation and ionization from pair
distributions in many-electron systems S. López-Rosa1,2,a
, J. C. Angulo2,3,b, A. L. Martín4, J. Antolín2
1 Departamento de Física Aplicada II, Universidad de Sevilla, 41012 Sevilla, Spain
2 Instituto Carlos I de Física Teórica y Computacional, Universidad de Granada, 18071 Granada, Spain
3 Departamento de Física Atómica, Molecular y Nuclear, Universidad de Granada, 18071 Granada, Spain
4 Berlin Institute for Medical Systems Biology, Max Delbrück Center for Molecular Medicine in the Helmholtz
Association, 10115 Berlin, Germany S. López-Rosa1,2,a
, J. C. Angulo2,3,b, A. L. Martín4, J. Antolín2
1 Departamento de Física Aplicada II, Universidad de Sevilla, 41012 Sevilla, Spain
2 Instituto Carlos I de Física Teórica y Computacional, Universidad de Granada, 18071 Granada, Spain
3 Departamento de Física Atómica, Molecular y Nuclear, Universidad de Granada, 18071 Granada, Spain
4 Berlin Institute for Medical Systems Biology, Max Delbrück Center for Molecular Medicine in the Helmholtz
Association, 10115 Berlin, Germany Received: 21 January 2021 / Accepted: 10 July 2021
© The Author(s) 2021 Received: 21 January 2021 / Accepted: 10 July 2021
© The Author(s) 2021 Abstract Jensen–Shannon divergence is used to quantify the discrepancy between the
Hartree–Fock pair density and the product of its marginals for different N-electron systems,
enclosing neutral atoms (with nuclear charge Z = N) and singly-charged ions (N = Z ±1). This divergence measure is applied to determine the interelectronic correlation in atomic
systems. A thorough study was carried out, by considering (i) both position and momentum
conjugated spaces, and (ii) systems with a nuclear charge as far as Z = 103. The correlation
among electrons was measured by comparing, for an arbitrary system, the double-variable
electron-pair density with the product of the respective one-particle densities. A detailed
analysis throughout the Periodic Table highlights the relevance not only of weightiness for
the systems considered, but also of their shell structure. Besides, comparative computations
between two-electron densities of different atomic systems (neutrals, cations, anions) quan-
tify their dissimilarities, patently governed by shell-filling patterns throughout the Periodic
Table. a e-mail: slopezrosa@us.es (corresponding author)
b e-mail: angulo@ugr.es
0123456 1 Introduction The two-electron density Γ (r1, r2) is the probability density of any two electrons that are
located at radii r1 and r2, respectively, and is a convenient starting point to study the spatial
interaction between two electrons in an explicit manner [1]. This density provides a quantum-
mechanical description of the distribution of electron pairs in the space and therefore rep-
resents a central tool, in quantum chemistry, to study chemical binding [2] or electronic
structure and correlation in molecular systems [3,4]. On the other hand, the exact interaction energy of a many-electron system is determined
by the electron pair density, but is not well-approximated in standard Kohn–Sham density
functional models [5,6]. Consequently, pair density functional theory is regarded as not only
an extension of the density functional theory but a reduced density matrix theory as well
[7,8]. The main feature of the pair density functional theory is that the pair density, as a basic a e-mail: slopezrosa@us.es (corresponding author)
b e-mail: angulo@ugr.es
0123456 123 763
Page 2 of 16 763
Page 2 of 16 Eur. Phys. J. Plus (2021) 136:763 763 variable, essentially contains more information on the electron-electron interaction than the
electron density [9,10]. The expectation value of an arbitrary two-particle operator can be obtained using only
the electron pair density. In addition to this merit, the pair density functional theory has
another, which is related to the expansion of the approximate functional. That is to say, in
the pair density functional theory, the only required selection is the approximate form of the
kinetic energy functional, because the exchange-correlation energy functional can rigorously
be expressed by means of the pair density [11]. As a recent example multiconfiguration pair
density functional theory has been developed, as a way to combine the advantages of both
wave function and density functional theories to provide a better treatment of strongly cor-
related systems, and its results have been compared to those of some Kohn–Sham density
functionals, including both traditional and modern functionals [12,13]. Additional applica-
tions have been performed recently with a diversity of molecular systems, on the basis of
their one- and two-electron RDMs (reduced density matrices), which comparison provides
measures of electron-correlation and entanglement among systems. These measures display
their sensitivity to the presence of electric fields [14]. 1 Introduction Therefore, given the above, it would be logical to assume that two-electron densities for
atoms and molecules have been studied both intensively and extensively. However, this is not
the case. The effort invested in interpretative studies of electron pair densities has been small
as compared to that expended on such studies for one-electron densities. Indeed, there is a
large amount of information coded inside monoelectronic densities; however, these densities
do not give us any hint on how the position (or the momentum) of an electron conditions the
positions (or the momenta) of the others. Besides it is worthy to note that within the Waller–
Hartree theory, the two-electron density is related to the total X-ray scattering intensity which
is accessible by experimental measurements [15,16]. One of the main difficulties rests in the interpretation of the electron pair density. Visual
representation is a powerful analytical tool; however, being a function of 6 coordinates,
graphical representation of the pair density is not feasible and manipulations to the two-
electron density must be carried out to extract useful information. One such manipulation
gives place to the intracule and extracule densities and many relevant results have been
obtained in this framework [17–21]. It is important to note that these densities consider the
pairs of electrons as a whole thing. Other simplification consists in dealing with spherically
averaged electron pair densities. Consequently, as a result of this reduction in dimensionality,
some two-electron information is lost in the transformation. The concepts of uncertainty, randomness, disorder and delocalization are basic ingredients
in the study, within an information-theoretic framework, of relevant structural properties for
many different probability distributions appearing as descriptors of several chemical and
physical systems and processes. The relevancy of these concepts has motivated new studies
that pursue quantification, giving rise to a variety of density functionals, such as Shannon
entropy [22], Fisher information [23], disequilibrium [24], complexity [25], and many others. These information measures have been widely employed for describing the information
content and behavior in a great variety of fields [26,27] and, in particular, for the study of
many-electron systems [27–34] Following the usual procedures carried out within information theory for quantifying
the uncertainty or disorder of individual distributions, some extensions have been made in
order to introduce the concepts of ‘distance’ and ‘divergence’ between two distributions, as
comparative measures of their dissimilarity. 1 Introduction Jensen–Shannon divergence (JSD) is a powerful divergence measure that represents the
difference between the Shannon entropy of the mean density and the mean value of the indi- 123 Eur. Phys. J. Plus (2021) 136:763 Page 3 of 16
763 vidual entropies [35–37]. Its non-negativity arises from the convex character of the Shan-
non entropy functional. This comparative measure has found deep applications in statistics
and many other fields, including entanglement characterization [38], analysis of symbolic
sequences or series, and in particular in the study of segmentation of DNA sequences [39]. JSD and some effective generalizations have been used for the study and comparison of
one-electron densities [40–44]. For the first time, to the best of our knowledge, this infor-
mation measure of dissimilarity is applied to the study of two-electron density functions. Let us emphasize, however, recent applications of strongly related functionals (e.g. mutual
information, cumulative residual entropies, interaction information) for uncoupled and inter-
acting harmonic oscillators [45,46]. Additional applications include the cumulant part of
two-electron RDMs, for the detection of van der Waals interactions between specific parts
of molecular systems [47]. Other measures enclosing also the gradient of the density, as the
relative Fisher information, were considered for the study of specific one-particle quantum
systems (Morse potential, isotropic oscillator or hydrogen-like atoms) [48,49]. In this paper, small correlation effects in atoms are analyzed, on the basis of discrepancies
between the Hartree–Fock one- and two-electron densities. In doing so, comparative den-
sity functionals are employed as quantifiers of dissimilarity among the above densities. So
differences arising from the exchange term are emphasized, as the main cause of them. A
detailed analysis throughout the Periodic Table reveals the relevance not only of weightiness
for the systems considered, but also of their shell structure. Besides, comparative computa-
tions between two-electron densities of different atomic systems (neutrals, cations, anions)
quantify the important differences patently governed by shell-filling patterns throughout the
Periodic Table. Thereby, we use an information-theoretic functional that has shown its univer-
sality and versatility in other computations: the Jensen–Shannon divergence (JSD) [35–37]. The organization of the paper is the following: in the next section, two-particle densities
are defined and presented for atomic systems and also the entropic divergence measure (JSD)
we are going to use for comparing them. Numerical computations and results are presented
in Sect. 3. 1 Introduction Finally, conclusions and open problems or future work are briefly described in the
last section. 2 Electron pair densities and related measures In terms of the N-electron wave function Ψ , and its Fourier transform Φ, the two-electron
densities are defined as Γ (r1, r2) =
Ψ (x1, x2, . . . , xN)Ψ ∗(x1, x2, . . . , xN)dσ1dσ2dx3 . . . dxN
(1)
the position space, and
Π(
)
Φ(
)Φ∗(
)d
d
d
d
(2) Γ (r1, r2) =
Ψ (x1, x2, . . . , xN)Ψ ∗(x1, x2, . . . , xN)dσ1dσ2dx3 . . . dxN
(1) Γ (r1, r2) =
Ψ (x1, x2, . . . , xN)Ψ ∗(x1, x2, . . . , xN)dσ1dσ2dx3 . . . dxN
(1)
the position space and (1) in the position space, and in the position space, and Π(p1, p2) =
Φ(y1, y2, . . . , yN)Φ∗(y1, y2, . . . , yN)dσ1dσ2dy3 . . . dyN
(2) (2) in the momentum space. The variables xi = riσi and yi = piσi are combined coordinates
which include the spin. It is well known that the physical meaning of these densities regards
the probability of finding an electron with a compatible state within the region r1dr1 if there
is another electron with allowed/compatible quantum numbers within the region r2dr2, and
similarly regarding the momentum regions p1dp1 and p2dp2. These densities are directly
related to the electron correlations, as the compatibility of an electron state is in organically in the momentum space. The variables xi = riσi and yi = piσi are combined coordinates
which include the spin. It is well known that the physical meaning of these densities regards
the probability of finding an electron with a compatible state within the region r1dr1 if there
is another electron with allowed/compatible quantum numbers within the region r2dr2, and
similarly regarding the momentum regions p1dp1 and p2dp2. These densities are directly
related to the electron correlations, as the compatibility of an electron state is in organically 123 763
Page 4 of 16 Eur. Phys. J. Plus (2021) 136:763 763 determined by the compatibility of its state with those of the others. They naturally give us
quantifiers of correlation between electrons. 2 Electron pair densities and related measures The two-electron densities are going to be calculated by the Hartree–Fock approach and
can be expressed as follows [50]: Γ (r1, r2) =
1
N −1 [Nρ(r1)ρ(r2) −Γx(r1, r2)]
(3) (3) in the position space, and Π(p1, p2) =
1
N −1
Nγ (p1)γ (p2) −Πx(p1, p2)
(4) (4) in the momentum space, respectively. The functions ρ(ri) and γ (pi) are the one-electron den-
sities and Γx(r1, r2) and Πx(p1, p2) are the exchange densities in the position and momentum
space respectively. Let us remark that, using Hartree–Fock functions, we are studying the
Fermi correlation between same-spin electrons which arises from the antisymmetry of the
wave function. In this sense, the term electron correlation, mentioned before, alludes to the
statistical correlation. This point must be clarified in order to not lead to confusion if one
considers a correlated system as one beyond the Hartree–Fock approximation (the Löwdin
definition of correlation energy). It should be noticed that the one-particle densities are marginal distributions of the two-
particle ones, because the former are obtained from the latter by integration, as ρ(r1) =
Γ (r1, r2)dr2 and γ (p1) =
Π(p1, p2)dp2. Thus, the information content of the marginals
is not enough to determine the double-variable distribution they are coming from. Different definitions of distance between two arbitrary distributions f (x) and g(x) appear,
where both are defined over the same domain. The interest in finding appropriate tools for
‘comparing’ two different systems or processes in terms of distribution functions is justified
by the strong connection between some of the above mentioned information measures and
many relevant physical and chemical properties of those systems. The term ‘distance’ should
be understood as a measure of how dissimilar the two functions are, not necessarily verifying
all mathematical properties required for a true distance, as the usual quadratic distance does. However, all of them keep as main distance properties the non-negativity, the symmetry
(invariance under exchange of functions) and saturation (minimal zero value only for identical
distributions). The relative entropy or Kullback–Leibler divergence (KL) [35] is one of the pioneering
global measures of the difference between two arbitrary uni- or multivariate probability
distributions. 2 Electron pair densities and related measures It expresses the amount of information supplied by the data for discriminating
among the distributions, being a ‘directed divergence’ and therefore not symmetric: K L ( f, g) =
f (x) ln f (x)
g(x) dx
(5) (5) Its applications for different procedures in obtaining minimum cross entropy estimations
and the determination of atomic and molecular properties [29], among others, make it to
constitute an essential tool within the information theory. Won and You introduced, some-
what implicity, a closely related information measure between two or more distributions, the
Jensen–Shannon divergence (JSD) [51,52]: J SD( f, g) = 1
2
K L
f, f + g
2
+ K L
g, f + g
2
. (6) (6) Consequently, JSD represents the mean dissimilarity (understood in terms of the KL
measure) of each density respect to the mean one. Consequently, JSD represents the mean dissimilarity (understood in terms of the KL
measure) of each density respect to the mean one. 123 Eur. Phys. J. Plus (2021) 136:763 Page 5 of 16
763 This measure is strongly related to the Shannon entropy, which is defined for an arbitrary
uni- or multivariate probability distribution as S( f ) = −
dx f (x) ln f (x). Shannon entropy
is well known to play a relevant role within an information-theoretic framework because it
constitutes a measure of spreading of the density over its domain [53]. Attending to the JSD
definition given above, the Jensen–Shannon divergence can be also expressed in terms of the
Shannon entropy as J SD( f, g) = S
f + g
2
−1
2 [S( f ) + S(g)] . (7) (7) The JSD is characterized for quantifying the “Shannon entropy excess” of a couple of
distributions with respect to the mixture of their respective entropies (mean density), defined
in Eq. (7) as a particular case of the so-called weighted generalized divergences [54] with
identical weights 1/2. Apart from preserving the global character of the Shannon entropy, the
JSD possesses the main properties required for a measure to be interpreted as an informational
distance. In particular, its non-negativity is a consequence of the convexity of the S( f )
functional. 2 Electron pair densities and related measures Other important advantages of this divergence are that: (i) does not require the
condition of absolute continuity for the probability distributions involved, (ii) weights of each
density can be different from 1/2 as appearing in Eq. (7), and (iii) can be generalized for
an arbitrary number of distributions. During the past years researchers have been interested
in work towards parametric generalizations of these classical measures of information. The
JSD also admits other kind of generalizations, as will be shown elsewhere. In this work we intend to perform an exhaustive analysis of atomic systems using JSD,
on the basis of their two-electron densities, in the following two ways: (i) Differences between the pair distribution Γ (r1, r2) and the product of marginals
ρ(r1)ρ(r2) reveal the existence of some interconnection between its variables and quan-
tify the effects from the exchange density on the systems under consideration. With JSD
those differences are obtained for atomic systems (neutrals and ions), in both conjugated
spaces (Sect. 3.1). (ii) The ionization of a neutral system, by adding or removing an electron, modifies its
structural characteristics. This was studied by means of the JSD between the one-particle
distributions of the neutral and the ionized systems [55]. Here, a step further is taken,
dealing with divergences among electron-pair densities, also in position and momentum
spaces (Sect. 3.2). 3 Numerical results Divergence measures allow us to quantify the differences between general systems, in par-
ticular by means of electron atomic densities. In this section we intend to employ Jensen–
Shannon divergence in order to analyse the particularities and characteristics of two-electron
and one-electron densities. The Near-Hartree–Fock wavefunctions of Koga et al [56,57] are
employed. Functions normalized to unity will be managed in what follows, for the sake of
their interpretation as probability distributions. 3.1 Interelectronic correlation analysis 3.1 Interelectronic correlation analysis We will compute the Jensen–Shannon divergence between two-electron density and the
product of one-electron densities, being one of the main reasons its property that its value is
zero when the systems compared have no correlation. Thus, the Jensen–Shannon divergence 123 763
Page 6 of 16 Eur. Phys. J. Plus (2021) 136:763 Fig. 1 Jensen–Shannon divergence in position space (JSDr ) between the electron pair density and the product
of monoelectronic densities, for neutral atoms with number of electrons N = 3−103. Labels are for systems
with closed shells (red), closed subshells (blue) or anomalous shell-filling (green). Atomic units (a.u.) are used Fig. 1 Jensen–Shannon divergence in position space (JSDr ) between the electron pair density and the product
of monoelectronic densities, for neutral atoms with number of electrons N = 3−103. Labels are for systems
with closed shells (red), closed subshells (blue) or anomalous shell-filling (green). Atomic units (a.u.) are used acts as a measure of the disparity between two- and one-electron densities. These comparisons
will be carried out on double variable distributions in both position (Γ (r1, r2) and ρ(r1)ρ(r2)
respectively) and momentum (Π(p1, p2) and γ (p1)γ (p2) respectively) spaces. Let us remark
that in this case, the last term in the JSD (see Eq. (7)) can be expressed as −
S(Γ )
2
+ S (ρ)
in
position space, and similarly for the momentum space. The systems under study are neutral
atoms with a number of electrons N ≤103 (nuclear charge Z = N), and singly-charged
ions (Z = N ± 1) with N ≤54. Firstly,wecomputedtheJensen–Shannondivergencebetweentheelectronpairdensityand
the product of the corresponding monoelectronic densities for neutral systems N = 2−103
in position space. The obtained results are shown in Fig. 1. It can be observed the decreasing
trend of JSD when the nuclear charge of the system increases, which indicates that the
correlation decreases as the number of electrons increases. This was an expected result, as
the number of electrons is known to determine how strong or weak is the relevance of the
correlation between the mentioned electrons. Thus, in the case that we have information
about the position of an electron within a group of three, this knowledge becomes much
more relevant than if we have a group of thirty electrons, when the information about just
one does not tell us much more about the other twenty nine. – Closedshells: N = 10, 18, 36, 54, 86.Thusnoblegasesappearasminimasystematically.
The same was observed in the K L divergence for systems up to N ≤36 [58]. 3.1 Interelectronic correlation analysis Nevertheless, additional remarks are in order. Comparing curves in position and momen-
tum space, it is observed that (i) the aforementioned ‘changes of slope’ at N = 24, 29 in
position space are displayed as maxima in momentum space; (ii) previous minimum N = 42
translates now to N = 43, also with outer d-subshell; (iii) instead of minimum N = 80 it is
now observed the maximum N = 78. No extrema appear in the heavy systems region. Notice that just mentioned systems N = 24, 29, 42, 43, 46, 78 have a non-filled inner s-
subshell and an outermost d-subshell (the inner one is empty for the system N = 46, clearly
highlighted in both figures). As observed in Ref. [58] for KL, most JSD values are higher in position than in momentum
space. The opposite applies for a few systems with anomalous shell-filling. Particularly
relevant are the cases N = 24, 29 with half-filled 4s inner orbital and outermost 3d subshell
(half-filled and completely filled, respectively). Their JSDp (Fig. 2) are greater than the
respective JSDr (Fig. 1) by 17% and 25%. Such amount belongs to the range 8−13% for
heavier systems with similar anomalies: N = 41, 42, 44, 45, 47 (subshells 5s4d), and N =
78, 79 (subshells 6s5d), all them with an inner half-filled s-orbital. It is concluded that, for
most systems, correlation effects are weaker in momentum space, the opposite being true
for the aforementioned anomalous systems, in a slight fashion with few exceptions. These
results support and extend the conclusions obtained in Ref. [58] for systems up to N = 36,
regarding the relative behaviors in position and momentum spaces. The above discussion was focused on neutral atoms and their properties, but in what
follows we will consider also some of their corresponding ions. Let us now use the Jensen–
Shannon divergence, JSD, in order to quantify the differences between the electron distribu-
tions at the one- and two-electron levels, now for ionized systems. In doing so, Fig. 3 shows
the JSD between the monoelectronic and the electron pair density of different ions, calculated
in both conjugated spaces. First let us focus on Fig. 3a where singly-charged cations with
number of electrons N = 2−54 have been taken into account. 3.1 Interelectronic correlation analysis Notwithstanding, some additional patterns are noticeable, most certainly related to the
shell filling structure. This fact was emphasized in a previous work [58] for systems within
the shorter range N ≤36, on the basis of the mutual information as quantified by the K L
divergence among distributions, as here done by means of JSD up to N = 103. The trend
is monotonous, without many exceptions for a long period, however some discontinuities
appear when changing from a period to another indicating a transition to a new electronic
shell. Each segment can be assigned to a particular shell, in the fashion the figure shows. Besides, some less relevant extrema appear, all of them related to anomalous shell filling
systems. Some specific systems have been labelled, as follows: – Closedshells: N = 10, 18, 36, 54, 86.Thusnoblegasesappearasminimasystematically. The same was observed in the K L divergence for systems up to N ≤36 [58]. 123 123 Eur. Phys. J. Plus (2021) 136:763 Page 7 of 16
763 Eur. Phys. J. Plus (2021) 136:763 – (i) Anomalous shell-filling: minima N = 42, 46 (outer d-subshell), with valence subshell
half-filled and filled, respectively. (ii) The same applies to N = 24, 29: in spite of not
being extrema, the curve displays a ‘change of slope’ at those points, to be contrasted later
with momentum space. (iii) Minor extrema appear for heavy systems, within a region of
anomalies. – (i) Anomalous shell-filling: minima N = 42, 46 (outer d-subshell), with valence subshell
half-filled and filled, respectively. (ii) The same applies to N = 24, 29: in spite of not
being extrema, the curve displays a ‘change of slope’ at those points, to be contrasted later
with momentum space. (iii) Minor extrema appear for heavy systems, within a region of
anomalies. – Closed subshells: minima N = 48, 80 with filled d-subshell. – Closed subshells: minima N = 48, 80 with filled d-subshell. The Jensen–Shannon divergence between electron pair density and the product of mono-
electronic densities in momentum space hardly provides additional information. The conclu-
sions we arrive are similar than before, in fact, the curves in both spaces almost overlap each
other, as it can be seen by comparing Fig. 2 (momentum space) with the previous one (note
the identical scales of both figures). Thus the electron pair densities in momentum space are
providing roughly the same information as the position ones. – As for neutral atoms, noble gases N = 10, 18, 36, 54 are displayed as lower peaks. On the
other hand, alkalines N = 3, 11, 19, 37 appear as maxima. Similarly occurs, in position 3.1 Interelectronic correlation analysis – Pairs of extrema 23−24 and 28−29 (minimum–maximum) are due to the anomalous
shell filling of both maxima N = 24, 29, with the inner half-filled 4s subshell, and with
half-filled or filled (respectively) outermost 3d subshell. – Similarly the pair of consecutive extrema 41 −42 has inner half-filled 5s subshell, and
outermost subshell 4d, half-filled for the maximum N = 42. – Similarly the pair of consecutive extrema 41 −42 has inner half-filled 5s subshell, and
outermost subshell 4d, half-filled for the maximum N = 42. – Particularly interesting is the strong minimum N = 46, as discussed for neutrals, with
empty 5s subshell and filled valence subshell (4d). – Particularly interesting is the strong minimum N = 46, as discussed for neutrals, with
empty 5s subshell and filled valence subshell (4d). – A few of the above systems display a higher JSD in momentum than in position space:
N = 24, 29, 42, 47 (with inner half-filled s-orbital and outermost half-filled or com-
pletely filled d-orbital), as well as N = 25 (half-filled outermost d-orbital). In Fig. 3b, where the anions have been considered, exactly the same pattern is observed,
but with a less elaborate structure, due to the fewer available systems as it was previously
mentioned. Some alkaline earths have been labelled, but let us emphasize the ‘absence from
the list’ of alkalines, which are displayed as maxima for cations. Position space JSD is
almost systematically greater than the momentum one. The only extremely light exceptions
are N = 41−42, with differences between spaces below 0.2%. 3.2 Atomic ionization 3.1 Interelectronic correlation analysis The curves show us a decreas-
ing behaviour when N increases and, in this case, the momentum space roughly presents
smaller values of JSD. However, an interesting pattern appears in the figure: cations corre-
sponding to N = 10, 18, 23, 28, 36, 41, 46, 48, 54 have a minor difference, a lower value
of JSD, in both conjugated spaces with the only exception N = 48 (extremely light mini-
mum, only in position space). Among maxima (located at almost identical positions in both
spaces), let us emphasize the presence of N = 24, 29, 42. Regarding these two collections,
some comments are in order: – As for neutral atoms, noble gases N = 10, 18, 36, 54 are displayed as lower peaks. On the
other hand, alkalines N = 3, 11, 19, 37 appear as maxima. Similarly occurs, in position 123 123 763
Page 8 of 16 Eur. Phys. J. Plus (2021) 136:763 Fig. 2 Jensen–Shannon divergence in momentum space (JSDp) between the electron pair density and the
product of monoelectronic densities, for neutral atoms with number of electrons N = 3−103. Labels are for
systems with closed shells (red) or anomalous shell-filling (green). Atomic units (a.u.) are used Fig. 2 Jensen–Shannon divergence in momentum space (JSDp) between the electron pair density and the
product of monoelectronic densities, for neutral atoms with number of electrons N = 3−103. Labels are for
systems with closed shells (red) or anomalous shell-filling (green). Atomic units (a.u.) are used space, with the closed-shell system N = 48 (minimum) and its neighbors N = 47, 49
(maxima). – Pairs of extrema 23−24 and 28−29 (minimum–maximum) are due to the anomalous
shell filling of both maxima N = 24, 29, with the inner half-filled 4s subshell, and with
half-filled or filled (respectively) outermost 3d subshell. – Pairs of extrema 23−24 and 28−29 (minimum–maximum) are due to the anomalous
shell filling of both maxima N = 24, 29, with the inner half-filled 4s subshell, and with
half-filled or filled (respectively) outermost 3d subshell. – Pairs of extrema 23−24 and 28−29 (minimum–maximum) are due to the anomalous
shell filling of both maxima N = 24, 29, with the inner half-filled 4s subshell, and with
half-filled or filled (respectively) outermost 3d subshell. 3.2 Atomic ionization Applying divergence measures to the analysis of ions will allow us to find which electrons
have a higher impact on the whole density. In order to do that, we inspect the Jensen–Shannon
divergence between the pair densities of an ion (cation or anion) and its respective neutral,
in both position and momentum spaces. For completeness, similar results on the basis of
one-particle densities, previously obtained [55], are also displayed. This is shown in Figs. 4
and 5. Next results are to be interpreted as measures of the ‘amount of change’ on the electron
cloud of a given system after its ionization takes place, either by ejecting or by capturing an Eur. Phys. J. Plus (2021) 136:763 Page 9 of 16
763 (b)
(a)
Fig. 3 Jensen–Shannon divergence, in position and momentum spaces, between the electron pair density and
the product of one-particle densities, for a cations and b anions with number of electrons 3 ≤N ≤55. Labels
are for systems with closed shells (red), closed subshells (blue) or anomalous shell-filling (green). Atomic
units (a.u.) are used ( ) (a) (b) (b) Fig. 3 Jensen–Shannon divergence, in position and momentum spaces, between the electron pair density and
the product of one-particle densities, for a cations and b anions with number of electrons 3 ≤N ≤55. Labels
are for systems with closed shells (red), closed subshells (blue) or anomalous shell-filling (green). Atomic
units (a.u.) are used electron. Thus, if the initial neutral system has N electrons, the resulting ion will have N −1
or N + 1 electrons, for the cation and the anion, respectively. This interpretation applies to
both conjugated spaces, as well as to one-particle and electron pair densities. Figure 4 encloses curves of JSD between the electron distribution of a neutral system and
that of its singly-charged cation, denoted in position and momentum spaces as JSDr(N, C) in
Fig. 4a, and JSDp(N, C) in Fig. 4b, respectively. A simple function of the atomic ionization
potential AIP needed for performing such process is also depicted. This function has been
considered in order to better compare all curves, which display a variety of local extrema,
most of them characterized accordingly with the shell-filling of neutrals and ions, as discussed
below. 123 123 763
Page 10 of 16 Eur. Phys. J. Plus (2021) 136:763 (b)
(a)
Fig. – Rest of peaks (N = 23, 27, 41, 44) correspond to ‘d’ valence subshells. 3.2 Atomic ionization The same analysis has been carried out in momentum space, and the results are displayed
in Fig. 4b. Conclusions are similar to those obtained from position-space divergences, and
the differences are emphasized below: – Position-space peaks N = 3, 11, 19, 23, 37, 55 (enclosing all alkalines) remain in
momentum space for both one- and two-particle comparisons. On the other hand, N = 31
is lost also in both cases. – N = 27, 45 are kept in the monoparticular case, while the slight shifts from maxima
located at (27, 44) to (28, 45) are observed for electron pair divergences. – Maximum N = 49 disappears for momentum one-electron distributions, while N =
41, 47 for the two-electron ones. Notice that the half-filled 5s subshell of systems N = 41
and N = 47 becomes filled and empty, respectively, after ionization. – On the other hand, the disappearance of N = 31, 49 from the set of maxima in going from
position to momentum space corresponds to systems which ‘p’ valence subshell becomes
empty. Instead, the divergence at the one-electron level displays maxima N = 8, 52,
characterized for getting a half-filled ‘p’ valence subshell after ionization. – Concerning asymptotic trends, it is observed that the monoelectronic JSD remains within
a low interval for large N, while the electron pair divergence appreciably increases,
roughly from N = 25 on. – Concerning asymptotic trends, it is observed that the monoelectronic JSD remains within
a low interval for large N, while the electron pair divergence appreciably increases,
roughly from N = 25 on. The differences between the JSD behaviors in position and momentum spaces arise from
the asymptotic behaviors of the involved densities. This fact has been shown in previous
studies on density functional analysis, particularly JSD among others [41]. In position space,
the atomic one- and two-electron densities have an exponential decreasing behavior, so that
the values of the involved integrals are mainly determined by the regions surrounding the
nuclei, those almost unaffected by ionization. However, the lowspeed regions are associated
with the outermost subshells, being consequently those mainly determining the JSD values. The differences between the valence subshells of the neutral and ionized systems are more
clearly reflected in the pair distribution, as compared to the monoelectronic one. Let us analyse the structure of the JSD in each period comparing their extremal values to
those of AI P. 3.2 Atomic ionization 4 Pair and one-electron Jensen–Shannon divergences between the neutral system with a number of
electrons 3 ≤N ≤55 and its singly charged cation, and the inverse of square root of the atomic ionization
potential (AIP), in a position and b momentum spaces. Atomic units (a.u.) are used (a) (a) (b) Fig. 4 Pair and one-electron Jensen–Shannon divergences between the neutral system with a number of
electrons 3 ≤N ≤55 and its singly charged cation, and the inverse of square root of the atomic ionization
potential (AIP), in a position and b momentum spaces. Atomic units (a.u.) are used Fig. 4 Pair and one-electron Jensen–Shannon divergences between the neutral system with a number of
electrons 3 ≤N ≤55 and its singly charged cation, and the inverse of square root of the atomic ionization
potential (AIP), in a position and b momentum spaces. Atomic units (a.u.) are used Let us start with the location of maxima of JSD in position space, when comparing a
neutral of N electrons with the resulting cation of N −1 electrons, namely JSDr(N, C),
as displayed in Fig. 4a. In the two-particle case, higher peaks are found at neutrals
N = 3, 11, 19, 23, 27, 31, 37, 41, 44, 47, 49, 55. Within the range of N here considered,
it is worthy to remark: – Alkalines N = 3, 11, 19, 37, 55 appear included, that is systems which outermost ‘s’
subshell becomes empty after ionization. – Alkalines N = 3, 11, 19, 37, 55 appear included, that is systems which outermost ‘s’
subshell becomes empty after ionization. – The same applies to the anomalous N = 47 regarding its valence subs – For systems N = 31, 49 the lost subshell is 4p and 5p, respectively. – Rest of peaks (N = 23, 27, 41, 44) correspond to ‘d’ valence subshells. 123 Eur. Phys. J. Plus (2021) 136:763 Page 11 of 16
763 – Comparing with the one-electron curve, almost identical structures are obtained regard-
ing their extrema and monotonic trends. The most prominent differential feature is the
systematic higher value of electron pair JSD as compared to the monoelectronic one. – Comparing with the one-electron curve, almost identical structures are obtained regard-
ing their extrema and monotonic trends. The most prominent differential feature is the
systematic higher value of electron pair JSD as compared to the monoelectronic one. 3.2 Atomic ionization – Alkalines N = 3, 11, 19, 37 remain as maxima, also for anions. Let us remark that the
next alkaline (N = 55) is not included in the managed anion set. – The anion obtained from neutral N = 31 does not give rise to a maximum, contrary to
the cation. – The anion obtained from neutral N = 31 does not give rise to a maximum, contrary to
the cation. – Two sets of slight shifts are observed: (23, 27 and 41) to (24, 29 and 42), and (44 or 45,
and 47) to (44 and 46). – Two sets of slight shifts are observed: (23, 27 and 41) to (24, 29 and 42), and (44 or 45,
and 47) to (44 and 46). The only exception to the above common patterns is N = 49, shifted to N = 51 in the
monoelectronic case, and lost for electron pair densities. In momentum space, Fig. 5b, (i) alkalines N = 3, 11, 19 are kept for anions, (ii) the one-
electron shifts of maxima (8,23,27 or 28,41,45,47,52) to (7, 24, 29, 42, 44, 46, 51) include
the two-electron ones, corresponding to N = 24, 29, 46, and (iii) in both cases, the new
maximum N = 15 appears, as well as N = 33 at the one-electron level, that is systems with
a half-filled ‘p’ valence subshell, as also N = 51 included in the above list of shifts. Differences between the global behaviors of the one- and two-electron curves appear
justified in the same way as for neutral-cation transitions. That is, both densities decrease
exponentially far from the nucleus in position space, while the asymptotic fall of momentum
densities is much more slow. 3.2 Atomic ionization In Table 1, values of number of electrons N of the neutral system for which
the Jensen–Shannon divergence display local extrema (maximum) are given. In addition,
minimum values of the AI P are shown, which are associated to systems with a single electron
in the valence subshell, making consequently such a subshell to disappear after ionization,
and the resulting system to strongly differ from the initial one. This relevant difference is
usually revealed in terms of a high divergence and dissimilarity between the initial and the
final system. There exist a strong correlation also with the structure displayed by the atomic
ionization potential AI P in the NC ionization process in which an ‘s’ electron is removed,
as shown in the corresponding column of the table. However, as it can be observed the same
is not true when the removed electron is of ‘p’ or ‘d’ type. It appears natural to wonder ourselves about all the above features when the initial neutral
system gets an additional electron, so becoming a singly-charged anion. The appropriate
calculations have been performed, and the results obtained are displayed in Fig. 5. Conclusions for anions are much more limited, as compared to those obtained for cations. The reason is the absence of numerous systems from the list of anions, not included in the
Ref. [56] considered for the present work. However, some interesting remarks are in order. 123 763
Page 12 of 16 Eur. Phys. J. Plus Eur. Phys. J. Plus (2021) 136:763 Table 1 Number of electrons N at local extrema for the atomic ionization potential AIP of neutral atoms and
the Jensen–Shannon divergences in position (JSDr ) and momentum (JSDp) spaces, between the electron pair
densities of a neutral system with a number of electrons 3 ≤N ≤54 and its singly charged anions (A) and
cations (C). Atomic units are used Measure
N −→C
A −→N
AIP
(s)
3,11,19,23,28,37,47,55
(p,d)
5,8,13,16,31,34,49
JSDr
(s)
3,11,19,23,27,37,41,44,47,55
3,11,19,24,29,37,42,44,46
(p,d)
31,49
JSDp
(s)
3,11,19,23,28,37,45,55
3,11,19,24,29,46
(p,d)
49
15,49,53 In position space, Fig. 5a, maxima for the divergence neutral-anion are roughly the same
as those of neutral-cation, for the one- and the two-electron distributions. In both cases, the
following patterns are displayed: – Alkalines N = 3, 11, 19, 37 remain as maxima, also for anions. Let us remark that the
next alkaline (N = 55) is not included in the managed anion set. 4 Conclusions and future work Jensen–Shannon divergence (JSD) has been employed as comparative functional for different
many-electron atoms on the basis of their respective one- and two-electron distributions,
enclosing neutral atoms (N = Z) and singly-charged ions (N = Z ± 1). Particularly JSD
is applied to quantify the interelectronic correlation in atomic systems. A thorough study
was carried out, by considering (i) both position and momentum conjugated spaces, and (ii)
systems with a nuclear charge as far as Z = 103. The correlation among electrons was measured by comparing, for an arbitrary system, the
double-variable electron pair density with the product of the respective one-particle densities. As expected, the absence of correlation translates into the extreme allowed value JSD = 0, 123 Eur. Phys. J. Plus (2021) 136:763 Page 13 of 16
763 Eur. Phys. J. Plus (a)
(b)
Fig. 5 Pair and one-electron Jensen–Shannon divergences between the neutral system with a number of
electrons 3 ≤N ≤54 and its singly charged anion, in a position and b momentum spaces. Atomic units (a.u.)
are used (a) (a) (a) (b) (b) (b) Fig. 5 Pair and one-electron Jensen–Shannon divergences between the neutral system with a number of
electrons 3 ≤N ≤54 and its singly charged anion, in a position and b momentum spaces. Atomic units (a.u.)
are used Fig. 5 Pair and one-electron Jensen–Shannon divergences between the neutral system with a number of
electrons 3 ≤N ≤54 and its singly charged anion, in a position and b momentum spaces. Atomic units (a.u.)
are used and any deviation from the above value is caused by the existence of correlation, more or
less strong, accordingly with the relevance of the aforementioned deviations. The main conclusion is that the dependence on the number of electrons only governs
general trends of JSD, being insufficient as single variable to justify the structural patterns
shown by those curves. Interelectronic correlation roughly decreases as the number of elec-
trons increases, as revealed by the decreasing behavior of JSD in terms of N. However, there
appear a number of local extrema, whose location and strength are clearly determined by
shell-filling patterns. Specially relevant changes in JSD are observed just after adding an
electron to a closed-subshell system. All this is valid in position and momentum spaces. The above comments on neutral atoms apply also for cations and anions, regarding struc-
tural patterns. 4 Conclusions and future work Once again the interelectronic correlation, as measured by JSD, decreases
along a given period (i.e. as adding electrons) but notably increases in passing to a new
period. 123 763
Page 14 of 16 Eur. Phys. J. Plus (2021) 136:763 Apart from correlation, additional properties of atomic systems were analysed by means
of JSD. Such is the case of ionization as described by the ‘resemblance’ among neutrals and
their respective singly-charged ions, in terms of the corresponding electron pair densities. While structural patterns are similar in position and momentum spaces, the same is not true
regarding monotonic behaviors. JSD in both spaces displays a decreasing trend for light
systems, that remains in the position case for the whole set of systems. However, it turns out
to an increasing one in momentum space for medium-heavy systems. So break of resemblance
after ionization appears greater in momentum space than in the position one, for both cations
and anions. Future work is planned by considering other functionals. For instance, relative Fisher
information (containing also the gradient of the density) was exhaustively applied for the
analysis of specific one-particle quantum-mechanical systems (Morse potential and harmonic
oscillator [48], hydrogen-like atoms [49]). Additionally, numerous studies of atomic [59] and
molecular [60] systems were performed by means of the quantum similarity index on the
basis of their one-particle densities. On the other hand, a study of different divergences and
similarity measures, particularly Jensen–Shannon divergence, between the exchange density
and the product of the one particular densities could be interesting. Acknowledgements This work was supported in part by the Spanish MINECO project FIS2014-59311-P
(cofinanced by FEDER). A.L.M., J.C.A. and J.A. belong to the Andalusian research group FQM-020, and
S.L.R. to FQM-239. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence,
and indicate if changes were made. The images or other third party material in this article are included in the
article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. 4 Conclusions and future work If material is
not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. References 1. T. Koga, Electron-Pair Densities of atom. in Many-Electron Densities and Reduced Density Matrices
(Springer, New York, 2000) 2. M. Rodríguez-Mayorga, M. Via-Nadal, M. Solà, J.M. Ugalde, X. Lopez, E. Matito, Electron-pair dis
bution in chemical bond formation. J. Phys. Chem. A 122, 1916–1923 (2018) R.P. Sagar, N.L. Guevara, Mutual information and correlation measures in atomic systems. J. Chem. Phys. 123, 044108 (2005) 4. J. Cioslowski, Topology of Electron Correlation. in Many-Electron Densities and Reduced Density Matri-
ces (Springer, New York, 2000) 5. P.W. Ayers, E.R. Davidson, Necessary conditions for the n-representability of pair distribution functions. Int. J. Quant. Chem. 1006, 1487–1498 (2006) 5. P.W. Ayers, E.R. Davidson, Necessary conditions fo
Int. J. Quant. Chem. 1006, 1487–1498 (2006) Int. J. Quant. Chem. 1006, 1487–1498 (2006) 6. M. Levy, P. Ziesche, The pair density functional of the kinetic energy and its simple scaling property. J. Chem. Phys. 115, 9110–9112 (2001) 6. M. Levy, P. Ziesche, The pair density functiona Chem. Phys. 115, 9110–9112 (2001) y
. Á. Nagy, Density-matrix functional theory. Phys. R 8. D.A. Mazziotti, Two-electron reduced density matrix as the basic variable in many-electron quantum
chemistry and physics. Chem. Rev. 112, 244–262 (2012) y
p y
9. M. Higuchi, K. Higuchi, Pair density functional theory. Comput. Theor. Chem. 1003, 91–96 (2013) 9. M. Higuchi, K. Higuchi, Pair density functional theory. Comput. Theor. Chem. 1003, 91–96 (2013)
10. M. Higuchi, K. Higuchi, A pair density functional theory utilizing the correlated wave function. J. Phys. Conf. Ser. 150, 042056 (2009) 10. M. Higuchi, K. Higuchi, A pair density functional theory utilizing the correlated wave function. J. Ph
Conf. Ser. 150, 042056 (2009) 11. P. Ziesche, Attempts toward a pair density functional theory. Int. J. Quantum Chem. 60, 1361–1374 (1996)
12
G Li M
i R K C l
S L
D M
J Ol
D G T hl
L G
li di M lti
fi
ti 11. P. Ziesche, Attempts toward a pair density functional theory. Int. J. Quantum Chem. 60, 1361–1374 (1996)
12. G. Li Manni, R.K. Carlson, S. Luo, D. Ma, J. Olsen, D.G. Truhlar, L. Gagliardi, Multiconfiguration
pair-density functional theory. J. Chem. Theory Comput. 10, 3669–3680 (2014) p
p
y
y
Q
(
)
12. G. Li Manni, R.K. Carlson, S. Luo, D. Ma, J. Olsen, D.G. Truhlar, L. Gagliardi, Multiconfiguration
pair-density functional theory. J. Chem. Theory Comput. 10, 3669–3680 (2014) 123 Page 15 of 16
763 Eur. Phys. J. Plus (2021) 136:763 13. J.L. Bao, Y. Wang, X. He, L. Gagliardi, D.G. Truhlar, Multiconfiguration pair-density functional the
is free from delocalization error. J. Phys. Chem. Lett. 8, 5616–5620 (2017) 14. M. Sajjan, K. Head-Marsden, D.A. Mazziotti, Entangling and disentangling many-electron quantum
systems with an electric field. Phys. Rev. A 97, 062502 (2018) 15. R. Benesch, V.H. Smith, Correlation and X-ray scattering. I. Density matrix formulation. Acta Cryst. 26,
579–586 (1970) 16. A.J. Thakkar, A.N. Tripathi, V.H. Smith, Molecular X-ray- and electron-scattering intensities. Phys. Rev. A 29, 1108–1113 (1984) 17. T. Koga, Y. Kasa, A.J. Thakkar, Accurate algebraic densities and intracules for heliumlike ions. Int. J. Quantum Chem. 46, 689 (1993) 18. J.M. Herbert, J.E. Harriman, Comparison of two-electron densities reconstructed from one-electron den-
sity matrices. Int. J. Quantum Chem. 90, 355 (2002) 18. J.M. Herbert, J.E. Harriman, Comparison of two-electr
sity matrices. Int. J. Quantum Chem. 90, 355 (2002) 19. C. Sarasola, J.M. Ugalde, R.J. Chem. Phys. 115, 9110–9112 (2001) Boyd, The evaluation of extracule and intracule densities in the first-row
hydrides, LiH, BeH, BH, CH, NH, OH and FH, from self-consistent field molecular orbital wavefunctions. J. Phys. B: At. Mol. Opt. 23, 1095 (1990) y
p
20. J. Cioslowski, B.B. Stefanov, A. Tan, C.J. Umrigar, Electron intracule densities with correct electron
coalescence cusps from Hiller–Sucher–Feinberg-type identities. J. Chem. Phys. 103, 6093 (1995) 21. A.J. Proud, D.E.C.K. Mackenzie, J.K. Pearson, Exploring electron pair behaviour in chemical bo
using the extracule density. Phys. Chem. Chem. Phys. 17, 20194–20204 (2015) 21. A.J. Proud, D.E.C.K. Mackenzie, J.K. Pearson, Exploring electron pair behaviour in chemical bonds
using the extracule density. Phys. Chem. Chem. Phys. 17, 20194–20204 (2015)
22. C.E. Shannon, W. Weaver, The Mathematical Theory of Communication (University of Illinois Press, 22. C.E. Shannon, W. Weaver, The Mathematical Theory of Communication (University of Illinois Pr
Urbana, 1949) 23. R.A. Fisher, Theory of statistical estimation. Proc. Cambridge Phil. Soc., 22, 700–725, Reprinted in
Collected Papers of R.A. Fisher, edited by J.H. Bennet (University of Adelaide Press, South Australia)
1972, 15–40 (1925) 24. J. Pipek, I. Varga, Universal classification scheme for the spatial-localization properties of one-part
states in finite, d-dimensional systems. Phys. Rev. A 46, 3148 (1992) 25. R. López-Ruiz, H.L. Mancini, X. Calbet, A statistical measure of complexity. Phys. Lett. A 209, 321–326
(1995) 26. B.R. Frieden, Science from Fisher Information (Cambridge University Press, Cambridge, 2004) 27. T.M. Cover, J.A. Thomas, Elements of Information Theory (Wiley, New York, 1991) S.R. Gadre, R.D. Bendale, S.P. Gejji, Analysis of atomic electron momentum densities: use of information
entropies in coordinate and momentum space. Chem. Phys. Lett. 117, 138 (1985) 28. S.R. Gadre, R.D. Bendale, S.P. Gejji, Analysis of atomic electron momentum densitie
entropies in coordinate and momentum space. Chem. Phys. Lett. 117, 138 (1985) 29. M. Ho, V.H. Smith, D.F. Weaver, C. Gatti, R.P. Sagar, R.O. Esquivel, Molecular similarity based on
information entropies and distances. J. Chem. Phys. 108, 5469 (1998) 30. K.D. Sen, C.P. Panos, K.C. Chatzisavvas, C.C. Moustakidis, Net Fisher information measure versus
ionization potential and dipole polarizability in atoms. Phys. Lett. A 364, 286–290 (2007) 31. S. Janssens, A. Borgoo, C. Van Alsenoy, P. Geerlings, Information theoretical study of chir
tiomers with one and two asymmetric centra. J. Phys. Chem. A 112, 10560–10569 (2008) 32. R.F. Nalewajski, Information principles in the theory of electronic structure. Chem. Phys. Lett. 372, 28
(2003) (
)
33. J.C. Chem. Phys. 115, 9110–9112 (2001) Angulo, J. Antolín, Atomic complexity measures in position and momentum spaces. J. Chem. Phys. 128, 164109 (2008) 34. J.C. Angulo, J. Antolín, K.D. Sen, Fisher-Shannon plane and statistical complexity of atoms. Phys. L
A 372, 670–674 (2008) 35. S. Kullback, A. Leibler, On information and sufficiency. Ann. Math. Stat. 22, 79–86 (1951) 36. J.H. Lin, Divergence measures based on the Shannon entropy. IEEE Trans. Inf. Theory 37, 145–151
(1991) 37. B. Fuglede, F. Topsoe, Jensen-Shannon divergence and Hilbert space embedding. IEEE Int. Sym. Inf. Theo. 31, 71–73 (2004) 38. P.W. Lamberti, A.P. Majtey, A. Borras, M. Casas, A. Plastino, Metric character of the quantum Jensen–
Shannon divergence. Phys. Rev. A 77, 052311 (2008) 39. I. Grosse, P. Bernaola-Galván, P. Carpena, R. Román-Roldán, J.L. Oliver, H.E. Stanley, Analysis of
symbolic sequences using the Jensen–Shannon divergence. Phys. Rev. E 65, 041905 (2002) 40. J.C. Angulo, J. Antolín, S. López-Rosa, R.O. Esquivel, Jensen-Shannon divergence in conjuga
the entropy excess of atomic systems and sets with respect to their constituents. Phys. A 389, 8 py
y
p
y
,
(
)
41. J. Antolín, J.C. Angulo, S. López-Rosa, Fisher and Jensen-Shannon divergences: quantitative comparisons
among distributions. Application to position and momentum atomic densities. J. Chem. Phys. 130, 074110
(2009) 42. J.C. Angulo, S. López-Rosa, J. Antolín, Generalized Jensen divergence analysis of atomic electron den-
sities in conjugated spaces. Int. J. Quantum Chem. 111, 297–306 (2011) 123 123 123 763
Page 16 of 16 Eur. Phys. J. Plus (2021) 136:763 763 Eur. Phys. J. Plus 43. A.L. Martín, S. López-Rosa, J.C. Angulo, J. Antolín, Jensen-Shannon and Kullback–Leibler divergen
as quantifiers of relativistic effects in neutral atoms. Chem. Phys. Lett. 75, 642 (2015) 44. J. Antolín, J.C. Angulo, S. Mulas, S. López-Rosa, Relativistic global and local divergences in hydrogenic
systems: a study in position and momentum spaces. Phys. Rev. A 90, 042511 (2014) y
y
p
p
y
(
)
45. S.J.C. Salazar, H.G. Laguna, R.P. Sagar, Higher-order statistical correlations in three-particle quantum
systems with harmonic interactions. Phys. Rev. A 101, 042105 (2020) y
y
p
p
y
45. S.J.C. Salazar, H.G. Laguna, R.P. Sagar, Higher-order statistical correla
systems with harmonic interactions. Phys. Rev. A 101, 042105 (2020) y
y
46. S.J.C. Salazar, H.G. Laguna, R.P. Sagar, Higher-order information measures from cumulative densities
in continuous variable quantum systems. Quantum Rep. 2, 560–578 (2020) 47. O. Werba, A. Raeber, K. Head-Marsden, D.A. Chem. Phys. 115, 9110–9112 (2001) Mazziotti, Signature of van der Waals interactions in the
cumulant density matrix. Phys. Chem. Chem. Phys. 21, 23900 (2019) 48. T. Yamano, Relative fisher information for Morse potential and isotropic quantum oscillators. J. Phys. Commun. 2, 085018 (2018) T. Yamano, Relative fisher information of hydrogen-like atoms. Chem. Phys. Lett. 691, 196–198 (2018) 50. P.O. Löwdin, Quantum theory of many-particle systems. I. Physical interpretations by means of density
matrices, natural spin-orbitals, and convergence problems in the method of configurational interaction. Phys. Rev. 97, 1474 (1955) y
51. A.K.C. Wong, M. You, Entropy and distance of random graphs with application to structural pattern-
recognition. IEEE Trans. Pattern Anal. Mach. Intell. 7, 599 (1985) 52. C.R. Rao, T. Nayak, Cross entropy, dissimilarity measures, and characterizations of quadratic entro
IEEE Trans. Inf. Theory 31, 589 (1985) 53. C.E. Shannon, A mathematical theory of communication. Bell Syst. Tech. J. 27, 379 (1948) 54. L.E. Riveaud, D. Mateos, S. Zozor, P.W. Lamberti, Generalized divergences from generalized entropies. Phys. A 68–76, 510 (2018) 55. S. López-Rosa, J. Antolín, J.C. Angulo, R.O. Esquivel, Divergence analysis of atomic ionization processes
and isoelectronic series. Phys. Rev. A 80(1–9), 012505 (2009) 56. T. Koga, K. Kanayama, S. Watanabe, A.J. Thakkar, Analytical Hartree–Fock wave functions subject to
cusp and asymptotic constraints: He to Xe, Li+ to Cs+, H- to I-. Int. J. Quantum Chem. 71, 491 (1999) 57. T. Koga, K. Kanayama, T. Watanabe, T. Imai, A.J. Thakkar, Analytical Hartree–Fock wave functions for
the atoms Cs to Lr. Theor. Chem. Acc. 104, 411 (2000) 57. T. Koga, K. Kanayama, T. Watanabe, T. Imai, A.J. Thakkar, Analy
the atoms Cs to Lr. Theor. Chem. Acc. 104, 411 (2000) he atoms Cs to Lr. Theor. Chem. Acc. 104, 411 (2000 58. R.P. Sagar, H.G. Laguna, N.L. Guevara, Electron pair density information measures in atomic syste
Int. J. Quantum Chem. 111, 3497 (2011) 58. R.P. Sagar, H.G. Laguna, N.L. Guevara, E
Int. J. Quantum Chem. 111, 3497 (2011) Int. J. Quantum Chem. 111, 3497 (2011) Q
,
(
)
59. J.C. Angulo, J. Antolín, Atomic quantum similarity indices in position and momentum spaces. J. Phys. Chem. 126, 044106 (2007) 60. R. Carbó-Dorca, X. Girones, P.G. Mezey (eds.), Fundamentals of Molecular Similarity (Kluwer Aca-
demic, Dordrecht, 2001) 123
|
https://openalex.org/W2092253809
|
https://zenodo.org/records/1616836/files/article.pdf
|
German
| null |
Ueber Derivate des α‐Amidoalizarins
|
Berichte der Deutschen Chemischen Gesellschaft
| 1,902
|
public-domain
| 1,465
|
9 06 9 06 (Eingegangen am 20. Pebriiar 1902.) (Eingegangen am 20. Pebriiar 1902.) H. R o e m e r I) erhielt durch Einwirkurig von Essigsaureanhydrid
and Benzoylchlorid auf p-Amidoalizarin ein Diacetyl- bezw. Diben-
zsyl-Amidoalizarin und die entsprechenden Anhydroderirate: 0
/O--COC,jH5
/\ -'.'."''..__OH
I
I
'
NH.CO CsH6
--..A
\ -,\/--
0 0
/O--COC,jH5
/\ -'.'."''..__OH
I
I
'
NH.CO CsH6
--..A
\ -,\/--
0 0
,O--COCH,
/\, /-..,.-\
\/\/L./NH.COCH3
! I
l-OH
0 0
/O--COC,jH5
/\ -'.'."''..__OH
I
I
'
NH.CO CsH6
--..A
\ -,\/--
0 0
,O--COCH,
/\, /-..,.-\
\/\/L./NH.COCH3
! I
l-OH
0
(Diacetyl-@-amidoalizarin). (Aethen ylacetylamidoalizarin). 0
/O--COC,jH5
/\ -'.'."''..__OH
I
I
'
NH.CO CsH6
--..A
\ -,\/--
0
(Dibenzopl-3- amidoalizarin). (Benzen ylbenzoy lamitloalizarin). (Dibenzopl-3- amidoalizarin). (Benzen ylbenzoy lamitloalizarin). (Benzen ylbenzoy lamitloalizarin). (Benzen ylbenzoy lamitloalizarin). (Aethen ylacetylamidoalizarin). Das Dibenzoyl-B-amidoalizarin konnte er nicht rein darstellen und
analysiren. Da vom M -Amidoalizarin die analogen Verbindungen noch nicht
bekannt waren, versuchten wir diese Liicke auszufiillen und theilen
im Folgenden das Resultat unserer Untersuchungen mit. Diese Berichte 18, 1666 [1685]. 136. G. Sohultz und J. Erber: Ueber Derivate
des (c -Amidoalizarins. [Mittheil. aus dem chern.-techn. Laborat. der Kgl. techn. Hochsch. zu Miinchen.] I. E in w ir k u n g v o n E ssig s iiu re a n h y d r i d
a u f a- A m i d o a l i z a r in. 0.2106 g Sbst.: 0.4915 g CO2, 0.0799 g HzO. - 0.2014g Sbst.: 0.4724g
ClsH1306N. Ber. C 63.71,
H 3.53,
N 4.13. Gef. )) 63.68, 63.97, * 4.22, 3 91,
)) 4.33. I. E in w ir k u n g v o n E ssig s iiu re a n h y d r i d
a u f a- A m i d o a l i z a r in. a-Amidoalizarin last sich in Essigsiiureanhydrid in der Kalte
schwer, beim Erwarmen leichter mit rothviolrtter Farbe. Kocht man
a- Amidoalizarin mit iiberschiissigem Essigsaureanhydrid, bis die Farbe
der Losung gelblich-braun geworden igt iind (dine Probe sich in kalter
Sodalosung nicht mehr violet fiirbt, so ist die Acetylirung beendet. Man saugt heies ab und lasst erkalten. Es scbeiden sich dabei roth-
braune Blattchen von fast reinem Diacetyl-fx-amidoalizarin vom
Schmp. 2450 au3. Das Diacetylprodiict 16st sich am besten in Essigsaureanhydr id,
Eisessig, Chloroform und Toluol, aos welch' Letxterem es am vortheil-
haftesten zu krystallisiren ist. Schwer liislich ist es in Aether, Aceton
und Benzol. Wird e3 mit Alkohol gekocht. so geht es in das i s o m e r e , bei
2030 schmelzende Diacetyl-a-amidoalizarin (s. u.) iiber. Durch heisse concentrirte Salzsaure wird es in das salzsaure
a-Amidoalizarin iibergefiihrt. g
Concentrirte Schwefelsaure 16st es mit rother Farbe. Von Soda-
losung wird es erst beim Kochen, yon Natronlange dagegen schon in
der Kalte zerlegt. Beim Suhlimiren zersetzt es sich vollstandig und hinterlasst vie1
Kohle. COa, 0.1285 g H&. - 0.2428 g Sbst.: 9.6 ocm N (19(), 721 mm). 0.328s g Sbst.: 0.7714 g COa, 0.1196 g H20. - 0.3540 g Sbst.: 0.8306 g C1SH1306N. Bor. C 63.71,
H 3.83,
N 4.13. Gef
)) 63.95, 63.65, )) 4.04, 4.04, * 4.32. 1SH1306N. Bor. C 63.71,
H 3.83,
N 4.13. Gef
)) 63.95, 63.65, )) 4.04, 4.04, * 4.32. Miissigt man die Einwirkung des Essigsiiureanhydrids in der
Weise, dass man das R -Amidoalizarin, mit geschniolzenem Natrium-
scetat gemischt, der Reaction des Essigsiiureanhydrids a n f d em
Was s e rb a d e unterwirft, bis eben eine Probe niit Sodaliisung die
vollendetca Reaction anzeigt, so erhalt man das oben erwahnte zweite
Diacetylproduct vom Schmp. 205". Das Verhalten dieses Kiirpers zu
den oben gensnnten Reagentien ist das gleiche will das des bei 2450
schmelzenden Korpers, nur mit dem Unterschiede, dass es leichter
Eiislich ist, aus Alkohol unverfndert krystallisirt und durch Kochen
mit Essigsiiureanhydrid in den zuvor beschriebenen Korper iibergeht. CO., 0.0708 g HzO. - 0.1702 g Sbst.: 6.5 rcm N (214 725 mm). 0.2106 g Sbst.: 0.4915 g CO2, 0.0799 g HzO. - 0.2014g Sbst.: 0.4724g
ClsH1306N. Ber. C 63.71,
H 3.53,
N 4.13. Gef. )) 63.68, 63.97, * 4.22, 3 91,
)) 4.33. CO., 0.0708 g HzO. - 0.1702 g Sbst.: 6.5 rcm N (214 725 mm). 11. E i 11 w i r k u n g vo n B e n z o y 1 c h 1 o r i d a u f a- A m i d o ali z a r in. )) 72.90, 72.75, )> 3.87, 3.71, A 3.42. 0.1432 g Sbst.: 0.3828 g COa, 0.0500 g HgO. - 0.1513 g Sbst.: 0.4036 g COa, 0.0506 g H20. - 0.1252 g Sbst.: 3.!) ccul X (16O, 715 mm). C9sHi7015N. Ber. C 74.57,
H 3.67,
N 3.03. Gef. )) 72.90, 72.75, )> 3.87, 3.71, A 3.42. 0.1432 g Sbst.: 0.3828 g COa, 0.0500 g HgO. - 0.1513 g Sbst.: 0.4036 g Miinchen, den 19. Februar 1902. 11. E i 11 w i r k u n g vo n B e n z o y 1 c h 1 o r i d a u f a- A m i d o ali z a r in. Die Benzoylirung nach der Methode von S c h o t t e n - B a n m a n n
fiihrt hier zu keinem Resuitat, ebensowenig das Eiwiirmen mit einem
Ueberschuss von Renzoylchlorid. Dagegen komrnt man zum Ziel,
wenn man das u-Amidoalizarin in kochendem Nitrobenzol Kist und lang-
sniu Benzoylchlorid zngiebt, nachdem man die Flamme entfernt hat. Zu etzt erwsrmt man noch atif dem Wasserbade untrr hauGgem
Srlttitteln,
I&
die Salzs~ureentwickelung nachgelassen hat und kein
Ausgangsmaterial mehr nachgewiesen werden kann. Heim Erkalten
k r j stallisirt ein Theil des Reactionsproductes aus; den Rest kann man
durch Zusatz von Alkohol und Aether zur Ausscheidung bringen. Der erhaltene Krysta1lbrt.i wird abgesaugt und niit Alkohol und
Aether ausqekocht. Einmalige Durchfiihruiig dieseir Operation geniigt,
um das Product analysenrein zu erhalten. Dass,elbe schmilzt iiber
310" und bildet rothbraune Nadeln. Es ist in Aikohol und Aether
fast uuliislicb und in den iibrigen anderen Liisungsmitteln sehr schwer
aiistich j zum Umkrystallisiren eignet sich am besten Nitrobenzol. 5s * 5s * 908 Kochende Salzsaure ist ohne Einwirkung. In concentrirter
Schwefelsaure 16st es sich niit rother Farbe. Von kochender Soda-
lijsung wird es nicht verandert, wohl aber von kochender Natronlauge. j
g
,
g
Das Benzoylproduct sublimirt unter theilweiser Zersetzung (Ab-
sgaltnng von Benzoesaure) in prachtvollen, rothen, gelbschillernden
Nadeln. Die Analyse lieferte Zahlen fiir ein M o n o b e n z o y l a m i d o -
alizarin. COa, 0.0894 g Hs0. - 0.1856 g Sbst.: 7.4 cem N (17O, 707 mm). 0 2330 g Sbst.: 0.6005 a COg, 0.0802 g HzO. - 0.2572 g Sbst.: 0.6604 a COa, 0.0894 g Hs0. - 0.1856 g Sbst.: 7.4 cem N (17O, 707 mm). 0 2330 g Sbst.: 0.6005 a COg, 0.0802 g HzO. - 0.2572 g Sbst.: 0.6604 a
C~lE11305N. Ber. C 70.19,
H 3.62,
N 3.90. Gef. x 70.32, 70.03, >> 3.83, 3.86, )) 4.0!1. Aus dem Filtrat des Monobenzoyl-a-amidoalizarins scheiden sich
nach einigen Tagen feine, helibraune Nadeln vorn Schmp. 2550 ab. Dieses Product ist l6slich in Chloroform, Toluol und Nitrobenzol und
verhalt sich im Uebrigen wie der oben beschriebene Korper. Die Analyse ergab Zahlen, die fiir ein D i b e n z o y 1 p rod u c t
stimmen. COa, 0.0506 g H20. - 0.1252 g Sbst.: 3.!) ccul X (16O, 715 mm). C9sHi7015N. Ber. C 74.57,
H 3.67,
N 3.03. Gef. 111. E i n w i r k u n g von B e n z o y l c h l o r i d auf p- Amidoalizarin. In Folge der verhaltnissmassig leichten Gewinnbarkeit von Ben-
zoylderivaten des a-Amidoalizarins nach der oben besehriebenen Me-
thode haben wir Letztere auch auf das #-Amidoalizarin angewendet. Wir verfuhren wie beim a-Amidoalizarin und erhielten dabei als
Hauptreactionsproduct ein Dibenzoylderivat, wahrend beim cc-Amido-
alizarin das Monobenzoylproduct den Haupttheil ausmachte. Das D i b e n z o y l - $ - A m i d o a l i z a r i n bildet gelbe, glanzende
Blattchen vom Schmp. 252O, die in hohem Grade elektrisch sind und
sich sehr gut au8 Chloroform und Toluol umkrystallisiren lassen. Den anderen, obea genannten Reagentieu gegeniiber verhalt es rrich
wie das Monobenzoylderivat des a-Amidoalizarins. GO,, 0.0923 g HsO. - 0.2512 g Sbst.: 7.9 ccm N (IS0, 711 mm). 0.2562 g Sbst.: 0.6796 g COa, 0.0894 g HaO. - 0.2604g Sbst.: 0.6914g GO,, 0.0923 g HsO. - 0.2512 g Sbst.: 7.9 ccm N (IS0, 711 mm). 0.2562 g Sbst.: 0.6796 g COa, 0.0894 g HaO. - 0.2604g Sbst.: 0.6914g CasHi7OsN. Ber. C 72.57,
H 3.67,
N 3.03. Gef. B 72.34, 72.42, )) 3.87, 3.71,
3.41. Miinchen, den 19. Februar 1902. Miinchen, den 19. Februar 1902.
|
https://openalex.org/W4250580542
|
https://www.duo.uio.no/bitstream/10852/47933/1/12932_2011_Article_125.pdf
|
English
| null |
On The Potential For Co2 Mineral Storage In Continental Floodbasalts—phreeqc Batch And 1d Diffusion-reaction Simulations
|
Apple Academic Press eBooks
| 2,014
|
cc-by
| 9,104
|
Abstract Continental flood basalts (CFB) are considered as potential CO2 storage sites because of their high reactivity and
abundant divalent metal ions that can potentially trap carbon for geological timescales. Moreover, laterally
extensive CFB are found in many place in the world within reasonable distances from major CO2 point emission
sources. Based on the mineral and glass composition of the Columbia River Basalt (CRB) we estimated the potential of CFB
to store CO2 in secondary carbonates. We simulated the system using kinetic dependent dissolution of primary
basalt-minerals (pyroxene, feldspar and glass) and the local equilibrium assumption for secondary phases
(weathering products). The simulations were divided into closed-system batch simulations at a constant CO2
pressure of 100 bar with sensitivity studies of temperature and reactive surface area, an evaluation of the reactivity
of H2O in scCO2, and finally 1D reactive diffusion simulations giving reactivity at CO2 pressures varying from 0 to
100 bar. Although the uncertainty in reactive surface area and corresponding reaction rates are large, we have estimated the
potential for CO2 mineral storage and identified factors that control the maximum extent of carbonation. The
simulations showed that formation of carbonates from basalt at 40 C may be limited to the formation of siderite
and possibly FeMg carbonates. Calcium was largely consumed by zeolite and oxide instead of forming carbonates. At higher temperatures (60 – 100 C), magnesite is suggested to form together with siderite and ankerite. The
maximum potential of CO2 stored as solid carbonates, if CO2 is supplied to the reactions unlimited, is shown to
depend on the availability of pore space as the hydration and carbonation reactions increase the solid volume and
clog the pore space. For systems such as in the scCO2 phase with limited amount of water, the total carbonation
potential is limited by the amount of water present for hydration of basalt. © 2012 Pham et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited. On the potential for CO2 mineral storage in
continental flood basalts – PHREEQC batch-
and 1D diffusion–reaction simulations Thi Hai Van Pham*, Per Aagaard and Helge Hellevang Thi Hai Van Pham*, Per Aagaard and Helge Hellevang Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Open Access * Correspondence: vtpham@geo.uio.no
Department of Geosciences, University of Oslo, Pb. 1047, Blindern, Oslo,
Norway Introduction
d
d Mineral storage of CO2 in basaltic rocks is favored
over siliciclastic reservoirs both by the higher abundance
of divalent metal ions in basalt and the faster reactivity
of basaltic glass or crystalline basalt [4]. Moreover,
basalts such as the Columbia River flood basalts (CRB)
are abundant and in many places close to CO2 point
source emissions [5]. During the last decade several
flood basalts around the world have been mapped for
the possibility of CO2 storage, and possible candidates
such as CRB in USA and the Deccan traps in India have
been identified [4-6]. Underground sequestration of carbon dioxide is a poten-
tially viable greenhouse gas mitigation option as it
reduces the release rate of CO2 to the atmosphere
[1]. CO2 can be trapped subsurface by four storage
mechanisms: (1) structural and stratigraphic trapping;
(2) residual CO2 trapping; (3) solubility trapping; and
(4) mineral trapping [2]. Mineral trapping has been con-
sidered as the safest mechanism in long-term storage of
CO2 [3]. To be a candidate for CO2 storage, the flood basalt
must have a proper sealing and sufficient injectivity, the
latter limited by the available connected pore space. In * Correspondence: vtpham@geo.uio.no
Department of Geosciences, University of Oslo, Pb. 1047, Blindern, Oslo,
Norway Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Page 2 of 12 Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 assemblages in batch and 1D advection–diffusion-reaction
mode. As the code can only model fully saturated systems,
natural systems must be simplified to end-member situa-
tions, such as given by constant pressure boundary condi-
tions as may be the case close to underground CO2
plumes, or the assumption of packages (batches) of water
reacting along a reaction path with a homogenous sedi-
ment or rock body. Based on these limitations we divided
the simulations into three systems representing different
parts of CO2 storage: (1) basalt alteration in the H2O-rich
phase at constant CO2 pressure; (2) basalt alteration in a
H2O saturated scCO2 phase, and (3) reactions at the
boundary of the CO2 plume where CO2 diffuses into the
aquifer from the boundary of the scCO2 plume (Figure 1). In the second case, we assumed that the CO2 phase had
swept through the systems and dried out residual water,
giving only dissolved water in the scCO2 phase. Introduction
d
d In this
case an upper limit of carbonation potential was estimated
as reactions were allowed to occur until (nearly) all water
was consumed, passing the upper 2 mol/Kgw theoretical
limit for the Truesdell-Jones activity model [9]. flood basalts, the connected pore space is typically found
at zones containing abundant vesicles or in breccias be-
tween basalt flows. Because central zones of flood basalts
commonly are dense and impermeable without vesicles,
and flows are laterally continuous over large areas and
commonly stacked vertically for hundreds of meters,
flow units can act as seals [5]. The non-porous inner
parts of flows may however be penetrated by networks
of vertical fractures. These fractures can be open and
conductive,
or
closed
by
mineralization
and
non-
conductive. The main objectives of this study were to performe
batch- and 1D diffusion–reaction numerical simulations
to determine the geochemical potential for secondary
carbonate formation and to estimate the volume changes
and the possibility of self-sealing following the basalt-
CO2 interactions. The CRB system was used as an ex-
ample case and our results were compared to earlier
reported laboratory experiments and numerical simula-
tions of CO2-basalt interactions. As CO2 stored under-
ground will distribute spatially in the reservoir to give a
range of reactive conditions, such as the potential of
reactions by H2O dissolved in supercritical CO2 [7,8] or
reactions in the H2O-rich phase from residually trapped
CO2, we divided the simulations into three systems
representing different parts of CO2 storage: (1) basalt al-
teration in the H2O-rich phase at constant CO2 pres-
sure; (2) basalt alteration in a H2O saturated CO2 phase,
and (3) reactions at the boundary of the CO2 plume
where CO2 diffuses into the aquifer from the boundary
of the CO2 plume (Figure 1). The standard state adopted in this study for the thermo-
dynamic calculations was that of unit activity for pure
minerals and H2O at any temperature and pressure. For
aqueous species other than H2O, the standard state was
unit activity of the species in a hypothetical 1 molal solu-
tion referenced to infinite dilution at any temperature and
pressure. For gases, the standard state was for unit fugacity
of a hypothetical ideal gas at 1 bar of pressure. All simula-
tions used the llnl.dat database based on the thermo.com. V8.R6.230 dataset prepared at the Lawrence Livermore
National Laboratory, with additions of thermodynamic
data for those phases not present (see description below). Methods CO2 fugacity coefficients were estimated according to
the modified Redlich-Kwong (SRK) equation of state
[10] and the solubility was adjusted for by a poynting
correction term (exp(vCO2(Psat - P)/RT) where v denotes
molar volume, P pressure, R the universal gas constant
and T absolute temperature) [11]. The density of CO2 at
40 C and 100 bar was approximated from Bachu and
Stewart [12] to be 600 Kg/m3 and the solubility of water
in scCO2 at the same conditions was approximated to
0.5 mole% [13,14] All thermodynamic and kinetic calculations were per-
formed using the geochemical code PHREEQC-2 [9]. This
code is capable of simulating complex interactions be-
tween dissolved gases, aqueous solutions, and mineral Figure 1 Sketch of possible reaction settings during CO2
storage in basalt. System 1 (S1) is close to the injector and
contains a wet CO2 (0.5 mole% H2O at 100 bar and 40 C) with no
residual water; system 2 (S2) is fully in the H2O rich phase with CO2
diffusing in from the plume boundary; and system 3 (S3) is at the
boundary of the CO2 plume with both sufficient non-wetting CO2 at
a constant CO2 partial pressure of 100 bar and with sufficient water
wetting the mineral surfaces and available for reactions. The simulations were divided into batch simulations
of the H2O rich and CO2 rich phases respectively,
and 1D diffusion of CO2 in the H2O rich phase to
obtain
information
on
the
CO2-basalt
interactions
over a continuous range of CO2 pressures. The latter
was solved by PHREEQC using
@tC ¼ DL@2
xC þ q ,
where C denotes molal (mol/Kgw) concentration, q
denotes the sink term, subscripts t and x refer to
derivatives in time and x-direction respectively, and
an efficient diffusion coefficient DL of 0.45x10-9 m2/s
was used for CO2 [15] and all solutes. Figure 1 Sketch of possible reaction settings during CO2
storage in basalt. System 1 (S1) is close to the injector and Figure 1 Sketch of possible reaction settings during CO2
storage in basalt. Methods The
use of mass or mass fractions of the individual basalt
components to estimate the release rates of elements
from the basalt is supported by a recent experimental
study which suggests that release rates estimated from
the sum of volume fractions of the constituent minerals
are within one order of magnitude from measured values
[22]. A list of kinetic parameters is given in Table 1. All
secondary phases were allowed to form according to the
local equilibrium assumption [23]. Ω ¼ exp ΔGr
RT
ð2Þ ð2Þ Where ΔGr is the Gibbs free energy of the reaction, R
is the gas constant, and T is absolute temperature. Reac-
tion rate constants for crystalline basalt (pyroxenes and
plagioclase) were obtained from Palandri and Kharaka
[16], and pH dependencies were taken from the same
source. The dissolution rate of basaltic glass was calcu-
lated according to the expression suggested by Gislason
and Oelkers [17]: rþ ¼ kþexp Ea
RT
S
a3
Hþ
aAl3þ
0:33
1 Ω
ð
Þ
ð3Þ Changes in solid-phase volumes and porosities ϕ caused
by the mineral reactions were calculated according to: ð3Þ Δφt ¼
1
P
i ni;tvi
Vtotal
φt¼0
ð6Þ ð6Þ where k+ is the far-from-equilibrium dissolution rate co-
efficient. The saturation state term 1-Ω was approxi-
mated to 1 (i.e., rate independent to distance from
equilibrium) supported by earlier numerical estimates of
glass-CO2 reactivity suggesting an approximately linear
relation between time and reaction progress for basaltic
glass [18]. This expression takes into account the effect
of pH as well as the effect of the concentration of solutes
such as fluoride as they complex with Al3+ and reduce
the Al3+ activity [19]. The specific surface area for basalt
(m2/g) was estimated by: where ϕt=0 is the initial porosity, n and v are moles and
molar volume of mineral i respectively, and Vtotal is the
total volume of the system. The basalt was defined to consist of a mixture of glass
and crystalline basalt with mineral and glass fractions
chosen based on reported data from CRB [6,24,25]. To
represent
the
crystalline
basalt,
plagioclase
(Ca0.5Na0.5Al1.5Si2.5O8)
and
the
pyroxenes
augite
(Ca0.7Fe0.6
Mg0.7Si2O6)
and
pigeonite
(Ca1.14Fe0.64
Mg0.22Si2O6) were chosen. Methods System 1 (S1) is close to the injector and
contains a wet CO2 (0.5 mole% H2O at 100 bar and 40 C) with no
residual water; system 2 (S2) is fully in the H2O rich phase with CO2
diffusing in from the plume boundary; and system 3 (S3) is at the
boundary of the CO2 plume with both sufficient non-wetting CO2 at
a constant CO2 partial pressure of 100 bar and with sufficient water
wetting the mineral surfaces and available for reactions. Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Page 3 of 12 Dissolution rates of minerals in the basaltic rock were
calculated according to a kinetic equation taking into ac-
count pH and the distance from equilibrium: and ρs (g/m3) is the density of the basalt solid estimated
from the fraction of the individual basalt components. A
Ssp value of 1.52 × 10-5 m2/g basalt (= 0.137 m2/Kg
water) was obtained for the CRB using an average basalt
solid density of 2.93 × 106 g/m3 with 10% connected pore
space and a Ap/Vp ratio of 400 m-1 [5]. The reactive sur-
face area was calculated from the mass of the glass and
minerals present according to: rþ ¼ S kHexp Ea;H
RT
a
nH
H þ kN exp Ea;N
RT
þ kOHexp Ea;OH
RT
a
nOH
H
1 Ω
ð
Þ
ð1Þ ð1Þ Si ¼ MiniSspXr
ð5Þ ð5Þ Si ¼ MiniSspXr where S is the reactive surface area (m2), ki are rate con-
stants (moles/m2s), aH is the H+ activity, n is the reac-
tion order with respect to H+ and OH-, and Ω is the
saturation state given by: where M and n are molar mass and moles of mineral i,
and Xr is the fraction of the total mineral surface that is
reactive. As Xr is highly uncertain and is suggested to
vary by orders of magnitude [20,21], we used a value of
0.1 for the base case and varied Xr from 1 to 10-3. Methods The hydrolysis equilibrium
constants of these phases were estimated using the
PHREEQC program assuming ideal solid solutions of
the end-members enstatite, ferrosilite and wollastonite
for the pyroxenes, and albite and anorthite for the Ssp ¼
ϕ
1 ϕ
Ap
ρsVp
ð4Þ ð4Þ where the ratio Ap/Vp denotes the ratio between pore
surface and pore volume (m-1), ϕ is connected porosity, where the ratio Ap/Vp denotes the ratio between pore
surface and pore volume (m-1), ϕ is connected porosity, Table 1 Kinetic parameters for dissolution of primary minerals based on empirical data given in Palandri and Kharaka
[16]
d f
b
l i
l
f
[17] eters for dissolution of primary minerals based on empirical data given in Palandri and Kharaka Table 1 Kinetic parameters for dissolution of primary minerals based on empirical data given in P
f
f Table 1 Kinetic parameters for dissolution of primary minerals based on empirical data given in P
[16] and for basaltic glass from [17] Table 1 Kinetic parameters for dissolution of primary minerals based on empirical data given in Palandri and Kharaka
[16] and for basaltic glass from [17]
k_H (mol/m2s)
EaH kJ/mol
nH
k_N (mol/m2s)
EaNkJ/mol
k_OH (mol/m2s)
EaOHkJ/mol
nOH
References
Augite
1.58e-7
78
0.7
1.07e-12
78
-
-
[16]
Pigeonite
1.58e-7
78
0.7
1.07e-12
78
-
-
[16]
Feldspar
1.58e-9
53.5
0.541
3.39e-12
57.4
4.78e-15
59
-0.57
[16]
glass
1e-10
25.5
1
-
-
-
-
-
[17]
Magnetite
2.57e-9
18.6
0.279
1.66e-11
18.6
-
-
-
[16] Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Page 4 of 12 Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 total amount of 10 moles/kgw formed after 10000 years
(Figure 2d). Other carbonates, such as ankerite, dolo-
mite, magnesite, and dawsonite did not form as elements
such as Mg and Ca was consumed by the non-carbonate
secondary phases. plagioclase. Equilibrium constants for the solid solutions
for temperatures up to 100 C were estimated with
PHREEQC and from these data coefficients a to e for
the PHREEQC built-in analytical expression (log K = a +
bT + c/T + dlog10(T) + e/T2) were estimated using non-
linear regression in MATLAB. The effect of temperature on the basalt hydration and
carbonation was investigated by simulating the system at
60, 80 and 100 C (Figure 3). Methods As for the 40 C simulation
we see that basaltic glass dissolves orders of magnitude
faster than the crystalline basalt components and the
glass is the major source for the secondary phases. The
dissolution rates of the basalt components scale expo-
nentially with temperature, and the glass is almost The glass composition (Ca0.015Fe0.095 Mg0.065Na0.025
K0.01Al0.105 S0.003Si0.5O1.35) was taken from [6] and
modified by adding a small fraction of sulfur which is a
common minor constituent of the CR basaltic glass [26]. The secondary mineral assemblage was chosen based
on reports on basalt weathering [27-30], with additional
carbonates that could potentially form at elevated CO2
pressures from the release of Fe, Mg and Ca. The anker-
ite composition chosen for this work was CaFe0.6 Mg0.4
(CO3)2 which corresponds to a solid solution of 0.6
ankerite (CaFe(CO3)2) and 0.4 dolomite (CaMg(CO3)2). Because ankerite (CaFe0.6 Mg0.4(CO3)2) was not listed in
the thermodynamic database, we estimated values using
the same approach as in [31]. The full list of secondary
minerals is given in Table 2. Table 2 Mineralogy included in the model
Initial
Weight %
Density
(g/cm3)
2,3Log K0
Primary minerals
1Augite (En0.35Fs0.3Wo0.35)
16
3.40
21.00
1Pigeonite (En0.57Fs0.32Wo0.11)
3
3.38
21.40
1Plagioclase (An50)
35
2.68
14.20
Glass Ca0.015Fe0.095Mg0.065
Na0.025K0.01Al0.105 S0.003Si0.5O1.35
45
2.92
-99.00
Magnetite
1
5.15
10.47
Secondary minerals
SiO2(am)
0
2.62
-2.71
Albite
0
2.62
2.76
Goethite
0
3.80
0.53
Calcite
0
2.71
1.85
Hematite
0
5.30
0.11
Kaolinite
0
2.60
6.81
Smec high Fe-Mg
0
2.70
17.42
Saponite-Mg
0
2.40
26.25
Celadonite
0
3.00
7.46
Stilbite
0
2.15
1.05
Dawnsonite
0
2.42
4.35
Siderite
0
3.96
-0.19
1Ankerite (Ank0.6Do0.4)
0
3.05
-19.51
Dolomite
0
2.84
4.06
Magnesite
0
3.00
2.29 Table 2 Mineralogy included in the model To simulate the CRB-CO2 interaction we used the
average concentrations of solutes reported for the Grand
Ronde Formation (Table 3). As supercritical CO2 (scCO2
at T > 31.1 C; P > 73.9 bar) is the preferred choice for
CO2 storage, based on higher density compared to gas-
eous
CO2, we simulated
aqueous-phase basalt-CO2
interaction at a depth of 800 meters at a CO2 pressure
of 100 bar and temperatures of 40 to 100 C. The reactiv-
ity of basalt and a H2O saturated scCO2 phase was
simulated using an estimated 0.5 mol% H2O and a CO2
density of 600 g/cc giving an initial mass of 0.003 Kg
H2O per 1 liter pore space. i) CRB mineral and glass dissolution and formation of
secondary minerals The
reaction
rates
increased
however
with
temperature
and
the
time
needed
to reach
the maximum
potential
therefore
decreased with temperature (Figure 4). i) CRB mineral and glass dissolution and formation of
secondary minerals Following the injection of CO2 into the system, pH im-
mediately decreased from 9.5 to below 4, and thereafter
gradually increased to 5.8 at the end of 10000 years
(Figure 2a). At the acidic pH secondary phases such as
saponite
(Ca0.165Mg3Al0.33Si3.67O10(OH)2),
celadonite
(KMgAlSi4O10(OH)2) and zeolite (stilbite) were thermo-
dynamically stable and formed (Figure 2b). Glass dis-
solved orders of magnitude faster than the crystalline
basaltic constituents and more than half dissolved after
10000 years (Figure 2c). The dissolution rate of glass
was not increased by the aqueous fluoride as the Al3+
activity was fixed by the kaolinite and amorphous silica
equilibria. The fluoride therefore only increased the total
soluble aluminium. The steady release of Fe from the
basalt saturated the water with respect to siderite and a The mineralogy of the CRB has been described in [25,32] and the weight
fraction of pyroxene, feldspar and glass was estimated as average values from
the reported data. 1Solid solutions. En (enstatite), Fs (ferrosilite), Wo (wollastonite), An (anorthite),
Ank (ankerite), Do (dolomite). For details on the calculations of the ankerite
solid solution see [31]. 2Superscript 0 denotes Standard state (T = 298K, P = 1 atm). The equilibrium
constant log K value is that for the forward dissolution reaction for one mole
unit of the mineral. 3All thermodynamic data (log K and coefficients for the PHREEQC analytical
temperature expression) from the llnl.dat PHREEQC database, except for the
solid-solutions first estimated in PHREEQC by ideal solid solutions and then
added to the PHREEQC database as new solid solution phases. Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Page 5 of 12 Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Table 3 Composition of initial formation water
Elements (totals)
Mol/kgw
Na
1.0 x 10-3
Ca
6.0 x 10-4
K
1.0 x 10-4
Mg
2.0 x 10-5
Fe
1.2 x 10-6
Alkalinity (HCO3
-)
2.0 x 10-3
Cl
3.0 x 10-4
S (SO4
2-)
1.0 x 10-4
Si
2.0 x 10-4
Al
1.0 x 10-6
Log(O2)
-10.68
pH
7.5 siderite was the only phase that formed. At 80 and
100 C, magnesite and later ankerite formed together
with siderite. Taking zero porosity as the maximum ex-
tent of possible reactions we see that the total amount of
CO2 trapped as solid carbonates did not change much
with
temperature
(Figure
4). iii) Reduction of pore-space as a function of reactive surface
area As the reactive surface area is a large uncertainty we
simulated the changes of porosity over a range of values
from a maximum being equal to the estimated physical
surface area S0 (equation (5) with Xr = 1) to a three
orders of magnitude reduction (Figure 6). The physical
conditions of the simulated system was the same as for
the base-case at 40 C and a CO2 pressure of 100 bars. At a reactive surface area that is equal to the estimated ii) On the limitation of pore-space for the basalt
carbonation Figure 3 Mass fractions of minerals following basalt alteration at 60, 80, and 100 C over 10000 years: a, d, g) primary basalt minerals and
glass; b,e, h) secondary phases except the carbonates; and c, f, j) carbonates. of minerals following basalt alteration at 60, 80, and 100 C over 10000 years: a, d, g) primary basalt minerals and
hases except the carbonates; and c, f, j) carbonates. basalt-CO2 alteration which shows loss of porosity and a
rapid reduction of permeability during CO2-basalt inter-
actions (e.g., [33]). S0 all porosity is lost after approximately 1000 years as
stilbite and siderite fills the pore space. If the reactive
surface area is reduced by one order of magnitude (i.e. the base case) nearly 1/10 of the original 10% porosity is
preserved. Further reductions by one and two orders of
magnitude lead to smaller changes and at three orders
of magnitude reduction relative to S0 almost no change
is observed (Figure 6). ii) On the limitation of pore-space for the basalt
carbonation Secondary phases such as stilbite and amorphous silica
have lower density than the basalt components and al-
teration therefore leads to a reduction of pore space. At
the presence of CO2, secondary carbonates further re-
duce the pore space. For the volume calculations we
used expression (6) with the molar volumes listed in
Table 2. At 40 C, the starting porosity of 10% is reduced
to 0.85% after 10000 years. At the higher temperatures
all porosity is lost after 2700, 1200, and 300 years re-
spectively at 60, 80, and 100 C (Figure 5). Taking the ex-
treme of 0% porosity as the limit for the reactions we
obtain a maximum carbonation potential (mol CO2
stored/Kgw) at the different temperatures of 13.5, 29.3,
and 28.5 moles for 60, 80, and 100 C (Figure 5). The
simulated clogging of the pore space fits well with short-
term laboratory percolation experiments on open-system completely dissolved after 10000 years at 60 C, whereas
the time for a complete dissolution takes 4000 and
1500 years at 80 and 100 C respectively (Figure 3a, d, g). The secondary mineral assemblages were largely the
same for all temperatures. Stilbite dominated together
with amorphous silica (40 and 60 C) and quartz (80 and
100 C) (Figure 3b, e, h). Saponite formed at 40 and
60 C, but not at higher temperatures. Other secondary
minerals such as albite, celadonite, and kaolinite formed
at all conditions. At 60 C, magnesite and dolomite were
still considered to be too slow to form (see [29]) and Figure 2 Basalt alteration at 40 C and 100 bar CO2 pressure over 10000 years. a) pH changes; b) mass fractions of basaltic glass and
crystalline basalt components; c) secondary phases formed; and d) moles of secondary carbonates (siderite) formed per kgw. Figure 2 Basalt alteration at 40 C and 100 bar CO2 pressure over 10000 years. a) pH changes; b) mass fractions of basaltic glass and
crystalline basalt components; c) secondary phases formed; and d) moles of secondary carbonates (siderite) formed per kgw. Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Page 6 of 12 Page 6 of 12 Figure 3 Mass fractions of minerals following basalt alteration at 60, 80, and 100 C over 10000 years: a, d, g) primary basalt minerals and
glass; b,e, h) secondary phases except the carbonates; and c, f, j) carbonates. System 2: The potential for carbonate growth in a H2O-
saurated scCO2 phase The reaction between H2O dissolved in scCO2 and bas-
alt was simulated at 100 bar pressure and 40 C. The ini-
tial amount of water was 0.003 Kg and no H2O was Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Page 7 of 12 Figure 4 CO2 trapped in solids for 40 to 100 C simulations at
100 bar CO2 pressure. A cut-off value is used when all pore space
is filled up with the secondary phases (see Figure 5). The simulations
suggest that the total amount of secondary carbonates that form is
dictated by the available pore space and the thermodynamic
stability of secondary phases rather than temperature, whereas
carbonate generation rates depend on the exponential increase of
basalt dissolution rates with temperature. Figure 6 Porosity changes caused by the basalt alteration at
40 C and specific surface areas ranging from So (estimated
total basalt surface area) down to a three orders of magnitude
reduction. The base case specific surface used was So/10. Figure 6 Porosity changes caused by the basalt alteration at
40 C and specific surface areas ranging from So (estimated
total basalt surface area) down to a three orders of magnitude
reduction. The base case specific surface used was So/10. Figure 4 CO2 trapped in solids for 40 to 100 C simulations at
100 bar CO2 pressure. A cut-off value is used when all pore space
is filled up with the secondary phases (see Figure 5). The simulations
suggest that the total amount of secondary carbonates that form is
dictated by the available pore space and the thermodynamic
stability of secondary phases rather than temperature, whereas
carbonate generation rates depend on the exponential increase of
basalt dissolution rates with temperature. Figure 4 CO2 trapped in solids for 40 to 100 C simulations at
100 bar CO2 pressure. A cut-off value is used when all pore space
is filled up with the secondary phases (see Figure 5). The simulations
suggest that the total amount of secondary carbonates that form is
dictated by the available pore space and the thermodynamic
stability of secondary phases rather than temperature, whereas
carbonate generation rates depend on the exponential increase of
basalt dissolution rates with temperature. (Figure 7a). At this point stilbite was unstable and sup-
plied water until all water was consumed after approxi-
mately 100 years (Figure 7a). System 2: The potential for carbonate growth in a H2O-
saurated scCO2 phase Following the basalt
hydration,
siderite
and
ankerite
formed
from
the
released Ca, Mg, and Fe, with a final total amount of 0.2
moles CO2 consumed per liter pore space after 100 years
(Figure 7b). If H2O had been allowed to dissolve into the
scCO2 phase from residual aqueous phases trapped in
the smaller pores, the carbonation potential would have
been larger. This process is however, to the knowledge
of the authors, not possible to simulate using the
PHREEQC code, and was therefore outside the scope of
this study. allowed to enter the system. This is an ideal end-
member case and serves to illustrate the carbonation po-
tential in a volume with limited hydration potential. As secondary phases such as stilbite formed, water
was rapidly consumed and most gone after 45 years Figure 5 Porosity changes caused by the basalt alteration at 40
to 100 C. Secondary hydrated species and carbonates incorporate
the H2O and CO2 masses into the solids and clogs the pore space. As reaction rates increase exponentially with temperature, the pore
space is filled up faster at the higher temperatures. System 3: 1D diffusion of CO2 into the CRB aquifer
To see how the basalt reacted under different CO2 pres-
sures, we defined a 1D diffusion–reaction simulation. This provided us with basalt-CO2 interactions over a
continuous range of CO2 pressures from the background
1 bar up to the maximum 100 bars. The system corre-
sponds to a stagnant zone presented as a column with
one end close to the boundary of the injected CO2
plume and the other end further away from the plume
(Figure 1). The distance reached for the CO2 into the
column is given by the balance between diffusive trans-
port and removal of carbon by secondary carbonate for-
mation. We therefore varied reaction rates from no
reaction giving the maximum lengtht of diffusive trans-
port, and up to the base-case rate given by a reaction
surface area 1 order of magnitude lower than the esti-
mated physical surface area. Figure 8 shows pH, dis-
solved
CO2
(mol/Kgw)
and
amount
of
secondary
carbonate formed in the 1D column. As CO2 diffuses
into the column pH drops to approximately 4 at full Figure 5 Porosity changes caused by the basalt alteration at 40
to 100 C. Secondary hydrated species and carbonates incorporate
the H2O and CO2 masses into the solids and clogs the pore space. System 2: The potential for carbonate growth in a H2O-
saurated scCO2 phase As reaction rates increase exponentially with temperature, the pore
space is filled up faster at the higher temperatures. Figure 5 Porosity changes caused by the basalt alteration at 40
to 100 C. Secondary hydrated species and carbonates incorporate
the H2O and CO2 masses into the solids and clogs the pore space. As reaction rates increase exponentially with temperature, the pore
space is filled up faster at the higher temperatures. Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Page 8 of 12 Figure 7 Basalt alteration in the scCO2 rich phase with initial 3 grams of water per liter pore space. A) as zeolites form H2O is consumed
and the water activity is reduced. After approximately 45 years most water is consumed, whereas all is gone before 100 years. B) siderite formed
the secondary carbonate initially followed by ankerite. Figure 7 Basalt alteration in the scCO2 rich phase with initial 3 grams of water per liter pore space. A) as zeolites form H2O is consumed
and the water activity is reduced. After approximately 45 years most water is consumed, whereas all is gone before 100 years. B) siderite formed
the secondary carbonate initially followed by ankerite. Uncertainty on the choice of secondary phases used in
the model saturation. The depth of diffusion into the 1D column is
approximately 40 meters at 1000 years if no carbonate
forming reactions occur (Figure 8a). The penetration
depth decreased rapidly if reactions were allowed as sid-
erite formed and pulled carbon out of aqueous system
(Figure 8b, c). At the highest reaction rate (base-case),
CO2 diffused less than 10 meters into the column as ap-
proximately 13 moles/Kgw of siderite formed at the end
of the 10000 years simulation (Figure 8d). Siderite
formed at greater depth if the reactive surface area was
reduced by another order of magnitude, but less formed
in total (Figure 8d). Growth rate experiments of carbonates such as magne-
site and dolomite have shown that the activation energy
is high and that growth is negligible at low temperatures
(e.g., [35-37]). Dissolution rate studies of siderite sug-
gests that the reaction rate is intermediate between cal-
cite and magnesite [38,39], and growth rate data suggest
that siderite may form down to room temperature [40]. Data on ankerite dissolution and growth is to the know-
ledge of the authors not known. The crystallographic
and physical characteristics of ankerite do resemble
those of dolomite and siderite, and the chemistry is
related to dolomite with the Mg2+ substituted by various
amounts of Fe2+ and Mn2+. If the growth rate is close to
the magnesian carbonates such as dolomite and magne-
site [41,42], the amount that may form during low-
temperature alteration is likely low. In this case, more
iron would be available for siderite growth. If on the
other hand the growth rate is closer to siderite, we
would expect ankerite or other FeMg solid solution car-
bonates to grow during low-temperature alteration. Discussion y
The reactive surface area is considered as a major source
of uncertainty (e.g., [20,34]) and this leads to corre-
sponding high uncertainties in timing and extent of
reactions as dissolution rates have a first order depend-
ence on reactive surface areas. Weathering rates in na-
ture are commonly observed to be 1–3 orders of
magnitude lower than in laboratory experiments (e.g.,
[20,21,34]), and this may in part be explained by differ-
ences in reactive and physical (total) surface area be-
tween experimental and natural systems. We assumed in
this study a base-case reactive surface area 1 order of
magnitude lower than the estimated physical surface
area for the basalt. A further two orders of magnitude
reduction in the reactive surface area, which is within
the
range
of
values
expected
for natural
systems,
resulted in little basalt alteration and only minor reduc-
tion of porosity (see Figure 5). A better understanding of
the surface area of porous basalt and the effect of time
(aging) on features such as dislocation densities and re-
active surface areas are therefore required to understand
the potential for CO2 mineral storage in basaltic rocks. One uncertainty related to the local-equilibrium as-
sumption is on the growth retention time for the sec-
ondary carbonates. The local-equilibrium assumption
predicts growth of the secondary phases as soon as an
infinitesimally small supersaturation is reached [23]. The
time it takes to nucleate sufficient mass to initiate a sig-
nificant growth may however be hundreds to thousands
of years for some secondary phases [31]. There are no
nucleation rate data for siderite and ankerite and the re-
tention time is hence unknown. Finally, the total potential for secondary carbonate
growth may be affected by the amount of magnesium
and iron that enters ferromagnesian calcites. As a signifi-
cant fraction of the metal cations may substitute for cal-
cium (e.g., [43]), a iron-magnesium rich calcite may Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Page 9 of 12 Figure 8 1D reaction–diffusion of CO2 into permeable basalt. Partial pressure of the inlet boundary was fixed at 100 bar and with
a column temperature of 40 C. As consumption of CO2 by siderite
growth affects the depth of CO2 diffusion, we ran a sensitivity study
on reactive surface area going from no reaction (a) up to the base
case (c). Discussion Finally the amount of CO2 trapped as solid carbonate
(siderite) was compared for the base case and reduced reactive
surface area (d). We see that the reaction rates strongly constraint
the depth of the column affected by the CO2 diffusion. potentially form rather than ankerite and thereby reduce
the amount of siderite formed. Comparisons to experiments, numerical simulations and
natural analogues of basalt-CO2 interactions The initial mineralogy was similar to our study
whereas the temperature of 60 C was slightly higher
than our base case 40 C. In [18] the CO2-basalt interac-
tions were stretched to last for more than 280000 years
compared to our 10000 years perspective. The main dif-
ferences between our model and [18] are on the choice
of secondary mineral assemblage, and on the focus of
limiting factors such as the availability of water for hy-
dration in the present work. The lack of zeolites and hy-
drous phases other than kaolinite and goethite in [18]
made Ca available for secondary carbonates and the total
potential for carbonate formation was higher than in our
work. Marini allowed dolomite and magnesite to form at
60 C, whereas our simulations only produced siderite at
the similar conditions. Moreover, the formation of daw-
sonite in [18] is still uncertain and possibly limited at
high silica activities and with an assemblage of stable
NaAl-silicates defined to form [45]. Based on two differ-
ent approaches, the reactive surface area for basalt
was estimated to quite similar values. We estimated a
specific surface area of approximately 1.5x10-5 m2/gbasalt
(= 0.14 m2/Kg water at 10% porosity) based on the Ap/Vp
values estimated by [46] and reported in [5], and reduced
this value by one order of magnitude to get the reactive
surface area. Marini used a geometric model giving a
reactive surface area of 0.41 m2/Kg water. The higher
reactive surface area and higher temperature of [18]
resulted in faster reactions and more rapid clogging of the
pore space (within a few years). Studies of natural basalt
systems at similar or higher temperatures may give some
insight into how fast pore space is clogged by basalt
hydration or carbonation, and this should be used to
improve the estimates of reactive surface areas of basalt
for future studies of a solid basalt. Although no inverse modeling was
done to estimate the reactive surface area of the basalt
in [44], fragmented basaltic rocks such as hyaloclastite
breccias are expected to have significantly higher react-
ive surface areas than porous solid basalts, and they are
therefore correspondingly more reactive. One example of a natural analogue that shed light on
CO2 basalt interactions is the CO2 charged basalt hosted
groundwaters at Hekla, Iceland. Comparisons to experiments, numerical simulations and
natural analogues of basalt-CO2 interactions Our simulations suggest that the potential for carbonate
growth is limited to siderite or FeMg carbonates at low
temperatures as secondary phases such as zeolites out-
competed the carbonates for calcium. We here compare
our simulated results with reported data on CO2-basalt
interactions from laboratory experiments, natural analo-
gues, and other reported numerical simulations. g
The reactivity of CRB and other continental flood
basalts are available from the long-term (months to
years) laboratory experiments done by Schaef and co-
workers [6,24]. In these experiments basalt samples from
USA, India, South Africa, and Canada were reacted with
CO2 at about 100 bars and 60 to 100 C. Reacted samples
from these experiments showed generation of Ca-rich
carbonates interpreted as calcites with minor siderite
and magnesite. In experiments on CRB using mixtures
of H2S and CO2 at 60 C and 100 bar and run for
181 days, pyrite (FeS2) formed together with Mg-Fe poor
calcite and a Ca-poor Fe-carbonate [6]. Our simulations
at the same temperatures show rapid formation of sider-
ite (60 C) or siderite and magnesite at higher tempera-
tures (Figure 3). Our simulations do not predict any
calcite growth as the calcium activity is lowered by zeo-
lite formation. Calcite would however form in our mod-
els if the zeolites were not allowed to form at local
equilibrium, and possibly if a magnesian ferroan (solid
solution) calcite was used in the model instead of the
pure end-member calcite. Therefore, the apparent differ-
ence between our model and the experiment may be
caused by our use of the local equilibrium assumption,
whereas the zeolites in the laboratory experiments did
not form at low temperatures due to slow kinetics. Re-
cent experiments on basalt dissolution support the pre-
ferential release of Mg and Fe over Ca at acidic
conditions [22], suggesting that the MgFe-carbonates
will dominate as secondary carbonates during CO2 stor-
age in basalt. Our numerical simulations share some similarities to
other works such as by Marini [18] and Gysi [44], but Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Page 10 of 12 Page 10 of 12 our model and hence the outcome is different in several
aspects. The most comprehensive work done earlier is
the numerical simulations done by Marini [18] on the
reactivity of crystalline and glassy CFB following CO2
storage. Comparisons to experiments, numerical simulations and
natural analogues of basalt-CO2 interactions Solution aqueous spe-
cies sampled from natural cold springs and rivers here
showed a drop in total inorganic carbon (TIC) that was
interpreted to result from considerable formation of sec-
ondary carbonate phases such as calcite [47]. Reaction
path modeling of the system suggests however that the
carbonate formation is associated with high pH in ac-
cordance with the low TIC in the sampled waters. This
system is therefore different from basalt CO2 storage
projects where higher CO2 pressures may be maintained
over time and the pH is lower. In addition to calcite,
dolomite was also suggested as a potential storage host
for the low temperature reactions in Hekla [47]. This
may however be questionable as long-term laboratory
experiments at room temperature have failed to form
dolomite even at significant super saturations [48],
explained by the high activation energy for dolomite
growth [32,41]. Another natural analogue that more
closely corresponds to industrial CO2 storage is the
basalt-hosted petroleum reservoir on Nuussuaq, West
Greenland. In this system the bulk carbonate formation
appears to have occurred as secondary weathering pro-
ducts. Other alteration products such as zeolites and
oxides were replaced by dolomite, magnesite, siderite,
and calcite at temperatures of 70–120 C [49]. Therefore,
taking into account the basalt weathering products and
not only primary basalt minerals appears to be vital in
estimating the total potential for secondary carbonate
formation and the long-term potential for CO2 storage
in basalt systems. Another numerical study on low-temperature (25 C,
30 bar CO2) basaltic glass alteration was presented by
Gysi et al. [44]. Again a main difference is on the choice
of secondary minerals. Gysi et al. [44] allowed dolomite,
magnesite, and Fe-Mg carbonate to form together with
calcite and siderite, whereas we did not allow other Mg-
Fe carbonates to form than ankerite. As previously sta-
ted, the low-temperature formation of dolomite and
magnesite is not likely because of the high apparent acti-
vation energy and small kinetic coefficients for the
growth of Mg-carbonates [35-37]. Other carbonates such
as siderite and potentially FeMg-calcites are more likely
to form at these low temperatures. The high reactive
surface area used in [44] is based on a geometric model
for glass fragments, and is hence not directly comparable
with the surface area estimated for a vesicle pore space References 1. Holloway S: Underground sequestration of carbon dioxide–a viable
greenhouse gas mitigation option. Energy 2004, 30:2318–2333. 2. Bachu S, Bonijoly D, Bradshaw J, Burruss R, Holloway S, Christensen NP,
Mathiassen OM: CO2 storage capacity estimation: Methodology and gaps. Int J Greenhouse Gas Control 2007, 1:430–443. 26. Blake S, Self S, Sharma K, Sephton S: Sulfur release from the Columbia
River Basalts and other flood lava eruptions constrained by a model of
sulfide saturation. Earth Planet Sci Lett 2010, 299:328–338. 3. IPCC: IPCC Special Report on Carbon Dioxide Capture and Storage. In In
Prepared by Working Group III of the Intergovernmental Panel on Climate
Change. Edited by Metz B, Davidson O, Coninck H, Loos M, Meyer L. Cambridge, United Kingdom and New York, NY, USA: Cambridge University
Press; 2005:442 pp. 27. Neuhoff PS, Fridriksson T, Arnorsson S, Bird DK: Porosity evolution and
mineral paragenesis during low-grade metamorphism of basaltic lavas
at Teigarhorn, eastern Iceland. Am J Sci 1999, 299:467–501. 28. Neuhoff PS, Rogers KL, Stannius LS, Bird DK, Pedersen AK: Regional very
low-grade metamorphism of basaltic lavas, Disko-Nuussuaq region, West
Greenland. Lithos 2006, 92:33–54. 4. Oelkers EH, Gislason SR, Matter J: Mineral Carbonation of CO2. Elements
2008, 4:333–337. 4. Oelkers EH, Gislason SR, Matter J: Mineral Carbonation of CO2. Elements
2008, 4:333–337. 29. Stefansson A, Gislason SR, Arnorsson S: Dissolution of primary minerals in
natural waters - II. Mineral saturation state. Chem Geol 2001, 172:251–276. 5. McGrail BP, Schaef HT, Ho AM, Chien Y-J, Dooley JJ, Davidson CL: Potential
for carbon dioxide sequestration in flood basalts. J Geophysical Res-Solid
Earth 2006, 111:1–13. 5. McGrail BP, Schaef HT, Ho AM, Chien Y-J, Dooley JJ, Davidson CL: Potential 30. Reidel SP, Johnson VG, Spane FA: Natural gas storage in basalt aquifers of
the Colombia Basin, Pacific Northwest USA: A guide to site characterization. H:
Report Pacific Northwest National Laboratory; PNNL-13962; 2002. 6. Schaef HT, McGrail BP, Owen AT: Carbonate mineralization of volcanic
province basalts. Int J Greenhouse Gas Control 2010, 4:249–261. 6. Schaef HT, McGrail BP, Owen AT: Carbonate mineralization of volcanic
province basalts. Int J Greenhouse Gas Control 2010, 4:249–261. 31. Pham VTH, Lu P, Aagaard P, Zhu C, Hellevang H: On the potential of CO2-
water-rock interactions for CO2 storage using a modified kinetic model. International Journal of Greenhouse Gas Control 2011, 5:1002–1015. 7. Authors’ contributions
h
h
d
b VTH Pham has made substantial contributions to conception and designs
the manuscript. Her contributions are to acquisition of data, analysis and
interpretation of data. PA did revising and has given final approval of the
version to be published. HH involved in drafting the manuscript, writing
methodology part and revising of the final version. All authors read and
approved the final manuscript. 19. Wolff-Boenisch D, Gislason SR, Oelkers EH: The effect of fluoride on the
dissolution rates of natural glasses at pH 4 and 25 degrees C. Geochim
Cosmochim Acta 2004, 68:4571–4582. 20. White AF, Peterson ML: Role of reactive-surface-area characterization in
geochemical kinetic models. In In. Chemical Modeling of Aqueous Systems II. Edited by Melchior D, et al. Washington, DC: ACS Symposium Series, ACS; 1990 Summary and conclusions Simulations of closed-system (PCO2 = 100 bar, 40 C) and
1D reaction–diffusion (PCO2 = 0–100 bar, 40 C) alter-
ation of basalt suggest that the potential of secondary
carbonate formation is limited to siderite at low tem-
peratures as divalent metal cations are preferentially
consumed by zeolites and oxides. Higher temperatures
60 – 100 C appear to be in favor of secondary carbonate
formation, allowing the precipitation of carbonates such
as magnesite, siderite and possibly dolomite and other
FeMg carbonates (ankerite). Given an unlimited source
of CO2 (fixed CO2 pressure), the total amount of CO2
stored as solid carbonates is orders of magnitude higher
than the 1–2 mol/Kg water solubility in the formation
water (Figure 4). The total amount trapped might Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Page 11 of 12 however be reduced if CO2, H2O or pore space are limit-
ing factors. The formation of secondary hydrous and
carbonate phases increases the volume of solids and the
porosity is correspondingly reduced (Figure 5). This to-
gether with the immobilization of CO2 by solid carbon-
ate formation is in favor of safe long-term storage of
CO2 in basaltic aquifers. 13. Coan CR, King AD: Solubility of Water in Compressed Carbon Dioxide,
Nitrous Oxide, and Ethane - Evidence for Hydration of Carbon Dioxide
and Nitrous Oxide in Gas Phase. J Am Chem Soc 1971, 93:1857–1862. 14. Sabirzyanov AN, Ilìn AP, Akhunov AR, Gumerov FM: Solubility of water in
Supercritical carbon dioxide. High Temp 2002, 40:203–206. 15. Bahar M, Lui K: Measurement of the diffusion coefficient of CO2 in formation
water under reservoir conditions: Implication for CO2 storage, Society of
Petroleum Engineers, SPE Asia Pacific Oil and Gas Conference and Exhibition,
20–22 October 2008. Perth, Australia:; 2008. 16. Palandri JL, Kharaka YK: A Complilation of Rate Parameters of Water-Mineral
Interaction Kinetics for Applicatin to Geochemical Modeling. U.S. Geological
survey, Open report.: ; 2004:1068. Competing interests
W h
i We have no competing interests with any organization to publish the
manuscript ‘On the potential for CO2 mineral storage in continental flood
basalts – PHREEQC batch- and 1D diffusion-reaction simulations‘. 17. Gislason SR, Oelkers EH: Mechanism, rates, and consequences of basaltic
glass dissolution: II. An experimental study of the dissolution rates of
basaltic glass as a function of pH and temperature. Geochim Cosmochim
Acta 2003, 67:3817–3832. 18. Marini L: Chapter 7: Reaction path Modelling of geological CO2 sequatration,
In Geological Sequestration of Carbon Dioxide Thermodynamics, Kinetics, and
Reaction Path Modeling.: Dev Geochem; 2006:319–394. Acknowledgements 21. White AF, Brantley SL: The effect of time on the weathering of silicate
minerals: why do weathering rates differ in the laboratory and field? Chem Geol 2003, 202:479–506. We highly appreciated constructive comments and suggestions from the
reviewers. This work has been funded by SSC-Ramore (Subsurface storage of
carbon dioxide - risk assessment, monitoring and remediation) project and
(partially) by SUCCESS centre for CO2 storage under grant 193825/S60 from
Research Council of Norway (RCN). SUCCESS is a consortium with partners
from industry and science, hosted by Christian Michelsen Research as. 22. Gudbrandsson S, Wolff-Boenisch D, Gislason SR, Oelkers EH: An experimental
study of crystalline basalt dissolution from 2 ≤pH ≤11 and temperatures
from 5 to 75 C. Geochim Cosmochim Acta 2011, 75:5496–5509. 23. Helgeson HC: Evaluation of Irreversible Reactions in Geochemical
Processes Involving Minerals and Aqueous Solutions .I. Thermodynamic
Relations. Geochim Cosmochim Acta 1968, 32:853–877. Received: 1 June 2011 Accepted: 14 June 2012
Published: 14 June 2012 24. Schaef HT, McGrail BP, Owen AT: Basalt-CO2-H2O Interactions and Variability
in Carbonate Mineralization Rates. Energy Procedia 2009, 1:4899–4906. 25. Schaef HT, McGrail BP: Dissolution of Columbia River Basalt under mildly
acidic conditions as a function of temperature: Experimental results
relevant to the geological sequestration of carbon dioxide. Appl Geochem
2009, 24:980–987. References Schaef HT, Windish CF, McGrail BP, Martin PF, Rosso KM: Brucite [Mg(OH)2]
carbonation in wet supercritical CO2: An in situ high pressure x-ray
diffration study. Geochim Cosmochim Acta 2011, 75:7458–7471. 32. Reidel SP: A lava flow without a source: the cohassett flow and its
compositional components, Sentinel Bluffs Member, Columbia River
Basalt Group. J Geol 2005, 113:1–21. 8. White MD, McGrail BP, Schaef HT, Hu JZ, Hoyt DW, Felmy AR, Rosso KM,
Wurstner SK: Multiphase sequestration geochemistry: Model for mineral
carbonation. Energy Procedia 2011, 4:5009–5016. 33. Peuble S, Godard M, Gouze P, Luquot L: CO2 sequestration in olivine rich
basaltic aquifers: a reactive percolation experimental study. Geochim
Cosmochim Acta 2009, 73:A1635. gy
9. Parkhurst DL, Appelo CAJ: User's guide to PHREEQC (version 2) - a computer
program for speciation, reaction-path, 1D-transport, and inverse geochemical
calculations.: U.S. Geological Survey, Water-Resources Investigation Report;
1999:312. 34. Velbel MA: Constancy of silicate-mineral weathering-rate ratios between
natural and experimental weathering: implications for hydrologic control
of differences in absolute rates. Chem Geol 1993, 105:89–99. 10. Soave G: Equilibrium constants from a modified Redlich-Kwrong
equation of state. Chem Eng Sci 1972, 27:1197–1203. 35. Saldi GD, Jordan G, Schott J, Oelkers EH: Magnesite growth rates as a
function of temperature and saturation state. Geochim Cosmochim Acta
2009, 73:5646–5657. 11. Hellevang H, Kvamme B: ACCRETE - Geochemistry solver for CO2-water-
rock interactions. Proceedings GHGT 8 conference 2006, :8p. 12. Bachu S, Stewart S: Geological sequestration of anthropogenic carbon
dioxide in the western Canada sedimentary basin: Suitability analysis. J Can Pet Technol 2002, 41:32–40. 12. Bachu S, Stewart S: Geological sequestration of anthropogenic carbon
dioxide in the western Canada sedimentary basin: Suitability analysis. J Can Pet Technol 2002, 41:32–40. 36. Arvidson RS, Mackenzie FT: The dolomite problem; control of precipitation
kinetics by temperature and saturation state. Am J Sci 1999, 299:257–288. 36. Arvidson RS, Mackenzie FT: The dolomite problem; control of precipitation
kinetics by temperature and saturation state. Am J Sci 1999, 299:257–288. Page 12 of 12 Page 12 of 12 Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 37. Arvidson RS, Mackenzie FT: Tentative kinetic model for dolomite
precipitation rate and its application to dolomite distribution. Aquat
Geochem 1996, 2:273–298. 38. Golubev SV, Bébnézeth P, Schott J, Dandurand JL, Castillo A: Siderite
dissolution kinetics in acidic aqueous solutions from 25 to 100 C and 0
to 50 atm pCO2. Chem Geol 2009, 265:13–19. 39. Pokrovsky OS, Golubev SV, Schott J, Castillo A: Calcite, dolomite and
magnesite dissolution kinetics in aqueous solutions at acid to
circumneutral pH, 25 to 150 C and 1 to 55 atm pCO2: New constraints
on CO2 sequestration in sedimentary basins. Chem Geol 2009, 265:20–32 40. Greenberg J, Tomson M: Precipitation and dissolution kinetics and
equilibria of aqueous ferrous carbonate vs temperature. Appl Geochem
1992, 7:185–190. 41. Arvidson RS, Mackenzie FT: Tentative Kinetic Model for Dolomite
Precipitation Rate and Its Application to Dolomite Distribution. Aquat
Geochem 1997, 2:273–298. 42. Saldi GD, Jordan G, Schott J, Oelkers EH: Magnesite growth rates as
function of temperature and saturation state: An HAFM study. Geochim
Cosmochim Acta 2009, 73:A1149. 43. Busenberg E, Plummer LN: Thermodynamics of magnesian calcite solid-
solutions at 25 C and 1 atm total pressure. Geochim Cosmochim Acta
1989, 53:1189–1208. 44. Gysi AP, Stefansson A: Numerical modelling of CO2-water-basalt
interaction. Mineral Mag 2008, 72:55–59. 45. Hellevang H, Declercq J, Aagaard P: Why is dawsonite absent in CO2
charged reservoirs? Oil & Gas Science and Technology - Re. IFP Energies
nouvelles 2011, 66:119–135. 46. Saar MO, Manga M: The relationship between permeability, porosity, and
microstructure in vesicular basalts.: Master Thesis, Univ. of Oregon, Eugene;
1998:91. 47. Flaathen TK, Gislason SR, Oelkers EH, Sveinbjornsdottir AE: Chemical
evolution of the Mt. Hekla, Iceland, groundwaters: a natural analogue for
CO2 sequestration in basaltic rocks. Appl Geochem 2009, 24:463–474. 48. Land LS: Failure to precipitate dolomite at 25 degrees C from dilute
solution despite 1000-fold oversaturation after 32 years. Aquat Geochem
1998, 4:361–368. 49. Rogers KL, Neuhoff PS, Pedersen AK, Bird DK: CO2 metasomatism in a
basalt-hosted petroleum reservoir, Nuussuaq, West Greenland. Lithos 2006,
92:55–82. doi:10.1186/1467-4866-13-5
Cite this article as: Van Pham et al.: On the potential for CO2 mineral
storage in continental flood basalts – PHREEQC batch- and 1D
diffusion–reaction simulations. Geochemical Transactions 2012 13:5. Van Pham et al. Geochemical Transactions 2012, 13:5
http://www.geochemicaltransactions.com/content/13/1/5 Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of: Submit your next manuscript to BioMed Central
and take full advantage of: • Convenient online submission
|
https://openalex.org/W2418287953
|
https://www.frontiersin.org/articles/10.3389/fphys.2016.00203/pdf
|
English
| null |
Responses of In vitro-Grown Plantlets (Vitis vinifera) to Grapevine leafroll-Associated Virus-3 and PEG-Induced Drought Stress
|
Frontiers in physiology
| 2,016
|
cc-by
| 12,443
|
Responses of In vitro-Grown
Plantlets (Vitis vinifera) to Grapevine
leafroll-Associated Virus-3 and
PEG-Induced Drought Stress Zhen-Hua Cui 1, 2, Wen-Lu Bi 1, Xin-Yi Hao 1, Yan Xu 1*, Peng-Min Li 1, M. Andrew Walker 2
and Qiao-Chun Wang 1* 1 State Key Laboratory of Crop Stress Biology for Arid Areas, Key Laboratory of Genetic Improvement of Horticultural Crops
of Northwest China, College of Horticulture, Northwest A&F University, Yangling, China, 2 Department of Viticulture and
Enology, University of California, Davis, Davis, CA, USA Stresses caused by viral diseases and drought have long threatened sustainable
production of grapevine. These two stresses frequently occur simultaneously in many of
grapevine growing regions of the world. We studied responses of in vitro-grown plantlets
(Vitis vinifera) to Grapevine leafroll associated virus-3 (GLRaV-3) and PEG-induced
drought stress. Results showed that stress induced by either virus infection or drought
had negative effects on vegetative growth, caused significant decreases and increases
in total soluble protein and free proline, respectively, induced obvious cell membrane
damage and cell death, and markedly increased accumulations of O·−and H
2
2O2. Co-stress by virus and drought had much severer effects than single stress on the said
parameters. Virus infection alone did not cause significant alternations in activities of
POD, ROS, and SOD, and contents of MDA, which, however, markedly increased in the
plantlets when grown under single drought stress and co-stress by the virus and drought. Levels of ABA increased, while those of IAA decreased in the plantlets stressed by virus
infection or drought. Simultaneous stresses by the virus and drought had co-effects on
the levels of ABA and IAA. Up-regulation of expressions of ABA biosynthesis genes
and down-regulation of expressions of IAA biosynthesis genes were responsible for
the alternations of ABA and IAA levels induced by either the virus infection or drought
stress and co-stress by them. Experimental strategies established in the present study
using in vitro system facilitate investigations on ‘pure’ biotic and abiotic stress on plants. The results obtained here provide new insights into adverse effects of stress induced
by virus and drought, in single and particularly their combination, on plants, and allow
us to re-orientate agricultural managements toward sustainable development of the
agriculture. Citation: Citation:
Cui Z-H, Bi W-L, Hao X-Y, Xu Y,
Li P-M, Walker MA and Wang Q-C
(2016) Responses of In vitro-Grown
Plantlets (Vitis vinifera) to Grapevine
leafroll-Associated Virus-3 and
PEG-Induced Drought Stress. Front. Physiol. 7:203. doi: 10.3389/fphys.2016.00203 Edited by:
Olivier Lamotte,
UMR Agroécologie, France g
g ,
Reviewed by:
Walter Chitarra,
National Research Council, Italy
Giorgio Gambino,
National Research Council, Italy
*Correspondence:
Yan Xu
yan.xu@nwsuaf.edu.cn;
Qiao-Chun Wang
qiaochunwang@nwsuaf.edu.cn Reviewed by:
Walter Chitarra,
National Research Council, Italy
Giorgio Gambino,
National Research Council, Italy Specialty section:
This article was submitted to
Plant Physiology,
a section of the journal
Frontiers in Physiology Received: 20 December 2015
Accepted: 19 May 2016
Published: 02 June 2016 Keywords: cell damage, drought, grapevine leafroll virus, physiological metabolism, plant hormones, Vitis vinif ORIGINAL RESEARCH
published: 02 June 2016
doi: 10.3389/fphys.2016.00203 ORIGINAL RESEARCH Keywords: cell damage, drought, grapevine leafroll virus, physiological metabolism, plant hormones, Vitis vinifera Citation:
Cui Z-H, Bi W-L, Hao X-Y, Xu Y,
Li P-M, Walker MA and Wang Q-C
(2016) Responses of In vitro-Grown
Plantlets (Vitis vinifera) to Grapevine
leafroll-Associated Virus-3 and
PEG-Induced Drought Stress.
Front. Physiol. 7:203.
doi: 10.3389/fphys.2016.00203 INTRODUCTION Stresses caused by abiotic and biotic factors have long threatened sustainable development of
agricultural production. Drought, one of the greatest abiotic stresses, was reported to cause
agricultural losses of 50 billion dollars between 1980 and 2012 in United States alone (Suzuki
et al., 2014). It is estimated that more than 6% of the world’s land and 30% of the world’s irrigated June 2016 | Volume 7 | Article 203 Frontiers in Physiology | www.frontiersin.org Stress of Virus and Drought to Grapevine Cui et al. study. In vitro shoots suspected of GLRaV-3 infection were
established in vitro, according to Cui et al. (2015). After 6
months of in vitro establishment, the suspected shoots were
screened again by RT-PCR for their virus status including GVA,
GVB, GRSPaV, GFkV, GFLV, and ArMV, which are among
the grapevine viruses reported in China (Ren et al., 2013). Samples showing a positive response only to GLRaV-3 were
maintained and others discarded. The GLRaV-3 infected shoots
(scions) were micrografted on to the healthy in vitro shoots
(rootstocks). Micrografts developed into plantlets after 3 months
of micrografting. GLRaV-3 was confirmed by RT-PCR in the
micrografted rootstocks (Cui et al., 2015) and shoot segments
were excised from the virus-infected rootstocks were proliferated
to establish the diseased in vitro stock shoots. Thus, the healthy
and diseased in vitro stock shoots were produced from the same
mother plants. Both the healthy and virus-infected the cultures
were maintained on a basal medium (BM) containing half-
strength Murashige and Skoog (1962) medium (MS) with 30 g
l−1 sucrose and 7 g l−1 agar. The pH of the medium was adjusted
to 5.8 prior to autoclaving at 121◦C for 20 min. The cultures were
maintained at a constant temperature of 24 ± 2◦C under a 16-h
photoperiod with a light intensity of 50 µmol s−1 m−2 provided
by cool-white fluorescent tubes. Subculture was done once every
6 weeks. areas also face salinity problems (UNESCO Water Portal, 2007). Viral disease, a biotic stress, causes serious economic losses to
agricultural crops (Hadidi and Barba, 2011). areas also face salinity problems (UNESCO Water Portal, 2007). Viral disease, a biotic stress, causes serious economic losses to
agricultural crops (Hadidi and Barba, 2011). INTRODUCTION The overall goal was to better
understand the effect of abiotic and biotic stress, in single and
particularly in combination, on growth, physiological metabolic
processes and cell damage, thus providing valuable data upon
which agricultural management strategies could be adjusted to
allow more sustainable viticultural production. Grapevine (Vitis vinifera) is an economically important fruit
crop worldwide. Virus disease and drought stress are the two key
factors limiting high yield and quality production of grapevine
(Martelli, 2012). Co-occurrence of drought and virus infection is
common in many of grapevine-growing regions. However, most
of the previous studies on the said issues addressed single stress,
while the co-stress by virus and drought has been far less studied. The objective of the present study was, therefore, to study
responses of grapevines (Vitis vinifera) to Grapevine leafroll-
associated virus-3 (GLRaV-3) and PEG-induced drought stress
using in vitro culture system. The overall goal was to better
understand the effect of abiotic and biotic stress, in single and
particularly in combination, on growth, physiological metabolic
processes and cell damage, thus providing valuable data upon
which agricultural management strategies could be adjusted to
allow more sustainable viticultural production. Analysis of Total Soluble Protein Analysis of Total Soluble Protein
Shoots with leaves of (0.5 g FW) were extracted with 3 ml of
50 mM phosphate buffer (pH 7.0), followed by centrifugation
at 10,000 rpm for 10 min. The supernatant was collected and
absorbance was recorded at 595 nm in a spectrophotometer
(Thermo Multiskan MK3, USA) with bovine serum albumin
(BSA) as a standard, according to Bradford (1976). Vitis vinifera L. ‘Cabernet Sauvignon’, a major red wine cultivar
grown worldwide and susceptible to GLRaV-3, was used in this Abbreviations: ABA, abscisic acid; BM, basal medium; CAT, catalase; DAB,
3, 3′-diaminobenzidine; DW, dry weight; EL, electrolytic leakage; FW, fresh
weight; GLRaV-3, Grapevine leafroll-associated virus-3; IAA, indoleactic acid;
MDA, Methane Dicarboxylic Aldehyde; MS, Murashige and Skoog medium; NBT,
nitroblue tetrazolium; PCD, Programmed cell death; PEG, polyethylene glycol;
POD, peroxidase; ROS, Reactive oxygen species; SOD, superoxide dismutase. Abbreviations: ABA, abscisic acid; BM, basal medium; CAT, catalase; DAB,
3, 3′-diaminobenzidine; DW, dry weight; EL, electrolytic leakage; FW, fresh
weight; GLRaV-3, Grapevine leafroll-associated virus-3; IAA, indoleactic acid;
MDA, Methane Dicarboxylic Aldehyde; MS, Murashige and Skoog medium; NBT,
nitroblue tetrazolium; PCD, Programmed cell death; PEG, polyethylene glycol;
POD, peroxidase; ROS, Reactive oxygen species; SOD, superoxide dismutase. MATERIALS AND METHODS Maintenance of In vitro Healthy and
GLRaV-3 Infected Stock Plantlets and
Establishment of PEG-Induced Drought
Stress Vegetative Growth of In vitro GLRaV-3
Infected Plantlets with and without
Drought Stress Vegetative growth including time required for axillary bud
elongation, shoot length, number of roots, length of the longest
root, and fresh weight (FW) and dry weight (DW) of shoots
and roots were measured after 6 weeks of culture. Axillary
bud elongation was defined when ≥50% of shoots showed bud
elongation. INTRODUCTION g
p
The influence of stress by either virus infection (Sampol
et al., 2003; Moutinho-Pereira et al., 2012; Li et al., 2013;
Cui et al., 2015) or drought (Zhu, 2002; Suzuki et al., 2014)
has been well-demonstrated on plant development, growth,
photosynthetic capability and various physiological metabolisms,
eventually resulting in reduction of crop yield and quality. However, studies on their combining effects have been quite
limited (Atkinson and Urwin, 2012; Prasch, 2013; Suzuki et al.,
2014). Plants grown in nature are exposed simultaneously to
abiotic and biotic stresses. Abiotic stress was shown to alter
the ability of plants resist/tolerate pathogens (Atkinson and
Urwin, 2012; Prasch, 2013). Similarly, biotic stress was found
to alter resistance/tolerance of hosts to abiotic stress (Xu et al.,
2008; Atkinson and Urwin, 2012; Prasch, 2013). Therefore,
responses of plants to single stress differ from those to multiple
stresses (Atkinson and Urwin, 2012; Suzuki et al., 2014), and the
simultaneous occurrence of multiple stresses can cause complex
plant responses (Atkinson and Urwin, 2012; Prasch, 2013; Suzuki
et al., 2014). Thus, better understanding of effects of combining
abiotic with biotic stress is of great significance and would allow
us to orientate agricultural management strategies to ensure
sustainable development of agricultural production. Shoot segments (1.5–2.0 cm in length) with 2 fully-opened
leaves were excised from in vitro 6-week-old healthy and
virus-infected stock plantlets, respectively, and cultured on BM
containing 0, 2, or 4% (w/v) of polyethylene glycol (PEG)
8000, according to Cui et al. (2015), under the same cultural
conditions as used for in vitro stock plantlets. Water potentials
of medium containing 2 and 4% PEG were −2.4 MPa and −2.7
MPa, respectively, as calculated by Michel (1983). Unless stated
otherwise, samples were taken after 4 weeks of culture and used
for the following experiments. p
g
p
Grapevine (Vitis vinifera) is an economically important fruit
crop worldwide. Virus disease and drought stress are the two key
factors limiting high yield and quality production of grapevine
(Martelli, 2012). Co-occurrence of drought and virus infection is
common in many of grapevine-growing regions. However, most
of the previous studies on the said issues addressed single stress,
while the co-stress by virus and drought has been far less studied. The objective of the present study was, therefore, to study
responses of grapevines (Vitis vinifera) to Grapevine leafroll-
associated virus-3 (GLRaV-3) and PEG-induced drought stress
using in vitro culture system. Analysis of Methane Dicarboxylic Aldehyde
(MDA) The MDA content was determined according to Li et al. (2011). Briefly, leaves (1.0 g FW) taken from 2-week-old plantlets
grown at 0 and 4% PEG were homogenized in 10 ml of 10%
trichloroacetic acid, followed by centrifugation at 10000 rpmg
for 10 min. After then, 2 ml of 0.6% thiobarbituric acid in 10%
trichloroacetic acid were added to 2 ml of the supernatant. The
mixture was heated in boiling water for 15 min, and then quickly
cooled in an ice bath. After centrifugation at 10,000 rpm for 10
min, the absorbance of the supernatant was recorded at 450 nm,
532 nm and 600 nm in a spectrophotometer (Thermo Multiskan
MK3, USA). The MDA concentration was calculated using the
following formula: 6.45(OD 532−OD 600)−0.56OD450. Analysis of Antioxidant Enzyme Activities Analysis of Antioxidant Enzyme Activities
The activities of superoxide dismutase (SOD), peroxidase (POD),
catalase (CAT) were measured, as described by Li et al. (2011). Fresh leaves (0.5 g FW) were harvested from 2-week-old plantlets
grown at 0 and 4% PEG, ground in liquid nitrogen and extracted
with following extraction media: 100 mM potassium phosphate
buffer (pH 7.8) containing 0.1 mM EDTA, 1% (w/v) PVP and
0.1% (v/v) Triton x100. The extracts were centrifuged at 10,000
rpm for 15 min at 4◦C. The supernatants were collected and used
for the enzyme activity assays. Analysis of Free Proline Free proline content was determined according to the method
of Bates et al. (1973). Shoots with leaves (0.5 g FW) were
homogenized with 10 ml of 3% (w/v) sulfosalicylic acid. The June 2016 | Volume 7 | Article 203 Frontiers in Physiology | www.frontiersin.org 2 Stress of Virus and Drought to Grapevine Cui et al. the decomposition of H2O2 at 240 nm in a spectrophotometer
(Thermo Multiskan MK3, USA). The reaction mixture (3 ml)
contained 0.05 M Na phosphate buffer (pH 7.0) supplemented
with 1 mM EDTA, H2O2 (3%) and 100 µl of enzyme extract. homogenate was centrifuged at 3000 rpm for 20 min. The
supernatant was treated with acid ninhydrin in boiling water
for 1 h. The reaction was terminated in a water bath at a room
temperature for 10 min. The reaction mixture was extracted
with 4 ml of toluene and vortexed for 15 s. The absorbance
was determined at 520 nm in a spectrophotometer (Thermo
Multiskan MK3, USA) using L-proline as a standard. Analysis of Cell Membrane Damage (CMD)
and Cell Death Roots (0.5 g FW) were used for analysis of cell membrane damage
(CMD) by measuring relative electrolytic leakage (EL), according
to the method of Sullivan (1972). Cell death was detected by
Evan’s blue staining method (Gaffand Okong’O-Ogola, 1971). Samples were immersed in 0.1% Evan’s blue solution for 30 min,
and then washed with distilled water for 3 times to stop the
color reaction. Staining reaction was photographed using a digital
camera (PowerShot G9, Canon, Japan). Frontiers in Physiology | www.frontiersin.org Analysis of Endogenous Hormones 2
Fresh leaves were used for O·−
2
and H2O2 localization in situ,
according to the method of Romero-Puertas et al. (2004), with
some modifications. For O·−
2 localization, leaves were immersed
in 0.1% solution of nitroblue tetrazolium (NBT) in 10 mM K-
phosphate buffer (pH 7.6), vacuum-infiltrated for 30 min and
illuminated for 2 h, followed by bleaching in 95% boiling ethanol
for 5 min. For H2O2 localization, leaves were immersed in a 0.1%
filtered solution of 3, 3′-diaminobenzidine (DAB) in 10 mM MES
buffer (pH 5), vacuum-infiltrated for 30 min and then incubated
at room temperature for 4 h in the dark, followed by bleached in
95% boiling ethanol for 5 min. Photographs were captured by a
digital camera (PowerShot G9, Canon, Japan). Analysis of Endogenous Hormones
Contents of endogenous ABA and IAA were analyzed in the in
vitro plantlets grown at 0 and 4% PEG. Shoots with leaves were
taken from 1-week-old plantlets and divided into two groups. One group was used for analysis of ABA and IAA, and the
other for measurement of transcription level of ABA and IAA
biosynthetic genes, as described below. Samples were frozen in
liquid nitrogen and stored at −80◦C until usage. Abscisic acid
(ABA) and indoleactic acid (IAA) were provided by Sigma (St. Louis, MO, USA). At the beginning of extraction, 1000 Bq of each
ABA and IAA, accordingly, were added to monitor the losses
during purification. The extraction and purification of ABA and
IAA were conducted, according to by Dobrev et al. (2005). The
purified extract solution was transferred into 2-ml centrifuge
tubes containing 0.3 g polyvinylpolypyrolidone (PVPP) and then
kept at −20◦C for the measurements of ABA and IAA in water
breeze HPLC system (Waters 2489 UV/Visible Detector), using
wavelength at 254 nm, velocity at 0.7 ml min−1 and sample
quantity of 10 µm and column temperature at 30◦C. The dried
fraction containing ABA and IAA were, respectively, injected
into the HPLC system for measurements, as described by Li et al. (2013). Analysis of Transcription Level of ABA and
IAA Biosynthetic Genes The SOD activity was determined by measuring its ability to
inhibit photochemical reduction of nitrobluetetrazolium (NBT). The reaction mixture (3 ml) contained 0.3 ml each of 20 µM
riboflavin, 150 mM l-methionine, 600 µM NBT and 0.1 ml
extract, and performed under irradiance of 170 µmol photons
m−2 s−1 provided by white fluorescent lamps. The absorbance
was determined at 560 nm. The extract volume causing 50%
inhibition of NBT reduction was taken as one unit of activity. The POD activity was measured in the reaction mixture (3 ml)
containing 50 mM phosphate buffer (pH 7.0), 0.2 mM guaiacol,
10 mM H2O2. The reaction was initiated by adding 200 µl of
enzyme extract to the reaction mixture. The oxidation of guaiacol
was measured upon an increase in absorbance at 470 nm for
1 min. The CAT activity was determined by directly measuring The transcription levels of several major genes responsible for
biosynthesis of ABA and IAA were analyzed in the in vitro
plantlets grown at 0 and 4% PEG at different time durations. RNA was isolated from shoots with leaves (1 g FW), as described
by Davies and Robinson (1996). RNA quality was evaluated
by optical density (OD) value (1.9–2.1) at 260/280 nm. After
removal of DNA using DNA Eraser (Takara, Japan), cDNA
was synthesized using the reagent kits (RR047A, Takara, Japan)
according to the manufacture’s instructions. All the candidate
primers and reference genes of ABA and IAA biosynthesis
pathway were selected according to the existing studies (Table 1). The reagent kits (RR047A, Takara) were used for qPCR reaction June 2016 | Volume 7 | Article 203 Frontiers in Physiology | www.frontiersin.org 3 Stress of Virus and Drought to Grapevine Cui et al. TABLE 1 | Primers used for qRT-PCR analysis of ABA and IAA in in vitro plantlets of grape ‘Cabernet Sauvignon’. Analysis of Transcription Level of ABA and
IAA Biosynthetic Genes Primer
References
Sequence 5′-3′(forward/reverse)
PN40024 12X V1 ID*
Coordinates*
names
VvNCED1
Wheeler et al., 2009
GAGACCCCAACTCTGGCAGG/ AAGGTGCCGTGGAATCCATAG
VIT_19s0093g00550
chr19:17645348..17647649
VvNCED2
Wheeler et al., 2009
AGTTCCATACGGGTTTCATGGG/ CCATTTTCCAAATCCAGGGTGT
VIT_10s0003g03750
chr10:6374432..6376728
VvZEP
Wheeler et al., 2009
TACCGGGTATTTTTGGGACA/ CTTCTTCATCCGTGGCAAGT
VIT_07s0031g00620
chr7:16795707..16804559
VvTAR2
Böttcher et al., 2013
CAGCAATGAAGCATATTGAAGG/ GAGTGAGAGCACCAGGAAATG
VIT_17s0000g08990
chr17:10559797..10565113
VvTAR3
Böttcher et al., 2013
CCCAAGATGACT TTGATATGCTG/ TGATCAACTGATTGTTGATTCCACT
VIT_18s0157g00090
chr18:18858918..18864125
VvTAR4
Böttcher et al., 2013
CAGCCTCATCAAGACCCAAGAT/ TGACGGTTGATTTCATTCTTCG
VIT_18s0157g00170
chr18:18980507..18985732
VvYUC1
Böttcher et al., 2013
CAGGAAACTGTCGCAATAGTGG/ CAAGAACTATGTTGGGTATTGAGAGG
VIT_07s0104g01250
chr7:2276803..2279543
VvACTIN2
Böttcher et al., 2011
GCACCCTTCG CACGATATGA/ TGACGCAAGGCAAGGACTGA
VIT_04s0044g00580
chr4:21427077..21431176
*Means V1 annotations obtained from CRIBI (http://genomes.cribi.unipd.it/grape/) by a BLAST search of the gene sequence provided by NCBI (http://www.ncbi.nlm.nih.gov). PN40024 12X V1 ID*
Coordinates* differences were found in this parameter in the healthy shoots
grown at 0% and 2% PEG, but PEG at 4% considerably delayed
the time duration for axillary bud elongation (5.8 days). A similar
pattern in axillary bud elongation was found in the diseased
shoots grown at 2–4% PEG, but the negative effect was much
stronger (Figure 1A). Shoot length was similar between the
healthy and diseased shoots without PEG stress (Figure 1B). In the healthy shoots, shoot length markedly decreased as
PEG concentrations increased from 0 to 4% (Figure 1B). This
inhibitory effect exerted by PEG stress was much stronger on
the infected shoots than on the healthy shoots (Figure 1B). When grown at 0% PEG, the healthy and virus infected shoots
produced a similar number of roots (Figure 1C). Number of
roots significantly decreased in the healthy shoots when grown at
2–4% PEG (Figure 1C). Negative effects of PEG concentrations
on the number of roots were much pronounced in the infected
shoots than in the healthy ones (Figure 1C). The healthy shoots
without PEG stress produced greater length of the longest roots
than the virus infected ones (Figure 1D). Length of the longest
roots was similar in the healthy shoots grown at 0 and 2%
PEG (Figure 1D), but 4% PEG resulted in much shorter length
of the longest roots (Figure 1D). Co-stress by the virus and
4% PEG resulted in the shortest length of roots (Figure 1D). Without PEG stress, fresh weight (FW) of shoots was significantly
greater in the healthy shoots than that of the virus infected ones
(Figure 1E), and markedly reduced in the healthy shoots as PEG
concentrations increased from 0 to 4%. xperimental Design and Data Analysis p
g
y
Since some of the leaves at the basal part of the diseased shoots
showed GLRaV-3 symptoms when grown under PEG-induced
stress (Cui et al., 2015), symptomless leaves were used in all
experiments. The experiments determining effects of virus and
drought stress on vegetative growth of in vitro plantlets were
designed as a completely randomized design. Ten samples were
included in each of three replicates and the whole experiment
was repeated at least twice. In all experiments measuring contents
of total soluble protein, proline and MDA, activities of SOD,
POD and CAT, endogenous hormones and transcript levels of
ABA and IAA biosynthetic genes, each treatment contained
five biological replicates and repeated three times in each
experiment. Statistical analysis of the data was performed with
the SPSS-19 for Windows statistics software package. Significant
differences among means were calculated by the least significant
difference (LSD) at P ≤0.05. Two-way ANOVA, including
GLRaV-3 infection and drought as factors, was performed to
analyze the combined effects of these two factors on some
selected parameters. Two-way MANOVA was performed for
the vegetative growth. Significant differences were analyzed at
P ≤0.05 and P ≤0.01, respectively. Analysis of Transcription Level of ABA and
IAA Biosynthetic Genes When grown at 0% PEG, the healthy and virus infected shoots
produced a similar number of roots (Figure 1C). Number of
roots significantly decreased in the healthy shoots when grown at
2–4% PEG (Figure 1C). Negative effects of PEG concentrations
on the number of roots were much pronounced in the infected
shoots than in the healthy ones (Figure 1C). The healthy shoots
without PEG stress produced greater length of the longest roots
than the virus infected ones (Figure 1D). Length of the longest
roots was similar in the healthy shoots grown at 0 and 2%
PEG (Figure 1D), but 4% PEG resulted in much shorter length
of the longest roots (Figure 1D). Co-stress by the virus and
4% PEG resulted in the shortest length of roots (Figure 1D). Without PEG stress, fresh weight (FW) of shoots was significantly
greater in the healthy shoots than that of the virus infected ones
(Figure 1E), and markedly reduced in the healthy shoots as PEG
concentrations increased from 0 to 4%. Influences exerted by co-
stress of the virus and drought on FW of roots were similar to
those of FW of shoots (Figure 1F). Without PEG, the dry weight
(DW) of the healthy shoots was greater than that of the virus
infected shoots (Figure 1G), and it decreased when grown at 2–
4% PEG. The DW of the infected shoots also decreased as PEG
concentrations increased from 2 to 4%. There was no significant
difference in DW of roots between the healthy and diseased
shoots without PEG stress (Figure 1H). The DW of roots in
both the healthy and infected shoots considerably reduced as
PEG concentrations elevated from 0 to 2–4% (Figure 1H). Two-
way MANOVA of all data on vegetative growth showed that
single GLRaV-3 infection and PEG-induced drought, singly,
and GLRaV-3 in combination with PEG all had significantly
(P ≤0.01) negative effects on vegetative growth (Table 2). Analysis of Transcription Level of ABA and
IAA Biosynthetic Genes Influences exerted by co-
stress of the virus and drought on FW of roots were similar to
those of FW of shoots (Figure 1F). Without PEG, the dry weight
(DW) of the healthy shoots was greater than that of the virus
infected shoots (Figure 1G), and it decreased when grown at 2–
4% PEG. The DW of the infected shoots also decreased as PEG
concentrations increased from 2 to 4%. There was no significant
difference in DW of roots between the healthy and diseased
shoots without PEG stress (Figure 1H). The DW of roots in
both the healthy and infected shoots considerably reduced as
PEG concentrations elevated from 0 to 2–4% (Figure 1H). Two-
way MANOVA of all data on vegetative growth showed that
single GLRaV-3 infection and PEG-induced drought, singly,
and GLRaV-3 in combination with PEG all had significantly
(P ≤0.01) negative effects on vegetative growth (Table 2). (IQ5, BIO-RAD, USA) in a 25 µl reaction mixture containing
12.5 µl of 2 × SYBR Green I Master Mix, 1 µl of each primer
(0.4 µM), 2 µl of cDNA (100 ng) and 8.5 µl of RNase-free water
according to the kits instructions. The following program was
used: an initial denaturation step at 94◦C for 30 s, 45 cycles at
94◦C for 10 s, 56◦C for 10 s and 72◦C for 30 s. A melting curve
analysis was carried out over the range 65–97◦C to verify the
specificity of amplicons. Two controls (no-RT and no-template)
were included in the designs. Transcript levels of each gene were
normalized according to the reference gene, using the 2-(-Delta
Delta C (T) method (Livak and Schmittgen, 2001). differences were found in this parameter in the healthy shoots
grown at 0% and 2% PEG, but PEG at 4% considerably delayed
the time duration for axillary bud elongation (5.8 days). A similar
pattern in axillary bud elongation was found in the diseased
shoots grown at 2–4% PEG, but the negative effect was much
stronger (Figure 1A). Shoot length was similar between the
healthy and diseased shoots without PEG stress (Figure 1B). In the healthy shoots, shoot length markedly decreased as
PEG concentrations increased from 0 to 4% (Figure 1B). This
inhibitory effect exerted by PEG stress was much stronger on
the infected shoots than on the healthy shoots (Figure 1B). RESULTS Vegetative Growth and Root Formation
Without PEG stress, time duration required for axillary bud
elongation was much shorter in the healthy shoots (4.5 days) than
in GLRaV-3 infected ones (6.1 days; Figure 1A). No significant June 2016 | Volume 7 | Article 203 Frontiers in Physiology | www.frontiersin.org 4 Stress of Virus and Drought to Grapevine Cui et al. Cui et al. FIGURE 1 | Effects of GLRaV-3 infection and drought stress on shoot growth and root formation of in vitro plantlets of grape ‘Cabernet Sauvignon’. (A) Time required for axillary bud elongation. (B) Shoot length. (C) Number of roots. (D) Length of the longest root. (E) Fresh weight of the shoots. (F) Fresh weight of
the roots. (G) Dry weight of the shoots. (H) Dry weight of the roots. Data were presented as means ± SE and with different letters within the same parameter are
significantly different at P ≤0.05. FIGURE 1 | Effects of GLRaV-3 infection and drought stress on shoot growth and root formation of in vitro plantlets of grape ‘Cabernet Sauvignon’. (A) Time required for axillary bud elongation. (B) Shoot length. (C) Number of roots. (D) Length of the longest root. (E) Fresh weight of the shoots. (F) Fresh weight of
the roots. (G) Dry weight of the shoots. (H) Dry weight of the roots. Data were presented as means ± SE and with different letters within the same parameter are
significantly different at P ≤0.05. Cell Membrane Damage and Cell Death Data are
presented as means ± SE and with different letters are significantly different at
P ≤0.05. FIGURE 2 | Effects of GLRaV-3 infection and drought stress on
contents of total soluble protein (A), proline (B), and relative electrolyte
leakage (C) of in vitro plantlets of grape ‘Cabernet Sauvignon’. Data are
presented as means ± SE and with different letters are significantly different at
P ≤0.05. FIGURE 2 | Effects of GLRaV-3 infection and drought stress on
contents of total soluble protein (A), proline (B), and relative electrolyte
leakage (C) of in vitro plantlets of grape ‘Cabernet Sauvignon’. Data are
presented as means ± SE and with different letters are significantly different at
P ≤0.05. Cell Membrane Damage and Cell Death Relative electrolytic leakage (EL) was similar in the healthy
and diseased shoots without any stress (Figure 2C). Stress
by PEG concentrations at 2 and 4% produced increased EL
comparing to 0% PEG in the healthy shoots. The similar pattern
of EL was found in the diseased shoots grown under 2–4%
PEG concentrations, but the negative effect was much more
obvious. Blue color forms when dead cells are stained by Evan’s blue
solution. Without PEG stress, almost no blue color was seen
in the healthy roots (Figure 3A), but some light blue color was
observed in GLRaV-3 infected roots (Figure 3B). Density and
areas of blue color formation increased with increasing PEG
concentrations from 2 to 4% in both the healthy (Figures 3C,E)
and virus-infected roots (Figures 3D,F). The virus-infected roots
had a much stronger blue color and larger areas than the healthy
roots that were treated by the same concentrations of PEG. When
grown at 4% PEG, almost the whole roots infected by GLRaV-3
became blue (Figure 3F). O·−
2 and H2O2 Localization In situ
Production of O·−
2
in the leaves is visualized with NBT
staining to give rise to dark blue spots. Without PEG
stress, O·−
2
accumulation was hardly detected in the healthy
leaves (Figure 4A), but it was easily seen in the diseased
leaves (Figure 4B). In the healthy leaves, O·−
2
accumulation
considerably increased with an increase in PEG concentrations
from 2% (Figure 4C) to 4% (Figure 4E). A similar pattern was
found in the diseased leaves, but staining density and stained
areas were much stronger (Figure 4D) and larger (Figure 4F)
than in the corresponding healthy samples. Production of H2O2
in the leaves is visualized with DAB staining to give rise to
brown spots. Without PEG, H2O2 accumulation in the virus-
infected leaves (Figure 5B) was stronger than that in the healthy
leaves (Figure 5A). Although, PEG stress at 2–4% markedly
increased H2O2 accumulation in both the healthy and infected
leaves, staining density and strained areas were much stronger
and larger in the virus-infected leaves than in the healthy leaves
that were stressed by the corresponding PEG concentrations
(Figures 5C–F). FIGURE 2 | Effects of GLRaV-3 infection and drought stress on
contents of total soluble protein (A), proline (B), and relative electrolyte
leakage (C) of in vitro plantlets of grape ‘Cabernet Sauvignon’. Content of Total Soluble Protein and Free
Proline protein (Figure 2A). Content of free proline was similar in the
healthy and diseased shoots without PEG stress (Figure 2B). In the healthy shoots, PEG at 2–4% significantly increased
accumulations of free proline (Figure 2B). Levels of free proline
considerably increased in the virus-infected shoots as PEG
concentrations increased from 0 to 4% (Figure 2B). Two-way
ANOVA showed that virus infection and PEG stress, and virus
infection in combination with PEG stress had significant negative Proline
A significant difference was found in content of total soluble
protein between the healthy and virus-infected shoots without
PEG stress (Figure 2A). In the healthy shoots, marked reductions
were found in content of total soluble protein only when grown
at 4% PEG. In the infected shoots, PEG concentrations at
2–4% caused considerable decreases in contents of total soluble June 2016 | Volume 7 | Article 203 Frontiers in Physiology | www.frontiersin.org 5 Stress of Virus and Drought to Grapevine Cui et al. FIGURE 2 | Effects of GLRaV-3 infection and drought stress on
contents of total soluble protein (A), proline (B), and relative electrolyte
leakage (C) of in vitro plantlets of grape ‘Cabernet Sauvignon’. Data are
presented as means ± SE and with different letters are significantly different at
P ≤0.05. Activities of Antioxidant Enzymes and
C
t
t
f MDA TABLE 2 | Two-way MANOVA of effects of GLRaV-3 and PEG-induced
drought, and their combing effects on all variables of vegetative growth of
in vitro plantlets of grape ‘Cabernet Sauvignon’. Parameters
Wilk’s Lambda
F
p
Virus
0.031
66.690
**
Drought
0.008
22.241
**
Virus-drought
0.003
37.855
**
**Significant difference at P ≤0.01. TABLE 2 | Two-way MANOVA of effects of GLRaV-3 and PEG-induced
drought, and their combing effects on all variables of vegetative growth of
in vitro plantlets of grape ‘Cabernet Sauvignon’. | Two-way MANOVA of effects of GLRaV-3 and PEG-induce effects on contents of both total soluble protein and free proline
(Table 3). Activities of Antioxidant Enzymes and
Content of MDA Without PEG stress, no significant differences were found in the
activities of SOD, POD, and CAT between the virus-infected and
healthy plantlets (Figures 6A–C). For the healthy plantlets, 4%
PEG resulted in significant increases in the activities of SOD and
POD, although not in CAT (Figures 6A–C). For the diseased
plantlets, the activities of SOD, POD, and CAT significantly
increased when stressed by PEG 4%. No differences were found June 2016 | Volume 7 | Article 203 Frontiers in Physiology | www.frontiersin.org 6 Stress of Virus and Drought to Grapevine Cui et al. TABLE 3 | Two-way ANOVA of effects of GLRaV-3 and drought, and their
combining effects on indexes of in vitro plantlets shoots of grape
‘Cabernet Sauvignon’. Contents of Influenced indexes
Virus
Drought
Virus-Drought
Total soluble protein
**
**
*
Proline
**
**
**
SOD
**
**
*
POD
**
**
**
CAT
**
**
**
MDA
**
**
**
ABA
**
**
**
IAA
**
**
ns
**Significant difference with levels of P ≤0.01; *Significant difference with levels of P ≤
0.05; ns = no significant difference. TABLE 3 | Two-way ANOVA of effects of GLRaV-3 and drought, and their
combining effects on indexes of in vitro plantlets shoots of grape
‘Cabernet Sauvignon’. FIGURE 3 | Histochemical detection of cell death in roots of in vitro
plantlets of grape ‘Cabernet Sauvignon’. Cell death in the roots was
revealed by the blue precipitates produced by staining with Evan’s blue. Roots
of the healthy shoots cultured at 0% (A), 2% (C), and 4% PEG (E). Roots of
GLRaV-3-infected shoots cultured at 0% (B), 2% (D), and 4% PEG (F), Bar
scale = 1 cm. in MDA contents between the healthy and diseased plantlets
without PEG stress (Figure 6D). MDA contents were found
much higher both in the healthy and the diseased plantlets grown
at 4% PEG, (Figure 6D). Two-way ANOVA showed that single
stress by either the virus or PEG significantly (P ≤0.01) changed
the activities of SOD, POD, and CAT, and increased MDA content
(Table 3). Co-stress by the virus and PEG had significantly (P ≤
0.01) negative effects on those of POD, CAT and MDA, as well as
on SOD (P ≤0.05) (Table 3). Activities of Antioxidant Enzymes and
Content of MDA Contents of ABA and IAA, and Expression
Levels of ABA and IAA Biosynthetic Genes
Without PEG stress, virus infection caused significant (P ≤
0.05) increase and decrease in the content of ABA and IAA
(Figures 7A,B), respectively. For the healthy plantlets, the
content of ABA was significantly higher, while that of IAA was
much lower, in the plantlets grown at 4% PEG than at 0% PEG
(Figures 7A,B). Similar patterns of ABA and IAA levels were
found in the virus-infected plantlets grown at 0 and 4% PEG
(Figures 7A,B). Two-way ANOVA showed that PEG stress or
the virus infection and their co-stress all significantly (P ≤0.01)
increased ABA content; the virus infection or drought stress
significantly (P ≤0.01) decreased IAA level, but there was no
interaction between them on IAA level (Table 3). FIGURE 3 | Histochemical detection of cell death in roots of in vitro
plantlets of grape ‘Cabernet Sauvignon’. Cell death in the roots was
revealed by the blue precipitates produced by staining with Evan’s blue. Roots
of the healthy shoots cultured at 0% (A), 2% (C), and 4% PEG (E). Roots of
GLRaV-3-infected shoots cultured at 0% (B), 2% (D), and 4% PEG (F), Bar
scale = 1 cm. (Figure 8B), for ZEP at 96 h in the two types of plantlets
(Figure 8C). For IAA biosynthesis genes, without PEG stress,
higher expression levels of TAR2, TAR3, TAR4, and YUC1 were
detected in the healthy than in the diseased plantlets during
the whole culture time durations (Figures 9A–D). Expression
levels of TAR2, TAR3, and TAR4 started to increase at 24 h,
peaked at 96 h of culture and then decreased, while those of
YUC1 peaked at 24 h and gradually decreased during 24 h to 1
weeks of culture (Figure 9D). PEG stress consistently reduced
the expression levels of all the four genes in both the healthy
and diseased plantlets during the whole culture durations, with
more obviously reduced levels found in the diseased plantlets
(Figures 9A–D). For ABA biosynthesis genes, without PEG stress, expression
levels of NCED1 were relatively stable in the healthy plantlets
during 1 week of culture, and peaked at 96 h of culture and then
reduced in the virus infected plantlets (Figure 8A). Patterns of
expression levels of NCED2 and ZEP were similar to those of
NCED1 in the healthy plantlets (Figures 8B,C). Activities of Antioxidant Enzymes and
Content of MDA In the diseased
plantlets, expression levels of NCED2 started to increase at 48
h of culture and the increased levels maintained up to 1 week
of culture (Figure 8B). Expression levels of ZEP were hardly
detected until 48 h of culture, and relatively increased during
96 h and 1 week of culture (Figure 8C). PEG stress markedly
induced expression levels of NCED1, NCED2, and ZE in both
the healthy and diseased plantlets (Figures 8A–C). The highest
expression levels appeared for NCED1 between 48 to 96 h in
both the healthy and diseased plantlets (Figure 8A), for NCED2
at 48 and 96 h in the infected and healthy plantlets, respectively Frontiers in Physiology | www.frontiersin.org DISCUSSION FIGURE 5 | Histochemical detection of H2O2 in leaves of in vitro
plantlets of grape ‘Cabernet Sauvignon’. H2O2 deposits were revealed by
the brown precipitates produced by staining with DAB. Leaves of the healthy
shoots cultured at 0% (A), 2% (C), and 4% PEG (E). Leaves of
LRaV-3-infected shoots cultured at 0% (B), 2% (D), and 4% PEG (F). Bar
scale = 1 cm. FIGURE 4 | Histochemical detection of O·−
2
in leaves of in vitro
plantlets of grape ‘Cabernet Sauvignon’. Deposits of O·−
2 were revealed
by the blue formazan precipitates produced by staining with NBT. Leaves of
the healthy in vitro shoots cultured at 0% (A), 2% (C), and 4% PEG (E). Leaves of GLRaV-3-infected shoots cultured at 0% (B), 2% (D), and 4% PEG
(F). Bar scale = 1 cm. for studies on the said subjects (Watanabe et al., 2000; Khristov
et al., 2001; Christov et al., 2007; Lu et al., 2007; Lokhande et al.,
2010, 2011; Faraloni et al., 2011; Li et al., 2013; Markovi´c et al.,
2014; Cui et al., 2015), mainly due to its advantages that can avoid
other variables such as light, temperature and nutrients, and by
attacks of other pathogens facing in the field, in addition to its
effects of efficient, rapid and low cost, thus facilitating studies
of ‘pure’ biotic or abiotic stress (Watanabe et al., 2000; Khristov
et al., 2001; Christov et al., 2007; Lokhande et al., 2010, 2011; Li
et al., 2013; Cui et al., 2015). A recent study of Faraloni et al. (2011) found a high accordance between the in vitro and in vivo
chlorophyll fluorescence measurements in monitoring changes in
photosynthetic activity in response to drought stress. ones. Similar results have been reported in in vitro grapevine
shoots infected by GLRaV-3 (Cui et al., 2015), Grapevine
leafroll (Tanne et al., 1996), and Grapevine fanleaf virus (GFLV)
(Abracheva et al., 1994). A number of examples of negative
effects of virus infection on plant growth can be found in the
publications of Watanabe et al. (2000), Li et al. (2013), and Cui
et al. (2015). Reduced vegetative growth and slow rooting were
observed in the in vitro grapevine shoots stressed by 2–6% PEG
(Dami and Hughes, 1997) and by 4% sorbitol or 2–4% mannitol
(Tanne et al., 1996). DISCUSSION Although, in vitro culture conditions largely differ from the field
environments and responses of in vitro cultures to abiotic or
biotic stress may differ from those of the field-grown plants in
some cases (Suzuki et al., 2014), the former has been widely used June 2016 | Volume 7 | Article 203 Frontiers in Physiology | www.frontiersin.org 7 Cui et al. Stress of Virus and Drought to Grapevine FIGURE 4 | Histochemical detection of O·−
2
in leaves of in vitro
plantlets of grape ‘Cabernet Sauvignon’. Deposits of O·−
2 were revealed
by the blue formazan precipitates produced by staining with NBT. Leaves of
the healthy in vitro shoots cultured at 0% (A), 2% (C), and 4% PEG (E). Leaves of GLRaV-3-infected shoots cultured at 0% (B), 2% (D), and 4% PEG
(F). Bar scale = 1 cm. FIGURE 5 | Histochemical detection of H2O2 in leaves of in vitro
plantlets of grape ‘Cabernet Sauvignon’. H2O2 deposits were revealed by
the brown precipitates produced by staining with DAB. Leaves of the healthy
shoots cultured at 0% (A), 2% (C), and 4% PEG (E). Leaves of
LRaV-3-infected shoots cultured at 0% (B), 2% (D), and 4% PEG (F). Bar
scale = 1 cm. FIGURE 5 | Histochemical detection of H2O2 in leaves of in vitro
plantlets of grape ‘Cabernet Sauvignon’. H2O2 deposits were revealed by
the brown precipitates produced by staining with DAB. Leaves of the healthy
shoots cultured at 0% (A), 2% (C), and 4% PEG (E). Leaves of
LRaV-3-infected shoots cultured at 0% (B), 2% (D), and 4% PEG (F). Bar
scale = 1 cm. FIGURE 4 | Histochemical detection of O·−
2
in leaves of in vitro
plantlets of grape ‘Cabernet Sauvignon’. Deposits of O·−
2 were revealed
by the blue formazan precipitates produced by staining with NBT. Leaves of
the healthy in vitro shoots cultured at 0% (A), 2% (C), and 4% PEG (E). Leaves of GLRaV-3-infected shoots cultured at 0% (B), 2% (D), and 4% PEG
(F). Bar scale = 1 cm. FIGURE 5 | Histochemical detection of H2O2 in leaves of in vitro
plantlets of grape ‘Cabernet Sauvignon’. H2O2 deposits were revealed by
the brown precipitates produced by staining with DAB. Leaves of the healthy
shoots cultured at 0% (A), 2% (C), and 4% PEG (E). Leaves of
LRaV-3-infected shoots cultured at 0% (B), 2% (D), and 4% PEG (F). Bar
scale = 1 cm. Frontiers in Physiology | www.frontiersin.org DISCUSSION Inhibitory effects exerted by drought or
osmotic stress on vegetative growth were much stronger in
the grapevine leafroll infected shoots than in the healthy ones
(Chaves et al., 2007; Cui et al., 2015). All of these results were
consistent with ours. Effects of single stress by either virus or drought on plant
growth have been well-documented, but co-stress by pathogens
like virus and drought to plants has been far less studied. Working on the several grapevine cultivars, Markovi´c et al. (2014) found that the healthy in vitro-grown shoots had much
better shoot growth than the GLRaV-3 or other virus infected Virus infection or drought stress has been found to influence
levels of soluble protein. An increased level of soluble protein was
noted in grapevine infected with GFLV and GLRaV-3 (Sampol
et al., 2003) or GLRV (Bertamini et al., 2004; Moutinho-Pereira
et al., 2012), in banana infected with BBTV (Haq et al., 2012) June 2016 | Volume 7 | Article 203 8 Stress of Virus and Drought to Grapevine Cui et al. Cui et al. FIGURE 6 | Activities of superoxide dismutase (SOD, A), peroxidase (POD, B), catalase (CAT, C), and content of methane dicarboxylic aldehyde (MDA,
D) of in vitro plantlets of grape ‘Cabernet Sauvignon’. Data were presented as means ± SE and with different letters within the same parameter are significantly
different at P ≤0.05. FIGURE 6 | Activities of superoxide dismutase (SOD, A), peroxidase (POD, B), catalase (CAT, C), and content of methane dicarboxylic aldehyde (MDA,
D) of in vitro plantlets of grape ‘Cabernet Sauvignon’. Data were presented as means ± SE and with different letters within the same parameter are significantly
different at P ≤0.05. but drought, singly or in combination with the virus, markedly
increased cell membrane damage and induced cell death,
particularly when GLRaV-3 infected shoots were stressed by 4%
PEG. Programmed cell death (PCD) has been widely observed
in response of the plants to pathogenic infection (Greenberg and
Yao, 2004). The HR in response to virus infection is a typical
example of PCD (Greenberg and Yao, 2004). Increased cell
membrane damage and cell death have been well established in
the plants infected with various viruses such as Tobacco mosaic
virus (TMV)-infected Nicotiana tabacum (del Pozo and Lam,
2003), Groundnut bud necrosis virus (GBNV)-infected Vigna
unguiculata (Permar et al., 2014) and Soybean mosaic virus
(SMV)-infected Glycine max (Zhang et al., 2014). DISCUSSION The HR is
typically induced upon virus infection as in interaction between
host encoded resistance (R) proteins and pathogen encoded
avirulence proteins and functions to limit virus replication and
thereby its eventual movement inside the host (Padder, 2014). Increased cell membrane damage and induced cell death by
drought stress has also been found in other plants including
Saccharum officinarum (Patade et al., 2008), Arabidopsis (Duan
et al., 2010), and Cucumis sativus (Zhang et al., 2013). Drought
stress-induced root cell death was suggested to be an adaptive
response to the stress (Duan et al., 2010). Expression of BAX
inhibitor-1 (AtBI1) increased in response to water stress in
roots of Arabidopsis, indicating that AtBI1 and the endoplasmic
reticulum (ER) response pathway reduced water stress-induced
PCD (Duan et al., 2010). and in potato infected with PLRV or PVY (Li et al., 2013). Similar results were also found in the present study. The decrease
in synthesis of ribulose-1,5-bisphosphate (RuBP) carboxylase
was found responsible for the marked reduction in the content
of soluble proteins in the GLRV-infected samples (Bertamini
et al., 2004). Decreased contents of total protein were observed
in drought-stressed grapevine (Maroco et al., 2002) and other
plants such as Zea mays (Hsiao, 1970), and Gossypium hirsutum
(Parida et al., 2007). These data were again confirmed in our
study. Drought stress caused marked changes in the protein
synthesizing apparatus of plant tissue, thus reducing capacity for
protein synthesis (Hsiao, 1970; Parida et al., 2007). p
y
Our study found that drought stress and GLRaV-3 infection,
singly or in combination, significantly increased proline contents. Increased proline contents by drought have been found in other
plant species such as Populus euphratic (Watanabe et al., 2000),
Eucalyptus camaldulensis (Woodward and Bennett, 2005) and
Capsicum annuum (Fu et al., 2009), or by virus such as Zucchini
yellow mosaic virus (ZYMV, Radwan et al., 2007). Proline
accumulation generally is considered as a drought tolerance
mechanism in plants (Chutia and Borah, 2012; Sun et al., 2013). Increased accumulation of proline was believed to be as an
adaptive response to the stress, thus helping to maintain cell
membrane integrity and protecting subcellular structures in
drought-stressed plants (Chutia and Borah, 2012; Sun et al.,
2013), and to activate a hypersensitive response (HR) to virus
infection (Radwan et al., 2007). Frontiers in Physiology | www.frontiersin.org DISCUSSION Our
results
showed
that
GLRaV-3
infection
did
not
significantly cause cell membrane damage and cell death, MDA, a marker for lipid peroxidation, is frequently used as an
indicator for measurement of cellular membrane damage (Masia, June 2016 | Volume 7 | Article 203 9 Stress of Virus and Drought to Grapevine Cui et al. FIGURE 8 | Relative expressions of the ABA biosynthesis genes
NCED1 (A), NCED2 (B), and ZEP (C) of in vitro plantlets of grape
‘Cabernet Sauvignon’ at different time points. H+0 and H+4 = healthy
plantlets with 0 and 4% PEG stress, respectively. V+0 and V+4 =
virus-infected plants with 0 and 4% PEG stress, respectively. Data were
presented as means ± SE and with different letters within the same time of
analysis are significantly different at P ≤0.05. FIGURE 7 | Effects of GLRaV-3 infection and drought stress on content
of ABA (A) and IAA (B) of in vitro plantlets of grape ‘Cabernet
Sauvignon’. Data were presented as means ± SE and with different letters
within the same parameter are significantly different at P ≤0.05. 2003). There have been many studies on drought-induced MDA,
but data on virus-induced MDA have been quite limited. Huang
et al. (2008) and Li et al. (2011) found the MDA level markedly
FIGURE 8 | Relative expressions of the ABA biosynthesis genes
NCED1 (A), NCED2 (B), and ZEP (C) of in vitro plantlets of grape
‘Cabernet Sauvignon’ at different time points. H+0 and H+4 = healthy
plantlets with 0 and 4% PEG stress, respectively. V+0 and V+4 = FIGURE 7 | Effects of GLRaV-3 infection and drought stress on content
of ABA (A) and IAA (B) of in vitro plantlets of grape ‘Cabernet
Sauvignon’. Data were presented as means ± SE and with different letters
within the same parameter are significantly different at P ≤0.05. FIGURE 7 | Effects of GLRaV-3 infection and drought stress on content FIGURE 7 | Effects of GLRaV-3 infection and drought stress on content
of ABA (A) and IAA (B) of in vitro plantlets of grape ‘Cabernet
Sauvignon’. Data were presented as means ± SE and with different letters
within the same parameter are significantly different at P ≤0.05. FIGURE 8 | Relative expressions of the ABA biosynthesis genes
NCED1 (A), NCED2 (B), and ZEP (C) of in vitro plantlets of grape
‘Cabernet Sauvignon’ at different time points. DISCUSSION H+0 and H+4 = healthy
plantlets with 0 and 4% PEG stress, respectively. V+0 and V+4 =
virus-infected plants with 0 and 4% PEG stress, respectively. Data were
presented as means ± SE and with different letters within the same time of
analysis are significantly different at P ≤0.05. 2003). There have been many studies on drought-induced MDA,
but data on virus-induced MDA have been quite limited. Huang
et al. (2008) and Li et al. (2011) found the MDA level markedly
increased in the plants under drought, indicating the occurrence
of damage to cell membranes. These data were consistent with
ours. In our study, virus infection did not cause significant
accumulation of MDA, while 4% PEG resulted in significant
increase in MAD and the increased MAD was much obvious in
plantlets co-stressed by drought and the virus, indicating extra
injury to cell membranes by the interaction of drought stress and
virus-infection. been shown to influence the expression of a number of genes
and therefore manipulate various processes including growth,
PCD, abiotic stress responses and pathogen defense (Gill and
Tuteja, 2010). The production of ROS is the first response
when plants are under biotic and abiotic stress (Gara et al.,
2003; Evans et al., 2006), and as the stress continues, massive
and prolonged ROS production, called an oxidative burst,
occurs in the stressed cells (Gara et al., 2003). Accumulation
of ROS in the virus-infected (Evans et al., 2006) or in abiotic-
stressed plants may be associated with stress-induced oxidative
bursts (Gara et al., 2003; Evans et al., 2006). Such oxidative
bursts serve a number of protective functions activating defense
responses of plants to biotic and abiotic stress (Gara et al.,
2003). Reactive oxygen species (ROS), particularly O·−
2
and H2O2,
are subproducts responding to abiotic and biotic stress (Bolwell
et al., 1999). A number of studies have shown that accumulation
of O·−
2
and H2O2 considerably increased in plants infected
by virus (Evans et al., 2006; Zhang et al., 2014) or stressed
by drought (Duan et al., 2010; Gill and Tuteja, 2010; Sun
et al., 2013; Noctor et al., 2014). In the present study, we
found that stress by GLRaV-3 infection or drought increased
the accumulation of O·−
2
and H2O2, and that accumulation
was much greater when the in vitro shoots were stressed
simultaneously by GLRaV-3 and PEG-induced drought. Frontiers in Physiology | www.frontiersin.org DISCUSSION ROS has June 2016 | Volume 7 | Article 203 Frontiers in Physiology | www.frontiersin.org 10 Stress of Virus and Drought to Grapevine Cui et al. Cui et al. FIGURE 9 | Relative expressions of the IAA biosynthesis genes
TAR2 (A), TAR3 (B), TAR4 (C), and YUC1 (D) of in vitro plantlets of grape
‘Cabernet Sauvignon’ at different time points. H+0 and H+4 = healthy
plants with 0 and 4% PEG stress, respectively. V+0 and V+4 = virus-infected
plants with 0 and 4% PEG stress, respectively. Data were presented as
means ± SE and with different letters within the same time of analysis are
significantly different at P ≤0.05. (Gill and Tuteja, 2010). SOD was believed to be the first
line of defense against the toxic effects of elevated levels
of ROS caused by abiotic or biotic stress (Gill and Tuteja,
2010). FIGURE 9 | R l ti
i
f th
IAA bi
th
i ABA generally inhibits plant growth and is thought to act as a
messenger when plants were stressed by either drought or virus
(Jameson and Clarke, 2002; Zhu, 2002). Increased accumulation
in ABA in plants under drought stress has been well-documented
(Zhu, 2002), whereas there existed only a few studies on virus-
induced ABA and no studies on changes of ABA levels in plants
co-stressed by drought and virus (Jameson and Clarke, 2002). The present study confirmed again grapevine in vitro plantlets
responded to drought stress by increasing ABA level, and found
the level of ABA was higher in the GLRaV-3 infected plantlets
than in the healthy ones. Similar results were also reported in
plants infected with other viruses (Fraser and Whenham, 1989;
Wang et al., 2011; Alazem et al., 2014). The increased levels of
ABA were associated with up-regulation of ABA biosynthesis
genes including NCED1, NCED2, and ZEP, which were the
major genes for ABA biosynthesis in grapevine (Wheeler et al.,
2009). These findings supported the speculation of Alazem et al. (2014) that the induction of ABA and up-regulation of the
ABA biosynthesis genes may be common features of RNA virus
infection. Increased levels of ABA in plants stressed by drought
or infected with virus may eventually result in reduced growth,
as found in the present study. IAA
can
regulate
diverse
processes
including
growth,
development and physiological metabolisms of plants (Kieffer
et al., 2010; Swarup and Péret, 2012). DISCUSSION Decreased IAA levels
induced by drought stress have been reported in Brassica napus
(Qaderi et al., 2006), Phaseolus vulgaris (Figueiredo et al., 2008),
and Cotinus coggygria (Li et al., 2011). Reduced levels of IAA by
virus infection were also found by Smith et al. (1968), Lockhart
and Semancik (1970), Rao and Narasimham (1974), Rajagopal
(1977), and Li et al. (2013). The results reported here were
consistent with those mentioned above. The present study further
found that co-stress by drought and virus resulted in much
lower levels of IAA than single stress by either drought or virus,
and down-regulation of expressions of IAA biosynthetic genes
including TAR2, TAR3, TAR4, and YUC1 was associated with
the decreased levels of IAA. Systemic virus infection was shown
to influence the intercellular transport of the infected plants
(Jameson and Clarke, 2002), which may prevent directional
transport of auxin out of its source tissues, thus resulting in
inhibition of shoot growth and root formation, as reported in the
present study. FIGURE 9 | Relative expressions of the IAA biosynthesis genes
TAR2 (A), TAR3 (B), TAR4 (C), and YUC1 (D) of in vitro plantlets of grape
‘Cabernet Sauvignon’ at different time points. H+0 and H+4 = healthy
plants with 0 and 4% PEG stress, respectively. V+0 and V+4 = virus-infected
plants with 0 and 4% PEG stress, respectively. Data were presented as
means ± SE and with different letters within the same time of analysis are
significantly different at P ≤0.05. Numerous studies have demonstrated that plants are able to
protect against abiotic stress (Gill and Tuteja, 2010; Li et al.,
2011) or biotic stress (Gonçalves et al., 2013; Kyseláková et al.,
2013) by increasing the activities of antioxidant enzymes such
as SOD, CAT, and POD. In the present study, we found that
drought stress significantly increased the activities of SOD, POD,
and CAT in both the healthy and diseased plantlets, but single
virus infection increased slightly the activity of antioxidant
enzymes. The increased activities of three antioxidant enzymes
were much higher in the plantlets co-stressed by the drought
and virus, indicating a strong need for ROS removal. Frontiers in Physiology | www.frontiersin.org REFERENCES Dami, I., and Hughes, H. G. (1997). Effects of PEG-induced water stress on in
vitro hardening of ‘Valiant’ grape. Plant Cell Tiss. Org. Cult. 47, 97–101. doi:
10.1007/BF02318944 Abracheva, P., Rozenova, L., and Todorova, M. (1994). The influence of grapevine
fanleaf virus and stem pitting on in vitro grapevine cultures. Vitis 33,
181–182. Davies, C., and Robinson, S. (1996). Sugar accumulation in grape berries: cloning
of two putative vacuolar invertase cDNAs and their expression in grapevine
tissues. Plant Physiol. 111, 275–283. Alazem, M., Lin, K.-Y., and Lin, N. S. (2014). The abscisic acid pathway has
multifaceted effects on the accumulation of Bamboo mosaic virus. Mol. Plant
Microbe Interact. 27, 177–189. doi: 10.1094/MPMI-08-13-0216-R del Pozo, O., and Lam, E. (2003). Expression of the baculovirus p35 protein
in tobacco affects cell death progression and compromises N gene-mediated
disease resistance response to Tobacco mosaic virus. Mol. Plant Microbe
Interact. 16, 485–494. doi: 10.1094/MPMI.2003.16.6.485 Atkinson, N. J., and Urwin, P. E. (2012). The interaction of plant biotic and
abiotic stresses: from genes to the field. J. Exp Bot. 63, 3523–3543. doi:
10.1093/jxb/ers100 Dobrev, P. I., Haviˇcek, L., Vágner, M., Malbeck, J., and Kaminek, M. (2005). Purification and determination of plant hormones auxin and
abscisic acid using solid phase extraction and two-dimensional high
performance liquid chromatography. J. Chromat. A. 1075, 159–166. doi:
10.1016/j.chroma.2005.02.091 Bates, L. S., Waldren, R. P., and Teare, I. D. (1973). Rapid determination
of free proline for water stress studies. Plant Soil 39, 205–208. doi:
10.1007/BF00018060 Bertamini, M., Muthuchelian, K., and Nedunchezhian, N. (2004). Effect of
grapevine leafroll on the photosynthesis of field grown grapevine plants (Vitis
vinifera L. cv. Lagrein). J. Phytopathol. 152, 145–152. doi: 10.1111/j.1439-
0434.2004.00815.x Duan, Y. F., Zhang, W. S., Li, B., Wang, Y. N., Li, K. X., Sodmergen, et al. (2010). An endoplasmic reticulum response pathway mediates programmed cell death
of root tip induced by water stress in Arabidopsis. New Phytol. 186, 681–695. doi: 10.1111/j.1469-8137.2010.03207.x Bolwell, G. P., Blee, K. A., Butt, V. S., Davies, D. R., Gardner, S. L., Gerrish,
C., et al. (1999). Recent advances in understanding the origin of the
apoplastic oxidative burst in plant cells. Free Radic. Res. 31, S137–S145. doi:
10.1080/10715769900301431 Evans, C., Malin, G., Graham, P., and Mills, G. P. (2006). Viral infection of
Emiliania huxleyi (Prymnesiophyceae) leads to elevated production of reactive
oxygen species. J. Phycol. 42, 1040–1047. doi: 10.1111/j.1529-8817.2006.00256.x Böttcher, C., Boss, P. K., and Davies, C. (2011). REFERENCES Acyl substrate preferences of an
IAA-amido synthetase account for variations in grape (Vitis vinifera L.) berry
ripening caused by different auxinic compounds indicating the importance
of auxin conjugation in plant development. J. Exp. Bot. 62, 4267–4280. doi:
10.1093/jxb/err134 Faraloni, C., Cutino, I., Petruccelli, R., Leva, A. R., Lazzeri, S., and Torzillo, G. (2011). Chlorophyll fluorescence technique as a rapid tool for in vitro screening
of olive cultivars (Olea europaea L.) tolerant to drought stress. Environ. Experi. Bot. 73, 49–56. doi: 10.1016/j.envexpbot.2010.10.011 Figueiredo, M. V., Burity, H. A., Martínez, C. R., and Chanway, C. P. (2008). Alleviation of drought stress in the common bean (Phaseolus vulgaris L.) by
co-inoculation with Paenibacillus polymyxa and Rhizobium tropici. Appl. Soil
Ecol. 40, 182–188. doi: 10.1016/j.apsoil.2008.04.005 Böttcher, C., Burbidge, C. A., Boss, P. K., and Davies, C. (2013). Interactions
between ethylene and auxin are crucial to the control of grape (Vitis
vinifera L.) berry ripening. BMC Plant Biol. 13:222. doi: 10.1186/1471-
2229-13-222 Fraser, R. S. S., and Whenham, R. J. (1989). Abscisic-acid metabolism in tomato
plants infected with tobacco mosaic virus: relationships with growth, symptoms
and the Tm-1 gene for TMV resistance. Physiol. Mol. Plant Pathol. 34, 215–226. doi: 10.1016/0885-5765(89)90045-3 Bradford, M. M. (1976). A rapid and sensitive method for the quantization of
microgram quantities of protein utilizing the principle of protein–dye binding. Anal. Biochem. 72, 248–254. doi: 10.1016/0003-2697(76)90527-3 Fu, Q. S., Li, H. L., Cui, J., Zhao, B., and Guo, Y. D. (2009). Effects of water stress on
photosynthesis and associated physiological characters of Capsicum annuum L. Sci. Agric. Sin. 42, 1859–1866. doi: 10.3864/j.issn.0578-1752.2009.05.046 Chaves, M. M., Santos, T., Souza, C. R., Ortuño, M. F., Rodrigues, M. L., Lopes, C.,
et al. (2007). Deficit irrigation in grapevine improves water-use efficiency while
controlling vigour and production quality. Ann. Appl. Biol. 150, 237–252. doi:
10.1111/j.1744-7348.2006.00123.x Gaff, D. F., and Okong’O-Ogola, O. (1971). The use of non-permeating pigments
for testing the survival of cells. J. Exp. Bot. 22, 756–758. Christov, I., Stefanov, D., Velinov, T., Goltsev, V., Georgieva, K., Abracheva, P.,
et al. (2007). The symptomless leaf infection with grapevine leafroll associated
virus 3 in grown in vitro plants as a simple model system for investigation
of viral effects on photosynthesis. J. Plant Physiol. 164, 1124–1133. doi:
10.1016/j.jplph.2005.11.016 Gara, L. D., de Pinto, M. C., and Tommasi, F. (2003). The antioxidant systems vis-
à-vis reactive oxygen species during plant–pathogen interaction. Plant Physiol. Biochem. 41, 863–870. DISCUSSION Transgenic
plants that showed increase in CAT or POD or SOD activity
exhibited enhanced tolerance to drought or pathogen resistance In conclusion, GLRaV-3 infection or drought stress reduced
vegetative growth, affected physiological metabolisms, enhanced
cell membrane damage and cell death, increased accumulation of
H2O2 and O·−
2 , Co-stresses by virus infection and drought cause
much more deleterious effects, than either single stress, on the
studied parameters. Analysis of total soluble protein, free proline,
activities of antioxidant enzymes, MDA, levels of ABA and IAA
and expressions of their biosynthetic genes provide insights into
better understanding of the adverse effects of virus infection
and drought, in single and in combination, on the grapevine
in vitro plantlets. Results obtained in the present study address June 2016 | Volume 7 | Article 203 Frontiers in Physiology | www.frontiersin.org 11 Stress of Virus and Drought to Grapevine Cui et al. manuscript revision. QW design of the experiment, revision of
the manuscript and the financial support. use of virus-free plants and good irrigation toward achieving
sustainable development of the grapevine. AUTHOR CONTRIBUTIONS The authors gratefully acknowledge financial supports through
“948” program from State Forestry Administration of China
(project No. 2013-4-41) and through from Department of
Science & Technology of Shaanxi Province of China (Project No. 2013KTCL02-01). ZC main work of all the experiments, data analysis and
manuscript. WB
phytohormone
analysis
and
ROS
gene
expression analysis. XH the activity of the antioxidant enzymes
YX design of the experiment and the financial support PL
valuable discussion. AW the guide of the experiment and REFERENCES Growth and physiological
responses of canola (Brassica napus) to three components of global climate
change: temperature, carbon dioxide and drought. Physiol. Plant. 128, 710–721. doi: 10.1111/j.1399-3054.2006.00804.x Li, Y., Zhao, H., Duan, B., Korpelainen, H., and Li, C. (2011). Effect of drought and
ABA on growth, photosynthesis and antioxidant system of Cotinus coggygria
seedlings under two different light conditions. Environ. Exp. Bot. 71, 107–113. doi: 10.1016/j.envexpbot.2010.11.005 Radwan, D. E. M., Fayez, K. A., Mahmoud, S. Y., Hamad, A., and Lu, G. Q. (2007). Physiological and metabolic changes of Cucurbita pepo leaves in
response to zucchini yellow mosaic virus (ZYMV) infection and salicylic acid
treatments. Plant Physiol. Biochem. 45, 480–489. doi: 10.1016/j.plaphy.2007. 03.002 Livak, K. J., and Schmittgen, T. D. (2001). Analysis of relative gene expression
data using real-time quantitative PCR and the 2(-Delta Delta C (T)) method. Methods 25, 402–408. doi: 10.1006/meth.2001.1262 Lockhart, B. E., and Semancik, J. S. (1970). Growth inhibition, peroxidase
and 3–indoleacetic acid oxidase activity and ethylene production in cowpea
mosaic virus infected cowpea seedlings. Phytopathology 60, 553–554. doi:
10.1094/Phyto-60-553 Rajagopal, R. (1977). Effect of tobacco mosaic virus infection on the endogenous
levels of indoleacetic, phenylacetic and abscisic acids of tobacco leaves
in various stages of development. Z. Pflanzenphysiol. 83, 403–409. doi:
10.1016/S0044-328X(77)80046-9 Lokhande, V. H., Nikam, T. D., Patade, V. Y., Ahire, M. L., and Suprasanna, P. (2011). Effects of optimal and supra-optimal salinity stress on antioxidative
defence, osmolytes and in vitro growth responses in Sesuvium portulacastrum
L. Plant Cell Tiss. Organ. Cult. 104, 41–49. doi: 10.1007/s11240-010-9802-9 Rao, M. R. K., and Narasimham, B. (1974). Endogenous auxin levels as influenced
by infectious variegation’ in Lisbon lemon leaves. South Ind. Hort. 22, 138–139. Ren, F., Dong, Y.-F., Zhang, Z.-P., Fan, X.-D., Hu, G.-J., and Zhu, H.-J. (2013). Review of advances and perspectives in virus-resistant transgenic grapevine
studies. J. Hortic. Sci. 40, 1633–1644. Lokhande, V. H., Nikam, T. D., and Penna, S. (2010). Biochemical, physiological
and growth changes in response to salinity in callus cultures of Sesuvium
portulacastrum L. Plant Cell Tiss. Organ. Cult. 102, 17–25. doi: 10.1007/s11240-
010-9699-3 Romero-Puertas, M. C., Rodríguez-Serrano, M., Corpas, F. J., Gomezó, M., del Río,
L. A., and Sandalio, L. M. (2004). Cadmium-induced subcellular accumulation
of O2−and H2O2 in pea leaves. Plant Cell Environ. 27, 1122–1134. doi:
10.1111/j.1365-3040.2004.01217.x Lu, S., Peng, X., Guo, Z., Zhang, G., Wang, Z., Wang, C., et al. (2007). REFERENCES Rapid changes in the levels of polyribosomes in Zea mays in
response to water stress. Plant Physiol. 46, 281–285. doi: 10.1104/pp.46.2.281 Huang, X., Yin, C., Duan, B., and Li, C. (2008). Interactions between drought and
shade on growth and physiological traits in Populus Cathayana populations. Can. J. Forest Res. 38, 1877–1887. doi: 10.1139/X08-040 Padder, B. A. (2014). Plant disease resistance genes: from perception to signal
transduction. Plant Signal. 20, 345–354. doi: 10.1007/978-81-322-1542-4_20 Jameson, P. E., and Clarke, S. F. (2002). Hormone-virus interactions in plants. Crit. Rev. Plant Sci. 21, 205–228. doi: 10.1080/0735-260291044241 Parida, A. K., Dagaonkar, V. S., Phalak, M. S., Umalkar, G. V., and Aurangabadkar,
L. P. (2007). Alterations in photosynthetic pigments, protein and osmotic
components in cotton genotypes subjected to short-term drought stress
followed by recovery. Plant Biotech. Rep. 1, 37–48. doi: 10.1007/s11816-006-
0004-1 Khristov, I. K., Stefanov, D., Goltsev, V. N., and Abrasheva, P. (2001). Effects
of grapevine fanleaf and stem pitting viruses on the photosynthetic activity
of grapevine plants grown in vitro. Russ. J. Plant Physiol. 48, 473–477. doi:
10.1023/A:1016799329371 Patade, V. Y., Suprasanna, P., and Bapat, V. A. (2008). Effects of salt stress
in relation to osmotic adjustment on sugarcane (Saccharum officinarum L.)
callus cultures. Plant Growth Regul. 55, 169–173. doi: 10.1007/s10725-008-
9270-y Kieffer, M., Neve, J., and Kepinski, S. (2010). Defining auxin response
contexts in plant development. Curr. Opin. Plant Biol. 13, 12–20. doi:
10.1016/j.pbi.2009.10.006 Kyseláková, H., Sedláˇrová, M., Kubala, M., Nožková, V., Piterková, J., Luhová, L.,
et al. (2013). Reactive oxygen and nitrogen species and hormone signalling in
systemic infection of pea by Pea enation mosaic virus. Plant Protect. Sci. 49,
105–119. Permar, V., Singh, A., Pandey, V., Alatar, A. A., Faisal, M., Jain, R. K., et al. (2014). Tospo viral infection instigates necrosis and premature senescence by micro
RNA controlled programmed cell death in Vigna unguiculata. Physiol. Mol. Plant Pathol. 88, 77–84. doi: 10.1016/j.pmpp.2014.09.004 Li, J. W., Wang, B., Song, X. M., Wang, R. R., Zhang, H., Zhang, Z. B., et al. (2013). Potato leafroll virus (PLRV) and Potato virus Y (PVY) influence vegetative,
physiological metabolism of in vitro-cultured shoots of potato (Solanum
tuberosum L.). Plant Cell Tiss. 114, 313–324. doi: 10.1007/s11240-013-0327-x Prasch, C. M. (2013). Simultaneous application of heat, drought, and virus to
Arabidopsis plants reveals significant shifts in signaling networks. Plant Physiol. 162, 1849–1866. doi: 10.1104/pp.113.221044 Qaderi, M. M., Kurepin, L. V., and Reid, D. M. (2006). REFERENCES doi: 10.1016/S0981-9428(03)00135-9 Gill, S. S., and Tuteja, N. (2010). Reactive oxygen species and antioxidant
machinery in abiotic stress tolerance in crop plants. Plant Physiol. Biochem. 48,
909–930. doi: 10.1016/j.plaphy.2010.08.016 j j
Chutia, J., and Borah, S. P. (2012). Water stress effects on leaf growth and
chlorophyll content but not the grain yield in traditional rice (Oryza
sativa Linn.) genotypes of Assam, India II. Protein and proline status in
seedlings under PEG induced water stress. Am. J. Plant Sci. 3, 971–980. doi:
10.4236/ajps.2012.37115 Gonçalves, L. S. A., Rodrigues, R., Diz, M. S. S., Robaina, R. R., do Amaral Júnior,
A. T., Carvalho, A. O., et al. (2013). Peroxidase is involved in Pepper yellow
mosaic virus resistance in Capsicum baccatum var. Pendulum. Genet. Mol. Res. 12, 1411–1420. doi: 10.4238/2013.April.26.3 Cui, Z. H., Bi, W. L., Chen, P., Xu, Y., and Wang, Q. C. (2015). Abiotic stress
improves in vitro biological indexing of Grapevine leafroll-associated virus-
3 in red grapevine cultivars. Aust. J. Grape Wine Res. 21, 490–495. doi:
10.1111/ajgw.12146 Greenberg, J. T., and Yao, N. (2004). The role and regulation of programmed
cell death in plant–pathogen interactions. Cell. Microbiol. 6, 201–211. doi:
10.1111/j.1462-5822.2004.00361.x June 2016 | Volume 7 | Article 203 Frontiers in Physiology | www.frontiersin.org 12 Stress of Virus and Drought to Grapevine Cui et al. Hadidi, A., and Barba, M. (2011). “Economic impact of pome and stone fruit
viruses and viroids,” in Virus and Virus-like Diseases of Pome and Stone Fruits,
eds A. Hadidi, M. Barba, T. Candresse, and W. Jelkman (St. Paul, MN:
American Phyto pathological Society), 1–7. Moutinho-Pereira, J., Correia, C. M., Gonclve, B., Bacelar, E. A., Coutinho, J. F.,
Ferreira, H. F., et al. (2012). Impacts of leafroll-associated viruses (GLRaV-
1 and -3) on the physiology of the Portuguese grapevine cultivar ‘Touriga
Nacional’ growing under field conditions. Annu. Appl. Biol. 160, 237–249. doi:
10.1111/j.1744-7348.2012.00536.x Haq, I. U., Nazia, P., Muhammad, T. R., and Muhammad, U. D. (2012). Comparative characteristics of micropropagated plantlets of banana from
BBTV-infected explants to its normal and saline stressed cultures. Pak. J. Bot. 44, 1127–1130. Murashige, T., and Skoog, F. (1962). A revised medium for rapid growth
and bioassays with tobacco cell cultures. Physiol. Plant. 15, 473–497. doi:
10.1111/j.1399-3054.1962.tb08052.x Noctor, G., Mhamdi, A., and Foyer, C. H. (2014). Roles of reactive oxygen
metabolism in drought: not so cut and dried. Plant Physiol. 164, 1636–1648. doi: 10.1104/pp.113.233478 Hsiao, T. C. (1970). REFERENCES In vitro
selection of salinity tolerant variants from triploid bermudagrass (Cynodon
transvaalensis x C. dactylon) and their physiological responses to salt and
drought stress. Plant Cell Rep. 26, 1413–1420. doi: 10.1007/s00299-007-0339-2 Sampol, B., Bota, J., Riera, D., Medrano, H., and Flexas, J. (2003). Analysis of the
virus-induced inhibition of photosynthesis in malmsey grapevines. New Phytol. 160, 403–412. doi: 10.1046/j.1469-8137.2003.00882.x Markovi´c, Z., Preiner, D., Bošnjak, A. M., Safner, T., Stupi´c, D., Andabaka, Ž.,
et al. (2014). In vitro introduction of healthy and virus-infected genotypes
of native Croatian grapevine cultivars. Cent. Eur. J. Biol. 9, 1087–1098. doi:
10.2478/s11535-014-0337-7 Smith, S. H., McCall, S. R., and Harris, J. H. (1968). Auxin transport in curly top
virus-infected tomato. Phytopathology 58, 1669–1670. Sullivan, C. Y. (1972). “Mechanism of heat and drought resistance in grain
sorghum and methods of measurement,” in Sorghum in the seventie, eds N. G. Sullivan, C. Y. (1972). “Mechanism of heat and drought resistance in grain
sorghum and methods of measurement,” in Sorghum in the seventie, eds N. G. P. Raoand and L. R. House(New Delhi: Oxford and IBH), 247–264. Maroco, J. P., Rodrigues, M. L., Lopes, C., and Chaves, M. (2002). Limitations
to leaf photosynthesis in field-grown grapevine under drought–metabolic and
modeling approaches. Funct. Plant Biol. 29, 451–459. doi: 10.1071/PP01040 Sun, J., Guo, J., Zeng, J., Han, S., Song, A., Chen, F., et al. (2013). Changes in leaf
morphology, antioxidant activity and photosynthesis capacity in two different
drought-tolerant cultivars of chrysanthemum during and after water stress. Sci. Hortic. 161, 249–258. doi: 10.1016/j.scienta.2013.07.015 Martelli, G. P. (2012). “Grapevine virology highlights: 2010–2012,” in Proceedings
of the 17th Congress of ICVG (Davis, CA), 13–31. Masia, A. (2003). “Physiological effects of oxidative stress in relation to ethylene in
postharvest produce,” in Postharvest Oxidative Stress in Horticultural Crops, ed
D. M. Hodges (New York, NY: Food Products Press), 165–197. Suzuki, N., Rivero, R. M., Shulaev, V., Blumwald, E., and Mittler, R. (2014). Abiotic
and biotic stress combinations. New Phytol. 203, 32–43. doi: 10.1111/nph. 12797 D. M. Hodges (New York, NY: Food Products Press), 165–197. Michel, B. L. (1983). Evaluation of the water potentials of solutions of polyethylene
glycol 8000 both in the absence and presence of other solutes. Plant Physiol. 72,
66–70. doi: 10.1104/pp.72.1.66 Swarup, R., and Péret, B. (2012). AUX/LAX family of auxin influx carriers – an
overview. Front. Plant Sci. 3:225. REFERENCES doi: 10.3389/fpls.2012.00225 June 2016 | Volume 7 | Article 203 Frontiers in Physiology | www.frontiersin.org 13 Stress of Virus and Drought to Grapevine Cui et al. Tanne, E., Spiegel-Roy, P., and Shlamovitz, N. (1996). Rapid in
vitro
indexing of grapevine viral diseases: the effect of stress-inducing agents
on the diagnosis of leafroll. Plant Dis. 80, 72–974. doi: 10.1094/P.D.-
80-0972 Xu, P., Chen, F., Mannas, J. P., Feldman, T., Sumner, L. W., and Roossinck, M. J. (2008). Virus infection improves drought tolerance. New Phytol. 180, 911–921. doi: 10.1111/j.1469-8137.2008.02627.x Zhang, K., Song, Y. P., Wang, Y., Li, K., Gao, L., Zhong, Y. K., et al. (2014). Differential necrotic lesion formation in soybean cultivars in response to
Soybean mosaic virus. Eur. J. Plant Pathol. 139, 525–534. doi: 10.1007/s10658-
014-0408-7 UNESCO Water Portal (2007). Available online at: http://www.unesco.org/water
(Accessed 25 October 2007). Wang, S.-M., Hou, X.-L., Ying, L., Cao, X.-W., Zhang, S., and Wang, F. (2011). Effects of Turnip mosaic virus (TuMV) on endogenous hormones
and transcriptional level of related genes in infected non-heading Chinese
cabbage. J. Nanjing Agric. Univ. 5, 13–19. doi: 10.7685/j.issn.1000-2030.2011. 05.003 Zhang, N., Zhao, B., Zhang, H. J., Weeda, S., Yang, C., Yang, Z. C., et al. (2013). Melatonin promotes water-stress tolerance, lateral root formation, and seed
germination in cucumber (Cucumis sativus L.). J. Pineal Res. 54, 15–23. doi:
10.1111/j.1600-079X.2012.01015.x 10.1111/j.1600-079X.2012.01015.x Watanabe, S., Kojima, K., Ide, Y., and Sasaki, S. (2000). Effects of saline
and
osmotic
stress
on
proline
and
sugar
accumulation
in
Populus
euphratica in vitro. Plant Cell Tiss. 63, 199–206. doi: 10.1023/A:1010619
503680 Zhu, J. K. (2002). Salt and drought stress signal transduction in plants. Annu. Rev. Plant Biol. 53, 247–273. doi: 10.1146/annurev.arplant.53.091401.143329 Conflict of Interest Statement: The authors declare that the research was
conducted in the absence of any commercial or financial relationships that could
be construed as a potential conflict of interest. Wheeler, S., Loveys, B., Ford, C., and Davies, C. (2009). The relationship between
the expression of abscisic acid biosynthesis genes, accumulation of abscisic
acid and the promotion of Vitis vinifera L. berry ripening by abscisic acid. Aust. J. of Grape Wine Res. 15, 195–204. doi: 10.1111/j.1755-0238.2008. 00045.x Copyright © 2016 Cui, Bi, Hao, Xu, Li, Walker and Wang. This is an open-access
article distributed under the terms of the Creative Commons Attribution License (CC
BY). Frontiers in Physiology | www.frontiersin.org June 2016 | Volume 7 | Article 203 REFERENCES The use, distribution or reproduction in other forums is permitted, provided the
original author(s) or licensor are credited and that the original publication in this
journal is cited, in accordance with accepted academic practice. No use, distribution
or reproduction is permitted which does not comply with these terms. Woodward, A. J., and Bennett, I. J. (2005). The effect of salt stress and abscisic
acid on proline production, chlorophyll content and growth of in vitro
propagated shoots of Eucalyptus camaldulensis. Plant Cell Tiss. 82, 189–200. doi: 10.1007/s11240-005-0515-4 June 2016 | Volume 7 | Article 203 Frontiers in Physiology | www.frontiersin.org 14
|
https://openalex.org/W3166896493
|
http://liu.diva-portal.org/smash/get/diva2:1574903/FULLTEXT01
|
English
| null |
“Teach more, but do not expect any applause”: Are Women Doubly Discriminated Against in Universities’ Recruitment Processes?
|
Journal of academic ethics
| 2,021
|
cc-by
| 8,548
|
Journal of Academic Ethics
https://doi.org/10.1007/s10805-021-09421-5 Journal of Academic Ethics
https://doi.org/10.1007/s10805-021-09421-5 Abstract Studies repeatedly find that women and men experience life in academia differently. Impor-
tantly, the typical female academic portfolio contains less research but more teaching and
administrative duties. The typical male portfolio, on the other hand, contains more research
but less teaching and administration. Since previous research has suggested that research
is a more valued assignment than teaching in academia, we hypothesise that men will be
ranked higher in the peer-evaluations that precede hirings to tenured positions in Swedish
academia. We analyze 861 peer review assessments of applicants in 111 recruitment pro-
cesses in Economics, Political Science, and Sociology at the six largest Swedish universi-
ties. Our findings confirm that the premises established in previous research are valid in
Sweden too: Women have relatively stronger teaching merits and men relatively stronger
research merits, and also that, on balance, research is rewarded more when applicants are
ranked by reviewers. Accordingly, male applicants are ranked higher compared to female
applicants. Keywords Higher education · Academia · Gender gap · Gendered division of labour ·
Recruitment “Teach more, but do not expect any applause”: Are Women
Doubly Discriminated Against in Universities’ Recruitment
Processes? Douglas Brommesson1 · Gissur Ó Erlingsson2 · Jörgen Ödalen3 · Mattias Fogelgren4 Accepted: 24 May 2021
© The Author(s) 2021 Accepted: 24 May 2021
© The Author(s) 2021 Introduction For the past 40 years, the share of female university students as well as doctoral students
has increased dramatically in most parts of the world. However, the transformation towards
gender equality in academia is not nearly as apparent when it comes to the share of women
among tenured lecturers and full professors. The share of women drops about ten percent-
age points at each promotional stage after graduate studies, so that ultimately the share of
women who are full professors is down to approximately one-third (McFarland et al., 2017; * Douglas Brommesson
douglas.brommesson@lnu.se
1
Linnaeus University, Växjö, Sweden
2
Linköping University, Linköping, Sweden
3
Mälardalen University, Västerås, Sweden
4
Government Offices, Ministry of Justice, Stockholm, Sweden 3456789)
1 3 D. Brommesson et al. Renwick Monroe et al., 2014). Thus, it is unsurprising that the road towards greater gender
equality in tenure rates has been described as ‘excruciatingly slow’ (Larrán George et al. 2016; see also Tessens et al., 2011; Marschke et al., 2007), and others maintain that the
gender gap in tenure rates is actually not closing at all (e.g. Perna, 2005). Renwick Monroe et al., 2014). Thus, it is unsurprising that the road towards greater gender
equality in tenure rates has been described as ‘excruciatingly slow’ (Larrán George et al. 2016; see also Tessens et al., 2011; Marschke et al., 2007), and others maintain that the
gender gap in tenure rates is actually not closing at all (e.g. Perna, 2005). g
g p
y
g
g
Arguably, this situation is intimately associated with the gendered division of labour in
academia, which several studies have observed. This division of labour inevitably hampers
women’s career opportunities since there is an undeniable and sharp difference in how men
and women experience academia. These differences are well characterised in The Wash-
ington Post’s “Monkey Cage´s” symposium on the gender gap in academia (Voeten, 2013). To name but a few such differences that have been reported in the literature: women teach
and perform lesser-ranked administrative services to a larger extent than men, while men
conduct more research (see e.g. Kalm, 2019; Guardino & Borden, 2017; Coate & Kandiko
Howson, 2016; Nature, 2016; European Commission, 2008). Furthermore, a gender gap in
citations has been observed1 (Dion et al. 1 This can, at least to some extent, be ascribed to the tendency of men to self-cite to a considerably larger
degree than women do (King et al., 2017). From the 1990s and onwards, men self-cited 70 percent more
than women.
2 Note that this particular result applies to conferences within the field of Economics. 2 Note that this particular result applies to conferences within the field of Economics. Introduction 2018); a gender bias is present in research grant
peer review (Tamblyn et al., 2018); male academics are more likely than women to be
accepted to peer-reviewed2 conferences (Times Higher Education, 2019); and gender plays
a significant role in influencing how students rate their instructors – to the disadvantage of
women (MacNell et al., 2015). Furthermore, even when women are just as scientifically
competent as men, there is still a significant gender gap in career advancement that is not
explained by gender differences in productivity (Filandra & Pasqua, 2019; see also Wullum
Nielsen, 2016, however compare Madison & Fahlman, 2020; Kulp, 2020), and women are
more likely to be steered into part-time positions, making it less likely for them to transfer
to full-time positions (Lundby & Warme, 1990).i In addition, studies have shown that male academics are more satisfied with their sal-
ary, their promotions, and experience greater job satisfaction compared to their female col-
leagues (Okbara et al., 2005). Finally, female academics experience higher overall levels of
job-related stress (Doyle & Hind, 2002). Taken together, these gender differences make it
reasonable to expect that they negatively affect the career opportunities of female scholars. For instance, even after controlling for productivity, women are less likely than men to
be promoted to full professors (August & Waltman, 2004; Perna, 2001), and furthermore,
salaries for women are generally lower compared to those of men with similar academic
track-records (Barbezat & Hughes, 2005; Carr et al., 2015).f It is fair to say, then, that gender differences in academia regarding division of labour,
availability of career opportunities and salaries are both real and significant. In this article,
we set out to perform an explorative study on the effect of these differences in terms of who
is recruited to tenured positions and based on what merits. The aim is to analyse whether
the gender differences of division of labour, described above, are reproduced in universi-
ties’ hiring processes, amounting to what we dub a potential ‘double discrimination’ in
academia. This will be studied in a Swedish context. Some previous qualitative studies of the
Swedish case indicate that there is indeed a gendered division of labour in Swedish
academia – to the disadvantage of women (Angervall & Beach, 2017, 2018; Angervall
et al., 2015). Introduction A review from the Swedish National Board of Higher Education, which was 1 3 “Teach more, but do not expect any applause”: Are Women Doubly… based on 22 different projects, concluded that the meritocratic ambitions of higher educa-
tion institutions in Sweden is hampered ‘by norms and values that confirm men as supe-
rior and that different conditions are active regarding women’s and men’s meriting and
carrier development.’ Such norms and values are fostered by informal power configura-
tions and procedures, making it hard for those who do not have access to informal net-
works to compete on equal terms (UHR, 2020: 29f, our translation). We can also tell from
large-n studies that variations along gender lines still exist to a rather high degree in Swe-
den, in terms of recruitment to higher education, recruitment to prestigious programs and
new demarcations between men and women who make it to higher education (Berggren,
2008, 2011, on the latter point see also Haley, 2018). All of this notwithstanding, Sweden is still consistently ranked among the most gen-
der equal countries (e.g. Equal Measures, 2019; The Global Gender Gap Report 2018;
cf. Madison & Fahlman, 2020). Against this backdrop we can draw the conclusion that
Sweden is apparently not a gender equal utopia, but nevertheless a country that has
taken important steps, and more steps compared to many other countries, towards gender
equality. Despite the remaining challenges related to gender equality we therefore view
Sweden as a ‘more likely’ case when it comes to finding gender equality for academics,
at least compared to most other countries. Much previous research on equality in academia has been carried out in somewhat less
gender-equal, Anglo-Saxon contexts. A crucial question is, therefore, whether gendered
differences can be observed when we turn our eyes to the Swedish setting, where gender
equality – according to several international comparisons – is regarded to have progressed
the most. Consequently, viewing Sweden as a relatively more likely setting to observe gen-
der equality, we argue that if reproduction of gender differences is observed in Swedish
academic hiring processes, it is plausible to conclude that similar differences persist, and
are presumably magnified, in other settings. A Double Discriminatory Effect? Assumptions and Hypotheses As concluded by the European Commission (2008), there still exists an often-held gender
stereotype that views female academics, first and foremost, as ‘talented teachers’, exhibit-
ing excellent soft skills such as communication and a sensitive approach to, and open ear
for, students. On the other hand, men in academia tend to be viewed as analytical, objec-
tive, hard thinking researchers. As stated in the EC-report, these images are mirrored in
a division of labour where female academics are typically stuck with teaching duties and
administrative tasks with lower status, whilst men are doing research, echoing the stereo-
type ‘women teach, men think’.i As already indicated, the descriptions in the EC-report have been confirmed in pre-
vious as well as later studies. A number of studies have shown that women and men
employed in academia experience and react to their work environments differently vious as well as later studies. A number of studies have shown that women and men
employed in academia experience and react to their work environments differently
– largely in ways unfavorable to women. Research indicates that female academics are
paid less than men (Carr et al., 2015; Toutkoushian & Conley, 2005). Moreover, it has
been found that female academics do more housework at home than their male counter-
parts (Scheibinger & Gilmartin, 2010), and they find it much harder to achieve a work-
life balance (O’Laughlin & Bischoff, 2005). Moreover, female academics spend more
time on teaching and on public-engagement tasks, and less time on research, than their
male counterparts, this according to a survey of UK university staff in science-based
subjects (Nature, 2017; see also Sax et al., 2002). Partly as a consequence of this, some-
what older studies found that female faculty are less likely to be promoted to the rank
of full professors – even after controlling for productivity and human capital (Perna,
2001; Toutkoushian, 1999, cf. Filandra & Pasqua, 2019), although it is unclear whether
this imbalance has been redressed in recent years (Guarino & Borden, 2017). However,
a ‘smoking gun’ that indicates that there still is some way to go even in such a gender-
equal country as Sweden, is that women constitute only 28 per cent of the country’s full
professors (Allbright, 2019). Research Questions We argue that our explorative study – which asks whether the gender differences of divi-
sion of labour are reproduced in hiring processes – has the potential to further the knowl-
edge on gender divisions in higher education in general, and within the field of recruitment
to tenured positions in particular. In order to study these gender differences of division of
labour, and whether these differences have the potential to be reproduced, we analyse hir-
ing processes to tenured positions as senior lecturers that took place in Sweden between
2003 and 2013 within Economics, Political science and Sociology. Our data was collected
from peer-review evaluation reports (sakkunnigutlåtanden, in Swedish) which include
rankings of the applicants as well as reviewers’ arguments for their ranking decisions. We
describe this unique data-set in more detail below, but it should be stressed that the data
gives us opportunities to study the division between women and men in terms of a) how
they are ranked by their peers in the evaluation reports, and b) the specific merits – e.g. teaching vs. research – ascribed to women and men respectively. To fulfil the paper’s aim,
we ask the following research questions: • To what extent are women and men ascribed different types of merits in the evaluation
process? Are these differences consistent across disciplines?f f
• How often are different types of merits the determinant factor in the recruitment pro-
cess? Are there differences between disciplines in this respect? f
• How often are different types of merits the determinant factor in the recruitment pro-
cess? Are there differences between disciplines in this respect? 1 3 1 3 3 D. Brommesson et al. The paper proceeds as follows. First, we spell out the theoretical points of departure
where we present our assumptions of why gender divisions in academic recruitment pro-
cesses are to be expected. Second, we present our methodological considerations and dis-
cuss case selection and the specifics of the Swedish higher education system, before mov-
ing on to present the data and how it is analysed. Third, we proceed to present our results
and interpret these, before ending the paper with a concluding discussion. A Double Discriminatory Effect? Assumptions and Hypotheses Against this backdrop, if previous findings from the gen-
eral international literature on gender and academic careers are valid for the Swedish
context too, four hypotheses are specified and will be tested in this article: H1: Female applicants to tenured positions as senior lecturers have stronger teaching
merits, than male applicants. H2: Male applicants have stronger research merits, than female applicants. H2: Male applicants have stronger research merits, than female applicants. H2: Male applicants have stronger research merits, than female applicants. H3: Candidates with stronger research merits are prioritized, before candidates with
stronger teaching merits. H3: Candidates with stronger research merits are prioritized, before candidates with
stronger teaching merits. H3: Candidates with stronger research merits are prioritized, before candidates with
stronger teaching merits. H4: It follows from H1-H3 that male applicants are ranked higher than female appli-
cants in peer-review evaluations that precede the hiring of senior lecturers. In our analysis, three academic disciplines are included (Economics, Political Science,
and Sociology). This is done to facilitate an analysis of a high number of hiring processes
within a limited scope of time. The three disciplines are related, all being close to the core 1 3 “Teach more, but do not expect any applause”: Are Women Doubly… of social sciences. However, each discipline can reasonably be expected to have developed
their own separate norms and routines, which may affect how merits and hiring processes
are viewed. This means that although the differences between academic disciplines are not
at the center of our analyses, our design still provides us with the opportunity to explore
potential variation between disciplines. This endeavor is motivated by results from previous research on ‘academic tribal-
ism’ (Becher, 1989, 1994; Neumann, 2001). According to this literature, we could
expect Economics to be closer to the ideal type of what Biglan (1973) labelled ‘hard
pure’ (the ideal of natural sciences) while Sociology comes closer to ‘soft pure’ (the
ideal of social sciences), with Political Science somewhere in-between. In ‘hard pure’
disciplines we are, according to Becher, expected to find a culture that is described as
‘competitive, gregarious; politically well-organised; high publication rate; task-ori-
ented’, while the ‘soft pure’ culture is described with characteristics such as ‘individ-
ualistic, pluralistic; loosely structured; low publication rate; person-oriented’ (Becher,
1994). From this follows, we argue, an expectation of a stronger focus within Eco-
nomics on output that can be more easily measured (i.e. A Double Discriminatory Effect? Assumptions and Hypotheses research publications), while
Sociology and to some extent Political Science can be expected to have stronger focus
on portfolios with a better balance between teaching and research. This also reveals
a non-universal understanding of academic meritocracy where the understanding of
meritocracy varies between contexts. One form of meritocracy can be expected to
discriminate against women, while another form may not be expected to do so. Hence,
if the ‘academic tribalism’ argument has any merit, the disadvantage of women is
expected to be most pronounced in Economics and least so in Sociology. These differences are supported by research on how we define research activi-
ties. As some have argued, even the conceptions of research differ between men and
women, and a narrower conception work in the favor of men (Healey & Davis, 2019). A narrower understanding of research comes close to the ‘hard pure’ type above. What we have referred to as ‘soft pure’ can instead, tentatively, be expected to be
stronger both within certain disciplines and among women. Although our aim here
is straightforward and empirical, these observations still motivate us to reflect also
on social relationships between and within different genders, disciplines and research
environments. In sum, based on the empirical studies discussed above, women are expected to be dis-
criminated against twice. Since they, as a rule, are stereotyped as talented teachers and
administrators, they are allocated relatively less time and other resources to do research. Then, when they apply for tenured positions, they are expected to be discriminated against
again, since their talents in the pedagogical and administrative fields are valued less, com-
pared with research merits. This is precisely the potential double discriminatory effect we
aim to explore. Data and Methods Our dataset comprises 861 data points. One data point equals one applicant in one evalu-
ation report (321 female applicants, 536 male applicants, and 4 missing cases). Most hir-
ing processes involve two external reviewers who are urged to independently evaluate the 3 D. Brommesson et al. applicants and then rank them.3 This implies that, in most cases, each applicant forms two
data points in our data set. The 861 data points concern a total of 111 hiring processes at
the six largest Swedish universities: Gothenburg; Linköping; Lund; Stockholm; Umeå; and
Uppsala.4 The data we acquired from the universities did include more than these 111 hir-
ing processes, but we choose to exclude some of these for varying reasons. For instance,
we excluded hiring processes that did not include any competition because there was only
one applicant. We also excluded those hiring processes which concerned disciplines out-
side the scope of this study, and those where the evaluation reports were incomplete. These
are rare exceptions that do not affect the overall results of our study.5 f
In each of the evaluation reports, for all applicants, we measured the space in the report
devoted to teaching merits and research merits, respectively. If a reviewer devoted one page
discussing teaching merits for a certain candidate, and four pages to discussing research
merits, this was coded as 20 per cent teaching merits and 80 per cent research merits. We
also coded the gender of the applicant, the gender of the reviewer, the academic discipline
of the position the recruitment process concerns, and finally the ranking of the applicants
for the specific position. i
We use descriptive statistics to describe the merits of the applicants, and OLS regres-
sion analyses to study the effects of gender and of different kinds of merits on the ranking
of applicants. We discuss the use of these methods in more detail when we present the
results. Before moving on to our analyses, a few additional comments are warranted. One
important reservation concerns the difficulty reviewers encounter when measuring teach-
ing/pedagogical merits. While research merits (or scientific merits, as they are typically
referred to in the evaluation reports) are more easily quantified (e.g. 5 We requested all evaluation reports concerning the review of applicants for all open positions as senior
lecturers in Economics, Political science and Sociology during 2003–2013. We cannot guarantee that we
received all reports, but our random checks indicate that we have received all or at least almost all of the
reports.
6 This rather parsimonious treatment of teaching merits in many of the evaluation reports is interesting in
itself and says something about the weight ascribed to the teaching experience of the applicants, despite
teaching being the main task for many senior lecturers.
3 The evaluation and the ranking are reported to the faculty office at the relevant university in the form of a
public evaluation report.
4 The number of recruitment cases for each university in our data was: Lund: 24; Uppsala: 21; Gothenburg:
18; Linköping: 13; Umeå: 9; and Stockholm 16. The Effects of Different Kinds of Merits on the Ranking of Applicants Based on our review of previous research – admittedly, primarily carried out in Anglo-
Saxon settings – there are strong reasons to expect that female applicants to academic posi-
tions will have stronger teaching merits, while male applicants will have stronger research
merits. Since research historically and typically has had a higher status than teaching (see
e.g. Fairweather, 2005), we also expect that candidates with stronger research merits will
be prioritised before candidates with stronger teaching merits when reviewers ultimately
rank candidates. For instance, this is what Parker (2008) found for evaluations preceding
promotions to the ranks of reader and professor in the UK: research excellence is priori-
tised over teaching activities (cf. Fairweather, 2005). To test these hypotheses, we initially
present a descriptive overview of the applicants and their merits. In Table 1, we see that evaluation reports concerning female candidates are generally
more occupied with teaching merits compared to evaluations of male candidates. As we
can see in the bottom rows of the table, this pattern holds even if we only focus on the
top ranked candidates. However, reports on applicants for senior lectureships in Sociology
differ somewhat. Here, reports on female candidates devote slightly less space to teach-
ing merits, compared with the reports on male candidates. However, this ‘sociology’-fac-
tor does not challenge the overall picture: as a rule, evaluation of teaching merits is more
extensive in the case of female candidates. From these descriptive statistics, where ‘vol-
ume in evaluation reports’ is used as a proxy, we draw the tentative conclusion that female
applicants have (relatively) stronger teaching merits and (relatively) weaker research mer-
its. Nonetheless, a possible interpretation is that reviewers are primed to associate female
candidates more with teaching.ff Let us move on to study the effect of these differences on the ranking of the applicants. We do so through linear regression analyses where the effect of different variables on
the ranking of the candidates is tested. We start out by studying the effects on ranking Table 1 Distribution of teaching merits of all applicants within all disciplines
The share of the evaluation reports devoted to a discussion on the teaching merits of the candidate, where
teaching merits and research merits total 100 per cent. Data and Methods number of publica-
tions, impact factor or ranking of the journals where articles are published, number of cita-
tions, etc.), teaching merits are, as a rule, perceived to be harder to quantify and also evalu-
ate qualitatively. After having read an abundance of evaluation reports, we can conclude
that due to perceived difficulties of evaluating teaching merits, such evaluations are typi-
cally summarised in a rather parsimonious way. The reviewer usually concludes by stat-
ing something along the lines that the applicant has been a teacher for so-and-so many
years (or has taught for so-and-so many teaching hours), and can therefore be considered
an experienced teacher.6 In other words, the quantity of teaching is used as an extremely
crude proxy for measuring quality in teaching. At least to us, this implies a very undifferen-
tiated and shaky strategy.fi Given the (perceived) difficulties of the reviewers to evaluate teaching merits, a rather
short discussion on teaching merits, compared with a lengthier discussion on research mer-
its, should not necessarily be understood as an indication that the reviewer values teaching “Teach more, but do not expect any applause”: Are Women Doubly… merits less. However, there is no reason to expect that these negative conditions vary
between universities, disciplines or gender. We can, therefore, expect the reviewers to have
similar conditions for their work and that variations in the space devoted to different merits
between different groups of applicants must say something substantial about how different
merits are valued by evaluators. Such variation between universities, disciplines and gen-
der can have different explanations, and it is outside the scope of this article to explain such
variation. Our initial theoretical discussion does, however, provide us with potential expla-
nations against which the results could be read. We return to these potential explanations in
the concluding section. But let us now turn to our empirical results. The Effects of Different Kinds of Merits on the Ranking of Applicants The evaluation reports concern recruitment matters
for tenured positions as senior lecturers in Economics, Sociology, and Political science
Teaching merits % Economics
Teaching merits % Political science
Teaching merits % Sociology
All
Women
23.4
All
Women
25.2
All
Women
23.9
Men
19.7
Men
22.1
Men
25.2
Rank #1
Women
25.0
Rank #1
Women
24.8
Rank #1
Women
23.0
Men
18.0
Men
21.9
Men
25.8 Table 1 Distribution of teaching merits of all applicants within all disciplines le 1 Distribution of teaching merits of all applicants within all disciplines The share of the evaluation reports devoted to a discussion on the teaching merits of the candidate, where
teaching merits and research merits total 100 per cent. The evaluation reports concern recruitment matters
for tenured positions as senior lecturers in Economics, Sociology, and Political science The share of the evaluation reports devoted to a discussion on the teaching merits of the candidate, where
teaching merits and research merits total 100 per cent. The evaluation reports concern recruitment matters
for tenured positions as senior lecturers in Economics, Sociology, and Political science 1 3 D. Brommesson et al. decisions of the reasons given by reviewers for rankings. In our coding of the evaluation
reports we assessed whether the reviewers, when deciding upon the ranking of a particu-
lar applicant, based their ranking decision equally on the applicant’s research and teaching
merits, primarily on their research merits, or primarily on their teaching merits. Sometimes
the reviewers explicitly state their reasons for particular ranking decisions in the reports,
but on other occasions a measure of interpretation of the evaluation reports was required
in order to assess the ultimate reason behind the ranking decisions. This resulted in a cat-
egorisation of all our data points into three groups: applicants whose ranking was deter-
mined by research and teaching merits equally; applicants whose ranking was determined
by research merits; and applicants whose ranking was determined by teaching merits.f Next, we estimate the effect of belonging to a certain group on the ranking one is
afforded. The method for this is to represent group membership with dummy variables
that take on values 0 and 1; membership in a particular group is coded one whereas non-
membership in the group is coded zero. In order to avoid introducing multicollinearity, the
general rule is to include one less dummy variable in the model than there are categories. The Effects of Different Kinds of Merits on the Ranking of Applicants Consequently, we constructed two dummy variables; one for the group of applicants whose
ranking was determined by research merits, and one for the group whose ranking was
determined by teaching merits. This means that the applicants whose ranking was deter-
mined by research and teaching merits to an equal extent will receive the value 0 on both
dummy variables and will function as the reference group (Alkharusi, 2012). When dummy coding is used in OLS regression analyses, the overall results indi-
cate whether there is a relationship between the dummy variables and the dependent
variable, which in our case is the ranking of an applicant (Alkharusi, 2012). The inter-
cept obtained using OLS estimation will represent the mean of the group coded 0 on
all the dummy variables which, in our case, is the group of applicants whose ranking was
determined by research and teaching merits to an equal extent. The regression coefficients
will represent the deviation from the mean of this reference group for the other two groups. By adding the value of the regression coefficients to the intercept we will obtain the mean
ranking of the two groups of applicants whose rankings were determined by either research
merits or teaching merits. Table 2 shows the results of our OLS regression model using the two dummy variables. What we see is a comparison between the three different groups in terms of the average
ranking afforded by the reviewers, where 1 signifies the ranking afforded to the strong-
est candidate. As can be seen in the table, applicants whose ranking was determined by
research and teaching merits to an equal extent were on average afforded a rank of around
2.2. Hence, belonging to this group yields the best outcome, on average. Belonging to the
group of applicants whose ranking was determined by research merits yields the second-
best outcome, as the applicants in this group were afforded a rank of around 2.7, on aver-
age. Lastly, belonging to the group of applicants whose ranking was determined by teach-
ing merits yields the worst outcome, as these applicants were on average given a rank of
around 3.1. This confirms our third hypothesis, i.e. that candidates with stronger research
merits are prioritised before candidates with stronger teaching merits. 7 We have also constructed multivariate models with a number of control variables, such as the gender of
the reviewer and which university is appointing the position, but none of the controls change the results
reported here in any substantial way.
8 From looking at the R-squared values in table 3 we can also see that the explanatory powers of the mod-
els are quite low; they explain below or slightly above one percent of the variance in the ranking variable.
This is not surprising given that we study very few variables.
9 An alternative way to analyse our data, which would not rely on this assumption, is to perform a logistic
regression analysis which models the probability of a particular outcome, such as the probability of a candi-
date with certain characteristics of becoming ranked number 1. When we perform such an analysis we see
that for female candidates in our dataset the odds ratio of becoming ranked number 1 is 0.89 compared to
male candidates. If we look at the three disciplines separately, we see that for female economists the odds
ratio of becoming ranked number 1 is 0.61 compared to male economists, for female political scientists the
odds ratio is 0.62 compared to male political scientists, and for female sociologists the odds ratio of becom-
ing ranked number 1 is 1.49 compared to male sociologists. This confirms the general patterns we saw in
the OLS regressions. The Effects of Gender on the Ranking of Applicants So far, we have established that 1) female applicants are generally deemed to have stronger
teaching merits and weaker research merits relative to men (except for in Sociology where
this pattern was reversed), and 2) that applicants with stronger teaching merits and weaker
research merits are ranked lower than applicants with stronger research merits and weaker 1 “Teach more, but do not expect any applause”: Are Women Doubly… “Teach more, but do not expect any applause”: Are Women Doubly… Table 2 OLS Regression Model for Ranking Decision
OLS Regression on the effect of different merits on the ranking
***indicates significance on 99% level of significance. N=842
Group
Ranking of Applicant
(1 = highest rank)
[Regression Coef-
ficient]
Applicants whose ranking was determined by research and teaching merits to an
equal extent (reference group)
2.19***
Applicants whose ranking was determined by research merits
2.74***
[0.55]
Applicants whose ranking was determined by teaching merits
3.10***
[0.91] teaching merits, and lower still than applicants with equally strong research and teaching
merits. This would lead us to expect that it is also the case that 3) female applicants are
generally ranked lower than male applicants. We use linear regression analyses to study
whether this is the case. We construct a basic bivariate regression model (A) with the gender of the applicant as
independent variable, and the ranking afforded by the reviewers as dependent variable.7
The results are reported in Table 3, first for all the disciplines taken together and then bro-
ken down for each discipline. The constants in the table indicate the average ranking of a
male candidate (remember that 1 is the highest rank). The regression coefficients on the
‘Gender of applicant’-row indicate how female candidates are ranked compared to male
candidates; a positive number means that female candidates are ranked lower than male
candidates, and a negative number that they are ranked higher. The values of the coef-
ficients tell us how many steps above (if negative) or below (if positive) female candidates
are ranked compared to male candidates. Since we are studying a total population, we are not going to attach much importance to
statistical significance. On the other hand, we will be quite cautious in our interpretations
and focus mainly on the direction of the effects on rankings, i.e. The Effects of Gender on the Ranking of Applicants whether women are ranked
higher or lower than men, and not pay much attention to differences in ranking steps.8
Another reason for being cautious is that we have to make the somewhat controversial
assumption that there is equidistance between different scale steps in the rankings, i.e. that
the distance between rankings 1 and 2 is the same as between rankings 2 and 3, and so on.9 3 3 D. Brommesson et al. Table 3 OLS regressions on the
effect of gender on ranking
OLS regressions on the effect on ranking. Standard error within paren-
thesis
** indicates significance on 95% level of significance
A
(all)
A
(econ)
A
(pol sci)
A
(socio)
Cons
2.547
(.069)
2.348
(.104)
2.817
(.126)
2.468
(.123)
Gender of applicant
.151
(.112)
.462**
(.200)
.230
(.197)
-.226
(.183)
R-squared
.002
.019
.004
.006
Number of observations
842
273
314
255 Table 3 OLS regressions on the
effect of gender on ranking When we analyse our data, we do indeed identify a tendency that female candidates
are, on average, ranked lower than male candidates. In the first column of Table 3 we can
see that female candidates are ranked slightly lower than male candidates when we study
all three academic disciplines together. When the results are broken down by disciplines a
more complicated and interesting, pattern emerges. We can see that the general pattern of
female candidates being ranked lower than male candidates holds for both Political Science
and Economics, and this tendency seems to be strongest within Economics. For Sociology
however, the pattern is the reverse: female candidates are on average ranked higher than
male candidates. Once again, we wish to stress that these results need to be interpreted with caution. If
we look at the effects on scale steps, we see that for Political science and Sociology, there
seems to be only a rather modest effect in that being a female candidate will put you about
a fifth of a scale step below (for Political Science) or above (for Sociology) male candi-
dates. For Economics, the difference is almost half a scale step. From this we can at least
tentatively conclude that there are tendencies in our data that confirms H4, i.e. that male
applicants are ranked higher than female applicants in peer-review evaluations that precede
the hiring of senior lecturers. As we have seen, however, this is not true for Sociology. 1 3 The Effects of Gender on the Ranking of Applicants How can we understand the fact that Sociology seems to deviate from the general pat-
tern? Although it falls outside the aim of our study to explain the differences between
disciplines, the literature on tribalism and different conceptions of research still gives us
the opportunity to reflect on the differences. What we have discussed above as a narrower
conception of research, more often favored by men, and the ‘hard pure’ type of academic
discipline, could be expected to be closer related to Economics, while a wider conception
of research, more often favored by women, and the ‘soft pure’ type of discipline could be
expected to be closer to Sociology, with Political Science somewhere in-between. At least
tentatively, our results lend credibility to this argument, where the ‘softer’ discipline of
Sociology provides women with greater career opportunities, while a ‘harder’ discipline
such as Economics provides women with fewer opportunities.i In sum, our findings support the premise that has found solid support in previous
research, i.e. the stereotype ‘women teach, men think’. Women tend to have (relatively)
stronger teaching merits and men (relatively) stronger research merits. Moreover, we have
also identified a tendency that female candidates are, on average, ranked lower than male
candidates. We have emphasized that these results should be interpreted with caution, but
the differences are still there, which makes it possible for us to at least say that the men and
women we have studied here had different opportunities, to a varying degree between dif-
ferent disciplines. Conclusion Based on previous research, which has univocally found a gendered division of labour
in academia, we set out to study who is awarded tenured positions as senior lectures in
recruitment processes, and on what merits such awarding is based. Based on previous research, which has univocally found a gendered division of labour
in academia, we set out to study who is awarded tenured positions as senior lectures in
recruitment processes, and on what merits such awarding is based. Our premise was that there is a gendered division of labour within academia where – as
the European Commission (2008) phrased the stereotype – ‘women teach, men think’; i.e. men conduct relatively more research. This assumption was verified in our empirical find-
ings. The typical ‘male academic portfolio’ contains stronger research merits compared
with those of women, while the typical ‘female academic portfolio’ contained relatively
stronger teaching merits. However, we observed interesting differences between disci-
plines: Sociology displayed the reverse pattern compared with the hypothesized one. This
answered our first research question: To what extent are women and men ascribed different
types of merits in the evaluation process, and are differences consistent across disciplines?f f
Our second research question was: How often are different types of merits the determin-
ing factor in recruitment processes, and are there differences between disciplines in this
respect? We demonstrated that reviewers devote more space in their reports to evaluate
research merits. However, our findings showed that the effect on the ranking of candidates
is strongest when reviewers make an assessment of research and teaching merits as being
of equal strength. Still, the positive effect on the ranking of research merits is stronger
compared with the positive effect of teaching merits.f f
In order to reach the aim – to analyze whether the gender differences of division of
labour are reproduced in universities’ hiring processes – we have also studied the effect
the differences discussed here have on the ranking of applicants. Here we have concluded
that male applicants are ranked higher than female applicants in peer-review evaluations
that precede the hiring of senior lecturers. This is at least the case in Economics and, to a
lesser degree, in Political Science. As we have seen, however, this is not true for Sociology,
where the reverse pattern is observed.i Let us return to our hypotheses. 1 3 1 3 1 “Teach more, but do not expect any applause”: Are Women Doubly… Conclusion H1 and H2 are verified: Female applicants do have
stronger teaching merits compared with their male competitors and male applicants do
have stronger research merits compared with their female competitors. We can to some
extent also verify H3: candidates with stronger research merits are, relatively speaking,
rewarded more than candidates with stronger teaching merits. But it is important to note
that the determining factor with the strongest effect on rankings is the combination of
strong merits in both teaching and research. H4, that male applicants are ranked higher in
peer-review evaluations that precede the hiring of senior lecturers can also be – somewhat
cautiously – verified, at least for Political Science and Economics, but not for Sociology. i
We should also remember that it is not unreasonable to expect that women who invest
time and energy, in addition to risking one’s prestige, in applying for senior lectureships
self-select from a pool of the most motivated women. They may not, therefore, be repre-
sentative of female academics in general. Potential female applicants, with typical ‘female
academic portfolios’ – a heavy teaching load, more low-valued administrative work – may
anticipate that the kind of experiences and qualities they possess will be discriminated
against, and hence, they abstain from applying. Therefore, while our study on this self-
selected pool of motivated applicants already confirm that there is discrimination against
typical ‘female academic portfolios’ – i.e. strong track-records in teaching and administra-
tive work, we can expect this effect to be even stronger in the full population of female
university teachers. 1 3 3 D. Brommesson et al. Lastly, a brief afterthought on the implications of our findings. We have seen how the
most competitive women tend to have typical ‘male academic portfolios’, i.e. they have
strong research as well as teaching merits; and that teaching – for all intents and purposes
– is downgraded in hiring processes. This result mirrors what Parker (2008) found study-
ing promotion processes in the UK. In the wake of the ‘massification’ of higher educa-
tion, this is as depressing as it is truly puzzling. In Sweden, as well as in the UK and sev-
eral other countries, there has been – at a rhetorical level, at least – a strong movement
towards strengthening rewards for teaching excellence. Conclusion Governments have urged universi-
ties to improve the status of teaching and it has become de facto mandatory for academics
to complete teacher training programmes to get promoted or hired. However, given our
results – and as Parker (2008) concluded more than a decade ago – despite widespread
public endorsements, the value of teaching is not reflected in reward-structures in higher
education; and much points to the fact that, for the majority of female academics, this fact
ultimately puts them at a disadvantage when it comes to hiring, promotions and salary
development. Funding Open access funding provided by Linnaeus University. This study was funded by the Institute for
Evaluation of Labour Market and Education Policy, Uppsala, Sweden. Funding Open access funding provided by Linnaeus University. This study was funded by the Institute for
Evaluation of Labour Market and Education Policy, Uppsala, Sweden. Data Availability Data available through the corresponding author. Data Availability Data available through the corresponding author. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Com-
mons licence, and indicate if changes were made. The images or other third party material in this article
are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly
from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. References Occupational Stress, Burnout and Job Status in Female Academics. Gen-
der Work and Organization, 5(2), 67–82. Equal Measures (2019). Harnessing the Power of Data for Gender Equality. Surrey: Equal Measures. Equal Measures (2019). Harnessing the Power of Data for Gender Equality. Surrey: Equal Measures. European Commission. (2008). Mapping the Maze: Getting more women to the top in research. Euro-
pean Commission p
Fairweather, J. S. (2005). Beyond the Rhetoric: Trends in the Relative Value of Teaching and Research
in Faculty Salaries. The Journal of Higher Education, 76(4), 401–422. Filandra, M., Pasqua, S. (2019). ‘Being Good isn’t Good Enough’: Gender Discrimination in Italian aca-
demia. Studies in Higher Education, early view online: https://www.tandfonline.com/doi/full/10.
1080/03075079.2019.1693990 Global Gender Gap Report. (2018). “The Global Gender Report 2018.” Cologne/Geneva: World Eco-
nomic Forum. Guardino, C., & Borden, V. (2017). Faculty Service Loads and Gender: Are Women Taking Care of the
Academic Family? Research in Higher Education, 58(6), 672–694. Haley, A. (2018). Returning to rural origins after higher education: gendered social space. Journal of
Education and Work, 31(4), 418–432. Healey, R., & Davies, C. (2019). Conceptions of ‘research’ and their gendered impact on research activ-
ity: a UK case study. Higher Education Research & Development, 38(7), 1386–1400. Kalm, S. (2019). Om akademiskt hushållsarbete och dess fördelning. Sociologisk Forskning, 56(1),
5–26. Kulp, A. M. (2020). Parenting on the Path to the Professoriate: A Focus on Graduate Student Mothe
Research in Higher Education, 61, 408–429. King, M. M., Bergstrom, C. T., Correll, S. J. Jacquet, J., West, J. D. (2017). Men Set Their Own Cites High:
Gender and Self-citation Across Fields and Over Time. Socius: Sociological Research for a Dynamic
World, 3: 1–22. Larrán George, M., Javier Andrades, F., & Gomez Cana, M. C. (2016). Gender Differences Between
Faculty Members in Higher Education. Educational Research Review, 18, 58–69. Lundby, K. L. P., & Warme, B. D. (1990). Gender and career trajectory: The case of part-time facu
Studies in Higher Education, 15(2), 207–222. g
MacNell, L., Driscoll, A., & Hunt, A. N. (2015). What’s in a Name: Exposing Gender Bias in Student
Ratings of Teaching. Innovation in Higher Education, 40, 291–303.fi Madison, G., Fahlman, P. (2020). Sex differences in the number of scientific publications and citations
when attaining the rank of professor in Sweden. Studies in Higher Education, published online 06
Feb 2020. https://doi.org/10.1080/03075079.2020.1723533 McFarland, J., Hussar, B., de Brey, C., Snyder, T., Wang, X. Wilkinson-Flicker, S., Gebrekristos. S. References Alkharusi, H. (2012). Categorical Variables in Regression Analysis: A Comparison of Dummy and Effect
Coding. International Journal of Education, 4(2) 202–210. Allbright. (2019). Vetenskapsmannen inte kvinna. Stockholm: Allbright foundation. ). Vetenskapsmannen inte kvinna. Stockholm: Allbrig Angervall, P., & Beach, D. (2017). Dividing Academic Work: Gender and Academic Career at Swedish
Universities. Gender and Education, 32(3), 347–362. https://doi.org/10.1080/09540253.2017.1401047 Angervall, P., & Beach, D. (2018). The Exploitation of Academic Work: Women in Teaching at Swedish
Universities. Higher Education Policy, 31(1), 1–17. Angervall, P., Beach, D., & Gustafsson, J. (2015). The Unacknowledged Value of Female Academic Labour
Power for Male Research Careers. Higher Education Research and Development, 34(5), 815–827. August, L., & Waltman, J. (2004). Culture, Climate, and Contribution: Career Satisfaction Among Female
Faculty. Research in Higher Education, 45, 177–192.f y
g
Barbezat, D. A., & Hughes, J. W. (2005). Salary Structure Effects and the Gender Pay Gap in Academ
Research in Higher Education, 46, 621–640. Becher, T. (1989). Academic Tribes and Territories. Open University Press.if Becher, T. (1994). The Significance of Disciplinary Differences. Studies in Higher Education, 19(2),
151–161.f ggren, C. (2008). Horizontal and Vertical Differentiation within Higher Education – Gender and
Class Perspectives. Higher Education Quarterly, 62, 20–39. Berggren, C. (2008). Horizontal and Vertical Differentiation within Higher Education Gender and
Class Perspectives. Higher Education Quarterly, 62, 20–39. B
C (2011) G
d
li
li i
d hi h
d
i
J
l
f Ed
i
d Class Perspectives. Higher Education Quarterly, 62, 20–39. Berggren, C. (2011). Gender equality policies and higher education careers. Journal of Education and ggren, C. (2011). Gender equality policies and higher education careers. Journal of Education and
Work, 24(1–2), 141–161. 1 3 “Teach more, but do not expect any applause”: Are Women Doubly… Biglan, A. (1973). The Characteristics of Subject Matter in Different Scientific Areas. Journal of
Applied Psychology, 57, 195–203. Carr, P. L., Gunn, C. M., Kaplan, S. A., Raj, A., & Freund, K. M. (2015). Inadequate Progress for
Women in Academic Medicine: Findings From the National Faculty Study. Journal of Women’s
Health, 24(3), 190–199. ate, K., & Kandiko Howson, C. (2016). Indicators of Esteem: Gender and Prestige in Academic Work. British Journal of Sociology of Education, 37(4), 567–585. Dion, M., Lawrence Sumner, J., & McLaughlin Mithcell, S. (2018). Gendered Citation Patterns across
Political Science and Social Science Methodology Fields. Political Analysis, 26(3), 312–327. Doyle, C., & Hind, P. (2002). References (2017). The Condition of Education 2017. NCES 2017–144. National Center for Education
Statistics. Marschke, R., Laursen, S., Nielsen, J. M., & Rankin, P. (2007). Demographic Inertia Revisited: An
Immodest Proposal to Achieve Equitable Gender Representation Among Faculty in Higher Educa-
tion. Journal of Higher Education, 78, 1–26. Nature. (2016). “Men Cite Themselves More Than Women Do.” July 16. https://www.nature.com/news/
men-cite-themselves-more-than-women-do-1.20176 Nature. (2017). “Teaching Load Could put Female Scientists at Career Disadvantage.” April 13. https://
www.nature.com/news/teaching-load-could-put-female-scientists-at-career-disadvantage-1.21839f Neumann, R. (2001). Disciplinary Differences and University Teaching. Studies in Higher Educati
26(2), 135–146.f Okbara, J. O., Squillace, M., & Erondu, E. A. (2005). Gender Differences and Job Satisfaction: a Study of
University Teachers in the United States. Women in Management Review, 20(3), 177–190.f O’Laughlin, E. M., & Bischoff, L. G. (2005). Balancing Parenthood and Academia Work/Family Stress as
Influenced by Gender and Tenure Status. Journal of Family Issues, 26, 79–106. 1 3 D. Brommesson et al. Parker, J. (2008). Comparing Research and Teaching in University Promotion Criteria. Higher Education
Quarterly, 62(3), 237–251.f Perna, L. W. (2001). Sex and Race Differences in Faculty Tenure and Promotion. Research in Higher Edu-
cation, 42(5), 541–567.f Perna, L. W. (2005). Sex Differences in Faculty Tenure and Promotion: The Contribution of Family Ties. Research in Higher Education, 46, 277–307. Renwick Monroe, K., Choi, J., Howell, E. Lampros-Monroe, C. Trejo, C., Perez, V. (2014). Gender Equality
in the Ivory Tower, and How Best to Achieve It. PS: Political Science & Politics, 47(2), 419. Sax, J. J., Hagedorn, L. S., Arredondo, M., & Dicrisi, F. A. (2002). Faculty Research Productivity: Explor-
ing the Role of Gender and Family-Related Factors. Research in Higher Education, 43, 423–446. g
y
g
iebinger, L., & Gilmartin, S. K. (2010). Housework is an Academic Issue. Academe, 96, 39–44. Tamblyn, R., Girard, N., Qian, C. J., & Hanley, J. (2018). Assessment of Potential Bias in Research Grant Peer
Review in Canada. CMAJ, 190(16), 489–499. Tessens, L., White, K., & Web, C. (2011). Senior Women in Higher Education Institutions: Perceived
Development Needs and Support. Journal of Higher Education Policy and Management, 33, 653–665. Times Higher Education. (2019). Male Economists More Likely to Accept Papers by Other Men: Gap in Acceptance
Rates for Major European Conferences Entirely Down to Behaviour of Male Referees. Times Higher Educa-
tion. https://www.timeshighereducation.com/news/male-economists-more-likely-accept-papers-other-men p
g
y
p p p
Toutkoushian, R. K. (1999). Parker, J. (2008). Comparing Research and Teaching in University Promotion Criteria. Higher Education
Quarterly, 62(3), 237–251.f References The Status of Academic Women in the 1990s No Longer Outsiders, But Not
Yet Equals. The Quarterly Review of Economics and Finance, 39(5), 679–698. Toutkoushian, R. K., & Conley, V. M. (2005). Progress for Women in Academe, Yet Inequities Persist. Research in Higher Education, 46(1), 1–28. UHR. (2020). Jämställdhet i högskolan – ska den nu ordnas en gång för alla? https://www.uhr.se/globalassets/_
uhr.se/lika-mojligheter/jamstalldhetsdelegationen/uhr-jamstalldhet-i-hogskolan.pdf (access February 18,
2021). Voeten, E. (2013). Introducing the Monkey Cage Gender Gap Symposium. Washington Post. https://www.
washingtonpost.com/news/monkey-cage/wp/2013/09/30/introducing-the-monkey-cage-gender-gap-
symposium/ Wullum Nielsen, M. (2016). Gender inequality and research performance: moving beyond individual-meri-
tocratic explanations of academic advancement. Studies in Higher Education, 41(11). Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations. 1 3
|
https://openalex.org/W2950596902
|
https://f1000research.com/articles/7-497/v1/pdf
|
English
| null |
TDG regulates cell cycle progression in human neural progenitors
|
F1000Research
| 2,018
|
cc-by
| 11,755
|
RESEARCH ARTICLE
TDG regulates cell cycle progression in human neural
progenitors [version 1; peer review: 2 approved with
reservations] Igal Germanguz1, Jenny C. Park2, Jessica Cinkornpumin1, Aryeh Solomon1,
Minori Ohashi1-3, William E. Lowry
1-3 1Eli and Edythe Broad Center for Regenerative Medicine, University of California, Los Angeles, Los Angeles, CA, 90095, USA
2Department of Molecular Cell and Developmental Biology, University of California, Los Angeles, Los Angeles, CA, 90095, USA
3Molecular Biology Institute, University of California, Los Angeles, Los Angeles, CA, 90095, USA Open Peer Review
Approval Status
1
2
version 1
26 Apr 2018
view
view
Alexey Ruzov
, University of Nottingham,
Nottingham, UK
Abdulkadir Abakir
, University of
Nottingham, Nottingham, UK
1. Rahul Prasad
, Fox Chase Cancer Center,
Philadelphia, USA
Timothy Yen, Fox Chase Cancer Center,
Philadelphia, USA
Alfonso Bellacosa
, Fox Chase Cancer
Center, Philadelphia, USA
2. Any reports and responses or comments on the
article can be found at the end of the article. First published: 26 Apr 2018, 7:497
https://doi.org/10.12688/f1000research.13801.1
Latest published: 26 Apr 2018, 7:497
https://doi.org/10.12688/f1000research.13801.1
v1 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 Abstract bst act
Background: As cells divide, they must both replicate their DNA and
generate a new set of histone proteins. The newly synthesized
daughter strands and histones are unmodified, and must therefore be
covalently modified to allow for transmission of important epigenetic
marks to daughter cells. Human pluripotent stem cells (hPSCs) display
a unique cell cycle profile, and control of the cell cycle is known to be
critical for their proper differentiation and survival. A major
unresolved question is how hPSCs regulate their DNA methylation
status through the cell cycle, namely how passive and active
demethylation work to maintain a stable genome. Thymine-DNA
glycosylase (TDG), an embryonic essential gene, has been recently
implicated as a major enzyme involved in demethylation. Methods: We use human pluripotent stem cells and their derivatives
to investigate the role of TDG in differentiation and proliferation. To
perform loss of function of TDG, RNA Interference was used. To study
the cell cyle, we engineered human pluripotent stem cells to express
the FUCCI tool which marks cells at various stages of the cell cycle with
distinct patterns of fluorescent proteins. We also used cell cycle
profiling by FACS, and DNA methylation analysis to probe a
connection between DNA demethylation and cell cycle. l
d
h
i
h
l
ll
l Background: As cells divide, they must both replicate their DNA and
generate a new set of histone proteins. The newly synthesized
daughter strands and histones are unmodified, and must therefore be
covalently modified to allow for transmission of important epigenetic
marks to daughter cells. Human pluripotent stem cells (hPSCs) display
a unique cell cycle profile, and control of the cell cycle is known to be
critical for their proper differentiation and survival. A major
unresolved question is how hPSCs regulate their DNA methylation
status through the cell cycle, namely how passive and active
demethylation work to maintain a stable genome. Thymine-DNA
glycosylase (TDG), an embryonic essential gene, has been recently
implicated as a major enzyme involved in demethylation. Methods: We use human pluripotent stem cells and their derivatives
to investigate the role of TDG in differentiation and proliferation. To
perform loss of function of TDG, RNA Interference was used. To study
the cell cyle, we engineered human pluripotent stem cells to express
the FUCCI tool which marks cells at various stages of the cell cycle with
distinct patterns of fluorescent proteins. Abstract Lowry (blowry@ucla.edu)
Author roles: Germanguz I: Conceptualization, Formal Analysis, Investigation, Methodology, Writing – Review & Editing; Park JC:
Investigation; Cinkornpumin J: Investigation; Solomon A: Investigation; Ohashi M: Investigation; Lowry WE: Conceptualization, Formal
Analysis, Funding Acquisition, Investigation, Project Administration, Resources, Supervision, Writing – Original Draft Preparation, Writing
– Review & Editing
Competing interests: No competing interests were disclosed. Grant information: This work was supported by NIH (P01GM9913), Allen Distinguished Investigator award from the Allen Frontiers
Group to WEL, A CIRM Basic Biology Award (RT-2), and pilot support from the BSCRC at UCLA (Rose Hills Scholar Award). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Copyright: © 2018 Germanguz I et al. This is an open access article distributed under the terms of the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Data
associated with the article are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public
domain dedication). How to cite this article: Germanguz I, Park JC, Cinkornpumin J et al. TDG regulates cell cycle progression in human neural
progenitors [version 1; peer review: 2 approved with reservations] F1000Research 2018, 7:497
https://doi.org/10.12688/f1000research.13801.1
First published: 26 Apr 2018, 7:497 https://doi.org/10.12688/f1000research.13801.1 Abstract We also used cell cycle
profiling by FACS, and DNA methylation analysis to probe a
connection between DNA demethylation and cell cycle. Results: Here we present data showing that TDG regulates cell cycle
dynamics in human neural progenitors (NPCs) derived from hPSCs,
leading to changes in cell cycle related gene expression and neural
differentiation capacity. These data show that loss of TDG function
can block differentiation by driving proliferation of neural progenitors. We also identify specific cell cycle related genes whose expression
changes upon loss of TDG expression. Conclusions: These observations suggest that TDG and active
demethylation play an important role in hPSC cell cycle regulation and Page 1 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 differentiation. Keywords
TDG, DNA methylation, human neural progenitors Corresponding author: William E. Lowry (blowry@ucla.edu)
Author roles: Germanguz I: Conceptualization, Formal Analysis, Investigation, Methodology, Writing – Review & Editing; Park JC:
Investigation; Cinkornpumin J: Investigation; Solomon A: Investigation; Ohashi M: Investigation; Lowry WE: Conceptualization, Formal
Analysis, Funding Acquisition, Investigation, Project Administration, Resources, Supervision, Writing – Original Draft Preparation, Writing
– Review & Editing
Competing interests: No competing interests were disclosed. Grant information: This work was supported by NIH (P01GM9913), Allen Distinguished Investigator award from the Allen Frontiers
Group to WEL, A CIRM Basic Biology Award (RT-2), and pilot support from the BSCRC at UCLA (Rose Hills Scholar Award). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Copyright: © 2018 Germanguz I et al. This is an open access article distributed under the terms of the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Data
associated with the article are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public
domain dedication). How to cite this article: Germanguz I, Park JC, Cinkornpumin J et al. TDG regulates cell cycle progression in human neural
progenitors [version 1; peer review: 2 approved with reservations] F1000Research 2018, 7:497
https://doi.org/10.12688/f1000research.13801.1
First published: 26 Apr 2018, 7:497 https://doi.org/10.12688/f1000research.13801.1 Corresponding author: William E. Introduction Because of this and previous data suggesting TDG could poten-
tially regulate the cell cycle, we stained neural progenitors
made from human pluripotent stem cells or derived from tissue
for TDG and Ki67. The cell cycle phase for this assay was
determined based on a previously reported KI67 staining
pattern within the nucleus11. TDG was also previously reported
to be tightly regulated during the progression of the cell cycle
as its level is rapidly downregulated by ubiquitination in the
S phase of the cell cycle in cellular models such as HeLa and
fibroblasts and re-expressed in G212. Here, we found a similar
result, namely that TDG protein levels appear to correlate
with G0/G1stages of the cell cycle (Figure 1C and D). We and others have shown that standard differentiation pro-
tocols of hPSC leads to derivation of an embryonic-like cell
rather than a mature, postnatal-like cell10. We previously identi-
fied a group of embryonic related genes which are differentially
expressed in the PSC progeny of all three lineages and in tissues
of the early gestation period rather than in their respec-
tive cell types of later developmental stages. Among these
genes, we identified Thymine DNA Glycosylase (TDG), a gene
that was recently implicated in active DNA demethylation10. Unlike other glycosylases, TDG is essential for embryonic via-
bility as TDG null embryos die around E11.5-12.5 of internal
hemorrhage8. The lethal phenotype was also associated with
aberrant promoter methylation and imbalanced histone modifica-
tions. In addition there is evidence that levels of this enzyme are
linked to progression through specific cell cycle stages. Here we
provide evidence that that TDG regulates cell cycle related gene
expression in human neural progenitors (NPCs) derived from
hPSCs and controls their capacity for differentiation towards
neurons and glia. These observations suggest that TDG and
active demethylation play an important role in hPSC cell cycle
regulation and differentiation. Silencing TDG by siRNA To investigate the role of TDG in early human development,
we used siRNA mediated knock-down (KD) of TDG in neural
progenitor cells (NPCs) derived from hPSC (Figure 2A and B). To determine whether TDG KD led to expected changes in
5-carboxylcytosine (5caC) and 5-formylcytosine (5fC) DNA
residues, immunostaining for these markers was performed. As
expected, silencing TDG led to an increased intensity of 5caC
and 5fC DNA, with no change in 5-Hydroxymethylcytosine
(5hmC) DNA (Figure 2C). Introduction but were highly expressed in tissue derived cells. TDG was
expressed significantly higher in neural progenitors gener-
ated from pluripotent stem cells as opposed to the same cell
type derived from fetal brain (Figure 1A, shown in Log2 scale). but were highly expressed in tissue derived cells. TDG was
expressed significantly higher in neural progenitors gener-
ated from pluripotent stem cells as opposed to the same cell
type derived from fetal brain (Figure 1A, shown in Log2 scale). Coordinated changes to the epigenome are known to be essen-
tial for lineage specification and maintenance of cellular
identity. DNA methylation and histone modifications critically
contribute to epigenetic maintenance of chromatin structures and
gene expression programs. DNA methylation can silence genomic
regions, directly or indirectly, and play an important role dur-
ing mammalian development. Loss of methylation in specific
locations is associated with differentiation towards specific germ
layers as binding of several transcription factors is strongly
associated with specific loss of DNA methylation in one germ
layer, and in many cases a reciprocal gain in the other layers. However, the mechanism for the lineage related site specific
demethylation is not currently known. A major open question is
whether this is the result of an active or passive demethylation
after repeated cell division1,2. Promoters with low CpG content are
more likely to be methylated in human embryonic stem cells
(ESCs). Conversely, these same promoters are demethyl-
ated and actively expressed during differentiation in a cell-
type-specific manner3–5. Demethylation can occur by a passive
mechanism in which the normal function of DNMT1/UHRF1 is
insufficient or disrupted2,6. Alternatively, evidence for the exist-
ence of an active mechanism in which the cytosine modifications
are enzymatically removed is accumulating1,7–9. Which of these
mechanisms is responsible for demethylation changes in early
human development is not currently known. The Allen Brain Atlas created by the Allen Institute provides
gene expression data from various brain regions across both
development and through adulthood. As shown in Figure 1B,
TDG is expressed most highly in the brain in utero, and then
falls after birth and stays low throughout adulthood. The same
was true for the Ten–Eleven Translocation (TET) family of
dioxigenases7, suggesting that DNA demethylation is primarily
performed in utero. It is also possible that DNA demethyla-
tion by TDG and TETs function is linked to proliferation, which
is known to decrease at birth relative to that found in utero. Corresponding author: William E. Lowry (blowry@ucla.edu) Author roles: Germanguz I: Conceptualization, Formal Analysis, Investigation, Methodology, Writing – Review & Editing; Park JC:
Investigation; Cinkornpumin J: Investigation; Solomon A: Investigation; Ohashi M: Investigation; Lowry WE: Conceptualization, Formal
Analysis, Funding Acquisition, Investigation, Project Administration, Resources, Supervision, Writing – Original Draft Preparation, Writing
– Review & Editing Author roles: Germanguz I: Conceptualization, Formal Analysis, Investigation, Methodology, Writing – Review & Editing; Park JC:
Investigation; Cinkornpumin J: Investigation; Solomon A: Investigation; Ohashi M: Investigation; Lowry WE: Conceptualization, Formal
Analysis, Funding Acquisition, Investigation, Project Administration, Resources, Supervision, Writing – Original Draft Preparation, Writing
– Review & Editing
Competing interests: No competing interests were disclosed. Grant information: This work was supported by NIH (P01GM9913), Allen Distinguished Investigator award from the Allen Frontiers
Group to WEL, A CIRM Basic Biology Award (RT-2), and pilot support from the BSCRC at UCLA (Rose Hills Scholar Award). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Copyright: © 2018 Germanguz I et al. This is an open access article distributed under the terms of the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Data
associated with the article are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public
domain dedication). Page 2 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 Introduction RT-PCR for genes typical of the NPC state showed essen-
tially no change in TDG KD cells (Figure 2D). Though
downregulating TDG levels showed no influence on NPC
identity, we further tested whether lower TDG levels affects dif-
ferentiation. Four days post TDG KD treatment, NPCs were
induced to further differentiate using the growth factor withdrawal
method (removal of self-renewal supporting growth factors EGF,
bFGF). Three weeks after induction of differentiation, we ana-
lyzed the percentage of MAP2/GFAP positive cells, which rep-
resent the differentiation expectancy towards the neuronal/glial
lineage respectfully. We found that though the neural/
glial ratio remained similar, the total differentiated cell
percentage was lower than in control (Figure 2E), indicating
a failure to properly differentiate upon silencing of TDG. Typi-
cally, such a differentiation block would be due to aberrant
differentiation or due to prolonged proliferative stimulus. Results Expression of DNA demethylases through neural
development Expression of DNA demethylases through neural
development We originally identified TDG in a screen for genes that were
consistently differently expressed between human pluripotent
derivatives and their in vivo counterparts10. This screen identified
a number of genes that were persistently expressed in pluripotent
derivatives, and therefore suggestive of an early embryonic
state, or genes that failed to be induced in pluripotent derivatives We also looked for gene expression changes following TDG-KD
in NPCs by RNA-SEQ. 355 genes were differentially expressed Page 3 of 19 Page 3 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 G is expressed during early human neuronal development. A, Average expression of the TDG microarray probe in
NPCs and neurons and tissue derived NPCs, adapted from 10. Data presented are the average of at least three indep
normalized by log2 according to 13. B, RNA-seq analyses for the expression of TDG, TET1, 2 and 3 adopted from th
inspan developmental transcriptome database displayed as log-scale reads per kilobase measured (log2 RPKM) acr
uman brain. C, Immunofluorescent (IF) staining for KI67 and TDG expression in NPCs derived from human pluripotent ste
or human brain tissue (Tissue-NPCs). D, Magnified images (from panel C) of representative nuclei for different cell cycle
taken at 20X magnification. Pag Figure 1. TDG is expressed during early human neuronal development. A, Average expression of the TDG microarray probe in hPCS,
hPSC derived NPCs and neurons and tissue derived NPCs, adapted from 10. Data presented are the average of at least three independent
samples and normalized by log2 according to 13. B, RNA-seq analyses for the expression of TDG, TET1, 2 and 3 adopted from the Allen
Institute’s Brainspan developmental transcriptome database displayed as log-scale reads per kilobase measured (log2 RPKM) across the
developing human brain. C, Immunofluorescent (IF) staining for KI67 and TDG expression in NPCs derived from human pluripotent stem cells
(PSC-NPCs) or human brain tissue (Tissue-NPCs). D, Magnified images (from panel C) of representative nuclei for different cell cycle stages. Images were taken at 20X magnification. Page 4 of 19 Figure 1. TDG is expressed during early human neuronal development. A, Average expression of the TDG microarray probe in hPCS,
hPSC derived NPCs and neurons and tissue derived NPCs, adapted from 10. Data presented are the average of at least three independent
samples and normalized by log2 according to 13. Expression of DNA demethylases through neural
development B, RNA-seq analyses for the expression of TDG, TET1, 2 and 3 adopted from the Allen
Institute’s Brainspan developmental transcriptome database displayed as log-scale reads per kilobase measured (log2 RPKM) across the
developing human brain. C, Immunofluorescent (IF) staining for KI67 and TDG expression in NPCs derived from human pluripotent stem cells
(PSC-NPCs) or human brain tissue (Tissue-NPCs). D, Magnified images (from panel C) of representative nuclei for different cell cycle stages. Images were taken at 20X magnification. F1000Research 2018, 7:497 Last updated: 31 MAR 2022 gure 2. TDG downregulation in NPCs affects differentiation capacity. A, TDG protein expression level at day 4 post siRNA transfec
easured by Western blot (top). B, Immunostaining for TDG in NPCs following siRNA transfection, and quantification (right). C,
presentative immunofluorescence of DNA methylation modifications. Bars represent 50µm. Bottom: ImageJ quantification of TDG
rmalized to CONT-KD. Quantification was performed for over 100 nuclei across at least 5 images. Error bars represent standard error o
ean normalized to CONT-KD. D, Expression levels of NPC markers measured by qRT-PCR, normalized against the relative levels of GA
d compared to CONT-KD, error bars represent standard error of the mean of 3 knockdown experiments. E, 4 Days post siRNA transfec
Cs were induced to terminally differentiate by growth factor withdrawal (GFWD). Left: representative IF of 3 Weeks neural differentia
ciency measured by the ratio of percentage of MAP2 (neuron)/GFAP (glia). Bars represent 50µm. Right: quantification of n = 3 from at
ee separate knockdown/differentiation experiments. p-values were calculated with Student’s t test: * = p<0.05, ns=not significant. Page 5 Figure 2. TDG downregulation in NPCs affects differentiation capacity. A, TDG protein expression level at day 4 post siRNA transfection,
measured by Western blot (top). B, Immunostaining for TDG in NPCs following siRNA transfection, and quantification (right). C, Top:
Representative immunofluorescence of DNA methylation modifications. Bars represent 50µm. Bottom: ImageJ quantification of TDG-KD
normalized to CONT-KD. Quantification was performed for over 100 nuclei across at least 5 images. Error bars represent standard error of the
mean normalized to CONT-KD. D, Expression levels of NPC markers measured by qRT-PCR, normalized against the relative levels of GAPDH
and compared to CONT-KD, error bars represent standard error of the mean of 3 knockdown experiments. E, 4 Days post siRNA transfection,
NPCs were induced to terminally differentiate by growth factor withdrawal (GFWD). Left: representative IF of 3 Weeks neural differentiation. Expression of DNA demethylases through neural
development Efficiency measured by the ratio of percentage of MAP2 (neuron)/GFAP (glia). Bars represent 50µm. Right: quantification of n = 3 from at least
three separate knockdown/differentiation experiments. p-values were calculated with Student’s t test: * = p<0.05, ns=not significant. Page 5 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 by 1.5 fold across 3 independent experiments (Figure 3A). Using
the DAVID annotation tool14, we classified those genes into func-
tional groups (Figure 3B). Of the most significantly enriched
functional annotations identified, we found 34 cell cycle related
genes. Among those, genes which are major players in mito-
sis, CDK1, CDK10, SKP2 were upregulated. In contrast, other
genes like CDC25B and CDKN1C (p57), which are inhibitors
of cell cycle progression, were downregulated (Figure 3C). by 1.5 fold across 3 independent experiments (Figure 3A). Using
the DAVID annotation tool14, we classified those genes into func-
tional groups (Figure 3B). Of the most significantly enriched
functional annotations identified, we found 34 cell cycle related
genes. Among those, genes which are major players in mito-
sis, CDK1, CDK10, SKP2 were upregulated. In contrast, other
genes like CDC25B and CDKN1C (p57), which are inhibitors
of cell cycle progression, were downregulated (Figure 3C). that the DNA glycosylation activity of TDG is used for other
substrates besides methylated cytosine, for instance in DNA
repair. It is also possible that another domain of TDG regu-
lates cell cycle progression by another unknown mechanism. To
attempt to link DNA demethylation by TDG to regulation of the
cell cycle, we needed a system that could allow for simul-
taneous labeling of DNA demethylation intermediates and
cell cycle markers. We generated hESCs which expressed the
Fluorescence Ubiquitination Cell Cycle Indicator (FUCCI)
transgene reporter system by lentiviral transduction16. In this
system, cells which are in the G1 stage express Ctd1, which is
conjugated to mCherry, while cells in S/G2 expressed Gemi-
nin which is conjugated to visible green protein mVENUS. Cells entering the DNA replication stage at the end of G1
express both markers and emit yellow light (Figure 6A). We
first verified that the level of TDG was tightly regulated dur-
ing cell cycle progression in hPSC as in other reported cell
systems since hPSC display a unique cell cycle pattern. We
found high levels of TDG in early G1 which are downregulated
with cell cycle progression (Figure 6B). Expression of DNA demethylases through neural
development Interestingly, we found
that 5caC and 5fC were both induced in early S phase cells,
while 5hmC was reduced (Figure 6B) as was reported before. This indicates that the global state of DNA demethylation is
tightly correlated to progression of the cell cycle. Furthermore,
all these data on the effect of TDG on cell cycle serve to
explain why silencing of TDG led to defective neuronal
and glial specification in NPCs (Figure 2D). To validate that silencing of TDG by siRNA led to changes
in DNA demethylation, we performed Methylase-assisted
bisulfite conversion PCR (MAB-PCR) to probe for the pres-
ence of the 5mC and 5hmC in a gene whose expression changed
upon siRNA-mediated knockdown of TDG. MAB-PCR takes
advantage of an enzyme and bisulfite-conversion sequencing
to identify the relative abundance of 5caC and 5fC nucle-
otides (Figure 4A). This allows for a measure of TDG activity,
as TDG is known to use its glycosylase activity to finish the
demethylation process to convert 5caC and 5fC to the fully
demethylated state. To determine whether TDG activity can
regulate the methylation and gene expression, we looked
specifically at a gene whose expression was affected in TDG-KD
cells. EGR1, the early growth response gene is known to be
dynamically regulated by a variety of mechanisms, includ-
ing DNA Methylation at an upstream CpG island (Figure 4B)15. We analyzed a segment of a CpG island upstream to the EGR1
transcription start site (TSS site). This locus was chosen as
it was reported to have the highest distribution of 5fC, 5caC
around the TSS. Dataset 1. TDG regulates cell cycle progression in human neural
progenitors
http://dx.doi.org/10.5256/f1000research.13801.d201379
Complete dataset of all underlying data divided into folders based
on relevant figures Dataset 1. TDG regulates cell cycle progression in human neural
progenitors This analysis showed that silencing of TDG by siRNA led
to a dramatic accumulation of 5caC and 5fC in a CpG island
directly upstream of the start site of EGR1 transcription
(Figure 4C and D). This experiment provided evidence that
TDG not only regulates DNA demethylation, but also that this
can influence gene expression. The proportion of genes differen-
tially regulated by TDG-mediated DNA demethylation remains
unclear until a genome-wide analysis can be performed. http://dx.doi.org/10.5256/f1000research.13801.d201379 Complete dataset of all underlying data divided into folders based
on relevant figures Complete dataset of all underlying data divided into folders based
on relevant figures Discussion The data presented here confirm and extends previous find-
ings that TDG and DNA demethylation can play a role in proper
progression through the cell cycle. These results could be
particularly relevant for the nervous system, where we provide
evidence that TDG and TET mediated demethylation appears
to diminish across development. This correlates with both
proliferative rate and TDG expression, and could have impor-
tant consequences to the rate of developmental progression. The
big question remaining from this work is how DNA demeth-
ylation plays a role in progression of the cell cycle. When
DNA is replicated it is thought that the new daughter strand
is methylated according to the hemi-methylation pattern on
the sister strand by maintenance DNA Methyltransferases
(DNMT1). Less clear is what happens to portions of the genome
that are hemi-methylated hydroxymethylated nucleotides. The
change in proportions of 5caC and 5fC across the cell cycle
could indicate that these modified nucleotides are simply
erased through the action of TET and TDG enzymes, and then
re-written. In this scenario, it is interesting that blocking TDG
appeared to promote the cell cycle rate, and could suggest that
demethylation is a rate limiting step in cell cycle progression to
ensure proper methylation of DNA in both daughter cells. We further tested whether TDG-KD in NPCs affects entrance
into the cell cycle by Ki67 staining, and found that down-
regulating TDG resulted in a higher percentage of proliferat-
ing cells when this enzyme was knocked down (Figure 5A). Co-staining of Ki67 with TDG showed that TDG is downregu-
lated with cell cycle progression, as higher TDG expression is
observed in G0/early G1 cells, and downregulated with cell
cycle progression (Figure 2D). We also measured cell cycle
dynamics by Flow Cytometry (FACS) upon TDG silencing. This high-throughput method allowed for an accurate deter-
mination of the effect of siRNA on TDG, and showed that the
proportion of cells in S phase was significantly decreased,
while the proportion in G2/M was increased (Figure 5C). Taken together, it seems clear that TDG plays a role in human
pluripotent stem cell cycle regulation. Development of FUCCI model for cell cycle analyses Despite all the analyses above, it was not clear whether silenc-
ing of TDG affects the cell cycle through its ability to regulate
the terminal step of DNA demethylation. It is formally possible Page 6 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 Figure 3. TDG-KD NPC differential gene expression analyzed by RNA-seq. A, Differential gene expression of n = 3 siRNA knockdown
experiments. Scatter plot of the group average FPKM (log2) for all genes mapped above the background cutoff, differentially expressed
genes (over 1.5 fold change; p<0.05) are highlighted in red and green. B, Functional annotation of differentially expressed genes shows
significant change in genes related to cell cycle, regulation of apoptosis and structural genes. C, Cell Cycle related differentially expressed
list of genes and the relative fold change. Figure 3. TDG-KD NPC differential gene expression analyzed by RNA-seq. A, Differential gene expression of n = 3 siRNA knockdown
experiments. Scatter plot of the group average FPKM (log2) for all genes mapped above the background cutoff, differentially expressed
genes (over 1.5 fold change; p<0.05) are highlighted in red and green. B, Functional annotation of differentially expressed genes shows
significant change in genes related to cell cycle, regulation of apoptosis and structural genes. C, Cell Cycle related differentially expressed
list of genes and the relative fold change. Figure 3. TDG-KD NPC differential gene expression analyzed by RNA-seq. A, Differential gene expression of n = 3 siRNA knockdown
experiments. Scatter plot of the group average FPKM (log2) for all genes mapped above the background cutoff, differentially expressed
genes (over 1.5 fold change; p<0.05) are highlighted in red and green. B, Functional annotation of differentially expressed genes shows
significant change in genes related to cell cycle, regulation of apoptosis and structural genes. C, Cell Cycle related differentially expressed
list of genes and the relative fold change. Figure 3. TDG-KD NPC differential gene expression analyzed by RNA-seq. A, Differential gene expression of n = 3 siRNA knockdown
experiments. Scatter plot of the group average FPKM (log2) for all genes mapped above the background cutoff, differentially expressed
genes (over 1.5 fold change; p<0.05) are highlighted in red and green. B, Functional annotation of differentially expressed genes shows
significant change in genes related to cell cycle, regulation of apoptosis and structural genes. C, Cell Cycle related differentially expressed
list of genes and the relative fold change. Development of FUCCI model for cell cycle analyses Page 7 of 19 Page 7 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 Figure 4. TDG downregulation results in elevation of 5caC, 5fC in a CpG island upstream to the EGR1 TSS site. A, Schematic illustration
of the sequencing of methylase treated compared to non-treated bisulfite converted transcripts. B, EGR1 expression levels are downregulated
following TDG knockdown as measured by RNA-seq (described in Figure 2). C, Sanger sequencing of the CpG island upstream to the
EGR1 TSS following either bisulfite conversion (BS only; top) or MAB treatment (bottom two) shows higher abundance of 5cAC, 5fC in TDG
deficient cells; numbers indicate CpG dinucleotide position. D, Summary visualization (left) and quantification (right) of the abundance of
non-converted residues described in Figure 4C. Error bars represent standard error of the mean methylation level of at least 5 sequenced
samples, p values were calculated with Student’s t test: *** = p<0.001. Figure 4. TDG downregulation results in elevation of 5caC, 5fC in a CpG island upstream to the EGR1 TSS site. A, Schematic illustration
of the sequencing of methylase treated compared to non-treated bisulfite converted transcripts. B, EGR1 expression levels are downregulated
following TDG knockdown as measured by RNA-seq (described in Figure 2). C, Sanger sequencing of the CpG island upstream to the
EGR1 TSS following either bisulfite conversion (BS only; top) or MAB treatment (bottom two) shows higher abundance of 5cAC, 5fC in TDG
deficient cells; numbers indicate CpG dinucleotide position. D, Summary visualization (left) and quantification (right) of the abundance of
non-converted residues described in Figure 4C. Error bars represent standard error of the mean methylation level of at least 5 sequenced
samples, p values were calculated with Student’s t test: *** = p<0.001. Page 8 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 Figure 5. TDG deficiency entails a change in NPC cell cycle. A, TDG downregulation results in higher fraction of cells entering mitotic cell cycle, based on KI67 positive cells
staining compared to CONT-KD. Left: representative IF (images were taken at 10X magnification). Right: quantification of over 400 cells from five separate fields. Error bars represent
standard error of the mean. p-values were calculated with Student’s t test: **=p<0.01. B, TDG downregulation results with a change in cell cycle progression as G2/M increased. Development of FUCCI model for cell cycle analyses Left: representative flow analysis (out of four independent TDG-KD experiments analyzed by flow cytometry) of TDG-KD NPC cell cycle compared to CONT-KD NPC, based on DNA
content staining with PI. Right: Quantification of cell cycle phases from 4 separate TDG knockdown experiments Figure 5. TDG deficiency entails a change in NPC cell cycle. A, TDG downregulation results in higher fraction of cells entering mitotic cell cycle, based on KI67 positive cells
staining compared to CONT-KD. Left: representative IF (images were taken at 10X magnification). Right: quantification of over 400 cells from five separate fields. Error bars represent
standard error of the mean. p-values were calculated with Student’s t test: **=p<0.01. B, TDG downregulation results with a change in cell cycle progression as G2/M increased. Left: representative flow analysis (out of four independent TDG-KD experiments analyzed by flow cytometry) of TDG-KD NPC cell cycle compared to CONT-KD NPC, based on DNA
content staining with PI. Right: Quantification of cell cycle phases from 4 separate TDG knockdown experiments Figure 5. TDG deficiency entails a change in NPC cell cycle. A, TDG downregulation results in higher fraction of cells entering mitotic cell cycle, based on KI67 positive cells
staining compared to CONT-KD. Left: representative IF (images were taken at 10X magnification). Right: quantification of over 400 cells from five separate fields. Error bars represent
standard error of the mean. p-values were calculated with Student’s t test: **=p<0.01. B, TDG downregulation results with a change in cell cycle progression as G2/M increased. Left: representative flow analysis (out of four independent TDG-KD experiments analyzed by flow cytometry) of TDG-KD NPC cell cycle compared to CONT-KD NPC, based on DNA
content staining with PI. Right: Quantification of cell cycle phases from 4 separate TDG knockdown experiments Page 9 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 ermediate modifications level changes throughout the cell cycle. A, Illustration of the FUCCI cell cyc
staining for particular cell cycle phase with antibody against either TDG or methylation intermediate mo
otide for each cell cycle stage (Pics were taken using a 20X magnification and a cropped single nu
y of staining was quantified using ImageJ for over 500 cells, values normalized to G1 (Y Axis) are
with Student’s t test: **=p< 0.01, ***=p<0.001, ns=not significant. Pa Figure 6. Methylation intermediate modifications level changes throughout the cell cycle. A, Illustration of the FUCCI cell cycle reporter
system expression. Development of FUCCI model for cell cycle analyses B, Co-staining for particular cell cycle phase with antibody against either TDG or methylation intermediate modifications. Left: representative nucleotide for each cell cycle stage (Pics were taken using a 20X magnification and a cropped single nucleotide is
presented) Right: intensity of staining was quantified using ImageJ for over 500 cells, values normalized to G1 (Y Axis) are presented. p values were calculated with Student’s t test: **=p< 0.01, ***=p<0.001, ns=not significant. F1000Research 2018, 7:497 Last updated: 31 MAR 2022 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 Because of the difference in expression levels of TDG
between pluripotent and tissue derived NPCs (Figure 1A), we
expected that silencing TDG would have a positive effect on
the progression of developmental maturity of the NPCs. The
LIN28/let-7 circuit was previously shown to be differentially
regulated between NPCs born from pluripotent stem cells
versus those derived from tissues, and resolution of this
discrepancy was sufficient to advance the developmental matu-
rity of NPCs in that context17. When the expression of TDG was
brought down to a level similar to that seen in tissue derived
NPCs, instead of advancing developmental maturation, the cells
appeared to be unable to efficiently differentiate (Figure 2). This was presumably due to the increased rate of prolifera-
tion of the NPCs, a cell type where forced exit of the cell cycle
is known to induce differentiation. Therefore, experimen-
tally regulating TDG levels does not facilitate differentiation
from pluripotent stem cells, as was the case with experi-
mental downregulation LIN28. Perhaps the more interesting
result from this work is that pluripotent derivatives probably
need to silence TDG expression or activity at a more
developmentally appropriate time point to proceed through
proper development. Because of the difference in expression levels of TDG
between pluripotent and tissue derived NPCs (Figure 1A), we
expected that silencing TDG would have a positive effect on
the progression of developmental maturity of the NPCs. The
LIN28/let-7 circuit was previously shown to be differentially
regulated between NPCs born from pluripotent stem cells
versus those derived from tissues, and resolution of this
discrepancy was sufficient to advance the developmental matu-
rity of NPCs in that context17. When the expression of TDG was
brought down to a level similar to that seen in tissue derived
NPCs, instead of advancing developmental maturation, the cells
appeared to be unable to efficiently differentiate (Figure 2). Development of FUCCI model for cell cycle analyses Antibodies used include the following: polyclonal
rabbit anti-TDG 1:100(Atlas; HPA052263); polyclonal chicken
anti-GFAP 1:1000(Abcam; ab4674); mouse monoclonal anti-
MAP2 1:500 (Abcam, ab11267), polyclonal rat anti-KI67 1:100
(eBioscience; 14-5698). For methylation modifications, per-
meabilized cells were denatured with 2N HCl for 15 min
and then neutralized with 100 mM Tris-HCl (pH 8.5) for
10min before blocking. The following Active-Motif Anti-
bodies were used: rabbit anti 5hmC 1:100(39770); rabbit
anti-5fC 1:2500 (61223); rabbit anti-5caC 1:1000 (61225). Image
analysis and quantification was performed using ImageJ ver-
sion 1.50i with the same threshold for each channel for all
samples. Western Blot analysis was performed using standard
procedures as described (Lowry et al., 2005) antibodies
used were rabbit anti-TDG as described above and mouse
anti-Actin 1:1000 (Santa Cruz Biotechnology, sc-47778). Methods Tissue culture and TDG knockdown
H9 hESCs and XFIPS2 were used in this study in accord-
ance with the UCLA Embryonic Stem Cell Research Oversight
committee (ESCRO, 2006-019-11A) and the Institutional
Biosafety Committee (IBC). This work was specifically described
in our ESCRO application (2006-019-11A), and approved on
an annual basis by the committee. This work is considered
not human subjects research by the UCLA IRB. RT-PCR Analysis. Total RNA was extracted using an RNeasy
Mini Kit (QIAGEN). cDNA synthesis was performed using
the Superscript III first-strand cDNA synthesis kit (Invitrogen). Real-time PCR was performed in triplicate using the SYBR
green real-time PCR MIX (Roche) in the Roche lightcycler
480 machine. Run was performed for 50 cycles and analysis
was performed in Microsoft Excel 2013 using the 2-ΔΔCT
method. Cells were cultured in feeder free conditions on Matrigel (Corn-
ing) using mTeSR1 (Stem Cell Technologies) and passaged
mechanically or with collagenase every 4–5 days. NPCs differ-
entiation was performed as described previously (10). Briefly, for
rosette induction, 80% confluent hPCS were cultured in DMEM/
F12 with N2 and B27 supplements (Invitrogen), 20 ng/ml basic
fibroblast growth factor (bFGF; R&D Systems), 1 µM retinoic
acid (Sigma), 1 µM Sonic Hedgehog Agonist (Purmorphamine;
Sigma), 10µM SB431542 (TGFß inhibitor; Cayman) and
0.1µM LDN193189 (BMP receptor type 1 inhibitor; Cayman). Small molecules were resuspended according to each manu-
facturer instructions. After about a week of culture neural-like
rosettes were mechanically picked and expanded in NPC
maintenance medium: DMEM/F12 supplemented with N2/B27,
bFGF and 500 ng/ml epidermal growth factor (EGF; GIBCO). For further differentiation, the growth factors bFGF and EGF
were withdrawn from the media (GFW) and cultured for 3
weeks. TDG and control knockdown in NPCs was performed
using a unique 27-mer siRNA duplexes (Trilencer, Origene) at
a final concentration of 20 nM using Lipofectamine RNAiMAX
transfection reagent (Invitrogen). RNA-Seq q
Total RNA was extracted using an RNeasy Mini Kit (QIAGEN). Libraries were constructed according to manufacturer instruc-
tions (TruSeq Stranded Total RNA with Ribo-Zero; Illumina). Following second strand PCR amplification, ~200bp sized
libraries were excised from agarose gel and pooled together
in
10mM
concentration
each. Samples
were
sequenced
using Illumina HiSeq2000 on single-end 50-bp reads and
aligned to human reference genome (Hg19) using Tophat
(version 2.0.6)18. Processing using Cufflinks and Cuffdiff was
performed to obtain differential fragments per kilobase of tran-
script per million mapped reads (FPKM)18. Three biological
replicates (i.e. 3 separate knockdown experiments in different
PSC clones) were grouped together. Further analysis was per-
formed using the cummeRbund suite (v2.0.0). Functional
annotation was performed using DAVID (V6.7)14. Development of FUCCI model for cell cycle analyses This was presumably due to the increased rate of prolifera-
tion of the NPCs, a cell type where forced exit of the cell cycle
is known to induce differentiation. Therefore, experimen-
tally regulating TDG levels does not facilitate differentiation
from pluripotent stem cells, as was the case with experi-
mental downregulation LIN28. Perhaps the more interesting
result from this work is that pluripotent derivatives probably
need to silence TDG expression or activity at a more
developmentally appropriate time point to proceed through
proper development. PBS for 20 min, washed and then permeabilized and blocked in
10% donkey serum, 0.01% Triton in PBS for 1 hour. Primary
antibodies in 5% donkey serum were incubated for 1–2 hours
at room temperature following 3X wash in PBST and incuba-
tion with conjugated secondary antibody for 1 hour in room
temperature. After 3X wash with PBST cover slips were
incubated with 300nM DAPI final concertation in PBST for
3 min in room temperature (dark), followed with 3X wash
with PBST. Antibodies used include the following: polyclonal
rabbit anti-TDG 1:100(Atlas; HPA052263); polyclonal chicken
anti-GFAP 1:1000(Abcam; ab4674); mouse monoclonal anti-
MAP2 1:500 (Abcam, ab11267), polyclonal rat anti-KI67 1:100
(eBioscience; 14-5698). For methylation modifications, per-
meabilized cells were denatured with 2N HCl for 15 min
and then neutralized with 100 mM Tris-HCl (pH 8.5) for
10min before blocking. The following Active-Motif Anti-
bodies were used: rabbit anti 5hmC 1:100(39770); rabbit
anti-5fC 1:2500 (61223); rabbit anti-5caC 1:1000 (61225). Image
analysis and quantification was performed using ImageJ ver-
sion 1.50i with the same threshold for each channel for all
samples. Western Blot analysis was performed using standard
procedures as described (Lowry et al., 2005) antibodies
used were rabbit anti-TDG as described above and mouse
anti-Actin 1:1000 (Santa Cruz Biotechnology, sc-47778). PBS for 20 min, washed and then permeabilized and blocked in
10% donkey serum, 0.01% Triton in PBS for 1 hour. Primary
antibodies in 5% donkey serum were incubated for 1–2 hours
at room temperature following 3X wash in PBST and incuba-
tion with conjugated secondary antibody for 1 hour in room
temperature. After 3X wash with PBST cover slips were
incubated with 300nM DAPI final concertation in PBST for
3 min in room temperature (dark), followed with 3X wash
with PBST. Cell cycle analysis Following trypsin dissociation, knocked-down NPCs were
fixed overnight in 70% ethanol at -20°C. Fixed cells were then
stained for half an hour at room temperature in the dark, with
Propidium Iodide (PI) for a final concertation of 50 µg/ml
supplemented with RNAse (final 1 µg/ml). DNA content was
analyzed on BD-Biosciences LSR-II flow cytometer and cell 1.
Kohli RM, Zhang Y: TET enzymes, TDG and the dynamics of DNA
demethylation. Nature. 2013; 502(7472): 472–479.
PubMed Abstract | Publisher Full Text | Free Full Text
2.
Sen GL, Reuter JA, Webster DE, et al.: DNMT1 maintains progenitor function in
self-renewing somatic tissue. Nature. 2010; 463(7280): 563–567.
PubMed Abstract | Publisher Full Text | Free Full Text
3.
Flinders C, Lam L, Rubbi L, et al.: Epigenetic changes mediated by polycomb
repressive complex 2 and E2a are associated with drug resistance in a mouse
model of lymphoma. Genome Med. 2016; 8(1): 54.
PubMed Abstract | Publisher Full Text | Free Full Text
4.
Mikkelsen TS, Ku M, Jaffe DB, et al.: Genome-wide maps of chromatin state in
pluripotent and lineage-committed cells. Nature. 2007; 448(7153): 553–560.
PubMed Abstract | Publisher Full Text | Free Full Text
5.
Guenther MG, Levine SS, Boyer LA, et al.: A chromatin landmark and
transcription initiation at most promoters in human cells. Cell. 2007; 130(1):
77–88.
PubMed Abstract | Publisher Full Text | Free Full Text
6.
Kristensen DG, Nielsen JE, Jørgensen A, et al.: Evidence that active
demethylation mechanisms maintain the genome of carcinoma in situ cells Statistical analysis Student’s t-test was performed using GraphPad 6.01. Results
were judged to be significant if the p-value was < 0.05. All
other statistical analysis described in this study were performed
using Microsoft Excel 2013 Data availability
Dataset 1: TDG regulates cell cycle progression in human
neural progenitors. Complete dataset of all underlying data
divided into folders based on relevant figures. 10.5256/
f1000research.13801.d20137919 Grant information This work was supported by NIH (P01GM9913), Allen Dis-
tinguished Investigator award from the Allen Frontiers Group
to WEL, A CIRM Basic Biology Award (RT-2), and pilot
support from the BSCRC at UCLA (Rose Hills Scholar Award). This work was supported by NIH (P01GM9913), Allen Dis-
tinguished Investigator award from the Allen Frontiers Group
to WEL, A CIRM Basic Biology Award (RT-2), and pilot
support from the BSCRC at UCLA (Rose Hills Scholar Award). The funders had no role in study design, data collection and
analysis, decision to publish, or preparation of the manuscript. FUCCI cell line generation The FUCCI reporter lentiviral plasmids, pCSII-EF-mCherry-
hCdt1(30/120)
and
pCSII-EF-mVenus-hGeminin(1/110)
were a generous gift of Dr. Atsushi Miyawaki (RIKEN Brain
Science Institute, Saitama, Japan). Lentiviral virions were gen-
erated in 293T cells using standard protocols as previously
described13 followed by concentration with Amicon Ultra-15
centrifugal units (100K; Millipore). hPSC were single celled
with TryplE (Thermo) and re-plated 24h prior to infection
supplemented with 10 µM Rho-associated kinase (ROCK)
inhibitor Y27632 (Stemgent). Cells were first infected with one
reporter lentivirus particles for overnight infection. Cells were
washed with fresh medium and grown for 2– 3 passages for
recovery and expansion. Next, we FACS sorted using FacsARIA
(Becton Dickinson) the cells for the corresponding reporter
to ensure that all cells are infected. Briefly: cells from a
whole 6 well plate were treated with ROCK inhibitor for one MAB-PCR We infected the cells with the second reporter using that same
procedure followed by a second FACS sorting for the reporter. Genomic DNA was extracted using the DNeasy Blood &
Tissue Kit (Qiagen; 69506). One µg genomic DNA was
treated by M.SssI (New England Biolabs; M0226s) in a 50 µl
reaction for three rounds. For each round DNA was incubated
with 4 Units of M.SssI CpG methyltransferase (NEB), supple-
mented with 160 mM final S-Adenosyl methionine for 3 hours
at 37°C. At the end of each round DNA was cleaned using phe-
nol/chloroform extraction. Bisulfite conversion was performed
using the EpiTect Bisulfite Kit (QIAGEN; 59104) and then
selected loci was PCR amplified with KAPA HiFi Hotstart
Uracil+ DNA polymerase (KAPABiosystems). The resulting
PCR product was cloned into the TOPO-Blunt (Invitrogen)
vector, and sent to Laragen for Sanger sequencing (GeneWiz). Analysis and visualization of sequence reads was done
using the online BISMA tool. Immunofluorescence and Western Blot Immunofluorescence and Western Blot
Immunofluorescent staining was performed using standard
protocol10,17. Briefly, cover slips were fixed with 4% PFA in l
Immunofluorescent staining was performed using standard
protocol10,17. Briefly, cover slips were fixed with 4% PFA in Page 11 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 cycle phases were determined using the FlowJo cell cycle
module (version 7.6.5). hour, single celled using trypLE and re-suspended in PBS. After sorting cells were a re-plated for recovery for 5–7 days
and then hour, single celled using trypLE and re-suspended in PBS. After sorting cells were a re-plated for recovery for 5–7 days
and then Acknowledgements We would like to acknowledge the support of various core
facilities and their staff at the core facilities sponsored by
the Eli and Edythe Broad Center for Regenerative Medicine
(EEBCRC) including: Flow Cytometry, Genomics, and the Stem
Cell Cores. References hypomethylated in the adult testis. Br J Cancer. 2014; 110(3): 668–678. PubMed Abstract | Publisher Full Text | Free Full Text
7. Kinney SR, Pradhan S: Ten eleven translocation enzymes and
5-hydroxymethylation in mammalian development and cancer. Adv Exp Med
Biol. 2013; 754: 57–79. PubMed Abstract | Publisher Full Text
8. Cortázar D, Kunz C, Selfridge J, et al.: Embryonic lethal phenotype reveals a
function of TDG in maintaining epigenetic stability. Nature. 2011; 470(7334):
419–423. PubMed Abstract | Publisher Full Text
9. Rai K, Huggins IJ, James SR, et al.: DNA demethylation in zebrafish involves
the coupling of a deaminase, a glycosylase, and gadd45. Cell. 2008; 135(7):
1201–1212. PubMed Abstract | Publisher Full Text | Free Full Text
10. Patterson M, Chan DN, Ha I, et al.: Defining the nature of human pluripotent
stem cell progeny. Cell Res. 2012; 22(1): 178–193. PubMed Abstract | Publisher Full Text | Free Full Text
11. Braun N, Papadopoulos T, Müller-Hermelink HK: Cell cycle dependent
distribution of the proliferation-associated Ki-67 antigen in human embryonic PubMed Abstract | Publisher Full Text | Free Full Text 6. Kristensen DG, Nielsen JE, Jørgensen A, et al.: Evidence that active
demethylation mechanisms maintain the genome of carcinoma in situ cells 11. Braun N, Papadopoulos T, Müller-Hermelink HK: Cell cycle dependent
distribution of the proliferation-associated Ki-67 antigen in human embryonic Page 12 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 lung cells. Virchows Arch B Cell Pathol Incl Mol Pathol. 1988; 56(1): 25–33. PubMed Abstract | Publisher Full Text
12. Wheldon LM, Abakir A, Ferjentsik Z, et al.: Transient accumulation of 5-
carboxylcytosine indicates involvement of active demethylation in lineage
specification of neural stem cells. Cell Rep. 2014; 7(5): 1353–1361. PubMed Abstract | Publisher Full Text
13. Germanguz I, Listgarten J, Cinkornpumin J, et al.: Identifying gene expression
modules that define human cell fates. Stem Cell Res. 2016; 16(3): 712–724. PubMed Abstract | Publisher Full Text | Free Full Text
14. Huang da W, Sherman BT, Lempicki RA: Systematic and integrative analysis of
large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009; 4(1):
44–57. PubMed Abstract | Publisher Full Text
15. Penner MR, Parrish RR, Hoang LT, et al.: Age-related changes in Egr1
transcription and DNA methylation within the hippocampus. Hippocampus. 2016; 26(8): 1008–1020. PubMed Abstract | Publisher Full Text | Free Full Text
16. Singh AM, Trost R, Boward B, et al.: Utilizing FUCCI reporters to understand
pluripotent stem cell biology. Methods. 2016; 101: 4–10. PubMed Abstract | Publisher Full Text | Free Full Text
17. Patterson M, Gaeta X, Loo K, et al.: let-7 miRNAs can act through notch to
regulate human gliogenesis. Stem Cell Reports. 2014; 3(5): 758–73. PubMed Abstract | Publisher Full Text | Free Full Text
18. Trapnell C, Roberts A, Goff L, et al.: Differential gene and transcript expression
analysis of RNA-seq experiments with TopHat and Cufflinks. Nat Protoc. 2012;
7(3): 562–578. PubMed Abstract | Publisher Full Text | Free Full Text
19. Germanguz I, Park J, Cinkornpumin J, et al.: Dataset 1 in: TDG regulates cell
cycle progression in human neural progenitors. F1000Research. 2018. Data Source lung cells. Virchows Arch B Cell Pathol Incl Mol Pathol. 1988; 56(1): 25–33. PubMed Abstract | Publisher Full Text 14. Huang da W, Sherman BT, Lempicki RA: Systematic and integrative analysis of
large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009; 4(1):
44–57. PubMed Abstract | Publisher Full Text 15.
Penner MR, Parrish RR, Hoang LT, et al.: Age-related changes in Egr1
transcription and DNA methylation within the hippocampus. Hippocampus. https://doi.org/10.5256/f1000research.15003.r38594 © 2018 Bellacosa A et al. This is an open access peer review report distributed under the terms of the Creative
Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium,
provided the original work is properly cited. Rahul Prasad
Fox Chase Cancer Center, Philadelphia, PA, USA
Timothy Yen
Fox Chase Cancer Center, Philadelphia, PA, USA
Alfonso Bellacosa
Fox Chase Cancer Center, Philadelphia, PA, USA 13/11/2018: This referee report has been updated from a Not Approved to an Approved with
Reservations, and an additional sentence added, to reflect additional feedback from the
referees after their report was published. PubMed Abstract | Publisher Full Text 15. Penner MR, Parrish RR, Hoang LT, et al.: Age-related changes in Egr1
transcription and DNA methylation within the hippocampus. Hippocampus. Page 13 of 19 Page 13 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 Open Peer Review 13/11/2018: This referee report has been updated from a Not Approved to an Approved with
Reservations, and an additional sentence added, to reflect additional feedback from the
referees after their report was published. Also, the authors should clarify why the overall DNA content appears to be reduced in
TDG knockdown cells. The reported increased fraction of G2/M cells following TDG knockdown using FACS merits further
investigation. The use of propidium iodide prevents noting a distinction between increased
proliferation (increased mitotic cells) or if the cells are simply arrested in G2. This can be resolved
by using DAPI with a mitotic marker such as phospho-histone H3. Additionally, the authors do not investigate the effects of TDG knockdown using the FUCCI
reporter system. This seems to be a technically feasible experiment that would serve as an
important point in the study’s overall narrative linking TDG effects to the cell cycle. The use of EGR1 as an example of methylation impacting gene expression is a point of concern. Selection of EGR1 as a gene of interest appears to be based on a single reference demonstrating
methylation dependent expression in a rat model. Selection of a gene with a more established
methylation dependent promotor region, such as Pax 6, to compare with the expression profile
would be a better indicator of whether TDG knockdown affects differentiation2. We would argue
that the authors should show changes in methylation intermediate levels by TDG knockdown,
using MAB-seq, in more than a single gene. A minimum of three genes would seem to be
appropriate to convincingly show an effect of TDG on gene expression via changes in methylation
intermediates. The following points would be helpful to be addressed: The reduction of 5hmC levels in early S phase, as demonstrated in figure 6B is mentioned as
previously reported but without citation in the main text. 1. The reduction of 5hmC levels in early S phase, as demonstrated in figure 6B is mentioned as
previously reported but without citation in the main text. 1. In Figure 6B, 5hmC levels are shown to reduce in early S relative to G1, consistent with the
coinciding increase in 5fC and 5caC levels. However, while 5fC and 5caC levels remain
elevated in S, G2, and M compared to G1, 5hmC levels return to G1 levels and statistically
higher levels than early S phase. This is a paradoxical finding and should be addressed. 2. In the methodology described for MAB-seq, the use of phenol/chloroform extraction is
noted. 13/11/2018: This referee report has been updated from a Not Approved to an Approved with
Reservations, and an additional sentence added, to reflect additional feedback from the
referees after their report was published. Germanguz et al. investigate the role of TDG in the differentiation and proliferation of human
neural progenitor cells using RNA interference, RT-PCR, bisulfite sequencing, and cell cycle
profiling with the FUCCI system. They show that TDG knockdown, using siRNA and assessed by
immunofluorescence and supported by increased levels of 5caC and 5fC but not 5hmC, correlates
with an overall decrease in proportion of differentiated cells (both neurons and glia) following
induction, in contrast to the authors’ expectation of increased developmental maturity following
TDG knockdown. RT-PCR data show expression of neural progenitor cell markers to be unaffected
by TDG knockdown. The authors then investigate cell cycle changes using FACS following TDG
knockdown and report an increased fraction of G2/M cells. The FUCCI cell cycle reporter system
was used to show that TDG levels are decreased outside of G1. Methylation-assisted bisulfite
sequencing (MAB-seq) was used to show evidence of changes in methylation intermediate
modification levels in a putative promoter region of EGR1. The authors conclude that the reduction
of TDG expression induces cell cycle specific changes that result in increased proliferation and
reduced differentiation of neural progenitor cells. Overall the study uses a strong combination of
experimental methods to show interesting findings but is likely a few experiments short of adding
to the growing narrative involving TDG, demethylation, and cellular differentiation. One point the authors do not account for is the potential for reprogramming of neural progenitor
cells to different, undifferentiated progenitors. DNMT3B knockdown has been shown to cause an
increase in neural crest cell markers, hypomethylation at the Sox 10 promoter region, and Page 14 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 ultimately prolonged neural crest emigration1. Therefore it seems important to confirm that the
neural progenitor cells that failed to fully differentiate remained in the neural cell lineage instead
of reprogramming to neural crest precursors – if reprogramming did occur this would be a
significant finding. The cell cycle data are not clear; there seems to be a discrepancy between the marked effect of
TDG knockdown as assessed by Ki67 staining (Fig. 5A), which suggests an increased fraction of
proliferating cells, and the rather subtle effect as assessed by FACS (Fig. 5B). In particular, the
authors should show the percent of cells in G1, S and G2M by FACS in control and TDG
knockdown. Minor points:
Normalized is mispelled in the figure legends. References 1. Hu N, Strobl-Mazzulla PH, Simoes-Costa M, Sánchez-Vásquez E, et al.: DNA methyltransferase 3B
regulates duration of neural crest production via repression of Sox10.Proc Natl Acad Sci U S A. 2014;
111 (50): 17911-6 PubMed Abstract | Publisher Full Text 1. Hu N, Strobl-Mazzulla PH, Simoes-Costa M, Sánchez-Vásquez E, et al.: DNA methyltransferase 3B
regulates duration of neural crest production via repression of Sox10.Proc Natl Acad Sci U S A. 2014;
111 (50): 17911 6 PubMed Abstract | Publisher Full Text 1. Hu N, Strobl-Mazzulla PH, Simoes-Costa M, Sánchez-Vásquez E, et al.: DNA methyltransferase 3B
regulates duration of neural crest production via repression of Sox10.Proc Natl Acad Sci U S A. 2014;
b
d
b
|
bl h
ll 1. Hu N, Strobl-Mazzulla PH, Simoes-Costa M, Sánchez-Vásquez E, et al.: DNA methyltransferase 3B
regulates duration of neural crest production via repression of Sox10.Proc Natl Acad Sci U S A. 2014;
111 (50): 17911-6 PubMed Abstract | Publisher Full Text
2. Verma N, Pan H, Doré LC, Shukla A, et al.: TET proteins safeguard bivalent promoters from de
novo methylation in human embryonic stem cells.Nat Genet. 2018; 50 (1): 83-95 PubMed Abstract |
Publisher Full Text
3. Claycamp H: Phenol sensitization of DNA to subsequent oxidative damage in 8-hydroxyguanine
assays. Carcinogenesis. 1992; 13 (7): 1289-1292 Publisher Full Text
4 Neri F Incarnato D Krepelova A Parlato C et al : Methylation assisted bisulfite sequencing to 111 (50): 17911-6 PubMed Abstract | Publisher Full 2. Verma N, Pan H, Doré LC, Shukla A, et al.: TET proteins safeguard bivalent promoters from de
novo methylation in human embryonic stem cells.Nat Genet. 2018; 50 (1): 83-95 PubMed Abstract |
Publisher Full Text 2. Verma N, Pan H, Doré LC, Shukla A, et al.: TET proteins safeguard bivalent promoters from de
novo methylation in human embryonic stem cells.Nat Genet. 2018; 50 (1): 83-95 PubMed Abstract | y
Publisher Full Text 3. Claycamp H: Phenol sensitization of DNA to subsequent oxidative damage in 8-hydroxyguanine
assays. Carcinogenesis. 1992; 13 (7): 1289-1292 Publisher Full Text
4. Neri F, Incarnato D, Krepelova A, Parlato C, et al.: Methylation-assisted bisulfite sequencing to
simultaneously map 5fC and 5caC on a genome-wide scale for DNA demethylation analysis.Nat
Protoc. 11 (7): 1191-205 PubMed Abstract | Publisher Full Text 3. Claycamp H: Phenol sensitization of DNA to subsequent oxidative damage in 8-hydroxyguanine
assays. Carcinogenesis. 1992; 13 (7): 1289-1292 Publisher Full Text
4. Neri F, Incarnato D, Krepelova A, Parlato C, et al.: Methylation-assisted bisulfite sequencing to
simultaneously map 5fC and 5caC on a genome-wide scale for DNA demethylation analysis.Nat
Protoc. 11 (7): 1191-205 PubMed Abstract | Publisher Full Text 13/11/2018: This referee report has been updated from a Not Approved to an Approved with
Reservations, and an additional sentence added, to reflect additional feedback from the
referees after their report was published. This method has an established risk of sensitizing DNA to oxidation and should be
addressed3, and therefore artifactually increase the levels of the oxidized cytosine species
5fC and 5caC. It is recommended that alternative methods of extracting DNA are used after
each M.SssI treatment, such as Agencourt AMPure XP Beads4. 3. Minor points:
Normalized is mispelled in the figure legends. Minor points:
Normalized is mispelled in the figure legends. Page 15 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 Several references are missing in the Introduction and throughout the paper, such a g
g
Cortellino et al. (2011) on TDG knockout and embryonic lethality. y
y
Hardeland et al (2007), Slenn et al. (2014) and Shibata et al. (2014), on the decrease of TDG levels in
S-phase. Hardeland et al (2007), Slenn et al. (2014) and Shibata et al. (2014), on the decrease of TDG levels in
S-phase. In the Introduction, what are the references for the sentence: “Loss of methylation in specific
locations…”? In the Introduction, what are the references for the sentence: “Loss of methylation in specific
locations…”? In Fig. 2E what are the p-values for the differentiation? In particular, is the difference in GFAP-
positive cells between control and TDG-KD significant? In the MAB-seq results, can the authors clarify the sentence: “This locus was chosen as it was
reported (which reference?) to have the highest distribution of 5fC, 5caC around the TSS”? For the MAB-seq analysis in Fig. 4, it would be useful to have a schematic map showing the
relationship of the CpG island to the TSS. In the MAB-seq results, can the authors clarify the sentence: “This locus was chosen as it was
reported (which reference?) to have the highest distribution of 5fC, 5caC around the TSS”? For the MAB-seq analysis in Fig. 4, it would be useful to have a schematic map showing the
relationship of the CpG island to the TSS. p
p
In the paragraph on the FUCCI results, “DNA glycosylation activity of TDG” should be changed to
“DNA glycosylase activity of TDG”. In the paragraph on the FUCCI results, “DNA glycosylation activity of TDG” should be changed to
“DNA glycosylase activity of TDG”. In the paragraph on the FUCCI results, it is not clear whether the FUCCI system was employed in
hESC or hPSC cells. 13/11/2018: This referee report has been updated from a Not Approved to an Approved with
Reservations, and an additional sentence added, to reflect additional feedback from the
referees after their report was published. In the paragraph on the FUCCI results, it is not clear whether the FUCCI system was employed in
hESC or hPSC cells. In the paragraph on the FUCCI results, the last sentence: “Furthermore, all these data…” is not
clear and should be rewritten. In the paragraph on the FUCCI results, the last sentence: “Furthermore, all these data…” is not
clear and should be rewritten. The Discussion in the present form is a bit unclear, and should be rewritten, e.g. the last sentence
is particularly imprecise. Also, references should be added to the Discussion. In Fig. 4C, change “5C” to “C”. The Discussion in the present form is a bit unclear, and should be rewritten, e.g. the last sentence
is particularly imprecise. Also, references should be added to the Discussion. In Fig 4C change “5C” to “C” Alexey Ruzov Alexey Ruzov
Division of Cancer and Stem Cells, Centre for Biomolecular Sciences (CBS), School of Medicine,
University of Nottingham, Nottingham, UK We confirm that we have read this submission and believe that we have an appropriate level
of expertise to confirm that it is of an acceptable scientific standard, however we have
significant reservations, as outlined above. Reviewer Report 14 June 2018 https://doi.org/10.5256/f1000research.15003.r34353 https://doi.org/10.5256/f1000research.15003.r34353 © 2018 Ruzov A et al. This is an open access peer review report distributed under the terms of the Creative
Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium,
provided the original work is properly cited. Are sufficient details of methods and analysis provided to allow replication by others?
Partly Are sufficient details of methods and analysis provided to allow replication by others? Partly If applicable, is the statistical analysis and its interpretation appropriate? Yes Are all the source data underlying the results available to ensure full reproducibility? Page 16 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 Yes
Are the conclusions drawn adequately supported by the results? Partly
Competing Interests: No competing interests were disclosed. We confirm that we have read this submission and believe that we have an appropriate level
of expertise to confirm that it is of an acceptable scientific standard, however we have
significant reservations, as outlined above. Yes Yes
Are the conclusions drawn adequately supported by the results? Partly
Competing Interests: No competing interests were disclosed. We confirm that we have read this submission and believe that we h
of expertise to confirm that it is of an acceptable scientific standard,
significant reservations, as outlined above. cycle dynamics of these modifications from the presented results. Multiple images of the cell nuclei should be shown in Figure 6 for each condition. 5. Multiple images of the cell nuclei should be shown in Figure 6 for each condition
5. The nuclei are incorrectly called “nucleotides” in the Figure 6 legend. 6. Both the introduction and discussion are rather short and extremely superficial. Some of
the references are cited incorrectly or in improper place. Furthermore, there is a large
number of factual sentences in these sections that are lacking any references. Taking this
into account, both Introduction and Discussion should be completely rewritten. 7. Abdulkadir Abakir Division of Cancer and Stem Cells, Centre for Biomolecular Sciences (CBS), School of Medicine,
University of Nottingham, Nottingham, UK The manuscript by Germanguz et al. investigates potential roles of active DNA demethylation and
the component of DNA base excision repair (BER) pathway, thymine DNA glycosylase (TDG) in cell
cycle progression using human pluripotent stem cells (hPSCs) as a model. The authors first
examined the expression levels of TDG in hPSCs, hPSC-derived neuronal progenitor cells (NPCs) as
well as in the tissue-derived NPCs. Although the authors report a significant increase in expression
of TDG in the hPSC-derived NPCs compared to the NPCs derived from tissues, it is unclear how
comparable these two NPCs derivatives are and, therefore, the rationale behind this comparison is
not fully clear. Interestingly, the authors find that expression of TET enzymes responsible for
active DNA demethylation mimic TDG expression pattern in hPSCs and NPCs. Taking this into
account together with the fact that cellular proliferation diminishes following birth in neural
tissues, the authors sought to examine whether this pattern of TDG expression is linked to the cell
cycle progression. To this end, the authors found that TDG levels increase during G0/G1 and
decrease following entry to S and G2/M phases of the cell cycle. Next, the authors enquired
whether the decrease in the differentiation potential of NPCs following TDG KD is due to
prolonged proliferative status and show 34 cell cycle related genes amongst the genes
differentially expressed upon the TDG knockdown. Finally, using a cell cycle reporter, FUCCI, the
authors attempt to link the dynamics of TDG levels during cell cycle with active demethylation of
the hPSC genome. Although, the authors show that TDG levels increase in G0/G1 cells and that 5fC
and 5caC begin to accumulate in S phase, given the absence of what happens to the levels of 5fC
and 5caC during cell cycle following TDG KD, it is not possible to conclude if TDG impacts the cell Page 17 of 19 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 cycle dynamics of these modifications from the presented results. cycle dynamics of these modifications from the presented results. Overall, I think that this manuscript presents some novel and interesting data and the authors
have a potential to develop this story much further. At the same time there is a number of issues
with the current paper that should be fixed before it can be indexed, in my opinion. In addition to the discrepancies mentioned above that need to be, at least explained in the text,
there are following points that would be necessary to address: In addition to the discrepancies mentioned above that need to be, at least explained in the text,
there are following points that would be necessary to address: The authors do not show how their findings on the involvement of TDG in cell cycle
regulation impact differentiation of NPCs to either neurons or glia (This can be done by
performing MAB-seq on the genes directly involved in neuronal/glial differentiation). It is
difficult to understand why TDG depleted NPC have the same ratio of neuron to glial
differentiation as the wild type ones (Figure 2E). This is especially strange since it is now well
established that affecting the methylation status of differentiating NPCs either directly
through modulation of DNMT proteins or by modulating the active DNA demethylation
pathway via TDG/TETs, impacts the neuronal to glial differentiation lineage switch. 1. It would make sense to confirm at least the most important findings using an independent
siRNA. Moreover, there is no catalogue number for the siRNA used in the manuscript. 2. It would make sense to confirm at least the most important findings using an independent
siRNA. Moreover, there is no catalogue number for the siRNA used in the manuscript. 2. The effects of TDG depletion on the cell cycle regulation shown in Figure 5 are rather mild. I
understand that this can be explained by transient nature of the siRNA mediated depletion
but the conclusion that “Taken together, it seems clear that TDG plays a role in human
pluripotent stem cell cycle regulation” is an overstatement from my point of view. 3. It is unclear from the Methods section what type of microscopy (conventional or confocal)
was used for the cell imaging. 4. It is unclear from the Methods section what type of microscopy (conventional or confocal)
was used for the cell imaging. 4. Are sufficient details of methods and analysis provided to allow replication by others?
Partly Are sufficient details of methods and analysis provided to allow replication by others? Partly If applicable, is the statistical analysis and its interpretation appropriate? Page 18 of 19 Yes
Are all the source data underlying the results available to ensure full reproducibility? Partly
Are the conclusions drawn adequately supported by the results? Partly
Competing Interests: No competing interests were disclosed. Reviewer Expertise: Epigenetics, DNA methylation and demethylation
We confirm that we have read this submission and believe that we have an appropriate level
of expertise to confirm that it is of an acceptable scientific standard, however we have
significant reservations, as outlined above. The benefits of publishing with F1000Research:
Your article is published within days, with no editorial bias
•
You can publish traditional articles, null/negative results, case reports, data notes and more
•
The peer review process is transparent and collaborative
•
Your article is indexed in PubMed after passing peer review
•
Dedicated customer support at every stage
•
For pre-submission enquiries, contact research@f1000.com
F1000Research 2018, 7:497 Last updated: 31 MAR 2022 F1000Research 2018, 7:497 Last updated: 31 MAR 2022 Yes Are the conclusions drawn adequately supported by the results?
Partly Competing Interests: No competing interests were disclosed. The benefits of publishing with F1000Research:
Your article is published within days, with no editorial bias
•
You can publish traditional articles, null/negative results, case reports, data notes and more
•
The peer review process is transparent and collaborative
•
Your article is indexed in PubMed after passing peer review
•
Dedicated customer support at every stage
•
For pre-submission enquiries, contact research@f1000.com The benefits of publishing with F1000Research:
Your article is published within days, with no editorial bias
•
You can publish traditional articles, null/negative results, case reports, data notes and more
•
The peer review process is transparent and collaborative
•
Your article is indexed in PubMed after passing peer review
•
Dedicated customer support at every stage
•
For pre-submission enquiries, contact research@f1000.com Page 19 of 19
|
https://openalex.org/W2561921160
|
https://pure.tue.nl/ws/files/72244550/10.1007_s00330_016_4693_8.pdf
|
English
| null |
Multiparametric dynamic contrast-enhanced ultrasound imaging of prostate cancer
|
European radiology
| 2,016
|
cc-by
| 7,492
|
Document status and date:
Published: 01/08/2017 Document status and date:
Published: 01/08/2017 ocument Version:
ublisher’s PDF, also known as Version of Record (includes final page, issue and volume number Document Version:
Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication: • A submitted manuscript is the version of the article upon submission and before peer-review. There can be
important differences between the submitted version and the official published version of record. People
interested in the research are advised to contact the author for the final version of the publication, or visit the
DOI to the publisher's website. p
• The final author version and the galley proof are versions of the publication after peer review. • The final published version features the final layout of the paper including the volume, issue and page
numbers. Link to publication Link to publication General rights
C
i ht
d General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study
Y
t f
th
di t ib t
th
t
i l
it f
fit
ki
ti it
i l
i • You may not further distribute the material or use it for any profit-making activity o
• You may freely distribute the URL identifying the publication in the public portal. If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please
follow below link for the End User Agreement: Multiparametric dynamic contrast-enhanced ultrasound
imaging of prostate cancer
Citation for published version (APA):
Wildeboer, R. R., Postema, A. W., Demi, L., Kuenen, M. P. J., Wijkstra, W. H., & Mischi, M. M. (2017).
Multiparametric dynamic contrast-enhanced ultrasound imaging of prostate cancer. European Radiology, 27(8),
3226-3234. https://doi.org/10.1007/s00330-016-4693-8 Citation for published version (APA):
Wildeboer, R. R., Postema, A. W., Demi, L., Kuenen, M. P. J., Wijkstra, W. H., & Mischi, M. M. (2017). Multiparametric dynamic contrast-enhanced ultrasound imaging of prostate cancer. European Radiology, 27(8),
3226-3234. https://doi.org/10.1007/s00330-016-4693-8 DOI:
10.1007/s00330-016-4693-8 DOI:
10.1007/s00330-016-4693-8 Download date: 24. Oct. 2024 www.tue.nl/taverne Take down policy
If you believe that this document breaches copyright please contact us at:
openaccess@tue.nl
providing details and we will investigate your claim. Download date: 24. Oct. 2024 Eur Radiol (2017) 27:3226–3234
DOI 10.1007/s00330-016-4693-8 UROGENITAL Multiparametric dynamic contrast-enhanced ultrasound imaging
of prostate cancer Rogier R. Wildeboer1 & Arnoud W. Postema2 & Libertario Demi1 &
Maarten P. J. Kuenen3 & Hessel Wijkstra1,2 & Massimo Mischi1 Received: 25 May 2016 /Revised: 28 November 2016 /Accepted: 1 December 2016 /Published online: 21 December 2016
# The Author(s) 2016. This article is published with open access at Springerlink.com supporting future targeting of biopsies or therapy. Application
in other types of cancer can also be foreseen. Key points supporting future targeting of biopsies or therapy. Application
in other types of cancer can also be foreseen. Key points Abstract Objectives The aim of this study is to improve the accuracy of
dynamic contrast-enhanced ultrasound (DCE-US) for prostate
cancer (PCa) localization by means of a multiparametric approach. Materials and Methods Thirteen different parameters related to
either perfusion or dispersion were extracted pixel-by-pixel from
45 DCE-US recordings in 19 patients referred for radical prosta-
tectomy. Multiparametric maps were retrospectively produced
using a Gaussian mixture model algorithm. These were subse-
quently evaluated on their pixel-wise performance in classifying
43 benign and 42 malignant histopathologically confirmed re-
gions of interest, using a prostate-based leave-one-out procedure. Results The combination of the spatiotemporal correlation (r),
mean transit time (μ), curve skewness (κ), and peak time (PT)
yielded an accuracy of 81% ± 11%, which was higher than the
best performing single parameters: r (73%), μ (72%), and wash-
in time (72%). The negative predictive value increased to
83% ± 16% from 70%, 69% and 67%, respectively. Pixel inclu-
sion based on the confidence level boosted these measures to
90% with half of the pixels excluded, but without disregarding
any prostate or region. • DCE-US can be used to extract both perfusion and
dispersion-related parameters. • Multiparametric DCE-US performs better in detecting PCa
than single-parametric DCE-US. • Multiparametric DCE-US might become a useful tool for
PCa localization. Keywords Prostate cancer . Ultrasound . Contrast agents . Classification . Multiparametric imaging Keywords Prostate cancer . Ultrasound . Contrast agents . Classification . Multiparametric imaging Introduction shunts cause an elevation in perfusion, the high tortuosity
and permeability, with rising interstitial pressure, lead to the
opposite effect [16, 17]. Prostate cancer (PCa) is the most prevalent form of cancer
among American men, representing 26% of the new cases
and approximately 10% of the cancer-related deaths [1]. Therefore, a reliable and minimally invasive diagnostic tool
for PCa is of paramount importance. During the last decade of
the 20th century, the introduction of prostate-specific antigen
(PSA) blood testing led to a dramatic increase of the number
of PCa diagnoses as well as a growing number of patients
exhibiting low-risk or indolent disease [2]. Overdiagnosis
and overtreatment are considered substantial problems due
to the limited positive predictive value (PPV) of screening
tools such as PSA level assessment and digital rectal exami-
nation [3, 4]. Therefore, the definitive diagnosis of PCa still
relies on ≥10 core systematic biopsy [5]. There is a high inci-
dence of biopsy-related complications [6, 7] and a consider-
able fraction of malignancies is identified only in repeat biop-
sy [8]. This stresses the demand for an imaging modality that
is able to localize or rule out prostatic malignancies. Such a
technique could eventually serve as a localization tool for
targeted biopsy [9, 10], or assist in patient selection and treat-
ment planning for organ-sparing focal therapy [11]. Tissue characterization by DCE-US thus requires a more
detailed assessment of the UCA kinetics in the prostate. UCAs
are composed of encapsulated micron-sized gas bubbles that
remain a few minutes in the vasculature [26–28], and their
behaviour can be assessed by looking at the time-intensity
curve (TIC), which is the evolution of echo intensity over time
in a certain area. For example, malignant areas in DCE-US
recordings are found to be marked by rapid and enhanced
inflow compared to similar benign regions in the prostate
[29, 30]. Several parameters have been extracted from the
TICs in order to mark relevant alterations in perfusion; these
are, e.g., the wash-in rate [31, 32], time to peak (PT) [33, 34],
time to appearance (AT) [33, 34], peak intensity (PI) [32–35],
and the area under the curve [33, 36]. In addition to these
perfusion-related parameters, angiogenic microvasculature
changes can also be detected by assessing UCA dispersion
[37]. Contrast US dispersion imaging (CUDI) was recently
developed to analyse the dispersion kinetics of a microbubble
contrast bolus in a DCE-US recording. Introduction The dispersive behav-
iour has been assessed with curve fitting [37], or similarity
analysis [38, 39]. In these methods, the TIC of each pixel in
the imaging plane is either fitted by a convective diffusion
model or compared to its neighbouring TICs, respectively. Currently, multiparametric magnetic resonance imaging
(mpMRI) seems the most promising imaging method for
PCa localization [12]. A recent meta-analysis reported an ap-
preciable average sensitivity of 74% and specificity of 88%,
with negative predictive values (NPVs) that ranged from 65%
to 94% [13]. In view of the advantages of transrectal ultra-
sound (TRUS) over MRI in terms of costs, time, resolution,
and practicality at bedside, this paper proposes a
multiparametric approach of TRUS. Though different in nature, the parameters acquired with these
methods were shown to have appreciable levels of sensitivity and
specificity for PCa detection. Hence, we hypothesize that a com-
bination of complementary parameters related to perfusion and
dispersion allows us to localize prostatic carcinoma with an even
higher level of accuracy. PCa is a multifocal and heterogeneous
disease whose appearance depends on cancer type, grade and
topography [24]. A multiparametric approach reduces the risk
of missing tumours that are invisible to one of the parameters
and may be able to discriminate prostatic diseases that mimic
malignant characteristics, like prostatitis [40]. Unfortunately, B-mode TRUS is not considered sufficient-
ly accurate for stand-alone tumour detection [14, 15]. Since
clinically relevant prostatic malignancies are characterized by
angiogenesis and neovascularization [16, 17], increased per-
fusion has been proposed as a marker for PCa. Moreover, it
was observed that the microvascular density correlates with
cancer aggressiveness [18]. However, Doppler imaging was
not found sufficiently accurate to capture these vascular
changes due to its limited sensitivity for small flows [19,
20]. The use of intravenously injected ultrasound contrast
agents (UCAs) in the TRUS procedure, that is, dynamic
contrast-enhanced ultrasound (DCE-US) imaging, also allows
the visualization of the vascular fraction and perfusion. Contrast-specific imaging modes are even able to image mi-
crocirculation at a capillary scale [21]. Again, despite the re-
ported improvements in tumour detection rate [22, 23],
targeted biopsies based on visual interpretation of contrast-
enhanced US alone are not considered viable to replace sys-
tematic biopsy [9, 24]. This might be explained by the incon-
sistent, ambiguous effect of angiogenesis on blood flow [25]. Abbreviations in time (72%). The negative predictive value increased to
83% ± 16% from 70%, 69% and 67%, respectively. Pixel inclu-
sion based on the confidence level boosted these measures to
90% with half of the pixels excluded, but without disregarding
any prostate or region. Conclusions Our results suggest multiparametric DCE-US anal-
ysis might be a useful diagnostic tool for PCa, possibly
Abbreviations
AT
Appearance time
CUDI
Contrast-ultrasound dispersion imaging
DCE-US
Dynamic contrast-enhanced ultrasound
FWHM
Full width half mximum
GMM
Gaussian mixture model
LDRW
Local density random walk
mpMRI
Multiparametric magnetic resonance imaging
NPV
Negative predictive value
PCa
Prostate cancer
PPV
Positive predictive value
PSA
Prostate-specific antigen
PT
Peak time
ROC
Receiver operating characteristic
ROI
Region of interest
TIC
Time-intensity curve
TRUS
Transrectal ultrasound
UCA
Ultrasound contrast agent
WIT
Wash-in time
* Rogier R. Wildeboer
r.r.wildeboer@tue.nl
1
Laboratory of Biomedical Diagnostics, Department of Electrical
Engineering, Eindhoven University of Technology, PO-Box 513,
5600 MB Eindhoven, The Netherlands
2
Department of Urology, Academic Medical Center University
Hospital, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands
3
Philips Research, Philips, Eindhoven, The Netherlands Conclusions Our results suggest multiparametric DCE-US anal-
ysis might be a useful diagnostic tool for PCa, possibly * Rogier R. Wildeboer
r.r.wildeboer@tue.nl Eur Radiol (2017) 27:3226–3234 3227 Histopathological analysis After prostatectomy, the prostates were fixed with a formalin
solution, deprived of the seminal vesicles, sectioned in slices
of ~4 mm and examined by the pathologist as described by
Montironi et al. [47]. The tumours were subsequently delin-
eated by the pathologist. Prior to the parametric analysis,
~0.5 cm2-sized ROIs were manually drawn on the B-mode
US scans to identify areas containing histologically confirmed
malignancy. Histological images were matched to the US
scans based on the position of the imaging plane in a trans-
versal sweep video performed before the contrast recordings. ROIs were only drawn in areas where the histopathologic
information persisted in the two adjacent slices. The prostatic
boundary was used to aid the localization. In the same way, we
positioned similarly sized ROIs in areas that were not depicted
as malignant. This resulted in a dataset containing about
174,000 pixels extracted from 85 ROIs in 45 DCE-US record-
ings in 19 patients, marked as either benign (43 ROIs) or
malignant (42 ROIs). In the end, we were able to include
one to four DCE-US imaging planes per patient. Data acquisition Nineteen PCa patients that were scheduled for radical prosta-
tectomy underwent a transrectal DCE-US scan prior to sur-
gery. The procedures were approved by the local ethics com-
mittee and carried out at the Academic Medical Center
(Amsterdam, The Netherlands). All patients signed an in-
formed consent. Patients below the age of 18 or with contra-
indications for the administration of contrast agents as defined
by the European Medicines Agency were excluded. Patients
with a tumour of Gleason score ≥3 + 3 and a size suitable for
our analysis (see section BHistopathological Analysis^) were
selected for this study. The patient and tumour characteristics
are summarized in Table 1. For the procedure, 2.4 mL of a SonoVue® UCA
microbubble suspension (Bracco, Milan, Italy) was intrave-
nously administered. This suspension consists of encapsulated
sulphur hexafluoride bubbles with an average diameter of
2.5 μm [46]. Two-minute recordings were subsequently per-
formed with an iU22 US scanner (Philips Healthcare, Bothell,
WA, USA), generally using a 3-MHz to 10-MHz-ranged
endocavity US probe (C10-3v). For one patient, an endocavity
probe with a range from 4 MHz to 8 MHz (C8-4v) was used Introduction Whereas diminished vasomotor control and formation of Since the performance of a multiparametric approach
is not dependent on a single threshold, it cannot be eval-
uated using conventional receiver operating characteristic
(ROC) analysis [41, 42]. Instead, a multiparametric ap-
proach requires a classification algorithm to combine the
parameters into a single parametric map. Many algo-
rithms have been used in biomedicine for classification;
in particular, Gaussian mixture models (GMMs), support
vector machines and artificial neural networks have been
extensively employed [43–45]. GMMs have been chosen
for our multiparametric evaluation as these are fast,
purely based on data (no need for additional physical
modelling) and facilitate the definition of classification
confidence. Moreover, GMMs were reported to perform
better than neural networks in mammographic tumour
identification [45]. 3228 Eur Radiol (2017) 27:3226–3234 due to availability issues. The measurements were carried out
in contrast-specific mode based on a power modulation pulse
scheme at a frequency of 3.5 MHz and with a mechanical
index of 0.06 to minimize bubble disruption. In this paper, we take into account thirteen perfusion- and
dispersion-related parameters as well as the echo intensity on
the TRUS image. Retrospectively, the most useful parameters
are selected and combined using histopathologically deter-
mined regions of interest (ROIs) in order to improve the ac-
curacy of DCE-US for the localization of PCa. Data processing and parameter extraction To assess the dispersive UCA behaviour, the DCE-US record-
ings were analysed by TIC fitting as well as by similarity
analysis. The parametric maps were produced using a
custom-made CUDI program running in Matlab® (2015b,
Mathworks, Natick, MA, USA) [48]. Following the method
described in [37], the data were pre-processed and the extract-
ed TICs were fitted by a modified local density random walk
(LDRW) model. This allowed us to estimate the area under the
curve (α), the mean transit time (μ), the skewness parameter
(κ), and the ratio between the diffusive and convective time
(λ = μκ). In addition, we looked at the variance (var) which is
the second moment of the curve [49], and the fitting interval
(int) between the PT and the truncation time where UCA re-
circulation occurs [37]. Table 1
List of patient and tumour characteristics
Characteristic
Mean
Median
Range
Age (yrs)
62.7
64
52 – 73
PSA level (ng/mL)
8.7
7.3
2.9 – 31.9
Prostate volume (mL)
35.8
30
20 – 83.5
Number
Clinical stage
T1
0
T2
11
T3
8
Gleason score
3 + 3
6
3 + 4
7
4 + 3
4
3 + 5
1
4 + 5
1 Table 1
List of patient and tumour characteristics For the similarity analysis, the data were pre-processed as
described in [39]. The spectral coherence (ρ) [38] and the
spatiotemporal correlation coefficient (r) [39] were then cal-
culated. For these parameters, the TIC of a single pixel is
compared to those in a ring-shaped kernel of 1.0 to 2.5 mm
in radius, as this size allows us to visualize similarity on the
scale of early angiogenesis [38]. Based on the processing in [37], we also extracted the PI,
the AT (where the TIC reaches 5% of the PI) and the PT (the
time where the intensity is the highest). In addition, the wash- Eur Radiol (2017) 27:3226–3234 3229 To evaluate the performance of the classifier, the procedure
was tested on each of the prostates whilst using the observa-
tions in other prostates as the training set. The outcomes of this
leave-one-out analysis were averaged over all prostates. We
quantified the classification performance by computing the
accuracy, sensitivity, specificity, PPV and NPV [41, 42]. Data processing and parameter extraction Whereas sensitivity and specificity indicate the percentage of
correctly classified malignant and benign pixels, respectively,
PPV and NPV reflect the percentage of pixels classified re-
spectively as malignant and benign that were correct. In addi-
tion, accuracy represents the overall correct classification per-
centage. Since the NPVis of paramount importance in order to
avoid missing clinically relevant PCa lesions, we optimized
the classifier based on this measure as well as on the accuracy. in time (WIT, the time period between AT and the point where
the TIC reaches 95% of the PI), and the full width half maximum
(FWHM) were considered as parameters of interest. A full list of
the investigated parameters is reported in Table 2. in time (WIT, the time period between AT and the point where
the TIC reaches 95% of the PI), and the full width half maximum
(FWHM) were considered as parameters of interest. A full list of
the investigated parameters is reported in Table 2. Classification procedure Gaussian mixture modelling is a widely known approach in clus-
ter analysis [50, 51] and data classification [52, 53]. GMMs
describe a set of observations (i.e. pixels) in (multi)parametric
space by a mixture of normal distributions. Using two pre-
determined training subsets of benign and malignant observa-
tions, the class-specific probability distributions can be comput-
ed. Subsequently, each pixel in the test set is classified according
to these distributions. We define a measure for the confidence of
classification by comparing the probabilities, p, of the observa-
tion being benign or malignant. This confidence level, P, conve-
niently ranging from 0 to 1, is described by P ¼
2pA
pAþpB −1, where
A denotes the class with highest probability. Pixel exclusion The TIC measurement quality usually differs from pixel to
pixel. To ensure the quality of the classification, it is important
to identify pixels that are likely to be misclassified. As stated,
the classification algorithm indicates the confidence of the
classification by the level P. Also, we consider the coefficient
of determination, R2, describing how well the TIC can be
fitted by the modified LDRW model, and the absolute proba-
bility of an observation belonging to its class. Large healthy
vessels are more likely to show early arrival of the bolus,
which complicates the use of perfusion parameters to mark
malignancy. Therefore, we also evaluated the classifier’s per-
formance after excluding the pixels with the lowest PT for
each plane. The GMM classification algorithm was implemented in
Matlab® using the statistical analysis toolbox. The parameters
were normalized to the 90th percentile to ensure equal
weighting, and training was performed using an iterative
expectation-maximization algorithm [54]. Since the GMM al-
gorithm is very fast, it was feasible to evaluate all possible
combinations of one to four distinct parameters. We did not
take into account more than four parameters to avoid
overfitting. Table 2
Full list of the parameters considered for multiparametric
analysis, with symbols and units Table 2
Full list of the parameters considered for multiparametric
analysis, with symbols and units
Symbol
Parameter name
Unit
B-mode ultrasound
Greylevel
Echo intensity
a.u. Contrast-enhanced ultrasound
WIT
Wash-in time
s
AT
Appearance time
s
PT
Peak time
s
PI
Peak intensity
a.u. FWHM
Full width half maximum
s
Fitting analysis
κ
Skewness parameter
s-1
μ
Mean transit time
s
λ
Convective-diffusion ratio
-
α
Area under the curve
a.u. var
Variance
a.u. int
Interval time
s
Similarity analysis
ρ
Spectral coherence
-
r
Correlation coefficient
- Results Based on accuracy, the combination of r, μ, κ, and PTyielded the
best accuracy (mean ± standard deviation = 81 ± 11%). The
highest NPV was found for the parameters var, μ, r, and int
(87 ± 15%), but with two of the other performance measures
being inferior compared to first set. The high NPVand sensitivity
can be explained by a low number of false negatives. We found
the parameter distributions best described by a single Gaussian
function per variable. The outcomes were compared to the per-
formance of individual parameters. The best-performing param-
eters of all three analyses—μ for curve fitting, r for similarity
analysis and WIT for conventional perfusion analysis—were
evaluated by a ROC-based threshold optimization as well as
GMM classification in one-dimensional parametric space. As
shown in Table 3, the multiparametric classification has a higher
performance than the ROC-analysed single parameters, irrespec-
tive of the measure used. Since not all individual parameters are
well described by a single Gaussian distribution, a non-tailored
GMM approach for the single parameters yielded less stable
results in terms of the balance between sensitivity and specificity. 3230 Eur Radiol (2017) 27:3226–3234 Table 3
Performance of the classification methods using specified
parameters
ROC Analysis
Single GMM
Multiparametric
GMM
WIT
μ
r
WIT
μ
r
r, μ,
κ, PT
var, μ,
r, int
Accuracy (%)
72
72
73
73
71
67
81
72
Sensitivity (%)
75
74
71
88
90
65
79
90
Specificity (%)
68
70
75
51
47
71
80
50
PPV (%)
76
75
76
70
63
70
85
65
NPV (%)
67
69
70
84
85
72
83
87
The used abbreviations are listed in the Abbreviations List and Table 2 Table 3
Performance of the classification methods using specified
parameters malignant pixels, as well as over patients and regions; exclu-
sion would, therefore, not result in extra PCa foci being
missed. To illustrate the results of whole-prostate classification and
the effect of pixel exclusion, the US and classification maps as
well as histological images of two patients are shown in
Figure 3. For the first set in Figure 2, the accuracy and NPV
have grown from 81 ± 11% to 90 ± 10% and from 83 ± 16% to
91 ± 13% with 51 ± 17% of the pixels remaining. Discussion According to today’s guidelines [5], reliable PCa diagnosis
requires a ≥10 core systematic biopsy under US guidance
and local anaesthesia. In recent years, an increasing emphasis
has been laid on imaging and targeted biopsy [9] in view of the
number of reported complications [6], over-diagnoses due to
the overestimation of pathologically insignificant lesions [3],
and under-diagnoses due to small high-risk PCa foci being
missed [8]. Contrast-enhanced US allows the extraction of
multiple parameters that have potential to serve as a diagnostic
marker for malignancy. The presented multiparametric ap-
proach combines perfusion-related parameters from conven-
tional DCE-US and dispersion-related parameters from CUDI
by means of a GMM classifier. Even though accuracy and NPVare considered the most impor-
tant performance measures, a reliable technique requires the other
measures to be sufficiently high as well. In the BMaterials and methods^ section, we mentioned
feature-based exclusion of pixels to decrease the number of
misclassifications in the multiparametric map. Figure 1 shows
the changes in accuracy after pixel exclusion based on classi-
fication confidence P, absolute probability, R2, and PT. It re-
veals that P correctly reflects the confidence and that it is the
most suitable measure to identify pixels with a high risk of
misclassification. The results of the multiparametric classifi-
cation after pixel exclusion based on this measure are depicted
in Figures 2a and b for a parameter set containing r, μ, κ, and
PT and a set containing var, μ, r, and int, respectively. The
exclusion of pixels is equally distributed over benign and The optimal subset of parameters comprises r, μ, κ, and PT
and thus features parameters from all analysis methods. Of
these parameters, r contributes most to the outcome, which is
consistent with previous publications on CUDI [39]. As μ and
κ jointly describe the shape of LDRW-modelled TIC [55], it is
not surprising the combination of these two has the greatest
added value to r. Finally, the addition of PT offers a slight
improvement to the accuracy (80% to 81%). Though early
enhancement is a strong marker of malignancy, the quantitative
use of the PT is normally complicated by its strong dependence
on operator and circulation time [34]. In combination with the
other parameters, however, the PT is able to further delineate
malignant and benign regions. Despite its good performance as
a single parameter, the WIT is not included in the
multiparametric sets. Results These
values are 72 ± 10% to 90 ± 7% and 87 ± 15% to 89 ± 15%
for the second set, again with 51 ± 14% of the pixels included. These exclusion percentages include the 4.5 ± 3.2% of pixels
that could not be fitted by the LDRW model. For reference,
Figure 4 depicts the individual, normalized parametric maps
that contribute to a multiparametric image. The best performing perfusion-related parameter as well as best parame-
ters of curve fitting and similarity analysis were evaluated using ROC
analysis and single-parameter GMM. Multiparametric results are shown
of the parameter sets with the highest accuracy (r, μ, κ, PT) and NPV (var,
μ, r, int) Discussion We expect that this is the result of the high
correlation between WITand μ (Pearson’s r: 0.88), making this
parameter redundant after inclusion of μ. Fig. 1 Accuracy of the Gaussian mixture model classifier by exclusion
of pixels based upon their confidence level, coefficient of determination,
absolute probability, and peak time. The classifier was run using a set of
parameters containing r, μ, κ, and PT We have shown that observations that are likely to be
misclassified can be recognized by their low confidence level. Excluding low-confidence pixels from the multiparametric
map increases the reliability of the classification. A confidence Fig. 1 Accuracy of the Gaussian mixture model classifier by exclusion
of pixels based upon their confidence level, coefficient of determination,
absolute probability, and peak time. The classifier was run using a set of
parameters containing r, μ, κ, and PT Eur Radiol (2017) 27:3226–3234 3231 Fig. 2 Performance of the Gaussian mixture model classifier upon
exclusion of pixels with the lowest confidence using an increasing
confidence threshold for (a) r, μ, κ, PT and (b) var, μ, int, r. The bars
represent the percentage of benign (grey) and malignant (dark) pixels that
are still included; the lines represent the evolution of accuracy (blue
squares ■), NPV (red triangles ▲), and PPV (green circles ●) represent the percentage of benign (grey) and malignant (dark) pixels that
are still included; the lines represent the evolution of accuracy (blue
squares ■), NPV (red triangles ▲), and PPV (green circles ●) Fig. 2 Performance of the Gaussian mixture model classifier upon
exclusion of pixels with the lowest confidence using an increasing
confidence threshold for (a) r, μ, κ, PT and (b) var, μ, int, r. The bars Fig. 2 Performance of the Gaussian mixture model classifier upon
exclusion of pixels with the lowest confidence using an increasing
confidence threshold for (a) r, μ, κ, PT and (b) var, μ, int, r. The bars represent the percentage of benign (grey) and malignant (dark) pixels that
are still included; the lines represent the evolution of accuracy (blue
squares ■), NPV (red triangles ▲), and PPV (green circles ●) threshold of 0.5 leads to an average pixel exclusion of 36
± 15% per prostate (ranging from 8 to 67%). In general, the
pixel exclusion approach resulted in disregarding pixels in the
areas where benign regions border on malignant ones rather than pixels in specific prostates. Discussion As can be seen in Figure 2,
exclusion does not favour malignant or benign pixels specif-
ically. Figure 3 shows typical examples of classification maps. There is a high correspondence between the maps and histology, Fig. 3 The B-mode transrectal
ultrasound, confidence-weighted
classification image, exclusion-
classification images with a
threshold of P > 0.5 and
histopathological images of
patient A (a t/m d) and patient B,
(e t/m h). In the classification
images, red regions are classified
as malignant (i.e. suspicious) and
green regions as benign (i.e. not
suspicious). In the
histopathological images,
malignant areas are indicated with
red. Parameters: r, μ, κ, and PT. ROIs are shown in overlay to the
B-mode images 3232 Eur Radiol (2017) 27:3226–3234 Fig. 4 Example of the four
normalized parametric maps that
serve as input for the best
performing multiparametric map
as shown below. All maps overlay
the B-mode TRUS image. Red
regions are classified as malignant
(i.e. suspicious) and green regions
as benign (i.e. not suspicious) of
which the transparency is scaled
with the confidence level. The
histology slice with tumour tissue
marked red is shown in the upper
right corner of the
multiparametric image Fig. 4 Example of the four
normalized parametric maps that
serve as input for the best
performing multiparametric map
as shown below. All maps overlay
the B-mode TRUS image. Red
regions are classified as malignant
(i.e. suspicious) and green regions
as benign (i.e. not suspicious) of
which the transparency is scaled
with the confidence level. The
histology slice with tumour tissue
marked red is shown in the upper
right corner of the
multiparametric image of malignancies with varying aggressiveness) leads to a higher
risk of misclassification in prostates containing very low-grade or
very high-grade PCa, as these are the most different from the
training set average. even though small regions remain misclassified after pixel exclu-
sion. Due to the preservation of the correctly classified malignant
regions, these results are clinically relevant for, e.g. targeted bi-
opsy. Because the current analysis faces limitations with respect
to the registration of US imaging planes with histology slices, we
made use of histologically proven ROIs. These ROIs were ~0.5-
cm2 sized, which resembles the critical size of clinically relevant
foci [47]. In the future, three-dimensional US models would
enable us to apply more accurate registration. Fig. 4 Example of the four
normalized parametric maps that
serve as input for the best
performing multiparametric map
as shown below. All maps overlay
the B-mode TRUS image. Red
regions are classified as malignant
(i.e. suspicious) and green regions
as benign (i.e. not suspicious) of
which the transparency is scaled
with the confidence level. The
histology slice with tumour tissue
marked red is shown in the upper
right corner of the
multiparametric image Discussion Loeb S, Vellekoop A, Ahmed HU et al (2013) Systematic review of
complications of prostate biopsy. Eur Urol 64:876–892 7. Loeb S, van den Heuvel S, Zhu X et al (2012) Infectious compli-
cations and hospital admissions after prostate biopsy in a european
randomized trial. Eur Urol 61:1110–1114 8. Ukimura O, Coleman JA, de la Taille A et al (2013) Contemporary
role of systematic prostate biopsies: indications, techniques, and
implications for patient care. Eur Urol 63:214–230 In conclusion, we see that combined evaluation of contrast-
enhanced ultrasonographic parameters has superior accuracy and
NPV in tumour localization compared to the individual paramet-
ric maps. The GMM-based multiparametric analysis is fast, ver-
satile and allows a reliable confidence estimation of its classifi-
cation. It was shown that pixel exclusion could boost the perfor-
mance even more without disregarding relevant areas of the
prostate. Like in computer-aided diagnosis of breast lesions
[66, 67], an extensive review of other algorithms is recommend-
ed to obtain an overview of the performance, advantages and
drawbacks of other classification methods for the detection of
PCa. In the future, parametric maps derived from other US mo-
dalities such as Doppler and elastography could also be included
[68], as well as the results from other diagnostic tools (e.g. PSA
assessment), but this is beyond the scope of the current study. Furthermore, this analysis is based on a small patient group, and
we recognize that a more extended validation is needed to derive
global measures for classification. We expect that this method
can also be employed to image other types of cancer. 9. van Hove A, Savoie P-H, Maurin C et al (2014) Comparison of
image-guided targeted biopsies versus systematic randomized biop-
sies in the detection of prostate cancer: a systematic literature re-
view of well-designed studies. World J Urol 32:847–858 10. Heijmink SW, van Moerkerk H, Kiemeney LALM et al (2006) A
comparison of the diagnostic performance of systematic versus
ultrasound-guided biopsies of prostate cancer. Eur Radiol 16:927–938 11. Ahmed HU, Moore C, Emberton M (2009) Minimally-invasive
technologies in uro-oncology: the role of cryotherapy, HIFU and
photodynamic therapy in whole gland and focal therapy of localised
prostate cancer. Surg Oncol 18:219–232 12. Scheenen TWJ, Rosenkrantz AB, Haider MA, Fütterer JJ (2015)
Multiparametric magnetic resonance imaging in prostate cancer
management: current status and future perspectives. Invest Radiol
50:594–600 13. Discussion The current performance is limited by the small training set in
this study, hampering the possibility to define subgroups accord-
ing to Gleason score. Since the microvascular density is a viable
marker in the staging of PCa [18], this might allow us to distin-
guish low-risk and high-risk PCa. Studies in contrast-enhanced
MRI suggest that perfusion-based discrimination between PCa
grades, and even prostatitis, is possible [60]. In this study, we did
not assess the differences in tumour classification of grades; an
extended dataset would allow such validation in the future. Another limitation of the study concerns the analysis of small
foci, which, considering the error margin in appointing the ROIs,
could not be included in the study. In the current analysis, 9 (20%) of the 44 regions predomi-
nantly (>50% of the pixels) classified as negative were
misclassified. In Table 1, Gleason score 3 + 3 represents most
of these false negative regions. The Gleason score, which com-
prises the grade of the two most prevalent histological patterns
found in a stained prostate tissue slice, is an indicator of the stage
and aggressiveness of prostatic carcinoma [56, 57]. Following
recent consensus in prostate grading, Gleason rate score 3 + 3 is
rated as grade group 1, indicating very low-risk disease with high
survival rates and virtually no chance of metastasis [58, 59]. We
believe that the use of a diverse training set (i.e. a set that consists DCE-US is not only a valuable modality for diagnosis of
PCa; its use is increasingly mentioned as a tool to monitor the
therapeutic effect of focal therapy. For instance, DCE-US was
found to map tissue devascularisation as well as DCE-MRI Eur Radiol (2017) 27:3226–3234 3233 4. Bangma CH, Roemeling S, Schröder FH (2007) Overdiagnosis and
overtreatment of early detected prostate cancer. World J Urol 25:3–9 after interstitial laser therapy [61, 62]. DCE-US has, therefore,
been used for interstitial laser therapy [63] and high-intensity
focused US treatment to visualize viable and devascularized
regions [64, 65]. Apart from the classifier’s aid for tumour
localization and monitoring, the ability of such classifier to
discriminate low-risk and high-risk disease would especially
improve the application of focal therapy and active surveil-
lance strategies. 5. Heidenreich A, Bastian PJ, Bellmunt J et al (2014) EAU guidelines
on prostate cancer. Part 1: screening, diagnosis, and local treatment
with curative intent—update 2013. Eur Urol 65:124–137 6. Discussion de Rooij M, Hamoen EHJ, Fütterer JJ et al (2014) Accuracy of
multiparametric MRI for prostate cancer detection: a meta-analysis. Am J Roentgenol 202:343–351 14. Loch AC, Bannowsky A, Baeurle L et al (2007) Technical and
anatomical essentials for transrectal ultrasound of the prostate. World J Urol 25:361–366 15. Smeenge M, de la Rosette JJMCH, Wijkstra H (2012) Current
status of transrectal ultrasound techniques in prostate cancer. Curr
Opin Urol 22:297–302 Acknowledgements
The scientific guarantor of this publication is
Massimo Mischi. The authors of this manuscript Rogier Wildeboer and
Maarten Kuenen declare relationships with the following company:
Philips. This study has received funding by the European Research
Council Starting Grant (#280209), an unrestricted grant from the Dutch
Cancer Society (#UVA2013-5941) and the IMPULS2-program within
Eindhoven University of Technology in collaboration with Philips. No
complex statistical methods were necessary for this paper. Institutional
Review Board approval was obtained. Written informed consent was
obtained from all patients in this study. Methodology: retrospective, ex-
perimental, performed at one institution 16. Russo G, Mischi M, Scheepens Wet al (2012) Angiogenesis in prostate
cancer: onset, progression and imaging. BJU Int 110:794–808 17. Carmeliet P, Jain RK (2000) Angiogenesis in cancer and other
diseases. Nature 407:249–257 18. Brawer MK (1996) Quantitative microvessel density: a staging and
prognostic marker for human prostatic carcinoma. Cancer 78:345–349 19. Halpern EJ, Strup SE (2000) Using gray-scale and color and power
doppler sonography to detect prostatic cancer. Am J Roentgenol
174:623–627 20. Mitterberger MJ, Aigner F, Horninger W et al (2010) Comparative
efficiency of contrast-enhanced colour Doppler ultrasound targeted
versus systematic biopsy for prostate cancer detection. Eur Radiol
20:2791–2796 Open Access This article is distributed under the terms of the Creative
Commons Attribution 4.0 International License (http://
creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided you give appro-
priate credit to the original author(s) and the source, provide a link to the
Creative Commons license, and indicate if changes were made. 21. Cosgrove D (2003) Angiogenesis imaging–ultrasound. Br J Radiol
60:324–330 22. Halpern EJ, McCue PA, Aksnes AK et al (2002) Contrast-enhanced
US of the prostate with sonazoid: comparison with whole-mount
prostatectomy specimens in 12 patients. Radiology 222:361–366 23. Matsumoto K, Nakagawa K, Hashiguchi A et al (2010) Contrast-
enhanced ultrasonography of the prostate with Sonazoid. Jpn J Clin
Oncol 40:1099–1104 1.
American Cancer Society (2015) Cancer Facts and Figures 2015.
Am Cancer Soc References Postema AW, Frinking PJA, Smeenge M et al (2015) Dynamic
contrast-enhanced ultrasound parametric imaging for the detection
of prostate cancer. BJU Int 117:598–603 53. Grim J, Somol P, Haindl M, Daneš J (2009) Computer-aided eval-
uation of screening mammograms based on local texture models. Image Process IEEE Trans 18:765–773 32. Elie N, Kaliski A, Péronneau P et al (2007) Methodology for quanti-
fying interactions between perfusion evaluated by DCE-US and hyp-
oxia throughout tumor growth. Ultrasound Med Biol 33:549–560 54. Redner R, Walker H (1984) Mixture densities, maximum likelihood
and the EM algorithm. SIAM Rev 26:195–239 33. Eckersley RJ, Sedelaar JP, Blomley MJK et al (2002) Quantitative
microbubble enhanced transrectal ultrasound as a tool for monitor-
ing hormonal treatment of prostate carcinoma. Prostate 51:256–267 55. Strouthos C, Lampaskis M, Sboros Vet al (2010) Indicator dilution
models for the quantification of microvascular blood flow with
bolus administration of ultrasound contrast agents. Ultrason
Ferroelectr Freq Control IEEE Trans 57:1296–1310 34. Tang J, Yang J-C, Luo Y et al (2008) Enhancement characteristics
of benign and malignant focal peripheral nodules in the peripheral
zone of the prostate gland studied using contrast-enhanced
transrectal ultrasound. Clin Radiol 63:1086–1091 56. Gleason DF (1992) Histologic grading of prostate cancer: a per-
spective. Hum Pathol 23:273–279 57. Humphrey PA (2004) Gleason grading and prognostic factors in
carcinoma of the prostate. Mod Pathol 17:292–306 35. Qi TY, Chen YQ, Jiang J et al (2013) Contrast-enhanced transrectal
ultrasonography: Measurement of prostate cancer tumor size and cor-
relation with radical prostatectomy specimens. Int J Urol 20:1085–1091 58. Kryvenko ON, Epstein JI (2015) Changes in prostate cancer grading:
Including a new patient-centric grading system. Prostate 76:427–433 36. Eckersley RJ, Cosgrove DO, Blomley MJ, Hashimoto H (1998)
Functional imaging of tissue response to bolus injection of ultra-
sound contrast agent. In: Ultrason. Symp. 1998. Proceedings., 1998
IEEE. IEEE, pp 1779–1782 59. Epstein JI, Egevad L, Amin MB et al (2016) The 2014 International
Society of Urological Pathology (ISUP) consensus conference on
Gleason grading of prostatic carcinoma: definition of grading patterns
and proposal for a new grading system. Am J Surg Pathol 40:244–252 37. Kuenen MPJ, Mischi M, Wijkstra H (2011) Contrast-ultrasound
diffusion imaging for localization of prostate cancer. IEEE Trans
Med Imaging 30:1493–1502 60. References 24. Li Y, Tang J, Fei X, Gao Y (2013) Diagnostic performance of
contrast enhanced ultrasound in patients with prostate cancer: a
meta-analysis. Acad Radiol 20:156–164 1. American Cancer Society (2015) Cancer Facts and Figures 2015. Am Cancer Soc 2. Cooperberg MR, Lubeck DP, Meng MVet al (2004) The changing
face of low-risk prostate cancer: trends in clinical presentation and
primary management. J Clin Oncol 22:2141–2149 25. Delorme S, Knopp MV (1998) Non-invasive vascular imaging:
assessing tumour vascularity. Eur Radiol 8:517–527 26. Trabulsi EJ, Sackett D, Gomella LG, Halpern EJ (2010) Enhanced
transrectal ultrasound modalities in the diagnosis of prostate cancer. Urology 76:1025–1033 3. Loeb S, Bjurlin MA, Nicholson J et al (2014) Overdiagnosis and
overtreatment of prostate cancer. Eur Urol 65:1046–1055 3. Loeb S, Bjurlin MA, Nicholson J et al (2014) Overdiagnosis and
overtreatment of prostate cancer. Eur Urol 65:1046–1055 3234 Eur Radiol (2017) 27:3226–3234 27. Wink M, Frauscher F, Cosgrove D et al (2008) Contrast-enhanced
ultrasound and prostate cancer; a multicentre european research
coordination project. Eur Urol 54:982–993 48. Schalk SG, Demi L, Smeenge M et al (2015) 4-D Spatiotemporal
analysis of ultrasound contrast agent dispersion for prostate cancer
localization : a feasibility study. IEEE Trans Ultrason Ferroelectr
Freq Control 62:839–851 28. Halpern EJ, Verkh L, Forsberg F et al (2000) Initial experience with
contrast-enhanced sonography of the prostate. Am J Roentgenol
174:1575–1580 49. Mischi M, Kalker A, Korsten HHM (2004) Moment method for the
local density random walk model interpolation of ultrasound contrast
agent dilution curves. In: Proc. 17th Int. EURASIP Conf. BIOSIGNAL
2004, Brno (Czech Rep.), Jun. 23–25. Citeseer, pp 33–35 29. Seitz M, Gratzke C, Schlenker B et al (2011) Contrast-enhanced
transrectal ultrasound (CE-TRUS) with cadence-contrast pulse se-
quence (CPS) technology for the identification of prostate cancer. Urol Oncol Semin Orig Investig 29:295–301 2004, Brno (Czech Rep.), Jun. 23–25. Citeseer, pp 33–35 50. Fraley C, Raftery AE (2002) Model-based clustering, discriminant
analysis, and density estimation. J Am Stat Assoc 97:611–631 51. Banfield JD, Raftery AE (1993) Model-based Gaussian and non-
Gaussian clustering. Biometrics 49:803–821 30. Qi T, Chen Y, Zhu Y et al (2014) Contrast-enhanced transrectal
ultrasonography for detection and localization of prostate index
tumor: correlation with radical prostatectomy findings. Urology
84:138–143 52. Han SH, Ackerstaff E, Stoyanova R et al (2013) Gaussian mixture
model-based classification of dynamic contrast enhanced MRI data
for identifying diverse tumor microenvironments: preliminary re-
sults. NMR Biomed 26:519–532 31. References Franiel T, Lüdemann L, Rudolph B et al (2008) Evaluation of nor-
mal prostate tissue, chronic prostatitis, and prostate cancer by quan-
titative perfusion analysis using a dynamic contrast-enhanced in-
version-prepared dual-contrast gradient echo sequence. Invest
Radiol 43:481–487 38. Mischi M, Kuenen MPJ, Wijkstra H (2012) Angiogenesis imaging
by spatiotemporal analysis of ultrasound contrast agent dispersion
kinetics. IEEE Trans Ultrason Ferroelectr Freq Control 59:621–629 61. Atri M, Gertner MR, Haider MA et al (2009) Contrast-enhanced ultra-
sonography for real-time monitoring of interstitial laser thermal therapy
in the focal treatment of prostate cancer. Can Urol Assoc J 3:125–130 39. Kuenen MPJ, Saidov TA, Wijkstra H et al (2013) Spatiotemporal
correlation of ultrasound contrast agent dilution curves for angio-
genesis localization by dispersion imaging. IEEE Trans Ultrason
Ferroelectr Freq Control 60:2665–2669 62. van den Bos W, Bruin DM, Randen A et al (2016) MRI and contrast-
enhanced ultrasound imaging for evaluation of focal irreversible elec-
troporation treatment: results from a phase I-II study in patients under-
going IRE followed by radical prostatectomy. Eur Radiol 26:2252–
2560 40. Clements R (2002) The role of transrectal ultrasound in diagnosing
prostate cancer. Curr Urol Rep 3:194–200 41. Fawcett T (2006) An introduction to ROC analysis. Pattern
Recognit Lett 27:861–874 63. Colin P, Mordon S, Nevoux P et al (2012) Focal laser ablation of
prostate cancer: definition, needs, and future. Adv Urol 2012:589160 42. Metz CE (1978) Basic principles of ROC analysis. Semin Nucl
Med 8:283–298 64. Rouvière O, Glas L, Girouin N et al (2011) Prostate cancer ablation
with transrectal high-intensity focused ultrasound: assessment of tissue
destruction with contrast-enhanced US. Radiology 259:583–591 43. Chang R-F, Wu W-J, Moon WK et al (2003) Support vector ma-
chines for diagnosis of breast tumors on US images. Acad Radiol
10:189–197 65. Crouzet S, Rouviere O, Martin X, Gelet A (2014) High-intensity
focused ultrasound as focal therapy of prostate cancer. Curr Opin
Urol 24:225–230 44. Loch T, Leuschner I, Genberg C et al (1999) Artificial neural net-
work analysis (ANNA) of prostatic transrectal ultrasound. Prostate
39:198–204 66. Nattkemper TW, Arnrich B, Lichte O et al (2005) Evaluation of radio-
logical features for breast tumour classification in clinical screening
with machine learning methods. Artif Intell Med 34:129–139 45. Acharya R, Ng EYK, Chang YH et al (2008) Computer-based
identification of breast cancer using digitized mammograms. J
Med Syst 32:499–507 67. References Jalalian A, Mashohor SBT, Mahmud HR et al (2013) Computer-
aided detection/diagnosis of breast cancer in mammography and
ultrasound: a review. Clin Imaging 37:420–426 46. Greis C (2004) Technology overview: SonoVue. Eur Radiol Suppl
14:P11–P15 68. Postema A, Mischi M, de la Rosette J, Wijkstra H (2015)
Multiparametric ultrasound in the detection of prostate cancer: a
systematic review. World J Urol 33:1651–1659 47. Montironi R, van der Kwast T, Boccon-Gibod L et al (2003)
Handling and pathology reporting of radical prostatectomy speci-
mens. Eur Urol 44:626–636
|
https://openalex.org/W2332428787
|
https://europepmc.org/articles/pmc4814907?pdf=render
|
English
| null |
pH Induced Conformational Transitions in the Transforming Growth Factor β-Induced Protein (TGFβIp) Associated Corneal Dystrophy Mutants
|
Scientific reports
| 2,016
|
cc-by
| 14,438
|
pH Induced Conformational
Transitions in the Transforming
Growth Factor β-Induced Protein
(TGFβIp) Associated Corneal
Dystrophy Mutants received: 17 June 2015
accepted: 12 February 2016
Published: 31 March 2016 Elavazhagan Murugan1,2, Anandalakshmi Venkatraman1, Zhou Lei3, Victoria Mouvet1,
Rayne Rui Yi Lim1, Nandhakumar Muruganantham4, Eunice Goh4, Gary Swee Lim Peh1,2,
Roger W. Beuerman2,4,5,6, Shyam S. Chaurasia1,2,5, Lakshminarayanan Rajamani2,4,5 &
Jodhbir S. Mehta1,2,5,6 Most stromal corneal dystrophies are associated with aggregation and deposition of the mutated
transforming growth factor-β induced protein (TGFβIp). The 4th_FAS1 domain of TGFβIp harbors ~80%
of the mutations that forms amyloidogenic and non-amyloidogenic aggregates. To understand the
mechanism of aggregation and the differences between the amyloidogenic and non-amyloidogenic
phenotypes, we expressed the 4th_FAS1 domains of TGFβIp carrying the mutations R555W (non-
amyloidogenic) and H572R (amyloidogenic) along with the wild-type (WT). R555W was more
susceptible to acidic pH compared to H572R and displayed varying chemical stabilities with
decreasing pH. Thermal denaturation studies at acidic pH showed that while WT did not undergo
any conformational transition, the mutants exhibited a clear pH-dependent irreversible conversion
from αβ conformation to β-sheet oligomers. The β-oligomers of both mutants were stable at
physiological temperature and pH. Electron microscopy and dynamic light scattering studies showed
that β-oligomers of H572R were larger compared to R555W. The β-oligomers of both mutants were
cytotoxic to primary human corneal stromal fibroblast (pHCSF) cells. The β-oligomers of both mutants
exhibit variations in their morphologies, sizes, thermal and chemical stabilities, aggregation patterns
and cytotoxicities. Corneal Dystrophies are inherited protein aggregation disorders characterized by the deposition of misfolded
proteins aggregates in various layers of the cornea1–3. Most dystrophies in the corneal stromal region are asso-
ciated with the mutations in the transforming growth factor β -induced protein (TGFβ Ip). TGFβ Ip aggregation
and deposition occurs only in the cornea, though it is present in abundance in various connective tissues4–7. The
mature TGFβ Ip, a 660aa protein has an N-terminal cysteine-rich EMILIN-like (EMI) domain, four fascicilin-like
(FAS1) domains and an integrin-binding RGD motif at the C-terminus8. TGFβ Ip-associated corneal dystrophies
are phenotypically heterogeneous, inherited in an autosomal dominant manner1,9,10 and are classified as lattice,
granular, combined lattice and granular, Reis-Buckler and Thiel-Behnke corneal dystrophies1,2,11,12. So far, 64
single amino acid mutations associated with distinct phenotypes have been reported4,13,14. Among the four FAS1
domains of TGFβ Ip, the 1st and 4th FAS1 domains carry the disease related mutations, with ~80% of the muta-
tions residing in the 4th_FAS1 domain11. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports received: 17 June 2015
accepted: 12 February 2016
Published: 31 March 2016 Resultsf Effects of biochemical and biophysical factors on WT and mutants. The native 4th_FAS1 domain
(Fig. 1a) of TGFβ Ip (Genbank_ID-NM_000358; Protein_ID:-NP_000349) and the mutants R555W and H572R
were cloned and purified (Fig. 1b) as described previously21. The estimated pI values of the WT, R555W and
H572R domains were 6.53, 6.32 and 6.65 respectively. The amino acid substitutions (R→ W and H→ R) were
associated with changes in charge and hydrophobicity as listed in the table (Table 1). Effects of biochemical and biophysical factors on WT and mutants. The native 4th_FAS1 domain
(Fig. 1a) of TGFβ Ip (Genbank_ID-NM_000358; Protein_ID:-NP_000349) and the mutants R555W and H572R
were cloned and purified (Fig. 1b) as described previously21. The estimated pI values of the WT, R555W and
H572R domains were 6.53, 6.32 and 6.65 respectively. The amino acid substitutions (R→ W and H→ R) were
associated with changes in charge and hydrophobicity as listed in the table (Table 1). Effect of pH on the secondary structures of the WT and mutants. The CD spectra for the WT and
the mutants at pH 7 (Fig. 1c–e) showed negative minima in the n – π * region (222 nm) and a weak shoulder
at the π –π * region (207 nm) corresponding to their mixed α -helical and β -sheet conformations. Under acidic
conditions, there were discernible differences between the WT and mutants in their secondary structure. The
non-amyloidogenic R555W was more sensitive to pH compared to the WT and amyloidogenic H572R. The WT
(Fig. 1c) and H572R (Fig. 1e) remained unchanged under neutral and acidic pH. The R555W mutant displayed
an increase in the CD intensity at 222 nm and 207 nm with decrease in pH (Fig. 1d). CD intensities at 222 nm at
varying pH values (Fig. 1f) showed that R555W, was more sensitive to pH and the pH-response of H572R was
similar to the WT protein. We also incubated the mutants in acidic pH for 1 week and followed their aggregation/
oligomerization by ThT fluorescence (Supplementary Fig. S1). However, no significant conversion was observed
as seen from the corresponding CD spectra for WT and H572R. At pH 2.75, R555W showed a partial conversion
to β -sheet. Compared to the amyloid fibril peptide pN622K, almost no increase in fluorescence was observed for
WT and R555W. www.nature.com/scientificreports/ non-amyloidogenic deposits11–16. Both phenotypes display significant differences in morphology, aggregation and
tinctorial properties. However, the mechanisms adopted by these phenotypes in forming highly distinct ultras-
tructures, remains to be elucidated. It has been reported that the ability to form highly ordered aggregates such as amyloids, resides within the
polypeptide chains rather than the whole protein17. Hence, we chose to explore the crucial 4th_FAS1 domain
(135aa) as a representative of the full-length TGFβ Ip. Bioinformatics analyses of the aggregation propensities
of various regions of TGFβ Ip have also shown that the 4th_FAS1 domain harbors regions of high aggregation
propensities18,19. Also, homology-based modelling studies have shown that the 4th_FAS1 domain displays the
properties of the full-length TGFβ Ip20.f p
p
g
β p
In our study, we aim to delineate the differences between the amyloidogenic and the non-amyloidogenic
mutants and between the mutants and wild-type (WT) TGFβ Ip to explain the physiological variations exhibited
by these phenotypes. We had previously reported the cloning, expression and purification of the 4th_FAS1 domain
of four TGFβ Ip mutants21. Our studies indicated that under physiological pH, the mutants are more stable than
the WT. Here, we chose to examine the effects of factors like temperature and pH on the properties of the 4th_
FAS1 domains of TGFβ Ip harboring mutations of the amyloidogenic and non-amyloidogenic phenotypes. The
amyloidogenic H572R (LCDI/IIIA), discovered in Thai22 and Chilean populations23 with ages of onset ranging
around mid-twenties, is characterized by central sub-epithelial needle-like lattice lines and polymorphic anterior
stromal opacities. The highly ubiquitous non-amyloidogenic R555W (GCDI/II)24–26 appears as rod shaped or
granular bodies with sharp borders found in the central corneal stroma27,28. We aimed to examine the effects of
these charge modifying mutations on the domains, their aggregation, their sensitivity to pH and temperature. g
y
g
gg g
y
p
p
Aggregating proteins are found to be more susceptible to acidic pH and increase in temperature29,30. Even in
TGFβ Ip, it has been shown that the R124H mutation induces localization of TGFβ Ip to lysosomes with an acidic
environment31. In the present study, we have investigated the effects of acidic pH, denaturants and temperature
on the secondary structure and conformational stability of the domains. The cytotoxicities of the aggregates were
also studied in primary human corneal stromal fibroblasts (pHCSF). pH Induced Conformational
Transitions in the Transforming
Growth Factor β-Induced Protein
(TGFβIp) Associated Corneal
Dystrophy Mutants In lattice corneal dystrophies (LCD), the protein aggregates appear as
lattice lines or amyloid fibrils (amyloidogenic). In granular corneal dystrophies (GCD), they appear as granular, 1Tissue Engineering and Stem Cell Group, Singapore Eye Research Institute, Singapore. 2Duke-NUS Graduate
Medical School, Singapore. 3Proteomics and Microanalysis laboratory, Singapore Eye Research Institute,
Singapore. 4Ocular Chemistry and Anti-Infectives, Singapore Eye Research Institute, Singapore. 5Department of
Ophthalmology, Yong Loo Lin School of Medicine, NUS, Singapore. 6Singapore National Eye Centre, Singapore. Correspondence and requests for materials should be addressed to J.S.M. (email: jodmehta@gmail.com) Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 1 www.nature.com/scientificreports/ Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 Resultsf The slight increase in fluorescence corresponding to pH 2.75 and pH 3.25 for H572R did not
show a corresponding conversion in the CD spectra. Effect pH on thermal denaturation of WT and mutants. Conformational transition of the domains
undergoing thermal denaturation was examined by heating them from 20 °C to 70 °C, at neutral and acidic pH
conditions. At pH 7 and pH 8, for both the mutants and the WT, no well-defined transition was observed with
increasing temperatures (Fig. 2a–f). For both the mutants, the amplitude of the negative minima at 222 nm
decreased upon heating (Fig. 2c–f) and a weak hysteresis was observed when cooled. With acidic pH, while the
WT showed no apparent changes in the secondary structure with increasing temperature (Fig. 3a–c), both the
mutants displayed a clear transition from monomeric α /β -structure to β -sheet (Fig. 3d–i). Single wavelength
scan at 222 nm indicated a clear sigmoidal transition for both mutants under acidic conditions when compared to
the WT, which was unperturbed by the changes in pH and temperature (Fig. 4a–e). Thermal denaturation exper-
iments were also done by heating the domains from 20 °C to 90 °C (Supplementary Fig. S2). Though previous
studies have observed denaturation of WT above 60 °C 32, we did not observe any conformational transitions to
β -sheet even after heating to 90 °C. The mid-point of the normalized sigmoidal curve defines the transition tem-
perature (Tt) wherein the conversion of a monomeric α /β -structure to the β -structured oligomers was observed. The transition for the domains at different pH conditions was similar to the samples heated to 70 °C and the Tt lied
between 35–58 °C. Hence all the subsequent thermal denaturation experiments were performed by heating the
domains upto 70 °C. The non-amyloidogenic R555W showed a marked sensitivity to pH and displayed a higher
thermal instability compared to the H572R. Though H572R mutant displayed a clear pH-dependent conversion
to β -structure when heated, the Tt was higher than R555W. A difference in Tt of 5–12 °C is observed at various
pH conditions. Mild precipitation was observed in the samples after heating, which correlates with the change in Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 2 www.nature.com/scientificreports/ Figure 1. Biochemical and biophysical properties of the native 4th_FAS1 domains of the wild-type and
mutant TGFβIp. Resultsf (a) Schematic representation of the domain arrangement and boundaries of the full-length
TGFβ Ip and 4th_FAS1 domains of the wild-type, non-amyloidogenic (R555W) and amyloidogenic (H572R)
mutants used in the study. (b) SDS-PAGE gel showing the purified fractions of the 4th_FAS1 domains of the
WT, R555W and H572R mutants. (c–e) Far UV CD spectra of the 4th_FAS1 domains of WT (d), R555W (e) and
H572R (f) incubated for 16 hours at acidic pH conditions (pH 3, 4.5, 5.5 and 7). The R555W mutant displayed
clear changes in the CD spectra at 222 nm and 207 nm with decrease in pH (f). The CD intensity at 222 nm
decreased with decrease in pH confirming the unfolding of the secondary structures. The CD spectra for the
WT (c) and H572R (e) mutant remained almost unchanged. Plotting the intensities at 222 nm at varying pH
(g) showed that the non-amyloidogenic phenotype, R555W, was more sensitive to pH and the amyloidogenic
phenotype, H572R, remained more stable to pH changes at room temperature. Figure 1. Biochemical and biophysical properties of the native 4th_FAS1 domains of the wild-type and
mutant TGFβIp. (a) Schematic representation of the domain arrangement and boundaries of the full-length
TGFβ Ip and 4th_FAS1 domains of the wild-type, non-amyloidogenic (R555W) and amyloidogenic (H572R)
mutants used in the study. (b) SDS-PAGE gel showing the purified fractions of the 4th_FAS1 domains of the
WT, R555W and H572R mutants. (c–e) Far UV CD spectra of the 4th_FAS1 domains of WT (d), R555W (e) and
H572R (f) incubated for 16 hours at acidic pH conditions (pH 3, 4.5, 5.5 and 7). The R555W mutant displayed
clear changes in the CD spectra at 222 nm and 207 nm with decrease in pH (f). The CD intensity at 222 nm
decreased with decrease in pH confirming the unfolding of the secondary structures. The CD spectra for the
WT (c) and H572R (e) mutant remained almost unchanged. Plotting the intensities at 222 nm at varying pH
(g) showed that the non-amyloidogenic phenotype, R555W, was more sensitive to pH and the amyloidogenic
phenotype, H572R, remained more stable to pH changes at room temperature. intensities observed in the CD spectra. When the concentration of the sample was increased (from 0.6 mg/ml to
1.2 mg/ml~75 μM), the turbidity and precipitation increases. Urea denaturation of non-amyloidogenic R555W. The tryptophan residue in R555W allowed us to
measure the emission fluorescence. Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 Resultsf A significant decrease in free energy of unfold-
ing from 13.8 ± 0.8 kJ/mole to 7.108 ± 1.5 kJ/mole was observed from pH 7.0 to pH 4.5 for the R555W mutant. However, for the H572R mutant, fluorescence studies could not be performed because of the absence of a tryp-
tophan residue. Characterization of the β-oligomers of amyloidogenic and non-amyloidogenic mutants. The
amyloidogenic and non-amyloidogenic mutants displayed a clear transition to an all β -sheet conformation when
heated under acidic conditions. Conversion to an all β -sheet conformation could indicate β -oligomer forma-
tion17. A detailed investigation of the β -oligomers from two mutants was therefore performed to validate and
characterize the proposed β -oligomers. Confirmation of β-oligomers formed by the mutants and characterization of the β-oligomers by
TEM and DLS. TEM examination of the β -oligomers revealed that while no particles were visible for the WT
(data not shown), both the mutants displayed particles validating our proposal. The β -oligomers of the H572R
and R555W displayed varying sizes and morphologies (Fig. 6a,b). The β -oligomers of the non-amyloidogenic
R555W were homogeneous, displayed smoother edges and measured ~4–8 nm (mean diameter ~5.1 ± 1.79 nm)
(Fig. 6a). The β -oligomers of the amyloidogenic H572R were more heterogeneous, displayed rugged edges and
were larger, measuring 10–40 nm (mean diameter ~19.1 ± 4.9 nm) (Fig. 6b). The β -oligomers formed from the
amyloidogenic and the non-amyloidogenic phenotypes are distinctly different from each other. y
g
y
g
y
yf
Dynamic light scattering (DLS) allows the examination of apparent hydrodynamic radius (RH) of a protein in
solution33,34. DLS analysis on the β -oligomers prepared under acidic conditions show that H572R β -oligomers
exhibit large variations in their RH (~89.95 nm|pH 3.0, ~69 nm|pH 4.5 and ~155 nm|pH 5.5), while R555W β -oligomers
are more uniform and almost similar across various acidic pH conditions (~39.58 nm|pH 3.0, ~51.9 nm|pH 4.5
and ~68.2 nm|pH 5.5). The distribution curves from the % intensity plots also show the homogeneity of the
non-amyloidogenic R555W, and relative heterogeneity of the amyloidogenic H572R. This is in conjunc-
tion with the results obtained from TEM, where we see more homogenous and smaller β -oligomers from the
non-amyloidogenic R555W mutant and heterogeneous and relatively larger β -oligomers from the amyloidogenic
H572R mutant. The β-oligomers of the amyloidogenic H572R shows stronger binding to Thioflavin T
(ThT). Resultsf Examination of the emission fluorescence at ~332 nm of R555W in acidic pH Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 3 www.nature.com/scientificreports/ WT
R555W
H572R
Charge of the mutation
–
R ( + 1) → W(0)
H (0) → R(+1) (at pH 7.0)
H (+1) → R(+1) (at pH < 6.0)
Net Charge (at pH 7.0)
−0.6
− 1.6
0.2
Net Charge (at pH 5.5)
1.9
0.9
2.0
Change in hydrophobicity due to the mutation (17)
–
3.95 → − 2.13 (at pH 7.0 and pH < 6.0)
0.64 → 3.95 (at pH 7.0)
2.87 → 3.95 (pH < 6.0)
Table 1. Effects of mutation on the charge and hydrophobicity. The change in the individual charges and
overall net charge in the mutants. The change in hydrophobicities were calculated using the equation described
previously (17). Table 1. Effects of mutation on the charge and hydrophobicity. The change in the individual charges and
overall net charge in the mutants. The change in hydrophobicities were calculated using the equation described
previously (17). Table 1. Effects of mutation on the charge and hydrophobicity. The change in the individual charges and
overall net charge in the mutants. The change in hydrophobicities were calculated using the equation described
previously (17). showed a significant decrease in emission intensity with decreasing pH, however the emission maxima remained
unchanged (Fig. 5a). To obtain a better insight into the effect of pH on R555W, we monitored the urea-induced
unfolding of R555W. Increasing the urea concentration progressively shifts the emission maxima (~332 nm) to
longer wavelengths (~352 nm), suggesting a clear transition from folded to unfolded conformations (5b–e). To
confirm refolding of unfolded R555W, the unfolded protein at different pH conditions (pH 3.0, pH 4.5, pH 5.5
and pH 7.0) was diluted appropriately and emission spectra were recorded. The emission maxima (λ max) plot-
ted with the unfolded and refolded domains were superimposable (Supplementary Fig. S3). (Fig. 5b–e). A clear
reversal in fluorescence maxima from ~352 nm to ~332 nm was observed thereby allowing us to estimate the
thermodynamic stability of the mutant protein in various pH. Figure 5f–i shows urea denaturation curves plotted
as ‘fraction unfolded (yU) vs increasing urea concentrations’ as monitored by the changes in emission maxima
(Δλ max) at various pH values (Supplementary Fig. S3) for the R555W mutant. The thermodynamic parameters
derived from urea denaturation are shown in the table (Table 2). Resultsf Amyloid fibrils bind to the dye ThT and display an emission fluorescence at 485 nm35,36. In Alzheimer’s
disease, ThT binds to Aβ -oligomers themselves, and has been proposed for early diagnosis37. We wanted to test if
the TGFβ Ip β -oligomers bind to ThT. The amyloid forming TGFβ Ip peptide pN622K (pN622K611–633) displayed
a high fluorescence intensity when bound to ThT (Fig. 6c). The fluorescence displayed by WT was comparable to
the background fluorescence. Relatively, the β -oligomers of H572R showed significant fluorescence on binding to
ThT (**P < 0.01) and atleast ~3 times more fluorescence compared to R555W. While the fluorescence intensities
were much lower compared to the fibril forming peptide, it was significant that the β -oligomers of the amyloido-
genic H572R were able to bind to ThT and this could aid further characterization of the β -oligomers. The mutants display differences in their ‘aggregation hotspots’. To determine the regions with
high aggregation propensities or ‘aggregation hotspots’ within the mutants and the β -oligomers, the domains
were digested with trypsin and the resulting peptides were examined using LC-MS/MS. The peptide map gen-
erated (Supplementary data) following the insilico trypsin digestion displayed a series of peptides (Fig. 6d) Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 4 www.nature.com/scientificreports/ entificreports/
formed following trypsin digestion. We aimed to identify the regions that were probably buried within the
Figure 2. Thermal denaturation of the 4th_FAS1 domains of the WT and mutants at neutral and basic
pH. (a,b) Far UV CD spectra of the 4th_FAS1 domain of WT at pH 7.0 and pH 8.0 before heating (black), after
heating to 70 °C (red) and cooling back to 20 °C (blue). (c,d) Far UV CD spectra of the 4th_FAS1 domains of
R555W at pH 7 (c) and pH 8 (d) before heating (black) and after heating to 70 °C (red) and cooling back to 20 °C
(blue). (e,f) Far UV CD spectra of the 4th_FAS1 domains of H572R a pH 7.0 (e) and pH 8.0 (f) before heating
(black) and after heating to 70 °C (red) and cooling back to 20 °C (blue). The WT and the mutants did not
display any significant changes in structure at pH 7.0 and pH 8.0. Figure 2. Thermal denaturation of the 4th_FAS1 domains of the WT and mutants at neutral and basic
pH. Resultsf (a,b) Far UV CD spectra of the 4th_FAS1 domain of WT at pH 7.0 and pH 8.0 before heating (black), after
heating to 70 °C (red) and cooling back to 20 °C (blue). (c,d) Far UV CD spectra of the 4th_FAS1 domains of
R555W at pH 7 (c) and pH 8 (d) before heating (black) and after heating to 70 °C (red) and cooling back to 20 °C
(blue). (e,f) Far UV CD spectra of the 4th_FAS1 domains of H572R a pH 7.0 (e) and pH 8.0 (f) before heating
(black) and after heating to 70 °C (red) and cooling back to 20 °C (blue). The WT and the mutants did not
display any significant changes in structure at pH 7.0 and pH 8.0. formed following trypsin digestion. We aimed to identify the regions that were probably buried within the
β -oligomers and hence resisted trypsin digestion. For WT, three short peptides (549ALPPR553, 591SLQGDK596
603NNVVSVNK610) were not detectable after tryptic digestions. For R555W, two short peptides (558LLGDAK563,
591SLQGDK596 were absent. However, for the H572R, peptides in the region E611-L632 were not observed in
addition to a short peptide (591SLQGDK596) that was absent in the WT and R555W. Mapping the generated pep-
tides to the 4th_FAS1 domain displayed interesting results (Fig. 6e). The entire C-terminal region encompassing
the residues 603NNVVSVNK610 and 611EPVAEPDIMATNGVVHVITNVL632 peptides was absent in the H572R
native protein and β -oligomers. This region was intact in WT and R555W. A long stretch of residues between
E534 and K563 containing the peptides 534EGVYTVFAPTNEAFR548, 549ALPPR553, 554EWSR557, 558LLGDAK563 was
not observed in R555W β -oligomers. The peptide 591SLQGDK596 was absent in all the proteins. The segment
549ALPPR553 was absent in H572R β -oligomers. The β -oligomers of R555W, displayed a region extending between
E534 and A562 that could possibly be buried. It is interesting to note that the mutation R555W resides within this Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 5 www.nature.com/scientificreports/ Figure 3. Thermal denaturation of the 4th_FAS1 domains of the WT and mutants at acidic pH. (a–c) Far
UV CD spectra of the 4th_FAS1 domain of WT at pH 3.0 (a), pH 4.5 (b) and pH 5.5 (c) before heating (black)
and after heating to 70 °C (red) and cooling back to 20 °C (blue). Resultsf (d–f) Far UV CD spectra of the 4th_FAS1
domain of R555W at pH 3 (d), pH 4.5 (e) and pH 5.5 (f) before heating (black) and after heating to 70 °C (red)
and cooling back to 20 °C (blue). (g–i) Far UV CD spectra of the 4th_FAS1 domain of H572R at pH 3.0 (g), pH
4.5 (h) and pH 5.5 (i) before heating (black) and after heating to 70 °C (red) and cooling back to 20 °C (blue). While the WT did not show any changes in structure, both the mutants displayed a very clear transition to
β -sheet under acidic conditions. Figure 3. Thermal denaturation of the 4th_FAS1 domains of the WT and mutants at acidic pH. (a–c) Far
h Figure 3. Thermal denaturation of the 4th_FAS1 domains of the WT and mutants at acidic pH. (a–c) Far
UV CD spectra of the 4th_FAS1 domain of WT at pH 3.0 (a), pH 4.5 (b) and pH 5.5 (c) before heating (black)
and after heating to 70 °C (red) and cooling back to 20 °C (blue). (d–f) Far UV CD spectra of the 4th_FAS1
domain of R555W at pH 3 (d), pH 4.5 (e) and pH 5.5 (f) before heating (black) and after heating to 70 °C (red)
and cooling back to 20 °C (blue). (g–i) Far UV CD spectra of the 4th_FAS1 domain of H572R at pH 3.0 (g), pH
4.5 (h) and pH 5.5 (i) before heating (black) and after heating to 70 °C (red) and cooling back to 20 °C (blue). While the WT did not show any changes in structure, both the mutants displayed a very clear transition to
β -sheet under acidic conditions. Figure 3. Thermal denaturation of the 4th_FAS1 domains of the WT and mutants at acidic pH. (a–c) Far
UV CD spectra of the 4th_FAS1 domain of WT at pH 3.0 (a), pH 4.5 (b) and pH 5.5 (c) before heating (black)
and after heating to 70 °C (red) and cooling back to 20 °C (blue). (d–f) Far UV CD spectra of the 4th_FAS1
domain of R555W at pH 3 (d), pH 4.5 (e) and pH 5.5 (f) before heating (black) and after heating to 70 °C (red)
and cooling back to 20 °C (blue). Resultsf (g–i) Far UV CD spectra of the 4th_FAS1 domain of H572R at pH 3.0 (g), pH
4.5 (h) and pH 5.5 (i) before heating (black) and after heating to 70 °C (red) and cooling back to 20 °C (blue). While the WT did not show any changes in structure, both the mutants displayed a very clear transition to
β -sheet under acidic conditions. region. The β -oligomers of H572R did not reveal such a region within the domain. It is likely that different regions
of the 4th_FAS1 domain may be involved in the formation and stabilization of these β -oligomers. The β-oligomers remained stable at physiological conditions. For studying the thermal stability
of the β -oligomers of the R555W (Fig. 7a,b) and H572R (Fig. 7c,d) the mutant domains at pH 5.5 were heated
to 70 °C and cooled back to 20 °C. The β -oligomers remained in their β -sheet conformation at 20 °C showing no
reversibility to the native α /β -conformation. To test their stabilities at physiological pH, the β -oligomers formed
at pH 5.5 were reconstituted to pH 7 and examined (Fig. 7e,f). The samples were also incubated for 4 weeks at
pH 7 and the CD spectra were recorded. In both cases, the β -oligomers remained in their stable β -sheet confor-
mation. This demonstrates that the β -oligomers were stable to thermal changes and at physiological conditions
allowing for further examination. Cytotoxicity of β-oligomers. The toxicity of the β -oligomers on primary human corneal stromal fibro-
blast (pHCSF) cells was monitored using xCELLigence system. The WT domain did not interfere with the cell
adhesion and proliferation and exhibited little or no toxic effect on the seeded cells (Fig. 8a). R555W also dis-
played no cytotoxicity. H572R, however, displayed higher cytotoxicity (**P < 0.01) on the fibroblasts compared
to R555W (Fig. 8b). Interestingly, β -oligomers from both R555W and H572R were cytotoxic (**P < 0.01). While,
the β -oligomers of R555W decreased the cell proliferation and the cytotoxic effect was visible after 12 hours,
the β -oligomers of H572R displayed the maximum cytotoxic effect as no proliferation was observed and the
xCELLigence showed minimum cell index. These results suggest that amyloidogenic mutant displayed significant Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 6 www.nature.com/scientificreports/ Figure 4. Difference in thermal denaturation induced transition between non-amyloidogenic and
amyloidogenic mutants at acidic pH. Resultsf (a–e) Variable temperature CD curves at 222 nm of WT (black), R555W
(red) and H572R (blue) proteins heated from 20 °C to 70 °C at various pH (3.0 [a], 4.5 [b], 5.5 [c], 7.0 [d] and 8.0
[e]) and the CD intensities at 222 nm were plotted as a function of temperature. The baseline subtracted curves
of the WT (black), R555W (red) and H572R (blue) proteins show that while there was no transition observed
in the WT in all the conditions as observed from the unchanged straight line in black, little or no changes were
seen in pH 7 and pH 8 for the mutants. However, clear transitions to β -sheet were observed at acidic pH (pH 3,
pH 4.5 and pH 5.5) for both the mutants. In all cases, we observe transition (Tt) is higher for R555W compared
to H572R. A clear shift in their thermal denaturation curves between the mutants at acidic pH (pH 3.0, 4.5 and
5.5) is observed. A difference in Tt of 5–12 °C is observed at various pH conditions. Figure 4. Difference in thermal denaturation induced transition between non-amyloidogenic and Figure 4. Difference in thermal denaturation induced transition between non-amyloidogenic and
amyloidogenic mutants at acidic pH. (a–e) Variable temperature CD curves at 222 nm of WT (black), R555W
(red) and H572R (blue) proteins heated from 20 °C to 70 °C at various pH (3.0 [a], 4.5 [b], 5.5 [c], 7.0 [d] and 8.0
[e]) and the CD intensities at 222 nm were plotted as a function of temperature. The baseline subtracted curves
of the WT (black), R555W (red) and H572R (blue) proteins show that while there was no transition observed
in the WT in all the conditions as observed from the unchanged straight line in black, little or no changes were
seen in pH 7 and pH 8 for the mutants. However, clear transitions to β -sheet were observed at acidic pH (pH 3,
pH 4.5 and pH 5.5) for both the mutants. In all cases, we observe transition (Tt) is higher for R555W compared
to H572R. A clear shift in their thermal denaturation curves between the mutants at acidic pH (pH 3.0, 4.5 and
5.5) is observed. A difference in Tt of 5–12 °C is observed at various pH conditions. Resultsf cytotoxic effect both in the native as well as in the β -oligomeric forms whereas the non-amyloidogenic mutant
was cytotoxic in the β -oligomeric form only. To obtain detailed information on the cytotoxicities, we prepared
the β -oligomers of the two mutants at various acidic pH (pH 3.0, pH 4.5, pH 5.5). We also included the insolu-
ble aggregates obtained after centrifuging the mildly precipitated samples. When we examined this precipitates
using CD, we found similar β -sheet curves as we found for the soluble oligomers (data not shown). Hence, we
investigated these “insoluble aggregates” along with the soluble β -oligomers. The populations were reconstituted
to pH 7 using buffer exchange. pHCSFs from 3 different donors (n = 3) were treated with the β -oligomers and
examined (Fig. 8c). The WT shows almost no cytotoxicity, similar to the control. The soluble β -oligomers derived
from both the mutants at various acidic pH conditions displayed potent cytotoxic effect (**P < 0.01) compared to
the insoluble aggregates and controls. The insoluble aggregates were also cytotoxic (*P < 0.05) but were relatively
lesser compared to the soluble β -oligomers. The results were also validated using an MTT assay (Fig. 8d). Similar
to xCELLigence, we could see that soluble β -oligomers derived from both the mutants displayed potent cytotoxic
effect (**P < 0.01) compared to the controls and insoluble aggregates. cytotoxic effect both in the native as well as in the β -oligomeric forms whereas the non-amyloidogenic mutant
was cytotoxic in the β -oligomeric form only. To obtain detailed information on the cytotoxicities, we prepared
the β -oligomers of the two mutants at various acidic pH (pH 3.0, pH 4.5, pH 5.5). We also included the insolu-
ble aggregates obtained after centrifuging the mildly precipitated samples. When we examined this precipitates
using CD, we found similar β -sheet curves as we found for the soluble oligomers (data not shown). Hence, we
investigated these “insoluble aggregates” along with the soluble β -oligomers. The populations were reconstituted
to pH 7 using buffer exchange. pHCSFs from 3 different donors (n = 3) were treated with the β -oligomers and
examined (Fig. 8c). The WT shows almost no cytotoxicity, similar to the control. The soluble β -oligomers derived
from both the mutants at various acidic pH conditions displayed potent cytotoxic effect (**P < 0.01) compared to
the insoluble aggregates and controls. Resultsf The insoluble aggregates were also cytotoxic (*P < 0.05) but were relatively
lesser compared to the soluble β -oligomers. The results were also validated using an MTT assay (Fig. 8d). Similar
to xCELLigence, we could see that soluble β -oligomers derived from both the mutants displayed potent cytotoxic
effect (**P < 0.01) compared to the controls and insoluble aggregates. Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 7 www.nature.com/scientificreports/ Figure 5. pH sensitivity and stability of the non-amyloidogenic (R555W) phenotype. (a) Fluorescence
emission spectra of R555W with decrease in pH. There was a clear decrease in emission maximum at 332 nm
with decrease in pH (indicated by the black arrow). (b–e) Fluorescence emission spectra of R555W showing
the reversibility to folded state after removal of urea. The R555W mutant was incubated with increasing
concentrations of urea from 0.25 M to 8 M at various acidic conditions (pH 3, pH 4.5, pH 5.5) and pH 7 and
the emission fluorescence before and after urea incubation was measured. The emission spectra before urea
incubation 332 nm (black), after incubating with 8 M urea (red) and after removing urea by buffer exchange
(blue). Unfolding of the protein is seen by the shifting of peaks (black arrow) from 332 nm to ~ 352 nm. The
refolding of the protein after removal of urea is seen by the return of the emission maximum to ~332 nm (green
arrow). (f–i) Investigation of the stability the non-amyloidogenic phenotype using urea denaturation studies. The R555W mutant was incubated with increasing concentrations of Urea from 0.25 M to 8 M at various acidic
pH (pH 3, pH 4.5, pH 5.5) and pH 7, and the emission fluorescence was measured. The denaturation plots
of ‘fraction unfolded vs urea concentration’ were plotted and fit into a two state model, with the parameters
calculated as described in the methods section. Figure 5. pH sensitivity and stability of the non-amyloidogenic (R555W) phenotype. (a) Fluorescence
emission spectra of R555W with decrease in pH. There was a clear decrease in emission maximum at 332 nm
with decrease in pH (indicated by the black arrow). (b–e) Fluorescence emission spectra of R555W showing
the reversibility to folded state after removal of urea. Resultsf The R555W 4th FAS1 domain
was incubated with increasing concentrations of urea from 0.25 M-8 M at various acidic pH conditions and the
emission fluorescence was measured. The free energies at various pH conditions were calculated and tabulated. Resultsf The R555W mutant was incubated with increasing
concentrations of urea from 0.25 M to 8 M at various acidic conditions (pH 3, pH 4.5, pH 5.5) and pH 7 and
the emission fluorescence before and after urea incubation was measured. The emission spectra before urea
incubation 332 nm (black), after incubating with 8 M urea (red) and after removing urea by buffer exchange
(blue). Unfolding of the protein is seen by the shifting of peaks (black arrow) from 332 nm to ~ 352 nm. The
refolding of the protein after removal of urea is seen by the return of the emission maximum to ~332 nm (green
arrow). (f–i) Investigation of the stability the non-amyloidogenic phenotype using urea denaturation studies. The R555W mutant was incubated with increasing concentrations of Urea from 0.25 M to 8 M at various acidic
pH (pH 3, pH 4.5, pH 5.5) and pH 7, and the emission fluorescence was measured. The denaturation plots
of ‘fraction unfolded vs urea concentration’ were plotted and fit into a two state model, with the parameters
calculated as described in the methods section. pH
ΔGH2OZ, kJ/mole
Cm, M
m, kJ/mole/M
ΔΔGH2O ,
kJ/mole
3.0
8.417 ± 1.03
4.2 ± 0.15
3.14 ± 0.32
− 5.383
4.5
7.108 ± 1.5
4.74 ± 0.91
1.85 ± 0.25
− 6.692
5.5
11.28 ± 2.02
3.93 ± 0.21
2.78 ± 0.48
− 2.52
7.0
13.8 ± 0.8
3.09 ± 0.1
3.8 ± 0.3
–
Table 2. Urea denaturation studies of the non-amyloidogenic phenotype. The R555W 4th FAS1 domain
was incubated with increasing concentrations of urea from 0.25 M-8 M at various acidic pH conditions and the
emission fluorescence was measured. The free energies at various pH conditions were calculated and tabulated. pH
ΔGH2OZ, kJ/mole
Cm, M
m, kJ/mole/M
ΔΔGH2O ,
kJ/mole
3.0
8.417 ± 1.03
4.2 ± 0.15
3.14 ± 0.32
− 5.383
4.5
7.108 ± 1.5
4.74 ± 0.91
1.85 ± 0.25
− 6.692
5.5
11.28 ± 2.02
3.93 ± 0.21
2.78 ± 0.48
− 2.52
7.0
13.8 ± 0.8
3.09 ± 0.1
3.8 ± 0.3
– Table 2. Urea denaturation studies of the non-amyloidogenic phenotype. The R555W 4th FAS1 domain
was incubated with increasing concentrations of urea from 0.25 M-8 M at various acidic pH conditions and the
emission fluorescence was measured. The free energies at various pH conditions were calculated and tabulated. Table 2. Urea denaturation studies of the non-amyloidogenic phenotype. Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 Discussionhi The RH values calculated from the distribution curves for R555W (~39.58 nm|pH 3.0, ~51.9 nm|pH 4.5 and
~68.2 nm|pH 5.5) and H572R (~89.95 nm|pH 3.0, ~69 nm|pH 4.5 and ~155 nm|pH 5.5) show that H572R b-oligomers
are larger in size and slightly more heterogeneous. (e,f) Mass Spectrometric analyses - Identification of the
aggregation hotspots in the mutants and β -oligomers. (d) Peptide map generated following the insilico trypsin
digestion of TGFβ Ip displaying a series of peptides. (e) The β -oligomers formed from the mutant 4th_FAS1
domains were digested with trypsin and the resulting fragments were analyzed by LC–MS/MS. The regions
inaccessible for trypsin digestion have been underlined in red. It is clearly seen that the non-amyloidogenic
R555W shows more regions inaccessible for trypsin digestion. the effects of various biophysical and biochemical factors like temperature and pH on them. In amyloid forming
proteins such as β -microglobulin39, thermal unfolding leads to formation of β -oligomers. Results from our ther-
mal denaturation studies show clear differences between both the phenotypes and the WT protein at acidic pH.l f
Our results suggest that changes in the physiological pH can influence the biochemical and biophysical char-
acteristics of TGFβ Ip mutants. A clear conversion from α /β -structure to β -sheet was observed when the mutants
were heated under acidic conditions (< pH 5.5). TEM and DLS studies confirmed the formation of β -oligomers
by both phenotypes that remain irreversible and stable at physiological pH and temperature. For the first time
we have shown that both the amyloidogenic and non-amyloidogenic phenotypes displayed marked differences
in their sensitivities to pH and β -oligomer formation. The two β -oligomers of different phenotypes exhibit varia-
tions in their morphologies, sizes, thermal and chemical stabilities, aggregation patterns and cytotoxicity. Supporting our results that acidic pH could play an important role in the aggregation of proteins in vivo, some
recent studies have provided insights on the possible influence of acidic pH on the aggregation of proteins40–42. Pfefferkorn et al., have reported that Pmel 17, a functional amyloid involved in the structural scaffolding for
melanin deposition in human skin and eyes, displays a strong tendency to form fibrils at acidic pH 5.0 43. The
Pmel 17 fibrils are formed inside the melanosomes, organelles related to lysosomes that have acidic environments
similar to lysosomes. Discussionhi Also, in studies involving TGFβ Ip associated dystrophies, Kim et al., have reported31 that
the R124H mutation in TGFβ Ip disrupts its binding to periostin and causes TGFβ Ip to localize in lysosomes. It
is well known that the lysosomal environment is highly acidic. While the mutation affects the binding of TGFβ Ip
to its functional partners, the localization of TGFβ Ip mutants to the lysosomes with acidic pH could possibly
induce the initiation of β -oligomer formation. While the 4th_FAS1 domain is not known to bind to periostin, it
is still distributed to lysosomes by the R124H mutation31. It is known that the proteolytic fragments in patient
samples include the 4th_FAS1 domain along with the 1st FAS1 domain44. Acidic environments could probably
induce the conversion to β -sheet in the 4th_FAS1 domain. We also speculate that under acidic pH, binding of
TGFβ Ip to periostin might be disrupted possibly due to the formation of β -oligomers in the presence of other
interacting factors. TGFβ Ip is known to interact with several extracellular matrix proteins like various types of
collagen, biglycan, decorin and fibronectin31 and it is possible that the acidic pH could probably have a disruptive
effect on the binding of the mutants to other ECM components. Therefore, it is reasonable to suggest that acidic
environments could play an important role in the aggregation of dystrophic mutants. At physiological conditions
(pH 7.0), there were no perturbations in structure from the native conformation. Under acidic conditions (pH
5.5 and below), we were able to show the ability of the mutants to form β -oligomers. Any change in the stromal
environment causing a lowering of pH to about 1.5 units could probably be a trigger to initiate the formation of
β -oligomers. Physiological events like lysosomal membrane disruption that can alter the pH values of the intra-
cellular and extracellular microenvironment by a pH reduction of ~1.5 (pH 7 → pH 5.5) during cellular processes
such as wound-healing and phagocytosis45–47 could possibly lead to conversion to β -sheet and possibly formation
of β -oligomers. f
d l
b
f h
T
d
d
d
d f ll
d h We performed long-term incubation of the WT and mutants in acidic pH conditions and followed their pos-
sible aggregation/oligomerization using ThT fluorescence and CD. Both the mutants did not display any aggre-
gation/oligomerization over 1 week (Supplementary Fig. S1). Discussionhi The most significant aspect of the TGFβ Ip associated corneal dystrophies is that the single amino acid substi-
tutions in the mutants are responsible for clinically distinct phenotypes. Here, we examined the properties of
two mutants, the amyloidogenic H572R22,23, and non-amyloidogenic R555W24–26. Previous attempts to study the
differences between the WT and the mutants at physiological conditions have revealed minimal information on
the mechanism of their aggregation38. To elucidate the inherent variations between these mutants, we examined Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 8 /scientificreports/
Figure 6. Characterization amyloidogenic and non-amyloidogenic β-oligomers. (a–b) Transmission
Electron Microscopy. TEM images of the β -oligomers of the 4th_FAS1 domains of R555W (a) and H572R (b)
mutants were acquired with a JEOL JEM-1010 transmission electron microscope using Digital Micrograph™
1.81.78 for GMS 1.8.0. The β -oligomers of the amyloidogenic phenotype were larger measuring between
10–40 nm, mean size ~19.1 nm ± 4.9 nm (a) compared to the non-amyloidogenic β -oligomers that measured
4–8 nm, with a mean size ~5.1 nm ± 1.79 nm (b). Inset figures – particle size distribution of the β -oligomers. www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 6. Characterization amyloidogenic and non-amyloidogenic β-oligomers. (a–b) Transmission
Electron Microscopy. TEM images of the β -oligomers of the 4th_FAS1 domains of R555W (a) and H572R (b)
mutants were acquired with a JEOL JEM-1010 transmission electron microscope using Digital Micrograph™
1.81.78 for GMS 1.8.0. The β -oligomers of the amyloidogenic phenotype were larger measuring between
10–40 nm, mean size ~19.1 nm ± 4.9 nm (a) compared to the non-amyloidogenic β -oligomers that measured
4–8 nm, with a mean size ~5.1 nm ± 1.79 nm (b). Inset figures – particle size distribution of the β -oligomers. Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 9 www.nature.com/scientificreports/ While the amyloidogenic β -oligomers were larger and displayed rugged edges and varying diameters, the
non-amyloidogenic β -oligomers were smaller in size with smoother edges and were more homogenous. (c) ThT assay. Emission fluorescence intensities at 485 nm recorded after incubating the WT TGFβ Ip and
the β -oligomers of R555W and H572R with ThT dye. An amyloid forming peptide (611-633aa - pN622K)
of TGFβ Ip was used as the positive control. The β -oligomers of the amyloidogenic mutant H572R showed
significant fluorescence intensity (**P < 0.01) almost 3 times more fluorescence compared to the non-
amyloidogenic R555W. (d) DLS. %intensity plots plotted against the apparent hydrodynamic radii (RH). Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 Discussionhi We also performed seeded fibrillation experiments
using β -oligomers as seeds and incubated them for 2 weeks. However, we did not observe any significant increase
in ThT fluorescence (Supplementary Fig. S4) within the period. There could be other cellular factors and inter-
acting molecules involved directly or indirectly with the aggregation that need to be studied along with the acidic
pH at physiological temperature to induce aggregation/oligomerization. Prolonged incubation under these con-
ditions could probably induce the formation of oligomerization in both the cases. Also, in diseases conditions,
patients harbor the mutations from birth, but the ages of onset range from few years to few decades. While we
emphasize the significance of acidic pH in the possible aggregation and β -oligomer formation of the 4th_FAS1
domain, we certainly believe that more detailed in vivo investigation is needed to delineate the aggregation mech-
anism and the molecules.h At neutral pH, we have seen that the R555W mutant is more stable than the WT. This was also seen in the
stability experiments by Runager et al.48. However, from our results we see that the non-amyloidogenic R555W
displays high susceptibility to acidic pH. In R555W, the mutation (R→ W) causes an overall reduction in net
charge from − 0.6 to − 1.6 at pH 7.0 (17, Fig. 1c). At pH 5.5, the charge reduces from 1.9 to 0.9 17. The substitution
of the hydrophobic tryptophan (− 2.13) in place of the polar arginine (3.95) also increases the hydrophobicity, Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 10 www.nature.com/scientificreports/ Figure 7. Stability of the β-oligomers at physiological conditions. (a–d) Thermal stability. Variable
temperature CD values at 222 nm of the 4th_FAS1 domains of R555W (a) and H572R (c) in buffer at pH 5.5
when heated from 20 °C to 70 °C. Far UV CD spectra of the 4th_FAS1 domains of R555W (b) and H572R
(d) after heating (red) to 70 °C and cooling (blue) back to 20 °C. (e,f) Stability at physiological pH. Far UV
CD spectra of the 4th_FAS1 domains (black) of R555W (e) and H572R (f) at pH 5.5 under native conditions
(black), after heating to 70 °C (red) and after reconstituting in buffer at pH 7.0 (blue) after incubation at room
temperature for 4 weeks. Figure 7. Stability of the β-oligomers at physiological conditions. (a–d) Thermal stability. Discussionhi (c) xCELLigence assay. The WT shows low cytotoxicity,
almost similar to the control. The soluble β -oligomers of both the mutants showed high cytotoxicity (**P < 0.01)
compared to the controls. The insoluble β -oligomers were relatively less cytotoxic compared to the soluble
β -oligomers. (d) MTT assay. The cytotoxicity of the β -oligomers was also tested using an MTT assay. Similar to
xCELLigence, we could see that soluble β -oligomers derived from both the mutants displayed potent cytotoxic
effect (**P < 0.01) compared to the controls and insoluble aggregates. Though the insoluble oligomers were
relatively less cytotoxic than the soluble oligomers, they displayed cytotoxicity compared to the controls. Figure 8. Cytotoxicity. (a) The growth and proliferation of the pHCSF cells following the treatment with the
β -oligomers and the mutant 4th_FAS1 domains monitored for 48 hours in an xCELLigence system. (b) The cell
survival after treatment for 24 hours plotted as area under the curve (AUC). While the WT and R555W show
low cytotoxicity, almost similar to the media control, the amyloidogenic H572R mutant was cytotoxic compared
to the control (**P < 0.01). Interestingly, the β -oligomers of both the amyloidogenic and non-amyloidogenic
mutants showed high cytotoxicity (**P < 0.01) with the approximately 7 times more for the non-amyloidogenic
mutant and 12 times more for the amyloidogenic mutant. To obtain detailed information on the cytotoxicity
of the soluble and insoluble fractions, β -oligomers of the two mutants were prepared at various acidic pH
conditions (pH 3.0, pH 4.5, pH 5.5) and their cytotoxicity were examined on pHCSFs from 3 different donors
(n = 3) using xCELLigence (c) and MTT assays (d). (c) xCELLigence assay. The WT shows low cytotoxicity,
almost similar to the control. The soluble β -oligomers of both the mutants showed high cytotoxicity (**P < 0.01)
compared to the controls. The insoluble β -oligomers were relatively less cytotoxic compared to the soluble
β -oligomers. (d) MTT assay. The cytotoxicity of the β -oligomers was also tested using an MTT assay. Similar to
xCELLigence, we could see that soluble β -oligomers derived from both the mutants displayed potent cytotoxic
effect (**P < 0.01) compared to the controls and insoluble aggregates. Though the insoluble oligomers were
relatively less cytotoxic than the soluble oligomers, they displayed cytotoxicity compared to the controls. Discussionhi Variable
temperature CD values at 222 nm of the 4th_FAS1 domains of R555W (a) and H572R (c) in buffer at pH 5.5
when heated from 20 °C to 70 °C. Far UV CD spectra of the 4th_FAS1 domains of R555W (b) and H572R
(d) after heating (red) to 70 °C and cooling (blue) back to 20 °C. (e,f) Stability at physiological pH. Far UV
CD spectra of the 4th_FAS1 domains (black) of R555W (e) and H572R (f) at pH 5.5 under native conditions
(black), after heating to 70 °C (red) and after reconstituting in buffer at pH 7.0 (blue) after incubation at room
temperature for 4 weeks. Figure 7. Stability of the β-oligomers at physiological conditions. (a–d) Thermal stability. Variable
temperature CD values at 222 nm of the 4th_FAS1 domains of R555W (a) and H572R (c) in buffer at pH 5.5
when heated from 20 °C to 70 °C. Far UV CD spectra of the 4th_FAS1 domains of R555W (b) and H572R
(d) after heating (red) to 70 °C and cooling (blue) back to 20 °C. (e,f) Stability at physiological pH. Far UV
CD spectra of the 4th_FAS1 domains (black) of R555W (e) and H572R (f) at pH 5.5 under native conditions
(black), after heating to 70 °C (red) and after reconstituting in buffer at pH 7.0 (blue) after incubation at room
temperature for 4 weeks. Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 11 www.nature.com/scientificreports/ Figure 8. Cytotoxicity. (a) The growth and proliferation of the pHCSF cells following the treatment with the
β -oligomers and the mutant 4th_FAS1 domains monitored for 48 hours in an xCELLigence system. (b) The cell
survival after treatment for 24 hours plotted as area under the curve (AUC). While the WT and R555W show
low cytotoxicity, almost similar to the media control, the amyloidogenic H572R mutant was cytotoxic compared
to the control (**P < 0.01). Interestingly, the β -oligomers of both the amyloidogenic and non-amyloidogenic
mutants showed high cytotoxicity (**P < 0.01) with the approximately 7 times more for the non-amyloidogenic
mutant and 12 times more for the amyloidogenic mutant. To obtain detailed information on the cytotoxicity
of the soluble and insoluble fractions, β -oligomers of the two mutants were prepared at various acidic pH
conditions (pH 3.0, pH 4.5, pH 5.5) and their cytotoxicity were examined on pHCSFs from 3 different donors
(n = 3) using xCELLigence (c) and MTT assays (d). Discussionhi thereby altering the solubility and possible folding of the mutant protein The reduction in charge coupled with
the increase in hydrophobicity might influence the sensitivity of R555W to acidic pH. In H572R, the histidine
residue is replaced by arginine residue. Histidine has a neutral charge at pH 7 and positive charge at acidic pH
(pH < 6.0). While there is an increase in the overall charge from − 0.6 to 0.2 by the substitution of arginine at pH
7, there is almost no change in the overall net charge for the H572R mutant under acidic conditions (1.9 to 2.0 at
pH 5.5). This may explain the relatively higher stability of the H572R mutation at lower pH when compared to
R555W. One of the important observations is the relative instability or sensitivity of R555W (pI 6.32) to pH 4.5. Our results from pH incubation studies, urea denaturation studies show that around pH 4.5, there is a decrease in Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 12 www.nature.com/scientificreports/ stability for R555W. Further characterization of R555W at pH 4.5 might provide more details on the significance
of this pH. p
It was also important to understand relative instability of R555W from a structural perspective. In recent
studies, (Runager et al.48 and Underhaug et al.49) have observed that the R555W mutant is more stable than the
WT48 and the 4th_FAS1 domain of R555W has a more compact structure49 at pH 7. Underhaug et al. also observed
that the α 3′ helix, harboring the Arg555 residue, shows an increase in connected surface of negative charge in the
R555W mutant compared to the WT which might directly influence its charged molecular interactions. They also
found that in the presence of acidic conditions (10% TFE) R555W mutant exhibits higher aggregation propensity
compared to the WT. The increased concentration of negative charges on the molecular surface of the R555W
4th_FAS1 domain could possibly explain its sensitivity to acidic pH and also its instability at pH 4.5.h p
y
p
y
p
y
p
Various protein aggregation diseases involve β -oligomeric intermediates50–52. These β -oligomers could poten-
tially act as seeds for the formation of amyloid fibrils53,54, and hence understanding the stability of β -oligomers
becomes vital. In our study, the β -oligomers from both the phenotypes formed at pH 5.5 remain stable at phys-
iological conditions for more than 4 weeks. Discussionhi Interestingly, the two β -oligomers seem to adopt different modes
of oligomerization and also exhibit differences in their sizes, ThT-binding and cytotoxicity. Pfefferkorn et al.43,
have shown in Pmel 17 that one of the first steps in fibril formation is the formation of β -oligomers that appear
as large laterally associated small fibrillar species. At pH 4.5–5.0, a mix of amorphous and fibrillar structures
were observed. Consistent with this, we have seen formation of β -oligomers that appear as spherical structures
with rugged edges of different sizes using TEM. The amyloidogenic H572R displays larger particles compared
to the non-amyloidogenic R555W. Previous studies in protein aggregation disorders like Parkinson’s syndrome
have shown that the β -oligomers are more toxic than the amyloid fibrils themselves53–55. This has however not
been studied in TGFBI-associated corneal dystrophies. When we assessed the cytotoxicity of the β -oligomers in
pHCSF, β -oligomers from amyloidogenic H572R were more cytotoxic than the non-amyloidogenic R555W. In
serum amyloid A (SAA) peptides55, the pathogenic and non-pathogenic strains appear as two distinct oligomeric
populations (SAA1.1 and SAA2.2). While the non-pathogenic SAA2.2 forms smaller oligomers and grows into
braided fibrils, the pathogenic SAA1.1 forms larger oligomers and grows into straight fibrils. Similar to this, the
larger β -oligomers from amyloidogenic H572R are more cytotoxic compared to the smaller β -oligomers of the
non-amyloidogenic R555W. y
g
We have seen that the soluble β -oligomers of both mutants were more cytotoxic compared to the insoluble
aggregates. This leads us to believe that while the soluble β -oligomers are responsible for cell death in the corneal
dystrophies, the insoluble aggregates are more responsible for scattering of light and hence loss of visual acuity. This could also mean that cytotoxicity due to β -oligomer formation precedes the loss of sight due to accumula-
tion of protein aggregates. This provides us an important understanding in studying the modes of treatments of
corneal dystrophies wherein, the hastening of the intermediary soluble β -oligomers to form insoluble and stable
higher order aggregates such as fibrils and granules, could minimize cell death. g
gg gi
g
In conclusion, these results suggest that changes in the pH may cause localized perturbations in the sec-
ondary structure of the mutant proteins and increase the propensity to form β -oligomers. We have shown that
under similar biophysical and biochemical conditions, the different mutant phenotypes exhibit differences in
morphology, size and cytotoxicity. Discussionhi However, we did not observe the conversion of β -oligomers into amyloid fibrils
even after incubation for 4 weeks. The soluble β -oligomers exhibit higher cytotoxicity compared to the insoluble
β -oligomers. The interaction of β -oligomers with other extracellular matrix proteins will shed light into the mech-
anism of amyloid formation in future studies. Methods
M t
i l The final spectrum
was the average of three scans. The CD data were expressed as mean residual weight ellipticity (deg cm2 dmol-1). The mean residual weight (MRW) ellipticity ([θ ]MRW) was estimated using the equation 2 °C. Variable temperature (VT) scanning was done by heating the sample to 70 °C at a rate of 1 °C/min and meas-
uring the ellipticity at 222 nm. Initial and final spectra before and after heating were recorded. The final spectrum
was the average of three scans. The CD data were expressed as mean residual weight ellipticity (deg cm2 dmol-1). The mean residual weight (MRW) ellipticity ([θ ]MRW) was estimated using the equation θ
= θ
×
×
×
λ
l
c
[ ]
[ ]
MRW/10
[1]
MRW θ
= θ
×
×
×
λ
l
c
[ ]
[ ]
MRW/10
MRW [1] where [θ ]λ is the observed ellipticity, MRW is mean residual weight and defined as M/N− 1 where M is the molecu-
lar mass and N is number of amino acid residues, l is the path length of the cuvette and c is concentration in mg/ml. The data obtained were fit into a two-state model using Origin 8.0 software and analysed using previously
reported methods56–58. For thermal stability experiments, the samples were heated to 70 °C and cooled back to 20 °C at a rate of 1 °C/min
and the corresponding ellipticities at 222 nm were recorded. Far UV-CD spectra of the samples before heating,
after heating to 70 °C and after cooling back to 20 °C were recorded. To test the effects of temperature ramping
rates on the thermal unfolding, the mutant proteins were heated at pH 5.5, at various rates of thermal ramping
(0.5 °C/min, 1 °C/min, 2 °C/min and 10 °C/min) and the CD spectra were recorded. VT curves were generated by
measuring and plotting the CD intensity at 222 nm from 20 °C–70 °C. g
p
g
y
For testing the stability of the β -oligomers at pH 7, the β -oligomers were formed by heating the mutants at
pH 5.5, to 70 °C. Far UV-CD spectra after heating were recorded for the samples. The heated samples were cen-
trifuged at 14,600 rpm (~20,000 g) and reconstituted in PBS buffer at pH 7.0 and kept at room temperature for
4 weeks and their CD spectra were recorded. Fluorescence spectroscopy. Methods
M t
i l The 4th_FAS1 domain of R555W was incubated at various acidic conditions
(pH 3, pH 4.5, pH 5.5 and pH 7.0) and the emission spectra from 300 to 400 nm were recorded following excita-
tion at 280 nm for each sample using the Quanta Star spectrofluorimeter (Photon Technology International, NJ)
using a 10 mm quartz cuvette. For urea denaturation studies, the R555W mutant protein (0.6 mg/ml – 37 μM) was
treated with 0.25 M to 8 M urea for 16 hours at 4 °C. The intrinsic fluorescence corresponding to the tryptophan
residue in R555W was measured as described above. To confirm the reversibility of the urea denatured proteins
to their native conformation after removal of urea, the R555W protein was first incubated in buffers of different
pH (pH 3.0, pH 4.5, pH 5.5 and pH 7.0) with 8 M urea for 24 hours. The unfolding of the proteins in the presence
of 8 M urea was confirmed using fluorescence spectroscopy. These unfolded proteins were then reconstituted in
buffers without urea by buffer exchange by centrifugation using Amicon Centriprep filter units. The emission
peaks were measured. Reversibility studies were performed by diluting the R555W domain from 8 M urea to
appropriate concentrations.hi pp
p
The denaturation data were fitted by non-linear least squares method, assuming a two state model59. To cal-
culate the thermodynamics parameters from the chemical denaturation data, we first plotted the emission fluo-
rescence spectra of the R555W mutant subjected to unfolding with 8 M urea. After confirming the shift in peaks
from ~332 nm to ~ 352 nm, the differences in the wavelengths (2) ∆λ
= λ
−λ
(2)
max
max1
max2 corresponding to the emission peaks were plotted against urea concentration to estimate the λ U (unfolded) and
λ F (folded). The fraction-folded corresponding to the emission peaks were plotted against urea concentration to estimate the λ U (unfolded) and
λ F (folded). The fraction-folded = λ −λ
λ
−λ
y
(
)/(
)
(3)
F
F
U
F (3) and fraction-unfolded = λ −λ
λ
−λ
y
(
)/(
)
(4)
U
U
U
F = λ −λ
λ
−λ
y
(
)/(
)
U
U
U
F (4) were calculated. The ΔG values were calculated by substituting values in the equation ∆
=
G
RT ln(y /y )
(5)
F
U (5) The resulting ΔG values from the linear range were plotted against urea concentration. Methods
M t
i l Materials. The cDNA constructs of the 4th_FAS1 domains of the WT TGFβ Ip, the mutants R555W and
H572R were bought from Genscript (Piscataway, NJ) in pUC57 vectors. E.coli BL21(DE3) expression competent
cells, Ek/LIC cloning kit and pCDF-2 vector system were purchased from Novagen (Novagen (EMD), PA). High-
performance Ni-Sepharose resin was purchased from GE Healthcare (GE healthcare Life Sciences, NJ). Amicon
Centriprep filter units were purchased from EMD Millipore (EMD Millipore, MA). Ampicillin, streptomycin, iso-
propyl β -D-thiogalactopyranoside (IPTG) and Thioflavin T were purchased from Sigma-Aldrich (Sigma-Aldrich
Inc., MO). Formvar-carbon coated nickel grids were bought from EMS (Electron Microscopy Sciences, PA). Cloning, Protein expression and purification. Recombinant 4th_FAS1 domains of the WT, H572R and
R555W TGFβ Ip proteins were cloned, expressed and purified as described previously21. Briefly, the TGFBI genes
were cloned in pCDF2 vectors and expressed in E.coli BL21DE3 cells. The proteins were purified by affinity chro-
matography using a manually prepared column with High Performance Ni-sepharose beads. Further purification
was done using RP-HPLC (C4 column, Phenomenex) with a linear gradient of 50% acetonitrile in 0.01% trifluo-
roacetic acid at a flow rate of 3 ml/min. The purified proteins were lyophilized and stored in − 20 °C. Far-UV Circular Dichroism spectropolarimetry. Far UV–CD spectra of the proteins were collected
using a Jasco J-810 spectropolarimeter (Jasco Inc., Easton, MD) using a quartz cuvette with a path length of
0.1 cm (Hellma, Müllheim, Germany). Purified 4th_FAS1 WT and mutant TGFβ Ip proteins were reconstituted in
buffers of varying pH (pH 3.0–8.0) to a final concentration of ~0.6 mg/ml (~37 μM) and their CD spectra were
recorded.f For thermal denaturation experiments, the samples in buffers of varying pH (pH 3.0, pH 4.5, pH 5.5, pH 7.0 and
pH 8.0) were heated from 20 °C to 70 °C at a rate of 1 °C/min using the inbuilt Peltier heating system. Thermal dena-
turation experiments were also performed by heating the samples from 20 °C–90 °C under the same conditions. The spectra were recorded from 260 nm to 190 nm with a step-size of 0.1 nm, at a scan rate of 50 nm/min for every Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 13 www.nature.com/scientificreports/ 2 °C. Variable temperature (VT) scanning was done by heating the sample to 70 °C at a rate of 1 °C/min and meas-
uring the ellipticity at 222 nm. Initial and final spectra before and after heating were recorded. Methods
M t
i l Dynamic light scattering (DLS). The β -oligomers were formed by heating the 4th_FAS1 domains of the
R555W and H572R in buffers of different pH (pH 3.0, pH 4.5, pH 5.5) to 70 °C. The buffers were filtered using a
0.2 μM syringe filter before the thermal denaturation experiments. These prepared samples were examined using
a DynaPro Plate ReaderII (Wyatt Technology, CA, US) equipped with a Peltier temperature controller. Each sam-
ples was measured 3 times and the average of the 3 scans is reported. The average and standard deviation values
of the sizes corresponding to the peak of interest in each of these 3 distributions provides the apparent hydrody-
namic radius and the experimental error for each samples respectively. Mass spectrometric analysis. The native and heated 4th_FAS1 domains of the WT, R555W and H572R
were digested with trypsin. The tryptic-digested peptides were analyzed by LC–MS/MS [Ultimate 3000 nanoLC
(Thermo Fisher Scientific/Dionex, Sunnyvale, CA), coupled with AB SCIEX Triple TOF TM 5600 mass spectrom-
eter]. The peptide mixture was first desalted and pre-concentrated in a trap column (Acclaim PepMap 100 C18,
75 μm × 3 μm, 100 Å from Thermo Fisher Scientific/Dionex) for 3 min at a flow rate of 5 μl/min. After desalting,
the system was switched into line with the reversed phase analytical capillary column (25 cm × 75 μm i.d. Acclaim
PepMap RSLC C18, 2 μm, 100 Å, Thermo Fisher Scientific/Dionex, Sunnyvale, CA). A 35 min gradient was used
at 300 nl/min. All data was acquired using information-dependent acquisition (IDA) mode with Analyst TF 1.5.1
software (AB Sciex, USA). Protein Pilot software (version 4.01, AB Sciex) was used to analyze the MS/MS data. Cytotoxicity. The xCELLigence system (ACEA, San Diego, CA, USA) was used to investigate the effects of
the β -oligomers on pHCSF proliferation and cell death. Their E-plate 96 is incorporated with sensor arrays at the
bottom of each well, which detects and translate cell attachment in the form of electronic impedance. A parameter
termed cell index (CI) is derived which corresponds to the relative density and adherence strength of cells in each
well60. pHCSF were seeded at a density of 5000 cells per well on an E-plate 96, in fibroblast media with 5% FBS. An
initial concentration of ~37 μM (0.6 mg/ml) of the mutant protein was subjected to thermal denaturation at acidic
pH to generate the β -oligomers. Methods
M t
i l The Y-intercept of the
resulting straight line was taken as ΔGH2O and the slope (m) was estimated. ∆∆
= ∆
−∆
. . G
G
G
(6)
H O at given pH
H O at given pH
H O at pH 7 0
2
2
2 (6) Transmission Electron Microscopy (TEM). TEM images of the β -oligomers prepared at pH 5.5 were
acquired with a JEOL JEM-1010 transmission electron microscope using Digital Micrograph™ 1.81.78 for GMS
1.8.0 (Gatan, Pleasanton, CA) in the National University of Singapore Electron Microscopy facility. The samples
were centrifuged and applied on Formvar-carbon coated nickel grids with bacitracin as the binding agent and
negatively stained with 10% phosphotungstic acid (PTA). The samples were dried and observed at magnifications
8000–50000X at 80 kV. The size of the particles were estimated using the relation, =
. Real size
Measured length/Magnification =
. Real size
Measured length/Magnification Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 14 www.nature.com/scientificreports/ Thioflavin T assay. The native and heated domains were treated with 30 μM thioflavin T (ThT) in PBS
buffer at pH 5.5 in a Greiner 96-well flat bottom polystyrol microplate (Greiner, Frickenhausen Germany). The samples were excited at 445 nm and the resulting emission fluorescence at 485 nm was measured using a
microplate reader (Tecan infinite M200 pro, Zanker Road, SJ). The emission fluorescence was monitored for
24 hours. A 20-amino acid long peptide (pN622K) from the 4th_FAS1 domain of TGFβ Ip with the mutation
N622K (EPVAEPDIMATKGVVHVITNVLQ) capable of forming amyloid fibrils was used as positive control.hf p
g
yi
p
The native and heated 4th_FAS1 domains of the WT, R555W and H572R were incubated in buffers of dif-
ferent pH (pH 2.75, pH 3.25, pH 4.0, pH 4.5, pH 5.5 and pH 7.0) along with 30 μM thioflavin T (ThT) and their
corresponding changes in fluorescence intensity at 485 nm were measured for a period of 7 days using the Tecan
microplate reader. The N622K peptide was used a positive control.i ph
p p
p
Seeded fibrillation studies were performed by treating the domains with their respective β -oligomers prepared
at pH 5.5 using thermal denaturation as described above. The samples were shaken in an orbital shaker at 37 °C
at 180 rpm. The samples were taken after 2 weeks and mixed with ThT for a final concentration of 30 μM and the
fluorescence intensities were measured as described before. ,
y
p
p
J
,
(
)
3. Surguchev, A. & Surguchov, A. Conformational diseases: looking into the eyes. Brain Res. Bull 81, 12–24 (2010). References
1. Weiss, J. S. et al. The IC3D classification of the corneal dystrophies. Cornea 27 (Suppl. 2) S1–S83 (2008). Methods
M t
i l The β -oligomers were mixed with DMEM media at a ratio of 1:1 (~18 μM/0.3 mg/
ml final concentration) and added to pHCSF cells. Culture was maintained for 48 hours in the xCELLigence
RTCA SP Station, placed in an incubator at 37o C with 5% CO2. No media change was done to reduce the fluctua-
tions generated from removal of plate from RTCA system. The area under the curve (AUC) values were calculated
from the cI values using the Mathematics module in OriginPro software and were used for further analysis. g
gt
y
To examine the cytotoxicities of the soluble and insoluble fractions in the β -oligomers, the β -oligomers of
the two mutants prepared at various acidic pH conditions (pH 3.0, pH 4.5, pH 5.5) were were centrifuged at
14,600 rpm (~20,000 g) to separate the insoluble fractions. The insoluble β -aggregates in the pellets were recon-
stituted in PBS buffers of pH 7.0. For the soluble β -oligomers in the supernatant, buffer exchange by centrif-
ugation using Amicon centriprep was done with PBS at pH 7.0. The soluble and insoluble populations were
added to pHCSF cells from 3 donors (n = 3) and was monitored in an xCELLigence system. Cell survival assay
using the xCELLigence system was performed as explained earlier. To test the cell viability, MTT assay was per-
formed in parallel. The cells were incubated with the β -oligomers for 48 hours and treated with MTT (3-(4,5-dim
ethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) and incubated for 4 hours. The resulting color change was
measured using the Tecan microplate reader (Tecan infinite M200 pro, Zanker Road, San Jose, USA). Statistics. All numeric data obtained were expressed as mean ± standard deviation. For both the ThT bind-
ing assays and cytotoxicity studies using xCELLigence, all the comparisons were done using one-way ANOVA
followed by post-hoc Bonferroni test (SPSS Statistics 22.0, IBM, Chicago, IL) for multiple comparisons. For both
the experiments, values were deemed to be significant (*) when a significance level with a p-value of less than
0.05 was achieved and very significant (**) when a significance level with a p-value of less than 0.01 was achieved. 4. Lakshminarayanan, R. et al. Biochemical properties and aggregation propensity of transforming growth factor-induced protein
(TGFβ Ip) and the amyloid forming mutants. Ocul Surf. 13(1), 9–25 (2015).ii (
β p)
y
g
f
( ),
(
)
5. Kitahama, S. et al. Expression of fibrillins and other microfibril-associated proteins in human bone and osteoblast-like cells. Bone
27, 61–67 (2000). 2. Klintworth, G. K. Corneal dystrophies. Orphanet J. Rare Dis. 4, 7 (2009).
3. Surguchev, A. & Surguchov, A. Conformational diseases: looking into the eyes. Brain Res. Bull 81, 12–24 (2010).
4. Lakshminarayanan, R. et al. Biochemical properties and aggregation propensity of transforming growth factor-induced protein
(TGFβ Ip) and the amyloid forming mutants. Ocul Surf. 13(1), 9–25 (2015).
5. Kitahama, S. et al. Expression of fibrillins and other microfibril-associated proteins in human bone and osteoblast-like cells. Bone
27, 61–67 (2000). 3. Surguchev, A. & Surguchov, A. Conformational diseases: looking into the eyes. Brain Res. Bull 81, 12 24 (2010).
4. Lakshminarayanan, R. et al. Biochemical properties and aggregation propensity of transforming growth factor-induced protein
(TGFβIp) and the amyloid forming mutants Ocul Surf 13(1), 9–25 (2015) References
1. Weiss, J. S. et al. The IC3D classification of the corneal dystrophies. Cornea 27 (Suppl. 2) S1–S83 (2008).
2. Klintworth, G. K. Corneal dystrophies. Orphanet J. Rare Dis. 4, 7 (2009).
3. Surguchev, A. & Surguchov, A. Conformational diseases: looking into the eyes. Brain Res. Bull 81, 12–24 (2010).
4. Lakshminarayanan, R. et al. Biochemical properties and aggregation propensity of transforming growth factor-induced protein
(TGFβ Ip) and the amyloid forming mutants. Ocul Surf. 13(1), 9–25 (2015).
5. Kitahama, S. et al. Expression of fibrillins and other microfibril-associated proteins in human bone and osteoblast-like cells. Bone
27, 61–67 (2000). www.nature.com/scientificreports/ & Chiti F. Searching for conditions to form stable protein oligomers with amyloid-like
characteristics: The unexplored basic pH. Biochim Biophys Acta. 1804(1), 223–34 (2010). C
S
l C
f
l
f h
f H
F N J M l B l
( )
(
) h
34. Campioni, S. et al. Conformational properties of the aggregation precursor state of HypF-N. J Mol Biol. 379(3), 554–67 (2008)
l
d
l
l
h
f h
fl
b
d
l
d f b l
h
h p
,
p
p
gg g
p
yp
J
( ),
(
)
35. Biancalana, M. & Koide, S. Molecular mechanism of thioflavin-T binding to amyloid fibrils. Biochim. Biophys. Acta. 1804,
1405–1412 (2010).li 36. Groenning, M. Binding mode of thioflavin T and other molecular probes in the context of amyloid fibrils. Current status. J. C
Biol. 3, 1–18 (2010).l (
)
37. Maezawa, I. et al. Congo red and thioflavin-T analogs detect Abeta oligomers. J Neurochem. 104(2), 457–68 (2008). l
38. Grothe, H. L. et al. Altered protein conformation and lower stability of the dystrophic transforming growth factor beta-induced
protein mutants. Mol Vis. 19, 593–603 (2013). 9. Sasahara, K., Yagi, H., Naiki, H. & Goto, Y. Heat-induced Conversion of β 2-Microglobulin and Hen Egg-white Lysozyme into
Amyloid Fibrils J. Mol. Biol. 372, 981–991 (2007). y
40. Cardoso, I. et al. Transthyretin fibrillogenesis entails the assembly of monomers: a molecular model for in vitro assembled
transthyretin amyloid-like fibrils. J Mol Biol 317, 683–95 (2002).f i
41. Smith, D. P., Jones, S., Serpell, L. C., Sunde, M. & Radford, S. E. A systematic investigation into the effect of protein destabilisation
on beta 2-microglobulin amyloid formation. J Mol Biol 330, 943–54 (2003). g
y
(
)
2. Aso, Y., Shiraki, K. & Takagi, M. Systematic analysis of aggregates from 38 kinds of non disease-related proteins: identifying the
intrinsic propensity of polypeptides to form amyloid fibrils. Biosci Biotechnol Biochem 71, 1313–21 (2007).ff i
3. Pfefferkorn, C. M., McGlinchey, R. P. & Lee, J. C. Effects of pH on aggregation kinetics of the repeat domain of a functional amyloid
Pmel17. Proc Natl Acad Sci USA 107(50), 21447–52 (2010).f 44. Karring, H. et al. Differential expression and processing of transforming growth factor beta induced protein (TGFBIp) in the normal
human cornea during postnatal development and aging. Exp Eye Res. 90(1), 57–62 (2010). g p
p
g
g
p y
5. Brunk, U. T. www.nature.com/scientificreports/ www.nature.com/scientificreports/ 6. Ferguson, J. W. et al. The extracellular matrix protein betaIG-H3 is expressed at myotendinous junctions and supports muscle cel
adhesion. Cell Tissue Res. 313, 93–105 (2003). 7. Escribano, J., Hernando, N., Ghosh, S., Crabb, J. & Coca-Prados, M. cDNA from human ocular ciliary epithelium homologous to
beta ig-h3 is preferentially expressed as an extracellular protein in the corneal epithelium. J. Cell. Physiol. 160, 511–521 (1994). 7. Escribano, J., Hernando, N., Ghosh, S., Crabb, J. & Coca-Prados, M. cDNA from human ocular ciliary epithelium h
beta ig h3 is preferentially expressed as an extracellular protein in the corneal epithelium J Cell Physiol 160 511 52 g
p
y
p
p
p
y
8. Skonier, J. et al. beta ig-h3: a transforming growth factor-beta-responsive gene encoding a secreted protein that inh g
y
y
8. Skonier, J. et al. beta ig-h3: a transforming growth factor-beta-responsive gene encoding a secreted protein that inhibit
attachment in vitro and suppresses the growth of CHO cells in nude mice DNA Cell Biol 13 571 584 (1994) 8. Skonier, J. et al. beta ig-h3: a transforming growth factor-beta-responsive gene encoding a secreted protein that inhibits cel
attachment in vitro and suppresses the growth of CHO cells in nude mice. DNA Cell Biol. 13, 571–584 (1994). J
g
g g
p
g
g
p
attachment in vitro and suppresses the growth of CHO cells in nude mice. DNA Cell Biol. 13, 571–584 (1994). pp
g
9. Mashima, Y. et al. Severe form of juvenile corneal stromal dystrophy with homozygous R124H mutation in the keratoepithelin gene
in five Japanese patients. Br J Ophthalmol. 82, 1280–1284 (1998). i
10. Okada, M. et al. Granular corneal dystrophy with homozygous mutations in the kerato-epithelin gene. Am J Ophthalmol. 126,
169–176 (1998). (
)
11. Kannabiran, C. & Klintworth, G. K. TGFBI gene mutations in corneal dystrophies. Human Mutation 27 615–625 (2006). 12. Aldave, A. J. & Sonmez, B. Elucidating the molecular genetic basis of the corneal dystrophies: are we there yet? Arch Ophthalmol. 125, 177–186 (2007). 3. Lakshminarayanan, R. et al. Structure-Function Relationship of TGFβ Ip (BIGH3) Associated Mutants in Corneal Dystrophy. Brit
J. Ophthalmol. 95(10), 1457–62 (2011). p
4. Ołdak, M. et al. Late-onset lattice corneal dystrophy without typical lattice lines caused by a novel mutation in the TGFBI gene
Cornea 33(3), 294–9 (2014). 5. Klintworth, G. K., Valnickova, Z. & Enghild, J. J. References Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 15 www.nature.com/scientificreports/ Wu, J. W. et al. Comparative analysis of human γ D-crystallin aggregati 29. Wu, J. W. et al. Comparative analysis of human γ D-crystallin aggregation under physiological and low pH conditions. PLoS One. 12, 9(11) (2014). 30. Noormägi A., Valmsen K., Tõugu V. & Palumaa P. Insulin Fibrillization at Acidic and Physiological pH Values is Controlled by
Different Molecular Mechanisms. Protein J. 2015 Oct 22 [Epub ahead of print]. 29. Wu, J. W. et al. Comparative analysis of human γ D crystallin aggregation under physiological and low pH conditions. PLoS One. 12, 9(11) (2
30. Noormägi A., Valmsen K., Tõugu V. & Palumaa P. Insulin Fibrillization at Acidic and Physiological pH Values is Controlle 30. Noormägi A., Valmsen K., Tõugu V. & Palumaa P. Insulin Fibrillization at Acidic and Physiological pH Values is Controlled by
Different Molecular Mechanisms Protein J 2015 Oct 22 [Epub ahead of print] 30. Noormägi A., Valmsen K., Tõugu V. & Palumaa P. Insulin Fibrillization at Acidic an
Different Molecular Mechanisms. Protein J. 2015 Oct 22 [Epub ahead of print]. A., Valmsen K., Tõugu V. & Palumaa P. Insulin Fibrillization at A g
,
K ,
g
Different Molecular Mechanisms. Protein J. 2015 Oct 22 [Epub ahead of print]. g
,
K ,
g
y
g
p
y
Different Molecular Mechanisms. Protein J. 2015 Oct 22 [Epub ahead of print]. olecular Mechanisms. Protein J. 2015 Oct 22 [Epub ahead of print] f
1. Kim, B. Y. et al. Corneal dystrophy-associated R124H mutation disrupts TGFBI interaction with Periostin and causes mislocalization
to the lysosome. J Biol Chem. 284(29), 19580–91 (2009).i f
31. Kim, B. Y. et al. Corneal dystrophy-associated R124H mut
to the lysosome. J Biol Chem. 284(29), 19580–91 (2009).i f
31. Kim, B. Y. et al. Corneal dystrophy-associated R124H muta y
p y
p
to the lysosome. J Biol Chem. 284(29), 19580–91 (2009). 32. Andreasen, M. et al. Polymorphic fibrillation of the destabilized fourth fasciclin-1 domain mutant A546T of the Transforming
growth factor-β -induced protein (TGFBIp) occurs through multiple pathways with different oligomeric intermediates. J Biol Chem. 287(41) 34730 42 (2012) to the lysosome. J Biol Chem. 284(29), 19580–91 (2009).i y
2. Andreasen, M. et al. Polymorphic fibrillation of the destabilized fourth fasciclin-1 domain mutant A546T of the Transforming
growth factor-β -induced protein (TGFBIp) occurs through multiple pathways with different oligomeric intermediates. J Biol Chem
287(41), 34730–42 (2012). 33. Ahmad B., Winkelmann J., Tiribilli B. www.nature.com/scientificreports/ Accumulation of beta ig-h3 gene product in corneas with granular dystrophy. Am
J Pathol 152, 743–748 (1998). 6. Klintworth, G. K., Bao, W. & Afshari, N. A. Two mutations in the TGFBI (BIGH3) gene associated with lattice corneal dystrophy in
an extensively studied family. Invest Ophthalmol Vis Sci. 45, 1382–8 (2004).f y
y
p
7. Chiti, F., Stefani, M., Taddei, N., Ramponi, G. & Dobson, C. M. Rationalization of the effects of mutations on peptide and protein
aggregation rates. Nature 424(6950), 805–8 (2003).i 18. Yuan, C., Berscheit, H. L. & Huang, A. J. Identification of an amyloidogenic region on keratoepithelin via synthetic peptides. FEBS
Lett 581, 241–247 (2007).l 9. Schmitt-Bernard, C. F. et al. BIGH3 (TGFBI) Arg124 mutations influence the amyloid conversion of related peptides in vitro. Eur
Biochem 269, 5149–5156 (2002). 0. Clout, N. J. & Hohenester, E. A model of FAS1 domain 4 of the corneal protein beta(ig)-h3 gives a clearer view on cornea
dystrophies. Mol Vis 9, 440–448 (2003). y
p
1. Elavazhagan, M. et al. Expression, purification and characterization of fourth FAS1 domain of TGFβ Ip-associated cornea
dystrophic mutants. Protein Expr Purif. 84(1), 108–15 (2012).h 22. Atchaneeyasakul, L. O. et al. Study Group. A novel H572R mutation in the transforming growth factor-beta-induced gene in a Thai
family with lattice corneal dystrophy type I. Jpn J Ophthalmol. 50(5), 403–8 (2006).f y
y
p y yp
p
p
23. Romero, P., Moraga, M. & Herrera, L. Different phenotypes of lattice corneal dystrophy type I in patients with 417C > T (R
and 1762A > G (H572R) mutations in TGFBI (BIGH3). Mol Vis. 16, 1601–9 (2010). 24. Munier, F. L. et al. BIGH3 mutation spectrum in corneal dystrophies. Invest Ophthalmol Vis Sci. 43, 949–954 (2002). 25. Fujiki, K., Nakayasu, K. & Kanai, A. Corneal dystrophies in Japan. J Hum Genet 46, 431–435 (2001). j
y
y
p
p
26. Chakravarthi, S. V. V. K., Kannabiran, C., Sridhar, M. S. & Vemuganti, G. K. TGFBI gene mutations causing lattice and granular
corneal dystrophies in Indian patients. Invest Ophthalmol Vis Sci 46, 121–125 (2005). 7. Dighiero, P. et al. Clinical, histologic, and ultrastructural features of the corneal dystrophy caused by the R124L mutation of the
BIGH3 gene. Ophthalmology 107(7), 1353–7 (2000).f g
p
gy
8. Kocak-Altintas, A. G., Kocak-Midillioglu, I., Akarsu, A. N. & Duman, S. BIGH3 gene analysis in the differential diagnosis of cornea
dystrophies. Cornea 20(1), 64–8 (2001). 29. www.nature.com/scientificreports/ Proc Natl Acad Sci USA 94, 14338–42 (1
59 Pace, C N & Shaw, K L Linear extrapolation method of analyzing solvent denaturation curves Proteins Suppl 4, 1–7 (2000) p
y
g
pp
(
)
0. Urcan, E. et al. Real-time xCELLigence impedance analysis of the cytotoxicity of dental composite components on human gingiva
fibroblasts. Dental Materials 26, 51–58 (2010). Acknowledgementsh g
The authors would like to thank Dr. Gary Yam for his helpful insights. The project was funded by NMRC
Singapore (CS-IRG grant R978/87/2012). The authors would like to thank Dr. Gary Yam for his helpful insights. The project was funded by NMRC
Singapore (CS-IRG grant R978/87/2012). Author Contributions E.M., L.R. and J.S.M. conceived the experiments and wrote the paper. E.M. cloned the genes, expressed the
proteins, performed the CD fluorescence, TEM experiments and prepared the oligomers for MS experiments
and cytotoxicity studies. A.V., V.M. and N.M. assisted in the expression and purification of the proteins. Z.L. performed the MS experiments. N.M. and E.G. assisted with the CD experiments on thermal denaturation. R.R.Y.L., G.S.L.P. and S.S.C. performed the cytotoxicity assays. R.W.B. and S.S.C. provided inputs to the
conception of the work and editing the article. Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 www.nature.com/scientificreports/ & Ericsson, J. L. Cytochemical evidence for the leakage of acid phosphatase through utrastructurally intact lysosoma
membranes. Histochem J 4, 479–491 (1972). 46. Schneider, L., Korber, A., Grabbe, S. & Dissemond, J. Influence of pH on wound-healing: a new perspective for wound
therapy ? Archives of Dermatological Research 298, 413–420 (2007). 47. Huynh, K. K. & Grinstein, S. Regulation of vacuolar pH and its modulation by some microbial species. Microbiol Mol Biol Rev. 71,
452–462 (2007). 8. Runager, K. et al. Human phenotypically distinct TGFBI corneal dystrophies are linked to the stability of the fourth FAS1 domain o
TGFBIp. J. Biol Chem. 286(7), 4951–8 (2011). Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 16 www.nature.com/scientificreports/ www.nature.com/scientificreports/ 49. Underhaug, J. et al. Mutation in transforming growth factor beta induced protein associated with granular corneal dystrophy type 1
reduces the proteolytic susceptibility through local structural stabilization. Biochimica et Biophysica Acta 1834, 2812–2822 (2013). l
fft
h
ll
β
l
h
l
l h
’ d
d
d 50. Klein, W. L., Krafft, G. A. & Finch, C. Targeting small Aβ oligomers: the solution to an Alzheimer’s disease conundrum? Tr
Neurosci. 24, 219–224 (2001). 1. Ferreira, S. T., Vieira, M. N. & De Felice, F. G. Soluble protein oligomers as emerging toxins in Alzheimer’s and other amyloid
diseases. IUBMB Life 59, 332–345 (2007).i f
(
)
2. Volles, M. J. & Lansbury, P. T. Jr. Vesicle permeabilization by protofibrillar alpha-synuclein is sensitive to Parkinson’s disease-linked
mutations and occurs by a pore-like mechanism. Biochemistry 41(14), 4595–602 (2002). f
52. Volles, M. J. & Lansbury, P. T. Jr. Vesicle permeabilization by protofibrillar alpha-synuclein is se
mutations and occurs by a pore-like mechanism. Biochemistry 41(14), 4595–602 (2002). 53. Sokolov, Y. et al. Soluble amyloid oligomers increase bilayer conductance by altering dielectric structure. J Gen Physiol. 12
637–47 (2006).f 54. Valincius, G. et al. Soluble amyloid beta-oligomers affect dielectric membrane properties by bilayer insertion and domain forma
implications for cell toxicity. Biophys J. 95(10), 4845–61 (2008).i p
y
p y
55. Srinivasan, S. et al. Pathogenic serum amyloid A 1.1 shows a long oligomer-rich fibrillation lag phase contrary to the h
amyloidogenic non-pathogenic SAA2.2. J Biol Chem. 288(4), 2744–55 (2013). 56. Pace, C. N. & Sholtz, J. M. Protein Structure: A Practical Approach (Creighton T. E. ed) pp. 299–321, IRL Press, Oxford, UK. 56. Pace, C. N. & Sholtz, J. M. Protein Structure: A Practical Approa pp
g
pp
57. Benjwal, S., Verma, S., Rohm, K. H. & Gursky, O. Monitoring protein aggregation during thermal unfolding in circular dich
experiments. Protein science: a publication of the Protein Society 15, 635–9 (2006).h p
p
f
y
58. Bullock, A. N. et al. Thermodynamic stability of wild-type and mutant p53 core domain. Proc Natl Acad Sci USA 94, 14338–42 (1997). p
p
f
y
58. Bullock, A. N. et al. Thermodynamic stability of wild-type and mutant p53 core domain. Additional Information upplementary information accompanies this paper at http://www.nature.com/srep Supplementary information accompanies this paper at http://www.nature.com/srepi Competing financial interests: The authors declare no competing financial interests. Competing financial interests: The authors declare no competing financial interests. How to cite this article: Murugan, E. et al. pH Induced Conformational Transitions in the Transforming
Growth Factor β-Induced Protein (TGFβIp) Associated Corneal Dystrophy Mutants. Sci. Rep. 6, 23836; doi:
10.1038/srep23836 (2016). How to cite this article: Murugan, E. et al. pH Induced Conformational Transitions in the Transforming
Growth Factor β-Induced Protein (TGFβIp) Associated Corneal Dystrophy Mutants. Sci. Rep. 6, 23836; doi:
10.1038/srep23836 (2016). This work is licensed under a Creative Commons Attribution 4.0 International License. The images
or other third party material in this article are included in the article’s Creative Commons license,
unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license,
users will need to obtain permission from the license holder to reproduce the material. To view a copy of this
license, visit http://creativecommons.org/licenses/by/4.0/ This work is licensed under a Creative Commons Attribution 4.0 International License. The images
or other third party material in this article are included in the article’s Creative Commons license,
unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license,
users will need to obtain permission from the license holder to reproduce the material. To view a copy of this
license, visit http://creativecommons.org/licenses/by/4.0/ Scientific Reports | 6:23836 | DOI: 10.1038/srep23836 17
|
https://openalex.org/W3153864474
|
https://europepmc.org/articles/pmc8459249?pdf=render
|
English
| null |
Overcoming establishment thresholds for peat mosses in human‐made bog pools
|
Ecological applications
| 2,021
|
cc-by
| 10,937
|
Ecological Applications, 31(6), 2021, e02359 Ecological Applications, 31(6), 2021, e02359 g
pp
,
( ),
,
© 2021 The Authors. Ecological Applications published by Wiley Periodicals LLC on behalf of Ecological Society of America. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any
medium, provided the original work is properly cited. Overcoming establishment thresholds for peat mosses in
human-made bog pools Key words:
alternative stable states; rewetting; raised bog; peat moss; Sphagnum; terrestrialization. surface (Yu et al. 2010), they store ~20% of the global
soil organic carbon (Scharlemann et al. 2014, Nichols
and Peteet 2019, G¨unther et al. 2020). Furthermore,
peatlands retain freshwater, provide clean drinking water
and biodiversity (Limpens et al. 2008, Lamers et al. 2015). However, peatlands are being drained at a large
scale to facilitate agriculture, forestry, and peat extrac-
tion (Swindles et al. 2019). Drainage leads to the emis-
sion of greenhouse gases, deterioration of groundwater Overcoming establishment thresholds for peat mosses in
human-made bog pools Smolders, T. J. Bouma, G. S. Fivash, W. Lengkeek,
K. Didderen, L. P. M. Lamers, and T. van der Heide. 2021. Overcoming establishment thresholds for peat
mosses in human-made bog pools. Ecological Applications 31(6):e02359. 10.1002/eap.2359 Abstract. Globally, peatlands have been affected by drainage and peat extraction, with
adverse effects on their functioning and services. To restore peat-forming vegetation, drained
bogs are being rewetted on a large scale. Although this practice results in higher groundwater
levels, unfortunately it often creates deep lakes in parts where peat was extracted to greater
depths than the surroundings. Revegetation of these deeper waters by peat mosses appears to
be challenging due to strong abiotic feedbacks that keep these systems in an undesired bare
state. In this study, we theoretically explore if a floating peat mat and an open human-made
bog lake can be considered two alternative stable states using a simple model, and experimen-
tally test in the field whether stable states are present, and whether a state shift can be accom-
plished using floating biodegradable structures that mimic buoyant peat. We transplanted two
peat moss species into these structures (pioneer sp. Sphagnum cuspidatum and later-
successional sp. S. palustre) with and without additional organic substrate. Our model suggests
that these open human-made bog lakes and floating peat mats can indeed be regarded as alter-
native stable states. Natural recovery by spontaneous peat moss growth, i.e., a state shift from
open water to floating mats, is only possible when the water table is sufficiently shallow to
avoid light limitation (<0.29 m at our site). Our experiment revealed that alternative stable
states are present and that the floating structures facilitated the growth of pioneer S. cuspida-
tum and vascular plants. Organic substrate addition particularly facilitated vascular plant
growth, which correlated to higher moss height. The structures remained too wet for the late-
successional species S. palustre. We conclude that open water and floating peat mats in
human-made bog lakes can be considered two alternative stable states, and that temporary
floating establishment structures can induce a state shift from the open water state to peat-
forming vegetation state. These findings imply that for successful restoration, there is a clear
water depth threshold to enable peat moss growth and there is no need for addition of large
amounts of donor-peat substrate. Correct species selection for restoration is crucial for success. Manuscript received 22 September 2020; revised 6 January
2021; accepted 4 February 2021. Corresponding Editor: Hinsby
Cadillo-Quiroz.
9 Overcoming establishment thresholds for peat mosses in
human-made bog pools RALPH J. M. TEMMINK
,1,2,9 PETER M. J. M. CRUIJSEN,1 ALFONS J. P. SMOLDERS,1,3 TJEERD J. BOUMA,4,5,6,7 GREGORY
S. FIVASH,4 WOUTER LENGKEEK,1,8 KARIN DIDDEREN,8 LEON P. M. LAMERS,1,3 AND TJISSE VAN DERHEIDE1,7,2 RALPH J. M. TEMMINK
,1,2,9 PETER M. J. M. CRUIJSEN,1 ALFONS J. P. SMOLDERS,1,3 TJEERD J. BOUMA,4,5,6,7 GREGORY
S. FIVASH,4 WOUTER LENGKEEK,1,8 KARIN DIDDEREN,8 LEON P. M. LAMERS,1,3 AND TJISSE VAN DERHEIDE1,7,2 1Aquatic Ecology and Environmental Biology, Institute for Water and Wetland Research, Radboud University, He
Nijmegen 6525 AJ, the Netherlands
2 1Aquatic Ecology and Environmental Biology, Institute for Water and Wetland Research, Radboud University, Heyendaalseweg 135,
Nijmegen 6525 AJ, the Netherlands j
g
2Department Coastal Systems, Royal Netherlands Institute of Sea Research and Utrecht University, Landsdiep 4, ’t Hortje (Texel)
1797 SZ, the Netherlands j
2Department Coastal Systems, Royal Netherlands Institute of Sea Research and Utrecht University, Landsdiep 4
1797 SZ, the Netherlands
3 3B-WARE Research Centre, Toernooiveld 1, Nijmegen 6525 ED, the Netherlands
4Department of Estuarine and Delta Systems, Royal Netherlands Institute of Sea Research and Utrecht University, Korringaweg 7,
Yerseke 4401 NT, the Netherlands 3B-WARE Research Centre, Toernooiveld 1, Nijmegen 6525 ED, the Netherlands
ne and Delta Systems, Royal Netherlands Institute of Sea Research and Utrecht University, Korringaweg 7,
Yerseke 4401 NT, the Netherlands 3B-WARE Research Centre, Toernooiveld 1, Nijmegen 6525 ED, the Netherlands
4Department of Estuarine and Delta Systems, Royal Netherlands Institute of Sea Research and Utrecht University, Korringaweg 7,
Yerseke 4401 NT, the Netherlands B-WARE Research Centre, Toernooiveld 1, Nijmegen 6525 ED, the Netherlands
4Department of Estuarine and Delta Systems, Royal Netherlands Institute of Sea Research and Utrecht Universi
Yerseke 4401 NT, the Netherlands
5 5Delta Academy Applied Research Centre, HZ University of Applied Sciences, Vlissingen, the Netherlands
6Faculty of Geosciences, Department of Physical Geography, Utrecht University, Princetonlaan 8a, Utrecht 3584 CB, the Netherlands
7Conservation Ecology Group, Groningen Institute for Evolutionary Life Sciences, University of Groningen, Nijenborgh 7, Groningen
9747 AG, the Netherlands 8Bureau Waardenburg, Varkensmarkt 9, Culemborg 4101 CK, the Netherlands 8Bureau Waardenburg, Varkensmarkt 9, Culemborg 4101 CK, the Netherlands Citation: Temmink, R. J. M., P. M. J. M. Cruijsen, A. J. P. Smolders, T. J. Bouma, G. S. Fivash, W. Lengkeek,
K. Didderen, L. P. M. Lamers, and T. van der Heide. 2021. Overcoming establishment thresholds for peat
mosses in human-made bog pools. Ecological Applications 31(6):e02359. 10.1002/eap.2359 Citation: Temmink, R. J. M., P. M. J. M. Cruijsen, A. J. P. INTRODUCTION Peatlands provide vital ecological and socioeconomic
services at a global scale (Joosten and Clarke 2002). While they only account for ~3% of the terrestrial Manuscript received 22 September 2020; revised 6 January
2021; accepted 4 February 2021. Corresponding Editor: Hinsby
Cadillo-Quiroz. 9 9 E-mail: r.temmink@science.ru.nl Article e02359; page 1 Ecological Applications
Vol. 31, No. 6 Ecological Applications
Vol. 31, No. 6 RALPH J. M. TEMMINK ET AL. Article e02359; page 2 and surface water quality, and subsidence of peat soils
(Schothorst 1977, Verhoeven and Setter 2010, Lamers
et al. 2015). However, extraction and drainage also cre-
ates unnatural landscapes with altered hydrology and
variable thickness of the remaining peat layer (Haapale-
hto et al. 2014). situations is difficult once the abundance of the habitat-
modifying species drops below a critical density or
patch-size threshold. Well-known examples of ecosys-
tems with habitat-modifying species generating strong
positive feedbacks are submerged aquatic vegetation in
shallow lakes and coastal seagrasses that reduce turbid-
ity through trapping of suspended particles and stimu-
late phytoplankton grazing by providing refuge to
zooplankton (Scheffer et al. 2001, van der Heide et al. 2007, 2010). To restore peat-forming vegetation in degraded peat
bogs (formerly dominated by peat mosses, Sphagnum
spp.), the construction of dams that reduce water loss by
lateral flow and maintain a permanently high groundwa-
ter table is an often applied approach (Schumann and
Joosten 2008, Parry et al. 2014, Altenburg et al. 2017). Although this successfully elevates groundwater tables, it
typically also causes deeply excavated areas to turn into
relatively deep bog lakes. These bog lakes are frequently
without vegetation even decades after rewetting. Numer-
ous examples of remnant bogs with large bodies of open
water can be found in the Netherlands, Germany, the
Baltic States, the United Kingdom (Fig. 1), and in
North America (Quinty and Rochefort 2000). This
implies that although the restoration measures are partly
successful, large parts of the previously drained areas
remain as persistent open water without the targeted
bog vegetation. ,
)
In bogs, peat mosses (Sphagnum spp.) are the domi-
nant habitat-modifying species (Van Breemen 1995). As
they increase in biomass, density, and patch-size, they
increasingly
alter
their
environment
by
retaining
nutrient-poor rainwater, acidification, and the accumu-
lation of organic material (Van Breemen 1995, Lamers
et al. 2000, Soudzilovskaia et al. 2010). In this way,
mosses improve their own growing conditions with
increasing moss abundance, yielding a positive feedback. INTRODUCTION Peat mosses colonize bog lakes through a process called
terrestrialization. This process naturally occurs via the
formation of a stable, permanent floating mat consisting
of live mosses, structuring vascular plants, and dead
organic material, which can initiate at the lake bottom,
from the shore or from free-floating mosses. The mosses
produce oxygen from photosynthesis, while degradation
of organic matter by its associated microbial community
produces methane (CH4) and carbon dioxide (CO2). Buoyance is generated by oxygen that accumulates inside
the mosses and the other gasses that are trapped inside
and beneath the peat mat (Smolders et al. 2002, Tomas-
sen et al. 2003b, 2004; Appendix S1: Fig. S1). Once
afloat, the peat mosses are exposed to atmospheric CO2
and have ample light for growth; conditions that further
stimulate their growth and subsequent organic matter
accumulation. In general, persistent degraded states often occur in
disturbed ecosystems that are naturally controlled by
positive feedback mechanisms. Such feedbacks are typi-
cally generated by spatially dominant habitat-forming
organisms, which modify their surroundings to their
own benefit by reducing physical stress or increasing
resource availability (Stachowicz 2001). However, these
beneficial modifications often only work beyond a cer-
tain minimum density or patch-size of the habitat modi-
fier, yielding a positive (i.e., self-reinforcing) feedback on
its own growth. Unsuccessful colonization (e.g., too low
density or patch-size) of the species may cause alterna-
tive stable states (Scheffer et al. 2001, van der Heide
et al. 2007). This implies that, either a state that is domi-
nated by the habitat-modifying species or an alternative
(often bare) state is stable under the same environmental
conditions. Importantly,
natural
recovery
in
such In deep (human-made) bog lakes, vegetation is often
persistently lacking, as peat moss colonization from the
bottom is hampered by light limitation due to humic
substances in the water column (Smolders et al. 2003). Additionally, colonization by free-floating mosses is also A) The Netherlands
B) Germany
C) United Kingdom
FIG. 1. Examples of unnatural open water in former natural raised bogs. (A) Bargerveen in the Netherlands, (B) Tister Bauern-
moor in Germany, and (C) Shapwick Heath in the United Kingdom. Maps from Google Earth. A) The Netherlands C) United Kingdom B) Germany C) United Kingdom B) Germany FIG. 1. Examples of unnatural open water in former natural raised bogs. (A) Bargerveen in the Netherlands, (B) Tister Bauern-
moor in Germany, and (C) Shapwick Heath in the United Kingdom. Maps from Google Earth. THRESHOLDS FOR BOG RESTORATION Article e02359; page 3 September 2021 hampered, as these mats typically do not become thick
enough during the growing season to sustain buoyancy
in winter when production of oxygen, CO2 and CH4 are
all decreased (Smolders et al. 2003, Tomassen et al. 2004). Once the mosses are at the bottom, light and dis-
solved CO2 levels can be too low to allow sufficient pho-
tosynthesis
and
oxygen
production
required
for
resurfacing in the next growing season (Paffen and Roe-
lofs 1991, Smolders et al. 2003). Often, the threshold of
5% light, the minimum level required for submerged peat
moss growth, occurs at a depth 0.2–0.5 m when waters
are highly colored (Streefkerk and Casparie 1989, Money
and Wheeler 1999, Smolders et al. 2003). This implies that
mat formation beyond this depth is not possible. In addi-
tion, peat moss growth is also limited by low dissolved
carbon availability (<750 µmol CO2/L; Paffen and Roe-
lofs 1991, Patberg et al. 2013). Apart from issues with
light and carbon limitation, early forming peat mats that
manage to become afloat can also be hampered by wind-
generated waves that break the mats apart, a factor that
may be particularly important in larger lakes where fetch
lengths are long (Wheeler and Shaw 1995, Smolders et al. 2003). These difficulties for re-establishment of the origi-
nal peat system raises the question to what extent positive
feedbacks, and the potential for alternative stable states,
play a role in the persistence of these human-made bog
lakes. and related system dynamics in isolation) based upon a
combination of literature data (Boatman 1977, Hayward
and Clymo 1983, Money 1995, Smolders et al. 2002, 2003,
Rochefort et al. 2003) and field measurements in Fochte-
lo¨erveen, the Netherlands. Specifically, the model focuses
on light limitation as the simplest possible explanation for
alternative stable states, thus ignoring other potentially
exacerbating factors such as CO2 limitation at the bottom
or waves once the mat is afloat. A) The Netherlands Hence, it describes the
relation between peat moss growth and white peat accu-
mulation, dependent on light availability, which in turn
depends on water depth, water transparency and on
whether the peat mat is buoyant or not. We hypothesize
that the model, with parameters calibrated using empirical
data, can generate alternative stable states in a range of
realistic water transparencies and depths. Next, to investigate whether alternative stable state
conditions are present in the field and if they can be
overcome, we tested a novel restoration framework
(Temmink et al. 2020). In an experiment, we use
biodegradable floating establishment structures (here-
after, structures for brevity) that mimic buoyant peat. In
the structures, peat mosses have ample light and CO2 for
growth, as well as structural reinforcement to increase
resistance against small waves (Fig. 2D, E). We expect
that the temporary structures allow mosses to proliferate
and ultimately form dense floating mats. The structures
can naturally degrade once the mats generate buoyancy
themselves (Tomassen et al. 2004), are sufficiently coher-
ent to resist waves, and can act as nuclei for lateral
growth. To test the restoration potential of our frame-
work, we selected a human-made bog lake with highly
colored (E450 = 0.17, 5% light penetration threshold at
0.3 m) and carbon limited water layer, which is too deep
(~0.6 m) to support peat moss growth at the bottom. We
selected the pioneer Sphagnum cuspidatum and late-
succession species S. palustre as model species (Daniels
and Eddy 1985, Frahm and Frey 2003) that were trans-
planted inside structures with and without organic sub-
strate. We hypothesize that (1) the lake has two
alternative state states as described by the model, (2) the
structure physically supports peat mosses and ensures
sufficient light and CO2 and enables to mosses to resist
waves, and (3) when enriched with organic material it
provides additional CO2 and nutrients, further stimulat-
ing growth and peat mat development. The idea for the existence of alternative stable open
water and floating mat states in deep bog lakes is not
only supported by observations. In a number of small-
scale bog-restoration research projects, aiming to revege-
tate deep bog lakes, this was achieved this by introducing
poorly humified peat mats with a living top layer (white
peat, German: Bunkerde). A) The Netherlands In these cases, the mats
became buoyant due to CO2 and CH4 formation,
thereby overcoming apparent CO2- and transparency-
related establishment thresholds for peat mosses (Money
and Wheeler 1999, Smolders et al. 2002, Tomassen et al. 2003b, 2004). However, this approach has clear down-
sides, because (1) pristine donor sites are damaged by
the harvest of transplant material and (2) the quality of
white peat can differ greatly; from slow degrading
nutrient-poor pristine bog peat dominated by peat
mosses to dry and eutrophic top soils dominated by
Molinia that rapidly degrade (Money and Wheeler 1999,
Lamers et al. 2000). Consequently, restoration measures
to aid the terrestrialization of deep human-made bog
pools, without using unsustainable black or white peat,
are currently lacking. MATERIALS AND METHODS In this study, we therefore (1) explore whether open
water and floating peat mats can indeed be considered
two alternative stable states, and (2) test if establishment
thresholds caused by low transparency, stress from CO2
limitation and wind-generated waves can be overcome
by using temporary floating support structures. First, to
test whether the hypothesized alternative stable states
are theoretically possible, we constructed a minimal
model (i.e., a simplified model for exploring processes Model description Many lakes in rewetted raised bogs have either been
fully colonized by vegetation or persist for decades with-
out any significant vegetation development (Smolders
et al. 2003, Patberg et al. 2013). We constructed a mini-
mal model that explores the basic nature of the terrestri-
alization process in a qualitative manner (Couwenberg
and Joosten 2005, van der Heide et al. 2007, 2010). Ecological Applications
Vol. 31, No. 6 RALPH J. M. TEMMINK ET AL. Article e02359; page 4 the Netherlands
A
C
D
B
E
G
Module 1
Sp. 1
F
Module 2
Sp. 2
FIG. 2. Field site and experimental units. (A) The Netherlands, where the Fochtelo¨erveen is indicated by a red circle, (B) open
water without terrestrialization, (C) floating peat mat consisting of Sphagnumcuspidatum, (D) peat moss application to the estab-
lishment structure, (E) setup of two modules containing different species, and (F) experimental plots after setup with caution tapes
to exclude birds. (G) Overview of the four original experimental treatments, where peat moss species are indicated with different col-
ors. The control with peat moss was omitted (see Materials and Methods: Field experiment for details). Pictures: R. J. M. Temmink. Map made with Natural Earth. C the Netherlands
A B C B D
G F E
Module 1
Sp. 1
F
Module 2
Sp. 2 D F E G FIG. 2. Field site and experimental units. (A) The Netherlands, where the Fochtelo¨erveen is indicated by a red circle, (B) open
water without terrestrialization, (C) floating peat mat consisting of Sphagnumcuspidatum, (D) peat moss application to the estab-
lishment structure, (E) setup of two modules containing different species, and (F) experimental plots after setup with caution tapes
to exclude birds. (G) Overview of the four original experimental treatments, where peat moss species are indicated with different col-
ors. The control with peat moss was omitted (see Materials and Methods: Field experiment for details). Pictures: R. J. M. Temmink. Map made with Natural Earth. fl ¼ 1e I=Ik
ð
Þ
(2) Specifically, this model describes the relation between light,
peat moss growth, and white peat accumulation. Peat moss
growth, and consequently white peat accumulation, is
dependent on light availability, which in turn depends on
water depth, water transparency and on whether the peat
mat is buoyant or not. Model parameter settings were
derived from direct field measurements or literature data
(Table 1, Appendix S1: Table S1). THRESHOLDS FOR BOG RESTORATION D ¼ Dmax if M <Mc andD ¼ 0 if M ≥Mc
(5) (5) where Mc is the critical mat thickness required for float-
ing. To investigate whether alternative stable states occur
in our minimal bog lake simulation model, we explored
its sensitivity to the depth of the lake through a bifurca-
tion analysis using GRIND for Matlab. In such an anal-
ysis,
potential
critical
transition
thresholds
and
hysteresis (i.e., parameter range where two states can
occur) are determined by a numerical procedure in
which a key parameter value is increased and subse-
quently decreased again in small steps (van Nes 2017). In our case, we decreased Dmax from 1 to 0 m in stepwise
increments of 0.01 m. At each step, the model was left to
stabilize for 10,000 yr. Next, values of S and M were
recorded, after which the value of Dmax was reduced and
the next 10,000 yr of simulation followed. Once a Dmax
of 0 was reached, we performed the same numerical pro-
cedure in the opposite direction. Experimental setup.—The experiment consisted of four
treatments, replicated eight times (randomized block
design). The treatments consisted of (1) an unmanipu-
lated control, (2) control with peat moss addition, (3)
structure with peat moss addition, and (4) structure with
organic material and peat moss (Fig. 2F, Appendix S1:
Fig. S3 for the layout of the eight replicate blocks in the
lake). Plots were constructed in March 2017 and were
harvested in July 2019, after a 28-month experimental
period. THRESHOLDS FOR BOG RESTORATION THRESHOLDS FOR BOG RESTORATION Article e02359; page 5 September 2021 TABLE 1. Variables and default parameter settings of the peat moss model. Default
value
Unit
Description
Sources
Variables
S
m
living peat moss layer thickness
M
m
mat thickness
Parameters
d
0.001
per day
mat decay rate
based on empirically measured mat thickness
of 0.5 m
Dmax
0.5
m
depth of the lake
estimated from field data
I0
500
molm−2s−1
growing-season-averaged irradiance at the
surface
Royal Dutch Meteorological Institute (2019)
Ik
200
molm−2s−1
half-saturation irradiance constant
Titus and Wagner (1984)
K
0.1
m
carrying capacity of S
Clymo (1970)
k
7.2
per m
light attenuation coefficient
estimated from field data
m
0.0052
per day
relative mortality rate
estimated from field data
Mc
0.1
m
critical mat thickness for floating
estimated from field data
r
0.02
per day
relative growth rate
derived from Clymo (1970)
Notes: We chose default parameter settings to mimic average conditions in raised bogs in the Netherlands, based on literature
and our field data. TABLE 1. Variables and default parameter settings of the peat moss model. Notes: We chose default parameter settings to mimic average conditions in raised bogs in the Netherlands, based on literature
d
fi ld d t a critical thickness of M is reached, causing the depth of
the live peat layer to change from Dmax to 0; i.e. field experiment was conducted in the Fochtelo¨erveen, a
rewetted bog remnant in the Netherlands (Fig. 2,
53°0’22.78" N, 6°22’24.75" E). The bog has been heavily
exploited for peat extraction, drainage-based agriculture,
and buckwheat fire culture during the last centuries. Rewetting via dams resulted in a raised water table, and
in a higher cover of peat mosses (Altenburg et al. 2017),
but also in the formation of large unvegetated lakes. We
selected a wave-sheltered part of a 3-ha bog lake (the
dominant wind direction originates from the southwest
in the Netherlands). The water layer in this area was
0.6 0.09 m deep (mean SE) at setup (see Appendix
S1: Fig. S2 for the water layer depth through time), was
highly colored (E450 = 0.17), and was low in dissolved
CO2 concentrations in the water layer (120 µmol/L). The
bottom consisted of a 0.69 0.04 m thick peat layer,
without any peat mosses, presumably due to light limita-
tion (see Appendix S1: Table S1 for characteristics of the
water layer). Model description The change in the living
peat moss layer over time is described as Specifically, this model describes the relation between light,
peat moss growth, and white peat accumulation. Peat moss
growth, and consequently white peat accumulation, is
dependent on light availability, which in turn depends on
water depth, water transparency and on whether the peat
mat is buoyant or not. Model parameter settings were
derived from direct field measurements or literature data
(Table 1, Appendix S1: Table S1). The change in the living
peat moss layer over time is described as (2) where I is the light availability, and Ik the saturation irra-
diance constant in PAR. Light availability (I) is calcu-
lated using the Lambert-Beer equation I ¼ Ið0ÞeðkDÞ
(3) (3) with k as the attenuation coefficient and D as the depth
at which the acrotelm (living top layer) occurs. The
change in peat matt thickness (M) is described by the
second differential equation dS
dt ¼ rflS
rm
ð
Þ=K
ð
ÞS2 mS:
(1) (1) Here, S describes the change in thickness of the living
peat moss layer, r is the relative growth rate per unit of
time, and fl is the P-I (photosynthetic irradiation) curve
of peat moss (see Eq. 2). Parameter K is carrying capac-
ity for the living peat moss layer, and m is its relative
mortality rate (see Table 1 for units and default values). The P-I curve of peat moss is described as dM
dt ¼ mS d M
(4) (4) where m is the relative mortality rate of the living peat
moss layer S, and d is the decay rate of the peat mat. Finally, we assume that a peat mat becomes floating once Field experiment The occurrence of alternative stable states highlighted by (A) the living peat moss
layer (S) and (B) the thickness of the peat mat (M) in relation to water depths. The solid green lines represent a floating vege-
tated/peat mat state and the red a bare state (both equilibria are stable beyond a depth of 0.22 m). The dashed black line with the
arrows indicates the direction of change, i.e., when water depth is decreased in our model, spontaneous recovery from the bare (red)
state towards the vegetated state occurs at 0.22 m. measured with an Ag/AgCl electrode (Orion, Thermo
Fisher Scientific, Waltham, Massachusetts, USA) cou-
pled to an 877 Titrino plus (Metrohm, Herisau, Switzer-
land). Total Inorganic Carbon (TIC) was measured
using an infrared carbon analyzer (IRGA; ABB Analyti-
cal, Frankfurt, Germany), after which bicarbonate
(HCO
3 ) and CO2 were calculated based on the pH equi-
librium (e.g., van Bergen et al. 2020). A 10-mL subsam-
ple of each filtered water sample (Glass microfiber
filters, outer diameter 47 mm, GF/C, Whatman, GE
Healthcare UK Limited, Little Chalfont, Bucking-
hamshire, UK) was conserved by adding 0.1 mL of
nitric acid (HNO
3 65%) and stored at 4°C until P analy-
sis by inductively coupled plasma emission spectropho-
tometry (ICP-OES; model IRIS Intrepid II XDL,
Thermo Fisher Scientific, Franklin, Tennessee, USA). The rest of each sample was stored in polyethylene bot-
tles at −20°C prior to analyses. Nitrate (NO
3 ), ammo-
nium (NHþ
4 ), and phosphate (PO3
4 ) were measured
colorimetrically with an auto analyzer (Auto Analyzer
III, Bran and Luebbe GmbH, Norderstedt, Germany). Potassium (K) was determined by flame photometry
(FLM3 Flame Photometer, Radiometer, Copenhagen,
Denmark). Extinction at 450 nm was measured (Double
beam, UV-Vis-spectrophotometer, UV-6300PC, VWR,
Amsterdam, the Netherlands) as an estimate of humic
substance concentration (Kirk 1994, Smolders et al. 2003). of the lowest sheet, after which the second sheet was
clicked on top. Next, we added 340 g fresh mass (~10 g
dry mass [DM]) S. cuspidatum or S. palustre, collected in
the Fochtelo¨erveen, on top of each two-sheet OM module
(Fig. 2D). Finally, we clicked a third sheet on top, result-
ing in a 6-cm thick, three-layer module. To create modules
without OM, we repeated the steps described above, but
without the addition of OM. Finally, we randomly com-
bined two modules containing different species into a
two-module plot (Fig. 2E). Field experiment As hypothesized, our model results suggest that artifi-
cial bog lakes can in theory be characterized by two
alternative stable states (Fig. 3): open water or a floating
peat mat. To further explore this in the field, we used
floating biodegradable structures
to
investigate (1)
whether a state shift from open water to floating peat
mats can be accomplished as a general test for the occur-
rence of alternative stable states and (2) if this approach
can be applied as a potential restoration measure to
overcome critical bottlenecks for peat moss growth. The Each structure consisted of three stacked biodegradable
BESE-sheets
(sheet
dimensions:
91.5 × 45.5 × 2 cm,
length × width × height; BESE Ecosystem Restoration
Products, Culemborg, The Netherlands; Temmink et al. 2020). To create organic matter (OM) modules, we added
2 kg of fresh peat (0.9 kg/L, origin Baltic States) on top Ecological Applications
Vol. 31, No. 6 RALPH J. M. TEMMINK ET AL. Article e02359; page 6 0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Water depth (m)
0
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
(
)
B
Vol. 31, No. 6 Water depth (m)
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
S (m)
0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Water depth (m)
0
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
M (m)
B
A
FIG. 3. Results of the bifurcation analysis. The occurrence of alternative stable states highlighted by (A) the living peat moss
layer (S) and (B) the thickness of the peat mat (M) in relation to water depths. The solid green lines represent a floating vege-
tated/peat mat state and the red a bare state (both equilibria are stable beyond a depth of 0.22 m). The dashed black line with the
arrows indicates the direction of change, i.e., when water depth is decreased in our model, spontaneous recovery from the bare (red)
state towards the vegetated state occurs at 0.22 m. Water depth (m)
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.10
S (m)
0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
A B A .2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Water depth (m) FIG. 3. Results of the bifurcation analysis. Field experiment Similar to the peat moss results, vascular plant biomass,
consisting almost exclusively of Agrostis canina, was
highest in structures with OM (334 90 g DM/m2,
Fig. 4B), 3.3 times lower in structures without OM
(100 50 g DM/m2), and absent in controls (0 0 g
DM/m2, F2,38 = 40.6, P < 0.001). Juncus effusus bio-
mass was absent in the controls (0 0) and was highest
in the structures with (31 17 g DM/m2) and without
OM (23 10 g DM/m2, F2,38 = 4.8, P = 0.0136). were situated on the sediment and not visible due to the
low water clarity. Next, we collected and sorted peat
mosses and vascular plants from each module of each
two-module plot, determined fresh mass and dry mass
(70°C until constant mass). Statistical analyses.—Since only one species survived the
treatments (S. cuspidatum) and no S. palustre was found
after 28 months (Appendix S1: Fig. S3), the effect of
peat moss species was not taken into account when ana-
lyzing the data. To test whether establishment structures
with or without organic matter stimulate peat moss peat
moss biomass, height, and vascular plant biomass
(Agrostis canina), we analyzed the data with general lin-
ear mixed models with a Gaussian distribution and
block as random effect. These analyses were followed by
Tukey post-hoc test (Lenth and Lenth 2018). Data were
square root, reciprocal and log transformed for peat
moss biomass, peat moss height and Agrostis canina bio-
mass, respectively. To test the effect of treatment on Jun-
cus effusus biomass, we used a zero-inflated Gaussian
mixed model, as the data had many zeros (package
NBZIMM). To test whether lateral expansion differed
between structures with or without OM, we analyzed the
data with a Student’s t test. Surface water quality data
were averaged, and the minimum and maximum values
were determined (Appendix S1: Table S1). All analyses
were performed in R (version 3.6) statistical and pro-
gramming environment (R Core Team 2020). All results
are shown with their standard error of the arithmetic
mean (SE) and the significance level is at P < 0.05. Peat
moss
(S. cuspidatum)
height
was
positively
affected by structures, and particularly by structures
with OM, while peat mosses in controls did not grow
vertically (Fig. 5, F2,38 = 1272.4, P < 0.001). Field experiment A two-module plot always
consisted of two species with either a + or −OM treat-
ment. We then attached a PVC tube (2 cm in diameter)
around
the
combined
two-module
plot
(91.5 × 91 × 6 cm, length × width × height) to ensure
floatation (Fig. 2F). We used non-degradable PVC-tubes
for experimental purposes. We placed the plots 2.5 m
apart and secured them between four iron pins, after
which we installed caution tape around each plot to mini-
mize bird disturbance (Gahlert et al. 2010). Plots without
structures were marked with four iron pins and caution
tape was installed. For the “peat moss control treatment,”
the same amount of material was introduced into each
plot, while no peat mosses were introduced in the “con-
trols.” In the “peat moss control,” the added peat mosses
immediately floated away and washed ashore on the same
day. Therefore, this treatment was omitted from further
analyses. Sample analyses.—At every field visit (n = 16, frequency
once every one to two months), surface water level was
determined using a fixed beacon. In addition, we took
surface water samples with a 1-L cup attached to a 2-m
long pole to prevent disturbances. The conductivity of
the surface water was measured in situ (TetraCon 925
probe connected to a Multi 3,420 m; WTW, Weilheim,
Germany). In the laboratory, the pH and alkalinity were After a period of 28 months, we took a photograph of
each plot to determine the lateral expansion of the vege-
tation using ImageJ. Furthermore, we measured the peat
moss lawn height relative to the water table in each plot,
after which we transported the intact plots to the lab in
July 2019. We harvested the homogeneous occurring
S. cuspidatum from the inundated peat in controls and
calculated the biomass for the subplots, because they THRESHOLDS FOR BOG RESTORATION Article e02359; page 7 September 2021 mean SE),
HCO
3
(0.9 0.09 µmol/L),
NHþ
4
(32 6 µmol/L) and was highly colored (0.17 0 E450,
Appendix S1: Table S1 for other parameters). In general,
peat moss biomass was lowest in controls with 9.4 1 g
DM/m2 (F2,38 = 153.8, P < 0.001, Fig. 4A), intermedi-
ate in structures without OM (138 8 g DM/m2) and
highest in structures with OM (197 21 g DM/m2). DISCUSSION Human-made bog lakes in many restored bog rem-
nants in Europe and North America, originate from for-
mer peat extraction and subsequent rewetting. These
waters are often largely without vegetation, most likely
due to a combination of unfavorable abiotic conditions
and establishment thresholds caused by a lack of positive
feedbacks normally generated by the vegetation itself
(Paffen and Roelofs 1991, Smolders et al. 2002, 2003,
Tomassen et al. 2003b). Using a minimal model, we
show how these bog lakes can display alternative stable
state behavior between contrasting bare open water and
vegetated floating mat states driven by light availability,
which is in agreement with our hypothesis. Given this
finding, we conducted a field experiment to investigate
whether such states exist in the field, and simultaneously
test a novel method to induce a state shift from open
water to peat-forming vegetation as a restoration mea-
sure. In accordance with our hypotheses, the floating
biodegradable structures aid in overcoming light and
CO2 limitation and wave stress bottlenecks, allowing the
re-establishment of self-facilitating mechanisms (Fig. 6). Moreover, organic material further stimulated vascular
plant and peat moss growth. In establishing floating peat
mosses, we also provide proof of concept for our
approach as a restoration method, which has as advan-
tages that it does not require large-scale white peat trans-
plantation. Model results The model shows that spontaneous peat moss growth
is only possible when the water depth is between 0 and
0.29 m (Fig. 3). Between 0.22 and 0.29 m depth the
model shows bistability between a 9-cm thick living peat
layer (Fig. 3A) on top of a 0.46-cm floating peat mat
(Fig. 3B) vs. a thin (maximum 0.02 cm) layer of living
peat moss with a thin (maximum 0.1 cm) peat mat that is
insufficient to float. Beyond a depth of 0.29 m, peat
moss is not able to grow at the bottom of the lake at all. Here, peat moss is only able to grow when already afloat,
yielding alternative stable states between a bare state and
a state with a floating mat that is able to sustain itself
irrespective of the water depth. Field experiment In struc-
tures without OM, peat mosses reached a height of
3.2 0.5 cm, while in those with OM the moss layer
was 1.6 times higher (5.1 0.4 cm;). Interestingly, peat
moss height was positively related to the vascular plant
biomass (R2 = 0.6, P < 0.001). A. canina accompanied
by S. cuspidatum laterally expanded 53 6 cm outside
of
the
structure,
irrespective
of
OM
treatment
(P = 0.36). Field results Structures positively affected peat moss and vascular
plant biomass in the unvegetated human-made lake
(Fig. 4), and this effect was enhanced by the addition of
organic material (OM). The surface water was character-
ized by low concentrations of CO2 (120 10 µmol/L, Ecological Applications
Vol. 31, No. 6 RALPH J. M. TEMMINK ET AL. Article e02359; page 8 Control
ES
ES + OM
0
100
200
300
400
500
Biomass (g DM/m2)
a
b
c
Control
ES
ES + OM
0
100
200
300
400
500
Biomass(g DM/m2)
Juncus effusus (*)
Agrostis canina (***)
0
a | a
b
c
b
b
E) Vascular plants
D) Peat moss
A) Control
B) ES
C) ES + OM
FIG. 4. Peat moss and vascular plant biomass. Pictures of the three treatments after 28 months (A, control; B, structure; C,
structure with organic material). (D) Peat moss Sphagnumcuspidatum and (E) vascular plant biomass (n = 16; DM, dry mass) in
controls, structures (ES) and structures with organic material (OM). Error bars represent SE. Significant contrasts are indicated by
different letters. Asterisks indicate significant treatment effect on vascular plant biomass for Agrostis canina (***P < 0.001) and
Juncus effusus (*P = 0.01) with different letters for significant contrasts for each species. B) ES
C) ES + OM B) ES C) ES + OM C) ES + OM A) Control D) E) FIG. 4. Peat moss and vascular plant biomass. Pictures of the three treatments after 28 months (A, control; B, structure; C,
structure with organic material). (D) Peat moss Sphagnumcuspidatum and (E) vascular plant biomass (n = 16; DM, dry mass) in
controls, structures (ES) and structures with organic material (OM). Error bars represent SE. Significant contrasts are indicated by
different letters. Asterisks indicate significant treatment effect on vascular plant biomass for Agrostis canina (***P < 0.001) and
Juncus effusus (*P = 0.01) with different letters for significant contrasts for each species. of 0.06), while this threshold occurs at only 0.2 m at
the Bargerveen in the Netherlands (E450 of 0.26) (Lim-
pens et al. 2019). As comparison, at our site, the E450
ranges between 0.1 and 0.27 (0.17 on average, Appen-
dix S1: Table S1). Alternative stable states in bogs THRESHOLDS FOR BOG RESTORATION September 2021 Article e02359; page 9 Control
ES
ES + OM
0
2
4
6
Peat moss height (cm)
0
a
b
c
0
200
400
600
800
1,000
1,200
0
2
4
6
8
10
Vascular plant biomass (g DM/m 2)
Peat moss height (cm)
ES
ES+OM
Control
R2 = 0.6, P < 0.001
C) Height vs. treatment
D) Height vs. biomass
A) ES
B) ES + OM
FIG. 5. Peat moss height in treatments and related to vascular plant biomass. Pictures of the difference in peat moss (Sphag-
umcuspidatum) height in the structures (A) without and (B) with organic material. (C) Peat moss height relative to the water level
cm) in controls, structures (ES), and structures with organic material (OM), and the (D) relationship between peat moss height
cm) and vascular plant biomass. Error bars represent SE. Significant contrasts are indicated by different letters. The regression line
was fitted using all data. B) ES + OM A) ES
B) ES + OM A) ES B) ES + OM Control
ES
ES + OM
0
2
4
6
Peat moss height (cm)
0
a
b
c
0
200
400
600
800
1,000
1,200
0
2
4
6
8
10
Vascular plant biomass (g DM/m 2)
Peat moss height (cm)
ES
ES+OM
Control
R2 = 0.6, P < 0.001
C) Height vs. treatment
D) Height vs. biomass Control
ES
ES + OM
0
2
4
6
Peat moss height (cm)
0
a
b
c
0
200
400
600
800
1,000
1,200
0
2
4
6
8
10
Vascular plant biomass (g DM/m 2)
Peat moss height (cm)
ES
ES+OM
Control
R2 = 0.6, P < 0.001
C) Height vs. treatment
D) Height vs. biomass ) Height vs. biomass D) FIG. 5. Peat moss height in treatments and related to vascular plant biomass. Pictures of the difference in peat moss (Sphag-
numcuspidatum) height in the structures (A) without and (B) with organic material. (C) Peat moss height relative to the water level
(cm) in controls, structures (ES), and structures with organic material (OM), and the (D) relationship between peat moss height
(cm) and vascular plant biomass. Error bars represent SE. Significant contrasts are indicated by different letters. The regression line
was fitted using all data. Alternative stable states in bogs In bog remnants, such as our study site, the forma-
tion of deep lakes after rewetting is unnatural. How-
ever, bog lakes can also be a natural phenomenon,
particularly in regions with an excess of moisture
(Weber 1902, Foster et al. 1988). They can deepen
because the rate of peat accumulation in the surround-
ing mire exceeds the rate of sedimentation/organic mat-
ter accumulation in the lake, and it can expand due to
wave action (Clymo 1970, Foster et al. 1988, Crushell
et al. 2009). In our model, natural colonization of open
water by peat mosses occurs at water depths <0.29 m. This is in agreement with findings from the field of
Smolders et al. (2003) and Money (1995). They indeed
indicated that at shallow water depth (<0.3 m), particu-
larly when the water is highly colored, submerged
growing peat mosses have sufficient access to light to
grow and form mats. However, this depth threshold for
natural recovery is site specific as local factors, includ-
ing suspended particles and dissolved organic com-
pounds,
control
light
extinction
and
thus
light
availability at a certain depth (Smolders et al. 2003). For instance, the 5% light threshold required for plant
growth may reach a bog-pool bottom of 5.4 m in Nor-
way (E450 of <0.05), 0.85 m in Clara bog Ireland (E450 )
Once light is sufficient, the concentration of dissolved
CO2 in the water column can become the next limiting
factor controlling submerged peat moss growth (Paffen
and Roelofs 1991, Smolders et al. 2003, Patberg et al. 2013). Moreover, once mats are afloat, they also need to
be sufficiently coherent to resist wind-driven waves. For
simplicity, however, these additional factors were not
addressed in our minimal model. Yet, even without these
potentially aggravating factors our model clearly demon-
strates the potential for occurrence of alternative stable
states as a function of water depth and light availability. This, to our knowledge, has not yet been described in lit-
erature for such systems. The occurrence of two stable
states typically results in systems that are challenging to
restore, because of strong feedbacks (Scheffer et al. 2001, van der Heide et al. 2010). Therefore, restoration
practitioners can utilize our work to either fine-tune
local hydrology to promote natural peat moss growth
and peat formation, or apply alternative approaches to
initiate a state shift from open water to a floating peat
mat. Initiating a state shift using establishment structures Initiating a state shift using establishment structures We found that the floating structures successfully ame-
liorated light and CO2 limitation, and wind-driven
waves, factors that suppresses the peat growth mosses
and may therefore induce alternative stable state dynam-
ics (Paffen and Roelofs 1991). At our site, water color
measurements indicate that 5% of the light can only
reach the bottom when water depth is below 0.29 m
(Smolders et al. 2003). In addition, CO2 concentrations
(120 10 µmol/L) lay substantially below the threshold
(>750 µmol/L) required to form a floating mat (Paffen
and Roelofs 1991). High levels of NO
3 , NHþ
4 , and
HCO
3 (>250–500 µmol/L) in the surface water have also
been found to negatively affect peat moss growth (Press
et al. 1986, Harpenslager et al. 2015, Koks et al. 2019). However, in our lake, NO
3 and HCO
3 (both <5 µmol/
L) were virtually absent, while water dissolved NHþ
4 con-
centrations (~32 µmol/L) remained below levels that
impair growth (>200 µmol/L, Appendix S1: Table S1;
Press et al. 1986, Rudolph and Voigt 1986). While not
the case at our site, guanotrophy from gulls and other
water birds is known to affect vegetation composition in
large bodies of open water in peat cuttings (Tomassen
et al. 2005). This may substantially lower water quality
and may lead to the accumulation of N and P and algal
growth (Leentvaar 1967). )
Our study shows that proper species selection in rela-
tion to the targeted environment is of utmost importance
for success as peat moss biomass consisted only of
S. cuspidatum at the end of the experimental period. The
habitat created, 3 cm below the water level, is more suit-
able to S. cuspidatum compared to S. palustre, as the lat-
ter species typically grows in drier conditions (Daniels
and Eddy 1985). The favorable conditions for S. cuspi-
datum growth, allowed this species to colonize the sub-
strate inoculated by S. palustre (Appendix S1: Fig. S4). S. palustre or other late-succession species may benefit
from drier conditions at onset (i.e., a shallower support
structure), or from a two-step restoration approach
(Money 1995). In the latter approach, faster growing
pioneer species are first facilitated by the structures
enabling the fast formation of a peat lawn and a floating
mat. Initiating a state shift using establishment structures After successful establishment and floating mat for-
mation, late-successional species can then be introduced
as a second step, demonstrated by Smolders et al. (2003)
for S. magellanicum on a well-developed carpet of
S. cuspidatum. In this way, late-successional species
profit from drier conditions, allowing them to gradually
create
a typical
raised
bog
vegetation
community
(Robroek et al. 2009). The formation of such a mat may
occur rather rapidly. For instance, Sphagnum is able to
form a dense layer consisting of living and dead material
of 18 cm thick in under 10 yr (Vroom et al. 2020). Interestingly, organic matter added to structures facili-
tated vascular plant biomass, and they in turn facilitated
peat moss biomass and vertical growth by providing
physical support for growth. Vascular plants are known
to facilitate peat moss length increment, as they provide
structure and shade (Malmer et al. 1994, Pouliot et al. 2011). The high vascular plants biomass in this treat-
ment may be explained by a higher supply of nutrients
as well as the availability of rooting substrate by the
organic material, combined with the high atmospheric
nitrogen deposition in the Netherlands (Velders et al. 2018). In addition, A. canina started growing outside the
structure
and
provided
structure
to
S. cuspidatum
(Fig. 4C). This may indicate that A. canina enables the
expansion of peat mosses outside the structure and
pushing the system into a vegetated state. Next to pro-
viding structure for peat mosses, these larger vegetation
patches can also break up waves that enables the
attached mosses to resist wave energy. Otherwise, wind-
induced waves in deep and large bog lakes might break
up the newly established mosses. This means that the
artificial structures can act as nucleus for the growth of
more extensive mats, without covering the entire lake
with structures, which may substantially lower restora-
tion costs. Based on this observation and materials used
for the current small-scale experiment, we created four
scenarios differing in initial coverage by the structure
and estimate that the cost to cover 10%, 25%, 50%, or
100% of a lake are 34.4, 86, 172, and 344 thousand $US/
ha, respectively (for details see Appendix S1: Table S2). Moreover, the peat moss to vascular plant ratio shifts RALPH J. M. TEMMINK ET AL. RALPH J. M. TEMMINK ET AL. Article e02359; page 10 from 1.05 in structures to 0.55 in structures with organic
material. Although vascular plants facilitated peat moss
height in the short term, this may well lead to competi-
tion and light limitation for the peat mosses in the long
term (Hayward and Clymo 1983, Hogg et al. 1995,
Tomassen et al. 2003a). When peat mosses are outcom-
peted by vascular plants, this might lower the potential
of the new system to store carbon, as vascular plants
degrade more readily than peat mosses (Mettrop et al. 2014). Alternative stable states in bogs Groundwater
CO2 supply
No light reaches bottom
Large waves
No organic matter (OM) accumulation
and CH4 and CO2 production
Atmospheric CO2
Light
Limiting
Sufficient
Peat-forming state
Bare state
CO2
OM
Structure
CH4
Produced OM
Light
Atmospheric
CO2
Terrestrialization
OM accumulation and
CH 4 production
Groundwater
CO2 supply
OM accumulation and
CH4 production
FIG. 6. Alternative stable states in artificial bog lakes. Natural terrestrialization (left), bare state (middle) and peat-forming veg-
etated state on floating mat mimics (right). Vegetation is able to develop as it has ample light and carbon for primary growth and is
not physically stressed by wave action (left), while these factor limit growth in a deep lake (middle). FIG. 6. Alternative stable states in artificial bog lakes. Natural terrestrialization (left), bare state (middle) and peat-forming veg-
etated state on floating mat mimics (right). Vegetation is able to develop as it has ample light and carbon for primary growth and is
not physically stressed by wave action (left), while these factor limit growth in a deep lake (middle). Ecological Applications
Vol. 31, No. 6 Ecological Applications
Vol. 31, No. 6 Conclusion and implications for restoration In this paper, we demonstrate the potential for alter-
native stable states, open water vs. floating peat mat, in
artificial lakes in bog remnants, using a minimal model,
and confirm this in our field experiment. Moreover, we
show that restoration by inducing a state shift from open
water to floating mats using biodegradable establish-
ment structures can be achieved without the need to
introduce
unsustainable
peat. By
introducing
peat
mosses close to the water surface in the structures light
and carbon limitation, and wave stress were ameliorated,
thereby overcoming establishment thresholds. The shift
from open water to peat-forming vegetation may eventu-
ally result in carbon storage, as bog vegetation seques-
ters
and
stores
more
carbon
than
open
water
(Couwenberg et al. 2011). Such as shift may aid in cli-
mate mitigation when it takes places on a sufficiently
large scale in many bog remnants worldwide. Moreover,
the model results can also be used to fine-tune local THRESHOLDS FOR BOG RESTORATION Article e02359; page 11 September 2021 Daniels, R. E., and E. Eddy. 1985. Handbook of European
Sphagna. Institute of Terrestrial Ecology, Huntingdon, UK. hydrology to optimize peat moss growth. We further
observed that the structures act as a nucleus from where
terrestrialization can occur (Fig. 4C). It is important to
note, that for widespread upscaling a sustainable donor
supply could be an issue, because peat mosses are rela-
tively scarce in the Netherlands and Western Europe. Mosses grown at Sphagnum farms can be a sustainable
alternative to mosses collected in natural peatlands
(Gaudig et al. 2017, Temmink et al. 2017, Vroom et al. 2020). In addition, natural succession of a floating peat
moss carpet towards an ombrotrophic bog vegetation
can be slow process (decades to a century; Lindsay and
Clough 2016), and the introduction of later-successional
peat mosses on top of floating mats may speed this up. Lastly, the results we show here may well be extended to
the restoration of other peat ecosystems that are charac-
terized by a lack of vegetation development due to unfa-
vorable environmental conditions. Clear examples are
fens where the colonization of open water by terrestrial
species is crucial to form floating mats (Sarneel 2010)
and facilitate brown moss (Scorpidium sp.) growth estab-
lishment (Lamers et al. 2002, Lamers et al. 2015). Conclusion and implications for restoration Restoration techniques such as the one presented here
may become important to overcome establishment
thresholds and achieve greater restoration success in
degraded peatlands. Foster, D. R., H. E. Wright, M. Thelaus, and G. A. King. 1988. Bog development and landform dynamics in Central Sweden
and South-Eastern Labrador, Canada. Journal of Ecology
76:1164–1185. Frahm, J.-P., and W. Frey. 2003. Moosflora. Ulmer, Stuttgart,
Germany. Gahlert, F., A. Prager, A. S. Quesada, S. Wichmann, and H. Joosten. 2010. Torfmooskultivierung auf schwimmf¨ahigen
Vegetationstr¨agern
f¨ur ein
nachhaltiges
und umweltfre-
undlichen Torfsubstitut im Erwerbsgartenbau - MOOS-
FARM. Universit¨at Greifswald, Greifswald, Germany. FARM. Universit¨at Greifswald, Greifswald, Germany Gaudig, G., M. Krebs, A. Prager, S. Wichmann, M. Barney, S. Caporn, M. Emmel, C. Fritz, M. Graf, and A. Grobe. 2017. Sphagnum farming from species selection to the production
of growing media. Mires and Peat 20:1–30. G¨unther, A., A. Barthelmes, V. Huth, H. Joosten, G. Jurasinski,
F. Koebsch, and J. Couwenberg. 2020. Prompt rewetting of
drained peatlands reduces climate warming despite methane
emissions. Nature Communications 11:1644. Haapalehto, T., J. S. Kotiaho, R. Matilainen, and T. Tahvanainen. 2014. The effects of long-term drainage and subsequent restora-
tion on water table level and pore water chemistry in boreal
peatlands. Journal of Hydrology 519:1493–1505. Harpenslager, S. F., E. van den Elzen, M. A. R. Kox, A. J. P. Smolders, K. F. Ettwig, and L. P. M. Lamers. 2015. Rewetting
former agricultural peatlands: topsoil removal as a prerequi-
site to avoid strong nutrient and greenhouse gas emissions. Ecological Engineering 84:159–168. Hayward, P. M., and R. S. Clymo. 1983. The growth of Sphag-
num: experiments on, and simulation of, some effects of light
flux and water-table depth. Journal of Ecology 71:845–863. ACKNOWLEDGMENTS The authors would like to thank all volunteers who joined in
setting up and monitoring the experiments, and Roy Peters,
Germa Verheggen and Sebastian Krosse for their help with
chemical analyses. We thank Natuurmonumenten for site
access. R. J. M. Temmink, G. S. Fivash, K. Didderen, and W. Lengkeek were funded by NWO/TTW-OTP grant 14424, in col-
laboration with private and public partners: Natuurmonu-
menten,
STOWA,
Rijkswaterstaat,
Van
Oord,
Bureau
Waardenburg, Enexio, and Rodenburg Biopolymers. T. van der
Heide was funded by NWO/TTW-Vidi grant 16588. The authors would like to thank all volunteers who joined in
setting up and monitoring the experiments, and Roy Peters,
Germa Verheggen and Sebastian Krosse for their help with
chemical analyses. We thank Natuurmonumenten for site
access. R. J. M. Temmink, G. S. Fivash, K. Didderen, and W. Lengkeek were funded by NWO/TTW-OTP grant 14424, in col-
laboration with private and public partners: Natuurmonu-
menten,
STOWA,
Rijkswaterstaat,
Van
Oord,
Bureau
Waardenburg, Enexio, and Rodenburg Biopolymers. T. van der
Heide was funded by NWO/TTW-Vidi grant 16588. Hogg, P., P. Squires, and A. H. Fitter. 1995. Acidification, nitro-
gen deposition and rapid vegetational change in a small valley
mire in Yorkshire. Biological Conservation 71:143–153. Joosten, H., and D. Clarke. 2002. Wise use of mires and peat-
lands—background and principles including a framework for
decision-making. International Mire Conservation Group
and International Peat Society, Saarij¨arven Offset Oy, Saar-
ij¨arvi, Finland. Kirk, J. T. O. 1994. Light and photosynthesis in aquatic ecosys-
tems. Cambridge University Press, Cambridge, UK. Koks, A., G. van Dijk, A. Smolders, L. Lamers, and C. Fritz. 2019. The effects of alkalinity and cations on the vitality of
Sphagnum palustre L. Mires and Peat 24:1–14. LITERATURE CITED The hydrology of bog
ecosystems. In Guidelines for management. Staatsbosbeheer,
Utrecht, the Netherlands. Parry, L. E., J. Holden, and P. J. Chapman. 2014. Restoration of
blanket peatlands. Journal of Environmental Management
133:193–205. Swindles, G. T., et al. 2019. Widespread drying of European
peatlands in recent centuries. Nature Geoscience 12:922–928. Temmink, R. J. M., et al. 2020. Mimicry of emergent traits
amplifies coastal restoration success. Nature Communica-
tions 11:1–9. Patberg, W., G. J. Baaijens, A. J. P. Smolders, A. P. Grootjans,
and J. T. M. Elzenga. 2013. The importance of groundwater
carbon dioxide in the restoration of small Sphagnum bogs. Preslia 85:389–403. Temmink, R. J. M., P. M. J. M. Cruijsen, A. J. P. Smolders, T. J. Bouma, G. S. Fivash, W. Lengkeek, K. Didderen, L. P. M. Lamers, and T. van der Heide. 2021. Data from: Overcoming
establishment thresholds for peat mosses in human-made bog
pools. DANS EASY. https://doi.org/10.17026/dans-xg4-9mz7 Pouliot, R., L. Rochefort, E. Karofeld, and C. Mercier. 2011. Initiation of Sphagnum moss hummocks in bogs and the
presence of vascular plants: Is there a link? Acta Oecologica
37:346–354. Press, M. C., S. J. Woodin, and J. A. Lee. 1986. The potential
importance of an increased atmospheric nitrogen supply to
the growth of ombrotrophic Sphagnum species. New Phytolo-
gist 103:45–55. Temmink, R. J. M., C. Fritz, G. van Dijk, G. Hensgens, L. P. M. Lamers, M. Krebs, G. Gaudig, and H. Joosten. 2017. Sphagnum farming in a eutrophic world: The importance of
optimal
nutrient
stoichiometry. Ecological
Engineering
98:196–205. Quinty, F., and L. Rochefort. 2000. Bare peat substrate instabil-
ity in peatland restoration: problems and solutions. Sustain-
ing our Peatlands 2:6–12. Titus, J. E., and D. J. Wagner. 1984. Carbon balance for two
Sphagnum mosses: water balance resolves a physiological
paradox. Ecology 65:1765–1774. R Core Team. 2020. R: a language and environment for statisti-
cal computing. R Foundation for Statistical Computing,
Vienna, Austria. Tomassen, H. B. M., A. J. P. Smolders, L. P. M. Lamers, and J. G. M. Roelofs. 2003a. Stimulated growth of Betula pubescens
and Molinia caerulea on ombrotrophic bogs: role of high
levels of atmospheric nitrogen deposition. Journal of Ecology
91:357–370. Robroek, B. J., J. van Ruijven, M. G. Schouten, A. Breeuwer, P. H. Crushell, F. Berendse, and J. Limpens. 2009. Sphagnum
re-introduction in degraded peatlands: The effects of aggrega-
tion, species identity and water table. Basic and Applied Ecol-
ogy 10:697–706. Tomassen, H. B. M., A. J. LITERATURE CITED 925: A review of the influence of
ombrotrophic peat depth on the successful restoration of bog
habitat. Edinburgh Scottish Natural Heritage, Edinburgh,
UK. Scheffer, M., S. Carpenter, J. A. Foley, C. Folke, and B. Walker. 2001. Catastrophic shifts in ecosystems. Nature 413:591–596. habitat. Edinburgh Scottish Natural Heritage, Edinburgh,
UK. Schothorst, C. J. 1977. Subsidence of low moor peat soils in the
western Netherlands. Geoderma 17:265–291. Malmer, N., B. M. Svensson, and B. Wall´en. 1994. Interactions
between Sphagnum mosses and field layer vascular plants in
the development of peat-forming systems. Folia Geobotanica
et Phytotaxonomica 29:483–496. Schumann, M., and H. Joosten. 2008. Global peatland restora-
tion manual. Institute of Botany and Landscape Ecology,
Greifswald University, Greifswald, Germany. Mettrop, I. S., C. Cusell, A. M. Kooijman, and L. P. M. Lamers. 2014. Nutrient and carbon dynamics in peat from rich fens
and Sphagnum-fens during different gradations of drought. Soil Biology and Biochemistry 68:317–328. Smolders, A. J. P., H. B. M. Tomassen, L. P. M. Lamers, B. P. Lomans, and J. G. M. Roelofs. 2002. Peat bog restoration by
floating raft formation: the effects of groundwater and peat
quality. Journal of Applied Ecology 39:391–401. Money, R. P. 1995. Re-establishment of a Sphagnum-dominated
flora on cut-over lowland raised bogs. Pages 405–422 in B. D. Wheeler, S. C. Shaw, W. J. Fojt, and R. A. Robertson, editors. Restoration of temperate wetlands. John Wiley & Sons,
Chichester, UK. Smolders, A. J. P., H. B. M. Tomassen, M. van Mullekom, L. P. M. Lamers, and J. G. M. Roelofs. 2003. Mechanisms involved
in the re-establishment of Sphagnum-dominated vegetation in
rewetted bog remnants. Wetlands Ecology and Management
11:403–418. Money, R. P., and B. D. Wheeler. 1999. Some critical questions
concerning the restorability of damaged raised bogs. Applied
Vegetation Science 2:107–116. Soudzilovskaia, N. A., J. H. C. Cornelissen, H. J. During, R. S. P. van Logtestijn, S. I. Lang, and R. Aerts. 2010. Similar
cation exchange capacities among bryophyte species refute a
presumed mechanism of peatland acidification. Ecology
91:2716–2726. Nichols, J. E., and D. M. Peteet. 2019. Rapid expansion of
northern peatlands and doubled estimate of carbon storage. Nature Geoscience 12:917–921. Stachowicz, J. J. 2001. Mutualism, facilitation, and the structure
of ecological communities. BioScience 51:235–246. Paffen, B. G. P., and J. G. M. Roelofs. 1991. Impact of carbon
dioxide and ammonium on the growth of submerged Sphag-
num cuspidatum. Aquatic Botany 40:61–71. Streefkerk, J., and W. Casparie. 1989. LITERATURE CITED Lamers, L. P. M., R. Bobbink, and J. G. M. Roelofs. 2000. Nat-
ural nitrogen filter fails in polluted raised bogs. Global
Change Biology 6:583–586. Altenburg, W., W. Bijkerk, R. Douwes, and N. Straathof. 2017. Fall and rise of the Fochterlo¨erveen: results of 30 years of
conservation efforts. De Levende Natuur 118:79–84. Lamers, L. P. M., A. J. P. Smolders, and J. G. M. Roelofs. 2002. The restoration of fens in the Netherlands. Hydrobiologia
478:107–130. Boatman, D. J. 1977. Observations on the growth of Sphagnum
cuspidatum in a bog pool on the Silver Flowe National Nat-
ure Reserve. Journal of Ecology 65:119–126. Lamers, L. P. M., et al. 2015. Ecological restoration of rich fens
in Europe and North America: from trial and error to an
evidence-based approach. Biological Reviews 90:182–203. Clymo, R. S. 1970. The growth of Sphagnum: methods of mea-
surement. Journal of Ecology 58:13–49. Couwenberg, J., and H. Joosten. 2005. Self-organization in
raised bog patterning: The origin of microtope zonation and
mesotope diversity. Journal of Ecology 93:1238–1248. Leentvaar, P. 1967. Observations in guanotrophic environments. Hydrobiologia 29:441–489. Lenth, R., and M. R. Lenth. 2018. Package ‘lsmeans’. Ameri-
can Statistician 34:216–221. Couwenberg, J., et al. 2011. Assessing greenhouse gas emissions
from peatlands using vegetation as a proxy. Hydrobiologia
674:67–89. Limpens, J., F. Berendse, C. Blodau, J. G. Canadell, C. Free-
man, J. Holden, N. Roulet, H. Rydin, and G. Schaepman-
Strub. 2008. Peatlands and the carbon cycle: from local pro-
cesses to global implications—a synthesis. Biogeosciences
5:1475–1491. Crushell, P. H., A. J. P. Smolders, M. G. C. Schouten, J. G. M. Roelofs, and G. van Wirdum. 2009. The origin and develop-
ment of a minerotrophic soak on an Irish raised bog: an inter-
pretation of depth profiles of hydrochemistry and peat
chemistry. Holocene 19:921–935. Limpens, J., H. B. M. Tomassen, and A. J. P. Smolders. 2019. Sturende factoren voor hoogveengroei op standplaatsschaal Ecological Applications
Vol. 31, No. 6 Ecological Applications
Vol. 31, No. 6 RALPH J. M. TEMMINK ET AL. Article e02359; page 12 Scharlemann, J. P. W., E. V. J. Tanner, R. Hiederer, and V. Kapos. 2014. Global soil carbon: understanding and manag-
ing the largest terrestrial carbon pool. Carbon Management
5:81–91. (Driving factors for raised bog growth on habitat scale). Uit-
geverij Noordboek, Gorredijk, the Netherlands. Lindsay, R., and J. Clough. 2016. Scottish Natural Heritage
Commissioned Report No. LITERATURE CITED P. Smolders, J. M. van Herk, L. P. M. Lamers, and J. G. M. Roelofs. 2003b. Restoration of cut-over
bogs by floating raft formation: an experimental feasibility
study. Applied Vegetation Science 6:141–152. Rochefort, L., F. Quinty, S. Campeau, K. Johnson, and T. Mal-
terer. 2003. North American approach to the restoration of
Sphagnum dominated peatlands. Wetlands Ecology and Man-
agement 11:3–20. Tomassen, H. B. M., A. J. P. Smolders, L. P. M. Lamers, and J. G. M. Roelofs. 2004. Development of floating rafts after the
rewetting of cut-over bogs: the importance of peat quality. Biogeochemistry 71:69–87. Royal Dutch Meteorological Institute. 2019. Daggegevens van
het weer in Nederland, Eelde: September 2009—September
2019. Tomassen, H. B., A. J. Smolders, L. P. M. Lamers, and J. G. Roelofs. 2005. How bird droppings can affect the vegetation
composition of ombrotrophic bogs. Canadian Journal of Bot-
any 83:1046–1056. Rudolph, H., and J. U. Voigt. 1986. Effects of NH+ 4-N and
NO+ 3-N on growth and metabolism of Sphagnum magellan-
icum. Physiologia Plantarum 66:339–343. Sarneel, J. M. 2010. Colonisation processes in riparian fen vege-
tation. Utrecht University, Utrecht, the Netherlands. van Bergen, T. J. H. M., et al. 2020. Self-facilitation and nega-
tive species interactions could drive microscale vegetation THRESHOLDS FOR BOG RESTORATION September 2021 Article e02359; page 13 mosaic in a floating fen. Journal of Vegetation Science
31:343–354. Verhoeven, J. T. A., and T. L. Setter. 2010. Agricultural use of
wetlands: opportunities and limitations. Annals of Botany
105:155–163. Van Breemen, N. 1995. How Sphagnum bogs down other
plants. Trends in Ecology & Evolution 10:270–275. Vroom, R. J., R. J. Temmink, G. van Dijk, H. Joosten, L. P. Lamers, A. J. Smolders, M. Krebs, G. Gaudig, and C. Fritz. 2020. Nutrient dynamics of sphagnum farming on rewetted
bog grassland in NW Germany. Science of the Total Environ-
ment 726:138470. van der Heide, T., E. H. Nes, M. M. Katwijk, M. Scheffer, A. Jan Hendriks, and A. J. P. Smolders. 2010. Alternative stable
states
driven
by density-dependent
toxicity. Ecosystems
13:841–850. van der Heide, T., E. H. van Nes, G. W. Geerling, A. J. P. Smold-
ers, T. J. Bouma, and M. M. van Katwijk. 2007. Positive feed-
backs in seagrass ecosystems: implications for success in
conservation and restoration. Ecosystems 10:1311–1322. Weber, C. A. 1902. Ueber die Vegetation und Entstehung des
Hochmoors von Augstumal im Memeldelta. Paul Parey, Ber-
lin, Germany. Wheeler, B., and S. Shaw. 1995. LITERATURE CITED Restoration of damaged peat-
lands: with particular reference to lowland raised bogs
affected by peat extraction. Department of the Environment,
HMSO, London, UK. van Nes, E. 2017. Grind for Matlab. http://www.sparcs-center. org/grind Velders, G. J. M., J. M. M. Aben, G. P. Geilenkirchen, H. A. den Hollander, L. Nguyen, E. van der Swaluw, W. J. de Vries,
and R. J. Wichink Kruit. 2018. Grootschalige concentratie-
en depositiekaarten Nederland: Rapportage 2018. Rijksinsti-
tuut voor Volksgezondheid en Milieu, Bilthoven, the Nether-
lands. Yu, Z., J. Loisel, D. P. Brosseau, D. W. Beilman, and S. J. Hunt. 2010. Global peatland dynamics since the Last Glacial Maxi-
mum. Geophysical Research Letters 37:1–5. SUPPORTING INFORMATION Additional supporting information may be found online at: http://onlinelibrary.wiley.com/doi/10.1002/eap.2359/full OPEN RESEARCH Data (Temmink et al. 2021) that support the main findings of this study are available via Data Archiving and Networked Services
(DANS) EASYat: https://doi.org/10.17026/dans-xg4-9mz7.
|
https://openalex.org/W3116885256
|
https://digitalcommons.mtu.edu/cgi/viewcontent.cgi?article=34279&context=michigantech-p
|
English
| null |
Frequency domain measurements of melt pool recoil force using modal analysis
|
Scientific reports
| 2,021
|
cc-by
| 9,256
|
Michigan Technological University
Michigan Technological University
Digital Commons @ Michigan Tech
Digital Commons @ Michigan Tech
Michigan Tech Publications
5-26-2021
Frequency domain measurements of melt pool recoil force using
Frequency domain measurements of melt pool recoil force using
modal analysis
modal analysis
Tristan Cullom
Missouri University of Science and Technology
Cody Lough
Missouri University of Science and Technology
Nicholas Altese
Missouri University of Science and Technology
Douglas Bristow
Missouri University of Science and Technology
Robert Landers
Missouri University of Science and Technology
See next page for additional authors
Follow this and additional works at: https://digitalcommons.mtu.edu/michigantech-p
Part of the Manufacturing Commons
Recommended Citation
Recommended Citation
Cullom, T., Lough, C., Altese, N., Bristow, D., Landers, R., Brown, B., Hartwig, T., Barnard, A., Blough, J.,
Johnson, K., & Kinzel, E. (2021). Frequency domain measurements of melt pool recoil force using modal
analysis. Scientific Reports, 11(1). http://doi.org/10.1038/s41598-021-90423-z
Retrieved from: https://digitalcommons.mtu.edu/michigantech-p/14977 Michigan Technological University
Michigan Technological University
Digital Commons @ Michigan Tech
Digital Commons @ Michigan Tech See next page for additional authors Follow this and additional works at: https://digitalcommons.mtu.edu/michigantech-p Part of the Manufacturing Commons Part of the Manufacturing Commons Follow this and additional works at: https://digitalcommons.mtu.edu/michigantech-p Recommended Citation
Recommended Citation Cullom, T., Lough, C., Altese, N., Bristow, D., Landers, R., Brown, B., Hartwig, T., Barnard, A., Blough, J.,
Johnson, K., & Kinzel, E. (2021). Frequency domain measurements of melt pool recoil force using modal
analysis. Scientific Reports, 11(1). http://doi.org/10.1038/s41598-021-90423-z
Retrieved from: https://digitalcommons.mtu.edu/michigantech-p/14977 Follow this and additional works at: https://digitalcommons.mtu.edu/michigantech-p Part of the Manufacturing Commons Part of the Manufacturing Commons Tristan Cullom, Cody Lough, Nicholas Altese, Douglas Bristow, Robert Landers, Ben Brown, Troy Hartwig,
Andrew Barnard, Jason Blough, Kevin Johnson, and Edward Kinzel This article is available at Digital Commons @ Michigan Tech: https://digitalcommons.mtu.edu/michigantech-p/
14977 This article is available at Digital Commons @ Michigan Tech: https://digitalcommons.mtu.edu/michigantech-p/
14977 Tristan Cullom1, Cody Lough1, Nicholas Altese1, Douglas Bristow1, Robert Landers1,
Ben Brown2, Troy Hartwig2, Andrew Barnard3, Jason Blough3, Kevin Johnson3 &
Edward Kinzel4* Recoil pressure is a critical factor affecting the melt pool dynamics during Laser Powder Bed Fusion
(LPBF) processes. Recoil pressure depresses the melt pool. When the recoil pressure is low, thermal
conduction and capillary forces may be inadequate to provide proper fusion between layers. However,
excessive recoil pressure can produce a keyhole inside the melt pool, which is associated with gas
porosity. Direct recoil pressure measurements are challenging because it is localized over an area
proportionate to the laser spot size producing a force in the mN range. This paper reports a vibration-
based approach to quantify the recoil force exerted on a part in a commercial LPBF machine. The
measured recoil force is consistent with estimates from high speed synchrotron imaging of entrained
particles, and the results show that the recoil force scales with applied laser power and is inversely
related to the laser scan speed. These results facilitate further studies of melt pool dynamics and have
the potential to aid process development for new materials. Understanding the complex melt pool dynamics in laser powder bed fusion (LPBF) processes is critical to
maintaining quality of printed parts. Tight quality control is necessary as parts created by LPBF are used in
demanding biomedical, aerospace, and defense applications. In the LPBF process a focused laser spot follows
through a computer determined path over a thin powder layer. The powder is heated by the absorbed laser irra-
diation to the point that it melts and fuses with the underlying part. Variations in the melt pool dynamics lead
to changes in microstructure, notably porosity. Specifically, insufficient laser fluence produces a lack of fusion
between layers while excessive power leads to gas entrapment as the melt pool solidifies. The influence of the
recoil pressure can be seen with the different melt pool modes. If the recoil pressure is too low, heat transfer in
the melt pool operates in the conduction mode1, potentially leading to poor fusion between the molten material
and the previous layer and result in brittle parts. Conversely, if the recoil pressure is too high, convection in the
melt pool is the dominant heat transfer mode, depressing the melt pool into multiple previous layers and can
potentially create keyhole porosity due to increased laser absorptivity and making the melt pool less stable2,3. This melt pool mode is referred to as the keyhole mode4. Authors
Authors www.nature.com/scientificreports www.nature.com/scientificreports Frequency domain measurements
of melt pool recoil force using
modal analysis
OPEN Tristan Cullom1, Cody Lough1, Nicholas Altese1, Douglas Bristow1, Robert Landers1,
Ben Brown2, Troy Hartwig2, Andrew Barnard3, Jason Blough3, Kevin Johnson3 &
Edward Kinzel4* Tristan Cullom1, Cody Lough1, Nicholas Altese1, Douglas Bristow1, Robert Landers1,
Ben Brown2, Troy Hartwig2, Andrew Barnard3, Jason Blough3, Kevin Johnson3 &
Edward Kinzel4* Tristan Cullom1, Cody Lough1, Nicholas Altese1, Douglas Bristow1, Robert Landers1,
Ben Brown2, Troy Hartwig2, Andrew Barnard3, Jason Blough3, Kevin Johnson3 &
Edward Kinzel4* Illustrations of (a) laser excitation of tuning fork with accelerometer (Accel) mounted on tuning fork
prong, (b) laser interaction with powder on top of tuning fork during laser excitation, and (c) example FRFs and
mode shapes. The effects of the recoil pressure in depressing the melt pool has been studied numerically7,14,15. While mod-
eling recoil pressure is computationally expensive, an estimate for the recoil pressure can be made using the
Clausius-Clapeyron model and via the simulation results in Khairallah et al.16. The recoil pressure was estimated
to be 86 kPa for a 316L stainless steel simulation with a laser power of 200 W and a scan speed of 1.5 m/s. Simula-
tions have also shown that as material is vaporized, changes in the thermal and fluid transport within the melt
pool lead to surface defects known as humps10,17. This is an effect of the backflow generated by the recoil pressure
as the exposed area of the melt pool is displaced away from the center. In addition, the recoil pressure driven
melt pool depression has been experimentally shown to be correlated with the formation of spatter18–25, as well
as influence the magnitude of the oscillations that occur within the melt pool26.h l
g
p
Recently, there have been a few attempts to measure recoil pressure using particle tracking techniques. This
requires in-situ high-speed imaging of the melt pool. Zhao et al. used a custom built 2D setup illuminated with
synchrotron radiation to estimate an average pressure above the melt pool of 60 kPa for Ti-6Al-4 V powder
melted with a laser power of 210 W and a scan speed 0.5 m/s27. Yin et al. calculated a vapor pressure of 49 kPa
for Inconel powder processed with a laser power of 1150 W, a scan speed of 1 m/s, and a spot size of 159 µm
by observing spatter tracks using high-speed visible camera imaging28. These studies pose instrumentation
challenges and require the assumption that the particles are moving parallel to the imaging plane. In addition,
significant complications arise with the presence of the gas that flows over the build plate (i.e., shielding gas) to
create an inert atmosphere as it substantially modifies particle velocities. Shielding gas was present during the
experiment in the Inconel study28 to prevent melt pool oxidation while the Ti-6Al-4V study was performed in
vacuum27. Tristan Cullom1, Cody Lough1, Nicholas Altese1, Douglas Bristow1, Robert Landers1,
Ben Brown2, Troy Hartwig2, Andrew Barnard3, Jason Blough3, Kevin Johnson3 &
Edward Kinzel4* Since part quality is primarily driven by these defects,
it is important to understand recoil pressure to operate the LPBF process between the conduction and keyhole
modes, thus, minimizing the potential for these defects. g
p
For most commercial processes, evaporation occurs at the melt pool surface, which produces a localized
recoil pressure. This pressure depresses the melt pool5 which drives the melt pool deeper and enhances the heat
transfer allowing for adequate layer-to-layer fusion.h g
y
y
The dynamics of the melt pool are driven by Marangoni6–8, capillary9,10 and buoyancy11 forces in addition to
the recoil pressure. These forces all significantly influence the shape of the melt pool7,8 as well as its stability6 and
overall fluid transport. However, the recoil force magnitude is significantly greater than the other forces acting
on the melt pool once it starts to vaporize. For example, using the definitions of these forces described in Ref. 12 and typical processing parameters for 304L stainless steel13, the recoil force is at least an order of magnitude
larger than the net Marangoni force, and more than eight orders of magnitude larger than the capillary and
buoyancy forces. 1Department of Mechanical and Aerospace Engineering, Missouri University of Science and Technology, Rolla,
MO 65409, USA. 2Kansas City National Security Campus, Kansas City, MO 64147, USA. 3Department of Mechanical
Engineering‑Engineering Mechanics, Michigan Technological University, Houghton, MI 49931, USA. 4Department
of Aerospace and Mechanical Engineering, University of Notre Dame, Notre Dame, IN 46556, USA. *email:
ekinzel@nd.edu | https://doi.org/10.1038/s41598-021-90423-z Scientific Reports | (2021) 11:10959 www.nature.com/scientificreports/ V
Accel
Laser
Beam
v
L
F
60°
Build Plate
V
Vapor
Plume
Modulated
Laser
Argon
Gas
Flow
T > TEvap
Powder
Recoil Pressure
L [mm] 24 30 36
e
d
utin
g
a
M
F
R
F
Frequency
(c)
(a)
(b)
Figure 1. Illustrations of (a) laser excitation of tuning fork with accelerometer (Accel) mounted on tuning fork
prong, (b) laser interaction with powder on top of tuning fork during laser excitation, and (c) example FRFs and
mode shapes. V
Vapor
Plume
Modulated
Laser
Argon
Gas
Flow
T > TEvap
Powder
Recoil Pressure
(b) V
Accel
Laser
Beam
v
L
F
60°
Build Plate
(a) L [mm] 24 30 36
e
d
utin
g
a
M
F
R
F
Frequency
(c) (c) (a) (b) Figure 1. Tristan Cullom1, Cody Lough1, Nicholas Altese1, Douglas Bristow1, Robert Landers1,
Ben Brown2, Troy Hartwig2, Andrew Barnard3, Jason Blough3, Kevin Johnson3 &
Edward Kinzel4* A possible alternative to measuring the laser spatter trajectories is to directly measure the reaction force
generated by the recoil pressure. Both experimental and numerical studies give an expectation for the recoil
pressure on the order of 50–90 kPa. This is equivalent to a recoil force acting on the melt pool in the sub mN
range. Measuring this force in the time domain is difficult given the noise inherent in LPBF environments (e.g.,
shielding gas, chiller). However, if the experiments can be conducted in the frequency domain, spectral filter-
ing techniques can be employed to significantly improve the Signal to Noise Ratio (SNR). This paper presents a
study in a commercial LPBF system using an accelerometer to measure part vibration and quantify recoil force
in the frequency domain. Modal analysis is used in this study by exciting resonant parts with a laser to calculate
the recoil force. This approach is used to measure the recoil force for various ranges of process parameters to
evaluate their relationship with recoil pressure. Finally, the dependence of the microstructure and melt-pool
depth on recoil pressure for typical LPBF of SS304L parts is presented. Experimental approachh The tuning forks that were used in experiments are tabulated in Table 1
in “Methods”. Figure 2. Tuning fork build with arrows indicating beginning and ending tuning forks for each row, where the
tuning fork numbers linearly increment. The tuning forks that were used in experiments are tabulated in Table 1
in “Methods”. Table 1. Tuning forks used in experiments, resonant frequencies, and prong lengths. Tuning fork
Resonant frequency (kHz)
L (mm)
Mode number
15
10.05
35
7
16
10.03
36
8
17
9.930
37
12
20
9.800
41
12
27
6.300
49
11
29
6.300
51
10
30
6.175
52
10
31
11.02
53
14
32
5.200
54
11
33
5.490
56
9
34
6.256
57
12
39
2.050
62
7 Table 1. Tuning forks used in experiments, resonant frequencies, and prong lengths. Tuning fork
Resonant frequency (kHz)
L (mm)
Mode number
15
10.05
35
7
16
10.03
36
8
17
9.930
37
12
20
9.800
41
12
27
6.300
49
11
29
6.300
51
10
30
6.175
52
10
31
11.02
53
14
32
5.200
54
11
33
5.490
56
9
34
6.256
57
12
39
2.050
62
7 Table 1. Tuning forks used in experiments, resonant frequencies, and prong lengths. Table 1. Tuning forks used in experiments, resonant frequencies, and prong lengths. produces a strong acceleration response when the part is forced at resonance. This requires both the force and
the accelerometer to be coupled to a common resonant mode. To accomplish this, the tuning forks are printed
at 60° relative to horizontal so the laser would excite the bending modes and the parts could be printed without
support structures. The force/acceleration coupling of the tuning forks was simulated using ANSYS as shown
Fig. 1c and is quantified by the Frequency Response Function (FRF) (1)
FRF
f
= a
f
F
f (1) where f is the frequency, a is the part acceleration, and F is the force applied to the part. The tuning fork modes
n Fig. 1c were using ANSYS. The mass of the accelerometer was not included in the ANSYS simulations. Experimentally, the force acting on the part consists of all of the melt pool forces; including capillary (i.e.,
surface tension), thermo-capillary (i.e., Marangoni), and recoil5,31,32. However, as discussed earlier, the recoil
force magnitude is expected to dominate those of the other forces. Experimental approachh p
pp
The experiments in this paper are conducted using a commercial LPBF machine (AM250, Renishaw). This
machine stabilizes the melt pool using an Acousto-Optic Modulator (AOM) to pulse the laser. During process-
ing, the laser is pulsed for duration τpulse ~ 75 µs while being held stationary at a point. At the conclusion of the
pulse, galvo scanners move the beam along the scan path by a point distance, PD29,30. The point distance and
pulse duration can both be adjusted to provide an estimation of the scanning velocity, V = PD/τpulse. Adjusting
the pulse length corresponds to specifying the pulse repetition frequency (PRF) of the laser modulation, 1/τpulse,
and can be adjusted from 1 to 25 kHz with τpulse adjustable in 10 µs increments. p
Figure 1 shows a part being excited by the laser. In the LPBF process, the laser energy is absorbed by the
layer of unfused powder on top of the part. The powder is melted and then vaporized. The metallic vapor exerts
a recoil pressure on the melt pool, deforming it and producing a net force normal to the surface. In addition to
the vaporized metal, particles can be entrained by the local pressure field and ejected away from the melt pool. This is depicted in Fig. 1b.h h
The experiments conducted in this paper use 304L stainless steel powder. Tuning forks are printed directly on
the build plate. After printing, the unfused powder is removed from the chamber and an accelerometer is fixed
to the tuning forks as illustrated in Fig. 1a and shown in Fig. 2. Experiments show that the powder ordinarily
surrounding the part increases structural damping by a factor of 14, which would significantly lower the sensitiv-
ity of the experiment. After fixing the accelerometer to the tuning fork with superglue, a new 50 µm thick layer
of powder is placed on the top surface of the part. The force generated by the laser interaction at the melt pool Scientific Reports | (2021) 11:10959 | https://doi.org/10.1038/s41598-021-90423-z www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 2. Tuning fork build with arrows indicating beginning and ending tuning forks for each row, where the
tuning fork numbers linearly increment. The tuning forks that were used in experiments are tabulated in Table 1
in “Methods”. Figure 2. Tuning fork build with arrows indicating beginning and ending tuning forks for each row, where the
tuning fork numbers linearly increment. Experimental approachh The FRF of each tuning fork was measured
experimentally with a modal impact test by striking each prong with an impact hammer and measuring the cor-
responding acceleration. The impact hammer strikes the same surface that is irradiated by the laser and as normal
to the surface as possible (see Fig. 1a). This test occurs with the build plate inside the LPBF chamber. Powder is
than added to the top surface of part, the chamber is evacuated and back filled with Argon, and the acceleration
measured while the part is scanned by the laser with the accelerometer in the same position. After the test, the
FRF was measured again to ensure that any changes resulting from the laser interaction (e.g., additional mass
fused to the tip of the tuning fork) are negligible. Multiple tuning forks are fabricated in order to have resonant frequencies spanning a wide range of laser PRFs. Because the FRF of the individual parts are known (after modal impact hammer testing), the PRF used to excite
an individual tuning fork is adjusted to match the tuning fork’s resonant frequency. Figure 2 shows a photograph
of 40 tuning forks, 12 of which used in the experiments. During the experiment, the accelerometer remained in
the same location during the laser excitation and the impact test afterwards. Since the location of the accelerom-
eter did not change during the test, the relative effect of the mass loading is effectively cancelled out. The tuning
forks are printed using a laser power P = 175 W, V = 0.8 m/s, exposure time of 75 µs, hatch spacing of 85 µm,
and PD = 60 µm. The resonant frequencies of the 12 turning forks are given in Table 1 of the Methods section. https://doi.org/10.1038/s41598-021-90423-z Scientific Reports | (2021) 11:10959 | www.nature.com/scientificreports/ τPulse
τDelay
τDelay + Corner
1
2
3
4
5
6
6
8
10
12
14
16
18
20
22
0.00
0.25
0.50
0.75
2f0
f0
9.5
10.0
10.5
0.00
0.25
0.50
0.75
(b)
τCorner = 0 μs τCorner = 10 μs
e
d
u
tin
g
a
M
f [kHz]
0
155
310
465
620
0.00
0.25
0.50
0.75
1.00
1.25
1.50
(a)
]
u
. a
[la
n
gi
S
e
d
oid
o
t
o
h
P
t [μs]
t = 310
Figure 3. Experimental approachh (a) Normalized experimental photodiode waveform with inset showing raster scan path for
τPulse = 100 µs and (b) corresponding signal in the frequency domain with and without cornering. Figure 3. (a) Normalized experimental photodiode waveform with inset showing raster scan path for
τPulse = 100 µs and (b) corresponding signal in the frequency domain with and without cornering. The laser power modulation in the AM250 is not an ideal rectangular wave. Figure 3a shows the response
from a photodiode exposed to scattered laser radiation. The measured laser/AOM response has rise and fall times
of ~ 10 µs. In the experiment shown in Fig. 3a, the laser is scanned over an alumina disk using τPulse = 100 µs. Between pulses, there is a delay time of τDelay = 10 µs. The duty cycle is 1 – τDelay/τPulse. The figure also shows that
at t = 310 µs the effect of the laser as it reverses its travel as part of the raster scan process (scan path illustrated in
the inset). This introduces a slightly longer delay (τDelay+Corner ~ 20 µs) before the next pulse. The variable τDelay is
a function of the point distance but is constant for PD < 60 µm, while τDelay+Corner is a function of point distance
and hatch spacing. Both of τDelay and τDelay+Corner are kept constant in this paper. The addition of the intermittent
delay associated with cornering introduces a phase delay in the frequency domain. Figure 3b shows the Fourier
transform of the measured laser/AOM response from a single line with and without cornering. The fundamen-
tal frequency without cornering occurs at f0 = 10 kHz with harmonics at integer multiples of the fundamental
frequency (red curve). However, the Fourier transform of a waveform including raster scanning (i.e., with cor-
nering) shows a shift in the fundamental frequency and a slight (0.5%) decrease in magnitude. This energy loss
is shifted to side bands (blue curve). t
A range of laser powers, pulse durations, and scan paths were recorded. As Fig. 3b shows, much of the laser
energy is outside of the measurement range. Assuming that the recoil force has the same frequency content as
the laser/AOM response, the measured recoil force at the fundamental frequency can be scaled by the ratio of the
laser energy at the fundamental frequency to the total laser energy. Results and discussion Single line scan path without powder. The simplest case occurs when no powder, i.e. h = 0 µm, is added
to the exposed surface of the tuning fork and the scan path consists of a single line to avoid cornering. A laser
PRF of f0 = 10 kHz and a point distance of PD = 1 µm, corresponding to a scan speed of V = 10 mm/s, are used. This scan speed is well below the range typical for 304L stainless steel13,33; however, it was selected to gather suf-
ficient data when processing a single line (i.e., the laser did not need to reverse its direction of travel ). Figure 4
shows experimental results for a Tuning Fork using various laser powers. Because of the atypically high linear
energy density, the FRF was measured after each experiment to determine if the sample was physically modified
(i.e., its frequency response changed). Figure 4a shows that the resonant frequency changed over a range of only
8.7 Hz during the five experiments. Further, the peak response at fr = 10.03 kHz is approximately 18 dB larger
than the next highest peak, demonstrating that most of the energy is contained at this frequency. The measured
acceleration during laser excitation is shown in Fig. 4b. While there is energy at other frequencies, this shows a
very sharp peak at the excitation frequency which matched the resonant frequency of the part. The value of the
acceleration also scales with the laser power, whereas the energy at other frequencies does not. This is due to the
fact that the laser is only providing energy at 10 kHz and the subsequent harmonics; therefore, the only portion
of the acceleration that should be changing with laser power is at 10 kHz. Figure 4c shows the force response cal-
culated by solving Eq. 1 for F(f) using the measured FRF and acceleration. An equivalent Noise Equivalent Force
(NEF), the ratio of the measured accelerometer noise to the difference between the FRF signal and the measured
FRF noise as described in the Methods section, is used to calculate an equivalent Signal to Noise Ratio (SNR). Experimental approachh This fraction varies with the duty cycle and
the scaling is described in the Methods section. Further, for the experiment conducted in this study, to prevent
aliasing of the PRF peak magnitude, the frequency resolution, df, is selected such that it is an integer multiple
of the PRF, i.e., rem(PRF/df) = 0. Results and discussion Figure 4c shows that this is much higher near the part resonance and supports the confidence of the magnitude
of the force at the laser PRF (more than five orders of magnitude greater SNR at f = 10 kHz than f = 5.25 kHz).f i
g
g
f
f
To examine the part independence of this result, the experiment is repeated with different tuning forks using
the same laser powers. The different tuning forks have slightly different FRFs with resonant frequencies listed in
Table 1 in the “Methods” section. Figure 5 shows the total recoil force as a function of laser power after correcting
for the fraction of the force at the fundamental PRF frequency. The corrective measure for the fraction of force https://doi.org/10.1038/s41598-021-90423-z Scientific Reports | (2021) 11:10959 | www.nature.com/scientificreports/ 0
10
20
30
40
50
60
10.0
5.25
a
]
g
m
[
PRF = 10 kHz
h = 0 μm
(b)
1
2
3
4
5
6
7
8
9
10
11
12
-60
-40
-20
0
20
40
(a)
P [W] 60 100 140 180 200
f [kHz]
F
R
F
])
N
/
g
(
B
d
[
1
2
3
4
5
6
7
8
9
10
11
12
0
1
2
3
4
f [kHz]
(c)
F
]
N
m
[
1E-5
1E-4
0.001
0.01
0.1
1
10
100
SNR
Figure 4. Results for Tuning Fork #16 (a) measured FRFs for various laser powers, (b) measured acceleration
spectrum with inset showing single line scanning strategy schematic, and (c) calculated forcing spectrum with
SNR spectrum (gray line). V = 10 mm/s. Figure 4. Results for Tuning Fork #16 (a) measured FRFs for various laser powers, (b) measured acceleration
spectrum with inset showing single line scanning strategy schematic, and (c) calculated forcing spectrum with
SNR spectrum (gray line). V = 10 mm/s. 40
60
80
100
120
140
160
180
0
5
10
15
20
25
Tuning Fork #
15
16
17
20
PRF = 10 kHz
V = 10 mm/s
F
]
N
m
[
P [W]
Figure 5. Recoil force versus applied laser power for single line scanning strategies using Tuning Forks #15,
#16, #17, and #20. Results and discussion 0
5
10
15
20
25
0
5
10
15
20
25
30
0
50
(a)
h [μm]
F
]
N
m
[
P/V [J/mm] 0
5
10
15
20
25
30
0
5
10
15
20
25
30
(b)
F
r
e
d
w
o
P
]
N
m
[
FNo Powder [mN] Figure 7. (a) Recoil force as function of linear energy density for both powder and no powder being and (b)
recoil force magnitude comparison for both powder and no powder. at the PRF frequency is shown in the Methods section in Fig. 12. Figure 5 shows that the recoil force scales with
the laser power with good agreement across multiple specimens. The increase in recoil force is due to the fact
that as the laser power increases a greater amount of material will be vaporized. at the PRF frequency is shown in the Methods section in Fig. 12. Figure 5 shows that the recoil force scales with
the laser power with good agreement across multiple specimens. The increase in recoil force is due to the fact
that as the laser power increases a greater amount of material will be vaporized. Rasters with powder. The slow scan speed and the lack of powder in Figs. 4 and 5 does not correspond to
typical LPBF processing conditions. Figure 6 shows the recoil force for a range of characteristic process param-
eters for 304L stainless steel34. A powder was spread across the surface of the tuning forks using two 50 µm thick
metal shims to create a uniform layer. The laser was rastered with scan speeds varied by changing the point
distance. The higher scan speeds, compared to that used in the previous section, required rastering the laser;
therefore, the recoil force was corrected using the measurement at the fundamental frequency to account for the
delay time introduced by cornering. The force correction for the additional delay time is shown in the Methods
section in Fig. 12. Figure 6a shows the variance in the measured recoil force with respect to scan speed and laser
power when the PRF is maintained at 6.25 kHz. The magnitude of the force is inversely proportional to scan
speed, which agrees with the results in Figs. 4 and 5 where an order of magnitude slower scan speed produced
recoil forces that were an order of magnitude higher. Results and discussion 40
60
80
100
120
140
160
180
0
5
10
15
20
25
Tuning Fork #
15
16
17
20
PRF = 10 kHz
V = 10 mm/s
F
]
N
m
[
P [W] Figure 5. Recoil force versus applied laser power for single line scanning strategies using Tuning Forks #15,
#16, #17, and #20. https://doi.org/10.1038/s41598-021-90423-z Scientific Reports | (2021) 11:10959 | www.nature.com/scientificreports/ 80
100
120
140
160
180
200
0.0
0.5
1.0
1.5
2.0
(a)
125
250
375
PRF = 6.25 kHz
h = 50 μm
V [mm/s]
F
]
N
m
[
P [W]
80
100
120
140
160
180
200
0.0
0.2
0.4
0.6
0.8
1.0
1.2
5.25
5.55
6.25
11.1
(b)
V = 316 mm/s
h = 50 μm
PRF [kHz]
F
]
N
m
[
P [W]
Figure 6. (a) Recoil force as function of laser power and scan speed for constant PRF and (b) recoil force as
function of applied laser power for constant scan speed. Uncertainty is detailed in “Methods” section. 80
100
120
140
160
180
200
0.0
0.2
0.4
0.6
0.8
1.0
1.2
5.25
5.55
6.25
11.1
(b)
V = 316 mm/s
h = 50 μm
PRF [kHz]
F
]
N
m
[
P [W] 80
100
120
140
160
180
200
0.0
0.5
1.0
1.5
2.0
(a)
125
250
375
PRF = 6.25 kHz
h = 50 μm
V [mm/s]
F
]
N
m
[
P [W] Figure 6. (a) Recoil force as function of laser power and scan speed for constant PRF and (b) recoil force as
function of applied laser power for constant scan speed. Uncertainty is detailed in “Methods” section. 0
5
10
15
20
25
0
5
10
15
20
25
30
0
50
(a)
h [μm]
F
]
N
m
[
P/V [J/mm]
0
5
10
15
20
25
30
0
5
10
15
20
25
30
(b)
F
r
e
d
w
o
P
]
N
m
[
FNo Powder [mN]
Figure 7. (a) Recoil force as function of linear energy density for both powder and no powder being and (b)
recoil force magnitude comparison for both powder and no powder. Results and discussion It is significant to note that the recoil force does not depend
appreciably on the PRF. This is illustrated in Fig. 6b where the scan speed is constant while the laser PRF and
laser power are varied. These results demonstrate that when the PRF is varied but the scan speed remains con-
stant, the material vaporization remains constant in the process parameter range considered in this study. p
p
p
g
y
Figure 7 shows the data in Figs. 4, 5 and 6 replotted as a function of the linear energy data, P/V. Single line
scans at slow scan speeds with powder are also included in the figure. There is minimal recoil force at very
low linear energy densities (notably, the data between P/V = 2 and 5 J/mm is generated using a laser power of
P = 40 W). However, lager linear energy densities, P/V > 9.2 J/mm, produce a significantly greater recoil force. This
is consistent with models of the LPBF process in Trapp et al.3, predicting that the greater temperatures produced
by higher linear energy densities increase the vaporization rate. The greater vaporization rate produces a higher
recoil pressure, forming a cavity which traps more laser energy to further increase the absorptivity of the melt
pool. This positive feedback on the recoil pressure leads to formation of keyholes. In general, the figure shows
that the addition of powder increases the recoil force by a factor of 1.33 from the case without powder. This can Scientific Reports | (2021) 11:10959 | https://doi.org/10.1038/s41598-021-90423-z www.nature.com/scientificreports/ Figure 8. Sample micrographs for different laser powers, scan speeds, and PRFs with melt pool dimensions
annotated in side picture that is annotated portion of melt pool indicated by dashed box. Figure 8. Sample micrographs for different laser powers, scan speeds, and PRFs with melt pool dimensions
annotated in side picture that is annotated portion of melt pool indicated by dashed box. 0.1
1
10
0
300
600
900
1200
1500
1800
2100
2400
h [μm]
0
50
0
200
400
600
800
0
200
400
600
800
δ
r
e
d
w
o
P
[μm]
δNo Powder [μm]
(a)
[μ
]
P/V [J/mm]
0.1
1
10
0
300
600
900
1200
1500
1800
h [μm]
0
50
0
200
400
600
0
200
400
600
w
r
e
d
w
o
P
[μm]
wNo Powder [μm]
(b)
w [μm]
P/V [J/mm]
Figure 9. Results and discussion Melt pool (a) depth and (b) width as function of linear energy density for experiments with powder
and no powder. Insets show melt pool dimensions for experiments with powder versus experiments without
powder. 0.1
1
10
0
300
600
900
1200
1500
1800
2100
2400
h [μm]
0
50
0
200
400
600
800
0
200
400
600
800
δ
r
e
d
w
o
P
[μm]
δNo Powder [μm]
(a)
δ [μm]
P/V [J/mm] 0.1
1
10
0
300
600
900
1200
1500
1800
h [μm]
0
50
0
200
400
600
0
200
400
600
w
r
e
d
w
o
P
[μm]
wNo Powder [μm]
(b)
w [μm]
P/V [J/mm] Figure 9. Melt pool (a) depth and (b) width as function of linear energy density for experiments with powder
and no powder. Insets show melt pool dimensions for experiments with powder versus experiments without
powder. be attributed to a greater absorptivity of the loose powder bed and the fact that isolated powder particles are
more readily vaporized because of reduced thermal conductance to surrounding material. Measurements of melt pool after solidification. The recoil force significantly affects the melt pool
morphology. Figure 8 shows melt pool metallographic micrographs for different process parameters obtained
from an auxiliary set of experiments (different substrates but the same processing conditions as the samples in
Figs. 2, 3, 4,5, 6 and 7, see “Methods” section). The melt pool depth, δ, and half-width, w/2, are defined in the
close up of the metallograph for P = 200 W and V = 200 mm/s in Fig. 8. As expected, the melt pool width and
depth increase with increasing laser powers and decrease with increasing scan speed. Most significantly, gas
porosity can be seen with lower scan speeds. These are plotted as a function of linear energy density in Fig. 9. It is interesting to note that the melt pool dimensions produced in experiments with and without powder are
linearly related. This is plotted in the insets of Fig. 9 where the melt pool depth for experiments with powder is
1.11 times the melt pool depth for experiments without powder, and the melt pool width for experiments with
powder 1.12 times the melt pool width for experiments without powder. This agrees with the increased recoil
forces measured for experiments with powder. Results and discussion The results indicate that increasing melt pool dimensions without
keyholing or significant recoil pressure for experiments with powder can be attributed to increased absorptance
of the powder bed, which leads to greater heating, and thus size, of the melt pool. https://doi.org/10.1038/s41598-021-90423-z Scientific Reports | (2021) 11:10959 | www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 10. Melt pool aspect ratio as function of recoil force with horizontal lines indicating dominant melting
modes and vertical lines representing recoil force magnitudes at melting mode boundaries with insets showing
micrographs that were collected at V = 120 mm/s and PRF = 2 kHz. Figure 10. Melt pool aspect ratio as function of recoil force with horizontal lines indicating dominant melting
modes and vertical lines representing recoil force magnitudes at melting mode boundaries with insets showing
micrographs that were collected at V = 120 mm/s and PRF = 2 kHz. Figure 10. Melt pool aspect ratio as function of recoil force with horizontal lines indicating dominant melting
modes and vertical lines representing recoil force magnitudes at melting mode boundaries with insets showing
micrographs that were collected at V = 120 mm/s and PRF = 2 kHz. The melt pool aspect ratio (AR), δ/w, has been shown to be correlated with different melting modes1. Specifi-
cally, Qi et al.35 gives ranges for the conduction, transition, and keyhole mode of AR ≤ 0.5, 0.5 < AR < 1.1, and
1.1 ≤ AR, respectively. Figure 10 shows the measured aspect ratio as a function of the measured recoil forces. Vertical dashed lines show the ranges of recoil forces producing each mode. Again, this agrees with the metal-
lography in Fig. 8 (close-ups shown in the insets of Fig. 10) showing the gas porosity resulting for slow scan
speeds and void size scaling with laser power. p
g
p
None of the process parameters produced AR < 0.5, which corresponds to conduction mode. For process
parameters with AR ~ 0.5, the recoil force was measured to be less than 0.5 mN. This appears to be a threshold
between conduction and transition melting modes, and recoil forces between 0.5 and 8 mN produced aspect
ratios in the range between 0.5 and 1.1, meaning these forces were in the transition region. Basing the definition
of the keyhole mode as the melt pool ratio of > 1.1 from Qi et al.35, it was empirically found that this corresponds
to a measured force of 10 mN. Results and discussion It is worth noting that with the results from Yin et al. 26, who estimated a pressure
of 49 kPa over a laser spot size of 159 µm, corresponding to a recoil force of 0.97 mN in the transition region
which agrees with our experimental findings. It is significant to note that the presence of powder during process-
ing does not significantly appear to change these thresholds. Summary and conclusionsh y
This paper demonstrated a method to measure the recoil force produced by the recoil pressure in LPBF using a
vibration response approach. While subject to well bounded error, the results correlate well with microstructure
analysis. The following conclusions can be drawn from the study: • Recoil force is proportionate to the linear energy density past a threshold where the powder begins to be
melted. • Recoil force is proportionate to the linear energy density past a threshold where the powder begins to be
melted. • The recoil force increases by 33% with the addition of a 50 µm layer of powder on the part surface compared
to a bare surface. • Melt pool depth and width scale with linear energy density for experiments both with and without powder. • The process is hypothesized to operate in the conduction, transition, and keyholing modes for recoil force
values less than 0.5 mN, between 0.5 and 8 mN, and greater than 10 mN, respectively. • Melt pool depth and width scale with linear energy density for experiments both with and without powder. • The process is hypothesized to operate in the conduction, transition, and keyholing modes for recoil force
values less than 0.5 mN, between 0.5 and 8 mN, and greater than 10 mN, respectively. Methods
i Equipment. The accelerometer used in this study was model 352C34 from PCB Piezotronics. The band-
width of the accelerometer was 0.005–12 kHz. The impact hammer used to measure the sample FRFs was model
086E80 from PCB Piezotronics. This impact hammer is rated to typically excite frequencies up to 12 kHz given
a metal tip. The coherence spectrum to support this is shown in Fig. 11, which was the coherence plot for the
response from Tuning Fork #15. The coherence response has notches which correlate to the anti-resonances in
the FRF response. Aside from the anti-resonances, the magnitude of the coherence is above 0.95, indicating good
correlation between the impact and the resulting acceleration.hhh p
g
The photodiode used to capture the waveforms was model PDA100A2 from Thorlabs. The bandwidth of the
photodiode was 11 MHz. Tuning fork dimensions. The prong lengths and resonant frequencies of the samples used in the experi-
ments are tabulated in Table 1. Materials. The material used in this study was 304L stainless steel purchased from LPW Technology. Scientific Reports | (2021) 11:10959 | https://doi.org/10.1038/s41598-021-90423-z www.nature.com/scientificreports/ 1
2
3
4
5
6
7
8
9
10
11
12
0.0
0.2
0.4
0.6
0.8
1.0
(a)
e
c
n
e
r
e
h
o
C
Frequency [kHz]
1
2
3
4
5
6
7
8
9
10
11
12
-40
-30
-20
-10
0
10
20
30
(b)
])
N
/
g
(
B
d
[
F
R
F
Frequency [kHz]
Figure 11. (a) Coherence spectrum and (b) FRF for tuning fork 15. 1
2
3
4
5
6
7
8
9
10
11
12
0.0
0.2
0.4
0.6
0.8
1.0
(a)
e
c
n
e
r
e
h
o
C
Frequency [kHz] 1
2
3
4
5
6
7
8
9
10
11
12
-40
-30
-20
-10
0
10
20
30
(b)
])
N
/
g
(
B
d
[
F
R
F
Frequency [kHz] Figure 11. (a) Coherence spectrum and (b) FRF for tuning fork 15. Figure 11. (a) Coherence spectrum and (b) FRF for tuning fork 15. 70
75
80
85
90
95
100
0
5
10
15
20
25
30
35
40
τCorner = 0 μs τCorner = 10 μs
MFund / MInput
]
%
[
DC [%]
1
Delay
Pulse
τ
DC
τ
Figure 12. Methods
i Percentage of energy located in fundamental frequency for various duty cycles and both straight
(red) and raster (blue) paths. 70
75
80
85
90
95
100
0
5
10
15
20
25
30
35
40
τCorner = 0 μs τCorner = 10 μs
MFund / MInput
]
%
[
DC [%]
1
Delay
Pulse
τ
DC
τ Figure 12. Percentage of energy located in fundamental frequency for various duty cycles and both straight
(red) and raster (blue) paths. Recoil force magnitude correction. For most of the samples in Table 1, the higher harmonics lie outside
of the bandwidth of the accelerometer used in this study. The total energy can be inferred from the energy in
the first harmonic assuming that the force response scales with the laser energy as measured by a photodiode
response. This gives the fractional energy in the first harmonic a dependence on the duty cycle of the laser’s pulse
train. The fractional energy was calculated by taking the ratio of the magnitude at the fundamental frequency to
the rms magnitude of the input signal i.e. MFund/MInput. Figure 12 shows the fraction of energy in the first har-
monic relative to the total energy of the measured photodiode signal for various laser duty cycles. gy
p
g
y y
Using the data in Fig. 12, the total estimated recoil force is, (2)
F =
af1
f
FRFf1
f
MInput
MFund (2) where a is the measured tuning fork acceleration magnitude during the laser excitation at the first harmonic,
and FRF is the experimentally measured tuning fork FRF magnitude at the first harmonic. where a is the measured tuning fork acceleration magnitude during the laser excitation at the first harmonic,
and FRF is the experimentally measured tuning fork FRF magnitude at the first harmonic. Uncertainty calculation. The uncertainty in the recoil force measurements was defined using the follow-
ng propagation of uncertainty (3)
FUnc
f
= |F|
EAccel
f
A
f
2
+
EForce
f
F
f
2 (3) where Eaccel and EForce are the margins of error for a 95% confidence interval, defined below, for the measured
FRFs and how the acceleration spectrums were processed. The margin of error is defined as where Eaccel and EForce are the margins of error for a 95% confidence interval, defined below, for the measured
FRFs and how the acceleration spectrums were processed. References 1. King, W. E. et al. Observation of keyhole-mode laser melting in laser powder-bed fusion additive manufacturing. J. Mater. Process
Technol. 214, 2915–2925 (2014). 2. Madison, J. D. & Aagesen, L. K. Quantitative characterization of porosity in laser welds of stainless steel. Scr. Mater. 67, 783–786
(2012). 3. Trapp, J., Rubenchik, A. M., Guss, G. & Matthews, M. J. In situ absorptivity measurements of metallic powders during laser powder
bed fusion additive manufacturing. Appl. Mater. Today 9, 341–349 (2017).ll g
pp
y
4. Rai, R., Elmer, J. W., Palmer, T. A. & Debroy, T. Heat transfer and fluid flow during keyhole mode laser welding of tantalum, Ti-6Al-
4V, 304L stainless steel and vanadium. J. Phys. D. Appl. Phys. 40, 5753–5766 (2007).l 4V, 304L stainless steel and vanadium. J. Phys. D. Appl. Phys. 40 y
pp
y
5. Mazumder, J., Ki, H. & Mohanty, P. S. Role of recoil pressure, multiple reflections, and free surface evolution during laser keyhole
welding. In ICALEO 2002-21st International Congress Application Laser Electro-Optics, Congress Proceedings Vol. 33 (2002). 6. Chen, Q., Guillemot, G., Gandin, C. A. & Bellet, M. Numerical modelling of the impact of energy distribution and M
surface tension on track shape in selective laser melting of ceramic material. Addit. Manuf. 21, 713–723 (2018). 7. Wu, Y. C. et al. Numerical modeling of melt-pool behavior in selective laser melting with random powder distribution and experi-
mental validation. J. Mater. Process. Technol. 254, 72–78 (2018). 8. Zhang, W. Probing Heat Transfer, Fluid Flow and Microstructural Evolution During Fusion Welding of Alloys. PhD Dissertation 312
(2004).l 9. Panwisawas, C. et al. On the role of thermal fluid dynamics into the evolution of porosity during selective laser melting. Scr. Mater
105, 14–17 (2015). 0. Tang, C., Le, K. Q. & Wong, C. H. Physics of humping formation in laser powder b e d fusion. Int. J. Heat Mass Transf. 149 (2020) 1. Xiao, B. & Zhang, Y. Marangoni and Buoyancy effects on direct metal laser sintering with a moving laser beam. Numer. Heat Transf
Part A Appl. 51, 715–733 (2007). pp
2. Heeling, T., Cloots, M. & Wegener, K. Melt pool simulation for the evaluation of process parameters in selective laser melting
Addit. Manuf. 14, 116–125 (2017). f
3. Lough, C. S., Wang, X., Smith, C. C., Landers, R. G., Bristow, D. A., Drallmeier, J. A., Brown, B. & Kinzel, E. C. Methods
i The margin of error is defined as (4)
E = Zα/2
σ
√
N
= 1.96 σ
√
N (4) where Z is the normal distribution indexed at the confidence level of interest, σ is the standard deviation of the
FRFs, and N is the number of FRFs taken. where Z is the normal distribution indexed at the confidence level of interest, σ is the standard deviation of the
FRFs, and N is the number of FRFs taken. https://doi.org/10.1038/s41598-021-90423-z https://doi.org/10.1038/s41598-021-90423-z https://doi.org/10.1038/s41598-021-90423-z Scientific Reports | (2021) 11:10959 | www.nature.com/scientificreports/ Auxiliary experimental procedure for melt pool dimensions. Rectangular prisms were printed of
dimensions 6.35 × 3.85 × 5 mm. The process parameters used to print the specimens were P = 200 W, V = 0.8 m/s,
exposure time of 75 µs, hatch spacing of 85 µm, and PD = 60 µm. After they were printed, their top surfaces were
scanned with the same laser powers, scan speeds, and PRFs as the ones seen in Figs. 6, 7 and 8a. After scanning,
the specimens were removed from the build plate, mounted using a Simplimet 1000, polished using an AutoMet
250 Grinder-Polisher, and then subsequently etched using a 40/60 nitric acid solution. The melt pool dimensions
were measured using a Hirox KH-8700 digital microscope. Received: 10 December 2020; Accepted: 4 May 2021 Received: 10 December 2020; Accepted: 4 May 2021 References Le, K. Q., Tang, C. & Wong, C. H. On the study of keyhole-mode melting in selective laser melting process. Int. J. Therm. Sci. 145
(2019).i 18. Zhang, M. J., Chen, G. Y., Zhou, Y., Li, S. C. & Deng, H. Observation of spatter formation mechanisms in high-power fiber laser
welding of thick plate. Appl. Surf. Sci. 280, 868–875 (2013). 19. Taheri Andani, M., Dehghani, R., Karamooz-Ravari, M. R., Mirzaeifar, R. & Ni, J. Spatter formation in selective laser melting
process using multi-laser technology. Mater. Des. 131, 460–469 (2017). p
g
gy
0. Ly, S., Rubenchik, A. M., Khairallah, S. A., Guss, G. & Matthews, M. J. Metal vapor micro-jet controls material redistribution in
laser powder bed fusion additive manufacturing. Sci. Rep. 7, 1–12 (2017). p
g
p
21. Zhao, C., Fezzaa, K., Cunningham, R. W., Wen, H., Carlo, F. D., Chen. L., Rollett, A. D. & Sun, T. Real-time monitoring of laser
powder bed fusion process using high-speed X-ray imaging and diffraction. Sci. Rep. 7, 1–11 (2017).f f
22. Zhang, Y., Fuh, J. Y. H., Ye, D. & Hong, G. S. In-situ monitoring of laser-based PBF via off-axis vision and image processing
approaches. Addit. Manuf. 25, 263–274 (2019).f pp
f
3. Taheri Andani, M., Dehghani, R., Karamooz-Ravari, M. R., Mirzaeifar, R. & Ni, J. A study on the effect of energy input on spatter
particles creation during selective laser melting process. Addit. Manuf. 20, 33–43 (2018). p
g
g p
f
24. Guo, Q., Zhao, C., Escano, L. I., Young, Z., Xiong, L., Fezzaa, K., Everhart, W., Brown, B., Sun, T. & Chen. L. Transient dynamics of
powder spattering in laser powder bed fusion additive manufacturing process revealed by in-situ high-speed high-energy X-ray
imaging. Acta Mater. 151, 169–180 (2018). g
g
(
)
25. Sutton, A. T., Kriewall, C. S., Leu, M. C., Newkirk, J. W. & Brown, B. Characterization of laser spatter and condensate generated
during the selective laser melting of 304L stainless steel powder. Addit. Manuf. 31. (2020). 26. Caprio, L., Demir, A. G., & Previtali, B. Observing molten pool surface oscillations durin
fusion as a novel method to estimate the penetration depth. Addit. Manuf. 36. (2020). 26. Caprio, L., Demir, A. G., & Previtali, B. Observing molten pool surface oscillations during keyhole processing in laser powder be
fusion as a novel method to estimate the penetration depth. Addit. Manuf. 36. (2020). p
p
f
27. References Correlation of SWIR
imaging with LPBF 304 L stainless steel part properties. Addit. Manuf. 35, 101359 (2020).l g
g
f
4. Aggarwal, A., Patel, S. & Kumar, A. Selective laser melting of 316L stainless steel: Physics of melting mode transition and its influ-
ence on microstructural and mechanical behavior. Jom 71, 1105–1116 (2019). 14. Aggarwal, A., Patel, S. & Kumar, A. Selective laser melting of 316L stainless steel: Phy
ence on microstructural and mechanical behavior. Jom 71, 1105–1116 (2019). 15. Sharma, S., Mandal, V., Ramakrishna, S. A. & Ramkumar, J. Numerical simulation of melt pool oscillations and protuberance in
pulsed laser micro melting of SS304 for surface texturing applications J Manuf Process 39 282 294 (2019) 15. Sharma, S., Mandal, V., Ramakrishna, S. A. & Ramkumar, J. Numerical simulation of melt pool oscillations and p
l
d l
l
f
f
f
l
f
(
) 15. Sharma, S., Mandal, V., Ramakrishna, S. A. & Ramkumar, J. Numerical simulation of melt pool oscillations and protuberance in
pulsed laser micro melting of SS304 for surface texturing applications. J. Manuf. Process. 39, 282–294 (2019). pulsed laser micro melting of SS304 for surface texturing applications. J. Manuf. Process. 39, 282–294 (2019). 16. Khairallah, S. A., Anderson, A. T., Rubenchik, A. & King, W. E. Laser powder-bed fusion additive manufact
l
lt fl
d f
ti
h
i
f
tt
d d
d ti
A t M t
108 36 45 pulsed laser micro melting of SS304 for surface texturing applications. J. Manuf. Process. 39, 282 294 (2019). 6. Khairallah, S. A., Anderson, A. T., Rubenchik, A. & King, W. E. Laser powder-bed fusion additive manufacturing: Physics o
l
lt fl
d f
ti
h
i
f
tt
d d
d ti
A t M t
108 36 45 (2016) 16. Khairallah, S. A., Anderson, A. T., Rubenchik, A. & King, W. E. Laser powder-bed fusion additive manufacturing: Physics of
complex melt flow and formation mechanisms of pores, spatter, and denudation zones. Acta Mater. 108, 36–45 (2016).h g
p
g
y
complex melt flow and formation mechanisms of pores, spatter, and denudation zones. Acta Mater. 108, 36–45 (2016). 17 Le K Q Tang C & Wong C H On the study of keyhole mode melting in selective laser melting process Int J Therm Sci 145 complex melt flow and formation mechanisms of pores, spatter, and denudation zones. Acta Mater. 108, 36 45 (2016). 7. References Zhao, C. et al. Bulk-explosion-induced metal spattering during laser processing. Phys. Rev. X 9, 21052 (2019). h
l
b
f p
p
g
g
p
g
y
28. Yin, J., Wang, D., Yang, L., Wei, H., Dong, P., Ke, L., Wang, G., Zhu, H. & Zeng, X. Correlation between forming quality and spatter
dynamics in laser powder bed fusion. Addit. Manuf. 31, 100958 (2019). 9. Fischer, P. et al. Microstructure of near-infrared pulsed laser sintered titanium samples. Appl. Phys. A Mater. Sci. Process. 78
1219–1227 (2004). h
l
f
ll
d
h
d
l
(
) 0. Fischer, P. et al. Sintering of commercially pure titanium powder with a Nd:YAG laser source. Acta Mater. 51, 1651–1662 (2003) 30. Fischer, P. et al. Sintering of commercially pure titanium powder with a Nd:YAG laser source. Acta Mater. 51, 1651–1662 (2003). 31. Qiu, C. et al. On the role of melt flow into the surface structure and porosity development during selective laser melting. Acta
Mater. 96, 72–79 (2015). 30. Fischer, P. et al. Sintering of commercially pure titanium powder with a Nd:YAG laser source. Acta Mater. 51, 1651–1662 (2003). 31. Qiu, C. et al. On the role of melt flow into the surface structure and porosity development during selective laser melting. Acta 2. Shrestha, S., Rauniyar, S. & Chou, K. Thermo-fluid modeling of selective laser melting: Single-track formation incorporating
metallic powder. J. Mater. Eng. Perform. 28, 611–619 (2019). g
f
33. West, B. M. et al. Modal analysis of metal additive manufactu
h
h
d y
p
f
(
)
34. Lough, C. S., Wang, X., Smith, C. C., Landers, R. G., Bristow, D. A., Drallmeier, J. A., Brown, B., & Kinzel, E. C. Correlation of
SWIR imaging with LBPF 304L stainless steel part properties. Addit. Manuf. 35 (2020). https://doi.org/10.1038/s41598-021-90423-z Scientific Reports | (2021) 11:10959 | www.nature.com/scientificreports/ 35. Qi, T. et al. Selective laser melting of Al7050 powder: Melting mode transition and comparison of the characteristics between the
keyhole and conduction mode. Mater. Des. 135, 257–266 (2017). Competing interests h p
g
The authors declare no competing interests. Acknowledgementsh g
This work was funded by Honeywell Federal Manufacturing & Technologies under Contract No. DE-NA0002839
with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the
article for publication, acknowledges that the United States Government retains a nonexclusive, paid up, irrevo-
cable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do
so, for the United States Government purposes. Author contributions T.C. R.L. and E.K. wrote the main manuscript text. T.C., C.L. and N.A. performed the experiments with T.C. pre-
paring the figures. B.B., T.H., and E.K. proposed the concept and T. H., A. B., J.B., K.J., E.K. and D.B. contributed
to modeling and development of the experiment. All authors reviewed the manuscript. Additional information Correspondence and requests for materials should be addressed to E.K. © The Author(s) 2021 Reprints and permissions information is available at www.nature.com/reprints. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2021 https://doi.org/10.1038/s41598-021-90423-z Scientific Reports | (2021) 11:10959 |
|
https://openalex.org/W4200300862
|
https://esurf.copernicus.org/preprints/esurf-2021-60/esurf-2021-60.pdf
|
English
| null |
Reply on RC1
| null | 2,021
|
cc-by
| 18,861
|
Linear stability analysis of plane beds under flows with suspended
load Koji Ohata1, Hajime Naruse1, and Norihiro Izumi2
1Division of Earth and Planetary Sciences, Graduate School of Science, Kyoto University, Japan
2Division of Field Engineering for the Environment, Faculty of Engineering, Hokkaido University, Japan
Correspondence: Koji Ohata (ohata.koji.24z@gmail.com) Koji Ohata1, Hajime Naruse1, and Norihiro Izumi2
1Division of Earth and Planetary Sciences, Graduate School of Science, Kyoto University, Japan
2Division of Field Engineering for the Environment, Faculty of Engineering, Hokkaido University, Japan
Correspondence: Koji Ohata (ohata.koji.24z@gmail.com) Abstract. Plane beds develop under flows in fluvial and marine environments; they are recorded as parallel lamination in
sandstone beds, such as those found in turbidites. However, whereas turbidites typically exhibit parallel lamination, they rarely
feature dune-scale cross lamination. Although the reason for the scarcity of dune-scale cross-lamination in turbidites is still
debated, the formation of dunes may be dampened by suspended load. Here, we perform, for the first time, linear stability
analysis to show that flows with suspended load facilitate the formation of plane beds. For a fine-grained bed, suspended load
5
can promote the formation of plane beds and dampen the formation of dunes. These results of theoretical analysis were verified
with observational data of plane beds under open-channel flows. Our theoretical analysis found that suspended load promotes
the formation of plane beds, which suggest that the development of dunes under turbidity currents is suppressed by the presence
of suspended load. 5 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. 1
Introduction
10 Our theoretical analysis reveals for the first time that suspended load promotes the formation of plane beds,
which has implications for interpreting sedimentary structures in turbidites. Izumi (2008) was extended in this study to evaluate plane bed formation under various conditions of sediment diameter and
35
flow depth and to consider the threshold condition of suspension. To evaluate the suspended load effect, linear stability analyses
were performed on flows both with and without suspended load. Further, we tested our stability diagrams against observational
data of plane beds. Our theoretical analysis reveals for the first time that suspended load promotes the formation of plane beds,
which has implications for interpreting sedimentary structures in turbidites. Izumi (2008) was extended in this study to evaluate plane bed formation under various conditions of sediment diameter and
35
flow depth and to consider the threshold condition of suspension. To evaluate the suspended load effect, linear stability analyses
were performed on flows both with and without suspended load. Further, we tested our stability diagrams against observational
data of plane beds. Our theoretical analysis reveals for the first time that suspended load promotes the formation of plane beds,
which has implications for interpreting sedimentary structures in turbidites. 2
Methods
40 Linear stability analysis of fluvial bedforms can provide the wavelengths of perturbations (i.e., bed waves) that grow over time
(Colombini, 2004; Bohorquez et al., 2019). The instability of a system, namely, whether the system develops a dune or antidune
under a given perturbation, can be illustrated as a contour diagram of the perturbation growth rate ωi (see Fig. 1), where ωi
is obtained as a function of the wavenumber k, Froude number Fr, sediment diameter ˜D, and flow depth of the uniform flow
q Linear stability analysis of fluvial bedforms can provide the wavelengths of perturbations (i.e., bed waves) that grow over time
(Colombini, 2004; Bohorquez et al., 2019). The instability of a system, namely, whether the system develops a dune or antidune
under a given perturbation, can be illustrated as a contour diagram of the perturbation growth rate ωi (see Fig. 1), where ωi
is obtained as a function of the wavenumber k, Froude number Fr, sediment diameter ˜D, and flow depth of the uniform flow
q Linear stability analysis of fluvial bedforms can provide the wavelengths of perturbations (i.e., bed waves) that grow over time
(Colombini, 2004; Bohorquez et al., 2019). The instability of a system, namely, whether the system develops a dune or antidune
under a given perturbation, can be illustrated as a contour diagram of the perturbation growth rate ωi (see Fig. 1), where ωi
is obtained as a function of the wavenumber k, Froude number Fr, sediment diameter ˜D, and flow depth of the uniform flow
q ˜h0. Here, k is defined as k = (2π˜h0)/˜λ, Fr is defined as Fr = ˜U0/
q
˜g˜h0, ˜λ denotes the wavelength of the perturbation, ˜U0 is
45
the depth-averaged flow velocity of the uniform flow, and ˜g is the gravitational acceleration. Thus, we can obtain the growth
rate ωi as a function of k for given combinations of Froude number, the sediment diameter, and flow depth
Fr, ˜D,˜h0
. We
employ the two-dimensional Reynolds-averaged Navier-Stokes equations as the governing equations for flows and the quasi-
steady assumption to neglect the unsteady terms in the flow equations. The eddy viscosity is evaluated using a mixing-length approach. In this study, bed-load discharge is estimated using the Meyer-Peter and Müller formula modified as described in
50
Wong and Parker (2006). The entrainement rate of suspended load is estimated using the relationship proposed in de Leeuw
et al. (2020). 1
Introduction
10 The interactions between fluids and erodible surfaces generate small-scale topographic features called bedforms both on ter-
restrial surfaces (e.g., riverbeds, deserts, and deep-sea floors) and on extra-terrestrial surfaces (Bourke et al., 2010; Gao et al.,
2015; Hage et al., 2018; Cisneros et al., 2020). Such bedforms are preserved in sedimentary rocks as sedimentary structures
such as cross- and parallel lamination (Harms, 1979). The types of sedimentary structures observed vary among different types of rocks. Turbidites typically exhibit parallel lamination (Bouma, 1962), whereas they rarely feature dune-scale cross-
15
lamination (Talling et al., 2012). However, the opposite is true for fluvial deposits; i.e., dune-scale cross laminae are often
observed in riverine sandstone (Miall, 2010). Although the reason for the paucity of dune-scale cross-lamination in turbidites is still debated (Lowe, 1988; Arnott, 2012;
Schindler et al., 2015; Tilston et al., 2015), it could be attributed to the presence of suspended load. For example, in the case of open-channel flows, nearly flat bed waves and low-angle dunes have been observed in suspension-dominated rivers (Smith
20
and McLean, 1977; Kostaschuk and Villard, 1996; Bradley et al., 2013; Ma et al., 2017). Additionally, flume experiments have
suggested that dune height decreases with increasing suspended load flux (Bridge and Best, 1988; Naqshband et al., 2017). Therefore, the influences of suspended load on the suppression of dune development and the formation of plane beds must be
considered. 1 1 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. The relationships between sediment transport modes and the formation of plane beds have received little attention in theoret-
25
ical works that performed linear stability analyses. The reason could be because previous studies have succeeded in predicting
the wavelength of dunes and antidunes without considering suspended load (Colombini, 2004; Di Cristo et al., 2006; Colombini
and Stocchino, 2008; Vesipa et al., 2012; Bohorquez et al., 2019). However, this assumption is not appropriate for analyzing
open-channel flows where the suspended load is not negligible, such as flows in rivers with a fine sediment bed (de Almeida The relationships between sediment transport modes and the formation of plane beds have received little attention in theoret-
25
ical works that performed linear stability analyses. 1
Introduction
10 The reason could be because previous studies have succeeded in predicting
the wavelength of dunes and antidunes without considering suspended load (Colombini, 2004; Di Cristo et al., 2006; Colombini
and Stocchino, 2008; Vesipa et al., 2012; Bohorquez et al., 2019). However, this assumption is not appropriate for analyzing
open-channel flows where the suspended load is not negligible, such as flows in rivers with a fine sediment bed (de Almeida et al., 2016; Sambrook Smith et al., 2016). Moreover, although some research has considered both bed- and suspended load
30
(Engelund, 1970; Nakasato and Izumi, 2008; Bose and Dey, 2009), the hydraulic conditions of these analyses were limited,
and the results were tested using only observational data of dunes and antidunes. Therefore, in order to investigate the effect of sediment transport mode on the formation of plane beds, we performed linear
stability analyses of bedforms under open-channel flows carrying suspended load. The model introduced in Nakasato and et al., 2016; Sambrook Smith et al., 2016). Moreover, although some research has considered both bed- and suspended load
30
(Engelund, 1970; Nakasato and Izumi, 2008; Bose and Dey, 2009), the hydraulic conditions of these analyses were limited,
and the results were tested using only observational data of dunes and antidunes. Therefore, in order to investigate the effect of sediment transport mode on the formation of plane beds, we performed linear
stability analyses of bedforms under open-channel flows carrying suspended load. The model introduced in Nakasato and et al., 2016; Sambrook Smith et al., 2016). Moreover, although some research has considered both bed- and suspended load
30
(Engelund, 1970; Nakasato and Izumi, 2008; Bose and Dey, 2009), the hydraulic conditions of these analyses were limited,
and the results were tested using only observational data of dunes and antidunes. Therefore, in order to investigate the effect of sediment transport mode on the formation of plane beds, we performed linear
stability analyses of bedforms under open-channel flows carrying suspended load. The model introduced in Nakasato and Izumi (2008) was extended in this study to evaluate plane bed formation under various conditions of sediment diameter and
35
flow depth and to consider the threshold condition of suspension. To evaluate the suspended load effect, linear stability analyses
were performed on flows both with and without suspended load. Further, we tested our stability diagrams against observational
data of plane beds. 2
Methods
40 See the following section for details. To test the results of linear stability analyses against the observational data of plane beds, we must plot stability diagrams
in the parametric space of hydraulic parameters that are independent of the wavelength; i.e., Froude number Fr, the sediment approach. In this study, bed-load discharge is estimated using the Meyer-Peter and Müller formula modified as described in
50
Wong and Parker (2006). The entrainement rate of suspended load is estimated using the relationship proposed in de Leeuw
et al. (2020). See the following section for details. diameter ˜D, and flow depth ˜h0. 55
In this study, we illustrate stability diagrams as contour maps of the dominant wavenumber kd on ˜h0-Fr plane with fixed
˜D to investigate the impact of suspended load on the formation of plane beds. Here, the dominant wavenumber kd is the diameter ˜D, and flow depth ˜h0. 55
In this study, we illustrate stability diagrams as contour maps of the dominant wavenumber kd on ˜h0-Fr plane with fixed
˜D to investigate the impact of suspended load on the formation of plane beds. Here, the dominant wavenumber kd is the diameter ˜D, and flow depth h0. 55
In this study, we illustrate stability diagrams as contour maps of the dominant wavenumber kd on ˜h0-Fr plane with fixed
˜D to investigate the impact of suspended load on the formation of plane beds. Here, the dominant wavenumber kd is the 2 2 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. wavenumber that provides the maximum growth rate. We assume that the system is stable if ωi is not positive for all k within
the domain [0.01,1.5] for a given
Fr, ˜D,˜h0
combination (Fig. 1). In contrast, the system is assumed to be unstable if ωi is
positive for some k. 60 wavenumber that provides the maximum growth rate. We assume that the system is stable if ωi is not positive for all k within
the domain [0.01,1.5] for a given
Fr, ˜D,˜h0
combination (Fig. 1). In contrast, the system is assumed to be unstable if ωi is
positive for some k. 60 60 2.1
Linear Stability Analysis Linear stability analysis of fluvial bedforms can provide the wavelengths of perturbations (i.e., bed waves) that grow over time
(Colombini, 2004; Bohorquez et al., 2019). Here we present the formulation of the problem and the method used to solve the
differential equations. 2.1.1
Formulation of the Problem
65 The governing equations for flows are the two-dimensional Reynolds-averaged Navier-Stokes equations. On erodible beds, the
flow adjustments occurs immediately relative to the bed adjustments. Therefore, we employ the quasi-steady assumption to
neglect the unsteady terms in the flow equations (Colombini, 2004; Yokokawa et al., 2016). Under the quasi-steady assumption, the dimensionless forms of the Reynolds-averaged Navier-Stokes equations and conti-
nuity equation for incompressible flow are described as:
70 u∂u
∂x + w∂u
∂z = −∂p
∂x + 1 + ∂Txx
∂x + ∂Txz
∂z
(1)
u∂w
∂x + w∂w
∂z = −∂p
∂z + S−1 + ∂Txz
∂x + ∂Tzz
∂z
(2)
∂u
∂x + ∂w
∂z = 0
(3) u∂u
∂x + w∂u
∂z = −∂p
∂x + 1 + ∂Txx
∂x + ∂Txz
∂z
(1)
u∂w
∂x + w∂w
∂z = −∂p
∂z + S−1 + ∂Txz
∂x + ∂Tzz
∂z
(2)
∂u
∂x + ∂w
∂z = 0
(3) u∂u
∂x + w∂u
∂z = −∂p
∂x + 1 + ∂Txx
∂x + ∂Txz
∂z
u∂w
∂x + w∂w
∂z = −∂p
∂z + S−1 + ∂Txz
∂x + ∂Tzz
∂z
∂u
∂x + ∂w
∂z = 0 (2) (3) where u and w are the flow velocities in x- and z- direction, respectively; p denotes the pressure; S is the bed slope; and
Tij (i,j = x,z) is the Reynolds stress tensor. 75 where u and w are the flow velocities in x- and z- direction, respectively; p denotes the pressure; S is the bed slope; and Tij (i,j = x,z) is the Reynolds stress tensor. In the above equations, the system is nondimensionalized as follows:
5 In the above equations, the system is nondimensionalized as follows:
85 85 (u,w) = (˜u, ˜w)/˜uf0
(x,z,h,Z,R,D) = (˜x, ˜z,˜h, ˜Z, ˜R, ˜D)/˜h0
(p,Tij) = (˜p, ˜Tij)/˜ρ˜h0
νT = ˜˜νT /(˜uf0˜h0) (u,w) = (˜u, ˜w)/˜uf0
(9)
(x,z,h,Z,R,D) = (˜x, ˜z,˜h, ˜Z, ˜R, ˜D)/˜h0
(10)
(p,Tij) = (˜p, ˜Tij)/˜ρ˜h0
(11)
νT = ˜˜νT /(˜uf0˜h0)
(12) (u,w) = (˜u, ˜w)/˜uf0
(9)
(x,z,h,Z,R,D) = (˜x, ˜z,˜h, ˜Z, ˜R, ˜D)/˜h0
(10)
(p,Tij) = (˜p, ˜Tij)/˜ρ˜h0
(11)
νT = ˜˜νT /(˜uf0˜h0)
(12) (u,w) = (˜u, ˜w)/˜uf0
(9)
(x,z,h,Z,R,D) = (˜x, ˜z,˜h, ˜Z, ˜R, ˜D)/˜h0
(10)
(p,Tij) = (˜p, ˜Tij)/˜ρ˜h0
(11)
νT = ˜˜νT /(˜uf0˜h0)
(12) (11)
(12) (11) (12) where ˜uf0 denotes the shear velocity in the basic flat-bed state, D is the diameter of a bed particle, and ˜ρ is the water density
90
(= 1000 kg/m3). Hereafter, we denote dimensional variables using a tilde (˜). The shear velocity in the basic flat-bed state ˜uf0
is obtained as: where ˜uf0 denotes the shear velocity in the basic flat-bed state, D is the diameter of a bed particle, and ˜ρ is the water density
90
(= 1000 kg/m3). Hereafter, we denote dimensional variables using a tilde (˜). The shear velocity in the basic flat-bed state ˜uf0
is obtained as: ˜uf0 =
q
˜g˜h0S ˜uf0 =
q
˜g˜h0S (13) As the flow is continuous, the system can be rewritten using the stream function ψ defined as: As the flow is continuous, the system can be rewritten using the stream function ψ defined as: (u,w) =
∂ψ
∂z ,−∂ψ
∂x
95 (u,w) =
∂ψ
∂z ,−∂ψ
∂x
(14)
95 (14) Then, Eqs. (1) and (2) are rearranged to: ∂ψ
∂z
∂2ψ
∂x∂z −∂ψ
∂x
∂2ψ
∂z2 = −∂p
∂x + 1 + ∂
∂x
2νT
∂2ψ
∂x∂z
+ ∂
∂z
νT
∂2ψ
∂z2 −∂2ψ
∂x2
∂ψ
∂x
∂2ψ
∂x∂z −∂ψ
∂z
∂2ψ
∂x2 = −∂p
∂z + S−1 −∂
∂z
2νT
∂2ψ
∂x∂z
+ ∂
∂x
νT
∂2ψ
∂z2 −∂2ψ
∂x2
(15) (16) 100 100 Eliminating p from Eqs. (15) and (16), we obtain: Eliminating p from Eqs. 2.1.1
Formulation of the Problem
65 75
We employ a Boussinesq-type assumption to close the flow equations: j
We employ a Boussinesq-type assumption to close the flow equations: We employ a Boussinesq-type assumption to close the flow equations: Txx = 2νT
∂u
∂x
(4)
Tzz = 2νT
∂w
∂z
(5)
Txz = νT
∂u
∂x + ∂w
∂z
(6) Txx = 2νT
∂u
∂x
(4)
Tzz = 2νT
∂w
∂z
(5)
Txz = νT
∂u
∂x + ∂w
∂z
(6) Txx = 2νT
∂u
∂x
Tzz = 2νT
∂w
∂z
Txz = νT
∂u
∂x + ∂w
∂z
(4) (5) (6) Then, the eddy viscosity νT is evaluated using a mixing-length approach:
80 Then, the eddy viscosity νT is evaluated using a mixing-length approach:
80 Then, the eddy viscosity νT is evaluated using a mixing-length approach:
80 νT = l2
∂u
∂z
(7)
l = κ(z −Z)
r
h + R −z
h
(8) (8) where l is the mixing length, κ is the Kármán coefficient (= 0.4), h is the flow depth, Z denotes the bed height, and R is the
height of the reference level at which the flow velocity is assumed to vanish in a logarithmic profile (Fig. A1). 3 3 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. In the above equations, the system is nondimensionalized as follows:
5 (15) and (16), we obtain: ∂ψ
∂z
∂
∂x∇2ψ −∂ψ
∂x
∂
∂z ∇2ψ −4 ∂2
∂x∂z
νT
∂2ψ
∂x∂z
+
∂2
∂x2 −∂2
∂z2
νT
∂2
∂z2 −∂2
∂x2
ψ
= 0
(17) ∂ψ
∂z
∂
∂x∇2ψ −∂ψ
∂x
∂
∂z ∇2ψ −4 ∂2
∂x∂z
νT
∂2ψ
∂x∂z
+
∂2
∂x2 −∂2
∂z2
νT
∂2
∂z2 −∂2
∂x2
ψ
= 0 (17) T
∂2
∂z2 −∂2
∂x2
ψ
= 0
(1 We also assume a quasi-steady state for the advection-diffusion equation for suspended sediment, which is formulated as: We also assume a quasi-steady state for the advection-diffusion equation for suspended sedimen ∂Fx
∂x + ∂Fz
∂z = 0 ∂Fx
∂x + ∂Fz
∂z = 0
(18)
5 (18) Here, Fx and Fz are the normalized fluxes of suspended sediment in x- and z- directions, respectively, given by: ∂Fx
∂x = uc −νT
∂c
∂x
(19)
∂Fz
∂z = (w −ws)c −νT
∂c
∂z
(20) ∂Fx
∂x = uc −νT
∂c
∂x
(19)
∂Fz
∂z = (w −ws)c −νT
∂c
∂z
(20) ∂Fx
∂x = uc −νT
∂c
∂x
∂F (19) ∂Fz
∂z = (w −ws)c −νT
∂c
∂z (20) https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. where c denotes the concentration of suspended sediment and ws is the settling velocity of sediment. We assume that the diffu-
sion coefficient of suspended sediment is equal to the eddy viscosity νT . Based on Eqs. (19) and (20), Eq. In the above equations, the system is nondimensionalized as follows:
5 (18) is reformulated
110
as: 110 u ∂c
∂x + (w −ws) ∂c
∂z = ∂
∂x
νT
∂c
∂x
+ ∂
∂z
νT
∂c
∂z
u ∂c
∂x + (w −ws) ∂c
∂z = ∂
∂x
νT
∂c
∂x
+ ∂
∂z
νT
∂c
∂z
(21) (21) velocity of sediment ws is calculated using a relationship given in Ferguson and Church (2004): The settling velocity of sediment ws is calculated using a relationship given in Ferguson and Church (2004): The settling velocity of sediment ws is calculated using a relationship given in Ferguson and ws =
˜ws
q
Rs˜g ˜D
(22)
˜ws =
Rs˜g ˜D2
C1˜ν + 0.75C2
q
Rs˜g ˜D3
(23)
115 (22) (23) 115 where the constants C1 and C2 are set to the values for smooth spheres: C1 = 18 and C2 = 0.4. The particle Reynolds number Rep is defined as: Rep =
q
Rs˜g ˜D3
˜ν Rep =
q
Rs˜g ˜D3
˜ν
(24) Rs˜g ˜D3
˜ν (24) where Rs is the submerged specific density and ˜ν is the kinematic viscosity of the fluid (= 1.0 × 10−6 m2/s). The submerged
specific density Rs is defined as: where Rs is the submerged specific density and ˜ν is the kinematic viscosity of the fluid (= 1.0 × 10−6 m2/s). The submerged
specific density Rs is defined as:
20 Rs = ˜ρs −˜ρ
˜ρ
(25)
h
˜ d
t
th d
it
f th b d
ti l
(
2650 k /
3) Rs = ˜ρs −˜ρ
˜ρ Rs = ˜ρs −˜ρ
˜ρ
(25) Rs = ˜ρs −˜ρ
˜ρ (25) where ˜ρs denotes the density of the bed particles (= 2650 kg/m3). where ˜ρs denotes the density of the bed particles (= 2650 kg/m3). In the above equations, the system is nondimensionalized as follows:
5 Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. Since (R −Z)/h ≪1, then we can obtain: Since (R −Z)/h ≪1, then we can obtain:
l
(h
R
Z)
p
1
(31) Since (R −Z)/h ≪1, then we can obtain: l = κ(hη + R −Z)
p
1 −η
(31) l = κ(hη + R −Z)
p
1 −η (31) The boundary conditions include a vanishing flow component normal to the water surface, and vanishing stresses normal
and tangential to the water surface as follows: 135 135 u · ens = 0
ens · T · ens = 0
ets · T · ens = 0
at
η = 1 u · ens = 0
ens · T · ens = 0
ets · T · ens = 0
at
η = 1
(32) u · ens = 0
ens · T · ens = 0
ets · T · ens = 0
at
η = 1
(32) = 0
= 0
= 0
at
η = 1 (32) where u = (u,w) is the velocity vector, e denotes the unit vector, and T is the stress tensor. The subscripts ns and ts denote
directions normal and tangential to the water surface, respectively. where u = (u,w) is the velocity vector, e denotes the unit vector, and T is the stress tensor. The subscripts ns and ts denote
directions normal and tangential to the water surface, respectively. where u = (u,w) is the velocity vector, e denotes the unit vector, and T is the stress tensor. The subscripts ns and ts denote
directions normal and tangential to the water surface, respectively. At the bed, the boundary conditions include the vanishing flow components normal and tangential to the bed. 140 140 u · enb = 0
u · etb = 0
at
η = 0
(33) (33) where the subscripts nb and tb denote directions normal and tangential to the bed, respectively. The vectors ens, ets, enb, and
etb, and the tensor T are defined as: where the subscripts nb and tb denote directions normal and tangential to the bed, respectively. In the above equations, the system is nondimensionalized as follows:
5 We employ the following transformation of variables to apply the boundary condition at the bed and flow surfaces: We employ the following transformation of variables to apply the boundary condition at the bed and flow surfaces: ξ = x
(26)
η = z −R(x)
h(x)
(27)
25 ξ = x
η = z −R(x)
h(x)
125 (26) (27) The derivatives with respect to x and z are described as follows: The derivatives with respect to x and z are described as follows: The derivatives with respect to x and z are described as follows: The derivatives with respect to x and z are described as follows: ∂
∂x = ∂
∂ξ −η∂xh + ∂xR
h
∂
∂η
(28)
∂
∂z = 1
h
∂
∂η
(29) (28) (29) where ∂x denotes the partial derivative with respect to x. Using the above transformation of variables approach, the height of
the water surface and the reference level correspond to η = 1 and η = 0, respectively. where ∂x denotes the partial derivative with respect to x. Using the above transformation of variables approach, the height of
the water surface and the reference level correspond to η = 1 and η = 0, respectively. 0 where ∂x denotes the partial derivative with respect to x. Using the above transformation of variables approach, the height of the water surface and the reference level correspond to η = 1 and η = 0, respectively. 130
Additionally, the dimensionless mixing length l (Eq. (8)) is rearranged as: the water surface and the reference level correspond to η = 1 and η = 0, respectively. 130
Additionally, the dimensionless mixing length l (Eq. (8)) is rearranged as: the water surface and the reference level correspond to η = 1 and η = 0, respectively. 130
Additionally, the dimensionless mixing length l (Eq. (8)) is rearranged as: the water surface and the reference level correspond to η = 1 and η = 0, respectively. 130
Additionally, the dimensionless mixing length l (Eq. (8)) is rearranged as: p
η
η
p
y
Additionally, the dimensionless mixing length l (Eq. (8)) is rearranged as: Additionally, the dimensionless mixing length l (Eq. (8)) is rearranged as: Additionally, the dimensionless mixing length l (Eq. (8)) is rearranged as: l = κ(hη + R −Z)
s
1 −η
1 + (R −Z)/h (30) 5 https://doi.org/10.5194/esurf-2021-60
Preprint. In the above equations, the system is nondimensionalized as follows:
5 The vectors ens, ets, enb, and
etb, and the tensor T are defined as: ens =
1
p
1 + ∂x(R + h)2
−∂x(R + h),1
ets =
1
p
1 + ∂x(R + h)2
1,∂x(R + h)
145
enb =
1
√1 + ∂xR2
−∂xR,1
etb =
1
√
1 + ∂xR2
1,∂xR
T =
−p + Txx
Txz
Txz
−p + Tzz
ens =
1
p
1 + ∂x(R + h)2
−∂x(R + h),1
(34)
ets =
1
p
1 + ∂x(R + h)2
1,∂x(R + h)
(35)
5
enb =
1
√1 + ∂xR2
−∂xR,1
(36)
etb =
1
√
1 + ∂xR2
1,∂xR
(37)
T =
−p + Txx
Txz
Txz
−p + Tzz
(38) (34) (35) 145 (38) The boundary conditions for the suspended sediment flux at the flow surface and bed are as follows: The boundary conditions for the suspended sediment flux at the flow surface and bed are as follows: F · ens = 0
at
η = 1
(39)
150
F · enb =
˜Es
˜uf0
at
η = 0
(40)
where F = (Fx,Fz) is the flux vector of suspended sediment and ˜Es is the entrainment rate of the sediment calculated as
˜Es = ˜wsEs. In this study, the dimensionless coefficient Es is estimated using the relationship proposed in de Leeuw et al. (2020): F · ens = 0
at
η = 1
(39)
150
F · enb =
˜Es
˜uf0
at
η = 0
(40)
where F = (Fx,Fz) is the flux vector of suspended sediment and ˜Es is the entrainment rate of the sediment calculated as F · ens = 0
at
η = 1
(39)
50
F · enb =
˜Es
˜uf0
at
η = 0
(40) 150 (39) (40) where F = (Fx,Fz) is the flux vector of suspended sediment and ˜Es is the entrainment rate of the sediment calculated as
˜Es = ˜wsEs. In this study, the dimensionless coefficient Es is estimated using the relationship proposed in de Leeuw et al. (2020): Es = 5.73 × 10−3
uf
ws
1.31
Fr1.59 Rep
−0.8
155 Es = 5.73 × 10−3
uf
ws
1.31
Fr1.59 Rep
−0.86
(41)
155 (41) 6 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. In the above equations, the system is nondimensionalized as follows:
5 (48) from η = 0 to η = 1: Cz =
˜U0
˜uf0
= 1
κ
(1 + R0)ln
1 + R0
R0
−1
Cz =
˜U0
˜uf0
= 1
κ
(1 + R0)ln
1 + R0
R0
−1
(49) (49) ic law of the open-channel flows as: Now, we consider the logarithmic law of the open-channel flows as:
170 Now, we consider the logarithmic law of the open-channel flows as:
170 Now, we consider the logarithmic law of the open-channel flows as:
170 Now, we consider the logarithmic law of the open-channel flows as:
170 u = 1
κ ln
z
ks
+ 8.5 = 1
κ ln
30z
mD
u = 1
κ ln
z
ks
+ 8.5 = 1
κ ln
30z
mD
(50) u = 1
κ ln
z
ks
+ 8.5 = 1
κ ln
30z
mD
(50) 30z
mD
(50) (50) where ks denotes the roughness height and m is the ratio of ks to D. The velocity U vanishes when z = mD/30, that is
R0 = mD/30. We set m as 2.5 in this study, which corresponds to R0 = D/12. Additionally, we set the origin of z-axis at a
distance of D/6 below the top of the bed particles (Fig. A2). By setting the top of the bed particles as z = D/6, the reference where ks denotes the roughness height and m is the ratio of ks to D. The velocity U vanishes when z = mD/30, that is
R0 = mD/30. We set m as 2.5 in this study, which corresponds to R0 = D/12. Additionally, we set the origin of z-axis at a
distance of D/6 below the top of the bed particles (Fig. A2). By setting the top of the bed particles as z = D/6, the reference
level R0 is positioned below the top of bed particles. Therefore, the domain in which the mixing-length approach cannot be
175
li d i
t i t d
th b d level R0 is positioned below the top of bed particles. Therefore, the domain in which the mixing-length approach cannot be
175
applied is restricted near the bed. Under the above uniform flow condition over a flat bed, Eq. In the above equations, the system is nondimensionalized as follows:
5 The basic flow state for linear stability analysis is a uniform flow over a flat bed. Under this condition, the hydraulic param-
eters u, w, h, Z, R, and c are described as: (u,w,h,Z,R,c) = (u0(η),0,1,0,R0,c0(η)) (42) where the subscript 0 denotes a parameter in the basic state. The governing equations of flows can be simplified as: where the subscript 0 denotes a parameter in the basic state. The governing equations of flows can be simplified as: 1 + ∂Txy0
∂η
= 0
160
Txy0 = νT 0
∂u0
∂η
νT 0 = l0
2 ∂u0
∂η
l0 = κ(η + R0)
p
1 − (43) (44) (45) (46) with the boundary conditions: u0 = 0, Txy0 = 1
at
η = 0
(47)
165
With Eqs. (43)–(47), we can obtain the following logarithmic law for the flow velocity:
u0(η) = 1
κ ln
η + R0
R0
(48) u0 = 0, Txy0 = 1
at
η = 0 u0 = 0, Txy0 = 1
at
η = 0
165 (47) (47), we can obtain the following logarithmic law for the flow velocity: With Eqs. (43)–(47), we can obtain the following logarithmic law for the flow velocity: With Eqs. (43)–(47), we can obtain the following logarithmic law for the flow velocity: u0(η) = 1
κ ln
η + R0
R0
u0(η) = 1
κ ln
η + R0
R0
(48) Then, the friction coefficient Cz is obtained by the direct integration of Eq. (48) from η = 0 to η = 1: Then, the friction coefficient Cz is obtained by the direct integration of Eq. (48) from η = 0 to η = 1: Then, the friction coefficient Cz is obtained by the direct integration of Eq. (48) from η = 0 to η = 1: Then, the friction coefficient Cz is obtained by the direct integration of Eq. In the above equations, the system is nondimensionalized as follows:
5 (21) can be rewritten as: Under the above uniform flow condition over a flat bed, Eq. (21) can be rewritten as: Under the above uniform flow condition over a flat bed, Eq. (21) can be rewritten as: Under the above uniform flow condition over a flat bed, Eq. (21) can be rewritten as: −ws
∂c0
∂η = ∂
∂η
νT 0
∂c0
∂η
−ws
∂c0
∂η = ∂
∂η
νT 0
∂c0
∂η
(51) (51) with the following boundary conditions: wsc0 + νT 0
∂c0
∂η = 0
at
η = 1
180
c0 = cb
at
η = 0 wsc0 + νT 0
∂c0
∂η = 0
at
η = 1
180 (52) c0 = cb
at
η = 0 (53) 7 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. Here, cb is the near-bed concentration of suspended sediment. Under the basic state, the entrainment and deposition rates of
the suspended sediment are balanced. Thus, cb is described as: cb = Es0
(54)
Es0 = 5.73 × 10−3
uf0
ws
1.31
Fr1.59 Rep
−0.86
(55) (54) cb = Es0 Es0 = 5.73 × 10−3
uf0
ws
1.31
Fr1.59 Rep
−0.86
185 3
uf0
ws
1.31
Fr1.59 Rep
−0.86 Es0 = 5.73 × 10−3
uf0
ws
1.31
Fr1.59 Rep
−0.86
(55) 31
Fr1.59 Rep
−0.86
(55) (55) c0(η) = cb
R0(1 −η)
η + R0
ws/κ(1+R0) c0(η) = cb
R0(1 −η)
η + R0
ws/κ(1+R0) (56) (1 −λp)∂˜B
∂˜t + αb
∂˜qB
∂˜x + αs ˜ws
Es −c[ξ,ηb]
= 0
190 αs ˜ws
Es −c[ξ,ηb]
= 0
(5 (57) where λp denotes the sediment porosity, ˜B denotes the height of the bed-load layer, ˜t is time, and ˜qB denotes the bed-load
discharge per unit width. The coefficients αb and αs take a value of 0 or 1 depending on the sediment transport regime. Equation (57) is nondimensionalized as: where λp denotes the sediment porosity, ˜B denotes the height of the bed-load layer, ˜t is time, and ˜qB denotes the bed-load
discharge per unit width. The coefficients αb and αs take a value of 0 or 1 depending on the sediment transport regime. In the above equations, the system is nondimensionalized as follows:
5 (Brownlie, 1981; Niño et al., 2003), respectively, as follows: 215 θch = 0.22Rep
−0.6 + 0.06exp(−17.77Rep)−0.6
(68)
uf
ws
c
=
21.2Rep
−1.2
(1 < Rep < 27.3)
0.4
(27.3 ≤Rep)
(69) θch = 0.22Rep
−0.6 + 0.06exp(−17.77Rep)−0.6
uf
ws
c
=
21.2Rep
−1.2
(1 < Rep < 27.3)
0.4
(27.3 ≤Rep) (68) (69) The coefficients αb and αs in Eq. (57) take the following values: αb =
0
(θ0 < θch)
1
(θch ≤θ0)
(70)
220
αs =
0
(uf/ws < (uf/ws)c)
1
((uf/ws)c ≤uf/ws)
(71) αb =
0
(θ0 < θch)
1
(θch ≤θ0)
220
αs =
0
(uf/ws < (uf/ws)c)
1
((uf/ws)c ≤uf/ws) αb =
0
(θ0 < θch)
1
(θch ≤θ0)
(70)
220
αs =
0
(uf/ws < (uf/ws)c)
1
((uf/ws)c ≤uf/ws)
(71) αb =
0
(θ0 < θch)
1
(θch ≤θ0)
(70)
220
αs =
0
(uf/ws < (uf/ws)c)
1
((uf/ws)c ≤uf/ws)
(71) (70) (71) In the case without suspension, the coefficient αs is set to 0. In the case without suspension, the coefficient αs is set to 0. In the case without suspension, the coefficient αs is set to 0. In the case without suspension, the coefficient αs is set to 0. In the above equations, the system is nondimensionalized as follows:
5 where ηb is the dimensionless thickness of the bed-load layer and is obtained as: where ηb is the dimensionless thickness of the bed-load layer and is obtained as: where ηb is the dimensionless thickness of the bed-load layer and is obtained as: ηb = B0 −R0 = hb + D
12 ηb = B0 −R0 = hb + D
12
(65) ηb = B0 −R0 = hb + D
12
(65) (65) where B0 and R0 denote the height of the top of the bed-load layer and the reference level in the basic state, respectively. 210
According to Colombini (2004), the thickness of the bed-load layer hb is estimated as follows: hb = lbD
(66)
lb = 1 + 1.3
τr −τc
τc
0.55
(67) (66) hb = lbD
lb = 1 + 1.3
τr −τc
τc
0.5 lb = 1 + 1.3
τr −τc
τc
0 (67) where lb denotes the relative saltation height, τr is the shear stress at the reference level, and τc is where lb denotes the relative saltation height, τr is the shear stress at the reference level, and τc is the critical shear stress. In this study, the sediment transport regimes are classified using the threshold conditions of sediment motion and the initia-
15
tion of suspension proposed in refs. (Brownlie, 1981; Niño et al., 2003), respectively, as follows: where lb denotes the relative saltation height, τr is the shear stress at the reference level, and τc is the critical shear stress. In this study, the sediment transport regimes are classified using the threshold conditions of sediment motion and the initia-
15 where lb denotes the relative saltation height, τr is the shear stress at the reference level, and τc is the critical shear stress. where lb denotes the relative saltation height, τr is the shear stress at the reference level, and τc is the critical shear stress. where lb denotes the relative saltation height, τr is the shear stress at the reference level, and τc is the critical shear stress. In this study, the sediment transport regimes are classified using the threshold conditions of sediment motion and the initia-
5
tion of suspension proposed in refs. (Brownlie, 1981; Niño et al., 2003), respectively, as follows: In this study, the sediment transport regimes are classified using the threshold conditions of sediment motion and the initia-
215
tion of suspension proposed in refs. In the above equations, the system is nondimensionalized as follows:
5 Equation (57) is nondimensionalized as: ∂B
∂t + αb
∂qB
∂ξ + αs
ws
D
Es −c[ξ,ηb]
= 0 (58) with
195 with
195 ˜t = (1 −λp)˜h0
2
q
Rs˜g ˜D3
t ˜t = (1 −λp)˜h0
2
q
Rs˜g ˜D3
t
(59) (59) In this study, dimensionless bed-load discharge per unit width is estimated using the Meyer-Peter and Müller formula mod-
ified as described in Wong and Parker (2006); this equation is given as: In this study, dimensionless bed-load discharge per unit width is estimated using the Meyer-Peter and Müller formula mod-
ified as described in Wong and Parker (2006); this equation is given as: qB =
˜qB
q
Rsg ˜D3
= 3.97(θb −θc)3/2
(60) θc)3/2
(60) qB =
˜qB
q
Rsg ˜D3
= 3.97(θb −θc)3/2
(60) (60) where θb is the Shields stress at the top of bed-load layer and θc is the critical Shields stress for particle motion. These variables
0
can be expressed as follows: θ0 =
S
RsD
(61)
θb = θ0τb
(62)
θc = θch −µ
S −∂B
∂x
(63) θ0 =
S
RsD
θb = θ0τb
θc = θch −µ
S −∂B
∂x
(61) (62) (63) where θ0 is the Shields stress of the base flow, τb denotes the shear stress at the top of the bed-load layer, θch denotes the critical
205
Shields stress under the flat-bed conditions, and µ is a constant set to 0.1 (Fredsøe, 1974). The shear stress τb is described as: where θ0 is the Shields stress of the base flow, τb denotes the shear stress at the top of the bed-load layer, θch denotes the critical
205
Shields stress under the flat-bed conditions, and µ is a constant set to 0.1 (Fredsøe, 1974). The shear stress τb is described as: τb = [etb · T · enb] η =ηb (64) τb = [etb · T · enb] η =ηb 8 8 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. 2.1.2
Linear Analysis We impose an infinitesimal perturbation on the basic state. All the variables are modified using a small amplification A and a
complex angular frequency of the perturbation ω as follows:
225 (ψ,p,h,Z,R,B,c) = (ψ0,p0,1,0,R0,B0,c0) (ψ,p,h,Z,R,B,c) = (ψ0,p0,1,0,R0,B0,c0)
+ A(ψ1,p1,H1,Z1,R1,B1,c1)exp[i(kξ −ωt)]
(72) = (ψ0,p0,1,0,R0,B0,c0)
+ A(ψ1,p1,H1,Z1,R1,B1,c1)exp[i(kξ −ωt)] 1,p1,H1,Z1,R1,B1,c1)exp[i(kξ −ωt)]
(72) (72) + A(ψ1,p1,H1,Z1,R1,B1,c1)exp[i(kξ −ωt)] The subscript 1 denotes a variable at O(A). By substituting Eq. (72) into the governing equations and boundary conditions, we
can obtain the following equations at O(A): The subscript 1 denotes a variable at O(A). By substituting Eq. (72) into the governing equations and boundary conditions, we
can obtain the following equations at O(A): η)R1 = 0
(73) Lψ(η)ψ1(η) + Lh(η)H1 + LR(η)R1 = 0
ikp1(η) + Pψ(η)ψ1(η) + Ph(η)H1 + PR(η)R1 = 0
230 η)H1 + PR(η)R1 = 0
(74) (74) 9 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. Here, Lφ and Pφ (φ = ψ,h,R) are linear operators. The specific forms of Lφ and Pφ are skipped herein. With the use of the
boundary conditions (Eqs. (32) and (33)), we get: ψ1(1) = 0
(75)
p1(1) = 0
(76)
ψ1(0) = 0
(77)
235
∂ψ1
∂η
η =0
= 0
(78) (75) ψ1(1) = 0
p1(1) = 0
ψ1(0) = 0
235
∂ψ1
∂η
η =0
= 0 (76) 235 235 (77) (78) Additionally, Eqs. (74) and (76) give:
Pψ(1)ψ1(1) + Ph(1)H1 + PR(1)R1 = 0
(79) Additionally, Eqs. (74) and (76) give:
Pψ(1)ψ1(1) + Ph(1)H1 + PR(1)R1 = 0
(79) Additionally, Eqs. (74) and (76) give:
Pψ(1)ψ1(1) + Ph(1)H1 + PR(1)R1 = 0 Additionally, Eqs. (74) and (76) give: Additionally, Eqs. (74) and (76) give:
Pψ(1)ψ1(1) + Ph(1)H1 + PR(1)R1 = 0
(79) PR(1)R1 = 0
(79) Pψ(1)ψ1(1) + Ph(1)H1 + PR(1)R1 = 0 (79) We employ a spectral collocation method using Chebyshev polynomials to solve the above differential equations. We expand
ψ1 using the Chebyshev polynomials as follows:
40 We employ a spectral collocation method using Chebyshev polynomials to solve the above differential equations. 2.1.2
Linear Analysis ˇLψ T0(ζN−2)
···
ˇLψ TN(ζN−2)
ˇLh
(84)
a = (a0,a1,...,aN,D1)
(85)
M =
0,0,0, ˇPR, ˇLh,..., ˇLh
(86)
255
https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. with
L =
T0(−1)
···
TN(−1)
0
ˇ⟨T0(−1)
···
ˇ⟨TN(−1)
0
T0(1)
···
TN(1)
0
ˇPψ T0(1)
···
ˇPψ TN(1)
ˇPh
ˇLψ T0(ζ2)
···
ˇLψ TN(ζ2)
ˇLh
... ... ... ... ˇLψ T0(ζN−2)
···
ˇLψ TN(ζN−2)
ˇLh
(84)
a = (a0,a1,...,aN,D1)
(85)
M =
0,0,0, ˇPR, ˇLh,..., ˇLh
(86)
255 (84) (85) (86) where a check mark (ˇ) denotes a linear operator associated with variable transformation from η to ζ. We obtain the following
solution from Eq. (83): (87) Additionally, Eqs. (80) and (87) give: ψ1 = ψ∗
1(η)R1
(88)
60
H1 = H∗
1R1
(89) ψ1 = ψ∗
1(η)R1
(88)
60
H1 = H∗
1R1
(89) ψ1 = ψ∗
1(η)R1
260
H1 = H∗
1R1 ψ1 = ψ∗
1(η)R1
260 (88) (89) Similarly, we solve the eigenvalue problems for the sediment transport equations. By substituting Eq. (72) into Eq. (21), we
obtain the following equations at the order of O(A): Ccc1(η) + Cψ(η)ψ1(η) + CHH1 + CRR1 = 0 (90) 2.1.2
Linear Analysis We expand
ψ1 using the Chebyshev polynomials as follows:
240 ψ1 =
N
X
n=0
anTn(ζ)
(80) ψ1 =
N
X
n=0
anTn(ζ)
(80) =
N
X
n=0
anTn(ζ) (80) where an is the coefficient for the n-th order Chebyshev polynomial Tn and ζ is the independent variable of the Chebyshev
polynomials defined in the domain [−1,1]. In this study, we transform ζ using the following equation to improve the calculation
accuracy: where an is the coefficient for the n-th order Chebyshev polynomial Tn and ζ is the independent variable of the Chebyshev
polynomials defined in the domain [−1,1]. In this study, we transform ζ using the following equation to improve the calculation
accuracy: where an is the coefficient for the n-th order Chebyshev polynomial Tn and ζ is the independent variable of the Chebyshev
polynomials defined in the domain [−1,1]. In this study, we transform ζ using the following equation to improve the calculation
accuracy: ζ = 2
ln[(η + R0)/R0]
ln[(1 + R0)/R0]
−1
245 ζ = 2
ln[(η + R0)/R0]
ln[(1 + R0)/R0]
−1
(81)
5 (81) The above functions are substituted into Eq. (73); then, we evaluate the equation at the Gauss-Labatte points, which are defined
as: The above functions are substituted into Eq. (73); then, we evaluate the equation at the Gauss-Labatte points, which are defined
as: as: ζj = cos
jπ
N + 2
,
j = 1,2,...,N + 1 os
jπ
N + 2
,
j = 1,2,...,N + 1 ζj = cos
jπ
N + 2
,
j = 1,2,...,N + 1
(82) (82) By combining the governing equations, boundary conditions, and closure assumptions, we obtain the following system of
linear algebraic equations:
250 By combining the governing equations, boundary conditions, and closure assumptions, we obtain the following system of
linear algebraic equations:
250 La = MR1
(83) La = MR1 (83) 10 with
L =
T0(−1)
···
TN(−1)
0
ˇ⟨T0(−1)
···
ˇ⟨TN(−1)
0
T0(1)
···
TN(1)
0
ˇPψ T0(1)
···
ˇPψ TN(1)
ˇPh
ˇLψ T0(ζ2)
···
ˇLψ TN(ζ2)
ˇLh
... ... ... ... Therefore, the following equation is obtained: c1(η) = c∗
1(η)R1 (100) By substituting Eqs. (88), (89), and (100) into Exner’s equation (Eq. (58)), the complex angular frequency ω is obtained in
the following form: ω = ω(k,Fr,Cz,Rep) = ωr + iωi
285 (101) Here, ωi corresponds to the growth rate of the perturbation. Here, ωi corresponds to the growth rate of the perturbation. Based on Eqs. (88) and (89), we obtain:
265 Based on Eqs. (88) and (89), we obtain:
265 Based on Eqs. (88) and (89), we obtain:
265 Ccc1(η) +
Cψ(η)ψ∗
1(η) + CHH∗
1 + CR
R1 = 0 (91) The boundary conditions give: The boundary conditions give: Scc1(1) +
Sψ(1)ψ∗
1(1) + SHH∗
1 + SR
R1 = 0
Bcc1(0) +
Bψ(0)ψ∗
1(0) + BHH∗
1 + BR
R1 = 0 (92) (93) Here, Cφ, Sφ and Bφ (φ = ψ,h,R,c) are the linear operators. 270
We expand c1 using Chebyshev polynomials as follows: Here, Cφ, Sφ and Bφ (φ = ψ,h,R,c) are the linear operators. 270
We expand c1 using Chebyshev polynomials as follows: Here, Cφ, Sφ and Bφ (φ = ψ,h,R,c) are the linear operators. 270
We expand c1 using Chebyshev polynomials as follows: Here, Cφ, Sφ and Bφ (φ = ψ,h,R,c) are the linear operators. 270
We expand c1 using Chebyshev polynomials as follows: c1 =
N
X
n=0
bnTn(ζ) (94) 11 11 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. The system is evaluated at the Gauss-Labatte points, then we obtain: Kb = NR1 (95) with
275 with
5
K =
ˇBcT0(−1)
···
ˇBcTN(−1)
ˇScT0(1)
···
ˇScTN(−1)
ˇCcT0(ζ1)
···
ˇCcTN(ζ1)
... ... ... ˇCcT0(ζN−1)
···
ˇCcTN(ζN−1)
(96) b = (b0,b1,...,bN)
N = −
ˇBψψ∗
1(−1) + ˇBhH∗
1 + ˇBR
ˇSψψ∗
1(1) + ˇShH∗
1 + ˇSR
ˇCψψ∗
1(ζ1) + ˇChH∗
1 + ˇCR
... ˇCψψ∗
1(ζN−1) + ˇChH∗
1 + ˇCR
(97) (98) The coefficient bn is derived as: b = K−1NR1
280 (99) Therefore, the following equation is obtained: Therefore, the following equation is obtained: 2.2
Governing Parameters It is expected that coarse sediment is less transported in suspension than fine sediment. Therefore, we employed two grades of
fine particles ( ˜D = 0.12 and 0.25 mm) and one grade of coarse particles ( ˜D = 1.2 mm) to investigate the effect of suspension It is expected that coarse sediment is less transported in suspension than fine sediment. Therefore, we employed two grades of
fine particles ( ˜D = 0.12 and 0.25 mm) and one grade of coarse particles ( ˜D = 1.2 mm) to investigate the effect of suspension
on the bed instability. The flow depth and Froude number range from 1 cm to 5.0 m and from 0.01 to 2, respectively. The
315
domain [kmin,kmax] was set as [0.01,1.5], corresponding to λ ranging from ∼4.2h to ∼628h. on the bed instability. The flow depth and Froude number range from 1 cm to 5.0 m and from 0.01 to 2, respectively. The
315
domain [kmin,kmax] was set as [0.01,1.5], corresponding to λ ranging from ∼4.2h to ∼628h. 2.2
Governing Parameters The instability of a system is illustrated as a contour diagram of the perturbation growth rate ωi (Fig. 1). Generally, theoretical
studies of bedforms based on linear stability analyses describe the transition of bedform phases in the parametric space of 12 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. wavenumber k and Froude number Fr, which are given by:
290 290 (102) (103) where ˜λ denotes the perturbation wavelength, ˜U0 is the depth-averaged flow velocity of the uniform flow, ˜g is the gravitational
acceleration (= 9.81 m2/s), and ˜h0 is the flow depth of the uniform flow. where ˜λ denotes the perturbation wavelength, ˜U0 is the depth-averaged flow velocity of the uniform flow, ˜g is the gravitational
acceleration (= 9.81 m2/s), and ˜h0 is the flow depth of the uniform flow. Stability diagrams described on the k-Fr plane have been commonly used to predict the development of dunes and antidunes
295
(Kennedy, 1963). A few studies have used other combinations of dimensionless numbers such as the friction coefficient C
versus Fr (Colombini and Stocchino, 2008) and the relative roughness ˜D/˜h0 on the k-Fr plane (Bohorquez et al., 2019). Although the classic k-Fr diagrams are widely accepted, we cannot use this approach to evaluate whether plane bed forma- Stability diagrams described on the k-Fr plane have been commonly used to predict the development of dunes and antidunes
295
(Kennedy, 1963). A few studies have used other combinations of dimensionless numbers such as the friction coefficient C
versus Fr (Colombini and Stocchino, 2008) and the relative roughness ˜D/˜h0 on the k-Fr plane (Bohorquez et al., 2019). Although the classic k-Fr diagrams are widely accepted, we cannot use this approach to evaluate whether plane bed forma-
tion can be predicted reliably because plane beds have extremely small wavenumber or have infinite wavelength (i.e., they are Although the classic k-Fr diagrams are widely accepted, we cannot use this approach to evaluate whether plane bed forma-
tion can be predicted reliably because plane beds have extremely small wavenumber or have infinite wavelength (i.e., they are flat). Therefore, we illustrate stability diagrams as contour maps of dominant wavenumber kd on the ˜h0-Fr plane with fixed ˜D
300
to investigate the impact of suspended load on the formation of plane beds, where the dominant wavenumber kd denotes the
wavenumber that provides maximum growth rate. 2.2
Governing Parameters ω = ω
k,Fr, ˜D,˜h0
(104) Thus, we can obtain the growth rate ωi as a function of k for a given combination of
Fr, ˜D,˜h0
. In this study, we assume
305
that the system is stable if ωi is not positive for all k within the domain [kmin, kmax] for a given
Fr, ˜D,˜h0
combination. In
contrast, the system is assumed to be unstable if ωi is positive for some k (Fig. 1). We describe stability diagrams as contour
maps of kd in the parametric space of
Fr, ˜D,˜h0
(Figs. 2 and 3) . Thus, we can obtain the growth rate ωi as a function of k for a given combination of
Fr, ˜D,˜h0
. In this study, we assume
305
that the system is stable if ωi is not positive for all k within the domain [kmin, kmax] for a given
Fr, ˜D,˜h0
combination. In
contrast, the system is assumed to be unstable if ωi is positive for some k (Fig. 1). We describe stability diagrams as contour
maps of kd in the parametric space of
Fr, ˜D,˜h0
(Figs. 2 and 3) .
We performed linear stability analyses of bedforms under open-channel flows with and witho We performed linear stability analyses of bedforms under open-channel flows with and without suspension. In the case with-
out suspension, the development of the bed configuration associated with suspended load is ignored by setting the coefficient
310
αs in Eq. (57) to 0. In the case with suspension, the coefficient αs take a value of 0 or 1 depending on the sediment transport
regime (Eq. (71)). It is expected that coarse sediment is less transported in suspension than fine sediment Therefore we employed two grades of out suspension, the development of the bed configuration associated with suspended load is ignored by setting the coefficient
310
αs in Eq. (57) to 0. In the case with suspension, the coefficient αs take a value of 0 or 1 depending on the sediment transport
regime (Eq. (71)). It is expected that coarse sediment is less transported in suspension than fine sediment. Therefore, we employed two grades of
fine particles ( ˜D = 0.12 and 0.25 mm) and one grade of coarse particles ( ˜D = 1.2 mm) to investigate the effect of suspension regime (Eq. (71)). 2.3
Compilation of published data To calculate the particle Reynolds number, the kinematic viscosity ν was assumed as
follows (van den Berg and van Gelder, 1993): The data of which sediment diameter range from 0.74 ˜D to 1.36 ˜D were chosen to plot on stability diagram, which corre-
sponds to the range log10 Rep ± 0.2. To calculate the particle Reynolds number, the kinematic viscosity ν was assumed as
follows (van den Berg and van Gelder, 1993): ν =
h
1.14 −0.031(T −15) + 0.00068(T −15)2i
10−6
(105)
30 ν =
h
1.14 −0.031(T −15) + 0.00068(T −15)2i
10−6
(105)
330 ν =
h
1.14 −0.031(T −15) + 0.00068(T −15)2i
10−6
330 (105) where T represents the water temperature in degrees Celsius. A value of 20◦C was assumed for data when T was not reported. where T represents the water temperature in degrees Celsius. A value of 20◦C was assumed for data when T was not reported. 3
Results The contour maps of ˜h0 versus Fr show that the stable region for fine sediments is wider in the diagram with suspension than
in that without suspension (Fig. 2). A stable region appears at 0.6 < Fr < 1.2 and for h < 1.2 m in the case without suspension
(Fig. 2a, c), and the dominant wavenumber increases with increasing flow depth. In the case with suspension, Froude number
335
and flow depth of the stable region ranges from 0.15 to 1.0 and from 0.01 to 5, respectively (Fig. 2b, d), while at Fr > 1, the
dominant wavenumber can fall below 0.3 (Fig. 2b, d). The contour maps of ˜h0 versus Fr show that the stable region for fine sediments is wider in the diagram with suspension than
in that without suspension (Fig. 2). A stable region appears at 0.6 < Fr < 1.2 and for h < 1.2 m in the case without suspension p
(
g
)
g
pp
p
(Fig. 2a, c), and the dominant wavenumber increases with increasing flow depth. In the case with suspension, Froude number
335
and flow depth of the stable region ranges from 0.15 to 1.0 and from 0.01 to 5, respectively (Fig. 2b, d), while at Fr > 1, the
dominant wavenumber can fall below 0.3 (Fig. 2b, d). (Fig. 2a, c), and the dominant wavenumber increases with increasing flow depth. In the case with suspension, Froude number
335
and flow depth of the stable region ranges from 0.15 to 1.0 and from 0.01 to 5, respectively (Fig. 2b, d), while at Fr > 1, the
dominant wavenumber can fall below 0.3 (Fig. 2b, d). (Fig. 2a, c), and the dominant wavenumber increases with increasing flow depth. In the case with suspension, Froude number
335
and flow depth of the stable region ranges from 0.15 to 1.0 and from 0.01 to 5, respectively (Fig. 2b, d), while at Fr > 1, the
dominant wavenumber can fall below 0.3 (Fig. 2b, d). Comparing the results where ˜D = 0.12 mm and the observational data, in the case without suspension, all the plane bed data
are within unstable region; most values plot in the region where kd > 1 (Fig. 2a). 2.3
Compilation of published data The stability diagrams were assessed using an observational dataset pertaining to open-channel flows compiled from the litera-
ture, as summarized in Tables A1–A3. We compiled from the literature a total of 369 sets of data. The dataset consisted of 286 13 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. sets of laboratory data and 83 sets of field data. The flow depth ranges from 0.02 to 19.5 m, and the flow velocity ranges from
320
0.2 to 2 m/s. sets of laboratory data and 83 sets of field data. The flow depth ranges from 0.02 to 19.5 m, and the flow velocity ranges from
320
0.2 to 2 m/s. We used the data of plane beds in which the sediment transport mode could be identified, i.e., plane bed without suspension,
with suspension, and with sheet flows. We identified whether sediment particles were transported as suspended load or not
based on the suspended sediment concentration. Plane bed without sediment movement were not included in this analysis. For sets of laboratory data and 83 sets of field data. The flow depth ranges from 0.02 to 19.5 m, and the flow velocity ranges from
320
0.2 to 2 m/s. We used the data of plane beds in which the sediment transport mode could be identified, i.e., plane bed without suspension,
with suspension, and with sheet flows. We identified whether sediment particles were transported as suspended load or not
based on the suspended sediment concentration. Plane bed without sediment movement were not included in this analysis. For comparison with the theoretical analysis results, we used the data of dunes and antidunes with wavenumbers with the range
325
0 < k ≤1.5 for comparison. The data of which sediment diameter range from 0.74 ˜D to 1.36 ˜D were chosen to plot on stability diagram, which corre-
sponds to the range log10 Rep ± 0.2. To calculate the particle Reynolds number, the kinematic viscosity ν was assumed as
follows (van den Berg and van Gelder, 1993): comparison with the theoretical analysis results, we used the data of dunes and antidunes with wavenumbers with the range
325
0 < k ≤1.5 for comparison. The data of which sediment diameter range from 0.74 ˜D to 1.36 ˜D were chosen to plot on stability diagram, which corre-
sponds to the range log10 Rep ± 0.2. 4.1
Effect of suspension on fine sediment bed Previous research has suggested that the formation of dunes is suppressed
due to the insufficient time for dune development (Walker, 1965), the hysteresis effect under waning flow conditions (Endo and
Masuda, 1997), the turbulence suppression by high suspended-sediment concentrations (Lowe, 1988), the lack of a sharp near- bed density gradient (Arnott, 2012), and the effect of clay-sized sediment on bed rheology (Schindler et al., 2015). Although
375
these interpretations could explain the absence of dune-scale cross-lamination in turbidites, We show that dune formation is
suppressed without considering above conditions. Therefore, the above conditions are not required to suppress dune formation
(Fig. 2b, d). Instead, we propose that the development of dune-scale bed waves under turbidity currents is restricted by the
presence of suspended load. bed density gradient (Arnott, 2012), and the effect of clay-sized sediment on bed rheology (Schindler et al., 2015). Although
375
these interpretations could explain the absence of dune-scale cross-lamination in turbidites, We show that dune formation is
suppressed without considering above conditions. Therefore, the above conditions are not required to suppress dune formation
(Fig. 2b, d). Instead, we propose that the development of dune-scale bed waves under turbidity currents is restricted by the
presence of suspended load. 3
Results In contrast, all the plane bed data plot in the Comparing the results where ˜D = 0.12 mm and the observational data, in the case without suspension, all the plane bed data
are within unstable region; most values plot in the region where kd > 1 (Fig. 2a). In contrast, all the plane bed data plot in the stable region in the case with suspension (Fig. 2b). When ˜D = 0.25 mm, the plane bed data without suspension plot below the
340
threshold of sediment motion (Fig. 2c, d). Moreover, although some observational data points of plane beds with suspension
plot within the stable region in both diagrams, more data agree with the stable region in the case with suspension than in that
without suspension (Fig. 2c, d). As expected, most dune and antidune data plot in the unstable region, whereas several data
points of dunes and antidunes plot in the stable region in both cases with and without suspension (Fig. 2). The stability diagrams considering flows with and without suspension for coarse sediment beds ( ˜D = 1.20 mm) do not differ
345
much (Fig. 3). The difference of dominant wavenumbers can be found just above the threshold of initiation of suspension (Fig. 3). The data of plane beds without suspended load plot around the threshold of sediment motion and in the stable region below
the line defined by the threshold of suspension (Fig. 3). The observational data of plane beds under sheet flows fall inside the
unstable region where 0.3 < kd < 0.5 and Fr > 1.6 (Fig. 3). 345 14 14 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. 4.1
Effect of suspension on fine sediment bed The role of suspended load in the formation of plane beds and suppressing dune-scale instabilities is quantitatively illustrated
as the broadening of the stable regions (Fig. 2). The stability diagrams for fine sediment beds show a good agreement with
the observational data of plane beds under flows with suspension (Fig. 2b, d). The transition from dunes to plane beds has been explained by the spatial lag δ between the bed topography and the local sediment transport rate (Naqshband et al.,
355
2014; van Duin et al., 2017). If the bed topography and sediment transport rate are entirely in-phase (δ = 0), dunes migrate
downstream without growth or decay. The dune height increases and decreases when the maximum sediment transport rate
occurs upstream (δ < 0) and downstream (δ > 0) of the dune crest, respectively. Kennedy (1963) introduced the spatial lag in
his flow model to account for the bedform growth and decay, and subsequent research has investigated the effect of spatial lag been explained by the spatial lag δ between the bed topography and the local sediment transport rate (Naqshband et al.,
355
2014; van Duin et al., 2017). If the bed topography and sediment transport rate are entirely in-phase (δ = 0), dunes migrate
downstream without growth or decay. The dune height increases and decreases when the maximum sediment transport rate
occurs upstream (δ < 0) and downstream (δ > 0) of the dune crest, respectively. Kennedy (1963) introduced the spatial lag in
his flow model to account for the bedform growth and decay, and subsequent research has investigated the effect of spatial lag on the bedform development (McLean, 1990; van Duin et al., 2017). Recently, Naqshband et al. (2017) quantitatively observed
360
the positive spatial lag under suspended load dominated flows in their flume experiments. Our analyses confirm that suspended
load dampens the development of bed waves, thereby facilitating the formation of plane beds, and thus cannot be neglected in
theoretical analyses for realistic predictions of bedforms. We found that dunes are deformed under flows with suspended load, although further work is needed to investigate the We found that dunes are deformed under flows with suspended load, although further work is needed to investigate the
amplitudes of dunes under such conditions. 4.1
Effect of suspension on fine sediment bed Field surveys have indicated the existence of low-angle dunes in suspended-
365
load dominated rivers (Smith and McLean, 1977; Kostaschuk and Villard, 1996; Hendershot et al., 2016); moreover, flume
experiments have indicated that dune height decreases with increasing suspended load flux (Naqshband et al., 2017; Bradley
and Venditti, 2019). Theoretical analyses in Fredsøe (1981) have also reasonably predicted a decrease of dune steepness under
unsteady flows with suspension where the flow discharges were being increased. In the future work, nonlinear analysis should
be done to obtain the amplitudes of dunes under flows with suspended load
370 amplitudes of dunes under such conditions. Field surveys have indicated the existence of low-angle dunes in suspended-
365
load dominated rivers (Smith and McLean, 1977; Kostaschuk and Villard, 1996; Hendershot et al., 2016); moreover, flume
experiments have indicated that dune height decreases with increasing suspended load flux (Naqshband et al., 2017; Bradley
and Venditti, 2019). Theoretical analyses in Fredsøe (1981) have also reasonably predicted a decrease of dune steepness under
unsteady flows with suspension where the flow discharges were being increased. In the future work, nonlinear analysis should be done to obtain the amplitudes of dunes under flows with suspended load. 370
Ultimately, our linear analyses provide a simple explanation for the absence of dunes in turbidites: suspended load suppresses
dune formation and facilitates plane-bed formation. Previous research has suggested that the formation of dunes is suppressed
due to the insufficient time for dune development (Walker, 1965), the hysteresis effect under waning flow conditions (Endo and
Masuda, 1997), the turbulence suppression by high suspended-sediment concentrations (Lowe, 1988), the lack of a sharp near- be done to obtain the amplitudes of dunes under flows with suspended load. 370
Ultimately, our linear analyses provide a simple explanation for the absence of dunes in turbidites: suspended load suppresses
dune formation and facilitates plane-bed formation. Previous research has suggested that the formation of dunes is suppressed
due to the insufficient time for dune development (Walker, 1965), the hysteresis effect under waning flow conditions (Endo and
Masuda, 1997), the turbulence suppression by high suspended-sediment concentrations (Lowe, 1988), the lack of a sharp near- be done to obtain the amplitudes of dunes under flows with suspended load. 370
Ultimately, our linear analyses provide a simple explanation for the absence of dunes in turbidites: suspended load suppresses
dune formation and facilitates plane-bed formation. 4.2
Effect of suspension on coarse sediment bed
380 Further, linear analyses considering sheet flows can be extended to analyses of debris flows and turbidity currents that have
390
collisional layers (Sohn, 1997; Lanzoni et al., 2017). These topics can be further explored in future works. Further, linear analyses considering sheet flows can be extended to analyses of debris flows and turbidity currents that have
390
collisional layers (Sohn, 1997; Lanzoni et al., 2017). These topics can be further explored in future works. 4.2
Effect of suspension on coarse sediment bed
380 In the diagram with ˜D = 1.20 mm, the data with sheet flows plotted much above the upper limit of Fr for the stable region
(Fig. 3). Sheet flows consist of a shear layer of bed-load that moves under high shear stress (Shields number is larger than 0.5) 15 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. (Gao, 2008). A few past experimental studies have observed that plane bed develops beneath sheet flows on coarse sediment
beds in open-channel flows (Williams, 1970; Hernandez-Moreira et al., 2020). The difference in hydraulic properties between (
,
)
p
p
p
p
beds in open-channel flows (Williams, 1970; Hernandez-Moreira et al., 2020). The difference in hydraulic properties between
standard bed-load and sheet flows could result in the disagreement between the stability diagrams and observational data. For
385
example, the vertical velocity profile of an open-channel flow takes a logarithmic form (Keulegan, 1938), whereas that of sheet
flows takes a power form (Sumer et al., 1996) or can be obtained by solving the differential equations (Egashira, 1997). In
addition, pressures of static interparticle contacts and inelastic particle collisions are not negligible in sheet flows (Egashira,
1997). Considering these differences in hydraulic conditions, the stability fields of perturbations are affected by sheet flows. standard bed-load and sheet flows could result in the disagreement between the stability diagrams and observational data. For
385
example, the vertical velocity profile of an open-channel flow takes a logarithmic form (Keulegan, 1938), whereas that of sheet
flows takes a power form (Sumer et al., 1996) or can be obtained by solving the differential equations (Egashira, 1997). In
addition, pressures of static interparticle contacts and inelastic particle collisions are not negligible in sheet flows (Egashira,
1997). Considering these differences in hydraulic conditions, the stability fields of perturbations are affected by sheet flows. standard bed-load and sheet flows could result in the disagreement between the stability diagrams and observational data. For
385
example, the vertical velocity profile of an open-channel flow takes a logarithmic form (Keulegan, 1938), whereas that of sheet
flows takes a power form (Sumer et al., 1996) or can be obtained by solving the differential equations (Egashira, 1997). In
addition, pressures of static interparticle contacts and inelastic particle collisions are not negligible in sheet flows (Egashira,
1997). Considering these differences in hydraulic conditions, the stability fields of perturbations are affected by sheet flows. 5
Conclusions We investigated the influence of suspended load on the formation of plane beds under open-channel flows. The stability dia-
grams show that the stable region for finer sediments is wider in the diagram with suspension than that without suspension. Further, the published data of plane beds with suspension coincide well with the stability diagrams where the suspension was
395
considered. Our theoretical analysis found that suspended load promotes the formation of plane beds and suppresses the for-
mation of dunes on the fine-grained bed. These results suggest that dune-scale cross lamination is absent in turbidites because We investigated the influence of suspended load on the formation of plane beds under open-channel flows. The stability dia-
grams show that the stable region for finer sediments is wider in the diagram with suspension than that without suspension. We investigated the influence of suspended load on the formation of plane beds under open-channel flows. The stability dia-
grams show that the stable region for finer sediments is wider in the diagram with suspension than that without suspension. Further, the published data of plane beds with suspension coincide well with the stability diagrams where the suspension was
395
considered. Our theoretical analysis found that suspended load promotes the formation of plane beds and suppresses the for-
mation of dunes on the fine-grained bed. These results suggest that dune-scale cross lamination is absent in turbidites because
the development of dunes in turbidity currents is restricted by the presence of suspended load. In addition, our analysis displays
that the data pertaining to sheet flows deviate from the stable region. Additional theoretical work is required in order to examine
whether the plane bed under sheet flow can be interpreted as a stable condition or not. 400 Further, the published data of plane beds with suspension coincide well with the stability diagrams where the suspension was
395
considered. Our theoretical analysis found that suspended load promotes the formation of plane beds and suppresses the for-
mation of dunes on the fine-grained bed. These results suggest that dune-scale cross lamination is absent in turbidites because
the development of dunes in turbidity currents is restricted by the presence of suspended load. In addition, our analysis displays
that the data pertaining to sheet flows deviate from the stable region. References Abdel-Fattah, S., Amin, A., and Van Rijn, L. C.: Sand Transport in Nile River, Egypt, Journal of Hydraulic Engineering, 130, 488–500,
https://doi.org/10.1061/(ASCE)0733-9429(2004)130:6(488), 2004. 410 https://doi.org/10.1061/(ASCE)0733-9429(2004)130:6(488), 2004. 10
Arnott,
R. W. C.:
Turbidites,
and
the
Case
of
the
Missing
Dunes,
Journal
of
Sedimentary
Research,
82,
379–384,
https://doi.org/10.2110/jsr.2012.29, 2012. Baird, D. C.: Field adjustments of bed form phase diagrams, in: 2nd Joint Federal Interagency Conference, pp. 1137–1145, Las Vegas, NV,
2010. Arnott,
R. W. C.:
Turbidites,
and
the
Case
of
the
Missing
Dunes,
Journal
of
Sedimentary
Research,
82,
379–384,
https://doi.org/10.2110/jsr.2012.29, 2012. Baird, D. C.: Field adjustments of bed form phase diagrams, in: 2nd Joint Federal Interagency Conference, pp. 1137–1145, Las Vegas, NV,
2010. Blom, A., Ribberink, J. S., and de Vriend, H. J.: Vertical sorting in bed forms: Flume experiments with a natural and a trimodal sediment
415
mixture, Water Resources Research, 39, https://doi.org/10.1029/2001WR001088, 1025, 2003. B h
P C ñ d P
i
P Ji
R i
P J
d d l M
l E
i
J D
Th
f
i
i
f
h ll
h
f
h Blom, A., Ribberink, J. S., and de Vriend, H. J.: Vertical sorting in bed forms: Flume experiments with a natural and a trimodal sediment
415
mixture, Water Resources Research, 39, https://doi.org/10.1029/2001WR001088, 1025, 2003. Bohorquez, P., Cañada-Pereira, P., Jimenez-Ruiz, P. J., and del Moral-Erencia, J. D.: The fascination of a shallow-water theory for the
formation of megaflood-scale dunes and antidunes, Earth-Science Reviews, 193, 91–108, https://doi.org/10.1016/j.earscirev.2019.03.021,
2019. Bohorquez, P., Cañada-Pereira, P., Jimenez-Ruiz, P. J., and del Moral-Erencia, J. D.: The fascination of a shallow-water theory for the
formation of megaflood-scale dunes and antidunes, Earth-Science Reviews, 193, 91–108, https://doi.org/10.1016/j.earscirev.2019.03.021,
2019. Bose, S. K. and Dey, S.: Reynolds averaged theory of turbulent shear flows over undulating beds and formation of sand waves, Physical
420
Review E, 80, 036 304, https://doi.org/10.1103/PhysRevE.80.036304, 2009. Bouma, A. H.: Sedimentology of some flysch deposits: A graphic approach to facies interpretation, Elsevier Scientific Publishing Company,
Amsterdam, 1962. Bouma, A. H.: Sedimentology of some flysch deposits: A graphic approach to facies interpretation, Elsevier Scientific Publishing Company,
Amsterdam, 1962. Bourke, M. C., Lancaster, N., Fenton, L. K., Parteli, E. J. R., Zimbelman, J. R., and Radebaugh, J.: Extraterrestrial dunes: An introduction to
the special issue on planetary dune systems, Geomorphology, 121, 1–14, https://doi.org/10.1016/j.geomorph.2010.04.007, 2010. 425 Bourke, M. C., Lancaster, N., Fenton, L. K., Parteli, E. J. R., Zimbelman, J. 5
Conclusions Additional theoretical work is required in order to examine
whether the plane bed under sheet flow can be interpreted as a stable condition or not. 400 Code and data availability. The datasets and codes used for this study can be found at [url to be updated at acceptance]. Unpublished data
used for the analysis were cited from the dataset of Brownlie (2018). Code and data availability. The datasets and codes used for this study can be found at [url to be updated at acceptance]. Unpublished data
used for the analysis were cited from the dataset of Brownlie (2018). Author contributions. KO and NI performed the linear stability analysis. HN and NI contributed to the interpretation of the results. KO wrote
the manuscript and prepared the figures, and then HN and NI provided feedback on the manuscript and figures. Author contributions. KO and NI performed the linear stability analysis. HN and NI contributed to the interpretation of the results. KO wrote
the manuscript and prepared the figures, and then HN and NI provided feedback on the manuscript and figures. Competing interests. The authors declare no competing interests. 405 Acknowledgements. This work was supported by the Japan Society for the Promotion of Science (JSPS) Grant-in-Aid (KAKENHI) Grant
Number 18J22211. We would like to express our gratitude to Robert Dorrell for his comments. Acknowledgements. This work was supported by the Japan Society for the Promotion of Science (JSPS) Grant-in-Aid (KAKENHI) Grant
Number 18J22211. We would like to express our gratitude to Robert Dorrell for his comments. 16 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. References 562-J, Professional Paper, Washington, DC, 1972. 5 445 de Almeida, R. P., Galeazzi, C. P., Freitas, B. T., Janikian, L., Ianniruberto, M., and Marconato, A.: Large barchanoid dunes in the Ama-
zon River and the rock record: Implications for interpreting large river systems, Earth and Planetary Science Letters, 454, 92–102,
https://doi.org/10.1016/j.epsl.2016.08.029, 2016. de Leeuw, J., Lamb, M. P., Parker, G., Moodie, A. J., Haught, D., Venditti, J. G., and Nittrouer, J. A.: Entrainment and suspension of sand
d
l E
h S
f
D
i
8 485 504 h
//d i
/10 5194/
f 8 485 2020 2020
450 de Leeuw, J., Lamb, M. P., Parker, G., Moodie, A. J., Haught, D., Venditti, J. G., and Nittrouer, J. A.: Entrainment and suspension of sand
and gravel, Earth Surface Dynamics, 8, 485–504, https://doi.org/10.5194/esurf-8-485-2020, 2020. 450 and gravel, Earth Surface Dynamics, 8, 485–504, https://doi.org/10.5194/esurf-8-485-2020, 2020. 450
Di Cristo, C., Iervolino, M., and Vacca, A.: Linear stability analysis of a 1-D model with dynamical description of bed-load transport, Journal
of Hydraulic Research, 44, 480–487, https://doi.org/10.1080/00221686.2006.9521699, 2006. Egashira, S.: Mechanics of sediment transport and sediment laden flows: 1. Mechanics of debris flows, Japanese journal of multiphase flow,
11, 151–156, https://doi.org/10.3811/jjmf.11.151, 1997. Di Cristo, C., Iervolino, M., and Vacca, A.: Linear stability analysis of a 1-D model with dynamical description of bed-load transport, Journal
of Hydraulic Research, 44, 480–487, https://doi.org/10.1080/00221686.2006.9521699, 2006. Egashira, S.: Mechanics of sediment transport and sediment laden flows: 1. Mechanics of debris flows, Japanese journal of multiphase flow, Di Cristo, C., Iervolino, M., and Vacca, A.: Linear stability analysis of a 1-D model with dynamical description of bed-load transport, Journal
of Hydraulic Research, 44, 480–487, https://doi.org/10.1080/00221686.2006.9521699, 2006. Di Cristo, C., Iervolino, M., and Vacca, A.: Linear stability analysis of a 1 D model with dynamical description of bed load transport, Journal
of Hydraulic Research, 44, 480–487, https://doi.org/10.1080/00221686.2006.9521699, 2006. Egashira, S.: Mechanics of sediment transport and sediment laden flows: 1. Mechanics of debris flows, Japanese journal of multiphase flow,
11, 151–156, https://doi.org/10.3811/jjmf.11.151, 1997. Egashira, S.: Mechanics of sediment transport and sediment laden flows: 1. Mechanics of debris flows, Japanese journal of multiphase flow,
11, 151–156, https://doi.org/10.3811/jjmf.11.151, 1997. Endo, N. and Masuda, F.: Small ripples in dunes regime and interpretation about dunes-missing in turbidite, The Journal of the Geological
455
Society of Japan, 103, 741–746, https://doi.org/10.5575/geosoc.103.741, 1997. Endo, N. References R., and Radebaugh, J.: Extrat the special issue on planetary dune systems, Geomorphology, 121, 1–14, https://doi.org/10.1016/j.geomorph.2010.04.007, 2010. 425
Bradley, R., Venditti, J. G., Kostaschuk, R. A., Church, M., Hendershot, M., and Allison, M. A.: Flow and sediment suspen-
sion events over low-angle dunes: Fraser Estuary, Canada, Journal of Geophysical Research: Earth Surface, 118, 1693–1709,
https://doi.org/10.1002/jgrf.20118, 2013. Bradley, R. W. and Venditti, J. G.: Transport scaling of dune dimensions in shallow flows, Journal of Geophysical Research: Earth Surface,
124, 526–547, https://doi.org/10.1029/2018JF004832, 2019. 430 124, 526–547, https://doi.org/10.1029/2018JF004832, 2019. 430
Bridge, J. S. and Best, J. L.: Flow, sediment transport and bedform dynamics over the transition from dunes to upper-stage plane beds:
implications for the formation of planar laminae, Sedimentology, 35, 753–763, https://doi.org/10.1111/j.1365-3091.1988.tb01249.x, 1988. Brownlie, W. R.: Prediction of flow depth and sediment discharge in open channels, Tech. Rep. KH-R43A, W. M. Keck Laboratory of
Hydraulics and Water Resources, California Institute of Technology, Pasadena, California, https://doi.org/10.7907/Z9KP803R, 1981. Brownlie,
W. R.:
Digitized
dataset
from
"Compilation
of
alluvial
channel
data:
laboratory
and
field"
(Version
1.0),
435
https://doi.org/10.22002/d1.943, caltechDATA, 2018. Cisneros, J., Best, J., van Dijk, T., de Almeida, R. P., Amsler, M., Boldt, J., Freitas, B., Galeazzi, C., Huizinga, R., Ianniruberto, M., Ma, H.,
Nittrouer, J. A., Oberg, K., Orfeo, O., Parsons, D., Szupiany, R., Wang, P., and Zhang, Y.: Dunes in the world’s big rivers are characterized
by low-angle lee-side slopes and a complex shape, Nature Geoscience, 13, 156–162, https://doi.org/10.1038/s41561-019-0511-7, 2020. Brownlie,
W. R.:
Digitized
dataset
from
"Compilation
of
alluvial
channel
data:
laboratory
and
field"
(Version
1.0),
435
https://doi.org/10.22002/d1.943, caltechDATA, 2018. Cisneros, J., Best, J., van Dijk, T., de Almeida, R. P., Amsler, M., Boldt, J., Freitas, B., Galeazzi, C., Huizinga, R., Ianniruberto, M., Ma, H.,
Ni
J A Ob
K O f
O P
D S
i
R W
P
d Zh
Y D
i
h
ld’ bi
i
h
i
d Colombini,
M.:
Revisiting
the
linear
theory
of
sand
dune
formation,
Journal
of
Fluid
Mechanics,
502,
1–16,
440
https://doi.org/10.1017/S0022112003007201, 2004. Colombini,
M. and
Stocchino,
A.:
Finite-amplitude
river
dunes,
Journal
of
Fluid
Mechanics,
611,
283–306,
https://doi.org/10.1017/S0022112008002814, 2008. 17 17 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. Culbertson, J. K., Scott, C. H., and Bennett, J. P.: Summary of alluvial-channel data from Rio Grande conveyance channel, New Mexico,
1965-69, Tech. Rep. References and Masuda, F.: Small ripples in dunes regime and interpretation about dunes-missing in turbidite, The Journal of the Geological
455
Society of Japan, 103, 741–746, https://doi.org/10.5575/geosoc.103.741, 1997. Engelund, F.: Instability of erodible beds, Journal of Fluid Mechanics, 42, 225–244, https://doi.org/10.1017/S0022112070001210, 1970. Endo, N. and Masuda, F.: Small ripples in dunes regime and interpretation about dunes-missing in turbidite, The Journal of the Geological
455
Society of Japan, 103, 741–746, https://doi.org/10.5575/geosoc.103.741, 1997. Engelund, F.: Instability of erodible beds, Journal of Fluid Mechanics, 42, 225–244, https://doi.org/10.1017/S0022112070001210, 1970. Ferguson, R. and Church, M.: A simple universal equation for grain settling velocity, Journal of Sedimentary Research, 74, 933–937,
https://doi.org/10.1306/051204740933, 2004. erodible beds, Journal of Fluid Mechanics, 42, 225–244, https://doi.org/10.1017/S0022112070001210, 1970. ngelund, F.: Instability of erodible beds, Journal of Fluid Mechanics, 42, 225–244, https://doi.org/10.1017/S002 Engelund, F.: Instability of erodible beds, Journal of Fluid Mechanics, 42, 225 244, https://doi.org/10.1017/S0022112070001210, 1970. Ferguson, R. and Church, M.: A simple universal equation for grain settling velocity, Journal of Sedimentary Research, 74, 933–937,
https://doi.org/10.1306/051204740933, 2004. Foley, M. G.: Scour and fill in ephemeral streams, Tech. Rep. KH-R:KH-R-33, W. M. Keck Laboratory of Hydraulics and Water Resources,
460
California Institute of Technology, Pasadena, California, 1975. Fredsøe,
J.:
On
the
development
of
dunes
in
erodible
channels,
Journal
of
Fluid
Mechanics,
64,
1–16, Foley, M. G.: Scour and fill in ephemeral streams, Tech. Rep. KH-R:KH-R-33, W. M. Keck Laboratory of Hydraulics and Water Resources,
460
California Institute of Technology, Pasadena, California, 1975. Fredsøe,
J.:
On
the
development
of
dunes
in
erodible
channels,
Journal
of
Fluid
Mechanics,
64,
1–16,
https://doi.org/10.1017/S0022112074001960, 1974. Fredsøe, J.: Unsteady flow in straight alluvial streams. Part 2. Transition from dunes to plane bed, Journal of Fluid Mechanics, 102, 431–453, California Institute of Technology, Pasadena, California, 1975. Fredsøe,
J.:
On
the
development
of
dunes
in
erodible
channels,
Journal
of
Fluid
Mechanics,
64,
1–16,
https://doi.org/10.1017/S0022112074001960, 1974. Fredsøe, J.: Unsteady flow in straight alluvial streams. Part 2. Transition from dunes to plane bed, Journal of Fluid Mechanics, 102, 431–453,
https://doi.org/10.1017/S0022112081002723, 1981. 465 Fredsøe, J.: Unsteady flow in straight alluvial streams. Part 2. Transition from dunes to plane bed, Journal https://doi.org/10.1017/S0022112081002723, 1981. 465
Fukuoka, S., Okutsu, K., and Yamasaka, M.: Dynamic and kinematic features of sand waves in upper regime [in Japanese],
in:
Proceedings
of
the
Japan
Society
of
Civil
Engineers,
vol. 323,
pp. 77–89,
Japan
Society
of
Civil
Engineers,
https://doi.org/10.2208/jscej1969.1982.323_77, 1982. References Gabel S L : Geometry and kinematics of dunes during steady and unsteady flows in the Calamus River Nebraska USA Sedimentology Fukuoka, S., Okutsu, K., and Yamasaka, M.: Dynamic and kinematic features of sand waves in upper regime [in Japanese],
in:
Proceedings
of
the
Japan
Society
of
Civil
Engineers,
vol. 323,
pp. 77–89,
Japan
Society
of
Civil
Engineers,
https://doi.org/10.2208/jscej1969.1982.323_77, 1982. Gabel, S. L.: Geometry and kinematics of dunes during steady and unsteady flows in the Calamus Rive Gabel, S. L.: Geometry and kinematics of dunes during steady and unsteady flows in the Calamus River, Nebraska, USA, Sedimentology,
40, 237–269, https://doi.org/10.1111/j.1365-3091.1993.tb01763.x, 1993. 470 40, 237–269, https://doi.org/10.1111/j.1365-3091.1993.tb01763.x, 1993. 470
Gao, P.: Transition between two bed-Load transport regimes: saltation and sheet flow, Journal of Hydraulic Engineering, 134, 340–349,
https://doi.org/10.1061/(ASCE)0733-9429(2008)134:3(340), 2008. Gao, X., Narteau, C., and Rozier, O.: Development and steady states of transverse dunes: A numerical analysis of dune pattern coarsening
and giant dunes, Journal of Geophysical Research: Earth Surface, 120, 2200–2219, https://doi.org/10.1002/2015JF003549, 2015. 40, 237 269, https://doi.org/10.1111/j.1365 3091.1993.tb01763.x, 1993. 470
Gao, P.: Transition between two bed-Load transport regimes: saltation and sheet flow, Journal of Hydraulic Engineering, 134, 340–349,
https://doi.org/10.1061/(ASCE)0733-9429(2008)134:3(340), 2008. Gao, X., Narteau, C., and Rozier, O.: Development and steady states of transverse dunes: A numerical analysis of dune pattern coarsening Gao, P.: Transition between two bed-Load transport regimes: saltation and sheet flow, Journal of Hydraulic Engineering, 134, 340–349,
https://doi.org/10.1061/(ASCE)0733-9429(2008)134:3(340), 2008. https://doi.org/10.1061/(ASCE)0733-9429(2008)134:3(340), 2008. Gao, X., Narteau, C., and Rozier, O.: Development and steady states of transverse dunes: A numerical analysis of dune pattern coarsening
and giant dunes, Journal of Geophysical Research: Earth Surface, 120, 2200–2219, https://doi.org/10.1002/2015JF003549, 2015. Gao, X., Narteau, C., and Rozier, O.: Development and steady states of transverse dunes: A numerical analysis of dune pattern coarsening
and giant dunes, Journal of Geophysical Research: Earth Surface, 120, 2200–2219, https://doi.org/10.1002/2015JF003549, 2015. Gee, D. M.: Bed form response to nonsteady flows, Journal of the Hydraulics Division, 101, 437–449, 1975. 475 Gee, D. M.: Bed form response to nonsteady flows, Journal of the Hydraulics Division, 101, 437–449, 1975. 475
Guy, H. P., Simons, D. B., and Richardson, E. V.: Summary of alluvial channel data from flume experiments, 1956–61, Tech. Rep. 462-I,
Professional Paper, Washington, DC, 1966. Hage, S., Cartigny, M. J. B., Clare, M. A., Sumner, E. J., Vendettuoli, D., Hughes Clarke, J. E., Hubbard, S. M., Talling, P. J., Lintern,
D. G., Stacey, C. D., Englert, R. G., Vardy, M. E., Hunt, J. References 490
Kennedy, J. F.: The mechanics of dunes and antidunes in erodible-bed channels, Journal of Fluid Mechanics, 16, 521–544,
https://doi.org/10.1017/S0022112063000975, 1963. Keulegan, G. H.: Laws of turbulent flow in open channels, Journal National Bureau of Standards, Research Paper 1151, 21, 707–741, 1938. Kostaschuk, R. and Villard, P.: Flow and sediment transport over large subaqueous dunes: Fraser River, Canada, Sedimentology, 43, 849–863, ,
gy,
,
,
p
g
Q
,
Kennedy, J. F.: The mechanics of dunes and antidunes in erodible-bed channels, Journal of Fluid Mechanics, 16, 521–544,
https://doi.org/10.1017/S0022112063000975, 1963. Keulegan, G. H.: Laws of turbulent flow in open channels, Journal National Bureau of Standards, Research Paper 1151, 21, 707–741, 1938. Kennedy, J. F.: The mechanics of dunes and antidunes in erodible-bed channels, Journal of Fluid Mechanics, 16, 521–544,
https://doi.org/10.1017/S0022112063000975, 1963. https://doi.org/10.1017/S0022112063000975, 1963. Keulegan, G. H.: Laws of turbulent flow in open channels, Journal National Bureau of Standards, Research Paper 1151, 21, 707–741, 1938. eulegan, G. H.: Laws of turbulent flow in open channels, Journal National Bureau of Standards, Research Pape Kostaschuk, R. and Villard, P.: Flow and sediment transport over large subaqueous dunes: Fraser River, Canada, Sedimentology, 43, 849–863,
https://doi.org/10.1111/j.1365-3091.1996.tb01506.x, 1996. 495 Kostaschuk, R. and Villard, P.: Flow and sediment transport over large subaqueous dunes: Fraser River, Canad Kostaschuk, R. and Villard, P.: Flow and sediment transport over large subaqueous dunes: Fraser River, Canada, Sedimentology, 43, 849–863,
htt
//d i
/10 1111/j 1365 3091 1996 tb01506
1996
95 https://doi.org/10.1111/j.1365-3091.1996.tb01506.x, 1996. 495
Lanzoni, S., Gregoretti, C., and Stancanelli, L. M.: Coarse-grained debris flow dynamics on erodible beds, Journal of Geophysical Research:
Earth Surface, 122, 592–614, https://doi.org/10.1002/2016JF004046, 2017. Lowe, D. R.: Suspended-load fallout rate as an independent variable in the analysis of current structures, Sedimentology, 35, 765–776,
https://doi.org/10.1111/j.1365-3091.1988.tb01250.x, 1988. p
g
j
9
99
, 99
95
Lanzoni, S., Gregoretti, C., and Stancanelli, L. M.: Coarse-grained debris flow dynamics on erodible beds, Journal of Geophysical Research:
Earth Surface, 122, 592–614, https://doi.org/10.1002/2016JF004046, 2017. L
D R
S
d d l
d f ll
t
t
i d
d
t
i bl
i
th
l
i
f
t t
t
S di
t l
35 765 776 Lowe, D. R.: Suspended-load fallout rate as an independent variable in the analysis of current structures, Sedimentology, 35, 765–776,
https://doi.org/10.1111/j.1365-3091.1988.tb01250.x, 1988. Ma, H., Nittrouer, J. A., Naito, K., Fu, X., Zhang, Y., Moodie, A. References E., Yokokawa, M., Parsons, D. R., Hizzett, J. L., Azpiroz-Zabala, M., and Gee, D. M.: Bed form response to nonsteady flows, Journal of the Hydraulics Division, 101, 437–449, 1975. 475
Guy, H. P., Simons, D. B., and Richardson, E. V.: Summary of alluvial channel data from flume experiments, 1956–61, Tech. Rep. 462-I,
Professional Paper, Washington, DC, 1966. Guy, H. P., Simons, D. B., and Richardson, E. V.: Summary of alluvial channel data from flume experiments, 1956–61, Tech. Rep. 462-I,
Professional Paper, Washington, DC, 1966. Hage, S., Cartigny, M. J. B., Clare, M. A., Sumner, E. J., Vendettuoli, D., Hughes Clarke, J. E., Hubbard, S. M., Talling, P. J., Lintern,
D. G., Stacey, C. D., Englert, R. G., Vardy, M. E., Hunt, J. E., Yokokawa, M., Parsons, D. R., Hizzett, J. L., Azpiroz-Zabala, M., and
Vellinga, A. J.: How to recognize crescentic bedforms formed by supercritical turbidity currents in the geologic record: Insights from
480
active submarine channels, Geology, 46, 563–566, https://doi.org/10.1130/G40095.1, 2018. Vellinga, A. J.: How to recognize crescentic bedforms formed by supercritical turbidity currents in the geologic record: Insights from
480
active submarine channels, Geology, 46, 563–566, https://doi.org/10.1130/G40095.1, 2018. 18 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. Harms,
J. C.:
Primary
Sedimentary
Structures,
Annual
Review
of
Earth
and
Planetary
Sciences,
7,
227–248,
https://doi.org/10.1146/annurev.ea.07.050179.001303, 1979. Hendershot, M. L., Venditti, J. G., Bradley, R. W., Kostaschuk, R. A., Church, M., and Allison, M. A.: Response of low-angle dunes to
variable flow, Sedimentology, 63, 743–760, https://doi.org/10.1111/sed.12236, 2016. variable flow, Sedimentology, 63, 743–760, https://doi.org/10.1111/sed.12236, 2016. Hernandez-Moreira, R., Jafarinik, S., Sanders, S., Kendall, C. G. S. C., Parker, G., and Viparelli, E.: Emplacement of massive deposits by
sheet flow, Sedimentology, 67, 1951–1972, https://doi.org/10.1111/sed.12689, 2020. , P. Y.: Study of bedform geometry in large rivers, Tech. Rep. Q1386, Delft Hydraulics, Emmeloord, Netherlan Kennedy, J. F.: Stationary waves and antidunes in alluvial channels, Tech. Rep. KH-R-2, W. M. Keck L Kennedy, J. F.: Stationary waves and antidunes in alluvial channels, Tech. Rep. KH-R-2, W. M. Keck Laboratory of Hydraulics and Water
Resources, California Institute of Technology, Pasadena, CA, https://doi.org/10.7907/Z9QR4V22, 1961. 90 Kennedy, J. F.: Stationary waves and antidunes in alluvial channels, Tech. Rep. KH-R-2, W. M. Keck Laboratory of Hydraulics and Water
C lif
i
i
f
h
l
d
CA h
//d i
/10 90 / 9Q 4 22 1961 Resources, California Institute of Technology, Pasadena, CA, https://doi.org/10.7907/Z9QR4V22, 1961. References J., Wang, Y., Wu, B., and Parker, G.: The exceptional sediment load of
500
fine-grained dispersal systems: Example of the Yellow River, China, Science Advances, 3, https://doi.org/10.1126/sciadv.1603114, 2017. McLean, S. R.: The stability of ripples and dunes, Earth-Science Reviews, 29, 131–144, https://doi.org/10.1016/0012-8252(0)90032-Q, 1990. Mezaki, S.: Bed forms in the Yoro river at Azu, Chiba prefecture [in Japanese], Geographical Review of Japan, 46, 516–532,
https://doi.org/10.4157/grj.46.516, 1973. Ma, H., Nittrouer, J. A., Naito, K., Fu, X., Zhang, Y., Moodie, A. J., Wang, Y., Wu, B., and Parker, G.: The exceptional sediment load of
500
fine-grained dispersal systems: Example of the Yellow River, China, Science Advances, 3, https://doi.org/10.1126/sciadv.1603114, 2017. McLean, S. R.: The stability of ripples and dunes, Earth-Science Reviews, 29, 131–144, https://doi.org/10.1016/0012-8252(0)90032-Q, 1990. McLean, S. R.: The stability of ripples and dunes, Earth-Science Reviews, 29, 131–144, https://doi.org/10.101 McLean, S. R.: The stability of ripples and dunes, Earth-Science Reviews, 29, 131–144, https://doi.org/10.1016/0012-8252(0)90032-Q, 1990. Mezaki, S.: Bed forms in the Yoro river at Azu, Chiba prefecture [in Japanese], Geographical Review of Japan, 46, 516–532,
https://doi.org/10.4157/grj.46.516, 1973. Miall, A.: Alluvial deposits, in: Facies models 4, edited by James, N. P. and Dalrymple, R. W., pp. 105–137, Geological Association of
505
Canada, Canada, 2010. Nakasato Y and Izumi N : Linear stability analysis of small scale fluvial bed waves with active suspended sediment load [in Japanese] Miall, A.: Alluvial deposits, in: Facies models 4, edited by James, N. P. and Dalrymple, R. W., pp. 105–137, Geological Association of
505
Canada, Canada, 2010. Nakasato, Y. and Izumi, N.: Linear stability analysis of small-scale fluvial bed waves with active suspended sediment load [in Japanese],
Journal of applied mechanics, 11, 727–734, https://doi.org/10.2208/journalam.11.727, 2008. Naqshband, S., Ribberink, J. S., Hurther, D., and Hulscher, S. J. M. H.: Bed load and suspended load contributions to migrating sand dunes Canada, Canada, 2010. Nakasato, Y. and Izumi, N.: Linear stability analysis of small-scale fluvial bed waves with active suspended sediment load [in Japanese],
Journal of applied mechanics, 11, 727–734, https://doi.org/10.2208/journalam.11.727, 2008. Nakasato, Y. and Izumi, N.: Linear stability analysis of small-scale fluvial bed waves with active suspended sediment load [in Japanese],
Journal of applied mechanics, 11, 727–734, https://doi.org/10.2208/journalam.11.727, 2008. Naqshband, S., Ribberink, J. S., Hurther, D., and Hulscher, S. J. M. H.: Bed load and suspended load contributions to migrating sand dunes
in equilibrium, Journal of Geophysical Research: Earth Surface, 119, 1043–1063, https://doi.org/10.1002/2013JF003043, 2013JF003043,
510
2014. Naqshband, S., Ribberink, J. References CC BY 4.0 License. Schindler, R. J., Parsons, D. R., Ye, L., Hope, J. A., Baas, J. H., Peakall, J., Manning, A. J., Aspden, R. J., Malarkey, J., Simmons, S.,
520
Paterson, D. M., Lichtman, I. D., Davies, A. G., Thorne, P. D., and Bass, S. J.: Sticky stuff: Redefining bedform prediction in modern and
ancient environments, Geology, 43, 399–402, https://doi.org/10.1130/G36262.1, 2015. Shen, H. W., Harrison, A. S., and Mellema, W. J.: Temperature and Missouri river stages near Omaha, Journal of the Hydraulics Division,
104, 1–20, 1978. Schindler, R. J., Parsons, D. R., Ye, L., Hope, J. A., Baas, J. H., Peakall, J., Manning, A. J., Aspden, R. J., Malarkey, J., Simmons, S.,
520
Paterson, D. M., Lichtman, I. D., Davies, A. G., Thorne, P. D., and Bass, S. J.: Sticky stuff: Redefining bedform prediction in modern and
ancient environments, Geology, 43, 399–402, https://doi.org/10.1130/G36262.1, 2015. Shen, H. W., Harrison, A. S., and Mellema, W. J.: Temperature and Missouri river stages near Omaha, Journal of the Hydraulics Division,
104, 1–20, 1978. Shinohara, K. and Tsubaki, T.: On the Characteristics of Sand Waves Formed Upon Beds of the Open Channels and Rivers, Report of
525
Research Institute for Applied Mechanics, VII, Kyushu University, Kyusyu, Japan, 1959. Shinohara, K. and Tsubaki, T.: On the Characteristics of Sand Waves Formed Upon Beds of the Open Channels and Rivers, Report of
525
Research Institute for Applied Mechanics, VII, Kyushu University, Kyusyu, Japan, 1959. Simons, D. B.: Theory and design of stable channels in alluvial material, Ph.D. thesis, Colorado State University, Fort Collins, Colorado,
1957. Simons, D. B., Richardson, E. V., and Albertson, M. L.: Flume Studies Using Medium Sand (0.45 mm), Tech. Rep. Water-Supply Paper Research Institute for Applied Mechanics, VII, Kyushu University, Kyusyu, Japan, 1959. Simons, D. B.: Theory and design of stable channels in alluvial material, Ph.D. thesis, Colorado State University, Fort Collins, Colorado,
1957. Simons, D. B.: Theory and design of stable channels in alluvial material, Ph.D. thesis, Colorado State University, Fort Collins, Colorado,
1957. Simons, D. B., Richardson, E. V., and Albertson, M. L.: Flume Studies Using Medium Sand (0.45 mm), Tech. Rep. Water-Supply Paper
1498-A, U. S. Geological Survey, Washington, DC, 1961. 530 Simons, D. B., Richardson, E. V., and Albertson, M. L.: Flume Studies Using Medium Sand (0.45 mm), T 1498-A, U. S. Geological Survey, Washington, DC, 1961. 530
Smith, J. D. and McLean, S. References Tanaka, Y.: An experimental study on anti-dunes, Disaster Prevention Research Institute Annuals, 13, 271–284, 1970. Talling, P. J., Masson, D. G., Sumner, E. J., and Malgesini, G.: Subaqueous sediment density flows: Depositional processes and deposit types,
Sedimentology, 59, 1937–2003, https://doi.org/10.1111/j.1365-3091.2012.01353.x, 2012. Y.: An experimental study on anti-dunes, Disaster Prevention Research Institute Annuals, 13, 271–284, 1970. Tanaka, Y.: An experimental study on anti dunes, Disaster Prevention Research Institute Annuals, 13, 271 284, 1970. Taylor, B. D.: Temperature effects in alluvial streams, Tech. Rep. KH-R-27, W. M. Keck Laboratory of Hydraulics and Water Resources, Taylor, B. D.: Temperature effects in alluvial streams, Tech. Rep. KH-R-27, W. M. Keck Laboratory of Hy Taylor, B. D.: Temperature effects in alluvial streams, Tech. Rep. KH-R-27, W. M. Keck Laboratory of Hydraulics and Water Resources,
California Institute of Technology, Pasadena, CA, https://doi.org/10.7907/Z93776PN, 1971. 545 California Institute of Technology, Pasadena, CA, https://doi.org/10.7907/Z93776PN, 1971. 545
Tilston, M., Arnott, R., Rennie, C., and Long, B.: The influence of grain size on the velocity and sediment concentration profiles and
depositional record of turbidity currents, Geology, 43, 839–842, https://doi.org/10.1130/G37069.1, 2015. Tilston, M., Arnott, R., Rennie, C., and Long, B.: The influence of grain size on the velocity and sediment concentration profiles and
depositional record of turbidity currents, Geology, 43, 839–842, https://doi.org/10.1130/G37069.1, 2015. van den Berg, J. H. and van Gelder, A.: A new bedform stability diagram, with emphasis on the transition of ripples to plane bed in flows
over fine sand and silt, in: Alluvial sedimentation, edited by Marzo, M. and Puigdefabregas, C., vol. 17, pp. 11–21, Blackwell Scientific
Publications, Special Publications, International Association of Sedimentologists, 1993. 550 van den Berg, J. H. and van Gelder, A.: A new bedform stability diagram, with emphasis on the transition of ripples to plane bed in flows
over fine sand and silt, in: Alluvial sedimentation, edited by Marzo, M. and Puigdefabregas, C., vol. 17, pp. 11–21, Blackwell Scientific 550 Publications, Special Publications, International Association of Sedimentologists, 1993. 550
van Duin, O. J. M., Hulscher, S. J. M. H., Ribberink, J. S., and Dohmen-Janssen, C. M.: Modeling of spatial lag in bed-load transport
processes and its effect on dune morphology, Journal of Hydraulic Engineering, 143, 04016 084, https://doi.org/10.1061/(ASCE)HY.1943-
7900.0001254, 2017. Vesipa, R., Camporeale, C., and Ridolfi, L.: A shallow-water theory of river bedforms in supercritical conditions, Physics of Fluids, 24, van Duin, O. J. M., Hulscher, S. J. M. H., Ribberink, J. References R.: Spatially averaged flow over a wavy surface, Journal of Geophysical Research, 82, 1735–1746,
https://doi.org/10.1029/JC082i012p01735, 1977. Sohn, Y. K.: On traction-carpet sedimentation, Journal of Sedimentary Research, 67, 502–509, https://doi.org/10.1306/D42685AE-2B26-
11D7-8648000102C1865D, 1997. 1498-A, U. S. Geological Survey, Washington, DC, 1961. Smith, J. D. and McLean, S. R.: Spatially averaged flow over a wavy surface, Journal of Geophysical Research, 82, 1735–1746,
https://doi.org/10.1029/JC082i012p01735, 1977. p
g
p
,
Sohn, Y. K.: On traction-carpet sedimentation, Journal of Sedimentary Research, 67, 502–509, https://doi.org/10.1306/D42685AE-2B26-
11D7-8648000102C1865D, 1997. Southard, J. B.: Representation of bed configurations in depth-velocity-size diagrams, Journal of Sedimentary Research, 41, 903–915,
535
https://doi.org/10.1306/74D723B0-2B21-11D7-8648000102C1865D, 1971. Sukhodolov, A. N., Fedele, J. J., and Rhoads, B. L.: Structure of flow over alluvial bedforms: an experiment on linking field and laboratory
methods, Earth Surface Processes and Landforms, 31, 1292–1310, https://doi.org/10.1002/esp.1330, 2006. Sumer, B. M., Kozakiewicz, A., Fredsøe, J., and Deigaard, R.: Velocity and concentration profiles in sheet-flow layer of movable bed, Journal Southard, J. B.: Representation of bed configurations in depth-velocity-size diagrams, Journal of Sedimentary Research, 41, 903–915,
535
https://doi.org/10.1306/74D723B0-2B21-11D7-8648000102C1865D, 1971. Sukhodolov, A. N., Fedele, J. J., and Rhoads, B. L.: Structure of flow over alluvial bedforms: an experiment on linking field and laboratory
methods, Earth Surface Processes and Landforms, 31, 1292–1310, https://doi.org/10.1002/esp.1330, 2006. Sukhodolov, A. N., Fedele, J. J., and Rhoads, B. L.: Structure of flow over alluvial bedforms: an experiment
methods, Earth Surface Processes and Landforms, 31, 1292–1310, https://doi.org/10.1002/esp.1330, 2006 Sumer, B. M., Kozakiewicz, A., Fredsøe, J., and Deigaard, R.: Velocity and concentration profiles in sheet-flow Sumer, B. M., Kozakiewicz, A., Fredsøe, J., and Deigaard, R.: Velocity and concentration profiles in sheet-flow layer of movable bed, Journal
of Hydraulic Engineering, 122, 549–558, https://doi.org/10.1061/(ASCE)0733-9429(1996)122:10(549), 1996. 540 Sumer, B. M., Kozakiewicz, A., Fredsøe, J., and Deigaard, R.: Velocity and concentration profiles in sheet-flow layer of movable bed, Journal of Hydraulic Engineering, 122, 549–558, https://doi.org/10.1061/(ASCE)0733-9429(1996)122:10(549), 1996. 540
Talling, P. J., Masson, D. G., Sumner, E. J., and Malgesini, G.: Subaqueous sediment density flows: Depositional processes and deposit types,
Sedimentology, 59, 1937–2003, https://doi.org/10.1111/j.1365-3091.2012.01353.x, 2012. Tanaka, Y.: An experimental study on anti-dunes, Disaster Prevention Research Institute Annuals, 13, 271–284, 1970. Taylor, B. D.: Temperature effects in alluvial streams, Tech. Rep. KH-R-27, W. M. Keck Laboratory of Hydraulics and Water Resources, Talling, P. J., Masson, D. G., Sumner, E. J., and Malgesini, G.: Subaqueous sediment density flows: Depositional processes and deposit types,
Sedimentology, 59, 1937–2003, https://doi.org/10.1111/j.1365-3091.2012.01353.x, 2012. References S., Hurther, D., and Hulscher, S. J. M. H.: Bed load and suspended load contrib in equilibrium, Journal of Geophysical Research: Earth Surface, 119, 1043–1063, https://doi.org/10.1002/2013JF003043, 2013JF003043,
510
2014. Naqshband, S., Hoitink, A. J. F., McElroy, B., Hurther, D., and Hulscher, S. J. M. H.: A sharp view on river dune transition to upper stage
plane bed, Geophysical Research Letters, 44, 11 437–11 444, https://doi.org/10.1002/2017GL075906, 2017. Neill C R : Bed forms in the Lower Red Deer River Alberta Journal of Hydrology 7 58–85 https://doi org/10 1016/0022-1694(68)90195- 2014. Naqshband, S., Hoitink, A. J. F., McElroy, B., Hurther, D., and Hulscher, S. J. M. H.: A sharp view on river dune transition to upper stage
plane bed, Geophysical Research Letters, 44, 11 437–11 444, https://doi.org/10.1002/2017GL075906, 2017. Naqshband, S., Hoitink, A. J. F., McElroy, B., Hurther, D., and Hulscher, S. J. M. H.: A sharp view on river dune transition to upper stage
plane bed, Geophysical Research Letters, 44, 11 437–11 444, https://doi.org/10.1002/2017GL075906, 2017. Neill, C. R.: Bed forms in the Lower Red Deer River, Alberta, Journal of Hydrology, 7, 58–85, https://doi.org/10.1016/0022-1694(68)90195-
9, 1969. 5 515 9, 1969. 515
Niño,
Y.,
Lopez,
F.,
and
Garcia,
M.:
Threshold
for
particle
entrainment
into
suspension,
Sedimentology,
50,
247–263,
https://doi.org/10.1046/j.1365-3091.2003.00551.x, 2003. Sambrook Smith, G. H., Best, J. L., Leroy, J. Z., and Orfeo, O.: The alluvial architecture of a suspended sediment dominated meandering
river: the Río Bermejo, Argentina, Sedimentology, 63, 1187–1208, https://doi.org/10.1111/sed.12256, 2016. Niño,
Y.,
Lopez,
F.,
and
Garcia,
M.:
Threshold
for
particle
entrainment
into
suspension,
Sedimentology,
50,
247–263,
https://doi.org/10.1046/j.1365-3091.2003.00551.x, 2003. Sambrook Smith, G. H., Best, J. L., Leroy, J. Z., and Orfeo, O.: The alluvial architecture of a suspended sediment dominated meandering
river: the Río Bermejo Argentina Sedimentology 63 1187 1208 https://doi org/10 1111/sed 12256 2016 Niño,
Y.,
Lopez,
F.,
and
Garcia,
M.:
Threshold
for
particle
entrainment
into
suspension,
Sedimentology,
50,
247–263,
https://doi.org/10.1046/j.1365-3091.2003.00551.x, 2003. https://doi.org/10.1046/j.1365-3091.2003.00551.x, 2003. Sambrook Smith, G. H., Best, J. L., Leroy, J. Z., and Orfeo, O.: The alluvial architecture of a suspended sediment dominated meandering
river: the Río Bermejo, Argentina, Sedimentology, 63, 1187–1208, https://doi.org/10.1111/sed.12256, 2016. Sambrook Smith, G. H., Best, J. L., Leroy, J. Z., and Orfeo, O.: The alluvial architecture of a suspended sediment dominated meandering
river: the Río Bermejo, Argentina, Sedimentology, 63, 1187–1208, https://doi.org/10.1111/sed.12256, 2016. 19 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. Wilbers, A.: The development and hydraulic roughness of subaqueous dunes, Ph.D. thesis, Utrecht University, Utrecht, Netherlands, 2004.
Williams, G. P.: Flume width and water depth effects in sediment transport experiments, United States Government Printing Office, Wash-
ington, DC, U. S. Geological Survey Professional Paper 562-H, 1970.
560
Wong, M. and Parker, G.: Reanalysis and Correction of Bed-Load Relation of Meyer-Peter and Müller Using Their Own Database, Journal
of Hydraulic Engineering, 132, 1159–1168, https://doi.org/10.1061/(ASCE)0733-9429(2006)132:11(1159), 2006.
Yokokawa, M., Izumi, N., Naito, K., Parker, G., Yamada, T., and Greve, R.: Cyclic steps on ice, Journal of Geophysical Research: Earth
Surface, 121, 1023–1048, https://doi.org/10.1002/2015JF003736, 2016. References S., and Dohmen-Janssen, C. M.: Modeling of spatial lag in bed-load transport
processes and its effect on dune morphology, Journal of Hydraulic Engineering, 143, 04016 084, https://doi.org/10.1061/(ASCE)HY.1943-
7900.0001254, 2017. Vesipa, R., Camporeale, C., and Ridolfi, L.: A shallow-water theory of river bedforms in supercritical conditions, Physics of Fluids, 24,
94 104, https://doi.org/10.1063/1.4753943, 2012. 555 555 94 104, https://doi.org/10.1063/1.4753943, 2012. 555
Walker, R. G.: The origin and significance of the internal sedimentay structures of turbidites, Proceedings of the Yorkshire Geological Society,
35, 1–32, https://doi.org/10.1144/pygs.35.1.1, 1965. 20 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. Wilbers, A.: The development and hydraulic roughness of subaqueous dunes, Ph.D. thesis, Utrecht University, Utrecht, Netherlands, 2004. Williams, G. P.: Flume width and water depth effects in sediment transport experiments, United States Government Printing Office, Wash-
ington, DC, U. S. Geological Survey Professional Paper 562-H, 1970. 60
Wong, M. and Parker, G.: Reanalysis and Correction of Bed-Load Relation of Meyer-Peter and Müller Using Their Own Database, Journal
of Hydraulic Engineering, 132, 1159–1168, https://doi.org/10.1061/(ASCE)0733-9429(2006)132:11(1159), 2006. Yokokawa, M., Izumi, N., Naito, K., Parker, G., Yamada, T., and Greve, R.: Cyclic steps on ice, Journal of Geophysical Research: Earth
Surface, 121, 1023–1048, https://doi.org/10.1002/2015JF003736, 2016. 21 21 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. Figure 1. Contour map of perturbation growth rate ωi without suspension. Sediment diameter and flow depth were set to ˜D = 0.12 mm and
˜h0 = 0.1201 m, respectively. The dotted line denotes the threshold of sediment motion. The dashed lines denote the critical Froude numbers
Frcd and Frca for instabilities. The region where the growth rate is positive is highlighted in grey. Figure 1. Contour map of perturbation growth rate ωi without suspension. Sediment diameter and flow depth were set to ˜D = 0.12 mm and
˜h0 = 0.1201 m, respectively. The dotted line denotes the threshold of sediment motion. The dashed lines denote the critical Froude numbers
Frcd and Frca for instabilities. The region where the growth rate is positive is highlighted in grey. 22 g/10.5194/esurf-2021-60
cussion started: 28 July 2021
2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. Figure 2. Contour maps of the dominant wavenumbers of perturbations with a fixed sediment diameter ˜D. Symbols are observational
data. a, ˜D = 0.12 mm without suspension. b, ˜D = 0.12 mm with suspension. References c, ˜D = 0.25 mm without suspension. d, ˜D = 0.25 mm with
suspension. a and b, The range of ˜D of observational data is from 0.0883 mm to 0.163 mm. c and d, The range of ˜D of observational data is
from 0.184 mm to 0.34 mm. Figure 2. Contour maps of the dominant wavenumbers of perturbations with a fixed sediment diameter ˜D. Symbols are observational
data. a, ˜D = 0.12 mm without suspension. b, ˜D = 0.12 mm with suspension. c, ˜D = 0.25 mm without suspension. d, ˜D = 0.25 mm with
suspension. a and b, The range of ˜D of observational data is from 0.0883 mm to 0.163 mm. c and d, The range of ˜D of observational data is
from 0.184 mm to 0.34 mm. 23 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. Figure 3. Contour maps of dominant wavenumbers of perturbations. Symbols are observational data. The sediment diameter ˜D was set to
1.20 mm. a, Without suspension. b, With suspension. The range of ˜D of observational data is from 0.883 mm to 1.63 mm. Figure 3. Contour maps of dominant wavenumbers of perturbations. Symbols are observational data. The sediment diameter ˜D was set to
1.20 mm. a, Without suspension. b, With suspension. The range of ˜D of observational data is from 0.883 mm to 1.63 mm. 24 24 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. Figure A1. Conceptual diagram of the flow. The dimensionless parameters u and w are the flow velocities in x- and z- directions, respectively,
h is the flow depth, Z denotes the bed height, and R is the height of the reference level at which the flow velocity is assumed to vanish in a
logarithmic law. Figure A1. Conceptual diagram of the flow. The dimensionless parameters u and w are the flow velocities in x- and z- directions, respectively,
h is the flow depth, Z denotes the bed height, and R is the height of the reference level at which the flow velocity is assumed to vanish in a
logarithmic law. Figure A1. Conceptual diagram of the flow. The dimensionless parameters u and w are the flow velocities in x- and z- directions, respectively,
h is the flow depth, Z denotes the bed height, and R is the height of the reference level at which the flow velocity is assumed to vanish in a
logarithmic law. 25 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. Figure A2. Conceptual diagram of the sediment bed. The origin of z-direction is denoted by O. The parameter D is the dimensionless
diameter of a bed particle, B0 is the height of the top of the bed-load layer in the basic state, and R0 is the height of reference level in the
basic state. Figure A2. Conceptual diagram of the sediment bed. The origin of z-direction is denoted by O. The parameter D is the dimensionless
diameter of a bed particle, B0 is the height of the top of the bed-load layer in the basic state, and R0 is the height of reference level in the
basic state. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. 26 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. Table A1. Summary of data used for the stability diagram with ˜D = 0.12 mm. Reference
# of
points
flow
depth
˜h [m]
flow
velocity
˜U [m/s]
particle
diameter
˜D [mm]
Froude
number
Fr
Source
Plane bed with suspension
Taylor (1971)
2
0.0783
0.585
0.138
0.668
Flume
Simons (1957)
1
1.83
0.585
0.096
0.138
Field
Culbertson et al. (1972)
8
0.494–0.957
1.06–1.42
0.16–0.2
0.415–0.524
Field
Dunes
Baird (2010)
1
2.24
0.744
0.16
0.159
Field
Shen et al. (1978)
2
2.94–3.07
1.55–1.61
0.208–0.218
0.288–0.294
Field
Antidunes
Tanaka (1970)
5
0.0443–0.11
0.658–1.14
0.145
0.852–1.38
Flume
27 27 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. le A2. Summary of data used for the stability diagram with ˜D = 0.25 mm. Reference
# of
points
flow
depth
˜h [m]
flow
velocity
˜U [m/s]
particle
diameter
˜D [mm]
Froude
number
Fr
Source
Plane bed without suspension
Taylor (1971)
7
0.0606–0.061
0.198–0.229
0.215–0.248
0.256–0.296
Flume
Plane bed with suspension
Bridge and Best (1988)
2
0.1
0.9–0.98
0.3
0.909–0.990
Flume
Guy et al. (1966)
10
0.155–0.241
0.881–1.29
0.19–0.33
0.686–1.05
Flume
Taylor (1971)
4
0.0788–0.114
0.692–0.878
0.228
0.778–0.838
Flume
Culbertson et al. (1972)
3
0.284–0.969
0.457–1.52
0.2–0.24
0.264–0.492
Field
Dunes
Bridge and Best (1988)
2
0.1
0.6–0.8
0.3
0.606–0.808
Flume
Gee (1975)
6
0.0454–0.105
0.305–0.920
0.305
0.325–1.04
Flume
Guy et al. (1966)
31
0.140–0.344
0.421–0.820
0.19–0.33
0.318–0.628
Flume
Naqshband et al. (2014)
2
0.25
0.64–0.8
0.29
0.409–0.511
Flume
Abdel-Fattah et al. (2004), Wilbers (2004)
6
4.03–5.72
0.31–0.75
0.239–0.322
0.0493–0.118
Field
Baird (2010)
2
1.67–2.33
0.597–1.33
0.21
0.148–0.278
Field
Gabel (1993)
4
0.4–0.43
0.61–0.65
0.31–0.33
0.301–0.320
Field
Julien (1992)
28
6.6–19.5
1.3–1.55
0.2–0.33
0.094–0.186
Field
Mezaki (1973)
8
0.9–1.35
0.49–0.83
0.21
0.154–0.272
Field
Neill (1969)
1
3.05
1.10
0.34
0.201
Field
Shen et al. (1978)
16
2.78–4.94
1.37–1.73
0.193–0.266
0.240–0.309
Field
Antidunes
Foley (1975)
3
0.0305–0.0473
0.546–0.692
0.28
0.813–1.26
Flume
Fukuoka et al. (1982)
15
0.0209–0.0569
0.349–0.93
0.19
0.760–1.45
Flume
Guy et al. (1966)
13
0.0914–0.204
1.06–1.62
0.19–0.33
0.892–1.30
Flume
Kennedy (1961)
15
0.0448–0.106
0.637–1.05
0.233
0.798–1.49
Flume 28 https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. https://doi.org/10.5194/esurf-2021-60
Preprint. Discussion started: 28 July 2021
c⃝Author(s) 2021. CC BY 4.0 License. Summary of data used for the stability diagram with ˜D = 1.20 mm. Reference
# of
points
flow
depth
˜h [m]
flow
velocity
˜U [m/s]
particle
diameter
˜D [mm]
Froude
number
Fr
Source
Plane bed without suspension
Guy et al. (1966)
6
0.149–0.314
0.381–0.454
0.93
0.229–0.368
Flume
Taylor (1971)
6
0.061
0.305–0.335
1.07
0.394–0.433
Flume
Williams (1970)
14
0.0283–0.155
0.332–0.558
1.35
0.401–0.986
Flume
Plane bed with sheet flow
Hernandez-Moreira et al. (2020)
2
0.091–0.1
1.58–1.66
0.85–1.09
1.59–1.75
Flume
Williams (1970)
6
0.0299–0.143
0.914–1.99
1.35
1.68–1.90
Flume
Dunes
Blom et al. (2003)
1
0.245
0.69
1.3
0.445
Flume
Gee (1975)
4
0.0622–0.146
0.449–0.584
1
0.488–0.575
Flume
Guy et al. (1966)
14
0.140–0.338
0.488–0.951
0.93
0.370–0.583
Flume
Williams (1970)
71
0.0872–0.223
0.448–0.917
1.35
0.343–0.825
Flume
Shinohara and Tsubaki (1959)
2
0.202–0.372
0.7–0.752
1.33
0.394–0.497
Field
Sukhodolov et al. (2006)
1
0.35
0.44
1
0.238
Field
Antidunes
Fukuoka et al. (1982)
1
0.0355
0.852
1.6
1.44
Flume
Tanaka (1970)
1
0.0807
1.74
0.91
1.95
Flume
Williams (1970)
43
0.0271–0.157
0.466–1.69
1.35
0.826–2
Flume 29
|
W3013825173.txt
|
https://www.frontiersin.org/articles/10.3389/fvets.2020.00157/pdf
|
en
|
Dietary Supplementation With Magnolia Bark Extract Alters Chicken Intestinal Metabolite Levels
|
Frontiers in veterinary science
| 2,020
|
cc-by
| 4,943
|
ORIGINAL RESEARCH
published: 24 March 2020
doi: 10.3389/fvets.2020.00157
Dietary Supplementation With
Magnolia Bark Extract Alters Chicken
Intestinal Metabolite Levels
Inkyung Park 1 , Sungtaek Oh 2 , Erik. P. Lillehoj 3 and Hyun S. Lillehoj 1*
1
Animal Bioscience and Biotechnology Laboratory, Beltsville Agricultural Research Center, Agricultural Research Service,
United States Department of Agriculture, Beltsville, MD, United States, 2 Neuroregeneration and Stem Cell Programs,
Institute for Cell Engineering, Johns Hopkins University School of Medicine, Baltimore, MD, United States, 3 Department of
Pediatrics, University of Maryland School of Medicine, Baltimore, MD, United States
Edited by:
Rajesh Jha,
University of Hawaii at Manoa,
United States
Reviewed by:
Fulvia Bovera,
University of Naples Federico II, Italy
Gang Liu,
Institute of Subtropical Agriculture,
Chinese Academy of Sciences, China
*Correspondence:
Hyun S. Lillehoj
hyun.lillehoj@ars.usda.gov
Specialty section:
This article was submitted to
Animal Nutrition and Metabolism,
a section of the journal
Frontiers in Veterinary Science
Magnolia bark extract administered as a dietary supplement to poultry confers
a performance and health benefit, but the mechanisms are unknown. Here, a
metabolomics approach was used to identify changes in intestinal metabolite levels
in chickens fed an unsupplemented diet or a diet supplemented with magnolia bark
extract. Total body weight gains of chickens fed magnolia bark-supplemented diets
were increased 2% (from 861 to 878 g/chicken), compared with chickens fed an
unsupplemented diet. Compared with unsupplemented controls, the levels of 278
intestinal biochemicals (metabolites) were altered (165 increased, 113 decreased) in
chickens given the magnolia-supplemented diet. Data for biochemicals of intestinal
contents of chickens fed the unsupplemented diet clustered on the left side of the
PCA score plot, while those of the magnolia-supplemented diet were separated
and clustered on the right side. The biochemicals included changes in the levels
of amino acids, fatty acids, peptides, and nucleosides, which provided a distinctive
biochemical signature unique to the magnolia-supplemented group, compared with
the unsupplemented group. These results provide the foundation for future studies
to identify naturally-produced biochemicals that might be used to improve poultry
growth performance.
Keywords: phytochemical, intestine, metabolomics, growth, chicken, magnolia
Received: 13 January 2020
Accepted: 04 March 2020
Published: 24 March 2020
Citation:
Park I, Oh S, Lillehoj EP and
Lillehoj HS (2020) Dietary
Supplementation With Magnolia Bark
Extract Alters Chicken Intestinal
Metabolite Levels.
Front. Vet. Sci. 7:157.
doi: 10.3389/fvets.2020.00157
INTRODUCTION
Regulatory agencies warn that the rise of multidrug-resistant bacterial pathogens could potentially
be the greatest threat to human health of our time (1, 2). Antibiotic-resistant bacteria are estimated
to kill more than 100 million people worldwide per year by 2050. Among the causes for the
development and spread of antibiotic-resistant human pathogens is the overuse of antibiotic
growth promoters (AGPs) in food animal production. Animal agriculture is responsible for more
than 70% of total antibiotic usage worldwide, principally as in-feed additives for growth promotion
Frontiers in Veterinary Science | www.frontiersin.org
1
March 2020 | Volume 7 | Article 157
Park et al.
Magnolia Efficacy on Gut Metabolites
and prophylactic disease control (3). In commercial poultry
production, AGPs have been used to increase the growth rate
and improve feed conversion (kg body weight gain per kg feed)
since the 1940’s (4). While it is clear that dietary supplementation
of chickens with antibiotics alters the intestinal microbiome and
increases the bioavailability of nutrients to the host, the exact
mechanism of action of AGPs is unknown (5–7). As a result,
formulating a rationale scientific approach to the identification
of non-antibiotic alternatives that provide a growth enhancing
effect without the potential for development of drug resistance
has been impeded.
The magnolia plant contains a variety of chemical compounds
with demonstrated bioactive properties, and both the bark
and flower buds of Magnolia officinalis have been used
for hundreds of years in traditional human medicine (8).
Magnolia extracts and its purified components have been
demonstrated to have anti-cancer effects, as well as for treating
a variety of neuronal, inflammatory, cardiovascular, and
gastrointestinal disorders (8–11). Few studies, however, have
investigated the medicinal effects of magnolia in veterinary
medicine (12). We previously reported that two commonlyused antibiotics, bacitracin and virginiamycin, alter the
intestinal metabolome when given as dietary AGPs to broiler
chickens (13). Therefore, the current study was performed to
characterize the metabolic changes in the chicken intestine
following dietary supplementation with an extract of M.
officinalis bark in order to identify biochemical compounds
that might serve as alternatives to AGPs for improving poultry
growth performance.
TABLE 1 | Composition of the basal diet.
Ingredient
%
Corn
59.01
Soybean meal
33.99
Soybean oil
2.75
Dicalcium phosphate
2.00
Calcium carbonate
1.40
Salt
0.35
Poultry vitamin mixa
0.20
Poultry mineral mixb
0.15
DL-Methionine
0.10
Choline chloride (60%)
0.05
Total
100.0
Calculated nutrient composition
%
Crude protein
18.0
Crude fiber
3.21
Crude oil
8.48
Ash
5.41
Calcium carbonate
1.19
Available phosphorus
0.54
Lysine
1.00
Methionine
0.42
Cysteine + Methionine
0.65
Metabolizable energy, Mcal/kg
3.59
a The
vitamin mixture provided the following nutrients per kg of diet: vitamin A, 2,000 IU;
vitamin D3 , 22 IU; vitamin E, 16 mg; vitamin K, 0.1 mg; thiamin, 3.4 mg; riboflavin, 1.8 mg;
vitamin B6 , 6.4 mg; vitamin B12 , 0.013 mg; biotin, 0.17 mg; pantothenic acid, 8.7 mg; folic
acid, 0.8 mg; niacin, 23.8 mg.
b The mineral mixture provided the following nutrients per kg of diet: Fe, 0.4 mg; Zn, 0.2 mg;
Mn, 0.18 mg; Co, 0.0013 mg; Cu, 0.021 mg; Se, 0.0002 mg.
MATERIALS AND METHODS
Animals and Ethics Statement
One-day-old male Ross 708 commercial broiler chickens
(Longenecker’s Hatchery, Elizabethtown, PA) were randomly
divided into two treatment groups (n = 16/group). The chickens
were housed in starter cages from day 1 to day 14 of age
prior to transfer to finisher cages where they were kept until
sacrifice on day 21. All chickens were housed in the same room
and provided ad libitum access to feed and water. Each cage
was 0.65 m in width and 0.75 m in length (14 chickens/m2 ).
The guidelines for the care and use of animals in agriculture
research were followed throughout the entire experiment, and all
animal protocols (# 16-001) were approved by the Institutional
Animal Care and Use Committee of the Beltsville Agriculture
Research Center.
Experimental Design, Growth
Performance, and Intestinal Metabolomics
Analysis
Chickens were fed a corn-soybean meal-based diet formulated
to meet the requirements for chickens as suggested by National
Research Council (Table 1). Animals in the unsupplemented,
control diet group were provided the basal diet alone, while those
in the magnolia bark extract group were provided with the basal
diet supplemented with 0.33 mg/kg of the magnolia extract. Feed
additions were weighed and recorded daily, and feeders were
shaken once per day. The chickens were weighed at 21 days of
age for calculation of growth performance. At 21 days of age, 8
chickens from the unsupplemented control group and 8 chickens
from magnolia-supplemented group were euthanized by cervical
dislocation and the intestine harvested. Intestinal contents were
collected aseptically by gently finger-stripping the ileal segment,
immediately placed on dry ice, and stored at −80◦ C. Global
metabolomic profiling of the intestinal contents was performed
by mass spectrometry (MS) (Metabolon, Durham, NC) as
described (13). Raw data was extracted and processed using the
DiscoveryHD4 global metabolomics platform. Compounds were
Preparation of Magnolia Bark Extract
An extract of M. officinalis bark (Pancosma, Geneva, Switzerland)
was prepared as previously described (12). Briefly, the bark
was washed, dried at 50◦ C to a dry matter content of
90%, and comminuted. The dried material was subjected
to supercritical CO2 fluid extraction at 1200–1400 L/h for
3.5 h at 25–30 MPa and 35–40◦ C, and the extract taken up
in ethanol.
Frontiers in Veterinary Science | www.frontiersin.org
2
March 2020 | Volume 7 | Article 157
Park et al.
Magnolia Efficacy on Gut Metabolites
by calculating the q-value to account for the false positives
that normally occur in metabolomics-based studies. Principle
component analysis (PCA) was measured, which is a commonly
used algorithm to analyze the metabolic profiling originating
from different samples (14). Random forest analysis (RFA) was
performed to identify metabolite signatures and the biochemical
importance of the 30 most significantly altered metabolites for
distinguishing the control vs. magnolia-supplemented groups.
identified by comparison to library entries of purified standards
or recurrent unknown entities based on retention index, accurate
mass match to the library within 10 ppm, and MS/MS forward
and reverse scores between experimental data and authentic
standards. MS/MS scores were based on comparison of the ions
present in the experimental spectrum to the ions present in the
library spectrum.
Statistical Analysis
Statistical analysis was performed as previously described (13).
Each chicken was considered the experimental unit. The type of
experimental diet was considered the treatment factor. Data were
analyzed using a mixed model methodology (PROC MIXED,
SAS Institute, Cary NC). For growth performance, mean ±
SEM values were calculated for final body weight at day 21.
Differences between means were compared using the two-tailed
Student’s t test with p ≤ 0.05 considered significantly different.
ANOVA was used to identify the biochemicals whose levels
were significantly altered between the unsupplemented and
magnolia-supplemented groups following median scaling, log
transformation, and imputation of missing values, if any, with
the minimum value observed for each compound. All data were
analyzed for outliers and one sample from the control group
was excluded in the biochemical analysis. Standard statistical
analyses of log-transformed data were performed using Array
Studio software (OmicSoft, Cary, NC). For analyses that were
not standard in Array Studio, the programs R (R Foundation for
Statistical Computing, Vienna, Austria) or JMP (SAS Institute)
were used. An estimate of the false discovery rate was obtained
RESULTS
Dietary Magnolia Bark Extract
Supplementation Increases Chicken Body
Weight Gain
No differences in initial body weights between the two dietary
groups at day 1 were observed (data not shown). Final body
weights (878 ± 5.8 g) at day 21 of chickens fed the magnoliasupplemented diet were greater (p < 0.05) than those (861 ±
5.8 g) of chickens fed the unsupplemented diet, confirming the
previously reported growth enhancing effect of chickens fed a diet
supplemented with 0.33 mg the magnolia extract/kg (12).
Dietary Magnolia Bark Extract
Supplementation Affects Principle
Component Analysis
PCA revealed a district separation of identified intestinal
metablolites between chickens fed the unsupplemented, control
diet and those fed the magnolia-supplemented diet (Figure 1).
FIGURE 1 | Principle component analysis score plot of identified metabolites in intestinal contents of chicken fed an unsupplemented, control diet (blue) or a diet
supplemented with magnolia (MG) bark extract (green). Each dot on the plot represents an individual sample.
Frontiers in Veterinary Science | www.frontiersin.org
3
March 2020 | Volume 7 | Article 157
Park et al.
Magnolia Efficacy on Gut Metabolites
113 were statistically significant (p < 0.05). Of 165 upregulated biochemicals, 62 biochemicals and 57 biochemicals
were related to lipid and amino acid metabolism, respectively.
Otherwise, 19, 16, and 27 biochemicals of 113 down-regulated
biochemicals were identified as metabolites related to amino
acids, carbohydrates, and lipids, respectively.
S values for components 1 and 2 were 31.90 and 13.37%,
respectively. Data for metabolites of intestinal contents of
chickens fed the unsupplemented diet clustered on the left side
of the PCA score plot, while those of the magnolia-supplemented
diet were separated and clustered on the right side, indicating
that feeding chickens with the magnolia bark extract could
induce significant changes in the intestinal metabolomic profile,
compared with feeding the unsupplemented diet.
Intestinal Metabolite Signatures and
Biochemical Importance Analyses
Dietary Magnolia Bark Extract
Supplementation Alters Intestinal
Metabolite Levels
In RFA, metabolites of amino acids (40.0%), lipids (26.7%),
nucleosides (13.3%), peptides (13.3%), vitamins and cofactors
(0.03%), and carbohydrates (0.03%) accounted for the majority of
biochemicals classified as the most important for distinguishing
between the two treatment groups (Figure 2A). RFA of the
control vs. magnolia groups gave a predictive accuracy of 73.3%,
suggesting that these metabolites are candidate biomarkers for
distinguishing between the two groups. Of 8 magnolia group
samples, 6 were predicted to belong to the magnolia group and 2
to the control group (Table 2). Of 7 control group samples, 5 were
A total of 706 biochemicals were identified in the intestinal
contents of chickens fed an unsupplemented, control diet or
a diet supplemented with magnolia bark extract. Of these, the
levels of 423 metabolites were increased, 278 were decreased,
and 5 were unchanged in the magnolia-supplemented vs.
control groups. Of the 423 increased biochemicals, 165 were
significantly increased, and of the 278 decreased metabolites,
FIGURE 2 | (A) Random forest plots of the top 30 biochemicals whose levels were altered in chickens fed the magnolia bark extract, compared with unsupplemented
controls. Biochemicals are listed from bottom to top in increasing order of importance for contributing to the biochemical signatures separating the two treatment
groups, and are plotted in color-coded symbols according to chemical classification. (B) Heat map showing hierarchical clustering using Ward’s algorithm for the top
30 biochemicals identified by random forest analysis. The clustering along the abscissa (x-axis) is by samples of intestinal contents of chickens fed the
magnolia-supplemented diet (green) or unsupplemented diet (blue) and along the ordinate (y-axis) by specific biochemicals indicated in (A).
Frontiers in Veterinary Science | www.frontiersin.org
4
March 2020 | Volume 7 | Article 157
Park et al.
Magnolia Efficacy on Gut Metabolites
g/chicken), by 10.6% the average body weight gains (50.9 to 56.3
g/chicken), and by 10.1% the final body weights (1828 to 2012
g/chicken), compared with chickens given an unsupplemented
diet (12). Additionally, the prior study demonstrated that
dietary magnolia supplementation improves antioxidant enzyme
activities in the liver, spleen, and breast muscle of chickens,
compared with unsupplemented controls. Pang et al. (15)
reported that catalase and superoxide dismutase activities in
plasma and liver increased following in vivo treatment of mice
with magnolol and honokiol, the major active components
of M. officinalis. Rajgopal et al. (16) showed that magnolia
bark extract activated Nrf2-dependent heme oxygenase-1 gene
expression in vitro in murine hepatocytes. Chang et al.
(17) demonstrated an anti-atherogenic effect of M. officinalis
associated with suppression of oxidative stress, measured by
free radical, malondialdehyde, and oxidative DNA damage,
and down-regulation of apoptosis-related gene expression,
in hyperlipidemic rabbits. To the best of our knowledge,
however, the current study is the first to characterize alterations
in intestinal metabolite levels in poultry fed a magnoliasupplemented diet, compared with unsupplemented controls.
Total U.S. poultry meat consumption more than doubled from
15.5 g/person/day in 1960 to 42.5 g/person/day in 2018 (18).
To meet this growing demand, U.S. broiler growth rates have
doubled and corresponding feed conversion ratios (kg of feed
per kg of body weight gain) have been reduced by approximately
50% over the past 60 years (19). Improvements in nutrition
and genetic breeding stocks, and the introduction of AGPs,
among others, have contributed to increasing the efficiency of
commercial poultry production. In fact, most antibiotics are
manufactured for use in food animal production. In the U.S., for
example, total annual animal antibiotic usage is about 30 billion
pounds, while total human use accounts for <10 billion pounds
per year (20). AGP overuse in food animal production has led
to the emergence of drug-resistant infectious pathogens that
threaten human health (20, 21). As a substitute for AGPs, dietary
supplementation with phytochemicals and micronutrients offers
a potential alternative to increase chicken growth and decrease
the appearance of drug-resistant bacteria (22–24). With the
current experimental results, magnolia showed the potential as
an antibiotic alternative and the intestinal biochemicals produced
due to dietary magnolia supplementation may play a practical
role in improving intestinal health, which may be linked to a
growth-promoting effect in the chicken.
The bark of the magnolia plant, particularly M. officinalis
and M. obovata, has been extensively used in traditional
Chinese medicine for treating respiratory congestion, depression,
and loss of appetite (8). Magnolol and honokiol activate
both the peroxisome proliferator-activated receptor γ (PPARγ)
homodimer and the PPARγ/retinoid X receptor heterodimer
(25). PPARs comprise a family of transcription factors that
bind to peroxisome proliferator hormone response elements
in gene promoters, thereby regulating cellular differentiation,
development, and metabolism (26). Synthetic and naturallyoccurring PPAR agonists decrease triglyceride and blood glucose
levels, and are used for treating metabolic disorders, including
diabetes (27). Magnolol and honokiol are neolignans, two C6 C3
TABLE 2 | Random forest analysis of the altered biochemicals distinguishing
between the control (n = 7) vs. magnola-supplemented (n = 8) dietary groups.
Predicted group
Actual group
Class error
Control
Magnolia
Control
5
2
28.6%
Magnolia
2
6
25.0%
Predictive accuracy = 73.3%.
predicted to belong to the control group and 2 to the magnolia
group. Hierarchical clustering showed two distinct clusters based
on their abundance in the control and magnolia supplementation
groups (Figure 2B).
Specific Intestinal Metabolites Altered
Following Dietary Magnolia Bark Extract
Supplementation
Of biochemicals related to amino acid metabolism, the
levels of tyramine, 2-methylbutyrylglycine, saccharopine,
5-aminovalerate, and methionine sulfoxide were increased
12.2-, 3.92, 2.40-, 2.15-, and 1.46-fold, respectively, in the
intestinal contents of chicken fed the magnolia-supplemented
diet, compared with unsupplemented controls (Figure 3). On
the other hand, 3-(4-hydroxyphenyl)lactate, phenyllactate,
3-hydroxyisobutyrate, 2-hydroxy-3-methylvalerate, 5-oxoprolin,
N2 N6-diacetyllysine, and N-acetylthreonine were measured
at levels 0.15-, 0.21-, 0.23-, 0.38-, 0.38-, 0.46-, and 0.59-fold
less than those in the unsupplemented controls (Figure 3). Of
biochemicals related to lipid metabolism, 2-hydroxypalmitate,
cholate sulfate, margarate, nervonate, and behenate levels
were increased 5.46-, 4.79-, 4.43-, 3.08-, and 2.34-fold,
respectively, in the magnolia-supplemented group, compared
with unsupplemented controls, while glycerophosphoglycerol,
glycerophosphoserine, and chiro-inositol were decreased 0.14-,
0.35-, and 0.45-fold, respectively (Figure 4A). Of biochemicals
related to peptide metabolism, enterodiol levels were increased
9.09-fold, and those of erythritol, secoisolariciresinol, and
naringenin were decreased 0.21-, 0.28-, and 0.39-fold in the
magnolia vs. control groups (Figure 4B). Of biochemicals
related to nucleoside metabolism, the levels of adenine, cytosine,
5-methylcystine, and orotidine were decreased 0.19-, 0.29-, 0.30-,
and 0.36-fold in the magnolia vs. control groups (Figure 4C).
Finally, erythronate, classified as carbohydrate metabolite,
was decreased 0.56-fold, whereas 5-(2-hydroxyethyl)-4methylthiazole, related to vitamin metabolism, was increased
31.1-fold in the intestinal contents of chicken fed the magnoliasupplemented diet, compared with unsupplemented controls
(data not shown).
DISCUSSION
We previously reported that dietary supplementation of broiler
chickens from 0 to 35 days of age with an extract of magnolia
bark increased by 8.2% the average daily feed intake (79.3 to 85.8
Frontiers in Veterinary Science | www.frontiersin.org
5
March 2020 | Volume 7 | Article 157
Park et al.
Magnolia Efficacy on Gut Metabolites
FIGURE 3 | Box-and-whisker plots of the levels of amino acids in the intestine of chickens fed an unsupplemented, control diet (green) or a diet supplemented with
magnolia bark extract (blue). The boxes represent the interquartile range (IQR) defined by the 25th and 75th percentiles. The horizontal line represents the median
value. The cross represents the mean value. The upper whisker represents Q3 + (1.5 × IQR), while the lower whisker represents Q1 – (1.5 × IQR). Circles represent
outliers.
bacitracin-supplemented diet, compared with unsupplemented
controls. Similar to the current investigation, the most common
metabolites altered by AGP supplementation included those of
amino acids, fatty acids, nucleosides, and vitamins/cofactors.
However, there was little overlap in the individual chemical
compounds that were identified as the most dramatically
increased or decreased between the two studies. For example,
among the metabolites of amino acids most highly elevated
in both the virginiamycin- and magnolia-supplemented diets
were those related to lysine. Individually, however, while the
levels of N6-formyllyisne, 5-hydroxylysine, and 2-aminoadipate
were increased in the intestine of virginiamycin-supplemented
chickens, the levels of saccharopine and 5-aminovalerate were
most elevated in chickens given the magnolia-supplemented diet.
A similar pattern was noted for the other chemical compounds
identified in the two studies. These results may imply that
intestinal metabolites have a more direct role of maintaining
intestinal homeostasis than the intestinal microbiome because
they are final products produced through the intestinal
microflora. Therefore, whereas these combined results suggest
a strategy that might be used to improve poultry growth
performance without the use of in-feed antibiotics, additional
studies are needed to identify specific metabolic alterations that
might be exploited to increase chicken growth in the absence of
AGP supplementation.
In summary, this report demonstrates that dietary
supplementation with magnolia bark extract has profound
effects on the levels of a wide variety of chemical metabolites
in the chicken intestine, particularly those related to amino
acids, fatty acids, peptides, and nucleosides. Compared
with unsupplemented controls, these altered metabolite
levels provide a biochemical signature unique to magnolia
units which are not β-β (8-8’)-linked (28). Magnolol (4-allyl-2(5-allyl-2-hydroxy-phenyl)phenol) exhibits a unique mechanism
of binding to PPARγ (29). Two magnolol molecules occupy
the PPARγ ligand-binding domain in a cooperative manner.
One hydroxyl group of the first magnolol molecule forms a
hydrogen bond with Ser-289 of PPARγ and hydrogen bonds
with Tyr-473, while another hydroxyl group of the second
magnolol molecule forms a hydrogen bond with Ser-342 with
additional hydrogen bonding. Downstream pathways that have
been identified following magnolol/honokiol binding to their
cognate receptors include the NF-κB/MAPK, Nrf2/HO-1, and
PI3K/Akt signaling cascades (30). Base on the aforementioned
statement, the intestinal biochemicals in the current study might
stimulate host intestinal epithelial and/or immune cells receptors.
Future in vitro and/or in vivo studies of these or other receptors
may be required to further understand a mechanism of these
biochemicals in the intestine.
We observed that chickens given a diet containing Bacillus
subtilis-based probiotics had a lower number of metabolic
changes compared with the present report (31). More specifically,
compared with unsupplemented controls, the levels of 83
metabolites were altered (25 increased, 58 decreased) in chickens
given a diet supplemented with B. subtilis strain 1781, while 50
were altered (12 increased, 38 decreased) with a B. subtilis strain
747-supplemented diet. On the other hand, the metabolite level
changes seen in the current study (165 increased, 113 decreased)
are comparable to those we previously reported when analyzing
intestinal metabolites in chickens fed a diet supplemented with
the AGPs, virginiamycin and bacitracin (13). In the AGP study,
the levels of 218 biochemicals were altered (156 increased, 62
decreased) in chickens given a virginiamycin-supplemented diet,
while 119 were altered (96 increased, 23 decreased) with the
Frontiers in Veterinary Science | www.frontiersin.org
6
March 2020 | Volume 7 | Article 157
Park et al.
Magnolia Efficacy on Gut Metabolites
FIGURE 4 | Box-and-whisker plots of the levels of (A) fatty acids, (B) peptides, and (C) nucleosides in the intestine of chickens fed an unsupplemented control diet
(green) or a diet supplemented with magnolia bark extract (blue). The boxes represent the interquartile range (IQR) defined by the 25th and 75th percentiles. The
horizontal line represents the median value. The cross represents the mean value. The upper whisker represents Q3 + (1.5 × IQR), while the lower whisker represents
Q1 – (1.5 × IQR). Circles represent outliers.
chicken growth. Through future in vitro and/or in vivo
studies, identification of the altered metabolites that confer
properties of AGPs would suggest their potential use as
antibiotic alternatives.
bark extract supplementation. Our result suggest that
altered metabolites can be used to maintain intestinal
homeostasis within epithelial or immune cells, which might
account for their effects on overall gut health as well as
Frontiers in Veterinary Science | www.frontiersin.org
7
March 2020 | Volume 7 | Article 157
Park et al.
Magnolia Efficacy on Gut Metabolites
DATA AVAILABILITY STATEMENT
AUTHOR CONTRIBUTIONS
All datasets generated for this study are included in the
article/supplementary material.
SO and HL designed the research. SO and HL conducted
research. IP, EL, and HL analyzed data, wrote the paper and have
responsibility for its content. All authors read and approved the
final manuscript.
ETHICS STATEMENT
FUNDING
The animal study was reviewed and approved by The
Institutional Animal Care and Use Committee of the Beltsville
Agriculture Research Center.
This
work
was
supported
Project 8042-32000-107-00D.
REFERENCES
ARS
CRIS
14. Worley B, Powers R. Multivariate analysis in metabolomics. Curr
Metabolomics.
(2013)
1:92–107.
doi:
10.2174/2213235X113010
10092
15. Pang YL, Han XF, Bamikole MA, Gong ZH, Tang SX, Tan ZL, et al. Antidiarrhea and anti-oxidant properties of magnolol. Trop J Pharm Res. (2013)
12:85–91. doi: 10.4314/tjpr.v12i1.14
16. Rajgopal A, Missler SR, Scholten JD. Magnolia officinalis (Hou
Po) bark extract stimulates the Nrf2-pathway in hepatocytes
and protects against oxidative stress. J Ethnopharmacol. (2016)
193:657–62. doi: 10.1016/j.jep.2016.10.016
17. Chang WC, Yu YM, Hsu YM, Wu CH, Yin PL, Chiang SY, et al. Inhibitory
effect of Magnolia officinalis and lovastatin on aortic oxidative stress and
apoptosis in hyperlipidemic rabbits. J Cardiovasc Pharmacol. (2006) 47:463–
8. doi: 10.1097/01.fjc.0000211708.03111.6e
18. Per Capita Consumption of Poultry and Livestock, 1960 to Forecast 2020,
in Pounds, National Chicken Council. (2019). Available online at: https://
www.nationalchickencouncil.org/about-the-industry/statistics/per-capitaconsumption-of-poultry-and-livestock-1965-to-estimated-2012-inpounds/ (accessed March 13, 2020).
19. Thiruvenkadan A, Prabakaran R, Panneerselvam S. Broiler breeding strategies
over the decades: An overview. World’s Poult Sci J. (2011) 67:309–
36. doi: 10.1017/S0043933911000328
20. Marshall BM, Levy SB. Food animals and antimicrobials: impacts on
human health. Clin Microbiol Rev. (2011) 24:718–33. doi: 10.1128/CMR.00
002-11
21. Kumar S, Singh BR. An overview of mechanisms and emergence of
antimicrobial drug resistance. Adv Anim Vet Sci. (2013) 1:7–14.
22. Ademola IO, Ojo PO, Odeniran PO. Pleurotus ostreatus extract inhibits
Eimeria species development in naturally infected broiler chickens. Trop Anim
Health Pro. (2019) 51:109–17. doi: 10.1007/s11250-018-1665-9
23. Wunderlich F, Al-Quraishy S, Steinbrenner H, Sies H, Dkhil MA.
Towards identifying novel anti-Eimeria agents: Trace elements, vitamins,
and plant-based natural products. Parasitol Res. (2014) 113:3547–56.
doi: 10.1007/s00436-014-4101-8
24. Quiroz-Castañeda RE, Dantán-González E. Control of avian coccidiosis:
Future and present natural alternatives. Biomed Res Int. (2015)
2015:430610. doi: 10.1155/2015/430610
25. Wang L, Waltenberger B, Pferschy-Wenzig EM, Blunder M, Liu X,
Malainer C. Natural product agonists of peroxisome proliferator-activated
receptor gamma (PPARγ): a review. Biochem Pharmacol. (2014) 92:73–
89. doi: 10.1016/j.bcp.2014.07.018
26. Michalik L, Auwerx J, Berger JP, Chatterjee VK, Glass CK, Gonzalez FJ, et al.
International Union of Pharmacology. LXI. Peroxisome proliferator-activated
receptors. Pharmacol Rev. (2006) 58:726–41. doi: 10.1124/pr.58.4.5
27. Staels
B,
Fruchart
JC.
Therapeutic
roles
of
peroxisome
proliferator-activated receptor agonists. Diabetes. (2005) 54:2460–
70. doi: 10.2337/diabetes.54.8.2460
28. Gottlieb OR. The rational search for natural neolignans. Mem Inst Oswaldo
Cruz. (1991) 86(Suppl. 2):25–9. doi: 10.1590/s0074-02761991000600009
1. Ventola CL. The antibiotic resistance crisis: Part 1: causes and threats. PT.
(2015) 40:277–83.
2. Ventola CL. The antibiotic resistance crisis: Part 2: management strategies and
new agents. PT. (2015) 40:344–52.
3. Martin MJ, Thottathil SE, Newman TB. Antibiotics overuse in
animal agriculture: a call to action for health care providers. Am
J Public Health. (2015) 105:2409–10. doi: 10.2105/AJPH.2015.
302870
4. Moore PR, Evenson A, Luckey TD, McCoy E, Elvehjem CA, Hart EB. Use of
sulfasuxidine, streptothricin, and streptomycin in nutritional studies with the
chick. J Biol Chem. (1946) 165:437–41.
5. Gadde U, Kim WH, Oh ST, Lillehoj HS. Alternatives to antibiotics
for maximizing growth performance and feed efficiency in poultry: a
review. Anim Health Res Rev. (2017) 18:26–45. doi: 10.1017/S146625231
6000207
6. Lillehoj H, Liu Y, Calsamiglia S, Fernandez-Miyakawa ME, Chi F, Cravens
RL, et al. Phytochemicals as antibiotic alternatives to promote growth
and enhance host health. Vet Res. (2018) 49:76. doi: 10.1186/s13567-0180562-6
7. Grant A, Gay CG, Lillehoj HS. Bacillus spp. as direct-fed microbial
antibiotic alternatives to enhance growth, immunity, and gut health in
poultry. Avian Pathol. (2018) 47:339–51. doi: 10.1080/03079457.2018.14
64117
8. Lee YJ, Lee YM, Lee CK, Jung JK, Han SB, Hong JT. Therapeutic applications
of compounds in the Magnolia family. Pharmacol Ther. (2011) 130:157–
76. doi: 10.1016/j.pharmthera.2011.01.010
9. Prasad R, Katiyar SK. Honokiol, an active compound of magnolia
plant, inhibits growth, and progression of cancers of different organs.
Adv Exp Med Biol. (2016) 928:245–65. doi: 10.1007/978-3-31941334-1_11
10. Poivre M, Duez P. Biological activity and toxicity of the Chinese herb
Magnolia officinalis Rehder & E. Wilson (Houpo) and its constituents.
J Zhejiang Univ Sci B. (2017) 18:194–214. doi: 10.1631/jzus.B16
00299
11. Ranaware AM, Banik K, Deshpande V, Padmavathi G, Roy NK, Sethi G,
et al. Magnolol: A neolignan from the magnolia family for the prevention
and treatment of cancer. Int J Mol Sci. (2018) 19:2362. doi: 10.3390/ijms1908
2362
12. Oh S, Gadde UD, Bravo D, Lillehoj EP, Lillehoj HS. Growth-promoting
and antioxidant effects of magnolia bark extract in chickens uninfected
or co-infected with Clostridium perfringens and Eimeria maxima as
an experimental model of necrotic enteritis. Curr Dev Nutr. (2018)
2:nzy009. doi: 10.1093/cdn/nzy009
13. Gadde UD, Oh S, Lillehoj HS, Lillehoj EP. Antibiotic growth
promoters virginiamycin and bacitracin methylene disalicylate
alter the chicken intestinal metabolome. Sci Rep. (2018)
8:3592. doi: 10.1038/s41598-018-22004-6
Frontiers in Veterinary Science | www.frontiersin.org
by
8
March 2020 | Volume 7 | Article 157
Park et al.
Magnolia Efficacy on Gut Metabolites
29. Zhang H, Xu X, Chen L, Chen J, Hu L, Jiang H, et al.
Molecular determinants of magnolol targeting both RXRα and
PPARγ. PLoS ONE. (2011) 6:e28253. doi: 10.1371/journal.pone.00
28253
30. Zhang J, Chen Z, Huang X, Shi W, Zhang R, Chen M, et al. Insights
on the multifunctional activities of magnolol. Biomed Res Int. (2019)
2019:1847130. doi: 10.1155/2019/1847130
31. Park I, Zimmerman NP, Smith AH, Rehberger TG, Lillehoj EP,
Lillehoj HS. Dietary supplementation with Bacillus subtilis direct-fed
microbials alters chicken intestinal metabolite levels. Front Vet Sci. (2020)
7:123. doi: 10.3389/fvets.2020.00123
Frontiers in Veterinary Science | www.frontiersin.org
Conflict of Interest: The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could be construed as a
potential conflict of interest.
Copyright © 2020 Park, Oh, Lillehoj and Lillehoj. This is an open-access article
distributed under the terms of the Creative Commons Attribution License (CC BY).
The use, distribution or reproduction in other forums is permitted, provided the
original author(s) and the copyright owner(s) are credited and that the original
publication in this journal is cited, in accordance with accepted academic practice.
No use, distribution or reproduction is permitted which does not comply with these
terms.
9
March 2020 | Volume 7 | Article 157
|
|
https://openalex.org/W1982927559
|
https://zenodo.org/records/1893869/files/article.pdf
|
German
| null |
Versuche über Entstehung von Nebel bei Wasserdampf und einigen anderen Dämpfen
|
Annalen der Physik
| 1,907
|
public-domain
| 8,913
|
1) Kurze Berichte von F. Richarz in den Marburger Sitz.-Ber. vom
Juli 1905. p. 92 und vom 11. Juli 1906. p. 123.
2) R. v. Helmholtz, Wied. Ann. 27. p. 508-43. 1886; vgl. auch
H.v.Helmholta, Vorl. 6. p. 334.
3) W. T h o m s o n , Phil. Mag. (4) 42. p. 448. 1871; vgl. auch
H. v. Helmholtz, Vorl. 6. p. 336.
4) J. J. Thomson, Elektrizitiitsdurchgang in Gasen.
Deutsch
vow E. Mars. Teubner, 1906. p, 150. Dort sind auch die anderen,
Literaturangaben zu finden. 317 317 1) J. K i e s s l i n g , Die Dlimmerungsersch. im Jahre 1883 und ihre
physik. Erkl. L. Voss, Hamburg 1885 u. 1888; Pogg. Lex. 4. p. 748. (Aus der Marburger Dissertation vorn 4. MPrz 1906.)') R. v. Helmholtzz) hat eine Formel fur die Spannung
von Dampfen uber beliebig stark gekriimmten Oberflachen
hergeleitet, aus welcher er durch vereinfachende Annahmen
zu der von Sir Will. Thomson3) fur achwache Krummung
abgeleiteten Formel gelangt. Umgekehrt kann man auch die
Tho m s on sche SchluBweise, in prinzipiell derselben Weise,
wie die barometrische HBhenmeBformel, so erweitern, daB
man zu der R. v. Helmholtzschen Formel gelangt, wie in
meiner Inauguraldissertation p. 8 entwickelt ist. g
p
Aus dieser Formel ist bekanntlich zu ersehen, daB sehr
kleine Tropfen auBerst schwer entstehen werden. Notwendig
ist daher die Anwesenheit von Kernen, urn die sich der Dampf
in Tropfenform kondensieren kann. Als solche Kerne kSnnen
zunachst Staubteilchen und Ionen dienen. Nach J. J. Thomsons") Rechnung mu8 jedes Ion in einer
mit Wasserdampf gesattigten Atmosphare einen Wassertropfen
urn sich bilden, dessen Radius auBerordentlich schnell bis zur
GJeichgewichtsgroBe T = 1 : (3,2.10') cm wachst, wenn man
als Ladung ein Elementarquantum und die Kapillarkonstante
T = 76 setzt. Dem Gewichte eines Tropfchens dieser GriiBe 318 B, Barkow, entsprechen, wie sich leicht ergibt, iiur 10 Wasserdampfmole-
kiile. Man kann also wohl mit gleichem Rechte diese Gebilde
als Wassertropfchen
oder als Molionen ansehen. Solche
Tropfchen sind natiirlich unsichtbar. Wenn aber dann hin-
reichende Ubersattigung erzeugt wird, konnen sie wachsen bis
zur Bildung sichtbaren Nebels. Nach C. 2'. R. Wilson kon-
densiert sich das Wasser leichter, bei geringerer Ubersattigung,
an negntiven Ionentropfchen, als an positiven , wahrend nach
I(. Przibram (vgl. diese Abh. pa 343) bei anderen Dampfen
das Urngekehrte eintreten kann. g
In den bisherigen Untersuchungen sind die Unterschiede
im optischen Verhalten der Nebel nicht geniigend beachtet
worden. Bei meitien experimentellen Untersuchungen achtete
ich daher besonders darauf, ob der Nebel nach der J. Kiess-
lin gschen Definition l) homogen war und infolgedessen dann
farbenprachtige Beugungsringe in durchfallendem Lichte zeigte. AuBerdem aber ergab sich als von besonderer Wichtigkeit die
Erscheinungsform eines auBerst feinen blauen Nebels, der vollig
verschieden ist von dichtem inhomogenem und homogenem
Nebel, zu diesen letzteren Formen aber fiihren kann. I. Versuchsenordnung. Zur Untersuchung der Nebelbildung wandte ich die Ent-
spannungsmethode an. Das NebelgefaiB (Fig. 1) bestand aus einer
Glaskugel A , die durch einen Gummistopfen mit drei Durch-
bohrungen verschlossen war. Die Entspannung wurde hervor-
gerufen durch Verbindung mit einem evakuierten GefaB 7
Das etwa 1 cm weite Verbindungsrohr zwischen P und A trug
einen Metallhahn B, mit weiter Bohrung. Der Druck in A
wurde an dem Quecksilbermanometer XI abgelesen. Das Ge-
fa6 7, das die verdiinnte Luft enthielt, war eine Kugelflasche
aus Glas von 40 cm Durchmesser; der Grad der Luftver-
diinnung konnte an dem Metallvakuummeter BZ abgelesen
werden. P stand durch R mit einem groBen luftverdiinnten
Raum, bestehend aus funf Kugelflaschen von 30 cm Durch-
messer, in Verbindung. In diesem wurde die Luft durch eine Entstehuny
Nebel bei T
d
f
t
319 319 Entstehuny von Nebel bei Tasserdampf etc. Luftpumpe verdunnt. Diese komplizierte Einrichtung wurde
getroffen, um eine langere Versuchsreihe unter denselben Be-
dingungen ausfuhren zu konnen, obne fiir Erneuerung der
Luftverdunnung Sorge tragen zu mussen. -2z
--L
Fig. 1. Fig. 1. Nachdem die Luft aus dem Vakuumreservoir R ausgepumpt
war, konnte durch einfache Offnung des Hahnes H3 das Ge-
fa6 B bis auf den gewunschten Druck gebracht werden. D a m
wurde H3 geschlossen und HI geoffnet, bei geschlossenem H2. Dadurch wurde der Druck in A vermindert. Durch Probieren
lieS sich leicht herausfinden, wie groB der Druck in P sein
muBte, urn in A die gewunschte Entspannung zu erhalten. In P wurde dann durch Offnung von H3 wieder der vor dem
Versuch herrschende Druck hergestellt. Die aus A abgecsaugte
Luft wurde nach einem Versuch durch Luft ersetzt, die durch
das Wattefilter I?! gegangen war. Bei allen folgenden Ver-
suchen war es von groBter Wichtigkeit, daB die Luft in A
vollkommen gesattigt war. Wenn die Versuche schnell hinter-
einander gemacht wurden, genugte dazu nicht, daB sich in B
eine Wassermenge B befand. Es wurde daher folgende Ein-
richtung getroffen. 320 Em Barkow. 1) Bei den Beugungsfarben in homogenem Nebel lassen sich drei
Teile unterscheiden:
I. Die Farbung der Lichtquelle selbst, die aber erst bei sehr dicker
Nebelschicht sichtbar wird:
11. Das zentrale Feld, d. h. die Farbung der Umgebung der Licht-
quelle bis zum ersten Beugungsringe, das zentrale Feld ist bei griileren
Tropfen > 9-10 p wei6, bei kleineren Tropfen farbig.
111. Die Beugungsringe selbst.
2) 0. Lummer, Verh. d. Deutsch. Physik. Ges. 6. p. 142. 1904.
3) J. Kiessling und E. Barkow, Marburger Sitzungsber. 1904.
Nr. 6; Verhandl. d. Deutsch. Physik. Gesellsch. 7. Nr. 1 u. 2. 1905.
Annslen der Phpdk. IV. Folge. 23.
21 Em Barkow. 32 1 differenzen von 40 cm und mehr trat bei mir nie ein so dichter
Nebel auf, daB das zentrale Feld gefarbt gewesen ware.l) Es
mag dies seine Ursache darin gehabt haben, daB Wilson sehr
kleine GefaBe benutzte, In solchen ist eine auBerordentlich
schnelle Entspannung moglich. Dabei kijnnen sehr viele Kerne
ergriffen werden, wahrend bei langsameren Entspannungen, wie
in meinen groBeren Gef aBen, die grijBeren der vorhandenen
Kerne rasch wachsen und die kleineren nicht mehr aufkommen
lassen, so da8 also der resultierende Nebel geringere Dichtig-
keit hat. Der Nachteil der kleinen GefaBe ist aber der, dal3
die Wande schnell zu Niederschlag der entstehenden Kon-
densation und Verschwinden der Kerne, besonders der Ionen,
Gelegenheit geben. Bei den groBen (etwa 20 cm Durch-
messer) GefaBen, wie ich sie verwandte, ist der EinfluB der
Wande zu vernachlassigen. Der Nachteil der langsameren
Entspannungen wird aufgehoben durch die Miiglichkeit, sekun-
dare Erscheinungen zu verfolgen, die sich in kleinen GefaBen
nicht ausbilden konnen. Meine Untersuchungen erstreckten
sich nicht so sehr auf die primare Ionenkondensation, als viel-
mehr auf diese sekundaren Erscheinungen. Daher waren fur
mich die groBen GefaiBe nur von Vorteil. Em Barkow. In einer griiBeren Kochflasche wurde Wasser zum Sieden
gebracht. Die von L her angesogene Luft mischte sich mit
dem heiBen Wasserdampfe. Das Gemisch strijmte weiter in
eine Woulffsche Flasche, wo sich ein gro6er Teil des kon-
densierten Wassers absetzen konnte, durch ein Glaswollefilter
in eine zweite Woulffsche Flasche, in der sich der Rest des
kondensierten Wassers sammelte, und schlieBlich durch 1; und
das Wattefilter P2 (Fig. 1) in das NebelgefaB 8. Auf ihrem
langen Wege hatte sich die heiB gesattigte Luft wieder auf
Zimmertemperatur abgekuhlt und war nun naturlich erst recht
vollkommen mit Wasserdampf gesattigt. Das Filter F, be-
stand aus einem 35 cm langen, 3 cm weiten Glasrohr, das auf
beiden Seiten mit durchbohrten Gummistopfen verschlossen und
mit Watte dicht vollgestopft war. Um die Wirksamkeit dieses
Filters zu prufen, saugte ich durch dasselbe Flammengase von
einem Bunsenbrenner in das NebelgefaB, ohne eine Spur ver-
starkter Nebelbildung erkennen zu konnen , obwohl bei einem
Versuche sobar die erste Watteschicht Feuer fing. Da Flammen-
gase sonst bekanntlich sehr kraftig auf die Nebelbildung ein-
wirken, so ergab sich, daB das Filter jede Spur von Konden-
sationskernen wegfing. g
g
Bei geringen Entspannungen bis ca. 15 cm Quecksilber -
entsprechend einem Expansionsverhaltnis von 1,25 - zeigen
sich nur vereinzelte Tropfen, die nach C. T. R. Wilson keine
Ionentropfen sind. Werden die Entspannungen etwas groBer
als 15 cm, so zeigen sich zahlreichere Tropfen zunachst durch
Kondensation an den negativen, dann auch an den positiven
Ionen. Bei sehr starken Entspannungen auch staub- und
ionenfreier Luft treten dichte Nebel auf. Diese starken Nebel
bilden sich nach C. T. R. Wilson bei 21 cm Entspannung -
entsprechend 1,38 Expansionsverhaltnis -, wiihrend C. B a r u s
erst bei 26 cm diese starke Vermehrung der Tropfenzahl be-
obachtete. Ich selbst konnte stets oberhalb 21 crn Druck-
erniedrigung diesen dichteren Nebel konstatieren, aber keine
weitere Steigerung bei 26 cm. Auch die aufierordentliche
Steigerung der Nebeldichte und damit (da bei gr66erer Anzahl
die einzelnen Tropfchen kleiner sind) den weit grijBerer Wert
des Radius der Beugungsringe, wie sie C. T. R. Wilson be-
schreibt, habe ich nie wahrnehmen konnen. Selbst bei Druck- Entsteiiung von Nebel 6ei Wasserdampf etc. 1) Bei den Beugungsfarben in homogenem Nebel lassen sich drei
Teile unterscheiden: 1) R. v. Helmholtz u. F. Richarz, Wied. Ann. 40. p. 175ff. 1890. 11. Versuche uber die Entstehung und die Art des Nebels
in Wasserdampf. 1. E i n w i r k u n g ein es e 1 e k t r is c h en We c h s elf el d es. Den AnlaB zu diesen Versuchen gab eine Bemerkung von
Lummer2), wonach ein ,,elektrischer Nebel" nicht homogen sei
und also auch keine Beugungsringe zeige. Bei der naheren Unter-
suchung dieser Frage fand ichS), daB zwar bei der ersten 21 322 E. Barkow. Entspannung der Nebel nicht homogen war, wohl aber bei den
folgenden. Entspannung der Nebel nicht homogen war, wohl aber bei den
folgenden. g
Das oben beschriebene NebelgefaB ruhte auf einer groBen
Hartgummiplatte zwischen zwei vertikal stehenden Metall-
platten. Zur Erzeugung des Wechselfeldes wurden sie mit
den Polen eines Induktoriums verbunden. Der Abstand der
Platten betrug 21 cm, die Schlagweite des Induktors 12 cm. Als Unterbrecher diente zuerst ein Quecksilberunterbrecher
und spater ein Wehneltunterbrecher, der in spaterer Zeit immer
benutzt wurde, weil seine Wirkung in derselben Zeit intensiver
war. Prinzipiell dieselbe Anordnung wurde iibrigens schon
von R o b e r t v. Helmholtz und F. Richarz') bei den Unter-
suchungen uber den Dampfstrahl angewandt , ebenso spater
auch von Barus. Nach dem Ingangsetzen des Induktoriums gehen zwischen
den Platten und den ihnen zunachst liegenden Teilen des
NebelgefaBes zahlreiche Lichtbuschel uber. Im Innern des
BeobachtungsgefaBes selbst sind keine Lichterscheinungen wahr-
nehmbar. Nachdem die Luft auf diese Weise etwa 10 Sek. elektri-
siert worden war, wurde eine Entspannung vorgenommen. Der
entstandene Nebel ist irisierend, d. h. wenn man durch ihn
gegen eine entfernte Lichtquelle sieht, so tritt ein einige Zeit
anhaltender, schneller Wechsel zwischen hauptsachlich roten
und gl'iinen Farbentonen ein. Dabei ist der Nebel in stark
wirbelnder Bewegung. Allmahlich horen die Wirbel und damit
auch das Irisieren auf, und es zeigt sich nur die Lichtquelle
von einem braunroten Ring umgeben. Diese Erscheinung zeigt
deutlich, daD nur in kleinen Partien des Nebels die Tropfen
gleiche GroBe haben, und jede von diesen zeigt eine besondere
Farbe; durch die wirbelnde Bewegung werden sie abwechselnd
in die Sehrichtung gebracht und rufen ao das Irisieren hervor. Dieser inhomogene Nebel setzte sich verhaltnismaBig schnell. g
g
Es blieb aber, nachdem dieser Nebel verschwunden war,
noch ein feiner blauer Nebel ubrig, der nur in dem Licht-
kegel einer Bogenlampe bei sonst verdunkeltem Zimmer sicht-
bar war. Der blaue Nebel trat hin und wieder auch ohne Entstehung von Nebel bei Jicsserdampf etc. 11. Versuche uber die Entstehung und die Art des Nebels
in Wasserdampf. 323 jegliche Entspannung auf, besonders dann , wenn die Elektri-
sierung sehr stark war. Er verschwand auch dann nicht, wenn
nach der Entspannung wieder filtrierte Luft zustromte, und
dadurch adiabatische Erwarmung stattfand. Ebenso senkte er
sich nicht merklich und war noch sehr lange nach seinem
Entstehen vorhanden. Er besteht aus auBerordentlich kleinen
Tropfchen, wie man aus seiner geringen Fallgeschwindigkeit,
seiner blauen Farbe und aus dem Fehlen von Beugungs-
erscheinungen erkennt ; seiner Dichtigkeit nach zu urteilen,
mii6te er Beugungserscheinungen zeigen, da vie1 weniger dichte
Nebel sie sonst hervorbringen. Auf seine Entstehungsart und
seine wahrscheinliche Zusammensetzung ferde ich noch zu
sprechen kommen. p
Jedenfalls hangt mit seiner Bildung die Starke des nachher
bei der Entspannung entstehenden Nebels zusammen; denn
wenn vorher der blaue Nebel deutlich zu sehen war, so ist
auch der Entspannungsnebel stark, im anderen Falle ist er
kaum oder gar nicht wahrnehmbar. Vermutlich ist immer
etwas Ahnliches vorhanden, denn sonst laBt sich die Entstehung
von Nebel bei geringen Entspannungen unter 15 cm schwer
erklaren. Nach den Untersuchungen C. T. R. Wilsons und
anderer tritt die Kondensation an negativen bez. positiven
Ionen nur bei Entspannungen gr66er als 15 cm bez. 21 cm
auf; aufierdem nimmt die Kondensationsfahigkeit der Ionen
sehr schnell mit der Zeit ab, einige Sekunden genugen
schon, sie inaktiv zu machen. Hier jedoch halten sich die
Kondensationskerne viele Stunden. Also konnen es nicht die
Ionen in eigentlichem Sinne sein, die hier die Kondensation
hervorrufen, sondern hiichstens ein sekundares Erzeugnis der
Ionisation. Dafiir spricht ferner die Tatsache, da6 die Zabl
der Kerne auch noch nach dem Aufhoren der Elektrisierung
einige Sekunden lang zunimmt, wahrend die Ionenzahl rapide
abnimmt, wie Ruther ford durch Leitf ahigkeitsuntersuchungen
gezeigt hat. Wenn ich namlich sofort nach dem Aufhiiren
des Elektrisierens entspannte, so war der entstehende Nebel
schwach, wartete ich aber einige Sekunden, so nahm die Dichte
des Nebels und die Zahl der Entspannungen, die nijtig war,
um die Kondensationskerne zu beseitigen, betrachtlich zu und
nahm schlieBlich einen maximalen Wert an. Ganz allmahlich
* 21 * 21 * 324 E. Barkow. nahm dann die Kernzahl ab, after waren aber noch nach etwa
15 Stunden eine groBere Anzahl Kerne vorhanden, so daB
noch ziemlich kraftiger Nebel auftrat. Um von der erzeugten
Kernzahl einen Begriff zu geben, sei angefuhrt, daB in einem
Falle 22 Entspannungen von 7-8
cm notig waren, um sie
samtlich zu entfernen. 1) F. Richarz, Mitteil. d. Naturw. Ver. Greifswald 1896; Wied.
Ann. 69. p. 592-594. 1896. 11. Versuche uber die Entstehung und die Art des Nebels
in Wasserdampf. Wenn daher auch zweifellos im Wechsel-
feld Ionen entstehen, wie ich in meiner Dissertation p. 22, 23
des naheren auseinandergesetzt habe, so konnen diese Ionen
doch primar nicht die Kerne der vorbeschriebenen Konden-
sation sein. Bei sfurkeren Expansionen unmittelbar nach
Wirkung des Wechselfeldes werden natiirlich auch Ionen
primar als Kondensationskerne dienen. p
1st der erste inhomogene Nebel verschwunden, so wird
wieder staubfreie Luft zugelassen und von neuem entspannt. Der dann entstehende Nebel ist meistens homogen und zeigt
prachtige Beugungsbilder. 1st auch dieser gesunken, so zeigt
sich wieder der permanente blaue Nebel, wenn auch schwacher. Die folgenden Entspannungen zeigen im allgemeinen dieselben
Erscheinungen, nur sind die Beugungsringe anders gef arbt
und kleiner (vgl. weiter unten p. 336), der blaue Nebel zeigt
immer abnehmende Intensitat und ist zuletzt nicht mehr wahr-
nehmbar. SchlieBlich ist der Normalzustand erreicht und das
Nebelgefab zu einem neuen Versuch vorbereitet. 2. Einwirkung von Riintgenstrahlen. Die Wirkung von Rontgenstrahlen auf die Kondensation
des Wasserdampfes stelltc zuerst F. Richarz l) fest. Weitere
Untersuchungen liegen vor von (3. T. R. W i 1 son, J. 6. Tho m s on
und Barus. Meine eigenen Versuche uber diesen Gegenstaud ergaben
folgendes: Zwischen dem Nebelgefab aus Glas und der Riintgen-
rohre befand sich ein zur Erde abgeleiteter Aluminiumschirm. Das Induktorium hatte eine Schlagweite von etwa 15 cm. Er-
folg hatte ich erst bei einer fiir Wehneltbetrieb konstruierten
Rontgenrohre mit verstiarkter Antikathode ; ich beanspruchte
sie bis zur Grenze ihrer Leistungsfahigkeit. Hierdurch gelang Entsiehung von Nebel bei Wasserdampf etc. 325 es mir, schon nach einer halben Minute Strahlungsdauer einen
kraftigen Nebel zu erhalten, der deutlich sichtbare, auch objektiv
darstellbare Beugungsringe zeigte. Die GroBe der Entspannung
betrug etwa 8 cm, war also weit geringer als bei Wilson. Ja, dieser sagt sogar ausdruclrlich, selbst sehr starke Strahlung
beeinflusse nur die Tropfenzahl aber nicht die GroBe der Ent-
spannung. Das war also bei meinen Versuchen anders. p
g
Da es sehr unwahrscheinlich ist, daB die Art der Ionen
von der Starke der Strahlung abhangig ist, so sind auch die
bei diesen Versuchen wirksamen Kondensationskerne vermut-
lich nicht Ionen, sondern, entsprechend dem blauen Nebel im
vorigen Abschnitt, etwas, das sekundar durch die Ionen entsteht. Der einzige Unterschied wiirde der sein, daB dort der blaue Nebel
sichtbar ist, wahrend hier ein solcher nicht sichtbar ist. 11. Versuche uber die Entstehung und die Art des Nebels
in Wasserdampf. Wenn
das Wechselfeld so kurze Zeit wirkt, daB der Nebel nur dem
bei Riintgenstrahlen entstehenden an Dichte gleich ist, so ist
auch dort der blaue Nebel noch nicht sichtbar. Auch hier ist der bei der ersten Entspannuhg entstehende
Nebel nicht ganz homogen. 1) P. Lenard u. M. Wolf, Wied. Ann. 37. p. 443-456.
1889.
2) R. v. Helmholtz u. F. Richarz, Wied. Ann. 40. p. 187. 1890. 1) P. Lenard u. M. Wolf, Wied. Ann. 37. p. 443-456.
1889.
2) R. v. Helmholtz u. F. Richarz, Wied. Ann. 40. p. 187. 1890. 1) P. Lenard u. M. Wolf, Wied. Ann. 37. p. 443-456.
1889. 1) R. v. Helmholtz, Wied. Ann. 32. p. 1-19. 1887; Verh. d. Phyeik.
Ceeellsch. Berlin 1886. p. 20-21;
Naturw. Rundschau 1. p. 69-71.
1886. 4. Ein w i r kung el e k t ria c h e r Spit Zen en t la dung. 3. Wirkung von ultrsviolettem Licht. DaB ultraviolettes Licht die Nebelbildung in Wasserdampf
begiinstige, fanden zuerst Lenard und Wolf. l) Sie schrieben
aber diese Wirkung einer durch das ultraviolette Licht hervor-
gerufenen Zerstaubung des Quarzfensters zu, das sie zwischen
Lichtquelle und Dampfstrahl bez. NebelgefaB eingeschaltet
hatten. R. v. Helmholtz und F. Richarz bestritten diese An-
sichta), fiihrten als Grund fur die Verstarkung der Konden-
sation die Ionisation der Luft an und zeigten, daB durch ultra-
violettes Licht Ozonpapier in Luft geblaut wird. Durch viele
Untersuchungea wurde dargetan, daB ultraviolettes Licht wirklich
die Luft ionisiere, unter anderen von Lenard selbst und
C. T. R. Wilson, welche auch die Kondensationswirkung genau
verfolgten. Die wirksamste Quelle ultravioletter Strahlen ist wohl die
Quecksilberbogenlampe aus Quarzglas. 26
E B 26
E B 326 E. Barkow. Das NebelgefBB A (Fig. 1) wurde durch eine dreifach
tubulierte Glaskugel A, wie in Fig. 2, p. 328, ersetzt; a ent-
spricht dem a in Fig. 1 ; b wurde im vorliegenden Palle nicht
gebraucht und war durch einen Gummistopfen verschlossen ;
c war durch eine mit Siegellack aufgekittete Quarzlinse Q be-
deckt. Die Quecksilberbogenlampe war etwa 15 cm yon Q
entfernt. Zuerst wurde A, in der bekannten Art und Weise staub-
frei gemacht und dann ultraviolett bestrahlt. Bei der Ent-
spannung entstand ein auBerordentlich dichter, nicht homogener
Nebel. Der bei weiteren Entspannungen auftretende Nebel
ist homogen und zeigt prachtige Beugungsfarben. Dauert die
Belichtung einige Minuten, so entsteht schon ohne Expansion
ein dichter, fein blaulich glanzender Nebel, der bei starker
Beleuchtung sichtbar ist. Im durchgehenden Lichte waren
bei diesem blaulichen Nebel keine Beugungserscheinungen zu
sehen; die Tropfchen sind also wohl zu klein dazu, vielleicht
kleiner als Lichtwellen ; denn mangelnde Zahl der Tropfchen
kann nicht Ursache des Ausbleibens der Beugungsfarben sein,
vielmehr ist die Dichtigkeit bedeutend groBer als bei vielen
Nebeln, die Beugungsringe hervorrufen, wie der Augenschein
zeigt. Der blaue Nebel ist geschichtet. Im allgemeinen sind
die etwa 1 mm breiten Schichten senkrecht zu den ultra-
violetten Strahlen angeordnet. Diese auch schon von C. T. R. Wilson und J. H. Vincent beobachtete Erscheinung ware
noch genauer zu untersuchen, um ihre Entstehung erklaren
zu konnen. Der blaue Nebel, der hier entsteht, hat dasselbe
Aussehen und wahrscheinlich auch dieselbe Entstehungsursache
wie der blaue Nebel, der durch das Wechselfeld entsteht. Auch hier ist die notwendige EntspannungsgroBe so gering,
daB man nicht Ionen als primare Kondensationskerne annehmen
ham, sondern sekundar durch Ionen gebildete Produkte. 1) R. v. Helmholtz u. F. Richarz, Wied. Ann. 40. p. 161 bis
2) W. Lemme, Dissertation Greifswald 1901; Mitt. d. Naturw. Ver.
3) H. Rebenstorff, Pbysik. Zeitschr. 6. p.571-574.
1904; Naturw.
203. 1890.
Greifswald 33. 1901.
Rundscbau 19. p. 629. 1904. 4. Ein w i r kung el e k t ria c h e r Spit Zen en t la dung. 4. Ein w i r kung el e k t ria c h e r Spit Zen en t la dung. Die erste Beobachtung uber die Wirkung der Spitzen-
entladung auf die Kondensation des Wasserdampfes machte
R. v. He1mholtz.l) Weiter untersucht wurde diese Erschei- Entsteilung von Nebel bei Wasserdampf etc. 327 nung besonders von R. v. Helmholtz und F. Richarzl),
Lemme2) und Rebenst0rff.g Schon R. v. Helmholtz und
F. Richarz hatten nachgewiesen, da% die Wirkung nicht
herruhre von Staub oder Metallteilchen, die von der Spitze
abgeschleudert wiirden, sondern da6 es sich nur um die Wir-
kung von Ionen handeln k6nne. Ferner zeigten sie, da8 das
wirksame Agens eine gewisse, wenn auch kurze Zeit sein
KondensationsvermSgen beibehalte. g
C. T. R. Wilsons Untersuchungen hatten folgendes Er-
gebnis. Wird eine Entspannung vorgenommen, wahrend die
Spitzenausstromung noch fortdauert, so ist die fur Ionen
charakteristische Entspannungsgrofle erforderlich, um Nebel-
bildung hervorzubringen. Wird dagegen erst entspannt, wenn
die Spitzenentladung etwas vorher aufgehort hat, so ist die
zur Nebelbildung notwendige Druckverminderung vie1 geringer. Es bilden sich also neue, anders geartete Kerne nach Auf-
hiiren der Spitzenentladung. Meine eigenen Untersuchungen ergaben folgendes. Der
Nebel bei der ersten Entspannung war nicht homogen, und
erst bei weiteren Entspannungen traten kraftige Beugungs-
bilder auf. Die notwendige GrijBe der Entspannung war sehr
gering, und die Kondensationskerne waren noch nach mehreren
Stunden vorhanden. Dies wurde auch wieder ein Beweis dafur
sein, daB es nicht primar die Ionen sind, die die Nebelbildung
hervorrufen. Einen blauen Nebel, wie er bei dem Wechselfeld und
dem ultravioletten Licht auftritt, konnte ich kaum bemerken. Vermutlich riihrt es daher, da6 er an seiner Ansammlung
durch den elektrischen Wind verhindert wird, der ihn gegen
die Wande wirft. Indessen laljt sich wohl die Beobachtung
von Lemme, daS im NebelgefaB ein feiner Nebel ubrig bleibt,
wenn sich der dichte Nebel gesetzt hat, mit dem blauen Nebel
in Zusammenhang bringen. 328 E. Barkow. 1) Joh. Kiessling, Marburger Sitzungeber. 1904. Nr. 7. 5. Einwirkung des elektrischen Funkens. Die Anregung zu diesen Versuchen gab eine am Vier-
waldstattersee gemachte Beobachtung von J oh. Kiessling,')
die er folgendermaBen beschreibt. ,,Unmittelbar nachdem aus
einer auf dem Uri-Rotstock lagernden, nach unten hin scharf
begrenzten Gewitterwolke ein sehr hell leuchtender Blitz fast
senkrecht zum Seeniveau sich entladen hatte, konnte ein genau
der Blitzbahn entsprechender , schmaler grauer Regenstreifen
beobachtet werden, der 8-10
Sek. lang sich sehr deutlich
von den dunkelviolett erscheinenden Felswanden abhob." Es handelte sich bei den Versuchen darum, diese Er-
scheinung experimentell nachzumachen und die Bedingungen
ihres Auftretens gennuer zu untersuchen. Als VersuchsgefaB diente eine Glaskugel, wie sie Fig. 2 zeigt. Der Tubulus a ist mit der Entspannungsvorrichtung, die Fig. 1 Fig. 2. Fig. 2. zeigt, verbunden. Durch b und c fuhren durch Gummistopfen
und Glasrijhren gut isoliert zwei Drahte, die im Beobachtungs-
gefab in 2 cm grogen Messingkugeln endigen. Die Kugeln sind
mit feinem Musselin oder FlieBpapier umhullt, die mit Wasser
getrankt waren, um jegliche Zerstaubung des Metalls auszu-
schlieben. Die beiden Drahte wurden mit den Polen eines
Induktoriums von etwa 16 cm Schlagweite verbunden. Der
selbsttiatige Unterbrecher wurde ausgeschaltet ; statt dessen Entstehung von Nebel bei Wasserdampf etc. 329 wurde der Prim&rstrom durch einen Quecksilberkontakt ge-
schlossen und geoffnet. LieB ich dann ohne weiteres einen lOcm langen Funken
in staubfreier Luft zwischen den Kugeln iibergehen, so trat
keine sichtbare Nebelbildung auf. Das NebelgefaB wurde durch
Bogenlicht hell beleuchtet, so daB jede Spur von Nebel zu
erkennen gewesen ware. Wurde die Luft nachtraglich ent-
spannt, so zeigte es sich, daB Kondensationskerne vorhanden
waren. Der Nebel war nicht homogen. Folgender Versuch aber hatte sofortigen Erfolg. Die
wieder kernfreie Luft wurde einer so kleinen Entspannung
-
-etwa 4-5
cm Quecksilber - unterworfen, da6 dadurch
kein Nebel entstehen konnte; gleichzeitig damit lie0 ich einen
Funken iibergehen. Es zeigte sich dann zwischen den Elek-
troden ein Biischel von feinen Nebelstreifen, die 1-2
Sek. lang
sichtbar blieben, bis der durch die Funkenwarme erzeugte
Luftstrom sie zerriB. DaB der Nebel-streifen nicht einfach
war, wie bei der Kiesslingschen Gewitterbeobachtung, erklart
sich leicht daraus, daB auch der Funke nicht einfach war,
sondern neben dem glanzenden Hauptfunken blauliche, schwache
Nebenfunken zwischen den Elektroden ubergingen , wie auch
durch direkte Beobachtung des Funkens im Dunkeln zu er-
kennen war. Fur die Nebelbildung wirken sowohl Baupt- als
Nebenfunken. 1) J. Eleter u. H. Geitel, Meteorol. Zeitschr. 1% p. 187-188.
1895; Beibl. 19. p. 874. 1895. 5. Einwirkung des elektrischen Funkens. Die Kernbildung durch den Funken ist so
kraftig, da8 in den Nebelstreifen sich nicht an jedem Kern
ein Tropfen bilden kann, wie die Entstehung von Nebel bei
weiteren Entspannungen zeigt. g Dai3 die Funkengase auf die Kondensation einwirken, ist
schon langere Zeit bekannt. Der Dampfstrahl zuckt jedesmal
auf, wenn ein Funke durch ihn schlagt. Einen ahnlichen Ver-
such wie den oben beschriebenen, haben, veranlaat durch eine
Bemerkung von W. v. Bezold, Elster und Geitel') gemacht. Sie hatten im wesentlichen dieselbe Versuchsanordnung, nur
war die Funkenlange so klein, daB sie keine Einzelheiten
daran erkennen konnten. Sie wollten durch ihren Versuch
auch nur zeigen, daB Luft mit Wasserdampf ubersattigt sein, 330 3: Barkow. und daB die Eondensation aus thersiittigung durch elektrische
Vorgange ausgelijst werden kiinne. W. v. Bezold hatte nam-
lich, um einige Erscheinungen bei Gewittern, z. €5. die soge-
nmnten Gewitternasen erklaren zu kijnnen, die Existenz von
iibersiiitigtem Wasserdamp f auch in der freien Atmosphare
angenommen. Ubrigens hatten fir Laboratoriumsbedingungen
schon R. v. Helmholtz1) und F. R i c h a r z 3 das Bestehen
eines solchen labilen Zustandes der Ubersattigung und dessen
Buslosung durch elektrische Krafte aus dem Verhalten des
Dampfstrahles geschlossen. p
g
Die von Kiessling beobachtete Erscheinung tritt nur
dann auf, wenn die staubfreie Luft mit Wasserdampf uber-
sattigt ist und wird auBerdem gut sichtbar nur, wenn ein
gunstiger Hintergrund vorhanden ist, von dem sich der Nebel-
streifen deutlich abheben kann. Diese Bedingungen kiinnen
in der Natur erfullt sein, wenn durch vorubergehenden Regen
der Staub niedergeschlagen ist und die so gereinigte Luft auf-
steigt und sich dabei adiabatisch abkuhlt. DaB die Luft an
Regentagen arm an Staub ist, ist bekannt. Und damit hangt
auch die hohe Durchsichtigkeit der Luft an solchen Tagen
zusammen. 6. Wirkung von radioaktiven Substanzen. 1) R. v. Helmholtz, Wied. Ann. 27. p. 512. 1886; 32. p. 11. 1887.
2) R. v. Helmholtz u. F. Richarz, Wied. Ann. 40. p. 164,
3) W. Lemme, Dissertation Greifswald 1901.
4) K. Schaum, Marhurger Sitzungsber. 1902. Nr. 8; Zeitschr. f.
Zeile 5 v. 0. 1890; p. 182, Zeile 10 v. u.
wissensch. Photographie 3. p. 239-241.
1905. 4) K. Schaum, Marhurger Sitzungsber. 1902. Nr. 8; Zeitschr. f.
wissensch. Photographie 3. p. 239-241.
1905. 1) E. Warburg, Sitzungsber. d. k. Akad. d. Wissensch. zu Berlin
2) Vgl. F. Richarz, Marburger Sitzungsber. Nr. 7 p. 47-53.
Juli
1903; K. Kliipfel, Dissertation Marburg 1903. p. 41; Ann. d. Physik
16. p. 578. 1905.
1903. p. 1011-1015. 6. Wirkung von radioaktiven Substanzen. Da die von den radioaktiven Substanzen ausgehenden
Strahlen die Luft zu ionisieren vermiigen, so war zu erwarten,
daB sie auch die Kondensation von Wasserdampf beeinflussen
wiirden. Hierher gehorende Versuche machten z. B. mit dem
Dampfstrahl Lemmes) und S ~ h a u m . ~ )
Wie schon Schaum
bei seinen Untersuchungen uber den Dampfstrahl bernerkte,
ist die Kondensationswirkung sehr schwach, vie1 schwacher,
als man nach der entladenden Fahigkeit erwarten sollte. Auch meine Versuche im Nebelgefa6 ergaben, da8 die
Verstarkung des Nebels sehr schwach, ja kaum bemerkbar 33 1 Bntstehung con Nebel hei Fasserdampf etc. war. Untersucht wurden Radiumbromid (in geschlossener
Kapsel, so daB es sich nur um 13-. und y-Strahlen handelte),
ein Polonium enthaltender Stab und eine Platte mit Radio-
tellur. Die Substanzen wurden in das NebelgefaB gebracht
und dann die Luft entspannt. Auch lieB ich das Praparat
mehrere Tage im NebelgefaB und entspannte dam, ohne
besseren Erfolg. Auch wenn ich feuchte Luft eine Woche
und langer in abgeschlossenen GefaBen der Strahlung aussetzte
und diese Luft in das NebelgefaB saugte, erhielt ich keine
anderen Resultate. In diesen Fallen ist also offenbar der Dampfstrahl
emp findlicher. 1) E. Warburg, Sitzungsber. d. k. Akad. d. Wissensch. zu Berlin
1903. p. 1011-1015. 2) Vgl. F. Richarz, Marburger Sitzungsber. Nr. 7 p. 47-53.
Juli
1903; K. Kliipfel, Dissertation Marburg 1903. p. 41; Ann. d. Physik
16. p. 578. 1905.
p 7. Versuche mit Ozon. Wie man sieht, ist die bloBe Anwesenheit von Ionen in
einem Gase nicht geniigend, um so starke Kondensation des
Wasserdampfes wie in den Fallen 1, 3, 4, 5 hervorzubringen,
sondern es kommt dabei wesentlich auf die Art der Ionen
oder auf sekundare Wirkungen yon ihnen an; denn sonst
miifiten z. B. radioaktive Substanzen kriiftiger auf die Nebel-
bildung einwirken. Es liegt nahe, unter den sekundaren Wir-
kungen an die Ozonbildung zu denken, die bei Spitzenaus-
stromung und Funkenbildung auftritt. Hierbei spielt nach
Warburgs Untersuchungenl) das ultraviolette Licht, das bei
der stillen Entladung auftritt, eine Rolle, wenn auch wohl nicht
die primare. 2,
Unter den angewandten Versuchsbedingungen
erzeugt die Quecksilberbogenlampe am meisten Ozon und auch
die starkste Nebelbildung. Es ist bei den oben beschriebenen
Versuchen ein Paralleli-smus zwischen Ozonentstehung und der
Starke der Nebelbildung vorhanden. g
Aus diesen Griinden wurden direkt Versuche mit Ozon
angestellt. Diese Versuche sind noch nicht abgeschlossen. Ich will daher vorlaufig nur folgendes mitteilen. Der blaue
Nebel tritt auch bei der Einwirkung fertigen Ozons sehr 332 E. Barkow. kraftig auf. Hat man in ozonhaltigem Sauerstoff durch einige
Entspannungen alle Kerne niedergeschlagen und iiberlaflt dann
das NebelgefiaB einige Zeit sich selbst im Dunkeln, so sind
wieder zahlreiche Kondensationskerne vorhanden. Starke Be-
lichtung - es muB nicht ultraviolettes Licht sein - erhoht
die spontane Kernbildung betrachtlich. Durch meine Versuche
bin ich zu der Vermutung gelangt, daB hierbei irgendwelche
Oxyde des Stickstoffs, dessen letzte Reste nur sehr schwer
aus dem Sauerstoff zu entfernen sind, eine wesentliche Rolle
spielen. Besfatigt wird diese Vermutung durch Versuche von
Hrn. E. Pringal, der im Marburger physikalischen Institut
damit beschaftigt ist, meine Untersuchungen fortzufuhren. J e
mehr man namlich den benutzten Sauerstoff von Stickstoff,
und je mehr man den ozonisierten Sauerstoff von seinen Ver-
unreinigungen durch Stickoxyde hefreit , urn so mehr nimmt
mit zunehmender Reinheit des Ozons d.ie Nebeldichte ab. 8. Un t er s u c h u n g v o n S a u er s t of f- S t i c ks t o f fv e r b in d ung e n. Von Wichtigkeit ist es unter diesen Urnstinden, die Wirk-
samkeit einiger fertiger Stickoxyde zu untersuchen. In bezug auf den Dampfstrahl ist dies bereits von
R. v. Helmholtz und F. Richarz geschehen; sie fanden, dafl
die Dampfe von rauchender Salpetersaure (NOSH mit N,O,)
sehr kraftig, sowie die nitrosen Gase NO, und N,O, stark wirkten. 7. Versuche mit Ozon. Folgendes sind die mit meiner Versuchsanordnung gewonnenen
Resultate. Ich loste Natriumnitrit in Wasser; diese Liisung zeigte
keine Wirkung. Fuhrte ich Schwefelsaure zu, so entwickelten
sich die rotbraunen Dampfe. Der Nebel, der hei einer Ent-
spannung entstand, war dichter als normal durch die spontane
Ionisation der Luft, aber nicht hesonders stark. Um besser
die VerEaltnisse iiberblicken zu kbnnen, entwickelte ich die
braunen Dampfe Ton StickstofTdi- bez. -tetroxyd nicht im Nebel-
geftlB selbst , sondern gesondert in einer Kugelflasche. Von
dort konnten sie dann leicht in das NebelgefaB gebracht
werden. Zuerst wurde eine Anzahl Entspannungen vorge-
nommen, die starke Nebel gaben, bis eine braune Farbung
im NebelgefaB kaum noch wahrzunehmen war. Reste von NO,
bez. N,O,
werden aber doch noch vorhanden gewesen sein. Entstehuny von Nebel bei Yasserdampf etc. 333 Es war anzunehmen, dab die Nebeldichte wachsen wiirde, wenn
den Nebelkernen Zeit bliebe, sich anzusammeln. Dies zeigte
sich in der Tat. Der Nebel ist nicht homogen; es treten
immer Tropfen von zweierlei GrijBe auf. De verschiedene
groBe Tropfen auch verschieden schnell fallen, so kann man
die beiden Tropfenarten nacheinander beobachten. Die klei-
neren Tropfen zeigen etwa doppelt so groBe Beugungsringe
wie die groaeren. Es war anzunehmen, dab die Nebeldichte wachsen wiirde, wenn
den Nebelkernen Zeit bliebe, sich anzusammeln. Dies zeigte
sich in der Tat. Der Nebel ist nicht homogen; es treten
immer Tropfen von zweierlei GrijBe auf. De verschiedene
groBe Tropfen auch verschieden schnell fallen, so kann man
die beiden Tropfenarten nacheinander beobachten. Die klei-
neren Tropfen zeigen etwa doppelt so groBe Beugungsringe
wie die groaeren. Wird das NebelgefaS mit Bogenlicht fortgesetzt belichtet,
so werden die zuerst wenigen Trijpfchen immer zahlreicher. Bei einer Entspannung ist der jetzt entstehende Nebel auBer-
ordentlich dicht und nicht homogen. Er zeigt oft irisierende
Farbungen. Zu dieser starken Vermehrung des Nebels ist
iibrigens nur auBerst wenig Nz04 notwendig. Bei einem Ver-
suche hatte ich das NebelgefaB mit dicken rotbraunen N,O,-
Dampfen gefiillt. Ich nahm dann im Dunkeln eine Anzahl
Entspannungen vor von etwa 'I3 Atmosphare und groBer. Die
abgesaugte Luft wurde natiirlich jedesmal durch filtrierte Luft
ersetzt. Nachdem ich dies 24 ma1 wiederholt hatte, belichtete
ich 5 Min. und entspannte wieder. Der entstehende Nebel
war augerst dicht. Nach fiinf Exhaustionen sind die Kerne
wieder beseitigt. Ich belichtete wieder 5 Min. und fand wieder
sehr dichten Nebel. Dies wiederholte ich noch oft. Nnch der
65. Entspannung belichtete ich 5 Min. 1) R. v. Helmholtz u. F. Richarz, Wied. Ann. 40. p. 161 bis
2) Vgl. auch d'Arcy, Phil. Mag. (6) 3. ,p. 42-52.
1902.
203. 1890. B. Barkow. Wir haben hier eine weitgehende Analogie zu der Wirkung
des Ozons mit Spuren yon Stickstoff oder nitrosen Gasen. Hier wie dort haben .wir eine spontane Kernbildung, sowie
bei starker Belichtung eine Vermehrung der Kernzahl. Zu
letzterem ist nur eine sehr geringe anfangliche N,O,- bez. 0,-
Menge (mit Stickstoff oder nitrosen Qasen yerunreinigt) nbtig. 7. Versuche mit Ozon. und fand einen Nebel,
dessen Dichtigkeit kaum geringer war als bei friiheren Be-
lichtungen. Ich iiberlieB dann das Gas etwa 20 Stunden sich
selbst am Fenster, also dem Tageslicht ausgesetzt. Durch
einige Expansionen wurden die Kerne beseitigt und das Gas
wieder 5 Min. dem Bogenlicht ausgesetzt. Bei einer Ent-
spannung trat sehr dichter Nebel auf, wenn auch seine Uichte
geringer war als am vorhergehenden Tage. Damit wurde
dieser Versuch abgebrochen. Wenn man jede Entspannung
zu
Atmosphare annimmt, so waren zuletzt nur noch etwa
(z/3)'o des urspriinglichen N,O,
vorhanden , und diese auSerst
geringe Menge geniigte, urn bei Belichtung noch kraftige Wir-
kung zu geben. Die zur Nebelbildung notwendige Entspannung braucht
ubrigens nicht so groS zu sein, wie bei dem obigen Versuch;
es geniigen vielmehr schon ganz geringe Expansionen. 334 B. Barkow. 9. Un t ersu c hung von W ass er B t of fs up er ox y d. Eine andere Verbindung, die sich in dem Luft- bez. Sauerstoff-Ozon-Wasserdampfgemisch unter Umstanden bilden
kbnnte, ist das Wasserstoffsuperoxyd. Es ist deshalb erforder-
lich, auch dieses zu untersuchen. Es ergab sich, daB sich
selbst iiberlassen Wasserstoffsuperoxyd die Nebelbildung nicht
vermehrt. Fur das Dampfstrahlphanomen haben bereits R. v. Helmholtz und F. Richarz l) dasselbe Resultat erhalten. Ai t ken fand, dab durch Sonnenlicht zersetzter H,O,-Dampf
recht kraftige Nebelbildung bei der Entspannung hervorrief,
und die entstandenen Kerne langen Bestnnd hatten. Diesen
Versuch konnte ich bestatigen. Es wirkt aber nur sehr inten-
sives Licht, wie Sonnenlicht, in diesem Sinn. Urn ultraviolettes
Licht handelt es sich hierbei nicht; denn die Strahlen muBten
die Wande des Glasballons passieren. Mit der elektrischen
Bogenlampe konnte ich keinen Nebel erhalten. Ob diese Be-
lichtung des H,O,-Dampfes durch Sonnenlicht auch elektrische
Leitfahigkeit hervorruft, ware noch zu untersuchen. ,) 111. Erkliirung der verschiedenen Formen des Nebels. 1. Homogener und inhomogener Nebel. 1. Homogener und inhomogener Nebel. Wir haben einen Nebel nach der Kiesslingschen Defi-
nition ,,homogen" genannt, wenn alle Tropfen so weit von der-
selben Gro5e sind, daB sie farbige Beugungsringe erkennen
lassen. In den vorhergehenden Versuchen zeigte sich stets,
da5 sehr starke Nebel, wie sie bei der ersten Entspannung
nach kraftigen Einwirkungen auftraten, auch inhomogen waren. Dies erklart sich folgenderma5en. 335 Entstehung von Nebel bei Wasserdampf etc. Ob ein Nebel homogen wird, hangt, wie man sofort sieht,
davon ab , ob die Kondensationskerne gleichartig und gleich-
maBig verteilt sind, so daB der Bereich, aus dem sie den
Wasserdampf an sich reiBen, fur jeden gleich grog ist. Wenn
wenig Kerne vorhanden sind, so werden die Tropfen verhaltnis-
mMig grog, weil ihr Kondensationsbereich groB ist. Kleine
Un-terschiede des letzteren haben dann hierbei wenig EinfluB
auf die TropfengroBe. Bei groBerer Zahl der Kerne werden
erstens die einzelnen Tropfen kleiner werden, zweitens werden
sie nicht mehr alle gleich groB werden, da hier die Verteilung,
die doch nie ganz regelmagig ist, einen bedeutend groberen
relativen EinfluB auf den Kondensationsbereich hat und damit
auch auf den Tropfenradius. Der so entstandene Unterschied
in der TropfengroBe wird seinen EinfluB auf das Beugungs-
bild dahin auBern, dab die von den einzelnen Tropfen er-
zeugten Beugungsbilder nicht mehr genau aufeinander fallen. Die Folge ist, daB das Beugungsbild an Scharfe verliert, seine
Farbenreinheit betrachtlich abnimmt und auaersten Falles
ganzlich verschwindet. Solche Nebel sind inhomogen. Wir
sehen also, daB auf diese Weise zu erklaren ist, weshalb gerade
sehr starke Nebel auch inhomogen sind. g
Eine Zwischenstufe sind die im vorhergehenden als ,,iri-
sierende" bezeichndten Nebel, die in kleinen Gebieten Farbungen
zeigen, aber keine Ringe. Bei ihnen kann man annehmen,
daB in kleinen Gebieten die annahernd gleichen Ursachen eine
annahernd gleiche Verteilung der Kerne, damit homogene Nebel
und wenigstens schwache Beugungsfarben hin und wieder er-
zeugen. Da sich diese gefarbten Wolken infolge yon Luft-
stromungen bewegen, entsteht der Eindruck des Irisierens. Bei einem dichten nicht homogenen Nebel wird durch
erstmalige Expansion , die eventuell wiederholt werden muB,
eine groBe Zahl der Kerne in den sich setzenden Tropfchen
niedergeschlagen. Fur folgende Kondensationen ist eine ge-
ringere, aber immer noch erhebliche Anzahl von Kernen iibrig
geblieben, oder nachtraglich wieder neu gebildet worden
(letzteres mehrfach bei den vorhergehenden Versuchen). Nach
den obigen oberlegungen ist daher bei den folgenden Konden-
sationen Homogenitat des Nebels zu erwarten und in der Tat,
wie beschrieben , gefunden. 1) Vgl. F. Richarz, Zur ErklBrung neutraler Tropfen; Marburger
Siteungsber. 1905. Nr. 3. 1. Homogener und inhomogener Nebel. Zuerst sind dann die Tropfchen, E. Barkow. wegen ihrer immer noch groBen Zahl, relativ klein und geben
daher weite Beugungsringe mit prachtigen Farbungen. Bei
den folgenden Expansionen sind die noch ubrig gebliebenen
Kerne immer weniger zablreich, die an ihnen niedergeschla.genen
Tropfchen wegen des grofieren ihnen zur Verfiigung stehenden
Kondensationsbereiches grober, die Beugungsringe immer kleiner
und weniger intensiv, alles wie beobachtet (vgl. z. B. oben p. 324). 2. Erk1arungsmb;glichkeiten der E n t s t e h u n g des blauen
Nebels. Ein ohne Expansion sichtbarer blauer Nebel entsteht bei
Einwirkung eines Wechselfeldes, von ultraviolettem Licht und
von Ozon bei Gegenwart von Stickstoff oder von nitrosen
Gasen. Er ist nicht identisch mit den Ionentropfchen, die
an und fur sich unsichtbar sind und erst bei 4-8facher
Ubersattigung wachsen, sondern mu6 etwas anderes sein , da
er spontan in gesattigtem, ja in ungesattigtem Wasserdampf
entsteht. DaB indessen die Ionentropfchen bei seiner Ent-
stehung einen Anteil haben, ist denkbar. Die eine Erklarungs-
moglichkeit seiner Entstehung ware die folgende. Wie schon auseinandergesetzt, bildet jedes Ion ein Wasser-
trijpfchen um sich, dessen Radius r = 1 : (3,2. lo7) cm ist. Die
ungleicbnamig geladenen ziehen sich gegenseitig an. I m
einfachsten Fall konnen ein positives und ein negatives Ionen-
tropfchen sich nahern und schlieBlich miteinander verschme1zen.l)
Es werden dann weiterhin aucb Vereinigungen von einer
gro6eren Anzahl zustande kommen. Der mittlere Tropfen-
radius wird einen Wert haben, der einem Gleichgewichts-
zustand entspricht und von der Starke der Ionisation und der
Menge des vorhandenen Wasserdampfes abhangen wird ; denn
je mehr Ionentropfchen vorhanden sind, desto groBere Kom-
plexe von ihnen werden sich dann bilden konnen. Diese so
entstandenen Tropfen sind elektrisch neutral und werden
infolgedessen leichter verdampfen. Wir werden also in dem
sich selbst uberlassenen Dampf eine groBe Anzahl feiner
W assertropfen bekommen, deren meiste kleiner als jene Crenze 3ntstehung von Nebel 6ei Vasserdampf etc. 337 sind , weil sie nach der gegenseitigen Neutralisierung wieder
verdampfen. Sollen dagegen weiterhin diese Tropfen durch
fJbersattigung bei einer Entspannung wachsen, so muB diese
sehr grog sein, wahrscheinlich gro6er als bei Ionentropfchen;
denn ihnen fehlt die elektrische Ladung, die die Dampf-
spannung herabsetzt , und dies wird nicht aufgehoben durch
den grOSeren Tropfenradius. Solche kleine Tropfen, die un-
geladen sind und etwa 8fache Ubersattigung zum Wachstum
erfordern, hat C. T. R. Wilson und Barus tatsachlich ge-
funden. Wenn die erste Ionisation sehr stark ist, so konnte
die mittlere GroBe der urspriinglichen, durch Vereinigung von
Ionentropfchen entstehenden neutralen Tropfen gr&Ber werden
und vielleicht die TropfengrOBe des blauen Nebels erreichen. Diese Tropfen kijnnen aber nicht stabil sein, da sie aus reinem
Wasser bestehen, ungeladen sind, und infolge ihrer groBeren
Dampfspannung rasch wieder verdampfen wiirden. Zur Er-
klarung der Permanenz des blauen Nebels mu6 daher eine
andere Erklarung herangezogen werden. 2. Erk1arungsmb;glichkeiten der E n t s t e h u n g des blauen
Nebels. g
g
g
Bei kraftiger Ionisation werden nicht alle Atomionen Ge-
legenheit haben, sich mit einer Wasserhiille zu umgeben, sondern
sie werden sich, wenn verschiedene Base vorhanden sind,
moglicherweise zu irgendwelchen chemischen Verbindungen
vereinigen. Diese kOnnen sich dann in den Wassertropfchen
&en und so deren Dampfspannung vermindern, so daB sie
weiter wachsen kbnnen, unter giinstigen Bedingungen, bis sie
im Bereich der Sichtbarkeit liegen. Wenn die TropfengroBe
von der GrO6e der Lichtwellen des blauen Lichtes ist, so
werden wir einen blau erscheinenden feinen Nebel sehen, wie
er oben beschrieben ist. Begunstigt wird dieser Vorgang,
wenn sich chemische Verbindungen besonders leicht bilden
kiinnen. Dieser Fall wird z. B. eintreten, wenn Sauerstoff
ionisiert wird. Haben wir freie Sauerstoffionen, so wird ein
Teil von ihuen zur Ozonbildung beitragen, ein anderer kann
chemische Verbindungen eingehen. Da6 Sauerstoff bei der
Bildung des blauen Nebels eine wesentliche Rolle spielt, geht
daraus hervor, daB er z. B. in reinem Wasserstoff nicht ent-
steht, wie C. T. R. Wilson und Vincent gezeigt haben. Auch
in reinem Wasserdampf, in Abwesenheit von anderen Gasen,
tritt der blaue Nebel nicht auf. Ebenso bilden kch bei 22 Annalen der Physik. IT. Folge. 23. 338 E. Barkom. Spitzenentladung in Wasserstoff keine Kerne, die bei geringen
Expansionen wachsen konnen. Dagegen war in allen oben
beschriebenen Beobachtungen des blauen Nebels Luft oder
Sauerstoff vorhanden. Spitzenentladung in Wasserstoff keine Kerne, die bei geringen
Expansionen wachsen konnen. Dagegen war in allen oben
beschriebenen Beobachtungen des blauen Nebels Luft oder
Sauerstoff vorhanden. Die Tropfchen des blauen Nebels werden zu ihrem Ent-
stehen aus den primaren Ionen Zeit gebrauchen. Im Ein-
klang hiermit steht die Erscheinung, die ich bei der Nebel-
bildung durch Wechselfeldwirkung beobachtete (vgl. oben p. 323). Von den elektrischen Eigenschaften der Tropfchen des
blauen Nebels wird sich folgendes vermuten lassen. Ein groBer
Teil wird ungeladen sein, ein anderer positiv oder negativ. Die Zahl der geladenen Tropfen wird nach dem Abstellen
des Ionisators rasch abnehmen. Dies sind alles Erscheinungen,
wie sie Vincent tatsachlich gefunden hat. 1) K. Schsum u. W. Braun, Physik. Zeitschr. 6. p. 73-74.
1905;
Marburger Sitzungsber. Jan. 1905. 3. Versuche und Erorterungen uber die Zusammensetzung
des blauen Nebels. 3. Versuche und Erorterungen uber die Zusammensetzung
des blauen Nebels. Nach obigem mug in den Tropfen des blauen Nebels eine
Substanz gelost sein, die sie stabil macht. Jede Substanz,
die den Dampfdruck genugend herabsetzt, auch ein gelostes
Gas, leistet das Verlangte. In unseren Fallen konnen diese
Stoffe Stickoxyde oder Wasserstoffsuperoxyd sein. C. T. R. Wilson vermutete als geloste Substanz das Wasserstoffsuper-
oxyd. Vincent glaubte diese Auffassung als unhaltbar nach-
gewiesen zu haben. Er leitete namlich die rnit dem blauen
Nebel geschwangerte Luft in einen Nessingkasten, der eine
photographische Platte enthielt. Bei der Entwickelung zeigte
sich keine Spur von Wirkung auf der Platte, wahrend H,Oz
recht krkftig die photographische Platte schwarzt. Ozon war
vorhanden, wie die Blauung von Jodkaliumstiirke bewies. Dieser Versuch ist jedoch nicht einwandfrei; denn wie I(. Schaum
und W. Braun l) nachgewiesen haben, gibt HBOz Solarisation
(Umkehr), kann also scheinbar unwirksam sein ; und die Wir-
kung von 0, auf die photographische Platte hangt von der
Plattensorte ab. 330 Entstehung von Nebel bei Fassedampf etc. Ich versuchte nun, auf eine andere Art zu kontrollieren,
ob H,O,
in den Tropfchen vorhanden sei. Lost man etwas
Titandioxyd in verdunnter Schwefelsaure und setzt eine Spur
von H,O,
zu, so tritt eine kraftige Gelbfarbung der Losung
auf. Diese Reaktion ist auBerordentlich empfindlich.1) Die
Versuchsanordnung war die folgende:
Durch den Tubulus b
(Fig. 2) fuhrte eine Qlasrohre mit einem Wattefilter, durch
das die atmospharieche Luft eintrat. Der Tubulus a war
durch einen Gummistopfen verschlossen; durch seine Bohrung
fiihrte eine Glasrohre sofort in eine Waschflasche, die Titan-
schwefelsaure enthiel t. Durch dieses System wurde mit Hilfe
einer Wasserstrahlluftpumpe die Luft langsam hindurchgesaugt. Die Luftmenge wurde an einer Gasuhr gemessen; Der Luft-
strom wurde durch ein mit Watte dicht gestopftes, enges
Glasrohr hindurchgesaugt und dadurch so verlangsamt , daB
stiindlich nur 10 -
15 Liter Luft hindurchgingen. Durch genugend langes Stehen wurde die Luft staubfrei. Dann wurde sie durch ein Quarzfenster Q, das den Tubulus c
verschloB, mit der Quecksilberbogenlampe bestrahlt , so dalj
sich der spontane blaue Nebel kraftig entwickeln konnte. Dann
wurde bei fortdauernder Strahlung die Luft hindurchgeleitet,
ofter auch die Lampe aus auBeren Griinden fur einige Minuten
ausgeschaltet und wieder eingeschaltet etc. Bei einem Versuch
dauerte die Gesamtstrahlungszeit 305 Min., und die durch
geleitete Luft betrug 134 Liter. 1) F. Richarz u. K. Lonnes, Zeitschr. f. pbysik. Chemie 20.
p. 147. 1896; W. Staedel, Zeitschr. f. angew. Chemie 16. p. 642. 1902.
* 3. Versuche und Erorterungen uber die Zusammensetzung
des blauen Nebels. Um die Reaktion moglichst empfindlich zu machen, fullte
ich ein hohes, schmales Beagensglas mit Titandioxydlosung ;
dann brauchte ich nur wenige ccm Losung, wahrend die
durchgerlte Schicht doch ziemlich hoch war. Das Reagens
zeigte keine bemerkbare Gelbfarbung. Die zur Erklarung des
blauen Nebels in seinem Tropfchen anzunehmende Substanz
scheint also nicht H,O, zu sein. Dagegen halte ich es nach den oben auf p. 332 bis 334
angefiihrten Versuchen und Uberlegungen fur hochstwahr-
scheinlich, daB jene Substanz eine Stickstoff- Sauerstoffver- 22 * 340 E. Barkow. bindung ist. Die Wirksamkeit des 0, tritt immer nur ein,
wenn Anwesenheit von Stickstoff, bez. Stickoxyden nicht vollig
ausgeschlossen ist. Von den nitrosen Gasen zeigte ich oben,
daB sie spontan Kerne bilden, die bei sehr geringer Expansion
starke, dichte Nebelbildung hervorrufen. Von rauchender Sal-
petersaure haben bereits R. v. Helmholtz und F. Richarz
mit dem Dampfstrahl nachgewiesen, da6 sie kraftige Kondcn-
sation hervorruft. Das dampfformige Anhydrid N,O, wird in
der Tat kraftig hygroskopisch sein, wie auch schon R. v. Helm-
holtz fur die minimalen Mengen von SO,H,-Dampfen nach-
wies, die in einem Luftstrom enthalten sind, der iiber Schwefel-
saure gestrichen ist. Alle Beobachtungen wurden sich also
volikommen erklaren lassen, wenn niedere Oxydationsstufen
des Stickstoffs als Grund der schwacheren Kondensation an-
gesehen werden; durch Ozon oder Belichtung erzeugte hohere
und noch starker hygroskopische Oxyde als Ursache der sehr
starken Kondensation. Wie erwahnt, wird dies von Hrn. P r i n g a 1 weiter untersucht. IV. Untersuchung von anderen Diimpfen und Gasen.
1 B
l i
W
t ff 1. Benzol im Wasserstoff. 1. Benzol im Wasserstoff. Bei alien bisherigen Versuchen war Sauerstoff vorhanden. Wie wir gesehen haben, spielt er bei der Nebelbildung eine
gro6e Rolle. Deshalb war es interessant, sauerstofffreie Gase
und Dampfe zu untersuchen. p
Die Versuchs-
anordnung war dieselbe wie friiher. Der Wasserstoff wurde
einer Stahlbombe entnommen; er passierte zuerst eine W oulff-
sche Flasche rnit Kaliumpermanganat, um etwa vorhandene
organische Stoffe zu entfernen, dann eine Waschflasche mit
Kalilauge, um Kohlensaure zu beseitigen, dann zwei Wasch-
flaschen rnit Pyrogallussaure zur Absorption des Sauerstoffs,
und zwecks Trocknung eine Waschflasche mit konzentrierter
Schwefelsaure, ein U-Rohr mit Chlorcalcium und zwei U-Rohren
mit Phosphorpentoxyd. Im NebelgefaB befand sich am Boden
eine Schicht reinen Benzols. Zuerst machte ich eine grij6ere An-
zahl son Entspannungen, um die im Nebelgefa6 noch vorhandene
Luft zu entfernen. Es zeigte sich, dat3 man in Benzoldampfen
Ich nahm Benzoldampfe in Wasserstoff. Entsteiiung von Nebel bei Wasserdampf etc. 341 nur sehr schwer Nebel erhalten kann. Es entstehen nur sehr
wenige groBe Trop fen, die wegen ihrer geringen Anzahl keine
Beugungsfarben zeigen. Erst bei sehr starken Entspannungen
(etwa 40 cm Quecksilbersaule) zeigt sich ein etwas starkerer
Nebel, der kleine Beugungsringe erkennen lii6t. ,
g
g
g
LieB ich das elektrische Wechselfeld wie p. 322 rangere
Zeit, bis 11 Min,, einwirken, so ergab sich keine verstarkte
Nebelbildung. Auch Rontgenstrahlen wirkten nicht auf die
Kondensation ein. Ebensowenig Erfolg hatte eine 15 Min. lange Bestrahlung mit ultraviolettem Licht der Quecksilber-
bogenlampe durch die Quarzlinse Q, Fig. 2 (an Stelle des
Tubulus c). Benzol in Wasserstoff vermag sich also nicht an
Ionen zu kondensieren; denn daB Ionen in Wasserstoff z. B. durch Rontgenstrahlen entstehen, ist ja nachgewiesen. 2. Schwefelkohlenstoff in Wasserstoff. 2. Schwefelkohlenstoff in Wasserstoff. 2. Schwefelkohlenstoff in Wasserstoff. Weitere Versuche ahnlicher Art machte ich mit reinem
Schmefelkohlenstoff in einer Wasserstoffatmosphare. Es zeigte sich, daB die Versuche nur im Dunkeln an-
gestellt werden diirfen, da sich im Lichte zahlreiche Konden-
sationskerne bilden. Das zur Beohachtung der Beugungsringe
dienende, 4,5 m entfernte Gasgluhlicht ist zu schwach, um
erhebliche Mengen von Kernen liefern zu konnen. Bogenlicht
wirkt sehr stark ein. Schon eine Belichtungsdauer von etwa
15 Sek, genugt, um soviel Kerne zu erzeugen, d& sieben Ent-
spannungen von etwa 28 cm Quecksilber notig sind, sie wieder
zu entfernen. Die Kerne ordnen sich so an, daB die grijBten
sich unten im GefaB befinden und die kleineren im oberen
Teile. Bei kleinen Entspannungen bildet sich deshalb auch
nur im untersten Teil des GefaBes eine diinne Nebelschicht
aus, und erst bei starkeren Expansionen wird der ganze Raum
mit Nebel erfullt. Aus diesem Grunde wurden auch die Ent-
spannungen so gro6 gewiihlt. Die wirksamen Strahlen sind
hier die sichtbaren Lichtstrahlen ; denn das Bogenlicht geht
zuerst durch die dicken Kondensorlinsen, durch das Kuhl-
wasser und durch die Wand des NebelgehBes, und hierbei
werden die ultravioletten sowie die ultraroten Strahlen ab-
sorbiert. Entsprechende Versuche mit der Quecksilberbogen-
lampe zeigten dasselbe, denn eine zwischen das Quarzfenster 34 2 E. Barkow. des NebelgefaBes und die Lichtquelle geschaltete Glasscheibe,
die die fur Luft und Wasserdampf wirksamen Strahlen ab-
schneidet, schwacht hier die Wirkung nur sehr wenig. g
g
Die von Schauml) gefundene Wirkung des Schwefel. kohlenstoffs auf den Dampfstrahl ist nach diesen Versuchen
wahrscheinlich nicht bloB eine rein thermische, sondern auch
wesentlich beeinflufit durch die Zersetzung des Schwefelkohlen-
stoffs im Licht. Rontgenstrahlen wirken hier kraftig auf die Nebelbildung
ein. Nach einer 10 Min. langen Strahlung gehorten zehn Ent-
spannungen von ca. 28 cm dazu, samtliche Kerne wieder zu
beseitigen. Auffallend ist, daB das elektrische Wechselfeld hier keine
Wirkung ausubt, selbst nnch 10 Min. langer Dauer. Be-
merkenswert ist, daB zwischen den mit dem Induktorium ver-
bundenen Platten und den nachstgelegenen Kugelkalotten des
NebelgefaBes keine Entladungen ubergehen, wie es bei den
fruheren Versuchen mit Luft und Wasser der Fall war (p. 322). Wahrscheinlich wurde bei jenen die dunne Wasserschicht, die
das NebelgefaB innen bedeckt, influenziert, und deswegen ent-
stehen zwischen Platten und Nebelgef'aB so starke Spannungen,
dafi dadurch die Lichtbiischel hervorgerufen werden. Bei
Schwefelkohlenstoff bleibt diese Erscheinung aus , weil er ein
Nichtleiter ist. Auch hier halten sich die Kerne sehr lange, wie in den
friiher beschriebenen Versuchen. 1) K. Schaum, Marb. Sitzungsber. 1902. Nr. 8. 1) K. Przibram, Sitzungsber. der k. Akad. der Wissensch. zu Wien,
Math-Nat. Klssse. p. 33-38.
1906. Kiirzlich ist eine neuere Arbeit von
ihm unter demselben Titel erschienen in der Adolf Lieben-Festschrift,
2) J. Campanile u. di Ciommo, Physik. Zeitschr. 4. p. 648 bis
3) R.v.Helmholtzu.F.Richarz, Wied. Ann. 40.p. 161-203.1890
pi 170-175. 1906.
651. 1903. 1) F.Richarz, Ann. d. Phya. 19. p. 639-642. 1906.
(Eingegangen 17. April 1907.) 2. Schwefelkohlenstoff in Wasserstoff. Das Nebelgefaf3 wurde 5 Min. lang belichtet und dann im Dunkeln sich selbst uberlassen. Nach 15 Stunden waren fiinf Entspannungen nMig, urn die
noch vorhandenen Kerne zu beseitigen. Der im Schwefelkohlenstoff entstehende Nebel zeigt lange
nicht die prachtigen Farben des Wassernebels. Die Beugungs-
ringe haben nie die GrbSe, daB sie die gefarbten zentralen
Felder zeigen (p. 321, Anmerk.). Der bei der ersten Entspannung auftretende Nebel ist
auch hier nicht immer vollkommen homogen. Die Tropfchen
fallen auch bedeutend schneller, als die gleich groBen Wasser- Bntstehung von Nebel bei Wasserdampf etc. 343 tropfchen in Luft, da die innere Reibung in Wasseratoff ge-
ringer ist. g
Die Frage, ob auch hier die Ionen die Kernbildner sind,
lBBt sich zurzeit noch nicht mit voller Sicherheit beantworten. Wahrscheinlich ist es ja, aber dann kann man schwer erklaren,
warum das elektrische Wechselfeld ganz unwirksam ist. Hier-
uber kiinnen erst Leitfiihigkeitsuntersuchungen und etwaige
Analogien bei anderen Dampfen von ahnlicher Zusammen-
setzung A.ufschluB geben. Wahrend der Drucklegung meiner Dissertation erhielt ich
Kenntnis von einer soeben erschienenen Arbeit K. Przibrams.')
E r benutzt dieselbe Versuchsanordnung wie C. T. R. Wilson
und auch Expansionsraume derselben GroBe wie dieser ; sie
sind also auch zu klein, um alle bei der Nebelbildung auf-
tretenden Erscheinungen beobachten zu konnen (vgl. p. 321). Es beruht ferner anf Irrtum, wenn er angibt, zuerst .J. Cam-
panile und di Ciommo2) hatten im Jahre 1903 andere
Dampfe a19 Wasserdampf in bezug auf Kondensationsauslosung
untersucht. Vielmehr haben schon R. v. Helmholtz und
F. Richar z 3, Dampfstrahlen von Alkohol, Ameisensaure, Essig-
saure und Anilindampfen untersucht. Przibrams geringe
ExpansionsgroBe fur Schwefelkohlenstoff ruhrt hochstwahr-
scheinlich davon her, daB das NebelgefaB nicht vor Licht ge-
schutzt war. Sehr wichtig ist aber das Hauptresultat der Arbeit von
Przibram, daB die Dampfe der Alkohole sich leichter auf die
positiven Ionen niederschlagen als auf die negativen, anders
als bei Wasserdampf. p
Von besonderem Interesse ist auch Przibrams Berech-
nung der Ubersattigung fur Ionenbondensation und fur kern- E. Barkow. 3ntstehung von Nebel 6ei Wasserdampf etc. 344 lose Nebelgrenze bei verschiedenen Dampfen. Die Werte finden
sich aus den Formeln von F. Richarz fur k eines Gasge-
misches l), und ergeben far verschiedene Dampfe bemerkens-
werterweise sehr verschiedene Werte. Auch Przibram kommt zu dem SchluE, daE man bei
der Erklarung der Kondensation nicht ausschliefilich die physi-
kalischen Eigenschaften der Gasionen zu beriicksichtigen habe,
wie es auch aus meinen Resultaten zu schlieEen ist. Marburg i. 2. Schwefelkohlenstoff in Wasserstoff. H., Physikalisches Institut, Marz 1906.
|
https://openalex.org/W4253740512
|
https://www.uco.es/ucopress/az/index.php/az/article/download/2798/1696
|
Portuguese
| null |
Produção de pacu em tanques-rede no reservatório de itaipu, brasil: retorno econômico
|
Archivos de zootecnia
| 2,011
|
cc-by-sa
| 5,405
|
RESUMO O objetivo deste trabalho foi avaliar a
rentabilidade econômica de um projeto aquícola
em tanques-rede para o cultivo de pacu (Piaractus
mesopotâmicus) com peso médio final de 1 kg,
propiciando assim informações para a tomada de
decisão de investidores, produtores, técnicos,
órgãos de fomento e demais instituições afetas à
área. Para avaliação dos cenários de rentabilidade
econômica utilizou-se o método determinístico
com auxílio de uma planilha eletrônica. Verificou-
se que os índices de retorno econômico na
produção de pacu em 160 tanques-rede no
reservatório de Itaipu apresentaram os piores
resultados no cenário I com índice de lucratividade
de -17,22%, devido à conversão alimentar de 3,58
e o preço da ração de R$ 0,75 kg. A análise do
retorno sobre o investimento (ROI) apresentou os
índices de 1,30; 1,36; 1,47 e 1,54 R$/kg entre os
cenários A a D, respectivamente. O ponto de
equilíbrio financeiro médio entre as projeções foi
de R$ 226 921,1 e a média do volume mínimo a ser
produzido foi de 55286,7 kg de pacu, baseando-
se no ponto de equilíbrio físico do projeto com 160
tanques-rede. Para o melhor retorno econômico
deve-se produzir no mínimo 40 946,4 kg de pacu em 160 tanques-rede com uma conversão alimen-
tar de 1:2,49 e preço de venda do peixe a 4,00
atingindo 31,78% de lucratividade sobre a receita
bruta com um ROI de 1,5393 R$ por quilo produzido. PALAVRAS CHAVE ADICIONAIS Agronegócio. Análise de rentabilidade. Aquicul-
tura. Espécies nativas. Agro business. Profitability analysis. Aquiculture. Native species. Agro business. Profitability analysis. Aquiculture. Native species. PRODUÇÃO DE PACU EM TANQUES-REDE NO RESERVATÓRIO
DE ITAIPU, BRASIL: RETORNO ECONÔMICO
ECONOMIC RETURN FROM CAGE PRODUCTION OF
PACU IN ITAIPU RESERVOIR, BRAZIL Silva, J.R.1, Rabenschlag, D.R.2, Feiden, A.3A, Boscolo, W.R.3B, Signor, A.A.4 e Bueno, G.W.5 1UNIOESTE-Universidade Estadual do Oeste do Paraná. Centro de Ciências Sociais Aplicadas. Toledo, PR. Brasil. josemar@unioeste.br 2UFSM-Universidade Federal de Santa Maria. Centro de Tecnologia/CT. Santa Maria, RS. Brasil. denis@ct.ufsm.br 3UNIOESTE-Universidade Estadual do Oeste do Paraná. Programa de Pós-Graduação em Recursos
Pesqueiros e Engenharia de Pesca. Toledo, PR. Brasil. Aaldifeiden@gmail.com; Bwilsonboscolo@hotmail.com
4IFPR-Instituto Federal do Paraná. Setor Aquicultura. Foz do Iguacu-Paraná. Brasil. arcangelo.signor@
ifpr edu br 3UNIOESTE-Universidade Estadual do Oeste do Paraná. Programa de Pós-Graduação em Recursos
Pesqueiros e Engenharia de Pesca. Toledo, PR. Brasil. Aaldifeiden@gmail.com; Bwilsonboscolo@hotmail.com
4IFPR-Instituto Federal do Paraná. Setor Aquicultura. Foz do Iguacu-Paraná. Brasil. arcangelo.signor@
ifpr.edu.br 4IFPR-Instituto Federal do Paraná. Setor Aquicultura. Foz do Iguacu-Paraná. Brasil. arcangelo.signor@
ifpr.edu.br 5MPA-Ministério da Pesca e Aquicultura. Brasil. guilherme.bueno@mpa.gov.br Recibido: 27-4-11. Aceptado: 7-11-11. SUMMARY The aim of this study was to evaluate the
economic profitability a project of aquiculture in
cage for the cultivation of pacu (Piaractus
mesopotamicus) with an initial medium weight of
1 kg. The study was done to provide information
for decision making investors, farmers, techni-
cians, foment organs and other institutions related
to the area. To evaluate the economic profitability
scenery, a deterministic method, with the help of
an electronic spread sheet, was used. It could be
verified that the economic outcome indexes in
pacu breeding in 160 cages in the Itaipu reservoir,
presented the worst results in scenery I with a
profitability index of -17.22%, due to diet conversion
of 3.58 and the ration price (R$ 0.75 kg). The
analysis of return on investment (ROI) presented
indexes of 1.30, 1.36, 1.47 and 1.54 R$/kg between
A and D sceneries, respectively. The medium Arch. Zootec. 61 (234): 245-254. 2012. Arch. Zootec. 61 (234): 245-254. 2012. Recibido: 27-4-11. Aceptado: 7-11-11. SILVA, RABENSCHLAG, FEIDEN, BOSCOLO, SIGNOR E BUENO água doce potenciais de cultivo (Godinho,
2007) e detém cerca de 5,5 milhões de hectares
em represas ou reservatórios naturais e
artificiais que podem ser utilizados para a
prática da aquicultura. Dentre estes
reservatórios, na bacia hidrográfica do
Paraná III, fronteira entre o Paraguai e Brasil
tem-se o reservatório de Itaipu que repre-
senta uma área inundada de 1350 km². Na
sua margem brasileira possui cerca de 20
braços, dentre estes, o São Francisco Falso,
São Francisco Verdadeiro e Ocoí que foram
destinados ao cultivo de pacu (Piaractus
mesopotâmicus) em tanques-rede. Nestas
áreas aquícolas estimam-se uma produção
de 4666 toneladas por ano e a possibilidade
de gerar 13 211 empregos diretos e 52 845
empregos indiretos (MPA, 2010). financial equilibrium point between the projections
was of R$ 226921.12 and the average of minimum
value to be produced was of 55 286.7 kg pacu,
based on the physical equilibrium point of the
project with one hundred and sixty net ponds. For
a better economic return it is necessary to breed
at least 40 946.4 kg pacu in one hundred and sixty
net ponds with a diet conversion of 1:2.49 and the
fish selling price of 4.00, reaching a profitability of
31.78% on the gross revenue with a ROI of 1.54
R$/kg produced. Archivos de zootecnia vol. 61, núm. 234, p. 246. INTRODUÇÃO O desenvolvimento da atividade pes-
queira, para fazer frente à constante elevação
da demanda mundial por alimentos,
engendrou uma série de políticas que
incentivaram a ampliação da produção
pesqueira mundial. Estas políticas de incen-
tivo geraram modelos de desenvolvimento
da atividade da pesca extrativa, que não
pesaram, tampouco se preocuparam com os
estoques pesqueiros, levando estes para
próximo do limite ou capacidade máxima de
exploração sustentável (Ostrensky et al.,
2008). p
g
(
)
O pacu é uma espécie nativa da Bacia da
Prata, apresenta maior distribuição nas
planícies alagadas da região Centro-Oeste,
no Pantanal do Mato Gtosso (Petrere, 1989),
destaca-se entre as espécies nativas
apresentando desejáveis características
zootécnicas para o cultivo como: hábito
alimentar onívoro, carne de excelente
qualidade e boa aceitação pelos consumi-
dores (Jomori et al., 2003). Rendimento do
processamento de 46,73% de filé sem pele;
16,57% de cabeça e 88,98% de rendimento
de carcaça segundo Faria et al. (2003),
destacando-se como uma interessante al-
ternativa para cultivo. Além de demonstrar
uma fácil adaptação à alimentação artificial,
elevada rusticidade e fecundidade (Castag-
nolli e Zuim, 1985). Desta forma, a principal alternativa para
suprir o déficit dos estoques pesqueiros
está no cultivo de organismos aquáticos,
isto pode ser observado com o aumento da
produção mundial que obteve um cresci-
mento de 8,4% em nível mundial (FAO, 2010),
destacando-se a piscicultura continental,
em diferentes modelos de sistemas produ-
tivos. Segundo o relatório Blue Frontiers:
Managing the environmental costs of
Aquaculture, a aquicultura é a melhor forma
de se produzir proteína animal, do ponto de
vista ambiental e econômico. Corroborando
com esta afirmativa, a Organização das
Nações Unidas para Agricultura e Alimen-
tação considera que as projeções para 2030
no consumo mundial de pescado será em
torno de 100 milhões de toneladas por ano. Portanto, os novos modelos de produ-
ção e o crescimento do consumo de peixes
promovem o avanço da aquicultura, que por
sua vez, demandam informações para toma-
da de decisão sobre a viabilidade econômica
e a rentabilidade dos empreendimentos
aquícolas, principalmente no que se refere
às espécies de peixes nativos. Neste contexto, o Brasil poderia atender
esta demanda e se tornar um dos maiores
produtores de peixe do mundo, pois o país
possui cerca de 40 espécies de peixes de Estudos desta natureza vêm sendo
desenvolvidos em cultivos de tilápia
(Oreochromis niloticus) em reservatórios
por Conte (2002); Medeiros (2002); Ono e Archivos de zootecnia vol. RETORNO ECONÔMICO DA PRODUÇÃO DE PACU EM TANQUES-REDE Kubitza (2003); Marengoni e Bueno (2007);
Kubitza e Campos (2005); Scorvo-Filho et
al. (2006); Furlaneto et al. (2006) e Ayroza
(2009). Porém, há uma grande lacuna de
informações de projetos de rentabilidade e
viabilidade econômica para espécies nati-
vas que também apresentam potencial
produtivo, como o pacu (Piaractus meso-
potâmicus), jundiá (Rhamdia sp.), surubim
(Pseudoplatystoma sp.), entre outras não
exóticas. zado para os cálculos o custo intermediário
de 0,7070 R$/kg, pois não houve diferenças
significativas no desempenho dos peixes
alimentados com as diferentes rações. Os índices zootécnicos foram obtidos
de um experimento conduzido em delinea-
mento inteiramente casualizado com seis
tratamentos e três repetições totalizando 18
tanques-rede de 5 m3 com densidade de 44
pacus/m3 com peso médio inicial de
293,38±5,67 g. Em seguida, os cálculos de
produção foram estipulados para um
empreendimento aquícola com 160 tanques-
rede (1,75 x 1,75 x 1,75 m), conforme
estabelecido pelo plano diretor da Itaipu
Binacional (2007) para produção em uma
área aquícola de 6400 m2 no reservatório de
forma a atender a capacidade de suporte do
local estimada pelo modelo Dillon e Rigler
(1974). Entretanto, este trabalho teve como
objetivo avaliar os cenários de retorno
econômico durante a produção de pacu (P. mesopotâmicus) em 160 tanques-rede ins-
talados em uma área aquícola no reservatório
de Itaipu. INTRODUÇÃO 61, núm. 234, p. 246. Archivos de zootecnia vol. 61, núm. 234, p. 247. 11,0 US$= 1,80 R$ SILVA, RABENSCHLAG, FEIDEN, BOSCOLO, SIGNOR E BUENO custos variáveis (CV); margem bruta do
custo total (MBCT) : definida pela relação
entre o lucro e custo total de produção;
margem bruta do custo operacional total
(MBCOT) : relação entre o lucro operacio-
nal (ou receita operacional) e custo opera-
cional total de produção; custo médio (CMe):
expresso pela relação entre os custos total
de produção, dividido pela quantidade
produzida; indice de lucratividade (IL):
relação entre o lucro (RB - CT) e demonstra
o percentual da receita obtida com a venda
da produção; retorno sobre o investimento
operacional (RIO): relação entre o lucro
operacional pelo investimento e demonstra
o percentual da receita obtida com a venda
da produção e o ponto de nivelamento ou
ponto de equilíbrio físico e financeiro: são
indicadores importantes para avaliação
econômica da atividade produtiva. Propi-
ciam a determinação da quantidade mínima
(Qmín) ou do preço mínimo (Pmín) para
cobertura dos custos. desenvolvimento da produção, assim como
equipamentos para manejo, biometrias e
acompanhamento da qualidade da água
foram definidos através de pesquisa de cam-
po e consulta as empresas especializadas
do ramo. A depreciação foi calculada pelo méto-
do linear, tomando-se o valor do equipamen-
to diminuído do valor residual e dividido
pela vida útil. Os gastos com a adequação das normas
e legislação pertinente ao licenciamento,
instalação e operação da atividade foram
levantadas de acordo com Filipetto (2004) e
pesquisas junto ao Instituto Ambiental do
Paraná (IAP, 2010) seguindo a Portaria IAP
nº 112, de 13 de julho de 2005. j
De posse destes dados, fez-se a
simulação dos resultados para diversos
cenários de rentabilidade econômica,
através de modelos determinísticos em
planilhas do Excel®. Assim, foram estabe-
lecidos os seguintes cenários possíveis de
acordo com a diferenciação nos indicadores/
parâmetros de conversão alimentar (CA),
custo de ração (R$/kg), preço de venda do
peixe (R$) e taxa de sobrevivência (%). Neste
contexto, definiu-se como A e B os cenários
mais pessimistas, C e D os cenários mais
otimistas e, por final, os pessimistas E a J. Esta relação pode ser verificada na tabela II. MATERIAL E MÉTODOS O trabalho foi realizado durante seis
meses em um cultivo experimental de pacu
(P. mesopotâmicus) em uma área de pes-
quisa com tanques-rede, instalada na
zona de transição do reservatório da Usina
Hidrelétrica Itaipu Binacional, no municí-
pio de Santa Helena região Oeste do Es-
tado do Paraná, Brasil, com as coordenadas
geográficas W54º21'196", S24º51'105",
W54º21'078", S24º51'192", e W54º21'224",
S24º51'143", pertencente à bacia hidrográfica
do Paraná III. O mercado consumidor de destino para
a produção foi estabelecido inicialmente
como o mercado de varejo, ou seja, feiras
livres, peixarias, restaurantes, supermerca-
dos e pesque-pague. Neste contexto,
considerou-se o preço pago por kg de pacu
vivo proposto por Richter (2004) do De-
partamento de Economia Rural da Secre-
taria de Estado Agricultura e Abastecimento
(SEAB-DERAL-DEB) no Paraná, o qual con-
sidera uma projeção na produção por
município do estado e preço médio estadual
de comercialização1, sendo de R$ 2,99/kg em
2004, de R$ 3,37/kg em 2005, de R$ 3,85/kg
em 2006 e de R$ 4,00/kg em 2007. Para a avaliação dos índices zootécnicos
do pacu, seguiu-se os dados avaliados por
Signor et al. (2010) que acompanhou os
parâmetros zootécnicos do cultivo. A partir
das características técnicas do projeto foram
levantadas informações relevantes, as quais
fundamentaram o estudo e as projeções da
produção e estimativas de custos e receitas. De acordo com as condições locais deste
estudo, não foi contabilizado a construção
de um galpão para armazenamento dos
equipamentos e acondicionamento da ração,
por já exisitir no local do experimento, não
obstante, em outros empreendimento deste
formato que não possua local para armaze-
namento e acondicionamento dos insumos,
este item deverá ser considerado nos cálcu-
los do investimento inicial. Para a realização do experimento de
avaliação de desempenho dos peixes foram
elaboradas seis rações experimentais
extrusadas com níveis de 25%, 30% e 35% de
proteína e dois níveis de energia 3000 e 3350
kcal/kg de energia digestível, as quais foram
cotadas a um custo mínimo de 0,6637 R$/kg,
custo máximo de 0,7502 R$/kg, sendo utili- Os equipamentos necessários para 11,0 US$= 1,80 R$ Archivos de zootecnia vol. 61, núm. 234, p. 247. Archivos de zootecnia vol. 61, núm. 234, p. 247. SILVA, RABENSCHLAG, FEIDEN, BOSCOLO, SIGNOR E BUENO Na tabela I, verificam-se as projeções
dos custos operacionais e o investimento
inicial para o projeto de criação de pacu em
160 tanques-rede, onde os itens para
investimento inicial foram pesquisados jun-
to às empresas da região oeste do Paraná a
fim de efetuar a implantação do projeto em
escala econômica para 160 tanques-rede em
um ciclo de produção de seis meses para
obtenção de um peixe com peso médio de 1 kg. p
Os dados do custo de produção, receitas
e montagem do fluxo de caixa foram baseados
na estrutura de Custo Operacional (COP)
proposta por Matsunaga et al. (1976), orien-
tando assim a aferição de dados primários e,
por conseguinte no estabelecimento dos
custos de produção. Contudo, foram
propostos os seguintes indicadores de
rentabilidade do empreendimento: receita
bruta (RB): receita obtida pela venda da
produção; lucro bruto (L): lucro obtido pela
diferença entre a RB e o custo total de
produção (CT); lucro operacional (LO) ou
receita líquida (RL): lucro obtido pela
diferença entre receita bruta (RB) e o custo
operacional total (COT); margem de
contribuição (MC): diferença entre a RB e os Archivos de zootecnia vol. 61, núm. 234, p. 248. RESULTADOS E DISCUSSÃO Através da simulação do retorno
econômico e rentabilidade do cultivo de
pacu com densidade de 44 peixes/m3 em 160
tanques-rede durante 6 meses, obteve-se a
produtividade de 75,03 kg/m3/ano e no
melhor cenário econômico (D) uma lucrati-
vidade sobre receita bruta de 31,78%, com
índice de lucratividade de 1,31 R$/kg (tabela
II). Demonstrando a potencialidade desta
espécie para produção em sistema intensi-
vo. Também avaliando o retorno econômico
em tanques-rede de 6 m3 no rio Parana- Archivos de zootecnia vol. 61, núm. 234, p. 248. SILVA, RABENSCHLAG, FEIDEN, BOSCOLO, SIGNOR E BUENO tividade dentro do empreendimento aquícola
(Kubitza e Ono, 2004). Na maioria dos casos,
o produtor adere aos modelos pré-estabe-
lecidos pelas empresas de rações e às
orientações dos técnicos do ramo, os quais
geralmente não fazem o planejamento
zootécnico específico para cada realidade,
ocasionando a menor rentabilidade do sis-
tema de cultivo, desperdício de estrutura e
maior poluição do ambiente. As projeções
de cenários mais pessimistas estão repre-
sentadas nos itens de E a J na tabela II,
diminuindo assim as possibilidades deste
produtor de tornar-se competitivo dentro
dos novos moldes da aqüicultura mundial
(Eler e Millani, 2007). Furuya (2007) e
Boscolo et al. (2008) vêm demonstrando
que algumas estratégias no manejo nu-
tricional dos peixes como a formulação de
rações através do conceito de proteína ideal,
da consideração do valor nutritivo dos ali-
mentos, determinação da exigências para
cada espécie, redução da proteína da dieta,
melhora do processamento, uso de pró-
nutrientes e fornecimento mínimo de fósfo-
ro disponível trazem maior eficiência dos
peixes. Relacionando-se a conversão ali-
mentar (CA), ganho de peso diário, decli-
nando o tempo de cultivo que irá perfazer
maior retorno econômico sobre o tempo do
investimento, além de menor emissão de
efluentes. panema, estado de São Paulo, Furlaneto et
al. (2006) produzindo tilápia do Nilo
(Oreochromis niloticus) encontraram uma
rentabilidade mínima de 173 t/ciclo/ha,
atingindo 10% de lucratividade em 6 meses. Valores acima dos obtidos pelos sistemas
tradicionais de cultivo de peixes tropicais
em viveiros escavados, que quando bem
manejados, representam rentabilidade de
12,7% em 12 meses (Crivelenti et al., 2006). Tal fato pode ser explicado devido as maiores
densidades empregadas e o melhor apro-
veitamento das dietas pelos peixes confina-
dos, as condições ambientais do local e a
biomassa produtiva utilizada por ciclo de
produção que tornam-se delimitadores do
tempo de retorno do investimento e da
capacidade de suporte do ambiente, fatores
que correlacionam-se a sustentabilidade do
sistema produtivo (Kubitza, 1999). Estes
fatores podem ser observados na tabela II,
quando simulado os cenários mais pessi-
mistas (G, H, I e J) onde o sistema produtivo
obteve um déficit na sua lucratividade e no
retorno sobre o investimento (ROI). RETORNO ECONÔMICO DA PRODUÇÃO DE PACU EM TANQUES-REDE A
hi
d
t
i
l 61
ú
234
249
Tabela I. Levantamento das variáveis financeiras para investimento inicial para produção do pacu (Piaractus mesopotamicus)
em 160 tanques-rede no reservatório de Itaipu. (Financial variables prospection for the initial investment in pacu production (Piaractus
mesopotamicus) in 160 cages located the Itaipu reservoir). Custo unitário
Custocapital fixo
Custo relativo
Vida útil
Depreciação
Total R$
Investimento inicial
R$
R$ (a)
(%) (a.1)
(anos)
R$ (b)
(a)+(b)
Construçao civil
Galpão e área administrativa (existente)
-
-
-
20
-
-
Equipamentos
160 Tanques-rede de engorda
674,94
107 990,40
76,31
5,00
21 598,08
129 588,48
1 Balança
150,00
150,00
0,11
10
15,00
165,00
1 Oxímetro
1900,00
1900,00
1,34
5
380,00
2280,00
2 Puçás para peixes alevinos e juvenis
30,00
60,00
0,04
2
30,00
90,00
2 Puçás para peixes adultos
45,00
90,00
0,06
2
45,00
135,00
4 Baldes (alumínio 45 L)
70,00
280,00
0,20
2
140,00
420,00
4 Coletes salva-vidas
70,00
280,00
0,20
2
140,00
420,00
1 Kit para análise da água
1100,00
1100,00
0,78
1
1100,00
2200,00
900m Corda em polietileno para amarração
3,80
3420,00
2,42
5
684,00
4104,00
36 Poita de fundeio em concreto (300 a 600 kg)
80,00
2880,00
2,04
5
576,00
3456,00
32 Bóias de fundeio para poita (vol. 100 L)
36,00
1152,00
0,81
5
230,40
1382,40
4 Bóias de sinalização e demarcação
37,00
148,00
0,10
5
29,60
177,60
Embarcaçoes
1 Barco de alumínio de 5 x 1,3 m
3650,00
3650,00
2,58
10
365,00
4015,00
1 Motor para barco com 15 HP e 2T
5400,00
5400,00
3,82
10
540,00
5940,00
1 Balça para despesca
4500,00
4500,00
3,18
10
450,00
4950,00
Serviços de documentação
1 Projeto técnico (5% de 1+2+3+4)
6650,02
6650,02
4,70
1 Licença prévia IAP
129,95
129,95
0,09
2
1 Licença de instalação IAP
847,274
847,27
0,60
2
1 Licença de operação IAP
582,176
582,18
0,41
5
1 Imprensa/Diário Oficial
50
50,00
0,04
5
1 Diário Oficial 2ª fase
200
200,00
0,14
5
1 Publicação em jornal de circulação
50
50,00
0,04
5
Total do capital fixo e da depreciação (R$)
141 509,82
100,00
26 323,08
159 323,48 Archivos de zootecnia vol. 61, núm. 234, p. 249. Archivos de zootecnia vol. 61, núm. 234, p. 250. SILVA, RABENSCHLAG, FEIDEN, BOSCOLO, SIGNOR E BUENO Esta pro-
jeção dos índices econômicos também foram
relatados por Marengoni e Bueno (2007)
avaliando a rentabilidade da tilápia do Nilo
em tanques de 4 m3 durante 135 dias sob as
densidades 250, 300, 350 e 400 peixes/m3
obtiveram o melhor lucro de 41,50 R$/m3 para
a densidade intermediária de 300 peixes/m3. De acordo com Signor et al. (2010), a CA
do pacu criado em tanques-rede apresen-
taram valores médios de 3,34; 3,51 e 3,34. Carneiro et al. (1992) encontrou conversão
alimentar de 1,54 a 2,01 para diferentes faixas
de peso de pacu, valores próximos ao de
Scorvo Filho et al. (1998) que obtiveram em
média uma CA de 1,80 para tilápia. De acordo com Ostrensky et al. (2008) a
rentabilidade e o retorno econômico na
produção de peixes diferenciam-se de
acordo com o sistema de cultivo o qual o
empreendedor irá utilizar, podendo este ser
intensivo, extensivo ou semi-intensivo. Um dos principais fatores limitantes no
cultivo de peixes está relacionado ao custo
com a ração e segundo Andrade et al. (2005),
a ração corresponde a 52,19% do custo de
produção. Para Crivelenti et al. (2006), este
item contribui com 41,07% no custo total. Entretanto, a utilização de novas estratégias
de manejo alimentar no cultivo de peixes
contribuirá com a queda nos custos com a
alimentação, ocasionando a melhor lucra- Na tabela II, verifica-se que nos cenários
A e B com a CA de 2,92 e custo da ração de
R$ 0,70, os valores de COE, COT e ponto de
equilíbrio financeiro foram semelhantes. O
ponto nivelador do retorno econômico en-
tre estes dois cenários (A e B) foi o preço de
venda do peixe, que ao aumentar em 0,15
centavos proporcionou um acréscimo de 9 RETORNO ECONÔMICO DA PRODUÇÃO DE PACU EM TANQUES-REDE Archivos de zootecnia vol 61 núm 234 p 25
Tabela II. Análise de sensibilidade através da simulação das variáveis de produção e economia durante o cultivo de pacu em 160
tanques-rede em seis meses (um ciclo) com peso final de 1 kg cultivado no reservatório de Itaipu. (Sensitivity analysis through the
simulation of production and economic variables during the growth phase of pacu cultivated in 160 cages located in the Itaipu reservoir during
six months (one cycle) with final weight of 1 kg). SILVA, RABENSCHLAG, FEIDEN, BOSCOLO, SIGNOR E BUENO Fonte: Dados de pesquisa
Cenário A
Cenário B
Cenário C
Cenário D
Cenário E
Cenário F
Cenário G
Cenário H
Cenário I
Cenário J
1
Ciclo de Produção/Ano
1,00
1,00
1,00
1,00
1,00
1,00
1,00
1,00
1,00
1,00
2
Conversão Alimentar (kg ração / kg de
peixe)
2,92
2,92
2,49
2,49
3,58
3,58
2,92
2,92
3,58
3,58
3
Custo da ração (R$/kg)
0,7070
0,7070
0,6637
0,6637
0,7502
0,7502
0,7502
0,7502
0,7502
0,7502
4
Nº de Tanques-rede
160,00
160,00
160,00
160,00
160,00
160,00
160,00
160,00
160,00
160,00
5
Preço de Venda (R$)
3,85
4,00
3,85
4,00
3,85
4,00
3,85
4,00
3,85
4,00
6
Volume útil Tanque-rede (m³)
5,00
5,00
5,00
5,00
5,00
5,00
5,00
5,00
5,00
5,00
7
Produtividade (kg/m³/ano)
8 / (4 x 6)
75,03
75,03
75,03
75,03
75,03
75,03
75,04
75,04
75,04
75,04
8
Produção kg/ciclo
60.023,76
60.023,76
60.023,76
60.023,76
60.023,76
60.023,76
60.032,00
60.032,00
60.032,00
60.032,00
9
Taxa de Sobrevivência (%)
97,95
97,95
97,95
97,95
97,95
97,95
80,00
80,00
80,00
80,00
10
Receita Bruta (R$)
1 x 5 x 8
231.091,48
240.095,04
231.091,48
240.095,04
231.091,48
240.095,04
231.123,20
240.128,00
231.123,20
240.128,00
11
COE (R$/Ciclo)
162.699,08
162.699,08
137.462,32
137.462,32
200.770,88
200.770,88
207.451,62
207.451,62
244.606,33
244.606,33
12
COT (R$/Ciclo)
189.022,16
189.022,16
163.785,40
163.785,40
227.093,96
227.093,96
233.774,70
233.774,70
270.929,41
270.929,41
13
COE Médio (R$/Ciclo)
11 / 8
2,71
2,71
2,29
2,29
3,34
3,34
3,46
3,46
4,07
4,07
14
COT Médio (R$/Ciclo)
12 / 8
3,15
3,15
2,73
2,73
3,78
3,78
3,89
3,89
4,51
4,51
15
Margem de Contribuição (R$)
68.392,39
77.395,96
93.629,15
102.632,72
30.320,60
39.324,16
23.671,58
32.676,38
-13.483,13
-4.478,33
16
Ponto de Equilíbrio Físico (kg)
12 / 5
49.096,67
47.255,54
42.541,66
40.946,35
58.985,44
56.773,49
60.720,70
58.443,68
70.371,27
67.732,35
17
Ponto de Equilíbrio Financeiro (R$)
5 x 16
189.022,16
189.022,16
163.785,40
163.785,40
227.093,96
227.093,96
233.774,70
233.774,70
270.929,41
270.929,41
18
Lucro Bruto (R$
10 - 11
68.392,39
77.395,96
93.629,15
102.632,72
30.320,60
39.324,16
15.273,50
24.278,30
-21.881,21
-12.876,41
19
Depreciação (R$)
26.323,08
26.323,08
26.323,08
26.323,08
26.323,08
26.323,08
26.323,08
26.323,08
26.323,08
26.323,08
20
Lucro Operacional (R$)
18 - 20
42.069,31
51.072,88
67.306,07
76.309,64
3.997,52
13.001,08
-2.651,50
6.353,30
-39.806,21
-30.801,41
21
Lucratividade sobre Rec. digestibilidade aparente da energia e nutrientes
da farinha de resíduos da indústria de filetagem
de tilápias, para a tilápia do Nilo (Oreochromis
niloticus). Cienc Rural, 38: 2579-2586. CONCLUSÃO O retorno econômico da produção de
pacu em tanques-rede no reservatório de
Itaipu demonstrou que os empreendedores
que pretendem investir no agronegócio pis-
cícola devem produzir no mínimo 40 946,35
kg de pacu em 160 tanques-rede com uma
conversão alimentar de 1:2,49 e preço de
venda do peixe a 4,00 R$/kg para atingir
31,78% de lucratividade sobre a receita bru-
ta obtendo retorno econômico de 1,5393 R$/
kg sobre o investimento. Ressaltando ainda
que existe a possibilidade de um retorno de
18% a 31% sobre a receita bruta. g
Diante das projeções realizadas, os
cenários E, F, G, H, I e J são os mais
pessimistas dentro da análise determinística. Estipulou-se valores menos eficientes para
conversão alimentar, custo com ração e taxa
de sobrevivência. Com isso, a lucratividade
sobre a receita bruta não ultrapassou 5,41%
representada no cenário F, obtendo o pior
índice de -17,22% no cenário I. Isto deve-se
a elevada conversão alimentar considerada,
custo com a ração, além do preço de venda
do peixe mais baixo e da taxa de sobrevi-
vência que reduziu-se para 80%. Pois
quando estocado juvenis com peso médio
de 293,38±5,67 g, como o utilizado neste
estudo, apenas em casos muito pessimista
o risco de mortalidade irá ultrapassa 20%. SILVA, RABENSCHLAG, FEIDEN, BOSCOLO, SIGNOR E BUENO Bruta (%)
(20/10)x100
18,20%
21,27%
29,13%
31,78%
1,73%
5,41%
-1,15%
2,65%
-17,22%
-12,83%
22
Índice de Lucratividade (Fator)
(20 / 11)
1,1820
1,2127
1,2913
1,3178
1,0173
1,0541
0,9885
1,0265
0,8278
0,8717
23
Retorno Sobre Investimento (ROI)
1,2973
1,3609
1,4756
1,5393
1,0282
1,0919
0,9813
1,0449
0,7187
0,7823
Cenários a partir dos dados da pesquisa de campo e índices zootécnicos
Cálculo
(itens)
Itens
Nº
Indicador Archivos de zootecnia vol. 61, núm. 234, p. 251. AGRADECIMENTOS Ao convênio AS/CT/0100/05 firmado en-
tre o Ministério da Pesca e Aquicultura,
Itaipu Binacional, Fundação Universitária
de Toledo e Universidade Estadual do Oeste
do Paraná, pela estrutura e insumos
fornecidos para a realização do experimento. A análise de retorno sobre o investi- SILVA, RABENSCHLAG, FEIDEN, BOSCOLO, SIGNOR E BUENO mil reais na margem de contribuição alteran-
do o ponto de equilíbrio positivamente em
11,63%. Ou seja, o fator preço de venda do
peixe ocasionou um incremento na lucrati-
vidade sobre a receita bruta de 2,53% maior
no cenário B em relação ao A. Esta proporção
também pode ser vista nos cenários C e D,
os quais foram simulados com o menor custo
com ração 0,6637 R$/kg e CA de 2,49, com
isto verificou-se que o índice de lucratividade
elevou-se a medida que o preço de venda do
peixe aumentou e consequentemente o
retorno sobre o investimento foi de 1,3609
R$/kg. mento (ROI) demonstrou que as melhores
projeções encontram-se para os índices de
1,2973; 1,3609; 1,4756; 1,5993 R$/kg nos
cenários A a D resaltando que para o
empreendedor que pretende investir no
agronegócio do peixe cada um real investi-
do é possível obter um ganho de 0,30 a 0,54
centavos. Castagnolli, N. e Zuim, S.M.F. 1985. Consolidação
do conhecimento adquirido sobre Colossoma Carneiro, D.J., Wagner, P.M., Dias, T.C.R. e
Carvalho, D.D.G. 1992. Efeito da densidade de
estocagem e do nível de proteína bruta no
desempenho de produção do pacu (Piaractus
mesopotamicus). Resultados preliminares.
Simpósio Brasileiro de Aqüicultura, 7. Anais...
SIMBRAQ. Peruíbe, SP. pp. 14. digestibilidade aparente da energia e nutrientes
da farinha de resíduos da indústria de filetagem
de tilápias, para a tilápia do Nilo (Oreochromis
niloticus). Cienc Rural, 38: 2579-2586.
Carneiro, D.J., Wagner, P.M., Dias, T.C.R. e
Carvalho, D.D.G. 1992. Efeito da densidade de
estocagem e do nível de proteína bruta no
desempenho de produção do pacu (Piaractus
mesopotamicus). Resultados preliminares.
Simpósio Brasileiro de Aqüicultura, 7. Anais...
SIMBRAQ. Peruíbe, SP. pp. 14. RETORNO ECONÔMICO DA PRODUÇÃO DE PACU EM TANQUES-REDE mitrei (Berg, 1985). FCAV/Unesp. Jaboticabal. Jomori, R., Carneiro, D.J. and Portella, M.C. 2003. Growth and survival of pacu Piaractus
mesopotamicus (Holmberg, 1887) juveniles
reaed in ponds or at different initial larviculture
periods indoors. Aquaculture, 221: 277-287. mitrei (Berg, 1985). FCAV/Unesp. Jaboticabal. Conte, L. 2002. Produtividade e economicidade da
tilapicultura em gaiolas na região sudoeste do
estado de São Paulo: Estudos de casos. Dissertação de Mestrado em Agronomia
(Ciência Animal e Pastagens). Universidade de
São Paulo. ESALQ/USP. Piracicaba. 59 pp. Kubitza, F. 1999. Tanques-rede, rações e impacto
ambiental. Rev Panorama Aquicult, 51: 44-50. Crivelenti, L.Z., Borin, S., Pirtouscheg, A., Neves,
J.E.G. e Abdão, E.M. 2006. Desempenho
econômico da criação de tilápias do Nilo
(Oreochromis niloticus) em sistema de
produção intensiva. Rev Vet Notícias, 12: 117-
122. Kubitza, F. e Ono, E.A. 2004. Projeto aqüícolas:
planejamento e avaliação econômica. Cultivo
de peixes em tanques-rede. 1ª ed. F. Kubitza. Jundiaí. 87 pp. Kubitza, F. e Campos, J.L. 2005. Desafios para a
consolidação da tilapicultura no Brasil. Rev
Panorama Aquicult, 15: 14-21. Dillon, P.J. and Rigler, F.H. 1974. Thephosphorus-
chlorophyllrelationshipin lakes. Limnol Oceanogr,
19: 767-773. Marengoni, N.G. e Bueno, G.W. 2007. Viabilidade
econômica na produção de tilápias utilizando
diferentes biomassas em tanques rede no
reservatório da UHE de Rosana-SP. Congresso
Internacional de Zootecnia, 4., Congresso
Brasileiro de Zootecnia, 17. Anais... UEL. Londrina. Eler, M.N. e Millani, T.J. 2007. Métodos de estudos
de sustentabilidade aplicados á aqüicultura. Rev Bras Zootecn, 36: 33-44. Faria, R.H.S., Souza, M.L.R., Wagner, P.M., Povh,
J.A. e Ribeiro, R.P. 2003. Rendimento do
processamento da tilapia do Nilo (Oreochroms
niloticus Linnaeus, 1757) e do pacu (Piaractus
mesopotamicus Holmberg, 1887). Acta Sci Anim
Sci, 25: 21-24. Matsunaga, M., Bemelmans, P.F., Toledo, P.E.N.,
Dulley, R.D., Okawa, H. e Pedroso, I.A. 1976. Metodologia de custo de produção utilizada pelo
IEA. Agr São Paulo, 23: 123-139. Medeiros, F.C. 2002. Tanque-rede: mais tecnologia
e lucro na piscicultura. Centro América. Cuiabá. 110 pp. FAO. 2010. Food and Agriculture Organization. The State of World Fisheries and Aquaculture
FAO. Rome. 197 pp. MPA-Ministério da Pesca e Aquicultura. 2010. Estatística da pesca e aquicultura no Brasil
2008/2009. http://www.mpa.gov.br/#info-
estatistica/estatistica-da-pesca-e-aquicultura
(25-11-2010). Filipetto, J.E.S. 2004. Normas para instalação de
pisciculturas. Em: B. Baldisserotto, J. Radünz
Neto (Orgs.). Criação de jundiá. 1ª ed. Editora da
UFSM. Santa Maria. pp. 13-34. Furuya, W.M. 2007. Redução do impacto ambiental
por meio de ração. Seminário de Aves e Suínos,
7. Anais... Avesui. Belo Horizonte. pp.121-139. Ono, E.A. BIBLIOGRAFIA Andrade, R.L.B., Wagner, R.L., Mahl, I. e Martins,
R.S. 2005. Custos de produção de tilápias
(Oreochromis niloticus) em um modelo de
propriedade da região oeste do Estado do Paraná,
Brasil. Cienc Rural, 35: 198-203. Ayroza, L.M.S. 2009. Criação de Tilápia-do-Nilo,
Oreochromis niloticus, em tanques-rede, na
Usina Hidrelétrica de Chavantes, Rio Parana-
panema. Tese (doutorado). Universidade
Estadual Paulista. Centro de Aquicultura. Jaboticabal. 92 pp. Boscolo, W.R., Hayashi, C., Feiden, C., Meurer, F. e Signor, A.A. 2008. Composição química e Castagnolli, N. e Zuim, S.M.F. 1985. Consolidação
do conhecimento adquirido sobre Colossoma Archivos de zootecnia vol. 61, núm. 234, p. 252. Archivos de zootecnia vol. 61, núm. 234, p. 253. Scorvo-Filho, J.D., Pinto, C.S.R.M., Verani, J.R. e
Silva, A.L. 2006. Custo operacional de produção
da criação de tilápias vermelha da flórida e
tailandesa em tanques-rede de pequeno volume.
Informações Econômicas, 36: 71-79. Archivos de zootecnia vol. 61, núm. 234, p. 254. RETORNO ECONÔMICO DA PRODUÇÃO DE PACU EM TANQUES-REDE e Kubitza, F. 2003. Cultivo de peixes em
tanques-rede. Jundiaí. 112 pp. Furlaneto, F.P.B., Ayroza, D.M.M.R. e Ayroza,
L.M.S. 2006. Custo e rentabilidade da produção
de tilápia (Oreochromis ssp.) em tanque-rede
no médio Paranapanema, estado de São Paulo,
Safra 2004/2005. Informações Econômicas,
SP, 36: 63-69. Ostrensky, A., Borghetti, J.R. e Sato, D. 2008. Aqüicultura no Brasil: o desafio é crescer. Brasília. 276 pp. Petrere, J.R.M. 1989. River fisheries in Brazil: A
review. Regul River, 4: 1-16. g
,
Richter, G.O. 2004. Pesca e aqüicultura. Secreta- Richter, G.O. 2004. Pesca e aqüicultura. Secreta-
ria de Estado e do Abastecimento (SEAB/
DERAL). Departamento de Economia Rural-
Divisão de Conjuntura Agropecuária do Estado
do Paraná (DCA). Curitiba. 35 pp. Godinho, H.P. 2007. Estratégias reprodutivas de
peixes aplicada à aquicultura: bases para o
desenvolvimento de tecnologias de produção. Rev Bras Rep Anim, 31: 351-360. IAP-Instituto Ambiental do Paraná. 2010. Licen-
ciamento e autorização. www.iap.pr.gov.br (15/
01/2010). Signor, A.A., Boscolo, W.R., Bittencourt, F.,
Coldebella, A. e Reidel, A. 2010. Proteína e
energia na alimentação de pacus criados em
tanques-rede. Rev Bras Zootecn, 39: 2336-
2341. Itaipu Binacional. 2007. Plano Diretor do Reservatório
de Itaipu. http://www.itaipu. gov.br (12/10/2010). Archivos de zootecnia vol. 61, núm. 234, p. 253. Archivos de zootecnia vol. 61, núm. 234, p. 253. Scorvo-Filho, J.D., Martin, N.B. e Ayroza, L.M.S.
1998. Piscicultura em São Paulo: custo e retor-
nos de diferentes sistemas de produção na
safra de 1996/1997. Informações Econômicas,
28: 41-61. Scorvo-Filho, J.D., Pinto, C.S.R.M., Verani, J.R. e
Silva, A.L. 2006. Custo operacional de produção
da criação de tilápias vermelha da flórida e
tailandesa em tanques-rede de pequeno volume.
Informações Econômicas, 36: 71-79. Scorvo-Filho, J.D., Martin, N.B. e Ayroza, L.M.S.
1998. Piscicultura em São Paulo: custo e retor-
nos de diferentes sistemas de produção na
safra de 1996/1997. Informações Econômicas,
28: 41-61.
|
https://openalex.org/W2142363124
|
https://ccsenet.org/journal/index.php/emr/article/download/21652/14108
|
English
| null |
Framework for Optimizing Team Performance and Project NPV: Enhancing the Probability of Success by Team Alignment
|
Engineering management research
| 2,012
|
cc-by
| 10,892
|
Abstract This study addresses a gap in enterprise risk management related to project team performance. Poorly
functioning teams may severely erode project net present value (NPV). The erosion of project NPV can be
quantified in terms of probability of success (POS). In the oil business POS is based on success criteria for
likelihood that exploration efforts for oil & gas prospects will realize the EMV for those assets. Similarly, POS
in team work and negotiations is based on success criteria for the likelihood that cooperation between team
individuals will be able to deliver the maximum value for the project. Practical rules are formulated to support
teams and team leaders in their efforts to optimize the alignment of team members in order to enhance the team’s
effectiveness. The probability of success (POS) is split into three fundamental factors of alignment: PCulture , PSkills
and PGoals. The dynamic effect of team learning on team alignment is graphed as the Cumulative POS. The cost
of failure is graphed for a range of POS values, and visualizes the impact on the EMV of extra Team OPEX,
each normalized by the project NPV. Applications are possible in all kinds of functional teams, including change
management teams that need to build coalitions to effectuate lasting change. The interaction between members
of engineering and other professional teams has been studied intensively, but the expression of team
performance in numbers as quantified here is a new direction. Keywords: NPV optimization, team performance, probability of success, probability of failure, oil and gas
projects Received: May 11, 2012 Accepted: August 10, 2012 Online Published: September 14, 2012
doi:10.5539/emr.v1n2p107 URL: http://dx.doi.org/10.5539/emr.v1n2p107 Received: May 11, 2012 Accepted: August 10, 2012 Online Published: September 14, 2012
doi:10.5539/emr.v1n2p107 URL: http://dx.doi.org/10.5539/emr.v1n2p107 Received: May 11, 2012 Accepted: August 10, 2012 Online Published: September 14, 2012
doi:10.5539/emr.v1n2p107 URL: http://dx.doi.org/10.5539/emr.v1n2p107 Engineering Management Research; Vol. 1, No. 2; 2012
ISSN 1927-7318 E-ISSN 1927-7326
Published by Canadian Center of Science and Education Engineering Management Research; Vol. 1, No. 2; 2012
ISSN 1927-7318 E-ISSN 1927-7326
Published by Canadian Center of Science and Education Framework for Optimizing Team Performance and Project NPV:
Enhancing the Probability of Success by Team Alignment Ruud Weijermars
1 Alboran Energy Strategy Consultants and Department of Geoscience and Engineering, Delft University of
Technology, Delft, Netherlands
Correspondence: Ruud Weijermars, Department of Geoscience and Engineering, Delft University of Technology,
Stevinweg
1,
PO
box
5048,
2600GA
Delft,
Netherlands. Tel:
31-15-2787-801. E-mail:
R.Weijermars@TUDelft.nl Technology, Delft, Netherlands
Correspondence: Ruud Weijermars, Department of Geoscience and Engineering, Delft University of Technology,
Stevinweg
1,
PO
box
5048,
2600GA
Delft,
Netherlands. Tel:
31-15-2787-801. E-mail:
R.Weijermars@TUDelft.nl Correspondence: Ruud Weijermars, Department of Geoscience and Engineering, Delft University o
Stevinweg
1,
PO
box
5048,
2600GA
Delft,
Netherlands. Tel:
31-15-2787-8
R.Weijermars@TUDelft.nl Received: May 11, 2012 Accepted: August 10, 2012 Online Published: September 14, 2012
doi:10.5539/emr.v1n2p107 URL: http://dx.doi.org/10.5539/emr.v1n2p107 1. Introduction The combination of knowledge from what are traditionally poorly connected disciplines, petroleum exploration
and social philosophy, provides a powerful tool for monitoring and improving team performance. Oil and gas
companies have invested considerable effort into the development of uncertainty and risk analysis tools to better
establish the anticipated net present value (NPV) of projects (Willigers & Majou, 2010). Although cognitive bias
in decisions for oil and gas projects has been highlighted as a human risk factor in recent studies (Welsh et al.,
2008), the role of professionals in realizing full project NPV is crucial and merits further study. The optimization
model outlined here rates the degree of team alignment and translates this into team effectiveness. Engineering and other functional specialist teams that prove unable to collaborate effectively are a corporate risk
factor that may jeopardize the project outcome and could cause shortfalls in the realization of the anticipated
NPV. A portion of the organizational resources and energy will go lost if the team process is ineffective (Figure
1). For example, if a team manages to collaborate only with 75% effectiveness, some 25% of the projected
project outcome may remain unrealized. If additionally a portfolio management team’s efforts remain also
suboptimum, capturing only 2/3 of the potential value, then 50% of the company’s value-adding potential
remains unrealized (Figure 1). Portfolio optimization is a corporate strategy planning responsibility and the
realization of the full portfolio value critically hinges on the quality of the portfolio team’s work. This study focuses on the performance of approved projects in the portfolio. Project execution is an operational
responsibility for business units and project teams. The generic vision for team work is to add value for their
organizations. The successful execution of the selected projects or portfolios must generate the anticipated cash
flows. This means capital expenditure (CAPEX) on any of such projects must result in project completion on 107 Engineering Management Research Vol. 1 No. 2; 2012 www.ccsenet.org/emr time, within budget and safely. The model outlined seeks to influence the success of project teams as an outcome
of the degree of collaborative performance development over time. time, within budget and safely. The model outlined seeks to influence the success of project teams as an outcome
of the degree of collaborative performance development over time. Figure 1. Optimization of the corporate strategy is critically dependent on operational project performance. 1. Introduction Portfolio balancing techniques ensure that project impact is buffered against adverse project execution conditions,
but the aim is that each investment project realizes the anticipated value Figure 1. Optimization of the corporate strategy is critically dependent on operational project performance. Portfolio balancing techniques ensure that project impact is buffered against adverse project execution conditions,
but the aim is that each investment project realizes the anticipated value The monitoring of the progressive team alignment can quantify the likelihood that a project NPV will be realized
as planned, or even optimized by the team when possible. The Expected Monetary Value (EMV) of a project in
progress at times T1, T2, T3, and TN will vary based on the effectiveness of team work (Figure 2). The projects
originally anticipated NPV should be fully realized at TN. Less effective teams will fail to optimize their
alignment on the maximum performance and are likely to overrun project budgets and timelines for project
completion. For project success, the EMV must match the original NPV, or even beat expectations. The model is
designed to proactively monitor and manage changes in the degree of alignment (or degree of collaboration)
between team members (Figure 2). Our approach supports industry in its ambition to manage talent and
performance by numbers (Baker, 2008). The monitoring of the progressive team alignment can quantify the likelihood that a project NPV will be realized
as planned, or even optimized by the team when possible. The Expected Monetary Value (EMV) of a project in
progress at times T1, T2, T3, and TN will vary based on the effectiveness of team work (Figure 2). The projects
originally anticipated NPV should be fully realized at TN. Less effective teams will fail to optimize their
alignment on the maximum performance and are likely to overrun project budgets and timelines for project
completion. For project success, the EMV must match the original NPV, or even beat expectations. The model is
designed to proactively monitor and manage changes in the degree of alignment (or degree of collaboration)
between team members (Figure 2). Our approach supports industry in its ambition to manage talent and
performance by numbers (Baker, 2008). Figure 2. Alignment of team members needs to develop and improve over time (T1->TN) in order to optimize
the collaboration between team members. Quantification of the effect of team alignment on project outcome is
the focus of this study Figure 2. 2. Previous Team Work Models Most organizations tend to be goal-oriented (here synonymous with task-oriented) and less people-oriented. Goal-oriented organizations run a risk of overlooking and disregarding the importance of improving team
collaboration. Previous studies have already demonstrated that goal-oriented teams (star teams being the extreme,
e.g., Fischer & Boynton, 2005) may become surprisingly ineffective if they do not manage to fully align their
talents into collaboration for the team task (Gratton & Erickson, 2007). A people-oriented approach, as newly
advocated in this study, is strongly recommended - if not a prerequisite for achieving an optimum project
outcome with teams. Team members will work goal-oriented only as long as people issues are properly
addressed. Team members that do not learn to understand each other cannot contribute super-effectively to the
goals of their team. Earlier studies focused on the importance of collaboration and concluded that members of large, complex teams
can still collaborate effectively if the right conditions are in place (Gratton & Erickson, 2007). Teams that are
equipped with attributes to be effective in collaboration (large skills base, virtually connected, from diverse
backgrounds and experienced professionals) are effective in their collaboration when team member alignment is
optimum. Decision-making theories for predicting team outcomes commonly focus on the quantification of
uncertainty in team outcomes, based upon the possible range of decisions of individual members (Axelrod, 1997;
Ridley, 1998). Other models study the impact of team collaboration on the trade-off between the team’s
creativity to ‘think-out-of-the-box’ and convergence of ideas toward project completion with consensus, rather
than divergence of ideas (Amabile, 1996, 1998; Thompson, 2003; Hoegl & Parboteeah, 2007). Figure 3. Cornelis generic concept of Feeling’s Logic with probabilities (chance factors) indicated for each
stability layer (modified from Cornelis, 1988) Figure 3. Cornelis generic concept of Feeling’s Logic with probabilities (chance factors) indicated for each
stability layer (modified from Cornelis, 1988) We advance the progressive consolidation of alignment in teams based on the work of Cornelis (1988, ‘Feelings
Logic’, written in Dutch and never translated into English; but briefly outlined in Weijermars, 2011). Cornelis
provided a rational framework for the role of emotions in communication across socio-cultural systems. The
great merit of the Cornelis 'Feelings Logic' lies in his recognition of the importance of emotional responses of
team members to build social stability layers (Figure 3). 1. Introduction Alignment of team members needs to develop and improve over time (T1->TN) in order to optimize
the collaboration between team members. Quantification of the effect of team alignment on project outcome is
the focus of this study The merits of our model are most critical when a team is newly assembled and faces tasks that are challenging
and need new solutions rather than repetitive application of an established workflow. Examples of tasks in an
established, rigid workflow are mostly blue collar projects. Problem solving teams are mostly engaged in white
collar projects. 108 Vol. 1 No. 2; 2012 Engineering Management Research www.ccsenet.org/emr 3. Cost of Failure in Optimizing Team Alignment The three major steps implied by the Cornelis model for the team alignment process are similar to the first three,
essential steps in the well-known Drexler-Sibbet team performance model (Figure 4), which is used here as a
reference frame for monitoring the probability of success (or failure) as teamwork. Figure 4. Effective team-functioning critically depends not only upon a spread of professional expertise: All
team members must succeed in aligning their attitude towards collaboration. Only optimized alignment will
enable positive knowledge sharing and team focus resulting in optimum outcomes (constructive, green track). Poor team spirit leads to progressive corruption of cooperation and torpedoes knowledge sharing activities
(destructive, red track). (Simplified from Drexler et al., 1988) Figure 4. Effective team-functioning critically depends not only upon a spread of professional expertise: All
team members must succeed in aligning their attitude towards collaboration. Only optimized alignment will
enable positive knowledge sharing and team focus resulting in optimum outcomes (constructive, green track). Poor team spirit leads to progressive corruption of cooperation and torpedoes knowledge sharing activities
(destructive, red track). (Simplified from Drexler et al., 1988) For effective team work, clearly non-alignment (destructive, red track in Figure 4) must be mitigated as it leads
to inefficiencies and waste in the company’s resources and energy. Assuming that the speed of correction for
non-alignment in teams can be positively affected by the team members, the following rules must be observed: For effective team work, clearly non-alignment (destructive, red track in Figure 4) must be mitigated as it leads
to inefficiencies and waste in the company’s resources and energy. Assuming that the speed of correction for
non-alignment in teams can be positively affected by the team members, the following rules must be observed: 1. The principal mechanism for optimizing the effectiveness of teams, once members are chosen, resides
foremost in the degree of team alignment. 1. The principal mechanism for optimizing the effectiveness of teams, once members are chosen, resides
foremost in the degree of team alignment. 2. All things being equal, the team can optimize its performance with the members’ skills and talent pool
only by maximizing collaborative alignment (including using the full range of personal creativity for
solving the challenges at hand). 3. 2. Previous Team Work Models Team members that understand and ‘grasp’ each other’s
cultural layer feel trusted, while they may still recognize culturally distinct habits of one another. Individuals in
social entities seek to share a common language, adhere to cultural traditions, and display a tendency to respect
similar customs and beliefs. The ‘Feelings Logic’ model of Cornelis (1988), partly covered by the better known
concept of building emotional intelligence in teams (Goleman, 1997, 2000; Druskat & Wolff, 2001), provides a
comprehensive intellectual framework for the development of emotional intelligence in groups. The Cornelis model of ‘Feelings Logic’ conceptualizes the embedding of emotions in so called ‘social layers of
stability’ (Figure 3); three such layers are distinguished: ‘Culture’, ‘Skills’ and ‘Self-actualization’. The social
stability layers of Cornelis ‘stack’ on top of one another -optimizing the efficiency in the skills layer is
contingent upon fulfilling the culture layer. Each layer needs conditional fulfilment for any action on the
subsequent level to be effectively undertaken and may be applied to micro-scale (social entities such as teams),
or to macro-scale (societal level). The theory of Cornelis (1988) has been previously expanded with a 109 Engineering Management Research Vol. 1 No. 2; 2012 www.ccsenet.org/emr probabilistic approach which improves the chances of success in negotiations and subsequent collaboration
between national (NOCs) and international oil companies (IOCs) (Weijermars et al., 2008). The theory has also
been applied to international cooperation between gas companies (Weijermars & De Jong, 2008). probabilistic approach which improves the chances of success in negotiations and subsequent collaboration
between national (NOCs) and international oil companies (IOCs) (Weijermars et al., 2008). The theory has also
been applied to international cooperation between gas companies (Weijermars & De Jong, 2008). 3. Cost of Failure in Optimizing Team Alignment Most organizations tend to be goal-oriented and less people-oriented, but alignment of team members
develops via shared core values and improves over time as the team focus grows The assumption is that teams will be best able to deliver an expected project net present value (NPV) when the
POS for the team is higher. A team’s probability of success (POS) can thus be determined by the following
product: (1) POS = PCulture * PSkills* PGoals POS = PCulture * PSkills* PGoals
(1) The Expected Monetary Value (EMV) that can be realized by a project team is given by (Weijermars et al., 2008;
reformulating Johnston, 2003): The Expected Monetary Value (EMV) that can be realized by a project team is given by (Weijermars et al., 2008;
reformulating Johnston, 2003): EMV = (POS * NPV) – [(1-POS)*Extra Team Cost] (2) (2) It is clear that a team with zero alignment, i.e. POS = PCulture * PSkills* PGoals = 0, will not realize any positive EMV. The probability of failure (POF) grows as POS nears zero, i.e. POF = 1-POS. The company is left with the cost
of the teamwork, a negative result. In contrast, the optimum EMV is achieved when instant alignment is realized,
i.e. POS = 1, so that EMV = NPV. However, this is not a likely outcome as in most cases team performance
needs time to improve. The chance of failure can be 100%, i.e. POF=1, or zero if POS is 100% or 1, and ranges between 1 and 0 for
POS values between 0 and 1. It is clear that team members that learn to optimize their POS in a people-oriented
team approach will thereby achieve the highest possible effectiveness for their team. Figure 6. Expected team project outcome in monetary terms as EMV (normalized by the maximum NPV of the
project) versus Team Cost (also normalized by the NPV) for POS values between 0 and 1
Figure 6 graphs the EMV, normalized by the NPV versus the Cost of Team Alignment, showing the impact of
POS for the degree of team alignment (POS=PCulture * PSkills* PGoals). The probability of success in the case of
moderate team alignment with POS= 0.5 implies an EMV less than half the calculated NPV of the project. Only Figure 6. 3. Cost of Failure in Optimizing Team Alignment The intrinsic possibility of success (POS) for the team can improve by a combination of individual
learning to improve the skills of team members and by team learning to enhance the shared values and
vision. 4. Cost of team work increases over time, but is offset by the quality and value of the expected project
outcome, both of which can be maximized in team efforts where team members succeed in optimizing
team alignment. Goal-oriented organizations run a risk of overlooking and disregarding the importance of improving team
collaboration, which requires a people-oriented approach based on a foundation of strong core values. When
people and functional teams do not share common understanding and do not passionately support their corporate
values and goals, this may result in that these organizational goals are not or ineffectively realized. The key to
optimizing team performance lies locked inside the degree of alignment between team members. Progressive
improvement of alignment in teams leads to better project outcomes. Goal-oriented organizations run a risk of overlooking and disregarding the importance of improving team
collaboration, which requires a people-oriented approach based on a foundation of strong core values. When
people and functional teams do not share common understanding and do not passionately support their corporate
values and goals, this may result in that these organizational goals are not or ineffectively realized. The key to
optimizing team performance lies locked inside the degree of alignment between team members. Progressive
improvement of alignment in teams leads to better project outcomes. Figure 5 provides a template for optimizing team alignment by capitalizing on insights from the Cornelis model
on the social factors that control the alignment between team members (Figure 5). This template can be used to
monitor the degree of team alignment as work is prepared, starts and progresses in teams. Progressive 110 Engineering Management Research Vol. 1 No. 2; 2012 www.ccsenet.org/emr improvement of alignment in teams by starting people-oriented instead of goal-oriented, commonly leads to
better outcomes. improvement of alignment in teams by starting people-oriented instead of goal-oriented, commonly leads to
better outcomes. Figure 5. Most organizations tend to be goal-oriented and less people-oriented, but alignment of team members
develops via shared core values and improves over time as the team focus grows Figure 5. 3. Cost of Failure in Optimizing Team Alignment Expected team project outcome in monetary terms as EMV (normalized by the maximum NPV of the
project) versus Team Cost (also normalized by the NPV) for POS values between 0 and 1 Figure 6 graphs the EMV, normalized by the NPV versus the Cost of Team Alignment, showing the impact of
POS for the degree of team alignment (POS=PCulture * PSkills* PGoals). The probability of success in the case of
moderate team alignment with POS= 0.5 implies an EMV less than half the calculated NPV of the project. Only 111 Engineering Management Research Vol. 1 No. 2; 2012 www.ccsenet.org/emr half the cost of team alignment is subtracted from the product of NPV*POS (see Eq. 2), because half the team’s
effort contributed to realize the EMV, but the other half did not. half the cost of team alignment is subtracted from the product of NPV*POS (see Eq. 2), because half the team’s
effort contributed to realize the EMV, but the other half did not. While Equation 2 expresses the stochastically correct EMV, an alternative Equation subtracts the full value of
the cost of team alignment, in analogy with the cost of exploration commonly used in mineral exploration
programs (e.g., Lord et al., 2001; Olivier et al., 2008): (3) Figure 7. Expected project outcome in monetary terms as EMV* (normalized by the maximum NPV of the
project) versus Team Cost (also normalized by the NPV) for POS values between 0 and 1 Figure 7. Expected project outcome in monetary terms as EMV* (normalized by the maximum NPV of the
project) versus Team Cost (also normalized by the NPV) for POS values between 0 and 1 Figure 8. Flow chart for tracking team alignment and intrinsic POS scores for the team at a particular time TN,
corresponding to arbitrary alignment stages shown in Figure 2 Figure 8. Flow chart for tracking team alignment and intrinsic POS scores for the team at a particular time TN,
corresponding to arbitrary alignment stages shown in Figure 2 The relationship between EMV* and Team Cost, both normalized by the NPV, is graphed in Figure 8, according
to Equation 3. Projects where the ratio Team Cost versus NPV is larger than 1 will yield negative EMV for all
cases. Therefore, for most organizations, Equation 3 (and Figure 7) may provide a criterion better than Equation
2 (and Figure 6) to decide whether to terminate or kill a project. 4. Guidelines for Team Leaders for Monitoring Team Alignment The rules for optimizing the POS in team performance that is made up of the product PCulture * PSkills* PGoals (see Eq. 1) are outlined in this section. The flow chart of Figure 8 serves as a practical guideline for tracking team
alignment and for tracking progress or regress in a team’s intrinsic POS. Assigning an experienced team leader who pays attention to people-skills as well as goal-setting issues is
essential for the team’s success. The regular assessment of the leadership styles of team leaders remains
important; they must be familiar with the mechanisms of group dynamics and basic leadership models (Blake &
Mouton, 1964; Fiedler, 1967; Kirkpatrick &Locke, 1991; Goleman, 1996; Collins, 2001). The most basic social stability layer is the ‘Culture’ level, which refers to cultural ‘embeddedness’ of social
groups, teams and companies alike, and the extent to which they have adopted a common language, norms and
customs. Building team trust by addressing PCulture requires skilful management from the team leader. Ideally,
individuals in social entities (teams) share a common language, adhere to cultural traditions, and display similar
customs and beliefs. One anecdote by George DeVries Klein (pers. comm. 7/11/ 2008) illustrates the point: “I
recall one meeting involving a team of which I was a member, and the client's team. We all gave our
presentations to the client team, but the geophysicist from the client-team was concerned about a difference
between our geophysicist's use of 180-degree phase and their preferred 90-degree phase. While they slugged it
out, the rest of us just sat there wondering why the two of them couldn't just work it out away from the team
meeting. Neither team leader stepped in to make that happen. Although eventually resolved, it left hard feelings
that weren't needed and made for a difficult working relationship between both groups. During a break, I met
with my counterpart to smooth things over and that at least led to some progress.” Factors that are seen to improve interaction between members of engineering teams and result in positive
collaboration are (Gratton & Erickson, 2007): role models of collaboration by the senior executives, investment
in facilities that foster open communication and collaboration, mentoring and coaching for cooperation in
networks across corporate boundaries, training to improve people skills (teamwork, emotional intelligence,
conflict resolution, networking and holding difficult conversations), building a sense of community (i.e. 3. Cost of Failure in Optimizing Team Alignment Alternatively, an argument against starting a
project could be made when the anticipated cost for the required team alignment renders the ratio Team Cost
versus NPV unattractive. For this consideration to be made, the effect of intrinsic POS augmentation as a result
of team learning, were quantified in Figures 6 and 7. The relationship between EMV* and Team Cost, both normalized by the NPV, is graphed in Figure 8, according
to Equation 3. Projects where the ratio Team Cost versus NPV is larger than 1 will yield negative EMV for all
cases. Therefore, for most organizations, Equation 3 (and Figure 7) may provide a criterion better than Equation
2 (and Figure 6) to decide whether to terminate or kill a project. Alternatively, an argument against starting a
project could be made when the anticipated cost for the required team alignment renders the ratio Team Cost
versus NPV unattractive. For this consideration to be made, the effect of intrinsic POS augmentation as a result
of team learning, were quantified in Figures 6 and 7. 112 Engineering Management Research Vol. 1 No. 2; 2012 www.ccsenet.org/emr 4. Guidelines for Team Leaders for Monitoring Team Alignment core
values an social bonding), assigning team leaders that are both people-oriented and task-oriented, and picking
teams that have a core of people that hold shared values, and then ensuring role clarity and task latitude. To help team leaders assess the POS of team performance, a workflow template is provided (Figure 8). The three
principal steps are outlined below: Step 1: As a first step, a team leader should assess the team’s cultural alignment (Figure 8, Step 1). The score for
POS of a team critically depends on the incremental success in overcoming the individual resistances to team
alignment. Team alignment at the cultural level is a first hurdle for optimizing the POS. The team score for
PCulture will be higher as individual scores Pj Culture of team members increase, according to:
n PCulture = (j=1;
n Pj Culture/)/N
(4) (4) Table 1. Chance Factors assigned to individual team member compliance with the criteria set of Table 2
Criteria Compliance
Chance Factors
Outstanding
0.9-1
Excellent
0.8-0.9
Good
0.7-0.8
Satisfactory
0.5-0.7
Insufficient
0.3-0.5
Weak
0.1-0.3
Poor
0-0.1 Table 1. Chance Factors assigned to individual team member compliance with the criteria set of Table 2
Criteria Compliance
Chance Factors
Outstanding
0.9-1
Excellent
0.8-0.9
Good
0.7-0.8
Satisfactory
0.5-0.7
Insufficient
0.3-0.5
Weak
0.1-0.3
Poor
0-0.1
With N the number of team members. The team’s current PCulture at time Tx can be succinctly scored using two
criteria: (A) the degree of integration of individual team members with the corporate culture and communication
style, and (B) intercultural experience of team members. The scale proposed for scoring the PCulture criteria is
shown in Table 1. The foundation for achieving further cultural alignment in the team (PCulture) lies in ensuring
that: With N the number of team members. The team’s current PCulture at time Tx can be succinctly scored using two
criteria: (A) the degree of integration of individual team members with the corporate culture and communication
style, and (B) intercultural experience of team members. The scale proposed for scoring the PCulture criteria is
shown in Table 1. The foundation for achieving further cultural alignment in the team (PCulture) lies in ensuring
that: With N the number of team members. The team’s current PCulture at time Tx can be succinctly scored using two
criteria: (A) the degree of integration of individual team members with the corporate culture and communication
style, and (B) intercultural experience of team members. 4. Guidelines for Team Leaders for Monitoring Team Alignment The scale proposed for scoring the PCulture criteria is
shown in Table 1. The foundation for achieving further cultural alignment in the team (PCulture) lies in ensuring
that: 113 Vol. 1 No. 2; 2012 Engineering Management Research www.ccsenet.org/emr
Core values are well-integrated into the company’s corporate culture and are shared by all team
members.
Core values are well-integrated into the company’s corporate culture and are shared by all team
members.
Teams can be formed and function without major resistances between team members. Team members
have an excellent social cohesion.
Teams can be formed and function without major resistances between team members. Team members
have an excellent social cohesion.
Teams can be formed from people most likely to enter into a knowledgeable dialogue about the task at
hand.
Teams can be formed from people most likely to enter into a knowledgeable dialogue about the task at
hand. The mechanisms for better alignment of PCulture in multicultural teams can be abstracted from the detail studies of
the impact of cultures on teams and negotiations (Trompenaars & Hampden-Turner, 1997; Hofstede & Hofstede,
2004). Further anecdotes of failed projects in the petroleum industry due to neglecting of PCulture have been
recounted by Klein (1999). In case any of the prerequisites for high team score of PCulture are not fulfilled, it is
important to not proceed until these prerequisites are in place by intervention and remediation of the situation
(Figure 8, track 1B). Step 2: The second step for the team is to work on raising their score for alignment at the skills level or PSkills. The ‘Skills’ level refers to the specific set of skills the people in a certain social group or team require for
successful functioning. The emotional responses related to the skill layer are 'confidence' for those team
members who have the appropriate skills - and ‘frustration’ for those that do not. Personal creativity and variety
in innate cognitive styles can be inventoried using Myers-Briggs type inventories and diversity spread can be
established using the models of Belbin (1993), Herrmann (1996) and Lumsdaine et al. (1999). These are all
useful analytical tools that help to improve insight in the personal traits that affect team chemistry and creative
skills. Skills include the recognition of - and adherence to - the economic, legal and political systems. 4. Guidelines for Team Leaders for Monitoring Team Alignment A team
leader (project manager) must ensure that team members have a professional attitude about their individual
strengths and weaknesses. The success of a project is based upon their mutual experience, skills, practitioner
record and education level (PSkills ):
The team will benefit if the team leader and team members jointly prepare a skill inventory needed for
the task at hand and compare this with the range of skills covered by team members.
The team will benefit if the team leader and team members jointly prepare a skill inventory needed for
the task at hand and compare this with the range of skills covered by team members.
Team members have a professional attitude about individual strengths and weaknesses, based upon an
inventory of individual experience (practitioner record & education level).
Team members have a professional attitude about individual strengths and weaknesses, based upon an
inventory of individual experience (practitioner record & education level).
Team members understand the skills needed for the team task at hand, and their preparedness is high to
share knowledge, learn and bridge skills gaps. Skill gaps can be plugged by attracting new team
members or by rapid continuing education of current members.
Team members understand the skills needed for the team task at hand, and their preparedness is high to
share knowledge, learn and bridge skills gaps. Skill gaps can be plugged by attracting new team
members or by rapid continuing education of current members.
Team members are willing to share quality assurance tools and common standards.
Skills needed by the team but not (yet) mastered by any member can be complemented by attracting
new team members, by inviting experts for advice, or by rapid continuing education of current
members.
Skills needed by the team but not (yet) mastered by any member can be complemented by attracting
new team members, by inviting experts for advice, or by rapid continuing education of current
members.
In fact, a team’s effort for an inventory of skills at hand and cross-checking with skills needed itself is
an exercise that may lead to high team scores for PSkills. There must be a preparedness to share
knowledge, to learn and to work actively to bridge skills gaps. Team members should also be willing to
share quality assurance tools and common industry standards. 4. Guidelines for Team Leaders for Monitoring Team Alignment Great leaders know how to inspire
and empower teams to formulate solutions (Collins, 2001):”First get on the bus - then we decide where we go.”
Effective goal- and direction-setting in the team (PGoals ), becomes possible only if the vision for the project’s Effective goal- and direction-setting in the team (PGoals ), becomes possible only if the vision for the project’s
objectives is shared by all team members:
A vision for the project’s objectives shared by all team members leads to effective teamwork.
A vision for the project’s objectives shared by all team members leads to effective teamwo
Specific goals are agreed, clear and supported by all team members.
Specific goals are agreed, clear and supported by all team members.
Specific goals are agreed, clear and supported by all team members.
A real sense of mission toward objectives and goals is shared by the team.
A real sense of mission toward objectives and goals is shared by the team. The team score for PGoals increases as individual scores Pj Goals of team members align, according to:
n
A real sense of mission toward objectives and goals is shared by the team. The team score for PGoals increases as individual scores Pj Goals of team members align, according to: team score for PGoals increases as individual scores Pj Goals of team members align, according to: e for PGoals increases as individual scores Pj Goals of team members align, according to: (6) PGoals = (j=1;
n Pj Goals/)/N PGoals = (j=1;
n Pj Goals/)/N
(6) The team should discuss the goals and mission objectives for the project (Figure. 8, track 3A) and make sure
there is agreement on a shared vision and goals. The team’s current PGoals at time Tx can be succinctly scored
using two criteria: (E) Vision for solutions for the project are adopted and supported, and (F) the goals are
understood by all team members. The scale proposed for scoring the PGoals criteria is as shown in Table 1. Table 2. 4. Guidelines for Team Leaders for Monitoring Team Alignment The team score for PSkills increases as individual scores Pj Skills of team members align, according to: The team score for PSkills increases as individual scores Pj Skills of team members align, according to: (5) PSkills = (j=1;
n Pj Skills/)/N
(5) PSkills = (j=1;
n Pj Skills/)/N The team should inventory its skills required for the project (Figure. 8, track 2A) and assess whether the skills
and experience base of team members are adequate. The team’s current PSkills at time Tx can be succinctly scored
using two criteria: (C) Skills adequacy, and (D) years of experience in applying the skills. The scale proposed for
scoring the PSkills criteria is as shown in Table 1. The team should inventory its skills required for the project (Figure. 8, track 2A) and assess whether the skills
and experience base of team members are adequate. The team’s current PSkills at time Tx can be succinctly scored
using two criteria: (C) Skills adequacy, and (D) years of experience in applying the skills. The scale proposed for
scoring the PSkills criteria is as shown in Table 1. Step 3: The third step is to work in teams on raising the score for alignment at the goals level or PGoals. In the
upper social stability layer, ‘self-actualization’ refers to people’s ability to formulate and realise their goals and
ambitions within a social entity. It is highly implausible for a member of any society, organization, or team to
realize their goals and ambitions (self-actualization) if not supported by a social culture (common understanding
of core values), and effective sharing of skills and knowledge. A team then may succeed in accomplishing
favourable alignment in the self-fulfilment layer. Team member are happy when this is achieved, and sad and
unsuccessful if this is not the case. Teams are commonly formed to assume responsibility for inventorying or
assessing particular types of problems and then proceed to suggest solutions. However, the method of solution 114 Engineering Management Research Vol. 1 No. 2; 2012 www.ccsenet.org/emr focus and vision for solving the challenge is entirely the team’s responsibility. Great leaders know how to inspire
and empower teams to formulate solutions (Collins, 2001):”First get on the bus - then we decide where we go.” focus and vision for solving the challenge is entirely the team’s responsibility. 4. Guidelines for Team Leaders for Monitoring Team Alignment •
Formal directions for strategic change are not supported by social structures and core values in the
company. Consequently, teams will lack a sense of direction and their poor core value adoption
handicaps the development of team focus.
The company’s elaborate procedures and well-structured hierarchy of the past do not fit the current
market dynamics. Consequently, alignment of members in teams suffers from mis-alignment between
the organizational focus and best-practice in the external business environment.
The company’s elaborate procedures and well-structured hierarchy of the past do not fit the current
market dynamics. Consequently, alignment of members in teams suffers from mis-alignment between
the organizational focus and best-practice in the external business environment.
Fast change in external environment cannot be coped with competitively by the organization’s current
technology, current people and current assets. Consequently, team members lack skills and tools to
achieve full alignment in team work.
Fast change in external environment cannot be coped with competitively by the organization’s current
technology, current people and current assets. Consequently, team members lack skills and tools to
achieve full alignment in team work.
Goals and strategy of the company’s dominant coalition are based on wrong assumptions about external
environment. Consequently, team work in such strategically disconnected organizations is frustrated by
inadequate leadership.
Goals and strategy of the company’s dominant coalition are based on wrong assumptions about external
environment. Consequently, team work in such strategically disconnected organizations is frustrated by
inadequate leadership. Figure 9. Anticipated NPV (= the cumulative cash flow over the life-cycle of the project) may be as planned in
well managed projects. Poorly managed projects result in lost cash flow. Exceptionally well managed projects
may beat the originally planned cash flow Figure 9. Anticipated NPV (= the cumulative cash flow over the life-cycle of the project) may be as planned in
well managed projects. Poorly managed projects result in lost cash flow. Exceptionally well managed projects
may beat the originally planned cash flow In case any, if the prerequisites for high team score PGoals are not fulfilled, it is important to not proceed until
these prerequisites are in place by intervention and remediation of the situation. In case any, if the prerequisites for high team score PGoals are not fulfilled, it is important to not proceed until
these prerequisites are in place by intervention and remediation of the situation. 4. Guidelines for Team Leaders for Monitoring Team Alignment Team alignment scorecard
Stability
Layer
Step 1: PCulture
Step 2: PSkills
Step 3: PGoals
Individual
POS
scores
Criteria
(A)
Corporate
Culture &
Communication
Style Fits
(B)
Level of
Multicultural
Experience
(C)
Skills
for
Task
Set
(D)
Skills
Experience
Level
(E)
Vision
Adopted
&
Supported
with
Passion
(F)
Project
Goals
Understood
&
Supported
with
Passion
Person 1
Score 1
Person 2
Score 2
Person 3
Score 3
Person N
Score N
Team
Total
Averages
PCulture(A) + Culture(B) =
(j=1;
n Pj Culture (A) + (Culture (B)/)/2N
PSkills(C) + Skills(D) =
(j=1;
n Pj Skills (C) + (Skills
(D)/)/2N
PGoals(E) + Goals(F) =
(j=1;
n Pj Goals (E) +
(Goals (F)/)/2N
Team
POS
Average
(Eq. 1) Table 2 provides a Team Alignment Scorecard which can be used by team leaders and evaluated episodically
with their team members. The scorecard outcome provides a single number which is a measure for the
probability of success (POS) of the team. The POS for the team’s efforts grows with their degree of alignment at
each social stability layer. Periodic appraisal of the team POS using the criteria of Table 2 can monitor team
progress. The POS for a team should increase as obstacles to collaboration are gradually removed as the team’s
collaboration efforts intensify. For team alignment, the sense that everyone succeeds or fails together remains
important. After completion of Steps 1 to 3, the team’s intrinsic POS can be calculated from (Eq. 1) as POS = PCulture * PSkills* PGoals. The fourth step (Figure 8, Step 4) consists of completion of the project in sustained
alignment of the team members on the project work. Even if the project goes well, it is still worthwhile to assess
POS at regular intervals (Figure 8, track 4A). If the project scope needs revision, team members should not
hesitate to revisit the three fundamental steps for team alignment (Figure 8, track 4B). Teams must be articulate
and define their objectives and scope of work. If objectives remain diffuse, the outcome will be diffuse Non-alignment in teams may also occur due to organizational problems (Kotter, 1978); this may occur when: 115 Engineering Management Research Vol. 1 No. 2; 2012 www.ccsenet.org/emr •
Formal directions for strategic change are not supported by social structures and core values in the
company. Consequently, teams will lack a sense of direction and their poor core value adoption
handicaps the development of team focus. 4. Guidelines for Team Leaders for Monitoring Team Alignment As a project moves forward, it is crucial to determine how objectives change with new findings and interim
project milestones. The EMV of the project can be episodically quantified based on the POS assessment using
Equation (3). Figure 9 shows the hypothetical outcomes for an asset investment development project. Cash flow
may be higher than expected if the EMV beats the NPV. Cash flow falls short of the scheduled cash flow when
EMV<NPV. Cash flows are as planned when EMV=NPV. 5. Examples of Team Alignment Challenges Some examples may serve to demonstrate the application of the previous theory as a team alignment
monitoring-tool. A range of survival games is widely used to examine team performance when strategic choices
have to be developed in teams. Commonly, the life of the team members - or a high reward - is at stake in such
games. Consider the Moon Survival Game (a classic game described on numerous websites), where the goal is
getting back to the designated rendez-vous point with the mothership after crash landing with an orbiter. Similarly, in the Alaskan Gold Mine Game, team members need a strategy plan to get back to the City from a
remote location with limited supplies to lay claim to the prospect before expiration of the mine’s exploration
license. Such crisis management simulation games impose a situation where teams are supposed to instantly align on
PGoals, whereas PCulture and PSkills need to be sorted out quickly in the team in order to reach the designated
location with limited supplies and within a limited time period. Playing such survival games in teams of
professionals provides a wealth of empirical evidence for potential obstacles to achieving perfect or 100%
alignment for PGoals, i.e. PGoals= 1. For example, a number of professionals may have difficulty identifying with
the authenticity of the game simulation and simply disengage. The impact on the team score for PGoals in a group 116 Engineering Management Research Vol. 1 No. 2; 2012 www.ccsenet.org/emr of ten members by one member disengaging with Pj Goals=0.1 is that PGoals dwindles to 0.91 for the team (see Eq. 5). Every next member that disengages lowers the PGoals for the team by nearly 10%. If half the team members
disengage from the purpose of the game, PGoals reduces to 0.55 or 55% of optimum team alignment for goals. of ten members by one member disengaging with Pj Goals=0.1 is that PGoals dwindles to 0.91 for the team (see Eq. 5). Every next member that disengages lowers the PGoals for the team by nearly 10%. If half the team members
disengage from the purpose of the game, PGoals reduces to 0.55 or 55% of optimum team alignment for goals. Assuming that the remaining team members succeed in reaching 100% alignment for both PCulture and PSkills, then
the probability of success POS = PCulture * PSkills* PGoals, gives at best a POS=0.55, according to Equation 1. 5. Examples of Team Alignment Challenges If lifes
are at stake, such as in the Moon Landing Game, a POS of 55% means 45% of the team members do not survive
because they do not reach the agreed pick-up location, which is the goal of the exercise. The team performance in survival games outlined above is exemplary for the performance of teams that need to
realize change in companies that are under severe competitive pressure or threat from the external environment. Change management teams commonly need to become convinced that there is a burning platform situation
(Conner, 1992). The fable by Kotter and Rathgeber (2006) replaces the burning platform by a melting iceberg,
and provides a powerful metaphor for companies under threat. The essence of change management is that
without rapid re-alignment by the company and a sense of urgency to adapt to changes in its external business
environment (Kotter, 1978), the company’s future may be at serious risk. In other words, achieving a high degree
of alignment for the agreed objectives (i.e. PGoals>0.8) is a prerequisite for successful change management. Research by PROSCI (2007) has revealed that the impact of effective change management is excellent if some
88% of the key players meet the agreed change objectives. Further, change management effects are good if some
75% meet the objectives, fair if only 45% meet the objectives and right-out poor if a mere 17% meet the change
management objectives. 6. Cumulative POS and Effect of Team Learning Team alignment on certain issues can accelerate if the team learns and the intrinsic POS grows over time with
each meeting. In reality, a team’s intrinsic POS will rarely remain constant over time as teams learn and the
degree of alignment commonly grows over time. A people-oriented approach, provides learning opportunities for
team members to optimize their POS. Explorationists in the oil industry know that drilling more wells with equal POS provides a cumulative POS
(Rhoads, 2005). In analogy, the number of meetings of a team, when effective, will increase the cumulative POS. The Cumulative probability of success expresses the cumulative changes in team alignment over time (e.g. using
equations from Rhoads, 2005): mulative POS= 1-[(1-POS1)*(1-POS2)*(1-POS3)*….(1-POSN)] (7a) Cumulative POS= 1-[(1-POS1)*(1-POS2)*(1-POS3)*….(1-POSN)] Cumulative POS= 1-[(1-POS1)*(1-POS2)*(1-POS3)*….(1-POSN)] (7a) (7a) where POSN is the probability of success at the N-th attempt. Alternatively, the team’s intrinsic POS does not
change over time but the Cumulative POS does grow, as follows (Johnston, 2003, Johnston and Johnston, 2002;
and Kjemperud, 2004): Cumulative POS= 1-(1-POS)N
(7b) Cumulative POS= 1-(1-POS)N
(7b)
Figure 10. Cumulative POS for alignment on a topic in team work increases as the number of meetings held
increases, even if the POS for the team remains constant. The blue line shows a team where the team’s intrinsic
POS grows from 0.1, to 0.15, to 0.2, to 0.3, to 0.4 over the period of five team meetings or attempts to tackle an
issue. Their Cumulative POS grows even faster from 0.1, to 0.28, to 0.49, to 0.75, to 0. 91 Cumulative POS= 1-(1-POS)N (7b) Cumulative POS= 1-(1-POS)N Figure 10. Cumulative POS for alignment on a topic in team work increases as the number of meetings held
increases, even if the POS for the team remains constant. The blue line shows a team where the team’s intrinsic
POS grows from 0.1, to 0.15, to 0.2, to 0.3, to 0.4 over the period of five team meetings or attempts to tackle an
issue. Their Cumulative POS grows even faster from 0.1, to 0.28, to 0.49, to 0.75, to 0. 91 Figure 10 graphs the Cumulative POS (acc. to Eq. 7b) versus the number of meetings for groups that are locked
in specific intrinsic POS scores for the team. 6. Cumulative POS and Effect of Team Learning For example, if the intrinsic POS for the team is 0.4, this means that
team members will align for only 40% of the decisions or actions to be taken for achieving an optimum project Figure 10 graphs the Cumulative POS (acc. to Eq. 7b) versus the number of meetings for groups that are locked
in specific intrinsic POS scores for the team. For example, if the intrinsic POS for the team is 0.4, this means that
team members will align for only 40% of the decisions or actions to be taken for achieving an optimum project 117 Engineering Management Research Vol. 1 No. 2; 2012 www.ccsenet.org/emr outcome. Teams that score a low intrinsic POS (as defined in Eq. 1) of say 0.1, and that do not learn to improve
alignment at any subsequent meeting, will have a chance to score an EMV that is only 40% of NPV, even after 5
meetings, based upon the cumulative POS for alignment (Figure 10, Eq. 7b). The NPV lost by suboptimum team
alignment has been quantified and graphed in Figures 6 and 7, respectively. In both calculations, the decline in
EMV and loss of NPV due to suboptimum team work justifies active investment to improve team alignment in
order to mitigate the NPV loss. outcome. Teams that score a low intrinsic POS (as defined in Eq. 1) of say 0.1, and that do not learn to improve
alignment at any subsequent meeting, will have a chance to score an EMV that is only 40% of NPV, even after 5
meetings, based upon the cumulative POS for alignment (Figure 10, Eq. 7b). The NPV lost by suboptimum team
alignment has been quantified and graphed in Figures 6 and 7, respectively. In both calculations, the decline in
EMV and loss of NPV due to suboptimum team work justifies active investment to improve team alignment in
order to mitigate the NPV loss. Figure 10 includes the Cumulative POS path for a team that learns, so that there is an additional effect on the
Cumulative POS: higher intrinsic POS for the team at each meeting adds onto the Cumulative POS effect on
team alignment at each new meeting. The enhanced effect of the Cumulative POS for learning teams benefits the
growth of EMV* (see Eq. 3). 7. Conclusions and Recommendations Starting project teams off in a people-oriented approach, rather than goal-oriented increases their probability of
success for an optimum project outcome. Checklist 1 provides a managerial checklist to promote the success of
teams. If the quality of the social stability layers remains poor in the team only a percentage of the optimum
project outcome can be realized. The method outlined here quantified how the loss in Net Present Value can be
mitigated by a positive learning curve for any team (Figure 10). If team members manage to align in the three
fundamental layers of team stability, their project outcome is optimized, as opposed to realizing only a fraction
of the required project outcome. As a generic rule, any team meeting needs a clear-cut agenda. Completion of the actions and decisions will make
the further difference, and the quantification of Cornelis’ Feelings Logic holds a powerful promise for tracking
the progressive alignment of team members as they work to realize a common goal. The key to optimizing team
performance lies locked in monitoring and influencing the degree of alignment between team members. Project
teams should first work people-oriented on common core values and skills inventories (Checklist 1), before
moving on toward a goal-oriented approach. An agenda always helps to start from a common expectation, a
prerequisite for building alignment in teams. Avoid an outcome saying “Why was the meeting called?” Be
careful with overemphasizing communication via modern ICT tools, as emotional signals are poorly captured
and ICT may filter out important information from the message. Checklist 1: Managerial checklist for success of teams (1) Successful teams need a core of people that already hold shared values from the outset, and team leaders
should act mostly people-oriented and then become more task-oriented only when the foundation of social
stability layers has firmed up. (1) Successful teams need a core of people that already hold shared values from the outset, and team leaders
should act mostly people-oriented and then become more task-oriented only when the foundation of social
stability layers has firmed up. (2) In effective teams, the team members align on common culture values and build social bonding through
mutual trust and acceptance. (2) In effective teams, the team members align on common culture values and build social bonding through
mutual trust and acceptance. 6. Cumulative POS and Effect of Team Learning After 5 meetings, the Cumulative POS for alignment over an issue may have risen
to 40% even when the intrinsic POS for the team remains unchanged at 10%. In the case of effective team
learning, that is the team’s intrinsic POS as assessed in subsequent meetings begins to exceed the team’s initial
POS, than the Cumulative POS (Eq. 7b) will rapidly grow to over 80% in a reasonable limited number of
meetings (less than 5, see Figure 10). The cost of the team work continues to grow with each meeting. The X axis in the graph of Figure 10 can be
scaled with a cumulative team OPEX of 5 kEuro per meeting (using an integral OPEX with 250 Euro for salary
& overhead per man hour, and a standardized 2 hours team meeting of 10 members). Regression is also possible,
if the Cumulative POS declines, and results in shrinking EMV*. Clearly, it is economically attractive to invest
time, resources and effort in building improved alignment in project teams. 7. Conclusions and Recommendations (7) A team becomes happy and successful if the self-fulfilment layer can be realized by alignment of team
members, and sad and unsuccessful if this is not the case. Effective goal setting for the team becomes possible
only when the vision for the project’s objectives is shared by all team members. (8) The most fundamental emotional response is related to the self-actualization or self-fulfilment level or
stability layer; ‘self-actualization’ refers to people’s ability to formulate and realise their goals and ambitions
within the social entity of a team. The model outlined in this study is powerful and preliminary tested in survival game simulations. Given the
positive impact of effective team learning on the Cumulative POS (Figure 10), monitoring team performance
seems worthwhile. Table 2 can be used for monitoring and improving team alignment in any organization – and
the payback can be attractive. Further benchmarking requires a client study to monitor team performance over an
extended period. Applications are possible in all kinds of functional teams, including change management teams
that need to build coalitions to effectuate lasting change. Value-adding by team excellence or
better-than-expected project outcomes remain worthwhile alternatives to value-lost by underperformance in
teams – truly competitive companies can no longer afford average team outcomes. Optimizing team outcomes
are a means to enhance organizational competitiveness. This study is not in disagreement with other team work models such as those that consider a degree of
self-organization, preceded by divergent-thinking or even chaos according to complexity theory, highlighted in
two recent books (Hazy et al., 2007; Uhl-Bien & Marion, 2008). Starting out with a laissez-faire approach in
team work as opposed to Venturi style management (e.g., Weijermars, 2008), does not necessarily mean that
alignment of team members, and the associated convergence of ideas, is postponed. Loosening the managerial
reins of goal-oriented functions and business operations can be an effective way to give teams room for
self-organization into a focused sharing of knowledge and skills. Self-organization is known to be pre-eminent in
the fertile region where novelty is created (Comfort, 1994). Effective top-down vision-sharing is still helpful for
teams when this involves both the adoption of a shared vision and the delegation of leadership to teams to help
realize the organizational vision and associated goals. 7. Conclusions and Recommendations (3) Teams benefit from consciously inventorying the skills needed for the task at hand and comparing these
with the range of skills mastered and covered by team members. The emotional responses related to the skills
layer of Cornelis are 'confidence' for those team members who have the appropriate skills - and ‘frustration’ for
those that do not. (4) An open dialogue about the team’s strengths in skills and weaknesses in is itself an exercise that will help
to build a stronger social Skills Layer for the team. Further strengthening of the skills layer is possible by
dedicated training and by inviting experts for advice or by recruiting new team members. (5) The three social stability layers of Cornelis ‘stack’ on top of one another. In other words, fulfilling the
Skills Layer is contingent on fulfilling the Culture Layer. It is highly implausible for a member of a team to
realize his or her goals and ambitions (self-actualization) if not supported by a social culture, common
understanding of core values, and effective sharing of skills and knowledge. (5) The three social stability layers of Cornelis ‘stack’ on top of one another. In other words, fulfilling the
Skills Layer is contingent on fulfilling the Culture Layer. It is highly implausible for a member of a team to
realize his or her goals and ambitions (self-actualization) if not supported by a social culture, common
understanding of core values, and effective sharing of skills and knowledge. 118 Engineering Management Research www.ccsenet.org/emr Vol. 1 No. 2; 2012 (6) Once the team aligns on core values and skills needed, the team can move on to define goals and work the
task (Figure. 4). When a team has defined the goals aspired, working of the tasks by the team follows naturally. The team will work with maximum collaboration between members to tap their creativity, passion, spontaneity,
feedback and productivity will be high. (6) Once the team aligns on core values and skills needed, the team can move on to define goals and work the
task (Figure. 4). When a team has defined the goals aspired, working of the tasks by the team follows naturally. The team will work with maximum collaboration between members to tap their creativity, passion, spontaneity,
feedback and productivity will be high. 7. Conclusions and Recommendations But a goal-oriented approach where top-down control is
too strong - an inherent danger in Venturi type or overly goal-oriented management - may make teams passive
and dependent (Argyris, 1957). Keeping a proper balance between core business goals that remain proactively
fuelled by innovation and new ideas remains important - modern management commonly knows how to utilize
the benefits of Venturi style and Complexity Leadership philosophies. Ultimately, team alignment (i.e., building
a high POS) is required to realize the team’s project outcome; and this is not necessarily counter to creating new
knowledge if the team optimizes skills and knowledge sharing for innovative solutions. Acknowledgements Dr. George DeVries Klein and Dr. Alfred Kjemperud provided helpful comments on the manuscript. This study
was sponsored by Alboran Energy Strategy Consultants, which specializes in independent strategy analysis and
training to support energy executives and professionals in strategy, leadership and change management. References Amabile, T. M. (1996). Creativity in Context. Westview Press, Boulder, Colorado. Amabile, T. M. (1996). Creativity in Context. Westview Press, Boulder, Colorado. Amabile, T. M. (1998). How to kill creativity. Harvard Business Review. Argyris, C. (1957). Personality and Organization. Harper Collins, New York. Argyris, C. (1957). Personality and Organization. Harper Collins, New York. Axelrod, R. (1997). The Complexity of Cooperation. Princeton University Press, Princeton, New Jerse Axelrod, R. (1997). The Complexity of Cooperation. Princeton University Press, Princeton, New Jersey. Baker, S. (2008). The Numerati. Houghton Mifflin Books. New York. Baker, S. (2008). The Numerati. Houghton Mifflin Books. New York. Belbin, M. (1993). Team Roles at Work. Butterworth Heinemann. Belbin, M. (1993). Team Roles at Work. Butterworth Heinemann. Blake, R., & Mouton, J. (1964). The managerial grid: key orientations for achieving production through people. Houston, Texas, Gulf. Blake, R., & Mouton, J. (1964). The managerial grid: key orientations for achieving production through people. Houston, Texas, Gulf. Collins, J. (2001). Good to Great: Why Some Companies Make the Leap... and Others Don’t. Harper Collins
Publishers Inc., New York. Collins, J. (2001). Good to Great: Why Some Companies Make the Leap... and Others Don’t. Harper Collins
Publishers Inc., New York. 119 Engineering Management Research Vol. 1 No. 2; 2012 www.ccsenet.org/emr Comfort , L. K. (1994). Self-Organization in Complex Systems. Journal of Public Administration Research
Theory, 4, 393-410. Conner, D. (1992). Managing at the speed of change: How resilient managers succeed and prosper where others
fail. New York, Villard. Cornelis, A. (1988). Logica van het gevoel, filosofie van de stabiliteitslagen in de cultuur als nesteling der
emoties. Uitgave Stichting Essence Amsterdam/Brussel/Middelburg, 9e druk, 2000. Drexler, A., Sibbet, D., & Forrester, R. (1988). The team performance model. In W. B. Reddy, & K. Jamieson
(Eds.), Teambuilding: blueprints for productivity and satisfaction (pp. 45-61). Alexandria, VA: NTL
Institute for the Applied Behavioral Sciences; San Diego, CA: University Associates. Druskat, V. U., & Wolff, S. B. (2001). Building the Emotional Intelligence of Groups. Harvard Business
Review. Fiedler, F. E. (1967). A Theory of Leadership Effectiveness. New York: McGraw-Hill E. (1967). A Theory of Leadership Effectiveness. New York: McGraw-Hill Fischer, B., & Boynton, A. (2005). Virtusoso Temas. Harvard Business Review. , & Boynton, A. (2005). Virtusoso Temas. Harvard Business Review. Goleman, D. (1996). Emotional Intelligence-Why it can Matter More Than IQ. Bloomsbury. Goleman, D. (1997). Emotional Intelligence. Bantam Books, New York. Goleman, D. (1997). Emotional Intelligence. References Bantam Books, New York. Goleman, D. (2000). Working with Emotional Intelligence. Bantam Books, New York. 2000). Working with Emotional Intelligence. Bantam Books, New York. Gratton, L., & Erickson, T. J. (2007). Eight ways to build collaborative teams. Harvard Business Review. Hazy, K., Goldstein, J., & Lichtenstein, B. (2007). Complex Systems Leadership Theory. ISCE Publishers,
Mansfield, MA. Herrmann, N. (1996). The whole brain business book. New York: Mc Graw Hill . Hoegl, M., & Parboteeah, K. P. (2007). Creativity in innovative projects: how teamwork matters. Journal of
Engineering and Technology Management, 24, 148-166. Hofstede, G., & Hofstede, G. J. (2004). Cultures and Organizations: Software of the Mind. McGraw-Hill, New
York. Johnston, D., & Johnston, D. (2002). Economic Modeling and Risk Analysis Handbook. CE
University of Dundee. ohnston, D. (2003). International Petroleum Economics, Risk, and Contract Analysis. Penwell Books Kirkpatrick, S. A., & Locke, E. A. (1991). Leadership: Do traits matter?. Academy of Management Executive, 5,
48-60. Kjemperud, S. A. (2004). Risk analysis and petroleum economics, The Bridge Group, 4th Workshop of
Cambodia PPM Case Study, Pnom Penn. Retrieved from http://www.ccop.or.th/ppm/ Klein, G. D. (1999). Navigating through an International Enhanced Oil Recovery Project: A Cross-Cultural
Odyssey. The Professional Geologist, 36(7), 4-7. Kotter, J. P. (1978). Organizational Dynamics: Diagnosis and Intervention. Addison-Wesley. New York. Kotter, J. P., & Rathgeber, H. (2006). Our Iceberg Is Melting: Changing and Succeeding Under Any Conditions. Macmillan, New York. Lord, D., Etheridge, M. A., Willson, M., Hall, G., & Uttley, P. J. (2001). Measuring exploration success: an
alternative to the discovery-cost-per-ounce method of quantifying exploration success. SEG Newsletter, 45,
10-16. Lumsdaine, E. M., Lumsdaine, & Shelnutt, J. W. (1999). Creative Problem Solving and Engineering Design. McGraw-Hill, New York, NY. Oliver, P., Kreuzer, O. P., Etheridge. M. A., Guj, P., McMahon, M. E., & Holden, D. J. (2008). Linking Mineral
Deposit Models to Quantitative Risk Analysis and Decision-Making in Exploration. Economic Geology,
103(4), 829-850. http://dx.doi.org/10.2113/gsecongeo.103.4.829 PROSCI. (2007). PROSCI best practices in change management 2007 report. Prosci, Colorado. PROSCI. (2007). PROSCI best practices in change management 2007 report. Prosci, Colorado. Rhoads, G. (2005). Managing the Probability of Making at Least One Discovery. Gulf of Mexico Shelf Business
Unit Chevron, Conference Presentation June 2005. 120 Vol. 1 No. 2; 2012 Engineering Management Research www.ccsenet.org/emr Ridley, M. (1998). The Origins of Virtue: Human Instincts and the Evolution of Cooperation. New York:
Viking. Thompson, L. (2003). Improving the creativity of organizational work groups. References Academy of Management
Executive, 17(1), 96-109. http://dx.doi.org/10.5465/AME.2003.9474814 Trompenaars, F., & Hampden-Turner, C. (1997). Riding the Waves of Culture. Nicholas Brealey Publishing,
Londen. Tuckman, B. W. (1965). Developmental sequence in small groups. Psychological Bulletin, 63, 384-399. http://dx.doi.org/10.1037/h0022100 Uhl-Bien, M., & Marion, R. (2008). Complexity & Leadership (Vol. I). Information Publishing Associ Welsh, M. B., Begg, S. H., & Bratvold, R. B. (2008). Modeling the Economic Effects of Cognitive Biases in Oil
and Gas Decisions. Journal of Petroleum Technology, 659-661. Weijermars, R. (2008). Balancing Venturi and Laissez-Faire Management Styles: Insights from Fluid
Mechanical Analogs. Journal of Systemics, Cybernetics and Informatics, 5(6), 1-10. Weijermars, R. (2011). Building Corporate IQ: Moving the Energy Business from Smart to Genius: Executive
Guide to preventing Costly Crises. Springer, London. Weijermars, R., & De Jong, V. (2008). Closing Communication Gaps can improve the Success of Oil & Gas
Ventures. Oil and Gas Business Journal, 1-24. Weijermars, R., De Jong, V., & Van der Kooi, K. (2008). Cultural challenges in oil and gas industry
management. World Oil, 229(4), 223-228. Willigers, B. J. A., & Majou, F. (2010). Creating efficient portfolios that match competing corporate strategies. SPE Paper 129259. 121
|
https://openalex.org/W2564529691
|
http://www.zora.uzh.ch/id/eprint/131138/1/ZORA131138.pdf
|
English
| null |
Social Phobia Is Associated with Delayed Onset of Chickenpox, Measles, and Mumps Infections
|
Frontiers in psychiatry
| 2,016
|
cc-by
| 5,777
|
Zurich Open Repository and
Archive Zurich Open Repository and
Archive
University of Zurich
University Library
Strickhofstrasse 39
CH-8057 Zurich
www.zora.uzh.ch University of Zurich
University Library
Strickhofstrasse 39
CH-8057 Zurich
www.zora.uzh.ch Social phobia is associated with delayed onset of chickenpox, measles, and mumps
infections Ajdacic-Gross, Vladeta ; Aleksandrowicz, Aleksandra ; Rodgers, Stephanie ; Müller, Mario ; Kawohl, Wolfram ;
Rössler, Wulf ; Castelao, Enrique ; Vandeleur, Caroline ; von Känel, Roland ; Mutsch, Margot ; Lieb, Roselind ;
Preisig, Martin DOI: https://doi.org/10.3389/fpsyt.2016.00203 Posted at the Zurich Open Repository and Archive, University of Zurich
ZORA URL: https://doi.org/10.5167/uzh-131138
Journal Article
Published Version
The following work is licensed under a Creative Commons: Attribution 4. Posted at the Zurich Open Repository and Archive, University of Zurich
ZORA URL: https://doi.org/10.5167/uzh-131138
Journal Article
Published Version ollowing work is licensed under a Creative Commons: Attribution 4.0 International (CC BY 4.0) License. Originally published at:
Ajdacic-Gross, Vladeta; Aleksandrowicz, Aleksandra; Rodgers, Stephanie; Müller, Mario; Kawohl, Wolfram;
Rössler, Wulf; Castelao, Enrique; Vandeleur, Caroline; von Känel, Roland; Mutsch, Margot; Lieb, Roselind;
Preisig, Martin (2016). Social phobia is associated with delayed onset of chickenpox, measles, and mumps in-
fections. Frontiers in Psychiatry, 7:203. DOI: https://doi.org/10.3389/fpsyt.2016.00203 Original research
published: 27 December 2016
doi: 10.3389/fpsyt.2016.00203 Vladeta Ajdacic-Gross1,2*, Aleksandra Aleksandrowicz1, Stephanie Rodgers1,
Mario Müller1, Wolfram Kawohl1, Wulf Rössler1,3,4, Enrique Castelao5, Caroline Vandeleur5,
Roland von Känel6, Margot Mutsch2, Roselind Lieb7 and Martin Preisig5 Vladeta Ajdacic-Gross1,2*, Aleksandra Aleksandrowicz1, Stephanie Rodgers1,
Mario Müller1, Wolfram Kawohl1, Wulf Rössler1,3,4, Enrique Castelao5, Caroline Vandeleur5,
Roland von Känel6, Margot Mutsch2, Roselind Lieb7 and Martin Preisig5 1 Department of Psychiatry, Psychotherapy and Psychosomatics, Psychiatric Hospital, University of Zürich, Zürich,
Switzerland, 2 Epidemiology, Biostatistics and Prevention Institute, University of Zürich, Zürich, Switzerland, 3 Collegium
Helveticum, Swiss Federal Institute of Technology, University of Zürich, Zürich, Switzerland, 4 Institute of Psychiatry,
Laboratory of Neuroscience (LIM27), University of Sao Paulo, Sao Paulo, Brazil, 5 Department of Psychiatry, University
Hospital of Lausanne, Lausanne, Switzerland, 6 Department of Neurology, Bern University Hospital, Clinic Barmelweid,
Barmelweid, Switzerland, 7 Department of Psychology, Division of Clinical Psychology and Epidemiology, University of Basel,
Basel, Switzerland Objective: Evidence showing that infectious diseases in childhood play an important
role in the etiopathogenesis of neurodevelopmental and other mental disorders is grow-
ing. The aim of this study was to explore the timing of common childhood diseases in
early-onset anxiety disorders. Edited by:
Karl Bechter,
University of Ulm, Germany
Reviewed by:
Carlo Lai,
Sapienza University of Rome, Italy
Michele Fornaro,
Columbia University, USA
*Correspondence:
Vladeta Ajdacic-Gross
vajdacic@dgsp.uzh.ch Edited by:
Karl Bechter,
University of Ulm, Germany Materials and methods: We analyzed data from PsyCoLaus, a large Swiss Population
Cohort Study (N = 3720). In this study, we regressed overanxious disorder, separation
anxiety disorder, social phobia, and specific phobias on the age of onset of several
childhood diseases, always adjusting for the other anxiety disorders listed above and
for sex. Reviewed by:
Carlo Lai,
Sapienza University of Rome, Italy
Michele Fornaro,
Columbia University, USA *Correspondence:
Vladeta Ajdacic-Gross
vajdacic@dgsp.uzh.ch results: The timing of viral childhood diseases (chickenpox, measles, and mumps)
was consistently delayed in social phobia, notably both in men and women. We found
no evidence for a reversed sequence of onset of phobia symptoms before that of the
infections included. Specialty section:
This article was submitted to
Mood and Anxiety Disorders,
a section of the journal
Frontiers in Psychiatry
Received: 12 September 2016
Accepted: 08 December 2016
Published: 27 December 2016 Specialty section:
This article was submitted to
Mood and Anxiety Disorders,
a section of the journal
Frontiers in Psychiatry conclusion: Social phobia was the only early anxiety disorder to show an association
with a delayed onset of common viral childhood diseases. Received: 12 September 2016
Accepted: 08 December 2016
Published: 27 December 2016 Keywords: social phobia, anxiety disorders, childhood diseases, infectious diseases, epidemiology Accepted: 08 December 2016 childhood Diseases/Disorders
and age of Onset The early-onset anxiety disorders are characterized by an early
age of onset of first symptoms, i.e., mostly before adolescence. While separation anxiety disorder is defined by excessive anxiety
of being separated from parents or significant others, overanxious
disorder was introduced as a childhood form of general anxiety
disorder and in DSM-IV subsumed within this disorder. Specific
phobias are a heterogeneous group of disorders (24) related to
specific animals, objects, or situations. In social phobia, these are
social situations. Typically, girls and women are more frequently
affected by early-onset anxiety disorders. The sex ratios can reach
a factor of 3 (specific phobias). There is a lack of research systematically examining how the
timing of infections and other inflammatory insults (17) impacts
the risk for mental disorders. This study focused on the timing of
childhood diseases with regard to early-onset anxiety disorders
(separation anxiety disorder, overanxious disorder, specific
phobias, and social phobia). The aim of the study was to examine
whether deviations in age of onset of childhood infectious dis-
eases are associated with the early-onset anxiety disorders. We
used a two-step approach. First, we examined whether the age
of onset of infections differs in cases with and without a specific
disorder. The analyses were adjusted for common confounding
factors and other early-onset anxiety disorders. In the case of
delayed onset, we examined whether the sequence of a childhood
infectious disease and the first symptoms of an anxiety disorder
might have induced the result. The following childhood infections were included in the
analyses: pertussis, chickenpox, measles, mumps, rubella, and
scarlet fever. The information on infectious diseases and other
related conditions was derived using an extended version of
the medical history part of the DIGS and was based on self-
reporting. In the interview, participants were asked questions
about ever having been diagnosed with various infectious
diseases, diseases of the nervous system, cardiovascular, respira-
tory, gastrointestinal, metabolic and dermatological conditions,
as well as allergies and hormonal problems. For each disease
group, a screening question was asked, followed by more specific
questions in the case of an affirmative response. In the section
on childhood diseases, chickenpox, measles, and mumps were
explicitely asked, whereas information regarding pertussis,
rubella, and scarlet fever was extracted from the free text field of
this section. These questions routinely included the age of onset
of a disease or the first occurrence of symptoms of a mental
disorder. inTrODUcTiOn Ajdacic-Gross V, Aleksandrowicz A,
Rodgers S, Müller M, Kawohl W,
Rössler W, Castelao E, Vandeleur C,
von Känel R, Mutsch M, Lieb R and
Preisig M (2016) Social Phobia Is
Associated with Delayed Onset
of Chickenpox, Measles,
and Mumps Infections. Front. Psychiatry 7:203. doi: 10.3389/fpsyt.2016.00203 Evidence showing that infectious diseases in childhood can play a crucial role in the etiopatho-
genesis of mental disorders is growing (1). This relation has been a long-standing issue in
psychiatric epidemiology, but research activity has recently intensified. Current topics include
prenatal infections with Toxoplasma gondii, rubella, and other infectious diseases (2, 3) as risk
factors for neurodevelopmental disorders and schizophrenia. Furthermore, a series of postnatal
infectious agents was reported in association with schizophrenia/psychosis (4–6). The pediatric
autoimmune neuropsychiatric disorders associated with streptococcal infections (PANDAS) December 2016 | Volume 7 | Article 203 Frontiers in Psychiatry | www.frontiersin.org 1 Social Phobia and Childhood Diseases Ajdacic-Gross et al. were randomly selected from the population of the city of
Lausanne (Switzerland). The assessment of the subjects took place
between 2003 and 2006 in an outpatient clinic (19, 20). It included
an interview with a semi-structured questionnaire, as well as the
collection of clinical data and blood samples. One year after their
initial assessment, CoLaus participants aged between 35 and 66
were asked to participate in the PsyCoLaus study. Subsequently,
a total of 3,720 individuals (67%) agreed to participate (18). One
major goal of PsyCoLaus was to collect data on the prevalence of
psychiatric syndromes/disorders. model also postulates a crucial role for autoantibodies
against basal ganglia tissue appearing after infections with
group A streptococci in the development of attention deficit
hyperactivity disorder, Gilles de la Tourette syndrome, and
obsessive–compulsive disorder (OCD) (7, 8). Finally, recent
analyses have shed new light on the role of infectious diseases
in anxiety disorders (9). y
On a formal level, there are four models of how infectious
diseases could affect the risk for mental disorders. The models
can be illustrated using the associations between Epstein–Barr
virus (EBV) and multiple sclerosis (MS) or psychotic disorders. The first model relates to the mere occurrence of an infection. For example, an EBV infection is a necessary precondition
for later MS onset: while the seroprevalence for EBV is high,
i.e., it covers 90% of the adult population or more (10), almost
no one from the negative seroprevalence group develops MS. inTrODUcTiOn The second model includes the severity of an infection, for
example, due to a dysfunctional immune response and/or due
to a combination with genetic or other types of vulnerabilities. The third model considers the timing of childhood infections. Deviations from typical timing, in particular delayed childhood
infectious diseases such as chickenpox, measles, mumps, or
rubella, are well known to lead to a more severe disease course
with more complications (11–14). This also applies to delayed
EBV infections, which increase the risk for mononucleosis and
probably also for MS (15). Finally, the fourth model focuses on
the sequence of childhood infections. This is a possible reason
for why infectious disease occurring earlier in life than typi-
cally expected might have harmful effects. For example, EBV
infections in the first years of life have been linked to psychotic
experiences (16). Both the timing and sequence of childhood
infections can be assumed to be subjected to evolutionary
optimization. A French version of the semi-structured diagnostic interview
for genetic studies (DIGS) (21, 22) was used in the PsyCoLaus
study to assess a broad spectrum of DSM-IV Axis I criteria. Moreover, the DIGS allowed for gathering additional informa-
tion about the course and chronology of comorbid features (18). However, the brief phobia section of the DIGS was replaced by the
corresponding sections from the Schedule for Affective Disorders
and Schizophrenia-Lifetime Version (23). The study was approved by the Ethics Committee of the
University of Lausanne. All participants gave their written
informed consent at study enrollment in accordance with the
Declaration of Helsinki. childhood Diseases/Disorders
and age of Onset Information such as “so early that I cannot remember”
or “since I can remember” or age below 2 was replaced by a
random value between 2 and 5. resUlTs • sex: in the analysis of overall data; The mean age of onset of first symptoms was 7.0 (SD 3.5) years
in overanxious disorder, 5.2 (SD 2.2) years in separation anxiety
disorder, 12.1 (SD 9.8) years in specific phobias and 10.9 (SD 8.2)
years in social phobia. Table 1 displays the average onset age of
common viral and bacterial childhood diseases in early anxiety
disorders. • other early-onset anxiety disorders (separation anxiety disor-
der, overanxious disorder, specific phobias, and social phobia)
apart from the dependent variable. Additional adjustments introduced in a one-by-one manner
included the education level (three categories), the age of the
participants (because of the possibility of recalling a higher age
for any events among older age groups), and childhood adversi-
ties (fear of parental punishment, parental quarrels, growing up
in children’s home). Table 2 shows the results for social phobia regressed on the age
of onset of chickenpox, measles, and mumps for the whole sample
and the results for separation anxiety disorders (only males)
regressed on the age of onset of mumps. The analyses in other
anxiety disorders did not yield significant estimates (results not
shown). Neither the adjustment for covariates (other early anxiety
disorders and sex) and potential confounders (education level,
age of subjects, and childhood adversities) nor the replication of
the analyses after excluding subjects born in 1963 or later yielded
any noteworthy differences. However, replication by sex-specific
analyses showed that in social phobia the onset of chickenpox,
measles, and mumps was consistently delayed in both sexes. The
odds ratios based on square root smoothed values for age were
1.8 (1.2–2.7), 2.1 (1.3–3.2), and 2.0 (1.3–3.0) in males, and 1.4
(1.0–1.9), 1.7 (1.2–2.2), and 1.4 (1.0–2.0) in females. In the case of statistically significant and consistent results, the
analysis design was supplemented by an additional comparison
of the observed and expected cases relating to the sequence of
a childhood disease and the first symptoms of an early-onset
anxiety disorder in order to ascertain that the sequence remained
unchanged. Since the relevant information on the age of onset
was available only for a subset of subjects (those reporting the
respective infectious disease and simultaneously indicating the
age of onset of symptoms of an early-onset anxiety disorder), a
consistent framework for inference statistics was missing and had
to be replaced by a bootstrapping procedure (25). The Psycolaus study The analysis was carried out within the framework of PsyCoLaus,
a large epidemiological study conducted in Switzerland. The
PsyCoLaus study (18) is the psychiatric part of the population-
based CoLaus study (19). The participants in the CoLaus study December 2016 | Volume 7 | Article 203 Frontiers in Psychiatry | www.frontiersin.org 2 Ajdacic-Gross et al. Social Phobia and Childhood Diseases statistical analysis Furthermore, in the case of statistically significant and consist-
ent results, we additionally examined the associations between
the early-onset anxiety disorders and the frequencies of reported
chickenpox, mumps, and measles infections. The analyses followed the conventional analysis design of univari-
ate, bivariate, and multivariate analyses. In bi- and multivariate
regression models, the age of onset of infectious diseases was
implemented as a predictor of any of the examined early-onset
anxiety disorders. In regression analyses, the onset age of infec-
tious diseases was limited to 16 years in order to exclude outliers
(14.8% in rubella, and 2.3–6.6% in other infectious diseases). In
addition, the age data were smoothed by square root transforma-
tion to approach a normal distribution. Each regression analysis
relied on a subsample of subjects who reported the infectious
disease in question (and not on the whole sample). The analyses
were routinely adjusted for: The analyses were routinely carried out with the overall data
as well as for males and females separately. Moreover, they were
repeated for measles after excluding those born in or after 1963
(decreasing endemicity of measles in part due to due to a national
vaccination policy). The basic analyses and programming were
carried out using SPSS Statistics (version 21). The bootstrapping
procedure was programed in a SAS (version 9.3) macro. December 2016 | Volume 7 | Article 203 resUlTs TaBle 2 | regression analysis models with social phobia regressed on age of onset in chickenpox (model 1), measles (model 2), and mumps (model
3); separation anxiety disorder on age of onset in mumps (only males; model 4); age smoothed by square root transformation; odds ratios and 95%
confidence interval adjusted in each model for other early anxiety disorders and sex. Model 1 social phobia on
chickenpox age
Model 2 social phobia on
measles age
Model 3 social phobia on
mumps age
Model 4 (males) separation anxiety
disorder on mumps age
Age at onset
1.57 (1.23–1.99)
1.76 (1.36–2.28)
1.55 (1.19–2.04)
2.89 (1.55–5.39)
Sex
1.44 (1.11–1.86)
1.39 (1.07–1.82)
1.54 (1.16–2.06)
*
Separation anxiety disorder
1.95 (1.28–2.97)
1.47 (0.93–2.33)
1.46 (0.88–2.42)
*
Overanxious disorder
2.42 (1.69–3.47)
2.47 (1.70–3.59)
2.69 (1.81–3.99)
*
Specific phobia
1.89 (1.31–2.71)
1.79 (1.22–2.63)
2.20 (1.46–3.31)
*
*not applied (see text and Table 3). TaBle 2 | regression analysis models with social phobia regressed on age of onset in chickenpox (model 1), measles (model 2), and mumps (model
3); separation anxiety disorder on age of onset in mumps (only males; model 4); age smoothed by square root transformation; odds ratios and 95%
confidence interval adjusted in each model for other early anxiety disorders and sex. TaBle 2 | regression analysis models with social phobia regressed on age of onset in chickenpox (model 1), measles (model 2), and mumps (model
3); separation anxiety disorder on age of onset in mumps (only males; model 4); age smoothed by square root transformation; odds ratios and 95%
confidence interval adjusted in each model for other early anxiety disorders and sex. Model 1 social phobia on
chickenpox age
Model 2 social phobia on
measles age
Model 3 social phobia on
mumps age
Model 4 (males) separation anxiety
disorder on mumps age
Age at onset
1.57 (1.23–1.99)
1.76 (1.36–2.28)
1.55 (1.19–2.04)
2.89 (1.55–5.39)
Sex
1.44 (1.11–1.86)
1.39 (1.07–1.82)
1.54 (1.16–2.06)
*
Separation anxiety disorder
1.95 (1.28–2.97)
1.47 (0.93–2.33)
1.46 (0.88–2.42)
*
Overanxious disorder
2.42 (1.69–3.47)
2.47 (1.70–3.59)
2.69 (1.81–3.99)
*
Specific phobia
1.89 (1.31–2.71)
1.79 (1.22–2.63)
2.20 (1.46–3.31)
*
*not applied (see text and Table 3). TaBle 3 | Observed and expected cases with onset of selected childhood disease before or in the same year as the onset of first symptoms of social
phobia (overall sample) and separation anxiety disorder (only males). social phobia vs. chickenpox
social phobia vs. measles
social phobia vs. resUlTs mumps
separation anxiety disorder vs. mumps (males)
Subsample N total
325
288
247
43
N (%) observed
198 (60.9)
185 (64.2)
135 (54.7)
8 (18.6)
N (%) expected
208.4 (64.1)
189 (65.7)
140.4 (56.9)
15.3 (35.7)
χ2
1.45 (n.s.)
0.27 (n.s.)
0.48 (n.s.)
5.41*
*p < 0.05. cases with onset of selected childhood disease before or in the same year as the onset of first symptoms of social
tion anxiety disorder (only males). TaBle 3 | Observed and expected cases with onset of selected childhood disease before or in the same year as the o
phobia (overall sample) and separation anxiety disorder (only males). and expected cases with onset of selected childhood disease before or in the same year as the onset of first sympto
ple) and separation anxiety disorder (only males). Additional analyses examining the associations between social
phobia and reported chickenpox, mumps, and measles infections
consistently yielded odds ratios above one, but reached significant
levels only regarding mumps (OR = 1.26, CI 1.01–1.57). With this interpretation, some viral infections are directly associ-
ated with the risk for social phobia and mediated by a more severe
course and symptoms of the infectious disease. The third interpretation is closely linked to the second one
mentioned earlier. It cannot be excluded that an unknown com-
mon factor simultaneously increases the risk for more severe
forms of childhood diseases and for social phobia. However, no
such factor is currently apparent. Frontiers in Psychiatry | www.frontiersin.org DiscUssiOn This is compatible with the increased frequencies of
at least mumps infections reported by subjects with social phobia. Thus, we offer four possible interpretations of our results. First,
the timing of the viral infections could interfere with the chang-
ing activity of the immune system, as hypothesized regarding
EBV infection in/after adolescence and the EBV–MS link (15, 26). However, social phobia symptoms typically have an early onset
before adolescence (27). Since the average delay of chickenpox,
measles and mumps infections found in persons with social pho-
bia is rather small (below 1 year), it cannot fill the gap between
childhood and adolescence. Therefore, there is no support for a
crucial role for adolescence and its concomitants in this disorder. Thus, the most plausible interpretation for the moment is that
viral infections in childhood not only precede but also contribute
to the risk for social phobia. The reporting of viral infections in
a survey such as PsyCoLaus depends on the perception of mani-
fest symptoms by the subjects. In turn, manifest symptoms are
related to a more severe course of these infections and thus also
to a stronger involvement of the immune system. Therefore, we
hypothesize that the increased risk for social phobia is mediated
by a relatively stronger involvement of the immune system. This
is in line with models linking infectious diseases with neurodevel-
opmental disorders—for example, the PANDAS model (8)—or
with postinfectious fatigue and depression (28–30). However,
in contrast to the PANDAS model, viral—and not, or not only,
bacterial—pathogens are involved in social phobia. Moreover, the
small delay suggests that apart from the severity of viral child-
hood diseases a specific age stage, i.e., a specific stage of brain
development, should also be considered in social phobia. The second interpretation is more trivial, suggesting that the
reporting of chickenpox, measles, and mumps basically relies on
the visibility of symptoms such as exanthema. It is noteworthy
that these childhood diseases had reached a high seroprevalence
of above 90%, before the vaccination campaigns began, thus
inducing a ceiling effect in the analyses (12, 14). As mentioned
earlier, a higher age at infection is a risk factor for more severe
symptoms. This is compatible with the increased frequencies of
at least mumps infections reported by subjects with social phobia. DiscUssiOn This study was the first to examine the timing of childhood
diseases in early-onset anxiety disorders (separation anxiety dis-
order, overanxious disorder, specific phobias, and social phobia). We found that social phobia was associated with a delayed age
of viral infectious diseases, in particular with respect to chicken-
pox, measles, and mumps. Notably, these results were separately
replicable in males and females and could not be explained by a
reciprocal sequence of reported conditions. This study was the first to examine the timing of childhood
diseases in early-onset anxiety disorders (separation anxiety dis-
order, overanxious disorder, specific phobias, and social phobia). Last but not least, the association between social phobia and
delayed age of viral infectious diseases could emerge as an impli-
cation of coping with early social phobia symptoms, for example,
by avoiding social contact in free time and out-of-school activi-
ties. A similar rationale would also apply to separation anxiety
and overanxious disorder. However, in these instances, it was not
supported by the data. We found that social phobia was associated with a delayed age
of viral infectious diseases, in particular with respect to chicken-
pox, measles, and mumps. Notably, these results were separately
replicable in males and females and could not be explained by a
reciprocal sequence of reported conditions. Thus, we offer four possible interpretations of our results. First,
the timing of the viral infections could interfere with the chang-
ing activity of the immune system, as hypothesized regarding
EBV infection in/after adolescence and the EBV–MS link (15, 26). However, social phobia symptoms typically have an early onset
before adolescence (27). Since the average delay of chickenpox,
measles and mumps infections found in persons with social pho-
bia is rather small (below 1 year), it cannot fill the gap between
childhood and adolescence. Therefore, there is no support for a
crucial role for adolescence and its concomitants in this disorder. The second interpretation is more trivial, suggesting that the
reporting of chickenpox, measles, and mumps basically relies on
the visibility of symptoms such as exanthema. It is noteworthy
that these childhood diseases had reached a high seroprevalence
of above 90%, before the vaccination campaigns began, thus
inducing a ceiling effect in the analyses (12, 14). As mentioned
earlier, a higher age at infection is a risk factor for more severe
symptoms. resUlTs In this proce-
dure, we created for each run a subsample of 1:1 matched controls
from the overall sample positively reporting the infectious disease
in question. For each run, the hypothetical number of controls
was determined in whom an infectious disease occurred before or
in the same year as in the corresponding case (relating to the year
of onset of first symptoms of an anxiety disorder). We performed
1,000 runs in order to fix the hypothetical sequence and treated it,
for the sake of simplicity, as the underlying population sequence. Therefore, the current sequence of infectious diseases and the first
symptoms of an early-onset anxiety disorder were tested as the
comparison between observed and theoretical frequencies. The results from analyses addressing the sequence of each
childhood disease and the onset of first symptoms of social
phobia (or separation anxiety disorder) are displayed in Table 3. In social phobia, the delay in the onset of chickenpox, measles,
and mumps did not induce any noteworthy shift between
observed and expected cases. However, in separation anxiety
disorders, the delay of the onset of mumps was accompanied by
an increase of cases with a mumps infection after encountering
preliminary separation anxiety symptoms thus indicating that
an altered sequence of onsets was involved and induced artificial
results. TaBle 1 | Onset age of childhood infectious diseases overall and per anxiety disorder; outliers above 16 years excluded; mean and 95% confidence
interval. chickenpox
N = 2566
Measles
N = 2390
Mumps
N = 1936
rubella
N = 201
Pertussis
N = 256
scarlet fever
N = 146
Overall
6.05 (5.95–6.14)
6.06 (5.97–6.16)
7.04 (6.93–7.16)
7.55 (7.01–8.01)
6.39 (6.06–6.72)
7.46 (6.97–7.95)
Separation anxiety disorder
5.76 (5.41–6.12)
5.93 (5.53–6.33)
7.46 (6.90–8.01)
7.11 (5.42–8.80)
7.14 (4.86–9.43)
7.85 (6.07–9.62)
Overanxious disorder
5.98 (5.64–6.33)
6.30 (5.98–6.63)
7.38 (6.93–7.82)
7.89 (6.01–9.78)
6.18 (4.87–7.49)
7.50 (4.47–10.53)
Specific phobia
5.99 (5.76–6.22)
6.16 (5.93–6.39)
7.15 (6.86–7.44)
7.40 (6.36–8.43)
6.39 (5.68–7.10)
6.64 (5.52–7.76)
Social phobia
6.50 (6.22–6.77)
6.67 (6.40–6.93)
7.60 (7.25–7.95)
7.54 (6.07–9.00)
6.83 (5.70–7.96)
8.15 (6.84–9.46)
3
Frontiers in Psychiatry | www.frontiersin.org
December 2016 | Volume 7 | Article 203 tious diseases overall and per anxiety disorder; outliers above 16 years excluded; mean and 95% confidence TaBle 1 | Onset age of childhood infectious diseases overall and per anxiety disorder; outliers above 16 years exclud
interval. 3 Social Phobia and Childhood Diseases Ajdacic-Gross et al. reFerences quality of life: a community study. Front Public Health (2014) 2:80. doi:10.3389/
fpubh.2014.00080 quality of life: a community study. Front Public Health (2014) 2:80. doi:10.3389/
fpubh.2014.00080 1. Hornig M. The role of microbes and autoimmunity in the pathogenesis of neu-
ropsychiatric illness. Curr Opin Rheumatol (2013) 25:488–795. doi:10.1097/
BOR.0b013e32836208de 10. Ruprecht K. [Multiple sclerosis and Epstein-Barr virus: new developments and
perspectives]. Nervenarzt (2008) 79:399–407. doi:10.1007/s00115-007-2335-8 11. Wittek M, Doerr HW, Allwinn R. [Varicella and herpes zoster. Part 1: virology,
epidemiology, clinical picture, laboratory diagnostics]. Med Klin (Munich)
(2010) 105:334–8. doi:10.1007/s00063-010-1061-3 2. Nickerson JP, Richner B, Santy K, Lequin MH, Poretti A, Filippi CG,
et al. Neuroimaging of pediatric intracranial infection – part 2: TORCH,
viral, fungal, and parasitic infections. J Neuroimaging (2012) 22:e52–63. doi:10.1111/j.1552-6569.2011.00700.x (2010) 105:334–8. doi:10.1007/s00063-010-1061-3 12. Perry RT, Halsey NA. The clinical significance of measles: a review. J Infect Dis
(2004) 189(Suppl 1):S4–16. doi:10.1086/377712 13. Stock I. [Rubella (German measles) – still a major infectious disease]. Med
Monatsschr Pharm (2012) 35:14–22. 3. Adams Waldorf KM, McAdams RM. Influence of infection during pregnancy
on fetal development. Reproduction (2013) 146:R151–62. doi:10.1530/
REP-13-0232 14. Hviid A, Rubin S, Muhlemann K. Mumps. Lancet (2008) 371:932–44. doi:10.1016/S0140-6736(08)60419-5 4. Sperner-Unterweger B. [Biological hypotheses of schizophrenia: possible
influences of immunology and endocrinology]. Fortschr Neurol Psychiatr
(2005) 73(Suppl 1):S38–43. doi:10.1055/s-2005-915544 15. Thacker EL, Mirzaei F, Ascherio A. Infectious mononucleosis and risk for mul-
tiple sclerosis: a meta-analysis. Ann Neurol (2006) 59:499–503. doi:10.1002/
ana.20820 5. Yolken RH, Torrey EF. Are some cases of psychosis caused by microbial
agents? A review of the evidence. Mol Psychiatry (2008) 13:470–9. doi:10.1038/
mp.2008.5 16. Khandaker GM, Stochl J, Zammit S, Lewis G, Jones PB. Childhood Epstein-
Barr virus infection and subsequent risk of psychotic experiences in adoles-
cence: a population-based prospective serological study. Schizophr Res (2014)
158:19–24. doi:10.1016/j.schres.2014.05.019 6. Torrey EF, Bartko JJ, Yolken RH. Toxoplasma gondii and other risk factors
for schizophrenia: an update. Schizophr Bull (2012) 38:642–7. doi:10.1093/
schbul/sbs043 17. Du Preez A, Leveson J, Zunszain PA, Pariante CM. Inflammatory insults and
mental health consequences: does timing matter when it comes to depression? Psychol Med (2016) 46:2041–57. doi:10.1017/S0033291716000672 7. Martino D, Defazio G, Giovannoni G. The PANDAS subgroup of tic disorders
and childhood-onset obsessive-compulsive disorder. J Psychosom Res (2009)
67:547–57. doi:10.1016/j.jpsychores.2009.07.004 18. Preisig M, Waeber G, Vollenweider P, Bovet P, Rothen S, Vandeleur C, et al. FUnDing The PsyCoLaus study was supported by research grants from
GlaxoSmithKline, the Faculty of Biology and Medicine of
Lausanne, and the Swiss National Science Foundation (grants
3200B0–105993, 3200B0-118308, 33CSCO-122661, 33CS30-
139468, and 33CS30-148401). The PsyCoLaus study was supported by research grants from
GlaxoSmithKline, the Faculty of Biology and Medicine of
Lausanne, and the Swiss National Science Foundation (grants
3200B0–105993, 3200B0-118308, 33CSCO-122661, 33CS30-
139468, and 33CS30-148401). The interpretation is solely based on association analyses and
results from regression analyses. Therefore, it includes a strong
theoretical or speculative component and should be adopted
with caution. Ongoing studies covering other infectious diseases cOnclUsiOn The analysis of the timing of childhood diseases in early anxiety
disorders has shed new light on the connection between viral
childhood diseases—chickenpox, measles, and mumps—and
social phobia. While the clinical implications of this study are
minor, its theoretical implications are challenging. The results
suggest a role for the immune system in impacting the develop-
ment of early onset anxiety disorders. The analysis of the timing of childhood diseases in early anxiety
disorders has shed new light on the connection between viral
childhood diseases—chickenpox, measles, and mumps—and
social phobia. While the clinical implications of this study are
minor, its theoretical implications are challenging. The results
suggest a role for the immune system in impacting the develop-
ment of early onset anxiety disorders. limitations As customary in population surveys, the information used was
based on self-reporting, including all information on child-
hood diseases. The participants in PsyCoLaus were adults up
to 66 years of age. Therefore, the analyses might be biased by
telescoping effects such as those found typically regarding onset
of substance use (27, 28) and display an age-dependent recall-bias
regarding symptoms and diseases occurring in childhood and
youth. However, no evidence for this was found in the analyses
adjusted for age. DiscUssiOn While the results of this study show a new facet of how the
immune system interferes in the etiopathogenesis of social December 2016 | Volume 7 | Article 203 4 Social Phobia and Childhood Diseases Ajdacic-Gross et al. and other mental disorders will facilitate a more precise and
comprehensive interpretation of the specific link between viral
childhood diseases and social phobia. phobia, they should not be generalized to the whole spectrum
of social phobia subtypes. As in other phobias (24) or OCD
(31), the analysis of epidemiological parameters, risk factors,
and comorbidity patterns split by sex and age at onset of first
symptoms indicates some hidden heterogeneity of this disorder,
as do clinical studies (32). aUThOr cOnTriBUTiOns MP, EC, and CV designed the PsyCoLaus study and acquired the
data. VA-G, AA, SR, and MM carried out the analysis. WK, WR,
RK, MargM, and RL discussed the preliminary results. VA-G and
AA wrote the paper. All the authors contributed to the interpreta-
tion of the results and to the critical revision of the manuscript. The diagnosis of social phobias and other disorders was based
on information from epidemiological instruments and not on
clinical assessment. In some instances, the analyses might also
have failed to reveal significant estimates due to small frequencies
of cases. While some infectious diseases were directly documented
in the DIGS, others were covered by a more general question. Thus, the recall bias might interfere differently, depending on the
question format. reFerences The PsyCoLaus study: methodology and characteristics of the sample of
a population-based survey on psychiatric disorders and their association
with genetic and cardiovascular risk factors. BMC Psychiatry (2009) 9:9. doi:10.1186/1471-244X-9-9 8. Murphy TK, Storch EA, Lewin AB, Edge PJ, Goodman WK. Clinical factors
associated with pediatric autoimmune neuropsychiatric disorders associated
with streptococcal infections. J Pediatr (2012) 160:314–9. doi:10.1016/
j.jpeds.2011.07.012 19. Firmann M, Mayor V, Vidal PM, Bochud M, Pecoud A, Hayoz D, et al. The
CoLaus study: a population-based study to investigate the epidemiology and 9. Witthauer C, Gloster AT, Meyer AH, Goodwin RD, Lieb R. Comorbidity of
infectious diseases and anxiety disorders in adults and its association with December 2016 | Volume 7 | Article 203 Frontiers in Psychiatry | www.frontiersin.org 5 Ajdacic-Gross et al. Social Phobia and Childhood Diseases genetic determinants of cardiovascular risk factors and metabolic syndrome. BMC Cardiovasc Disord (2008) 8:6. doi:10.1186/1471-2261-8-6 28. Anders S, Tanaka M, Kinney DK. Depression as an evolutionary strategy for
defense against infection. Brain Behav Immun (2013) 31:9–22. doi:10.1016/
j.bbi.2012.12.002 20. Novy J, Castelao E, Preisig M, Vidal PM, Waeber G, Vollenweider P, et al. Psychiatric co-morbidities and cardiovascular risk factors in people with
lifetime history of epilepsy of an urban community. Clin Neurol Neurosurg
(2012) 114:26–30. doi:10.1016/j.clineuro.2011.08.019 29. Katz BZ, Jason LA. Chronic fatigue syndrome following infections in
adolescents. Curr Opin Pediatr (2013) 25:95–102. doi:10.1097/MOP. 0b013e32835c1108 21. Nurnberger JI Jr, Blehar MC, Kaufmann CA, York-Cooler C, Simpson
SG, Harkavy-Friedman J, et al. Diagnostic interview for genetic studies. Rationale, unique features, and training. NIMH genetics initiative. Arch Gen
Psychiatry (1994) 51:849–59; discussion 63–4. doi:10.1001/archpsyc.1994. 03950110009002 30. Dantzer R, Heijnen CJ, Kavelaars A, Laye S, Capuron L. The neuroimmune
basis of fatigue. Trends Neurosci (2014) 37:39–46. doi:10.1016/j.tins.2013. 10.003 31. Rodgers S, Ajdacic-Gross V, Kawohl W, Muller M, Rossler W, Hengartner
MP, et al. Comparing two basic subtypes in OCD across three large com-
munity samples: a pure compulsive versus a mixed obsessive-compulsive
subtype. Eur Arch Psychiatry Clin Neurosci (2015) 265(8):719–34. doi:10.1007/
s00406-015-0594-0 22. Preisig M, Fenton BT, Matthey ML, Berney A, Ferrero F. Diagnostic interview
for genetic studies (DIGS): inter-rater and test-retest reliability of the French
version. Eur Arch Psychiatry Clin Neurosci (1999) 249:174–9. doi:10.1007/
s004060050084 32. Knappe S, Beesdo-Baum K, Fehm L, Stein MB, Lieb R, Wittchen HU. Social
fear and social phobia types among community youth: differential clinical fea-
tures and vulnerability factors. J Psychiatr Res (2011) 45:111–20. Frontiers in Psychiatry | www.frontiersin.org reFerences doi:10.1016/
j.jpsychires.2010.05.002 23. Endicott J, Spitzer RL. A diagnostic interview: the schedule for affective disor-
ders and schizophrenia. Arch Gen Psychiatry (1978) 35:837–44. doi:10.1001/
archpsyc.1978.01770310043002 24. Ajdacic-Gross V, Rodgers S, Muller M, Hengartner MP, Aleksandrowicz
A, Kawohl W, et al. Pure animal phobia is more specific than other specific
phobias: epidemiological evidence from the Zurich study, the ZInEP and
the PsyCoLaus. Eur Arch Psychiatry Clin Neurosci (2016) 266(6):567–77. doi:10.1007/s00406-016-0687-4 Conflict of Interest Statement: The authors declare that the research was con-
ducted in the absence of any commercial or financial relationships that could be
construed as a potential conflict of interest. Copyright © 2016 Ajdacic-Gross, Aleksandrowicz, Rodgers, Müller, Kawohl, Rössler,
Castelao, Vandeleur, von Känel, Mutsch, Lieb and Preisig. This is an open-access
article distributed under the terms of the Creative Commons Attribution License (CC
BY). The use, distribution or reproduction in other forums is permitted, provided the
original author(s) or licensor are credited and that the original publication in this
journal is cited, in accordance with accepted academic practice. No use, distribution
or reproduction is permitted which does not comply with these terms. 25. Chernick MR. Bootstrap Methods: A Guide for Practitioners and Researchers. Hoboken: John Wiley & Sons (2007). 26. Ascherio A, Munger KL. Epstein-Barr virus infection and multiple scle-
rosis: a review. J Neuroimmune Pharmacol (2010) 5:271–7. doi:10.1007/
s11481-010-9201-3 27. Stein MB, Stein DJ. Social anxiety disorder. Lancet (2008) 371:1115–25. doi:10.1016/S0140-6736(08)60488-2 December 2016 | Volume 7 | Article 203 Frontiers in Psychiatry | www.frontiersin.org 6
|
https://openalex.org/W4319593305
|
https://hull-repository.worktribe.com/file/4193656/1/Published%20article
|
English
| null |
Modulation of transient receptor potential (TRP) channels by plant derived substances used in over-the-counter cough and cold remedies
|
Respiratory research
| 2,023
|
cc-by
| 15,754
|
Stinson et al. Respiratory Research (2023) 24:45
https://doi.org/10.1186/s12931-023-02347-z Stinson et al. Respiratory Research (2023) 24:45
https://doi.org/10.1186/s12931-023-02347-z Respiratory Research Open Access Modulation of transient receptor potential
(TRP) channels by plant derived substances
used in over‑the‑counter cough and cold
remedies Rebecca J. Stinson1, Alyn H. Morice2 and Laura R. Sadofsky1* © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Abstract Background Upper respiratory tract infections (URTIs) impact all age groups and have a significant economic and
social burden on society, worldwide. Most URTIs are mild and self-limiting, but due to the wide range of possible
causative agents, including Rhinovirus (hRV), Adenovirus, Respiratory Syncytial Virus (RSV), Coronavirus and Influenza,
there is no single and effective treatment. Over-the-counter (OTC) remedies, including traditional medicines and
those containing plant derived substances, help to alleviate symptoms including inflammation, pain, fever and cough. Purpose This systematic review focuses on the role of the major plant derived substances in several OTC remedies
used to treat cold symptoms, with a particular focus on the transient receptor potential (TRP) channels involved in
pain and cough. Methods Literature searches were done using Pubmed and Web of Science, with no date limitations, using the prin-
ciples of the PRISMA statement. The search terms used were ‘TRP channel AND plant compound’, ‘cough AND plant
compound’, ‘cough AND TRP channels AND plant compound’, ‘cough AND P2X3 AND plant compound’ and ‘P2X3 AND
plant compound’ where plant compound represents menthol or camphor or eucalyptus or turpentine or thymol. Results The literature reviewed showed that menthol activates TRPM8 and may inhibit respiratory reflexes reducing
irritation and cough. Menthol has a bimodal action on TRPA1, but inhibition may have an analgesic effect. Eucalyptus
also activates TRPM8 and inhibits TRPA1 whilst down regulating P2X3, aiding in the reduction of cough, pain and
airway irritation. Camphor inhibits TRPA1 and the activation of TRPM8 may add to the effects of menthol. Activation of
TRPV1 by camphor, may also have an analgesic effect. Conclusions The literature suggests that these plant derived substances have multifaceted actions and can interact
with the TRP ‘cough’ receptors. The plant derived substances used in cough and cold medicines have the potential to
target multiple symptoms experienced during a cold. Keywords Menthol, Camphor, Eucalyptus, Thymol, TRP channels, Cough, Cold, Over-the-counter *Correspondence:
Laura R. Sadofsky
L.R.Sadofsky@hull.ac.uk
1 Centre for Biomedicine, Hull York Medical School, The University of Hull,
Cottingham Road, Hull HU6 7RX, UK
2 Clinical Sciences Centre, Hull York Medical School, Castle Hill Hospital,
Cottingham, Hull HU16 5JQ, UK Introduction There is encouraging evidence to suggest that transient
receptor potential (TRP) cation channels may play a role
in cough and airway inflammation. Furthermore, several
of the plant derived substances used in traditional and
herbal cough and cold remedies are known to modulate Upper respiratory tract infections (URTIs) represent a
significant global burden on society from both a social
and economic perspective owing to high morbidity lev-
els across all age groups [1]. The exact cost is difficult
to estimate however, extrapolation from direct cost of
healthcare, over-the-counter (OTC) cough and cold
remedy sales, and loss of income suggest URTIs cost in
excess of $20–40 billion annually in the United States
(US) [2–5] and approximately £11 billion in the United
Kingdom (UK) [6]. With a range of potential causative
agents including human Rhinovirus (hRV), Adenovi-
rus, Respiratory Syncytial Virus (RSV), Coronavirus and
Influenza, URTIs cause a variety of symptoms [5]. Com-
mon symptoms include cough, nasal congestion and
excessive mucus production [7]. Although the majority
of URTIs are mild and self-limiting in nature [8], there is
no single effective treatment for the troublesome symp-
toms [9]. The desire to alleviate symptoms has led to a
number of OTC remedies, such as anti-inflammatories,
analgesics and antipyretics to target fever and muscle
pain, alongside H1 receptor antagonists, decongestants
and nasal sprays which target nasal congestion. Further-
more, cough can be targeted through specific antitussive
medicines [10–12]. A number of herbal and traditional
remedies also exist to help alleviate symptoms includ-
ing honey as an antitussive agent [13], saline solutions
for nasal congestion and throat irritation [14], vitamins,
mineral supplements and remedies purported to boost
immune function [12, 15–17] and topical vapour rub
ointments containing menthol, camphor and eucalyp-
tus, which release therapeutic vapours aimed at reducing
cough and congestion and making breathing easier [18]. © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Stinson et al. Respiratory Research (2023) 24:45 Stinson et al. Respiratory Research (2023) 24:45 Page 2 of 17 Stinson et al. Respiratory Research Role of TRP channels in cough Comprised of 28 members, mammalian TRP chan-
nels are a family of conserved transmembrane proteins,
divided into six subfamilies; vanilloid (TRPV), melasta-
tin (TRPM), mucolipin (TRPML), canonical (TRPC),
ankyrin (TRPA) and polycystic (TRPP), each of which are
further subdivided into individuals members [23]. TRP
channels share a common structure, comprising of six
transmembrane spanning proteins, assembled as a tetra-
meric channel. With differences between family mem-
bers being derived from variations in the cytosolic N- and
C-terminals [24]. Cough and airway hypersensitivity have
been linked to upregulation of TRP channels on the sen-
sory nerves in the respiratory tract [25], however, not all
TRP channels are thought to function as cough receptors,
those most noticeably involved include TRPV1, TRPV4,
TRPM8 and TRPA1 (Fig. 1a). These receptors all play a
role in airway sensation, responding to changes in tem-
perature, pH, osmolarity, irritants and mechanical stretch
[26] (Fig. 1b). Activation of these TRP channels occurs on
reaching a threshold of tolerance to a stimulus, this opens
the channels, enabling the movement of ions across the
membrane and generation of an action potential, leading
to the observed response [27, 28]. However, this activa-
tion can often be attenuated through the use of suitable
antagonist compounds, which alter the channel response
[29]. The threshold at which an action potential is propa-
gated varies, with factors such as underlying respiratory
disease and hyperstimulation having the potential to lead
to hypersensitisation, desensitisation and down-regula-
tion of TRP receptors [30]. For example, the TRPV1 role
in cough is linked to the increased sensitivity to capsaicin
seen in asthmatics and COPD patients, which suggests
that during inflammation there is upregulation of TRPV1
expression and function [31]. g
g
g
g
[
]
Menthol, eucalyptus and camphor form the main
ingredients in many herbal cough and cold remedies [19–
21] and are the focus of this review. Whilst, historically
these plants were chosen for their medicinal properties
and the relief they provide from a number of symptoms
when infused, steeped or heated to create herbal drinks,
the investigation of the pharmacological mechanism of
action is a more recent development. Scientific investiga-
tion into these mechanisms has been sparse and is still
not fully understood [22]. Furthermore, each of these
plant derived substances have a range of specific phar-
macological activities that are potentially beneficial in the
alleviation of cold symptoms and other diseases (Table 1).h Introduction substances found in OTC cough and cold remedies
potentially interact with TRP channels involved in the
cough reflex creating the clinical effects observed. Upper respiratory tract infections (URTIs) represent a
significant global burden on society from both a social
and economic perspective owing to high morbidity lev-
els across all age groups [1]. The exact cost is difficult
to estimate however, extrapolation from direct cost of
healthcare, over-the-counter (OTC) cough and cold
remedy sales, and loss of income suggest URTIs cost in
excess of $20–40 billion annually in the United States
(US) [2–5] and approximately £11 billion in the United
Kingdom (UK) [6]. With a range of potential causative
agents including human Rhinovirus (hRV), Adenovi-
rus, Respiratory Syncytial Virus (RSV), Coronavirus and
Influenza, URTIs cause a variety of symptoms [5]. Com-
mon symptoms include cough, nasal congestion and
excessive mucus production [7]. Although the majority
of URTIs are mild and self-limiting in nature [8], there is
no single effective treatment for the troublesome symp-
toms [9]. The desire to alleviate symptoms has led to a
number of OTC remedies, such as anti-inflammatories,
analgesics and antipyretics to target fever and muscle
pain, alongside H1 receptor antagonists, decongestants
and nasal sprays which target nasal congestion. Further-
more, cough can be targeted through specific antitussive
medicines [10–12]. A number of herbal and traditional
remedies also exist to help alleviate symptoms includ-
ing honey as an antitussive agent [13], saline solutions
for nasal congestion and throat irritation [14], vitamins,
mineral supplements and remedies purported to boost
immune function [12, 15–17] and topical vapour rub
ointments containing menthol, camphor and eucalyp-
tus, which release therapeutic vapours aimed at reducing
cough and congestion and making breathing easier [18]. Menthol, eucalyptus and camphor form the main
ingredients in many herbal cough and cold remedies [19–
21] and are the focus of this review. Whilst, historically
these plants were chosen for their medicinal properties
and the relief they provide from a number of symptoms
when infused, steeped or heated to create herbal drinks,
the investigation of the pharmacological mechanism of
action is a more recent development. Scientific investiga-
tion into these mechanisms has been sparse and is still
not fully understood [22]. Furthermore, each of these
plant derived substances have a range of specific phar-
macological activities that are potentially beneficial in the
alleviation of cold symptoms and other diseases (Table 1). TRPA1 TRPA1, previously ANKTM1, is the only TRPA channel
expressed in mammalian cells and acts as a non-selec-
tive cation channel. Found in fibroblasts, epithelial cells
(including airway epithelia), melanocytes, smooth mus-
cle cells and neurons [23, 32, 33]. TRPA1 is frequently
co-localised on afferent neurons alongside TRPV1 how-
ever, activation of either receptor is dependent on the
presence of specific stimuli [34]. TRPA1 is predominately
activated by a wide range of chemicals including those
with strong odours and tastes including garlic, horserad-
ish, cinnamon oil, mustard oil and wasabi [23, 35]. Acti-
vation can occur via covalent modification of cysteine Stinson et al. Respiratory Research (2023) 24:45 Page 3 of 17 residues on the N-terminus of the receptor [34, 36]. In
addition, TRPA1 is a thermoreceptor, being activated by
noxious cold temperatures below 17 °C [37] and nocicep-
tor, creating a potential target for pain relief and reduc-
ing bronchial hyperresponsiveness in asthmatics when
of TRPA1 agonists such as cinnamaldehyde are known to
cause cough, confirming its role as a cough receptor [39]. TRPA1 TRPM8
TRPM channels were first identified in tumour cells
Table 1 Medicinal properties and chemical structures of the plant derived substances frequently found in traditional herbal cough
and cold remedies [22, 96, 97, 104, 114, 146–160]
✓denotes medicinal properties most relevant to the management of cold symptoms
*denotes additional medicinal properties not specifically related to cold symptoms
Menthol
Camphor
Eucalyptus oil
Turpentine oil
Thymol
Cedarleaf oil
Nutmeg oil
Plant
Mentha x piperita
(Peppermint)
and other mem-
bers of the mint
family
Cinnamomum
camphora (Cam-
phor Laurel)
Eucalyptus globu-
lus (Tasmanian
blue gum) and
other members
of the eucalyptus
family
Pinus Pinaster
(Maritime pine)
and other mem-
bers of the pine
family
Thymus vulgaris
Thuja orientalis
(Arbor vitae) and
other members
of the Cupres-
saceae family
Myristica fragrans
(Fragrant nut-
meg)
Chemical
Structure of Main
Pharmacologi-
cally Compound
Medicinal properties described
Antibacterial
✓
–
✓
–
–
✓
–
Analgesic
✓
✓
✓
–
–
–
✓
Anti-inflamma-
tory
–
✓
✓
✓
–
–
✓
Antioxidant
–
–
✓
✓
–
✓
✓
Antiviral
–
–
–
✓
–
✓
–
Antimicrobial
–
–
✓
✓
✓
–
✓
Antitussive
✓
✓
–
–
✓
–
–
Antipyretic
–
–
–
–
✓
–
–
Expectorant
–
✓
–
✓
✓
–
–
Sedative
–
–
–
–
✓
–
–
Cooling effect
✓
–
–
–
–
–
–
Counter irritant
–
✓
–
–
–
–
–
Antipruritic
*
*
–
–
–
–
–
Antifungal
*
–
*
–
*
*
–
Antiseptic
*
–
–
*
–
–
*
Antispasmodic
–
*
–
–
*
–
–
H1 receptor
antagonist
–
–
*
–
–
–
–
Antiparasitic
–
–
–
*
*
–
–
Astringent
–
–
–
–
–
*
–
Disinfectant
–
–
*
–
–
–
– Table 1 Medicinal properties and chemical structures of the plant derived substances frequently found in traditional herbal cough
and cold remedies [22, 96, 97, 104, 114, 146–160] Table 1 Medicinal properties and chemical structures of the plant derived substances frequently found in traditional herbal cough
and cold remedies [22, 96, 97, 104, 114, 146–160] Menthol
Camphor
Eucalyptus oil
Turpentine oil
Thymol
Cedarleaf oil
Nutmeg oil
Plant
Mentha x piperita
(Peppermint)
and other mem-
bers of the mint
family
Cinnamomum
camphora (Cam-
phor Laurel)
Eucalyptus globu-
lus (Tasmanian
blue gum) and
other members
of the eucalyptus
family
Pinus Pinaster
(Maritime pine)
and other mem-
bers of the pine
family
Thymus vulgaris
Thuja orientalis
(Arbor vitae) and
other members
of the Cupres-
saceae family
Myristica fragrans
(Fragrant nut-
meg)
Chemical
Structure of Main
Pharmacologi-
cally Compound Chemical
Structure of Main
Pharmacologi-
cally Compound y ✓denotes medicinal properties most relevant to the management of cold symptoms
*denotes additional medicinal properties not specifically related to cold symptoms
Medicinal properties described
Antibacterial
✓
–
✓
–
–
✓
–
Analgesic
✓
✓
✓
–
–
–
✓
Anti-inflamma-
tory
–
✓
✓
✓
–
–
✓
Antioxidant
–
–
✓
✓
–
✓
✓
Antiviral
–
–
–
✓
–
✓
–
Antimicrobial
–
–
✓
✓
✓
–
✓
Antitussive
✓
✓
–
–
✓
–
–
Antipyretic
–
–
–
–
✓
–
–
Expectorant
–
✓
–
✓
✓
–
–
Sedative
–
–
–
–
✓
–
–
Cooling effect
✓
–
–
–
–
–
–
Counter irritant
–
✓
–
–
–
–
–
Antipruritic
*
*
–
–
–
–
–
Antifungal
*
–
*
–
*
*
–
Antiseptic
*
–
–
*
–
–
*
Antispasmodic
–
*
–
–
*
–
–
H1 receptor
antagonist
–
–
*
–
–
–
–
Antiparasitic
–
–
–
*
*
–
–
Astringent
–
–
–
–
–
*
–
Disinfectant
–
–
*
–
–
–
– ✓denotes medicinal properties most relevant to the management of cold symptoms
*denotes additional medicinal properties not specifically related to cold symptoms of TRPA1 agonists such as cinnamaldehyde are known to
cause cough, confirming its role as a cough receptor [39]. TRPA1 of TRPA1 agonists such as cinnamaldehyde are known to
cause cough, confirming its role as a cough receptor [39]. TRPM8 residues on the N-terminus of the receptor [34, 36]. In
addition, TRPA1 is a thermoreceptor, being activated by
noxious cold temperatures below 17 °C [37] and nocicep-
tor, creating a potential target for pain relief and reduc-
ing bronchial hyperresponsiveness in asthmatics when
exposed to inhaled irritants [38]. Furthermore, inhalation TRPV1 Involvement of P2X3 (TRPV4–ATP–P2X3 pathway) nvolvement of P2X3 (TRPV4–ATP–P2X3 pathway) TRPV1 is activated by vanilloid compounds most mark-
edly that of capsaicin, in addition to camphor, black
pepper, ethanol and garlic [35, 46–48]. Other chemi-
cal and physical stimuli include temperatures above
43 °C, low pH and spider toxin [24, 33, 35, 38]. TRPV1
is expressed in the liver, heart, pancreas and lungs [49]
however, the most predominant expression is in affer-
ent nerve fibres throughout the skin and gut, thus act-
ing as both a thermoreceptor and nociceptor whereby
it plays a key role in pain detection [23, 32, 33, 50]. The
activation of afferent nerve fibres also plays a role in
the airways, insofar as activation of TRPV1 by inhaled
irritants results in increased mucus secretion, bron-
choconstriction and an urge to cough. Furthermore,
increased expression of TRPV1 is linked to chronic
cough and hypersensitivity in chronic airway diseases
[49, 51]. The activation of TRPV1 as well as TRPA1 can
also occur via intracellular calcium, furthermore, the
co-expression of the receptors can result in one channel
sensitising the other [67]. Thus it is possible that during
airway inflammation both receptors may be activated
simultaneously [68], as such antitussive agents may be
better targeted to both receptors rather than individual
ones [34]. As a common symptom of URTIs, cough is an area of sig-
nificant interest as for some individuals acute cough can
become chronic, lasting in excess of 8 weeks however,
the mechanism involved has not been fully elucidated
[55]. Of particular interest in this mechanism are the P2
purinergic receptors (P2R), specifically the P2XR, trans-
membrane cationic channels on sensory neurons, which
are mediated by ATP [60]. The receptor of most interest
is P2X3, whereby activation of TRPV4 causes the release
of ATP, through pannexin-1, which subsequently acti-
vates P2X3 eliciting a cough response (Fig. 2) [59]. The
cough response can be partially attenuated using a P2X3
antagonist, which shows promise for the treatment of
chronic cough [61, 62] and thus may provide a potential
target for the treatment of cough as a symptom of URTIs. Indeed, recently hRV-16 has been shown to increase
ATP release by airway epithelial cells with and without
secondary TRPV4 stimulation suggesting a role for ATP
release in URTIs [63]. TRPV4 TRPV4 is widely expressed throughout mammalian tis-
sue including in the nervous system, heart, skin, kid-
neys, sweat and salivary glands, and lungs [52]. TRPV4
is activated by a number of mechanical stimuli includ-
ing changes to osmolarity, mechanical stretch, shear
stress and temperatures between 24 and 38 °C [33, 53]. Suggested TRPV4 functionality includes regulation of
blood flow, ciliary action control, osmotic regulation,
vasodilation and nociception [54]. The osmoregulation
function of TRPV4 is of particular interest in relation
to viral URTIs which can result in the upregulation
of mucus and changes in viscosity, thus altering the
hypotonicity of the mucus in airways, leading to the
activation of TPRV4 [55, 56]. Furthermore, TRPV4 is
implicated in respiratory function and disease, play-
ing a role in endothelial and epithelial barrier integrity,
smooth muscle constriction and regulation of inflam-
mation, which if compromised can result in alveolar
oedema [57, 58], whilst gene polymorphisms are linked
to chronic obstructive pulmonary disease (COPD) [59]. TRPV4 has also been linked to the cough reflex owing
to the production of adenosine triphosphate (ATP) in
response to activation, which activates other receptors,
namely purinergic receptor P2X3 [59]. TRPM8 TRPV
is expressed in the liver, heart, pancreas and lungs [4
however, the most predominant expression is in affe
ent nerve fibres throughout the skin and gut, thus ac
ing as both a thermoreceptor and nociceptor whereb
it plays a key role in pain detection [23, 32, 33, 50]. T
activation of afferent nerve fibres also plays a role
the airways, insofar as activation of TRPV1 by inhale
irritants results in increased mucus secretion, bro
choconstriction and an urge to cough. Furthermor
increased expression of TRPV1 is linked to chron
cough and hypersensitivity in chronic airway diseas
[49, 51]. The activation of TRPV1 as well as TRPA1 ca
also occur via intracellular calcium, furthermore, th
co-expression of the receptors can result in one chann
sensitising the other [67]. Thus it is possible that durin
airway inflammation both receptors may be activate
simultaneously [68], as such antitussive agents may b
better targeted to both receptors rather than individu
ones [34]. TRPV4
TRPV4 is widely expressed throughout mammalian ti
sue including in the nervous system, heart, skin, ki
neys, sweat and salivary glands, and lungs [52]. TRPV
is activated by a number of mechanical stimuli inclu
ing changes to osmolarity, mechanical stretch, she
stress and temperatures between 24 and 38 °C [33, 53
Suggested TRPV4 functionality includes regulation
blood flow, ciliary action control, osmotic regulatio
vasodilation and nociception [54]. The osmoregulatio
function of TRPV4 is of particular interest in relatio
to viral URTIs which can result in the upregulatio
of mucus and changes in viscosity, thus altering th
hypotonicity of the mucus in airways, leading to th
activation of TPRV4 [55, 56]. Furthermore, TRPV4
implicated in respiratory function and disease, pla
ing a role in endothelial and epithelial barrier integrit
smooth muscle constriction and regulation of inflam
mation, which if compromised can result in alveol
oedema [57, 58], whilst gene polymorphisms are linke
to chronic obstructive pulmonary disease (COPD) [59
TRPV4 has also been linked to the cough reflex owin
to the production of adenosine triphosphate (ATP)
response to activation, which activates other receptor
namely purinergic receptor P2X3 [59]. TRPV1 Of the aforementioned TRP channels, TRPA1 and
TRPV1 have the most significant link to the cough mech-
anism, although TRPV4 has also been postulated to play
some role being co-localised on the same sensory neu-
rons and through the TRPV4-ATP-P2X3 pathway [59]. Interestingly, TRPA1 and TRPV1 antagonist have been
shown to inhibit cough induced by irritants and agonists
of the channels e.g. citric acid or capsaicin. However,
therapies such as SB-705498 (TRPV1 antagonist) and
GRC 17536 (TRPA1 antagonist) failed to reduce cough in
chronic cough patients [64–66]. Importantly, the role of
these TRP channel modulators in URTI associated cough
have not yet been proven. TRPM8 TRPM channels were first identified in tumour cells
with expression linked to metastatic potential [23, 35]. Page 4 of 17 Stinson et al. Respiratory Research (2023) 24:45 Stinson et al. Respiratory Research Fig. 1 a Structural representation of the potential TRP channels involved in cough and airway hypersensitivity. Example chemicals which stimulate
each TRP channel are shown above and the mechanical and physical stimuli below. b Flow diagram representing the steps involved in the cough
reflex pathway in relation to the activation of TRP channels via relevant stimulants Fig. 1 a Structural representation of the potential TRP channels involved in cough and airway hypersensitivity. Example chemicals which stimulate
each TRP channel are shown above and the mechanical and physical stimuli below. b Flow diagram representing the steps involved in the cough
reflex pathway in relation to the activation of TRP channels via relevant stimulants increases in osmolarity [41]. Furthermore, activation of
TRPM8 by cooling compounds such as menthol, makes
it a potential analgesic target, as activation can allevi-
ate pain from inflammation and noxious heat [44]. Fur-
thermore, activation could have an anti-inflammatory
effect, whereby pro-inflammatory cytokine release is
inhibited [32] which may limit activation of nerve fibres
involved in cough, potentially providing an antitussive
effect [45]. TRPM8 is predominately expressed in neurons, but
also taste papillae, testis, prostate, lungs [40], cornea
[41], skin and bladder, weak expression is also observed
in pulmonary smooth muscle and liver [42]. TRPM8 is
activated by a number of chemical compounds, with
the most extensively studied being menthol and euca-
lyptus [43, 44]. Additionally, TRPM8 is thermoregu-
lated, being activated by cool temperatures between
23 and 28 °C [35] and shows evidence of responding to TRPM8 is predominately expressed in neurons, but
also taste papillae, testis, prostate, lungs [40], cornea
[41], skin and bladder, weak expression is also observed
in pulmonary smooth muscle and liver [42]. TRPM8 is
activated by a number of chemical compounds, with
the most extensively studied being menthol and euca-
lyptus [43, 44]. Additionally, TRPM8 is thermoregu-
lated, being activated by cool temperatures between
23 and 28 °C [35] and shows evidence of responding to Page 5 of 17 Stinson et al. Respiratory Research (2023) 24:45 Stinson et al. Respiratory Research (2023) 24:45 Stinson et al. Respiratory Research Stinson et al. TRPM8 Respiratory Research (2023) 24:45
TRPV1
TRPV1 is activated by vanilloid compound
edly that of capsaicin, in addition to cam
pepper, ethanol and garlic [35, 46–48]. O
cal and physical stimuli include temper
43 °C, low pH and spider toxin [24, 33, 35
is expressed in the liver, heart, pancreas a
however, the most predominant expressio
ent nerve fibres throughout the skin and g
ing as both a thermoreceptor and nocicep
it plays a key role in pain detection [23, 32
activation of afferent nerve fibres also pl
the airways, insofar as activation of TRPV
irritants results in increased mucus secr
choconstriction and an urge to cough. increased expression of TRPV1 is linked
cough and hypersensitivity in chronic air
[49, 51]. The activation of TRPV1 as well a
also occur via intracellular calcium, furth
co-expression of the receptors can result in
sensitising the other [67]. Thus it is possibl
airway inflammation both receptors may
simultaneously [68], as such antitussive ag
better targeted to both receptors rather th
ones [34]. TRPV4
TRPV4 is widely expressed throughout ma
sue including in the nervous system, hea
neys, sweat and salivary glands, and lungs
is activated by a number of mechanical st
ing changes to osmolarity, mechanical s
stress and temperatures between 24 and 3
Suggested TRPV4 functionality includes
blood flow, ciliary action control, osmoti
vasodilation and nociception [54]. The osm
function of TRPV4 is of particular intere
to viral URTIs which can result in the
of mucus and changes in viscosity, thus
hypotonicity of the mucus in airways, le
activation of TPRV4 [55, 56]. Furthermo
implicated in respiratory function and d
ing a role in endothelial and epithelial bar
smooth muscle constriction and regulatio
mation, which if compromised can resul
oedema [57, 58], whilst gene polymorphism
to chronic obstructive pulmonary disease
TRPV4 has also been linked to the cough
to the production of adenosine triphosph
response to activation, which activates oth
namely purinergic receptor P2X3 [59]. TRPV1
TRPV1 is activated by vanilloid compounds most mar
edly that of capsaicin, in addition to camphor, bla
pepper, ethanol and garlic [35, 46–48]. Other chem
cal and physical stimuli include temperatures abov
43 °C, low pH and spider toxin [24, 33, 35, 38]. Systematic reviewh The mechanisms involved in the action of some plant
derived substances used in herbal cough and cold rem-
edies are not well documented or explored. As such the
interactions between TRP channels and the plant derived
substances will be elucidated from the existing literature. To elucidate how plant derived substances, which form
the major ingredients of cold remedies, may interact
with TRP channels to alleviate the symptoms of the com-
mon cold, specifically cough, we searched Pubmed and
Web of Science for existing studies, with no date limita-
tions, using the principles of the PRISMA statement [69]
(Fig. 3).h The following search terms were used: ‘TRP chan-
nel AND plant compound’, ‘cough AND plant com-
pound’, ‘cough AND TRP channels AND plant
compound’, ‘cough AND P2X3 AND plant com-
pound’ and ‘P2X3 AND plant compound’ where plant Page 6 of 17 Stinson et al. Respiratory Research (2023) 24:45 Stinson et al. Respiratory Research Fig. 2 TRPV4-ATP-P2X3 pathway involved in the cough response. Proposed mechanism involves the activation of TRPV4 by hypotonic solutions,
irritant or known agonist, leading to the influx of calcium ions into the cytosol (1). Activation of TRPV4 and the increase in intracellular calcium, leads
to the release of ATP into the extracellular space mediated by pannexin-1 (2). Extracellular ATP activates P2X3 on sensory neurons creating an action
potential in the sensory neurons of the airways (3) which may subsequently trigger the cough reflex (4) Fig. 2 TRPV4-ATP-P2X3 pathway involved in the cough response. Proposed mechanism involves the activation of TRPV4 by hypotonic solutions,
irritant or known agonist, leading to the influx of calcium ions into the cytosol (1). Activation of TRPV4 and the increase in intracellular calcium, leads
to the release of ATP into the extracellular space mediated by pannexin-1 (2). Extracellular ATP activates P2X3 on sensory neurons creating an action
potential in the sensory neurons of the airways (3) which may subsequently trigger the cough reflex (4) Fig. 3 PRISMA flowchart, including exclusion criteria, utilised to screen identified citation to produce the final number of studies considered in the
systematic literature search Fig. 3 PRISMA flowchart, including exclusion criteria, utilised to screen identified citation to produce the final number of studies considered in the
systematic literature search in the systematic literature search (Table 2). Systematic reviewh When thy-
mol was searched for using the same search terms as
previously outlined, four articles were included, with
the main observations outlined below. compound represents menthol or camphor or euca-
lyptus or turpentine or thymol. Only peer reviewed,
primary research articles published in English which
included specified key terms were selected for inclusion Stinson et al. Respiratory Research (2023) 24:45 Page 7 of 17 Table 2 Publications included in the final systematic literature search
Citation
Publication type
Title
Plant derived substance and drug target
Alpizar et al. 2013 [99]
Original research
Bimodal effects of cinnamaldehyde and
camphor on mouse TRPA1
Camphor as a TRPA1 agonist and antagonist
Andrè et al. 2009 [100]
Original research
Transient receptor potential ankyrin receptor
1 is a novel target for pro-tussive agents
Camphor as a TRPA1 antagonist and antitus-
sive
Behrendt et al. 2004 [78]
Original research
Characterization of the mouse cold-menthol
receptor TRPM8 and vanilloid receptor
type-1 VR1 using a fluorometric imaging
plate reader (FLIPR) assay
Menthol as a TRPM8 agonist
Ben-Arye et al. 2011 [15]
Randomised controlled trial
Treatment of Upper Respiratory Tract Infec-
tions in Primary Care: A Randomized Study
Using Aromatic Herbs
Menthol, camphor and eucalyptus as an
antitussive
Bödding et al. 2007 [81]
Original research
Characterisation of TRPM8 as a pharmaco-
phore receptor
Menthol as a TRPM8 agonist
Buday et al. 2018 [121]
Randomised controlled trial
Modulation of cough response by sensory
inputs from the nose—role of trigeminal
TRPA1 versus TRPM8 channels
Menthol as an antitussive
Cohen et al. 2012 [126]
Randomised controlled trial
Effect of Honey on Nocturnal Cough and
Sleep Quality: A Double-blind, Randomized,
Placebo-Controlled Study
Eucalyptus honey as an antitussive
Ghosh et al. 2020 [109]
Original research
Essential Oils from Monarda fistulosa: Chem-
ical Composition and Activation of Transient
Receptor Potential A1 (TRPA1) Channels
Thymol as a TRPA1 agonist
Johnson et al. 2018 [122]
Observational study
Menthol Cough Drops: Cause for Concern? Menthol as an antitussive
Karashima et al. 2007 [93]
Comparative study
Bimodal action of menthol on the transient
receptor potential channel TRPA1
Menthol as a TRPA1 agonist and antagonist
Kenia et al. 2008 [20]
Randomised controlled trial
Does inhaling menthol affect nasal patency
or cough? Menthol and eucalyptus oil as an antitussive
Kumar et al. 2012 [124]
Original research
Effect of nitrogen insertion on the antitus-
sive properties of menthol and camphor
Menthol and camphor as an antitussive
Kurohane et al. Systematic reviewh 2013 [94]
Original research
Lack of transient receptor potential mel-
astatin 8 activation by phthalate esters that
enhance contact hypersensitivity in mice
Menthol as a TRPM8 agonist
Laude et al. 1994 [125]
Original research
The antitussive effects of menthol, camphor
and cineole in conscious guinea-pigs
Menthol, camphor and eucalyptus as an
antitussive
Lee et al. 2008 [111]
Original research
Thymol and related alkyl phenols activate
the hTRPA1 channel
Thymol as a TRPA1 agonist
Macpherson et al. 2006 [92]
Comparative study
More than cool: Promiscuous relationships
of menthol and other sensory compounds
Menthol and camphor as TRP channel
agonist and antagonist
Mahieu et al. 2007 [83]
Original research
TRPM8-independent menthol-induced
Ca2 + release from endoplasmic reticulum
and Golgi
Menthol as a TRPM8 agonist
Mälkiä et al. 2007 [161]
Original research
Bidirectional shifts of TRPM8 channel gat-
ing by temperature and chemical agents
modulate the cold sensitivity of mammalian
thermoreceptors
Menthol as a TRPM8 agonist
Marsakova et al. 2012 [101]
Comparative study
Pore Helix Domain Is Critical to Camphor
Sensitivity of Transient Receptor Potential
Vanilloid 1 Channel
Camphor as a TRPV1 agonist
McKemy et al. 2002 [84]
Original research
Identification of a cold receptor reveals a
general role for TRP channels in thermosen-
sation
Menthol as a TRP channel agonist
Mergler et al. 2013 [79]
Comparative study
Functional significance of thermosensitive
transient receptor potential melastatin
channel 8 (TRPM8) expression in immortal-
ized human corneal endothelial cells
Menthol and eucalyptus as an agonist for
TRPM8
Millqvist et al. 2013 [118]
Randomised controlled trial
Inhalation of menthol reduces capsaicin
cough sensitivity and influences inspiratory
flows in chronic cough
Menthol as an antitussive Table 2 Publications included in the final systematic
Citation
Publication type
Alpizar et al. 2013 [99]
Original research
Andrè et al. 2009 [100]
Original research
Behrendt et al. 2004 [78]
Original research
Ben-Arye et al. 2011 [15]
Randomised controlled trial
Bödding et al. 2007 [81]
Original research
Buday et al. 2018 [121]
Randomised controlled trial
Cohen et al. 2012 [126]
Randomised controlled trial
Ghosh et al. 2020 [109]
Original research
Johnson et al. 2018 [122]
Observational study
Karashima et al. 2007 [93]
Comparative study
Kenia et al. 2008 [20]
Randomised controlled trial
Kumar et al. 2012 [124]
Original research
Kurohane et al. 2013 [94]
Original research
Laude et al. 1994 [125]
Original research
Lee et al. 2008 [111]
Original research
Macpherson et al. 2006 [92]
Comparative study
Mahieu et al. Systematic reviewh 2007 [83]
Original research
Mälkiä et al. 2007 [161]
Original research
Marsakova et al. 2012 [101]
Comparative study
McKemy et al. 2002 [84]
Original research
Mergler et al. 2013 [79]
Comparative study
Millqvist et al. 2013 [118]
Randomised controlled trial Table 2 Publications included in the final systematic literature search Plant derived substance and drug target Menthol as a TRP channel agonist Stinson et al. Respiratory Research (2023) 24:45 Page 8 of 17 Table 2 (continued)
Citation
Publication type
Morice et al. 1994 [120]
Clinical trial
Paschke et al. 2017 [87]
Original research
Paul et al. 2010 [18]
Randomised controlled t
Peier et al. 2002 [75]
Original research
Pertusa et al. 2014 [76]
Original research
Pertusa et al. 2018 [77]
Comparative study
Plevkova et al. 2013 [80]
Original research
Sabnis et al. 2008 [40]
Original research
Selescu et al. 2013 [102]
Original research
Takaishi et al. 2012 [106]
Original research
Voets et al. 2004 [91]
Original research
Wang et al. 2020 [112]
Original research
Weil et al. 2005 [82]
Comparative study
Willis et al. 2011 [85]
Original research
Wise et al. 2012 [119]
Randomised controlled t
Xiao et al. 2008 [95]
Original research
Xing et al. 2006 [90]
Original research
Xu et al. 2005 [46]
Original research
Xu et al. 2015 [110]
Original research
Zhang et al. 2018 [117]
Original research Table 2 (continued)
Citation
Publication type
Title
Plant derived substance and drug target
Morice et al. 1994 [120]
Clinical trial
Effect of inhaled menthol on citric acid
induced cough in normal subjects
Menthol as an antitussive
Paschke et al. 2017 [87]
Original research
Activation of the cold-receptor TRPM8 by
low levels of menthol in tobacco products
Menthol as a TRPM8 agonist
Paul et al. 2010 [18]
Randomised controlled trial
Vapor Rub, Petrolatum, and No Treatment
for Children With Nocturnal Cough and Cold
Symptoms
Menthol, camphor and eucalyptus as an
antitussive
Peier et al. 2002 [75]
Original research
A TRP channel that senses cold stimuli and
menthol
Menthol as a TRPM8 agonist
Pertusa et al. 2014 [76]
Original research
Bidirectional modulation of thermal and
chemical sensitivity of TRPM8 channels by
the initial region of the N-terminal domain
Menthol as a TRPM8 agonist
Pertusa et al. 2018 [77]
Comparative study
Critical role of the pore domain in the cold
response of TRPM8 channels identified by
ortholog functional comparison
Menthol as a TRPM8 agonist
Plevkova et al. Plant extract interactions with TRP channels
Menthol activation of TRPM8 in nasal trigeminal afferent neurons
may go some way to explain the perceived congestion
relief experienced when menthol is inhaled, creating a
cooling effect and greater perception of airflow within the
nose, albeit not to a measurable extent [20, 80]. The abil-
ity of menthol to inhibit respiratory irritation is not only
of benefit in OTC medication for cold symptoms and as
an analgesic [82], but also formed one of the major addi-
tives in cigarettes whereby, it was used to reduce the irri-
tation caused by tobacco smoke [87]. However, it is worth
noting that this practice has been banned in Europe and
USA since 2020 and 2021, respectively [88, 89]. Menthol is a cyclic terpene alcohol derived from the
plants of the Mentha species, such as natural pepper-
mint, additionally, it can be synthesised from other essen-
tial oils. Known for its distinctive flavour and fragrance,
it has been widely used medicinally for over 2000 years
and is widely used in other products including confec-
tionary, toothpaste and cold medication [70–72]. Natural
menthol exists in two isomer, d- or (+)-menthol and l- or
(-)-menthol, with the former lacking medicinal proper-
ties [73]. Additionally, menthol has cooling and analge-
sic properties however, some adverse effects are linked
to over exposure including irritation, skin allergies and
burning sensations. Furthermore, in young children and
chronic obstructive pulmonary disease (COPD) patients,
over exposure has been linked to upper airway spasms,
reflex apnoea and breathing difficulties [73, 74]. The dosage of menthol also plays some role in the sen-
sation experienced, with low dosages causing a cooling
sensation and higher dosages a burning sensation [90]. The extent of sensation experienced could reflect the
number of menthol sensitive or insensitive neurons, with
sensitive neurons having higher expression of TRPM8
and lower thresholds to cold stimuli [90, 91]. In addition,
menthol also acts on the warm receptor TRPV3, thus
activation of this receptor may explain why high concen-
tration of menthol cause burning sensations [84, 92]. The
activation of TRPA1 by menthol may explain the role it
can play as an analgesic, as activation of TRPM8 alone
does not cause increased skin sensitisation. Plant extract interactions with TRP channels
Menthol Instead acti-
vation of TRPA1 by menthol can have a bimodal effect,
with low doses causing pain and inflammation and high
doses acting as an antagonist hence the analgesic effect
[93–95], thus analgesic effects are potentially a conse-
quence of TRPM8 activation and TRPA1 inhibition [82]. l
p
gfi
Menthol is a widely recognised TRPM8 agonist, as
the cooling effect generated by its inhalation or topical
application activates the cold sensitive channels on sen-
sory neurons [75]. This cooling effect occurs as a result
of the menthol binding to the N-terminal domain of the
channel [76, 77] with (–)-menthol being a more effec-
tive agonist than (+)-menthol [78–80]. Being a voltage
dependent channel, menthol binding leads to the depo-
larisation of the ion channel, shifting the voltage depend-
ence of the channel to the left, nearer to the membrane
potentials which are physiologically relevant for opening
[81, 82]. However, it should also be noted that the C-ter-
minal domain also plays some role in channel activation
[77]. This activation results in the influx of calcium from
both extracellular and intracellular sources, with the lat-
ter being via a TRPM8 independent mechanism [79, 83]. This method of activation occurs in both neuronal and
non-neuronal TRPM8 expressing cells, including the
bronchial epithelial cells. Activation of TRPM8 channels
in the lungs via cold sensitisation may play a role in man-
aging airway homeostasis in response to changes induced
by exposure to cold air or cooling agents [40]. The cooling
sensation has the benefit of inhibiting respiratory reflexes
and irritation, hence the antitussive effect, but the initial
activation of TRPM8 in nasal trigeminal afferent neurons
may play more of a role than those situated in the bron-
chopulmonary vagal afferent neurons [80, 84–86]. The Systematic reviewh 2013 [80]
Original research
The role of trigeminal nasal TRPM8-express-
ing afferent neurons in the antitussive
effects of menthol
Menthol as a TRPM8 agonist and antitussive
Sabnis et al. 2008 [40]
Original research
Human lung epithelial cells express a func-
tional cold-sensing TRPM8 variant
Menthol as a TRPM8 agonist
Selescu et al. 2013 [102]
Original research
Camphor Activates and Sensitizes Transient
Receptor Potential Melastatin 8 (TRPM8) to
Cooling and Icilin
Camphor as a TRPM8 agonist
Takaishi et al. 2012 [106]
Original research
1,8-cineole, a TRPM8 agonist, is a novel
natural antagonist of human TRPA1
Eucalyptus as agonist and antagonist for
TRPM8 and TRPA1
Voets et al. 2004 [91]
Original research
The principle of temperature-dependent
gating in cold- and heat-sensitive TRP
channels
Menthol as a TRPM8 agonist
Wang et al. 2020 [112]
Original research
Thymol activates TRPM8-mediated
Ca2 + influx for its antipruritic effects
and alleviates inflammatory response in
Imiquimod-induced mice
Thymol as a TRPM8 agonist
Weil et al. 2005 [82]
Comparative study
Conservation of functional and pharmaco-
logical properties in the distantly related
temperature sensors TRVP1 and TRPM8
Menthol as a TRPM8 agonist
Willis et al. 2011 [85]
Original research
Menthol attenuates respiratory irritation
responses to multiple cigarette smoke
irritants
Menthol and eucalyptus as an agonist for
TRPM8
Wise et al. 2012 [119]
Randomised controlled trial
Sweet taste and menthol increase cough
reflex thresholds
Menthol as an antitussive
Xiao et al. 2008 [95]
Original research
Identification of transmembrane domain
5 as a critical molecular determinant of
menthol sensitivity in mammalian TRPA1
channels
Menthol as a TRPA1 agonist
Xing et al. 2006 [90]
Original research
Chemical and cold sensitivity of two distinct
populations of TRPM8-expressing soma-
tosensory neurons
Menthol as a TRPM8 agonist
Xu et al. 2005 [46]
Original research
Camphor activates and strongly desensitizes
the transient receptor potential vanilloid
subtype 1 channel in a vanilloid-independ-
ent mechanism
Camphor as a TRPV1 agonist
Xu et al. 2015 [110]
Original research
Action of thymol on spontaneous excitatory
transmission in adult rat spinal substantia
gelatinosa neurons
Thymol as a TRPA1 agonist
Zhang et al. 2018 [117]
Original research
1,8-cineole decreases neuropathic pain
probably via a mechanism mediating P2X3
receptor in the dorsal root ganglion
Use of eucalyptus in targeting pain via P2X3
pathway Plant derived substance and drug target Eucalyptus as agonist and antagonist for
TRPM8 and TRPA1
Menthol as a TRPM8 agonist
Thymol as a TRPM8 agonist Stinson et al. Systematic reviewh Respiratory Research (2023) 24:45 Page 9 of 17 Stinson et al. Respiratory Research Table 2 (continued)
Citation
Publication type
Title
Plant derived substance and drug target
Zhou et al. 2011 [86]
Comparative study
Sensitivity of bronchopulmonary receptors
to cold and heat mediated by transient
receptor potential cation channel subtypes
in an ex vivo rat lung preparation
Menthol as a TRPM8 agonist Camphor p
Camphor is a derived from the wood of Camphor laurel
and other trees of the laurel family. Native to East Asia,
the distillation and purification of the wood creates an
essential oil with a distinct aroma and flavour, that has
long been utilised in traditional medicines [96]. Cam-
phor relieves irritation and itch, alongside acting as an
antiseptic and analgesic. It is used in topical pain relief
ointments and balms, or as an inhalant to ease nasal con-
gestion [97]. Whilst camphor has a number of benefits
there are also potential risks associated with accidental
ingestion or intranasal application of liquid or semi-solid
camphor products, these typically include gastrointesti-
nal symptoms, seizures and neurological changes [98]. Stinson et al. Respiratory Research (2023) 24:45 Page 10 of 17 Stinson et al. Respiratory Research (2023) 24:45 Camphor has been less widely explored in relation to
its effect on TRP channels than menthol, nevertheless
camphor has been implicated as an agonist or antagonist
in three different channels, namely TPRA1, TRPM8 and
TRPV1. Camphor has been identified as having a bimodal
effect on TRPA1, whereby, higher concentrations cre-
ate an antagonistic effect and lower concentrations cre-
ate an agonist response. However, these concentrations
being relatively close (600 μM and 300 μM, respectively)
mean that the antagonist effect may mask the agonist
response [99]. Although this antagonist effect is widely
documented, there is limited evidence to suggest this
has any impact on cough [100]. Alongside being able to
inhibit TRPA1, camphor has also been shown to activate
TRPV1 via the outer pore domain of the N-terminus
[101]. However, the exact mechanism by which this bind-
ing and activation of TRPV1 occurs is less clear [46]. A
few potential mechanisms have been proposed includ-
ing direct binding to the channel resulting in its opening
or indirect opening as a response to camphor initiating
a signalling pathway [46, 101]. Although camphor acti-
vates TRPV1, its efficacy is lower than other agonists,
such as capsaicin, requiring concentrations in the milli-
molar range. Furthermore, increased temperatures also
increase the activity of the channel, thus if utilised dur-
ing events of inflammation or irritation, the activation of
TRPV1 could be effective in creating the burning sensa-
tion, desensitisation and analgesic effect experienced
when applying camphor containing balms [46]. Camphor Camphor
has also been shown to activate the TRPM8 channel, at
temperatures in the physiological range of the cool acti-
vated channel and with concentrations similar to those
of menthol. Furthermore, camphor appears to have a
bimodal effect, blocking menthol activation of TRPM8,
in addition to activating the channel [102]. numerous chemical components, including predomi-
nately 1,8-cineole and some 1,4-cineole, which affects the
manner in which it interacts with TRP channels. 1,8-cin-
eole has been shown to activate TRPM8 whilst acting
as an antagonist of TRPA1, this is potentially due to the
chemical structure being similar to menthol. Conversely,
1,4-cineole activates both TRPA1 and TRPM8 however,
neither have any effect on TRPV1. This bimodal effect of
1,8-cineole may indicate that eucalyptus oil could form
a useful analgesic and anti-inflammatory as it does not
activate TRPA1 in the same manner as menthol [106]. As with menthol and camphor, how this plays a role in
the respiratory system is less clear however, inhalation
of eucalyptus oil vapours gives the sensation of a clearer
nose and may reflect activation of TRPM8 in the nasal
passages [107]. Thymolh Thymol has antitussive, antibacterial and expectorant
properties [108]. Thymol is often included in herbal rem-
edies and has been shown to activate TRP channels in a
similar manner to other plant derived substances utilised
in these remedies. Although not extensively researched
in relation to TRP channel activation, thymol has been
shown to activate TRPA1 at micromolar concentrations,
leading to intracellular calcium flux [109, 110]. Whilst
the actual mechanism involved is not clear, thymol
appears to directly activate TRPA1 and the action can
be blocked by camphor. Furthermore, thymol appears to
have a faster activation than other TRPA1 agonists such
as cinnamaldehyde, suggesting thymol acts via a different
mechanism, or binding site [111]. In addition, thymol has
a bimodal effect, both activating and inhibiting TRPA1
receptor at high concentrations [111]. This activation of
TRPA1 may explain the role thymol can play in pain relief
[110]. Alongside TRPA1, TRPM8 has also been shown to
be activated by thymol in a manner similar to menthol. Activation of TRPM8 by thymol may mean it also has an
anti-inflammatory effect [112] Eucalyptus oil
l
l Eucalyptus oil, derived from the native Australia tree foli-
age of Eucalyptus species, has been utilised for hundreds
of year [103]. It has a number of medicinal properties
including antimicrobial, analgesic, antioxidant, anti-
inflammatory and H1 receptor antagonism, as well as
potential cancer therapy [104]. Over exposure can lead to
dry itchy skin and burning sensations, whilst accidental
ingestion can lead to gastrointestinal upset, central nerv-
ous system depression, plus cardiovascular and respira-
tory complications [105]. lf
Whilst not included in our systematic review, cedarleaf
oil has traditional use in the treatment of URTI symp-
toms and wounds, acting as an antiviral and antibacte-
rial agent [113]. In addition, nutmeg oil functions as an
anti-inflammatory, antiseptic, antimicrobial, analgesic
and antioxidant [114, 115]. It is also worth noting that
the systematic search of the literature did not produce
any evidence of interactions between turpentine oil and
any of the TRP channels included. However, turpentine
oil has numerous beneficial medicinal properties includ-
ing acting as a disinfectant, expectorant, antiseptic and
antiparasitic, it is used in the treatment of bronchitis, and
may aid in transdermal drug delivery [116]. Eucalyptus oil has been linked to activation of TRP
channels however, the extent of research is limited and
often done in conjunction with menthol [78, 79, 84]. Eucalyptus oil is a TRPM8 agonist, activating recep-
tors on sensory neurons, albeit to a lesser extent than
on menthol [78, 79]. Eucalyptus oil is comprised of Stinson et al. Respiratory Research (2023) 24:45 Page 11 of 17 Stinson et al. Respiratory Research (2023) 24:45 eucalyptus honey showed improved sleep, reduced cough
frequency and severity after consuming a measured dose
of honey, 30 min prior to the onset of sleep. However,
similar results were seen with citrus and labiatae honey,
suggesting the honey was causing the most pronounced
effect [126]. Interestingly, thymol also has some antitus-
sive effect. Nasal application of thymol has been shown
to cause a reduction in the number of coughs when chal-
lenged with capsaicin [127] and high concentrations have
been shown to exhibit some antispasmodic properties
on smooth muscle of the trachea [128]. Decongestion effect of plant derived substancesh Decongestion effect of plant derived substances
These plant derived substances also have the potential to
ease nasal congestion. As with studies relating to cough,
menthol features most frequently as aromatic compounds
in the treatment of URTI symptoms. Inhalation of men-
thol activates TRPM8 receptors within the nasal mucosa,
producing a cooling sensation and giving the effect of a
clearer nose [129]. Similarly, orally administered men-
thol causes a subjective easing of nasal congestion but no
marked changes in nasal airflow measurements [70]. The
absence of actual change in nasal patency [20] suggests
TRPM8 may not be involved either directly or indirectly
in this mechanism of nasal patency [130]. Although this
sensation of decongestion appears mainly subjective,
this change may be a result of small reductions in ven-
tilation, albeit only transiently, immediately after inhala-
tion, coupled with cold receptor adaptation either locally
or centrally [131]. When camphor is mixed with other
aromatics including menthol, a similar effect is seen
whereby, the subjective sensation of nasal decongestion
is experienced but no changes to nasal airflow resistance
were observed [132]. Eucalyptus oil, thyme oil and men-
thol, when delivered nasally has been shown to increase
ciliary beat frequency which has the benefit of improving
mucociliary clearance [133]. Similarly, Myrtol® a mixture
of aromatic essential oils, including eucalyptus oil, which
is taken orally for respiratory disorders, showed evidence
of improving mucociliary clearance and increased cili-
ary beat frequency both in vitro and ex vivo [134]. Thy-
mol also plays a role in increasing mucociliary clearance
[128] and has been shown to have anti-inflammatory
effects in allergic disorders of the respiratory system
[135]. Whilst no single aromatic compound directly eases
the congestion associated with a URTI, the combined
use of these aromatic compounds provides a sensation Camphor also appears to have an antitussive effect,
albeit studied to a much lesser extent. Nevertheless, cam-
phor has been included in treatments for cough since
the eighteenth century, with the first commercial inhaler,
patented by John Mudge in 1778, utilising it as part of the
mixture (with opium) inhaled by patients to treat catar-
rhous cough [123]. Eucalyptus oil
l
l Although each
of these aromatic compounds have their own properties
and degree of effectiveness in treating cough, when com-
bined, in either spray or rubbing ointment form, they also
appear to have a beneficial effect on this URTI symptom,
reducing the severity and incidence of nocturnal cough
thus improving sleep [15, 18]. Furthermore, none of the plant derived substances
included in the systematic search provided evidence of
interacting with TRPV4. However, eucalyptus oil or more
specifically 1,8-cineole, has been shown to interact with
P2X3, resulting in the downregulation of P2X3 expres-
sion on dorsal root ganglia which subsequently creates an
analgesic effect [117]. Given the potential role of P2X3 in
cough, specifically the TRPV4-ATP-P2X3 pathway, the
use of eucalyptus oil could potentially interact with P2X3
in the airway reducing the effect of the ATP released
by TRPV4 and attenuating the cough response to some
extent. Therapeutic potential of plant derived substances
Antitussive effect of plant derived substances Menthol features widely in cough research and has
repeatedly been shown to have an antitussive effect. Inhaled menthol, at concentrations of approximately
1% is effective at reducing capsaicin cough sensitivity,
whereby high concentrations of capsaicin are required
to cause a cough response [118, 119]. When delivered
repeatedly in a measured dose, via an inhaler, menthol
acts as an effective antitussive, reducing cough frequency
[120]. Similarly, when delivered nasally, menthol also
appears to suppress airway irritation, inhibit cough and
decrease sensitivity to capsaicin. However, whether this
is an effect of TRPM8 activation in the nasal passages or
due to the high volatility of menthol enabling it to reach
the airways is not clear [121]. However, the effectiveness
of menthol as an antitussive is supported by sufferers
of acute cough, whereby consumption of mentholated
cough drops reduces cough symptoms, with individu-
als increasing the number of cough drops consumed as
cough severity increases [122]. Decongestion effect of plant derived substancesh Treatment of cough with both men-
thol and camphor for 5 min prior to the commencement
of a citric acid cough challenge caused a reduction in
cough response and latency in awake guinea pigs, dem-
onstrating the potential role of camphor in the treatment
of cough [124, 125].hf The effectiveness of eucalyptus oil as an antitussive is
less clear, as when used alone does not cause any notice-
able reduction in cough response [20, 125]. However,
eucalyptus oil is used as a carrier for menthol and so
features in a number of studies, whether this alters the
effectiveness of menthol is not clear [20, 120]. Further-
more, the treatment of childhood nocturnal cough with Stinson et al. Respiratory Research (2023) 24:45 Page 12 of 17 Page 12 of 17 Stinson et al. Respiratory Research may provide beneficial relief from a sore throat in the
early stages of URTIs. of reduced congestion, which when considered along-
side the increased ciliary beat frequency and improved
mucociliary clearance that a number of the plant derived
substances provide and the potential anti-inflammatory
effect of thymol, may mean that herbal remedies may not
only provide the sensation of clearer nasal passages but
actually clear some of the excess mucus experienced dur-
ing a URTI. Conclusionsh The use of plant derived substances for their medicinal
properties have a long and varied history, being utilised
in not only traditional remedies for a variety of aliments
but also in a number of widely available OTC treat-
ments, most noticeably for the treatment of the symp-
toms of cold and flu. Although there is no clearly defined
mechanism of action for many of these traditional herbal
remedies, the individual plant derived substances have
properties which when combined may explain how these
remedies help to alleviate symptoms of URTIs. Of most
interest is the interaction between the plant derived sub-
stances and the TRP channels. The cooling effect of men-
thol as a result of TRPM8 activation appears to have the
potential to inhibit respiratory reflexes, thus reducing
irritation and acting as an antitussive [80, 84–86], whilst
the bimodal action leading to the inhibition of TRPA1
may have an analgesic effect [79, 93, 95], thus targeting
two of the main symptoms experienced during a cold. In
contrast, camphor inhibits TRPA1, yet there is little evi-
dence to indicate whether this influences the cough reflex
[100] however, the activation of TRPM8 may build on
the effects of menthol, providing the sensation of easing
breathing [107]. Additionally, the activation of TRPV1
by camphor, may have an analgesic effect [46] thus con-
tinuing to enhance the effect of menthol by targeting the
muscle pain associated with some cold symptoms. Euca-
lyptus also activates TRPM8 [78, 79] and inhibits TRPA1
[106], thus further aiding in the reduction of nasal and
airway irritation, the sensation of nasal clearing and the
analgesic effect relieving muscle aches. Beyond the TRP
channels, the downregulation of P2X3 by eucalyptus
[117] may also have the potential to disrupt the TRPV4-
ATP-P2X3 pathway and attenuate the cough reflex to
some extent. These plants extracts also have the potential
to provide a mild sedative effect [137–139] and analge-
sic effect [144, 145] which may help improve sleep qual-
ity, which can be disrupted by the congestion and cough
experienced during a cold. Whether the action of the
plant derived substances is a consequence of interaction
with the TRP ‘cough’ receptors or some other mechanism
that is yet to be fully elucidated remains to be seen how-
ever, there is evidence to suggest that these plant derived
substances directly target a number of cold symptoms. When taken together, the aromatic compounds Sedative effect of plant derived substances Menthol, camphor and eucalyptus oil when combined
and applied topically appear to have a sedative effect
which may explain the benefit of topical application of
these substances prior to sleep, thus aiding in nocturnal
rest and the sensation of reduced symptoms overnight
[18, 136]. Both menthol and camphor have been shown
to reduce spontaneous motor activity in mice who are
exposed to the vapours, this is potentially due to men-
thol and camphor having similar chemical structures to
other known sedatives. The exact mechanisms of action
are unclear however, their mode of action may be a result
of either interactions with olfactory nerves or the nasal
mucosa when inhaled, which subsequently act on the
human γ-aminobutyric acid type A (GABAA) neurotrans-
mitters creating the sedative effect [137, 138]. Yomogi oil,
a traditional Japanese herbal medicine, extracted from
plants of the Artemisia species contains camphor and
1,8-cineole and has been shown to have a sedative effect
akin to the use of lavender oil, with 1,8-cineole having the
most potential for creating this effect [139]. Thymol also
has a sedative effectand appears to interact with GABAA
receptors leading to increased function of the neuro-
transmitter [140, 141]. Thus, this may explain why the use
of topical vapour rub ointment may help improve sleep
quality during common cold infections. Received: 6 October 2022 Accepted: 27 January 2023 them. As such, it is less clear whether the beneficial
effects observed when utilising these aromatic com-
pounds are due to direct or indirect interactions
between the receptors and the aromatic compounds,
or if the effects are owing to a single ingredient or the
cumulative effect of all the aromatic compound. Whilst
a number of questions are still pertaining to the mecha-
nism of action of these aromatic compounds, it is clear
that the plant derived substances used in traditional
herbal remedies have a multifaceted action and the
potential to target multiple symptoms experienced dur-
ing a cold. Thus, these plant derived substances and the
therapeutic vapours they release are as relevant today
as they have been for the treatment of cough and cold
symptoms for centuries. them. As such, it is less clear whether the beneficial
effects observed when utilising these aromatic com-
pounds are due to direct or indirect interactions
between the receptors and the aromatic compounds,
or if the effects are owing to a single ingredient or the
cumulative effect of all the aromatic compound. Whilst
a number of questions are still pertaining to the mecha-
nism of action of these aromatic compounds, it is clear
that the plant derived substances used in traditional
herbal remedies have a multifaceted action and the
potential to target multiple symptoms experienced dur-
ing a cold. Thus, these plant derived substances and the
therapeutic vapours they release are as relevant today
as they have been for the treatment of cough and cold
symptoms for centuries. Analgesic effect of plant derived substances Topical application of menthol ointments is widely used
to treat muscle pain as it has been shown to work by
decreasing pain sensation in the skin [71, 92, 142] and
altering blood flow to the underlying tissue [143], thus
application can create an analgesic effect [144, 145]. Although, topical menthol ointment application is typi-
cally used for the treatment of minor muscle injuries, the
analgesic effect it creates could also be beneficial in man-
aging the muscle pain experienced during some URTIs. The use of plant derived substances as an analgesic may
have potential beyond the treatment of URTI associated
muscle pain. The use of throat spray containing menthol
and eucalyptus has been shown to provide targeted and
localised relief of the sore throat sensation often expe-
rienced during a URTI [15]. As such the use of menthol
and eucalyptus in the forms of teas, sprays and lozenges When taken together, the aromatic compounds
appear to have the potential to interact with the TRP
‘cough’ receptors and have a beneficial effect on the
treatment of cold symptoms. However, it is worth con-
sidering the complexity of this interplay between TRP
receptors and the aromatic compounds which modulate Stinson et al. Respiratory Research (2023) 24:45 Stinson et al. Respiratory Research (2023) 24:45 Page 13 of 17 Page 13 of 17 Received: 6 October 2022 Accepted: 27 January 2023 Received: 6 October 2022 Accepted: 27 January 2023 Availability of data and materials 16. Cohen HA, Varsano I, Kahan E, Sarrell EM, Uziel Y. Effectiveness of an
herbal preparation containing echinacea, propolis, and vitamin C
in preventing respiratory tract infections in children: a randomized,
double-blind, placebo-controlled, multicenter study. Arch Pediatr
Adolesc Med. 2004;158:217–21. Abbreviations 8. Peters S, Rowbotham S, Chisholm A, Wearden A, Moschogianis S,
Cordingley L, Baker D, Hyde C, Chew-Graham C. Managing self-limiting
respiratory tract infections: a qualitative study of the usefulness of the
delayed prescribing strategy. Br J Gen Pract. 2011;61:e579–89.f 8. Peters S, Rowbotham S, Chisholm A, Wearden A, Moschogianis S,
Cordingley L, Baker D, Hyde C, Chew-Graham C. Managing self-limiting
respiratory tract infections: a qualitative study of the usefulness of the
delayed prescribing strategy. Br J Gen Pract. 2011;61:e579–89. 9. Abuelgasim H, Albury C, Lee J. Effectiveness of honey for symptomatic
relief in upper respiratory tract infections: a systematic review and
meta-analysis. BMJ Evid Based Med. 2021;26:57. 9. Abuelgasim H, Albury C, Lee J. Effectiveness of honey for symptomatic
relief in upper respiratory tract infections: a systematic review and
meta-analysis. BMJ Evid Based Med. 2021;26:57. y
10. Hart AM. An evidence-based approach to the diagnosis and manage-
ment of acute respiratory infections. J Nurse Practit. 2007;3:607–11. 10. Hart AM. An evidence-based approach to the diagnosis and manage-
ment of acute respiratory infections. J Nurse Practit. 2007;3:607–11. 11. Ozdemir S, Ozturk TC, Metiner Y, Ak R, Ocal O. Evaluation of the pre-
scriptions written for upper respiratory tract infections. Northern Clin
Istanbul. 2015;2:107–14. 11. Ozdemir S, Ozturk TC, Metiner Y, Ak R, Ocal O. Evaluation of the pre-
scriptions written for upper respiratory tract infections. Northern Clin
Istanbul. 2015;2:107–14. 12. Thinesse-Mallwitz M, Maydannik V, Keller T, Klement P. A homeopathic
combination preparation in the treatment of feverish upper respiratory
tract infections: an international randomized controlled trial. Comple-
ment Med Res. 2015;22:163–70. Author contributions Conceptualization, AM and LS; Methodology, RS; Validation, RS, AM and LS;
Formal Analysis, RS; Investigation, RS; Resources, LS; Data Curation, RS; Writ-
ing—Original Draft Preparation, RS; Writing—Review and Editing, RS, AM
and LS; Visualization, RS; Supervision, LS; Project Administration, LS; Funding
Acquisition, AM and LS. All authors read and approved the final manuscript. 13. Paul IM, Beiler J, McMonagle A, Shaffer ML, Duda L, Berlin CM Jr. Effect
of honey, dextromethorphan, and no treatment on nocturnal cough
and sleep quality for coughing children and their parents. Arch Pediatr
Adolesc Med. 2007;161:1140–6. 13. Paul IM, Beiler J, McMonagle A, Shaffer ML, Duda L, Berlin CM Jr. Effect
of honey, dextromethorphan, and no treatment on nocturnal cough
and sleep quality for coughing children and their parents. Arch Pediatr
Adolesc Med. 2007;161:1140–6. 14. Cabaillot A, Vorilhon P, Roca M, Boussageon R, Eschalier B, Pereirad B. Saline nasal irrigation for acute upper respiratory tract infections in
infants and children: a systematic review and meta-analysis. Paediatr
Respir Rev. 2020;36:151–8. Funding g
This work was supported by Procter & Gamble International Operations SA, 47
Route de Saint-Georges, 1213 Petit-Lancy, Geneva, Switzerland. 15. Ben-Arye E, Dudai N, Eini A, Torem M, Schiff E, Rakover Y. Treatment of
upper respiratory tract infections in primary care: a randomized study
using aromatic herbs. Evid Based Complement Altern Med. 2011;2011:
690346. Declarations 17. Vavilova VP, Abramov-Sommariva D, Steindl H, Wonnemann M, Ryzhova
EG, Rusova TV, Lebedenko AA, Kolchenko II. Effectiveness and toler-
ability of Tonsilgon® N in the treatment of recurrent upper respiratory
tract infections in children: a non-interventional study in Russia. Clin
Phytosci. 2016;2:6. References 1. Fendrick AM, Monto AS, Nightengale B, Sarnes M. The economic
burden of non-influenza-related viral respiratory tract infection in the
United States. Arch Intern Med. 2003;163:487–94. 2. Bertino JS. Cost burden of viral respiratory infections: issues for formu-
lary decision makers. Am J Med. 2002;112:42–9. 3. Dixon RE. Economic costs of respiratory tract infections in the United
States. Am J Med. 1985;78:45–51. 4. Nieman DC, Henson DA, Austin MD, Sha W. Upper respiratory tract
infection is reduced in physically fit and active adults. Br J Sports Med. 2011;45:987. 5. Thomas M, Bomar PA. Upper Respiratory Tract Infection. In: StatPearls. Treasure Island (FL): StatPearls Publishing, Copyright© 2021, StatPearls
Publishing LLC.; 2021. 6. Burki TK. The economic cost of respiratory disease in the UK. Lancet
Respir Med. 2017;5:381. 7. Esposito C, Garzarella EU, Bocchino B, D’Avino M, Caruso G, Buonomo
AR, Sacchi R, Galeotti F, Tenore GC, Zaccaria V, Daglia M. A standardized
polyphenol mixture extracted from poplar-type propolis for remission
of symptoms of uncomplicated upper respiratory tract infection (URTI):
a monocentric, randomized, double-blind, placebo-controlled clinical
trial. Phytomedicine. 2021;80: 153368. 7. Esposito C, Garzarella EU, Bocchino B, D’Avino M, Caruso G, Buonomo
AR, Sacchi R, Galeotti F, Tenore GC, Zaccaria V, Daglia M. A standardized
polyphenol mixture extracted from poplar-type propolis for remission
of symptoms of uncomplicated upper respiratory tract infection (URTI):
a monocentric, randomized, double-blind, placebo-controlled clinical
trial. Phytomedicine. 2021;80: 153368. Abbreviations
ANKTM1
Ankyrin-like with transmembrane domains protein 1
COPD
Chronic obstructive pulmonary disease
GABAA
γ-Aminobutyric acid type A
hRV
Human Rhinovirus
OTC
Over-the-counter
P2R
Purinergic P2 receptors
P2X3
Purinergic receptor 3
P2XR
Purinergic P2X receptors
RSV
Respiratory syncytial virus
TRP
Transient receptor potential
TRPA1
Transient receptor potential cation channel subfamily A (Ankyrin)
member 1
TRPV1
Transient receptor potential cation channel subfamily V (Vanilloid)
member 1
TRPV4
Transient receptor potential cation channel subfamily V member 4
TRPM8
Transient receptor potential cation channel subfamily M (Melasta-
tin) member 8
URTI
Upper respiratory tract infection Competing interests Naunyn Schmiede-
bergs Arch Pharmacol. 2015;388:421–36. 27. Zheng J. Molecular mechanism of TRP channels. Compr Physiol. 2013;3:221–42. 54. Cao S, Anishkin A, Zinkevich NS, Nishijima Y, Korishettar A, Wang Z, Fang
J, Wilcox DA, Zhang DX. Transient receptor potential vanilloid 4 (TRPV4)
activation by arachidonic acid requires protein kinase A-mediated
phosphorylation. J Biol Chem. 2018;293:5307–22. 28. Kaneko Y, Szallasi A. Transient receptor potential (TRP) channels: a clini-
cal perspective. Br J Pharmacol. 2014;171:2474–507. 29. Koivisto A-P, Belvisi MG, Gaudet R, Szallasi A. Advances in TRP channel
drug discovery: from target validation to clinical studies. Nat Rev Drug
Discovery. 2022;21:41–59. 55. Atkinson SK, Sadofsky LR, Morice AH. How does rhinovirus cause the
common cold cough? BMJ Open Respir Res. 2016;3: e000118. 30. Xiao R, Liu J, Xu XZS. Thermosensation and longevity. J Comp Physiol A
Neuroethol Sens Neural Behav Physiol. 2015;201:857–67. 56. Fernández-Fernández JM, Andrade YN, Arniges M, Fernandes J, Plata
C, Rubio-Moscardo F, Vázquez E, Valverde MA. Functional coupling of
TRPV4 cationic channel and large conductance, calcium-dependent
potassium channel in human bronchial epithelial cell lines. Pflügers
Arch Eur J Physiol. 2008;457:149. 31. Geppetti P, Patacchini R, Nassini R, Materazzi S. Cough: the emerging
role of the TRPA1 channel. Lung. 2010;188(Suppl 1):S63-68. 32. Khalil M, Alliger K, Weidinger C, Yerinde C, Wirtz S, Becker C, Engel MA. Functional role of transient receptor potential channels in immune cells
and epithelia. Front Immunol. 2018;9:174. y
57. Morty RE, Kuebler WM. TRPV4: an exciting new target to promote
alveolocapillary barrier function. Am J Physiol Lung Cell Mol Physiol. 2014;307:L817–21. 33. Li H. TRP channel classification. In: Wang Y, editor. Transient receptor
potential canonical channels and brain diseases. Dordrecht: Springer;
2017. p. 1–8. 58. Michalick L, Kuebler WM. TRPV4—a missing link between mechanosen-
sation and immunity. Front Immunol. 2020;11:413. 34. Taylor-Clark TE, Nassenstein C, McAlexander MA, Undem BJ. TRPA1:
a potential target for anti-tussive therapy. Pulm Pharmacol Ther. 2009;22:71–4. 59. Bonvini SJ, Birrell MA, Grace MS, Maher SA, Adcock JJ, Wortley MA,
Dubuis E, Ching Y-M, Ford AP, Shala F, et al. Transient receptor potential
cation channel, subfamily V, member 4 and airway sensory afferent
activation: role of adenosine triphosphate. J Allergy Clin Immunol. 2016;138:249-261.e212. 35. Venkatachalam K, Montell C. TRP channels. Annu Rev Biochem. 2007;76:387–417. 36. Sadofsky LR, Boa AN, Maher SA, Birrell MA, Belvisi MG, Morice AH. Competing interests 18. Paul IM, Beiler JS, King TS, Clapp ER, Vallati J, Berlin CM. Vapor rub, pet-
rolatum, and no treatment for children with nocturnal cough and cold
symptoms. Pediatrics. 2010;126:1092. p
g
LS received funding from Procter & Gamble to support this work. 19. DeGeorge KC, Ring DJ, Dalrymple SN. Treatment of the common cold. Am Fam Phys. 2019;100:281–9. Page 14 of 17 Page 14 of 17 Page 14 of 17 Stinson et al. Respiratory Research (2023) 24:45 20. Kenia P, Houghton T, Beardsmore C. Does inhaling menthol affect nasal
patency or cough? Pediatr Pulmonol. 2008;43:532–7. 48. McNamara FN, Randall A, Gunthorpe MJ. Effects of piperine, the
pungent component of black pepper, at the human vanilloid receptor
(TRPV1). Br J Pharmacol. 2005;144:781–90. 21. Love JN, Sammon M, Smereck J. Are one or two dangerous? Camphor
exposure in toddlers. J Emerg Med. 2004;27:49–54. 49. Du Q, Liao Q, Chen C, Yang X, Xie R, Xu J. The role of transient receptor
potential vanilloid 1 in common diseases of the digestive tract and the
cardiovascular and respiratory system. Front Physiol. 2019;10:1064. 22. McKay DL, Blumberg JB. A review of the bioactivity and potential
health benefits of peppermint tea (Mentha piperita L.). Phytother Res. 2006;20:619–33. 50. Elokely K, Velisetty P, Delemotte L, Palovcak E, Klein ML, Rohacs T,
Carnevale V. Understanding TRPV1 activation by ligands: insights from
the binding modes of capsaicin and resiniferatoxin. Proc Natl Acad Sci. 2016;113:E137. 23. Samanta A, Hughes TET, Moiseenkova-Bell VY. Transient receptor
potential (TRP) channels. In: Harris JR, Boekema EJ, editors. Membrane
protein complexes: structure and function. Singapore: Springer; 2018. p. 141–65. 51. Lee L-Y, Gu Q. Role of TRPV1 in inflammation-induced airway hypersen-
sitivity. Curr Opin Pharmacol. 2009;9:243–9. 24. Gaudet R. TRP channels entering the structural era. J Physiol. 2008;586:3565–75. 52. Jia Y, Wang X, Varty L, Rizzo CA, Yang R, Correll CC, Phelps PT, Egan
RW, Hey JA. Functional TRPV4 channels are expressed in human
airway smooth muscle cells. Am J Physiol Lung Cell Mol Physiol. 2004;287:L272-278. 25. Guan M, Ying S, Wang Y. Increased expression of transient receptor
potential channels and neurogenic factors associates with cough
severity in a guinea pig model. BMC Pulm Med. 2021;21:187. 26. Bonvini SJ, Belvisi MG. Cough and airway disease: the role of ion chan-
nels. Pulm Pharmacol Ther. 2017;47:21–8. 53. Goldenberg NM, Ravindran K, Kuebler WM. TRPV4: physiological role
and therapeutic potential in respiratory diseases. Competing interests TRPA1
is activated by direct addition of cysteine residues to the N-hydroxysuc-
cinyl esters of acrylic and cinnamic acids. Pharmacol Res. 2011;63:30–6. 60. Pelleg A, Schulman ES, Barnes PJ. Extracellular adenosine 5’-triphos-
phate in obstructive airway diseases. Chest. 2016;150:908–15. 61. Abdulqawi R, Dockry R, Holt K, Layton G, McCarthy BG, Ford AP, Smith
JA. P2X3 receptor antagonist (AF-219) in refractory chronic cough:
a randomised, double-blind, placebo-controlled phase 2 study. The
Lancet. 2015;385:1198–205. y
y
37. Belvisi MG, Dubuis E, Birrell MA. Transient receptor potential A1 chan-
nels: insights into cough and airway inflammatory disease. Chest. 2011;140:1040–7. 38. Patapoutian A, Tate S, Woolf CJ. Transient receptor potential channels:
targeting pain at the source. Nat Rev Drug Discov. 2009;8:55–68. 62. McGarvey LP, Birring SS, Morice AH, Dicpinigaitis PV, Pavord ID, Schelf-
hout J, Nguyen AM, Li Q, Tzontcheva A, Iskold B, et al. Efficacy and safety
of gefapixant, a P2X3 receptor antagonist, in refractory chronic cough
and unexplained chronic cough (COUGH-1 and COUGH-2): results from
two double-blind, randomised, parallel-group, placebo-controlled,
phase 3 trials. The Lancet. 2022;399:909–23. 39. Birrell MA, Belvisi MG, Grace M, Sadofsky L, Faruqi S, Hele DJ, Maher SA,
Freund-Michel V, Morice AH. TRPA1 agonists evoke coughing in guinea
pig and human volunteers. Am J Respir Crit Care Med. 2009;180:1042–7. 40. Sabnis AS, Shadid M, Yost GS, Reilly CA. Human lung epithelial cells
express a functional cold-sensing TRPM8 variant. Am J Respir Cell Mol
Biol. 2008;39:466–74. 63. Atkinson SK, Morice AH, Sadofsky LR. Rhinovirus-16 increases ATP
release in A549 cells without concomitant increase in production. ERJ
Open Res. 2020;6:00159. 41. Quallo T, Vastani N, Horridge E, Gentry C, Parra A, Moss S, Viana F, Bel-
monte C, Andersson DA, Bevan S. TRPM8 is a neuronal osmosensor that
regulates eye blinking in mice. Nat Commun. 2015;6:7150. 64. Khalid S, Murdoch R, Newlands A, Smart K, Kelsall A, Holt K, Dockry R,
Woodcock A, Smith JA. Transient receptor potential vanilloid 1 (TRPV1)
antagonism in patients with refractory chronic cough: a double-blind
randomized controlled trial. J Allergy Clin Immunol. 2014;134:56–62. 42. Noyer L, Grolez GP, Prevarskaya N, Gkika D, Lemonnier L. TRPM8 and
prostate: a cold case? Pflugers Arch. 2018;470:1419–29. l
43. Blackshaw LA, Brierley SM, Hughes PA. TRP channels: new targets for
visceral pain. Gut. 2010;59:126. 65. Mazzone SB, McGarvey L. Mechanisms and rationale for targeted thera-
pies in refractory and unexplained chronic cough. Clin Pharmacol Ther. 2021;109:619–36. p
44. 68. Long L, Yao H, Tian J, Luo W, Yu X, Yi F, Chen Q, Xie J, Zhong N, Chung
KF, Lai K. Heterogeneity of cough hypersensitivity mediated by TRPV1 Competing interests Kurohane K, Sahara Y, Kimura A, Narukawa M, Watanabe T, Daimon T,
Imai Y. Lack of transient receptor potential melastatin 8 activation by
phthalate esters that enhance contact hypersensitivity in mice. Toxicol
Lett. 2013;217:192–6. compound. J Am Acad Dermatol. 2007;57:873–8. 73. Farco JA, Grundmann O. Menthol—pharmacology of an important
naturally medicinal “cool.” Mini Rev Med Chem. 2013;13:124–31. 95. Xiao B, Dubin AE, Bursulaya B, Viswanath V, Jegla TJ, Patapoutian A. Identification of transmembrane domain 5 as a critical molecular
determinant of menthol sensitivity in mammalian TRPA1 channels. J
Neurosci. 2008;28:9640–51. 74. Gavliakova S, Buday T, Shetthalli VM, Plevkova J, Gavliakova S, Buday T,
Shetthalli VM, Plevkova J. Analysis of pathomechanisms involved in side
effects of menthol treatment in respiratory diseases. Open J Mol Integr
Physiol. 2013;03:21–6. y
75. Peier AM, Moqrich A, Hergarden AC, Reeve AJ, Andersson DA, Story GM,
Earley TJ, Dragoni I, McIntyre P, Bevan S, Patapoutian A. A TRP channel
that senses cold stimuli and menthol. Cell. 2002;108:705–15. 96. Chen W, Vermaak I, Viljoen A. Camphor—A Fumigant during the Black
Death and a Coveted Fragrant Wood in Ancient Egypt and Babylon—A
Review. Molecules. 2013;18:5434–54. 76. Pertusa M, González A, Hardy P, Madrid R, Viana F. Bidirectional modula-
tion of thermal and chemical sensitivity of TRPM8 channels by the initial
region of the N-terminal domain. J Biol Chem. 2014;289:21828–43. 97. Zuccarini P. Camphor: risks and benefits of a widely used natural prod-
uct. J Appl Sci Environ Manage. 2009;13. 98. Love JN, Sammon M, Smereck J. Are one or two dangerous? camphor
exposure in toddlers1 1Series Editors: Jeffrey N. Love, md, The George-
town University Emergency Department, Washington, DC; Wendy
Klein-Schwartz, PharmD, MPH, The Maryland Poison Center, Baltimore,
MD; Liesl Curtis, md, The Georgetown University Emergency Depart-
ment, Washington, DC. J Emerg Med. 2004; 27:49–54. 77. Pertusa M, Rivera B, González A, Ugarte G, Madrid R. Critical role of the
pore domain in the cold response of TRPM8 channels identified by
ortholog functional comparison. J Biol Chem. 2018;293:12454–71. 78. Behrendt HJ, Germann T, Gillen C, Hatt H, Jostock R. Characterization
of the mouse cold-menthol receptor TRPM8 and vanilloid receptor
type-1 VR1 using a fluorometric imaging plate reader (FLIPR) assay. Br J
Pharmacol. 2004;141:737–45. 99. Alpizar YA, Gees M, Sanchez A, Apetrei A, Voets T, Nilius B, Talavera K. Bimodal effects of cinnamaldehyde and camphor on mouse TRPA1. Pflügers Arch Eur J Physiol. 2013;465:853–64. 79. Competing interests Dussor G, Cao YQ. TRPM8 and migraine. Headache. 2016;56:1406–17. 66. Mukhopadhyay I, Kulkarni A, Aranake S, Karnik P, Shetty M, Thorat S,
Ghosh I, Wale D, Bhosale V, Khairatkar-Joshi N. Transient receptor poten-
tial ankyrin 1 receptor activation in vitro and in vivo by pro-tussive
agents: GRC 17536 as a promising anti-tussive therapeutic. PLoS ONE. 2014;9: e97005. 45. Plevkova J, Biringerova Z, Gavliakova S. Thermo sensitive TRPM8 chan-
nel and its role in cold induced airway symptoms. Open J Mol Integr
Physiol. 2012;2:26. 46. Xu H, Blair NT, Clapham DE. Camphor activates and strongly desensi-
tizes the transient receptor potential vanilloid subtype 1 channel in a
vanilloid-independent mechanism. J Neurosci. 2005;25:8924–37. 67. Rhyu M-R, Kim Y, Lyall V. Interactions between chemesthesis and taste:
role of TRPA1 and TRPV1. Int J Mol Sci. 2021;22:3360. 47. Macpherson LJ, Geierstanger BH, Viswanath V, Bandell M, Eid SR, Hwang
S, Patapoutian A. The pungency of garlic: activation of TRPA1 and TRPV1
in response to allicin. Curr Biol. 2005;15:929–34. 68. Long L, Yao H, Tian J, Luo W, Yu X, Yi F, Chen Q, Xie J, Zhong N, Chung
KF, Lai K. Heterogeneity of cough hypersensitivity mediated by TRPV1 Stinson et al. Respiratory Research (2023) 24:45 Page 15 of 17 Page 15 of 17 and TRPA1 in patients with chronic refractory cough. Respir Res. 2019;20:112.f and TRPA1 in patients with chronic refractory cough. Respir Res. 2019;20:112. 91. Voets T, Droogmans G, Wissenbach U, Janssens A, Flockerzi V, Nilius
B. The principle of temperature-dependent gating in cold- and heat-
sensitive TRP channels. Nature. 2004;430:748–54. 69. Moher D, Liberati A, Tetzlaff J, Altman D. Preferred reporting items for
systematic reviews and meta-analyses: the PRISMA statement. Ann
Intern Med. 2009;151:264–9. 92. Macpherson LJ, Hwang SW, Miyamoto T, Dubin AE, Patapoutian A, Story
GM. More than cool: promiscuous relationships of menthol and other
sensory compounds. Mol Cell Neurosci. 2006;32:335–43. 70. Eccles R. Menthol and related cooling compounds. J Pharm Pharmacol. 1994;46:618–30. 93. Karashima Y, Damann N, Prenen J, Talavera K, Segal A, Voets T, Nilius B. Bimodal action of menthol on the transient receptor potential channel
TRPA1. J Neurosci. 2007;27:9874–84. 71. Galeotti N, Di Cesare ML, Mazzanti G, Bartolini A, Ghelardini C. Menthol:
a natural analgesic compound. Neurosci Lett. 2002;322:145–8. a natural analgesic compound. Neurosci Lett. 2002;322:145– g
72. Patel T, Ishiuji Y, Yosipovitch G. Menthol: A refreshing look at this ancient
compound. J Am Acad Dermatol. 2007;57:873–8. 94. Competing interests Mergler S, Mertens C, Valtink M, Reinach PS, Castelo Székely V, Slavi N,
Garreis F, Abdelmessih S, Türker E, Fels G, Pleyer U. Functional signifi-
cance of thermosensitive transient receptor potential melastatin chan-
nel 8 (TRPM8) expression in immortalized human corneal endothelial
cells. Exp Eye Res. 2013;116:337–49. 100. Andrè E, Gatti R, Trevisani M, Preti D, Baraldi P, Patacchini R, Geppetti P. Transient receptor potential ankyrin receptor 1 is a novel target for pro-
tussive agents. Br J Pharmacol. 2009;158:1621–8. 101. Marsakova L, Touska F, Krusek J, Vlachova V. Pore helix domain is critical
to camphor sensitivity of transient receptor potential vanilloid 1 chan-
nel. Anesthesiology. 2012;116:903–17. 80. Plevkova J, Kollarik M, Poliacek I, Brozmanova M, Surdenikova L, Tatar
M, Mori N, Canning BJ. The role of trigeminal nasal TRPM8-expressing
afferent neurons in the antitussive effects of menthol. J Appl Physiol. 2013;115:268–74. 102. Selescu T, Ciobanu AC, Dobre C, Reid G, Babes A. Camphor activates
and sensitizes transient receptor potential melastatin 8 (TRPM8) to
cooling and icilin. Chem Senses. 2013;38:563–75. 81. Bödding M, Wissenbach U, Flockerzi V. Characterisation of TRPM8 as a
pharmacophore receptor. Cell Calcium. 2007;42:618–28. 103. Batish DR, Singh HP, Kohli RK, Kaur S. Eucalyptus essential oil as a natural
pesticide. For Ecol Manage. 2008;256:2166–74. 82. Weil A, Moore SE, Waite NJ, Randall A, Gunthorpe MJ. Conservation
of functional and pharmacological properties in the distantly related
temperature sensors TRPV1 and TRPM8. Mol Pharmacol. 2005;68:518. 104. Dhakad AK, Pandey VV, Beg S, Rawat JM, Singh A. Biological, medicinal
and toxicological significance of Eucalyptus leaf essential oil: a review. J
Sci Food Agric. 2018;98:833–48. 83. Mahieu F, Owsianik G, Verbert L, Janssens A, De Smedt H, Nilius B, Voets
T. TRPM8-independent menthol-induced Ca2+ release from endoplas-
mic reticulum and golgi. J Biol Chem. 2007;282:3325–36. 105. Darben T, Cominos B, Lee C. Topical eucalyptus oil poisoning. Australas
J Dermatol. 1998;39:265–7. 84. McKemy DD, Neuhausser WM, Julius D. Identification of a cold receptor
reveals a general role for TRP channels in thermosensation. Nature. 2002;416:52–8. 106. Takaishi M, Fujita F, Uchida K, Yamamoto S, Sawada Shimizu M, Hatai
Uotsu C, Shimizu M, Tominaga M. 1,8-cineole, a TRPM8 agonist, is a
novel natural antagonist of human TRPA1. Mol Pain. 2012;8:86–86. 85. Willis DN, Liu B, Ha MA, Jordt S-E, Morris JB. Menthol attenuates respira-
tory irritation responses to multiple cigarette smoke irritants. FASEB J. 2011;25:4434–44. 107. Burrow A, Eccles R, Jones AS. 113. Selvarani V, James H. The activity of cedar leaf oil vapor against respira-
tory viruses: practical applications. In.; 2013 Competing interests Morice AH, Marshall AE, Higgins KS, Grattan TJ. Effect of inhaled
menthol on citric acid induced cough in normal subjects. Thorax. 1994;49:1024–6. 143. Topp R, Winchester LJ, Schilero J, Jacks D. Effect of topical menthol on
ipsilateral and contralateral superficial blood flow following a bout
of maximum voluntary muscle contraction. Int J Sports Phys Ther. 2011;6:83–91. 121. Buday T, Brozmanova M, Biringerova Z, Gavliakova S, Poliacek I, Calko-
vsky V, Shetthalli MV, Plevkova J. Modulation of cough response by
sensory inputs from the nose—role of trigeminal TRPA1 versus TRPM8
channels. Cough. 2012;8:11. 144. Stefanelli L, Lockyer EJ, Collins BW, Snow NJ, Crocker J, Kent C, Power KE,
Button DC. Delayed-onset muscle soreness and topical analgesic alter
corticospinal excitability of the biceps brachii. Med Sci Sports Exerc. 2019;51:2344–56. 122. Johnson D, Mead R, Kennelty K, Hahn D. Menthol cough drops: cause
for concern? J Am Board Family Med. 2018;31:183. 123. Gourd K. John Mudge. Lancet Respir Med. 2016;4:16–7. 145. Whalen A, Farrell K, Roberts S, Smith H, Behm DG. Topical analgesic
improved or maintained ballistic hip flexion range of motion with
treated and untreated legs. J Sports Sci Med. 2019;18:552–8. 124. Kumar N, Nepali K, Sapra S, Bijjem KRV, Kumar R, Suri OP, Dhar KL. Effect
of nitrogen insertion on the antitussive properties of menthol and
camphor. Med Chem Res. 2012;21:531–7.f 146. Aditi S, Lalit S, Rohit G. A review on himalayan pine species: ethnop-
harmacological. Phytochem Pharmacolog Aspects Pharmacogn J. 2018;10:611. 125. Laude EA, Morice AH, Grattan TJ. The antitussive effects of menthol,
camphor and cineole in conscious Guinea-pigs. Pulm Pharmacol. 1994;7:179–84. 147. Barbosa LCA, Filomeno CA, Teixeira RR. Chemical variability and
biological activities of Eucalyptus spp. essential oils. Molecules (Basel,
Switzerland). 2016;21:1671. 126. Cohen HA, Rozen J, Kristal H, Laks Y, Berkovitch M, Uziel Y, Kozer E,
Pomeranz A, Efrat H. Effect of honey on nocturnal cough and sleep
quality: a double-blind, randomized, placebo-controlled study. Pediat-
rics. 2012;130:465. 148. Chelliah DA. Biological activity prediction of an ethno medicinal plant
cinnamomum camphora through bio-informatics. Ethnobotanical
Leaflets. 2008;2008:22. 127. Gavliakova S, Biringerova Z, Buday T, Brozmanova M, Calkovsky V,
Poliacek I, Plevkova J. Antitussive effects of nasal thymol challenges in
healthy volunteers. Respir Physiol Neurobiol. 2013;187:104–7. l
149. Farco J, Grundmann O. Menthol–pharmacology of an important natu-
rally medicinal “cool.” Mini Rev Med Chem. 2013;13:124–31. y
p
y
128. Begrow F, Engelbertz J, Feistel B, Lehnfeld R, Bauer K, Verspohl EJ. Competing interests The effects of camphor, eucalyptus and
menthol vapour on nasal resistance to airflow and nasal sensation. Acta
Otolaryngol. 1983;96:157–61. 86. Zhou Y, Sun B, Li Q, Luo P, Dong L, Rong W. Sensitivity of bronchopul-
monary receptors to cold and heat mediated by transient receptor
potential cation channel subtypes in an ex vivo rat lung preparation. Respir Physiol Neurobiol. 2011;177:327–32. 108. Ziment I. Herbal antitussives. Pulm Pharmacol Ther. 2002;15:327–33. 109. Ghosh M, Schepetkin IA, Özek G, Özek T, Khlebnikov AI, Damron DS,
Quinn MT. Essential oils from monarda fistulosa: chemical composition
and activation of transient receptor potential A1 (TRPA1) channels. Molecules (Basel, Switzerland). 2020;25:4873. 87. Paschke M, Tkachenko A, Ackermann K, Hutzler C, Henkler F, Luch A. Activation of the cold-receptor TRPM8 by low levels of menthol in
tobacco products. Toxicol Lett. 2017;271:50–7. 110. Xu Z-H, Wang C, Fujita T, Jiang C-Y, Kumamoto E. Action of thymol on
spontaneous excitatory transmission in adult rat spinal substantia
gelatinosa neurons. Neurosci Lett. 2015;606:94–9. 88. Ban on menthol cigarettes: European Union member states shall pro-
hibit the placing on the market of tobacco products with a characteris-
ing flavour. Tob Prev Cessat. 2020; 6:40. 111. Lee SP, Buber MT, Yang Q, Cerne R, Cortés RY, Sprous DG, Bryant RW. Thymol and related alkyl phenols activate the hTRPA1 channel. Br J
Pharmacol. 2008;153:1739–49. 89. Levy DT, Meza R, Yuan Z, Li Y, Cadham C, Sanchez-Romero LM, Travis N,
Knoll M, Liber AC, Mistry R, et al. Public health impact of a US ban on
menthol in cigarettes and cigars: a simulation study. Tobacco Control. 2021:tobaccocontrol-2021-056604. 112. Wang W, Wang H, Zhao Z, Huang X, Xiong H, Mei Z. Thymol activates
TRPM8-mediated Ca2+ influx for its antipruritic effects and alleviates
inflammatory response in imiquimod-induced mice. Toxicol Appl
Pharmacol. 2020;407: 115247. 90. Xing H, Ling J, Chen M, Gu JG. Chemical and cold sensitivity of two
distinct populations of TRPM8-expressing somatosensory neurons. J
Neurophysiol. 2006;95:1221–30. 113. Selvarani V, James H. The activity of cedar leaf oil vapor against respira-
tory viruses: practical applications. In.; 2013 Page 16 of 17 Page 16 of 17 Stinson et al. Respiratory Research (2023) 24:45 Stinson et al. Respiratory Research (2023) 24:45 114. Matulyte I, Jekabsone A, Jankauskaite L, Zavistanaviciute P, Sakiene V,
Bartkiene E, Ruzauskas M, Kopustinskiene DM, Santini A, Bernatoniene
J. Competing interests Impact of thymol in thyme extracts on their antispasmodic action and
ciliary clearance. Planta Med. 2010;76:311–8. 150. Hudson J, Kuo M, Vimalanathan S. The antimicrobial properties of cedar
leaf (Thuja plicata) oil; a safe and efficient decontamination agent for
buildings. Int J Environ Res Public Health. 2011;8:4477–87. y
129. Liu SC, Lu HH, Cheng LH, Chu YH, Lee FP, Wu CC, Wang HW. Identifica-
tion of the cold receptor TRPM8 in the nasal mucosa. Am J Rhinol
Allergy. 2015;29:e112-116. 151. Jasuja ND, Sharma S, Choudhary J, Joshi SC. Essential oil and important
activities of Thuja orientalis and Thuja occidentalis. J Essential Oil Bearing
Plants. 2015;18:931–49. gy
130. Pereira EJ, Sim L, Driver H, Parker C, Fitzpatrick M. The effect of inhaled
menthol on upper airway resistance in humans: a randomized con-
trolled crossover study. Can Respir J. 2013;20:e1–4. 152. Kamatou GPP, Vermaak I, Viljoen AM, Lawrence BM. Menthol: a simple
monoterpene with remarkable biological properties. Phytochemistry. 2013;96:15–25. 131. Nishino T, Tagaito Y, Sakurai Y. Nasal inhalation of l-menthol reduces
respiratory discomfort associated with loaded breathing. Am J Respir
Crit Care Med. 1997;156:309–13. 153. Kearns GL, Chumpitazi BP, Abdel-Rahman SM, Garg U, Shulman RJ. Systemic exposure to menthol following administration of peppermint
oil to paediatric patients. BMJ Open. 2015;5: e008375. 132. Eccles R, Lancashire B, Tolley NS. The effect of aromatics on inspiratory
and expiratory nasal resistance to airflow. Clin Otolaryngol Allied Sci. 1987;12:11–4. 154. Mercier B, Prost J, Prost M. The essential oil of turpentine and its major
volatile fraction (alpha- and beta-pinenes): a review. Int J Occup Med
Environ Health. 2009;22:331–42. 133. Neher A, Gstöttner M, Thaurer M, Augustijns P, Reinelt M, Schobers-
berger W. Influence of essential and fatty oils on ciliary beat frequency
of human nasal epithelial cells. Am J Rhinol. 2008;22:130–4. 155. Pan M, Lei Q, Zhang H. Prediction and confirmation of active ingre-
dients in Eucalyptus globulus Labill leaves. Ind Crops Prod. 2020;154:
112631. 134. Lai Y, Dilidaer D, Chen B, Xu G, Shi J, Lee RJ, Cohen NA. In vitro studies
of a distillate of rectified essential oils on sinonasal components of
mucociliary clearance. Am J Rhinol Allergy. 2014;28:244–8. 156. Salehi B, Mishra AP, Shukla I, Sharifi-Rad M, Contreras MDM, Segura-
Carretero A, Fathi H, Nasrabadi NN, Kobarfard F, Sharifi-Rad J. Thymol,
thyme, and other plant sources: health and potential uses. Phytothera
Res. 2018;32:1688–706. 135. Competing interests The essential oil and hydrolats from myristica fragrans seeds with
magnesium aluminometasilicate as excipient: antioxidant, antibacterial,
and anti-inflammatory activity. Foods (Basel, Switzerland). 2020;9:37. actigraphically assessed aspects of sleep in common cold patients. J
Respir Dis. 2017;07:83–101. actigraphically assessed aspects of sleep in common cold patients. J
Respir Dis. 2017;07:83–101. 137. Oshima T, Ito M. Sedative effects of l-menthol, d-camphor, phenylethyl
alcohol, and geraniol. J Nat Med. 2021;75:319–25. 138. Watt EE, Betts BA, Kotey FO, Humbert DJ, Griffith TN, Kelly EW, Veneskey
KC, Gill N, Rowan KC, Jenkins A, Hall AC. Menthol shares general
anesthetic activity and sites of action on the GABAA receptor with the
intravenous agent, propofol. Eur J Pharmacol. 2008;590:120–6. l
y
y
115. Olajide OA, Makinde JM, Awe SO. Evaluation of the pharmacological
properties of nutmeg oil in rats and mice. Pharm Biol. 2000;38:385–90. 116. Khan NR, Khan GM, Wahab A, Khan AR, Hussain A, Nawaz A, Akhlaq M. Formulation, and physical, in vitro and ex vivo evaluation of trans-
dermal ibuprofen hydrogels containing turpentine oil as penetration
enhancer. Die Pharmazie Int J Pharmaceut Sci. 2011;66:849–52. 139. Kunihiro K, Myoda T, Tajima N, Gotoh K, Kaneshima T, Someya T, Toeda
K, Fujimori T, Nishizawa M. Volatile components of the essential oil of
artemisia montana and their sedative effects. J Oleo Sci. 2017;66:843–9. 140. García DA, Bujons J, Vale C, Suñol C. Allosteric positive interaction of
thymol with the GABAA receptor in primary cultures of mouse cortical
neurons. Neuropharmacology. 2006;50:25–35. 117. Zhang Y-L, Liu Y-G, Li Q, Wang X-D, Zheng X-B, Yang B-L, Wan B, Ma J-M,
Liu Z-X. 1,8-cineole decreases neuropathic pain probably via a mecha-
nism mediating P2X3 receptor in the dorsal root ganglion. Neurochem
Int. 2018;121:69–74. 141. García DA, Vendrell I, Galofré M, Suñol C. GABA released from cultured
cortical neurons influences the modulation of t-[35S]butylbicyclophos-
phorothionate binding at the GABAA receptor: effects of thymol. Eur J
Pharmacol. 2008;600:26–31. 118. Millqvist E, Ternesten-Hasséus E, Bende M. Inhalation of menthol
reduces capsaicin cough sensitivity and influences inspiratory flows in
chronic cough. Respir Med. 2013;107:433–8. g
p
119. Wise PM, Breslin PA, Dalton P. Sweet taste and menthol increase cough
reflex thresholds. Pulm Pharmacol Ther. 2012;25:236–41. 142. Behm DG, Herat N, Power GMJ, Brosky JA, Page P, Alizadeh S. Menthol-
based topical analgesic induces similar upper and lower body
pain pressure threshold values: a randomized trial. J Sport Rehabil. 2021;31:1–7. 120. Competing interests Kilic K, Sakat MS, Yildirim S, Kandemir FM, Gozeler MS, Dortbudak MB,
Kucukler S. The amendatory effect of hesperidin and thymol in allergic
rhinitis: an ovalbumin-induced rat model. Eur Arch Otorhinolaryngol. 2019;276:407–15. 157. Silinski MAR, Licause J, Uenoyama T, Blake JC, Fernando RA, Robinson
VG, Waidyanatha S. Development and validation of an analytical
method for quantitation of alpha-pinene in rodent blood and mam-
mary gland by headspace GC-MS. J Anal Toxicol. 2020;46:128. 136. Santhi N, Ramsey D, Phillipson G, Hull D, Revell V, Dijk D-J. Efficacy of a
topical aromatic rub (Vicks VapoRub ® ) on effects on self-reported and Page 17 of 17 Stinson et al. Respiratory Research (2023) 24:45 Stinson et al. Respiratory Research (2023) 24:45 158. Tümen İ, Akkol EK, Taştan H, Süntar I, Kurtca M. Research on the
antioxidant, wound healing, and anti-inflammatory activities and the
phytochemical composition of maritime pine (Pinus pinaster Ait). J
Ethnopharmacol. 2018;211:235–46. 159. Vriens J, Nilius B, Vennekens R. Herbal compounds and toxins modulat-
ing TRP channels. Curr Neuropharmacol. 2008;6:79–96. 160. Woolf A. Essential oil poisoning. J Toxicol Clin Toxicol. 1999;37:721–7. 161. Mälkiä A, Madrid R, Meseguer V, de la Peña E, Valero M, Belmonte C,
Viana F. Bidirectional shifts of TRPM8 channel gating by temperature
and chemical agents modulate the cold sensitivity of mammalian
thermoreceptors. J Physiol. 2007;581:155–74. Publisher’s Note
S
i
N
i Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations. •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from:
|
https://openalex.org/W2947447797
|
https://europepmc.org/articles/pmc6538164?pdf=render
|
English
| null |
Spatial-temporal characteristics and causes of changes to the county-level administrative toponyms cultural landscape in the eastern plains of China
|
PloS one
| 2,019
|
cc-by
| 11,069
|
RESEARCH ARTICLE Yingying Wang1,2, Yingjie WangID1,2*, Lei Fang1,2, Shengrui Zhang1,2, Tongyan Zhang1,2,
Daichao Li3, Dazhuan Ge4 Yingying Wang1,2, Yingjie WangID1,2*, Lei Fang1,2, Shengrui Zhang1,2, Tongyan Zhang1,2,
Daichao Li3, Dazhuan Ge4 1 State Key Laboratory of Resources and Environmental Information System, Institute of Geographic
Sciences and Natural Resources Research, CAS, Beijing, China, 2 University of Chinese Academy of
Sciences, Beijing, China, 3 Spatial Information Research Center of Fujian, Fuzhou University, Fuzhou,
China, 4 College of Geography, Nanjing Normal University, Nanjing, China a1111111111
a1111111111
a1111111111
a1111111111
a1111111111 * wangyj@lreis.ac.cn Editor: Tzai-Hung Wen, National Taiwan University,
TAIWAN Editor: Tzai-Hung Wen, National Taiwan University,
TAIWAN Copyright: © 2019 Wang et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited. Data Availability Statement: All relevant data are
within the paper and its Supporting Information
files. OPEN ACCESS As part of the cultural landscape, administrative toponyms do not only reflect natural and
sociocultural phenomena, but also help with related management and naming work. Histori-
cally, county-level administrative districts have been stable and basic administrative regions
in China, playing a role in the country’s management. We explore the spatio-temporal evolu-
tionary characteristics of the county-level administrative toponyms cultural landscape in Chi-
na’s eastern plains areas. A Geographical Information System (GIS) analysis, Geo-
Informatic Tupu, Kernel Density Estimation, and correlation coefficients were conducted. We constructed a GIS database of county-level administrative toponyms from the Sui
dynasty onward using the Northeast China, North China, and Yangtze Plains as examples. We then summarized the spatio-temporal evolutionary characteristics of the county-level
administrative toponyms cultural landscape in China’s eastern plains areas. The results indi-
cate that (1) the number of toponyms has roughly increased over time; (2) toponym densi-
ties on the three plains are higher than the national average in the corresponding timeframe
since the Sui; and (3) county-level administrative toponyms related to mountains and hydro-
logical features accounted for more than 30% of the total in 2010. However, the percentage
of county-level administrative toponyms related to natural factors on the three plains has
decreased since the Sui. To explore the factors influencing this spatio-temporal evolution,
we analyzed the correlations between the toponyms and natural factors and human/social
factors. The correlation degree between toponym density and population density is the high-
est, and that between toponym density and Digital Elevation Model (DEM) the lowest. Tem-
perature changes were important in toponym changes, and population changes have
influenced toponym changes over the last 400 years in China. Citation: Wang Y, Wang Y, Fang L, Zhang S, Zhang
T, Li D, et al. (2019) Spatial-temporal
characteristics and causes of changes to the
county-level administrative toponyms cultural
landscape in the eastern plains of China. PLoS ONE
14(5): e0217381. https://doi.org/10.1371/journal. pone.0217381 Citation: Wang Y, Wang Y, Fang L, Zhang S, Zhang
T, Li D, et al. (2019) Spatial-temporal
characteristics and causes of changes to the
county-level administrative toponyms cultural
landscape in the eastern plains of China. PLoS ONE
14(5): e0217381. https://doi.org/10.1371/journal. pone.0217381 Introduction A place begins to exist when people assign it a name and meaning, differentiating it from the
larger, undifferentiated space [1]. Thus, toponyms (or place names) are names given to places
by people. Toponyms are also defined as names for natural and manmade geographic features
assigned to particular spaces [2]. As a cultural landscape, toponyms are the preservation and
mark of regional culture on the surface level, and directly reveal the spatial distribution of lan-
guages and ethnic origins. The term “toponyms cultural landscape” refers to the group charac-
teristics of toponyms formed by consistent factors in a certain area [3]. Moreover, the
toponyms cultural landscape does not only reflect the characteristics of the historical and cur-
rent natural environment, but also records information such as major political changes, the
prosperity or decline of a nation, ethnic migration, religious beliefs, and military activities [4,
5]. Toponyms are frequently used to label, identify, and locate sites in space, and are subjective
explanations by local inhabitants from the time of naming [6]. In China, studies concerning
toponyms at the earliest stage appeared alongside academic works on China’s ancient geogra-
phy, such as Shang ShuYu Gong, Classic of Mountains and Seas (Shan Hai Jing), Geographical
Records on the History of the Han Dynasty, and Commentary on the Water Classic (Shui Jing
Zhu), which all recorded important information on toponyms [7]. Because of their increasing
significance, studies on toponyms attracted the attention of the government and academia. During the Republic of China (ROC) (1912–1949), The Great Dictionary of Chinese Toponyms
and Historical Records Test Names were representative works of Chinese toponymy. Following
the founding of the People’s Republic of China (PRC) in 1949, the periodical China Toponyms
was published in 1984, and featured many scholars’ academic achievements regarding topo-
nyms [7, 8]. After the 1st Toponyms Census of China in the 1980s, many works were published
that interpreted toponyms from the perspectives of linguistics, historiography, culturology,
and geography [9–15]. That established the theoretical foundation for the study of toponyms. th Toponymy developed in the occidental countries in the late 19th century as an independent
discipline. For example, the United States established the Board on Geographic Names in 1890
[16]; Henry Gannett introduced the origin of toponyms of the United States in 1905 [17];
George Rippey Stewart published the book entitled Names on the land: a historical account of
placenaming in USA in 1982 [18]. Data Availability Statement: All relevant data are
within the paper and its Supporting Information
files. Funding: This work was supported by Ministry of
Civil Affairs of the People’s Republic of China
Y77L0200AJ to YW. The funders had no role in
study design, data collection and analysis, decision
to publish, or preparation of the manuscript. Competing interests: The authors have declared
that no competing interests exist. 1 / 24 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 County-level administrative toponyms cultural landscape PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 Case areas As linguistic fossils, toponyms have a mark function for the changes of natural environment
and social environment in different historical periods. Administrative toponyms are an inte-
gral part of the toponyms cultural landscape, and an important tool for state administration. Further research on administrative toponyms can reveal natural and sociocultural phenomena,
and be used for managing and naming places. County-level administrative districts are the most stable and basic administrative regions in
China’s history, and have played a key role in the management of the country. As the generic
part of administrative toponyms, “county” dates to China’s Spring and Autumn Period (770–
476 BC). Since the nationwide implementation of the prefecture and county system during the
Qin dynasty (221–207 BC), “county” became widespread throughout the country [14]. According to preliminary statistics from the Dictionary of Chinese Historical Toponyms [30]
and The New Century Chinese City View [31], less than 3% of county-level administrative
regions still use their names from the Qin and Han dynasties. As shown in Fig 1, 795 county-
level administrative toponyms are more than 1,000 years old, of which 331 are more than
1,500 years old and 215 older than 2,000 years. Furthermore, the Hu Line divides China into
an eastern and western part. Nearly 98% of county-level administrative toponyms are located
in the eastern part, with a concentrated distribution in the plains areas. The dots and polygons
data of administrative divisions of China at a scale of 1:1,000,000 (Fig 1) were obtained from
National Catalogue Service For Geographic Information [32]. In addition, the plains areas have the most intimate human-land relationship, and because
humans have recognized and reconstructed the natural environment, administrative topo-
nyms in these areas account for a sizable proportion of the total. Furthermore, we employed
county-level administrative toponyms located in the Northeast China Plain, North China
Plain, and Yangtze Plain (Fig 2) as our study object for the following reasons. First, the North-
east China Plain, North China Plain, and Yangtze Plain are three great plains located in the
eastern part of China, and have relatively high representativeness. Second, we selected these
three plains as the case study areas because of data availability and processing workload. Case areas Third,
considering that toponyms are in the category of linguistics, we selected the Northeast China
Plain to analyze the influence of minority languages on toponyms to compare it with the
North China Plain and Yangtze Plain, even though the Northeast China Plain developed later. The polygons data of China (Fig 2) were obtained from National Catalogue Service For Geo-
graphic Information [32]. Introduction In the 21st century, modern research on toponyms tends to
focus on quantitative and visualized analyses. For example, the Geographic Information Sys-
tem (GIS) is widely applied in the study of geographical names, and a research method com-
bining cartography with geo-visualization techniques is playing an important role in the
quantification and visualization of toponyms. In general, research has focused on exploring
the correlation between toponyms and landscape changes [19]; using toponyms to interpret
historical land use changes or reconstruct past land use [6, 20]; constructing GIS databases of
ancient and modern toponyms [21]; utilizing toponyms to indicate recent climate changes
[22]; analyzing the spatial distribution characteristics of toponyms based on a GIS approach,
including dialectic toponyms, rural toponyms, and toponyms in ethnic minority languages
[23–26]; and using GIS technology to study the cultural landscape of toponyms [3, 27–29]. Although scholars worldwide have investigated toponym features, few have paid attention
to the characteristics and intrinsic mechanisms of the evolution of the toponyms cultural land-
scape at different spatial and temporal scales, and few researches were visualized using maps to
intuitively and accurately clarify this evolution. Thus, this study focuses on the changes to the
county-level administrative toponyms cultural landscape in China’s eastern plains areas to
explore their influencing factors. This study provides theoretical and practical operational
guidance for rationally optimizing administrative divisions and toponym management, which
could help protect the cultural heritage of toponyms in China. PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 2 / 24 County-level administrative toponyms cultural landscape PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 Data sources and preparation Other data were sorted from the Dictionary of Chi-
nese Historical Toponyms [30], The New Century Chinese City View [31], The Changes of
Administrative Divisions in Chinese History [37], Encyclopedic Dictionary of Ancient and Mod-
ern Chinese Geographical Names [38] and China's Toponyms Anecdotes Dictionary [39]. Data sources and preparation For the purposes of this study, the historical toponyms data were acquired from The Historical
Atlas of China [33]. We vectorized the paper version of the atlas. The dots and polygons data
of administrative divisions of China at a scale of 1:1,000,000 were obtained from National Cat-
alogue Service For Geographic Information [32]. Moreover, to explore the influencing factors
of the spatio-temporal evolutionary characteristics of these toponyms, we collected GDP data
in 2010 at the county level from the RESDC [34]and population data at the county level from
the 6th Census of China (2010) [35]. In addition, DEM data at a resolution of 30 m and grid
data of China’s population and GDP in 2010 (at a 1-km gridded resolution) were provided by
the RESDC [34]. These grid data were used to calculate the correlation coefficients, as
explained in the following sections. For historical population data, we referred to statistical
data from the National Earth System Science Data Sharing Infrastructure, National Science & PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 3 / 24 County-level administrative toponyms cultural landscape Fig 1. Distribution map of county-level administrative toponyms with the time of being named from 1,000 years ago. The dots
represent data on county-level administrative toponyms with naming time from 1,000 years ago. The green dots represent county-
level administrative toponyms being named for 1,000–1,499 years, blue dots those being named for 1,500–1,999 years, and red dots
those being named for 2,000 years ago. Fig 1. Distribution map of county-level administrative toponyms with the time of being named from 1,000 years ago. The dots
represent data on county-level administrative toponyms with naming time from 1,000 years ago. The green dots represent county-
level administrative toponyms being named for 1,000–1,499 years, blue dots those being named for 1,500–1,999 years, and red dots
those being named for 2 000 years ago Fig 1. Distribution map of county-level administrative toponyms with the time of being named from 1,000 years ago. The dots
represent data on county-level administrative toponyms with naming time from 1,000 years ago. The green dots represent county-
level administrative toponyms being named for 1,000–1,499 years, blue dots those being named for 1,500–1,999 years, and red dots
those being named for 2,000 years ago. https://doi.org/10.1371/journal.pone.0217381.g001 https://doi.org/10.1371/journal.pone.0217381.g001 Technology Infrastructure of China [36]. PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 Research methods Kernel density estimation method. The Kernel Density Estimation (KDE) method uses a
moving cell (equivalent to a window) to estimate the density of a point or line pattern. It is
generally defined as x1, x2. . ., where xn are independent and identically distributed samples
extracted from the distribution density function f, and f(x) is the value estimated at given point
x. Usually, researchers adopt the Rosenblatt-Parzen kernel estimation [40]. fðxÞ ¼ 1
nh
X
n
i¼1
k dðx; xiÞ
h
ð1Þ ð1Þ In this expression, n denotes the number of toponyms in the range scale, while k() is the PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 4 / 24 County-level administrative toponyms cultural landscape Fig 2. Map of the case study areas. https://doi.org/10.1371/journal.pone.0217381.g002 Fig 2. Map of the case study areas. https://doi.org/10.1371/journal.pone.0217381.g002 kernel density function of the toponyms, h the distance threshold, and d(x, xi) the Euclidean
distance between the estimated point x and sample xi. The KDE method uses the distance attenuation bandwidth to represent an object’s spatial
influence domain, which is suitable for situations that consider the influence of regions [41]. Therefore, we used the KDE method to study changes in the toponym density of the adminis-
trative toponyms cultural landscape in China’s eastern plains areas since the Sui dynasty. Correlation coefficients. We selected data on DEM, GDP, and the population spatial dis-
tribution (at a 1-km gridded resolution) in 2010 as experimental data to calculate the correla-
tion coefficients between toponym density and natural, economic, and demographic factors. Research methods RDEM ¼
X
n
i¼1
ðxi xÞðai aÞ
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
X
n
i¼1
ðxi xÞ
2X
n
i¼1
ðai aÞ
2
s
ð2Þ
RGDP ¼
X
n
i¼1
ðxi xÞðbi bÞ
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
X
n
i¼1
ðxi xÞ
2X
n
i¼1
ðbi bÞ
2
s
ð3Þ
RPOP ¼
X
n
i¼1
ðxi xÞðci cÞ
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
X
n
i¼1
ðxi xÞ
2X
n
i¼1
ðci cÞ
2
s
ð4Þ RDEM ¼
X
n
i¼1
ðxi xÞðai aÞ
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
X
n
i¼1
ðxi xÞ
2X
n
i¼1
ðai aÞ
2
s
ð2Þ
RGDP ¼
X
n
i¼1
ðxi xÞðbi bÞ
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
X
n
i¼1
ðxi xÞ
2X
n
i¼1
ðbi bÞ
2
s
ð3Þ
RPOP ¼
X
n
i¼1
ðxi xÞðci cÞ
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
X
n
i¼1
ðxi xÞ
2X
n
i¼1
ðci cÞ
2
s
ð4Þ
RDEM, RGDP, and RPOP respectively refer to the correlation coefficient between x (x denotes ð2Þ ð3Þ ð4Þ RDEM, RGDP, and RPOP respectively refer to the correlation coefficient between x (x denotes RDEM, RGDP, and RPOP respectively refer to the correlation coefficient between x (x denotes 5 / 24 County-level administrative toponyms cultural landscape Fig 3. Generation and transmission mode of Geo-Informatic Tupu. Fig 3. Generation and transmission mode of Geo-Informatic Tupu. https://doi.org/10.1371/journal.pone.0217381.g003 https://doi.org/10.1371/journal.pone.0217381.g003 toponym density) and a (a denotes elevation), b (b denotes GDP), and c (c denotes population
density). Furthermore, xi, ai, bi, and ci respectively denote the toponym density, elevation,
GDP, and population density of raster i. Geo-Informatic Tupu. Geo-Informatic Tupu is a group of maps, curves, graphs, or tables
arranged by certain indicators gradient rules or systematization laws, it utilizes the methods
(such as GIS, RS, mathematical models, etc) to display spatial dynamic changes [42–47]. Although the theory and method of Geo-Informatic Tupu have been successfully applied to
many branches of geoscience, it has not been applied to the study of spatio-temporal evolution
of administrative toponyms or administrative divisions. In this paper, we used the generated
maps and adopt the event-based state amendments model to establish the spatio-temporal
database of administrative toponyms, then selected various indicators (such as toponym den-
sity, toponym naming time or life cycle, naming type, etc) to derive a series of spatio-temporal
composite Geo-Informatic Tupu according to time sequence. PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 Analysis of the spatio-temporal evolutionary characteristics of the county-
level administrative toponyms cultural landscape Spatio-temporal evolutionary features of the county-level administrative toponyms cul-
tural landscape. The statistical years are in keeping with The Historical Atlas of China [33]
from Sui to Qing dynasties, used for the ROC is 1925 AD, and 2010 for the PRC. As shown in Fig 4, the number of toponyms in the Northeast China Plain increased rapidly
from the Qing dynasty onward, because of its role as the dynasty’s birthplace. Furthermore,
the core of toponym densities gradually spread from south to north, finally causing the accu-
mulation of areas at the intersections of the Lower Liao, Songhua, and Nen Rivers. Different
types of county-level administrative toponyms are discussed in the next section. Fig 5 shows that the northwestern part of the North China Plain (i.e., the regions between
the northern part of the Yellow River and eastern part of Taihang Mountain) always had a rela-
tively high toponym density, except during the Yuan dynasty. Because ancient China moved
its economic center from north to south during the Song dynasty, thereafter, especially during
the Yuan, the density of toponyms gradually declined and the core area moved southward in
the North China Plain. However, the Ming dynasty had a higher toponym density than the
Yuan for complex reasons including the following. In terms of economic and demographic
factors, the population and economic recovery increased during the Ming. Regarding political
factors, the Ming dynasty’s Emperor Zhu Di moved the capital from Nanjing to Beijing, trig-
gering the economic resurgence of the North China Plain. The density of toponyms did not
change much after the Ming and Qing, because the Qing used the same administrative topo-
nym system as the Ming. The increasing prosperity and development of the maritime Silk Road during the Song and
Yuan lead to rapid economic development in the southern part of China. In addition, Liujia-
gang Harbor, located in the Yangtze Plain, was the starting point of Zheng He’s voyage during
the Ming. As shown in Fig 6, the center of county-level administrative toponyms began shift-
ing eastward in the Yangtze Plain from the Song and Yuan. Today, urban agglomeration in the
Yangtze River Delta, which is located in the Yangtze Plain, has lead it to become the most
developed and urbanized region of China. Furthermore, the Yangtze River Delta has the dens-
est concentration of county-level administrative toponyms in the Yangtze Plain. Research methods Fig 3 shows the generation
and transmission mode of Geo-Informatic Tupu [48]. In this paper, we adopted the spatio-
temporal data model on account of shapefile and base state with amendments model [49] to
establish the spatio-temporal database, the spatio-temporal evolution Tupu of county-level PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 6 / 24 County-level administrative toponyms cultural landscape administrative toponyms densities in the eastern plains of China since the Sui was generated
by using the KDE method, and extracted manually the naming time, naming basis and naming
type information of county-level administrative toponyms since the Sui by relying on the
expert knowledge from the Dictionary of Chinese Historical Toponyms [30], The New Century
Chinese City View [31], The Changes of Administrative Divisions in Chinese History [37], Ency-
clopedic Dictionary of Ancient and Modern Chinese Geographical Names [38] and China's Top-
onyms Anecdotes Dictionary [39]. In view of the advantage of Geo-Informatic Tupu in
analyzing and expressing the spatio-temporal changes, it is suitable to the study of spatio-tem-
poral evolution characteristics of administrative toponyms in this paper. PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 https://doi.org/10.1371/journal.pone.0217381.g004 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 Analysis of the spatio-temporal evolutionary characteristics of the county-
level administrative toponyms cultural landscape As seen in Figs 4–6 and Table 1, the number of county-level administrative toponyms in
the Northeast China Plain, North China Plain, and Yangtze Plain fluctuates based on the vicis-
situdes of dynasties. However, the trend in the number of toponyms is roughly increasing,
mainly because new counties are being separated from old ones over time. Therefore, the PRC
era has the most toponyms and the highest toponym density in the three plains and nation-
wide. Moreover, the number of toponyms is decreased in the Song and Yuan dynasties and the
reasons are complex, may be relevant to the administrative division system and demographic/ PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 7 / 24 County-level administrative toponyms cultural landscape Fig 4. Spatio-temporal evolution of county-level administrative toponyms in the Northeast China Plain since the Sui
dynasty. Fig 4. Spatio-temporal evolution of county-level administrative toponyms in the Northeast China Plain since the Sui
dynasty https://doi.org/10.1371/journal.pone.0217381.g004 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 8 / 24 County-level administrative toponyms cultural landscape Fig 5. Spatio-temporal evolution of county-level administrative toponyms in the North China Plain since the Sui dynasty. Spatio-temporal evolution of county-level administrative toponyms in the North China Plain since the Sui dynasty. Fig 5. Spatio-temporal evolution of county-level administrative toponyms in the North China Plain https://doi.org/10.1371/journal.pone.0217381.g005 https://doi.org/10.1371/journal.pone.0217381.g005 economic factors. For example, the Yuan dynasty implemented a multi-level compound
administrative division system, and most of the prefecture-level administrative toponyms
decreased to the county-level during the Ming and Qing dynasties [50]. Furthermore, since the Sui dynasty, the toponym densities in the three plains are almost all
greater than the national average density in the corresponding period. In particular, the PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 9 / 24 County-level administrative toponyms cultural landscape Fig 6. Spatio-temporal evolution of county-level administrative toponyms in the Yangtze Plain since the Sui dynasty. https://doi.org/10.1371/journal.pone.0217381.g006 e toponyms in the Yangtze Plain since the Sui dynasty. Fig 6. Spatio-temporal evolution of county-level administrative toponyms in the Yangtze Plain since the Sui dynasty. Fig 6. Spatio-temporal evolution of county-level administrative toponyms in the Yangtze Plain since the Sui dynasty. h
//d i
/10 13 1/j
l
021 381 006 Fig 6. Spatio-temporal evolution of county-level administrative toponyms in the Yangtze Plain since the Sui dynasty. Analysis of the spatio-temporal evolutionary characteristics of the county-
level administrative toponyms cultural landscape https://doi.org/10.1371/journal.pone.0217381.g006 https://doi.org/10.1371/journal.pone.0217381.g006 https://doi.org/10.1371/journal.pone.0217381.g006 toponym densities in the North China Plain and Yangtze Plain are more than triple the
national average during the corresponding periods since the Sui toponym densities in the North China Plain and Yangtze Plain are more than triple the
national average during the corresponding periods since the Sui. Naming time of county-level administrative toponyms. Fig 7 indicates that in 2010, the
naming time of county-level administrative toponyms in the Northeast China Plain were com-
paratively concentrated, but relatively scattered in the North China Plain and Yangtze Plain. However, naming during the PRC period has the highest proportion in all three plains. Specifi-
cally, over 97% of county-level administrative toponyms in the Northeast China Plain were 10 / 24 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 County-level administrative toponyms cultural landscape Table 2. Distribution characteristics of naming types of county-level administrative toponyms in 2010. Naming type
Northeast China Plain
North China Plain
Yangtze Plain
Count
(%)
Count
(%)
Count
(%)
Relevant to mountains
9
7.2
24
6.2
39
13.1
Relevant to hydrological features
30
24.0
94
24.2
102
34.2
Relevant to products
6
4.8
10
2.6
8
2.7
Relevant to terrain
4
3.2
15
3.9
6
2.0
Relevant to orientation
13
10.4
37
9.5
18
6.0
Relevant to blessings
19
15.2
70
18.0
38
12.8
Relevant to ancient relics
22
17.6
53
13.7
44
14.8
Relevant to historical figures and events
7
5.6
53
13.7
26
8.7
Relevant to surnames
3
2.4
8
2.1
5
1.7
Relevant to taboos
0
0
6
1.6
5
1.7
Other forms
12
9.6
18
4.6
7
2.3
Total
125
100
388
100
298
100
https://doi.org/10.1371/journal.pone.0217381.t002 Table 2. Distribution characteristics of naming types of county-level administrative toponyms in 2010. temporal evolution. Based on scholarly research achievements [3, 39, 51], we divided the 811
county-level administrative toponyms into 11 categories according to the characteristics of
their proper names in China’s plains areas (Table 2). For example, the naming type of Dang-
shan County is relevant to mountains, Lishui District relevant to hydrological features, Yandu
District relevant to products, Wanli District relevant to terrain, Ningxiang County relevant to
blessings, Dantu District relevant to historical figures and events, etc. Moreover, the naming
types relevant to mountains, hydrological features, products, and terrain belong to natural fac-
tors, while relevant to orientation, blessings, ancient relics, historical figures and events, sur-
names, taboos, and other forms belong to human/social factors. temporal evolution. Based on scholarly research achievements [3, 39, 51], we divided the 811
county-level administrative toponyms into 11 categories according to the characteristics of
their proper names in China’s plains areas (Table 2). For example, the naming type of Dang-
shan County is relevant to mountains, Lishui District relevant to hydrological features, Yandu
District relevant to products, Wanli District relevant to terrain, Ningxiang County relevant to
blessings, Dantu District relevant to historical figures and events, etc. Moreover, the naming
types relevant to mountains, hydrological features, products, and terrain belong to natural fac-
tors, while relevant to orientation, blessings, ancient relics, historical figures and events, sur-
names, taboos, and other forms belong to human/social factors. County-level administrative toponyms cultural landscape Table 1. Statistical characteristics of county-level administrative toponyms. Period
Number of toponyms
Average toponym density (per 104 km2)
Northeast China Plain
North China Plain
Yangtze Plain
Whole nation
Northeast China Plain
North China Plain
Yangtze Plain
National
average
Sui
1
279
132
1271
0.03
8.51
5.01
0.92
Tang
5
284
119
1575
0.16
8.67
4.51
1.13
Song
8
206
98
1129
0.26
6.29
3.72
0.81
Yuan
5
142
105
1123
0.16
4.33
3.98
0.80
Ming
7
267
178
1427
0.23
8.15
6.75
1.03
Qing
24
270
183
1549
0.79
8.24
6.94
1.11
ROC
76
280
195
2044
2.49
8.54
7.39
1.79
PRC
125
388
298
2856
4.10
11.84
11.30
2.98
https://doi org/10 1371/journal pone 0217381 t001 named since the Qing (125 in total). The Qing is the second-most common, accounting for
more than 34%. Furthermore, 46% of county-level administrative toponyms in the North
China Plain have had their names for more than 1,000 years (388 in total), with most coming
from the Han. In the Yangtze Plain, 41% of county-level administrative toponyms have had
their names for more than 1,000 years (298 in total), with the Wei-Jin making up the second-
largest proportion. The evolution of county-level administrative toponyms can also indicate local landscape
changes. For example, Hekou District in the North China Plain was named in 1984 based on
its location at the estuary of the Yellow River. However, Hekou District is located on what was
once a sea area. People originally settled there from the Qing onward. Qidong City in the
Yangtze Plain was named in 1989, since it was a developing eastern coastal area. Qidong City
was located by the estuary of the Yangtze River, which was also once a sea area. It initially
emerged as a shoal made by the sands of the Yangtze River [39]. Naming basis of county-level administrative toponyms. After analyzing the naming
basis of county-level administrative toponyms, we inferred the factors influencing their spatio- Fig 7. Distribution characteristics of naming time of county-level administrative toponyms in 2010. https://doi.org/10.1371/journal.pone.0217381.g007 Fig 7. Distribution characteristics of naming time of county-level administrative toponyms in 2010. Fig 7. Distribution characteristics of naming time of county-level administrative toponyms in 2010. https://doi.org/10.1371/journal.pone.0217381.g007 11 / 24 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 However, the Northeast China Plain does not demonstrate the
same rule, because of the small number of samples and influence of ethnic minority languages. Therefore, we contend that natural factors (such as mountains and hydrological features)
always played a vital role in the naming of county-level administrative toponyms since the Sui,
yet human/social factors will play a more critical role than natural factors in the evolution of
the naming basis of the county-level administrative toponyms cultural landscape in China’s
eastern plains areas. Therefore, we contend that natural factors (such as mountains and hydrological features)
always played a vital role in the naming of county-level administrative toponyms since the Sui,
yet human/social factors will play a more critical role than natural factors in the evolution of
the naming basis of the county-level administrative toponyms cultural landscape in China’s
eastern plains areas. Table 2 clarifies that in 2010, county-level administrative toponyms relevant to hydrological
features constituted the largest proportion in the Northeast China Plain, North China Plain,
and Yangtze Plain. This demonstrates that China’s plains areas were relatively abundant in
water resources during historical periods; for example, the Liao, Yellow, Huai, and Yangtze
Rivers all flow across these three plains. Historically, people tended to settle nearby a water
source. Next, county-level administrative toponyms related to ancient relics constitute a rela-
tively large proportion, indicating that the plains areas have a long history and abundant cul-
tural deposits. Then, county-level administrative toponyms relevant to blessings account for
the second-largest percentage in the North China Plain. This could be attributed to the fre-
quent hazards caused by the Yellow River’s shifting course, floods, and overflows, which made
people seek peaceful and prosperous lives using toponyms containing good wishes. In addi-
tion, based on the existence of ethnic minority regimes during historical periods, more topo-
nyms in the Northeast China Plain are derived from the Mongolian or Manchu languages
than in the other two plains. Furthermore, the number of county-level administrative toponyms relevant to mountains
and hydrological features in the Northeast China Plain, North China Plain, and Yangtze Plain
in 2010 accounted for more than 30% of all toponyms, with the largest proportion in the Yang-
tze Plain and lowest in the North China Plain. This demonstrates that natural landscapes with
mountains and rivers greatly influence the naming of administrative units in China’s plains
areas, as in the rest of China. Besides, we used Python to write the statistical word frequency
code of administrative toponyms since the Sui, and ran the results with the help of Python 2.7. The results show that the words of “Yang” (the south of mountain or north of river), “Shan”
(mountain), and “Jiang” (river) always had high frequencies. Especially, “Shan” had the highest 12 / 24 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 County-level administrative toponyms cultural landscape Table 3. Statistical characteristics of county-level administrative toponyms relevant to natural factors since the Sui dynasty. Period
Northeast China Plain
North China Plain
Yangtze Plain
Total
Count
(%)
Count
(%)
Count
(%)
Count
(%)
Sui
0
0
151
54.1
89
67.4
240
58.3
Tang
1
20.0
140
49.3
75
63.0
216
52.9
Song
3
37.5
98
47.6
57
58.2
158
50.6
Yuan
1
20.0
65
45.8
59
56.2
125
49.6
Ming
3
42.9
116
43.4
100
56.2
219
48.5
Qing
5
20.8
111
41.1
98
53.6
214
44.9
ROC
28
36.8
109
38.9
103
52.8
240
43.6
PRC
43
34.4
141
36.3
153
51.3
337
41.6 cal characteristics of county-level administrative toponyms relevant to natural factors since the Sui dynasty. Table 3. Statistical characteristics of county-level administrative toponyms relevant to nat frequency in 2010. According to Professor Shi Nianhai, the naming of administrative units
focused on geographic factors, and toponyms named after natural landscapes such as moun-
tains and rivers were universal and stable [51]. Therefore, to further analyze the correlations
between naming rules and natural factors, we sorted the statistics of county-level administra-
tive toponyms relevant to natural factors since the Sui (Table 3). As shown in Table 3, the percentages of county-level administrative toponyms in the North
China Plain and Yangtze Plain relevant to natural factors has been declining since the Sui, as
have those for the total of three plains in spite of always staying above 40%. Conversely, the
percentage of toponyms concerned with human/social factors has increased. This is also evi-
dent in Figs 4–6. Moreover, 58.4% of the total toponyms in the three plains were relevant to
human/social factors in 2010. However, the Northeast China Plain does not demonstrate the
same rule, because of the small number of samples and influence of ethnic minority languages. As shown in Table 3, the percentages of county-level administrative toponyms in the North
China Plain and Yangtze Plain relevant to natural factors has been declining since the Sui, as
have those for the total of three plains in spite of always staying above 40%. Conversely, the
percentage of toponyms concerned with human/social factors has increased. This is also evi-
dent in Figs 4–6. Moreover, 58.4% of the total toponyms in the three plains were relevant to
human/social factors in 2010. PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 https://doi.org/10.1371/journal.pone.0217381.g009 Factors influencing the spatio-temporal evolution of the county-level
administrative toponyms cultural landscape Factors influencing spatial differentiation. Changes caused by the allocation and distri-
bution of political districts are dynamic indicators of economic prosperity and population
changes [52]. Generally, a more developed district has a larger number of administrative topo-
nyms, and vice versa. To explore the influencing factors of the spatial differentiation of
county-level administrative toponyms in the plains areas of China, we analyzed the coupling
relationships between toponyms and natural, economic, and demographic factors. First, we
demonstrated the spatial distribution of these elements by using data on toponym density,
DEM, county GDP, and county population density in 2010, as shown in Figs 8–14. Subse-
quently, we also conducted a quantitative analysis of the spatial correlations between toponyms
and natural, economic, and demographic factors at the grid level using grid data on China’s
GDP and population in 2010. The correlation coefficients listed in Table 4 were calculated
using formulae (2)–(4). Figs 8–14 show the spatial distributions of toponym density and DEM, county GDP, and
county population density in the Northeast China Plain, North China Plain, and Yangtze
Plain. However, these figures can only demonstrate spatial correlations between toponyms PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 13 / 24 County-level administrative toponyms cultural landscape Fig 8. DEM data of the case study areas. DEM data at a resolution of 30 m were provided by the RESDC. https://doi org/10 1371/journal pone 0217381 g008 Fig 8. DEM data of the case study areas. DEM data at a resolution of 30 m were provided by the RESDC. Fig 8. DEM data of the case study areas. DEM data at a resolution of 30 m were provided by the R https://doi.org/10.1371/journal.pone.0217381.g008 Fig 9. Spatial distribution of toponym density and county GDP of the Northeast China Plain. The statistical data
on China’s GDP in 2010 (Figs 8, 10, 12) were obtained from the RESDC. Fig 9. Spatial distribution of toponym density and county GDP of the Northeast China Plain. The statistical data
on China’s GDP in 2010 (Figs 8, 10, 12) were obtained from the RESDC. https://doi org/10 1371/journal pone 0217381 g009 Fig 9. Spatial distribution of toponym density and county GDP of the Northeast China Plain. The statistical data
on China’s GDP in 2010 (Figs 8, 10, 12) were obtained from the RESDC. https://doi.org/10.1371/journal.pone.0217381.g009 https://doi.org/10.1371/journal.pone.0217381.g009 14 / 24 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 County-level administrative toponyms cultural landscape Fig 10. PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 Factors influencing the spatio-temporal evolution of the county-level
administrative toponyms cultural landscape Spatial distribution of toponym density and county population density in the Northeast China Plain. Population data at the county level (Figs 9, 11, 13) were obtained from the 6th Census of China (2010). https://doi.org/10.1371/journal.pone.0217381.g010 Fig 10. Spatial distribution of toponym density and county population density in the Northeast China Plain. Population data at the county level (Figs 9, 11, 13) were obtained from the 6th Census of China (2010). Fig 10. Spatial distribution of toponym density and county population density in the Northeast China Plain. Population data at the county level (Figs 9, 11, 13) were obtained from the 6th Census of China (2010). https://doi.org/10.1371/journal.pone.0217381.g010 https://doi.org/10.1371/journal.pone.0217381.g010 and natural, economic, and demographic factors in a simple and intuitive way. Consequently,
we calculated the correlation coefficients between the density of county-level administrative
toponyms and DEM, GDP, and population density at the grid level using formulae (2)–(4), as
shown in Table 4. The absolute values of the Pearson correlation coefficients shown in Table 4 are all larger
than 0.50, indicating that county-level administrative toponym density is correlated with
DEM, GDP, and population density. Moreover, the correlation coefficients between toponym
density and DEM were negative, meaning that regions in the plains areas at a higher elevation
had fewer administrative toponyms. Conversely, the correlation coefficients between toponym
density and GDP and population density were positive; thus, regions with higher GDPs or Fig 11. Spatial distribution of toponym density and county GDP of the North China Plain. https://doi.org/10.1371/journal.pone.0217381.g011 Fig 11. Spatial distribution of toponym density and county GDP of the North China Plain. https://doi.org/10.1371/journal.pone.0217381.g011 Fig 11. Spatial distribution of toponym density and county GDP of the North China Plain. https://doi.org/10.1371/journal.pone.0217381.g011 15 / 24 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 County-level administrative toponyms cultural landscape Fig 12. Spatial distribution of toponym density and county population density in the North China Plain. https://doi.org/10.1371/journal.pone.0217381.g012 g 12. Spatial distribution of toponym density and county population density in the North China Plain. Fig 12. Spatial distribution of toponym density and county population density in the North China Plain. htt
//d i
/10 1371/j
l
0217381 012 Fig 12. Spatial distribution of toponym density and county population density in the North China Plain. https://doi.org/10.1371/journal.pone.0217381.g012 https://doi.org/10.1371/journal.pone.0217381.g012 population densities had more toponyms. In addition, the results showed that the correlation
degree between toponym density and population density was the highest in these three plains,
and that between toponym density and DEM the lowest. PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 Factors influencing the spatio-temporal evolution of the county-level
administrative toponyms cultural landscape Based on the research
achievement from some scholars [53], we calculated the GDP, and population density in these
county-level districts by using the the spatial distribution rules of the relief degree of land sur-
face at county level and its correlation between GDP and population density. In addition, we
also calculated the correlation coefficients between township-level toponym density and DEM,
GDP, and population density to verify the applicability of county-level correlations by down-
scaling analysis (Table 5), and also got similar results: the correlation degree between toponym
density and population density was the highest in these three plains, and that between topo-
nym density and DEM the lowest, the correlation degree between toponym density and DEM
in the Northeast China Plain was the highest, and lowest for the Yangtze Plain. were across some county-level districts, the data of GDP, and population density in these
county-level districts could not be simply calculated by the area ratios. Based on the research
achievement from some scholars [53], we calculated the GDP, and population density in these
county-level districts by using the the spatial distribution rules of the relief degree of land sur-
face at county level and its correlation between GDP and population density. In addition, we
also calculated the correlation coefficients between township-level toponym density and DEM,
GDP, and population density to verify the applicability of county-level correlations by down-
scaling analysis (Table 5), and also got similar results: the correlation degree between toponym
density and population density was the highest in these three plains, and that between topo-
nym density and DEM the lowest, the correlation degree between toponym density and DEM
in the Northeast China Plain was the highest, and lowest for the Yangtze Plain. In conclusion, the spatial distribution maps (see Figs 8–14) and correlation coefficients
(Tables 4 and 5) illustrate the factors influencing county-level administrative toponym spatial
differentiation. The results intuitively showed that DEM, GDP, and population density were
key elements in the spatial differentiation of toponyms cultural landscape in China’s plains
areas, although population density played a greater role in these changes than natural factors. Factors influencing temporal evolution. Considering the availability of historical data,
we only analyzed the coupling relationship between the number of county-level administrative
toponyms and historical temperatures, and between the number of county-level administrative
toponyms and historic population levels. Factors influencing the spatio-temporal evolution of the county-level
administrative toponyms cultural landscape This suggests that population has a
greater effect on the distribution of toponyms than altitude. Furthermore, Table 4 shows that
the correlation degree between toponym density and DEM in the Northeast China Plain was
the highest, and lowest for the Yangtze Plain. The correlation degree between toponym density
and GDP and population density in the Northeast China Plain was the highest, and lowest in
the North China Plain. The correlation coefficients in Table 4 were calculated by using the grid data of county-level
administrative toponym density, DEM, GDP, and population density, then Table 5 demon-
strates that we used the statistical data of GDP, and population density to evaluate the applica-
bility of these correlations. Considering the data characteristics, the county-level
administrative toponym density was the ratio between the number of county-level administra-
tive toponyms and the area of the county-level administrative district, the DEM data was the
mean value of grid data in the county-level district. Because the boundaries of the three plains The correlation coefficients in Table 4 were calculated by using the grid data of county-level
administrative toponym density, DEM, GDP, and population density, then Table 5 demon-
strates that we used the statistical data of GDP, and population density to evaluate the applica-
bility of these correlations. Considering the data characteristics, the county-level administrative toponym density was the ratio between the number of county-level administra-
tive toponyms and the area of the county-level administrative district, the DEM data was the
mean value of grid data in the county-level district. Because the boundaries of the three plains Fig 13. Spatial distribution of toponym density and county GDP of the Yangtze Plain. https //doi org/10 1371/jo rnal pone 0217381 g013 Fig 13. Spatial distribution of toponym density and county GDP of the Yangtze Plain. https://doi.org/10.1371/journal.pone.0217381.g013 https://doi.org/10.1371/journal.pone.0217381.g013 https://doi.org/10.1371/journal.pone.0217381.g013 16 / 24 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 County-level administrative toponyms cultural landscape Fig 14. Spatial distribution of toponym density and county population density in the Yangtze Plain. https://doi.org/10.1371/journal.pone.0217381.g014 Fig 14. Spatial distribution of toponym density and county population density in the Yangtze Plain. https://doi.org/10.1371/journal.pone.0217381.g014 were across some county-level districts, the data of GDP, and population density in these
county-level districts could not be simply calculated by the area ratios. indicated the value was significant at the 0.01 level (2-tailed). The grid data on China’s population and GDP in
2010 (at a 1-km gridded resolution) were provided by the RESDC. indicated the value was significant at the 0.05 level (2-tailed). https://doi.org/10.1371/journal.pone.0217381.t004 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 For example, although county-level administra-
tive districts are the most stable and basic administrative regions in China’s history, the ranks
of county-level administrative toponyms range from the second to the fifth. Political factors
mainly affect the evolution of county-level administrative toponyms ranks. Specifically,
county-level administrative toponyms are often in different ranks in the administrative topo-
nym systems of different dynasties. For example, one might belong to the second level in the
Qin and Western Han dynasties; third level in the Eastern Han and Southern and Northern
dynasties; third, fourth, or fifth levels in the Yuan; third or fourth levels in the Ming; and so
on. Some scholars [55, 56] consider the fall of a dynasty a complicated event, and while tem-
perature and changes in dynasties may be correlated, simple correlation statistics may not be
appropriate for complex paleoclimates and histories. Thus, we conducted a preliminary study
of the factors influencing the spatio-temporal evolution of the county-level administrative top-
onyms cultural landscape based on the availability of historical data. The factors influencing the temporal evolution of county-level administrative toponyms are
complex and sometimes have a delayed effect. For example, although county-level administra-
tive districts are the most stable and basic administrative regions in China’s history, the ranks
of county-level administrative toponyms range from the second to the fifth. Political factors
mainly affect the evolution of county-level administrative toponyms ranks. Specifically, county-level administrative toponyms are often in different ranks in the administrative topo-
nym systems of different dynasties. For example, one might belong to the second level in the
Qin and Western Han dynasties; third level in the Eastern Han and Southern and Northern
dynasties; third, fourth, or fifth levels in the Yuan; third or fourth levels in the Ming; and so
on. Some scholars [55, 56] consider the fall of a dynasty a complicated event, and while tem-
perature and changes in dynasties may be correlated, simple correlation statistics may not be
appropriate for complex paleoclimates and histories. Thus, we conducted a preliminary study
of the factors influencing the spatio-temporal evolution of the county-level administrative top-
onyms cultural landscape based on the availability of historical data. Fig 15. Trend chart of toponym, population, and temperature changes. The temperature values in Fig 15 are the change values
with the current temperature. Factors influencing the spatio-temporal evolution of the county-level
administrative toponyms cultural landscape For the temperature data, we referred to Zhu’s tem-
perature change curve for 5,000 years in China [54]. Fig 15 shows that the change in the number of toponyms is consistent with temperature
changes from the Qin to the ROC and changes in the population size since the Ming. Table 4. Correlation coefficients between county-level administrative toponym density and DEM, GDP, and pop-
ulation density. Plain
RDEM
RGDP
RPOP
Northeast China Plain
-0.73
0.90
0.92
North China Plain
-0.66
0.78
0.81
Yangtze Plain
-0.60
0.88
0.90 Table 4. Correlation coefficients between county-level administrative toponym density and DEM, GDP, and pop-
ulation density. 17 / 24 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 County-level administrative toponyms cultural landscape Table 5. Correlation coefficients between administrative toponym density and DEM, GDP, and population density. Plain
County-level administrative toponym density
Township-level administrative toponym density
RDEM
RGDP
RPOP
RDEM
RGDP
RPOP
Northeast China Plain
-0.68
0.75
0.89
-0.78
0.80
0.85
North China Plain
-0.62
0.81
0.83
-0.59
0.89
0.91
Yangtze Plain
-0.56
0.76
0.85
-0.54
0.93
0.95
indicated the value was significant at the 0.05 level (2-tailed). indicated the value was significant at the 0.01 level (2-tailed). The GDP data in 2010 at the county level were from the RESDC and population data at the county level
from the 6th Census of China (2010). Table 5. Correlation coefficients between administrative toponym density and DEM, GDP, and population density. Plain
County-level administrative toponym density
Township-level administrative toponym density
RDEM
RGDP
RPOP
RDEM
RGDP
RPOP
Northeast China Plain
-0.68
0.75
0.89
-0.78
0.80
0.85
North China Plain
-0.62
0.81
0.83
-0.59
0.89
0.91
Yangtze Plain
-0.56
0.76
0.85
-0.54
0.93
0.95 efficients between administrative toponym density and DEM, GDP, and population density. Table 5. Correlation coefficients between administrative toponym density and DEM, GDP, and population density. indicated the value was significant at the 0.05 level (2-tailed). indicated the value was significant at the 0.05 level (2-tailed). indicated the value was significant at the 0.05 level (2-tailed). indicated the value was significant at the 0.05 level (2-tailed). indicated the value was significant at the 0.01 level (2-tailed). The GDP data in 2010 at the county level were from the RESDC and population data at the county level
from the 6th Census of China (2010). 05 level (2 tailed). 01 level (2-tailed). The GDP data in 2010 at the county level were from the RESDC and population data at the county level https://doi.org/10.1371/journal.pone.0217381.t005 Furthermore, toponym density demonstrates a relative growth since the Ming (Table 1). We
contend that temperature changes were important in historic toponym changes, and popula-
tion changes have influenced toponym changes over the last 400 years in China. Furthermore, toponym density demonstrates a relative growth since the Ming (Table 1). We
contend that temperature changes were important in historic toponym changes, and popula-
tion changes have influenced toponym changes over the last 400 years in China. g
p
y
g
y
The factors influencing the temporal evolution of county-level administrative toponyms are
complex and sometimes have a delayed effect. Conclusions This paper intuitively demonstrated the spatial-temporal characteristics and causes of changes
in the county-level administrative toponyms cultural landscape in China’s eastern plains areas
through the application of GIS spatial analysis, Geo-Informatic Tupu, KDE method, and cor-
relation coefficients. The conclusions are summarized as follows. 1. In China’s eastern plains areas, the number of administrative toponyms has roughly
increased over time since the Sui, the number of toponyms is decreased in Song and Yuan
dynasties. 2. The results strongly indicate that high population density and exploitation intensity can
trigger spatial and temporal changes to administrative toponyms. This can also be applied
in interpreting the moving of China’s political center, economic barycenter, and population
gravity center in historical periods. 3. In 2010, the naming time distribution of county-level administrative toponyms in the
Northeast China Plain was comparatively concentrated, but relatively scattered in the
North China Plain and Yangtze Plain. However, the highest proportion of naming time in
all three plains was during the PRC period. 4. County-level administrative toponyms related to mountains and hydrological features
accounted for more than 30% of all toponyms in 2010. However, the percentage of total
county-level administrative toponyms in the three plains related to natural factors has been
declining since the Sui. Conversely, the percentage of toponyms concerned with human/
social factors is increasing. Therefore, we contend that natural factors (such as mountains
and hydrological features) always played a vital role in the naming of county-level adminis-
trative toponyms since the Sui, yet human/social factors will play a more critical role than
natural factors in the evolution of the naming basis of the county-level administrative topo-
nyms cultural landscape in China’s eastern plains areas. 5. To explore the factors influencing the spatial differentiation of county-level administrative
toponyms cultural landscape, we further analyzed the correlations between it and DEM,
GDP, and population density. We negatively correlated county-level toponym density and
DEM in all three plains, and positively correlated toponym density with GDP and popula-
tion density. Furthermore, the correlation degree between toponym density and population
density was the highest, and that between toponym density and DEM the lowest. We pre-
liminarily explored the factors influencing the temporal evolution of the county-level
administrative toponyms cultural landscape by analyzing the change trends of toponyms,
populations, and temperatures since the Qin. For historical population data, we referred to the statistical data from the National Earth System
Science Data Sharing Infrastructure, National Science & Technology Infrastructure of China, and made some adjustments. https://doi.org/10.1371/journal.pone.0217381.g015 Fig 15. Trend chart of toponym, population, and temperature changes. The temperature values in Fig 15 are the change values
with the current temperature. For historical population data, we referred to the statistical data from the National Earth System
Science Data Sharing Infrastructure, National Science & Technology Infrastructure of China, and made some adjustments. https://doi.org/10.1371/journal.pone.0217381.g015 Fig 15. Trend chart of toponym, population, and temperature changes. The temperature values in Fig 15 are the change values
with the current temperature. For historical population data, we referred to the statistical data from the National Earth System
Science Data Sharing Infrastructure, National Science & Technology Infrastructure of China, and made some adjustments. https://doi.org/10.1371/journal.pone.0217381.g015 Fig 15. Trend chart of toponym, population, and temperature changes. The temperature values in Fig 15 are the change values
with the current temperature. For historical population data, we referred to the statistical data from the National Earth System
Science Data Sharing Infrastructure, National Science & Technology Infrastructure of China, and made some adjustments. https://doi.org/10.1371/journal.pone.0217381.g015 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 18 / 24 County-level administrative toponyms cultural landscape PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 Conclusions We confirmed that the change trend regard-
ing toponym numbers was consistent with temperature changes from the Qin to the ROC
and with population size changes since the Ming. Furthermore, toponym density demon-
strated a relative growth since the Ming. The approach used in this paper had limitations regarding the data and methodology. For
example, our research on the spatio-temporal distribution characteristics of toponyms in his-
torical periods was confined to the county level, and smaller dimensions such as the township
and village levels from the Sui to the ROC were not included in the study. Furthermore, we
only analyzed the spatio-temporal features of administrative toponym changes and influencing
factors in the plains areas. Thus, more researches are required on the evolutionary characteris-
tics of toponyms in different geomorphic settings such as mountains, hills, plateaus, and basin 19 / 24 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 County-level administrative toponyms cultural landscape regions, and other factors which could affect the spatio-temporal features of administrative
toponyms. Nevertheless, this study breaks through the restriction of administrative regions to reveal
the rules and causes of the spatio-temporal evolutionary characteristics of toponyms by using
China’s eastern plains areas as a study object, because it features the most intimate correlation
between humans and the environment. Importantly, through a quantitative analysis and visu-
alization of the correlations between toponyms and natural, economic, and demographic fac-
tors through Geo-Informatic Tupu, this study provides a new perspective for the future study
of toponyms. S1 Table. Statistical characteristics of county-level administrative toponyms.
(PDF) S2 Table. Distribution characteristics of naming types of county-level administrative topo-
nyms in 2010. (PDF) S2 Table. Distribution characteristics of naming types of county-level administrative topo-
nyms in 2010. (PDF) S3 Table. Statistical characteristics of county-level administrative toponyms relevant to
natural factors since the Sui dynasty. (PDF) S4 Table. Correlation coefficients between county-level administrative toponym density
and DEM, GDP, and population density. (PDF) S5 Table. Correlation coefficients between administrative toponym density and DEM,
GDP, and population density. (PDF) S1 File. Chinese population census data of 2010. (RAR) S1 File. Chinese population census data of 2010. (RAR) S2 File. Grid data of China’s population and GDP in 2010. (RAR) S2 File. Grid data of China’s population and GDP in 2010. (RAR) Supporting information S1 Fig. Distribution map of county-level administrative toponyms with the time of being
named from 1,000 years ago. (TIF)
S2 Fig. Map of the case study areas. (TIF)
S3 Fig. Generation and transmission mode of Geo-Informatic Tupu. (TIF)
S4 Fig. Spatio-temporal evolution of county-level administrative toponyms in the North-
east China Plain since the Sui dynasty. (TIF)
S5 Fig. Spatio-temporal evolution of county-level administrative toponyms in the North
China Plain since the Sui dynasty. (TIF)
S6 Fig. Spatio-temporal evolution of county-level administrative toponyms in the Yangtze
Plain since the Sui dynasty. (TIF)
S7 Fig. Distribution characteristics of naming time of county-level administrative topo-
nyms in 2010. (TIF)
S8 Fig. DEM data of the case study areas. (TIF)
S9 Fig. Spatial distribution of toponym density and county GDP of the Northeast China
Plain. (TIF)
S10 Fig. Spatial distribution of toponym density and county population density in the
Northeast China Plain. (TIF)
S11 Fig. Spatial distribution of toponym density and county GDP of the North China
Plain. (TIF) S1 Fig. Distribution map of county-level administrative toponyms with the time of being
named from 1,000 years ago. S1 Fig. Distribution map of county-level administrative toponyms with the time of being
named from 1,000 years ago. (TIF)
S2 Fi
M
f th
t d S6 Fig. Spatio-temporal evolution of county-level administrative toponyms in the Yangtze
Plain since the Sui dynasty. (TIF) S10 Fig. Spatial distribution of toponym density and county population density in the
Northeast China Plain. (TIF) S11 Fig. Spatial distribution of toponym density and county GDP of the North China
Plain. (TIF) S11 Fig. Spatial distribution of toponym density and county GDP of the North China
Plain. (TIF) 20 / 24 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 County-level administrative toponyms cultural landscape S12 Fig. Spatial distribution of toponym density and county population density in the
North China Plain. (TIF)
S13 Fig. Spatial distribution of toponym density and county GDP of the Yangtze Plain. (TIF)
S14 Fig. Spatial distribution of toponym density and county population density in the
Yangtze Plain. (TIF)
S15 Fig. Trend chart of toponym, population, and temperature changes. (TIF)
S1 Table. Statistical characteristics of county-level administrative toponyms. (PDF)
S2 Table. Distribution characteristics of naming types of county-level administrative topo-
nyms in 2010. (PDF)
S3 Table. Statistical characteristics of county-level administrative toponyms relevant to
natural factors since the Sui dynasty. (PDF)
S4 Table. Acknowledgments We thank Shengsheng Gong and Tao Zhang from Central China Normal University, who pro-
vided the historical geography data. We thank all participants in this study for their time. We
are also thankful to National Catalogue Service For Geographic Information, National Earth
System Science Data Sharing Infrastructure, and RESDC for basic geographic data. S13 Fig. Spatial distribution of toponym density and county GDP of the Yangtze Plain.
(TIF) S14 Fig. Spatial distribution of toponym density and county population density in the
Yangtze Plain. (TIF) Supporting information Correlation coefficients between county-level administrative toponym density
and DEM, GDP, and population density. (PDF)
S5 Table. Correlation coefficients between administrative toponym density and DEM,
GDP, and population density. (PDF)
S1 File. Chinese population census data of 2010. (RAR)
S2 File. Grid data of China’s population and GDP in 2010. (RAR)
Acknowledgments
We thank Shengsheng Gong and Tao Zhang from Central China Normal University, who pro-
vided the historical geography data. We thank all participants in this study for their time. We
are also thankful to National Catalogue Service For Geographic Information, National Earth
System Science Data Sharing Infrastructure, and RESDC for basic geographic data. References 1. Tuan YF. Space and Place: The Perspective of Experience. Minneapolis: University of Minnesota
Press; 1977. 2. China Encyclopedia Editor Committee “Geography” Editorial Board. The Encyclopedia of China.Geog-
raphy, Article Name. Beijing: The Encyclopedia of China Press; 1992. 3. Hu HN, Zhang LR. Cultural landscapes of toponyms of administrative regions in Shaanxi province. Areal Research And Development. 2013; 32(1): 170–174. 4. Situ SJ. Historical geography study of toponyms of Guangdong province. Collections of Essays on Chi-
nese Historical Geography. 1992;(1): 21–55. 5. Wang JT. Introduction of Toponymy. Beijing: China Social Sciences Publishing; 1993. 6. Conedera M, Vassere S, Neff C, Meurer M, Krebs P. Using toponymy to reconstruct past land use: a
case study of ‘bru¨sa´da’ (burn) in southern Switzerland. Journal of Historical Geography. 2007; 33: 729–
748. 7. Hua LF. Textual Research on the History of Toponymy in China. Beijing: Social Sciences Literature
Press; 2002. 8. Gu Y. A review of the researches on toponyms in modern China. Theory. 2017;(6): 147–149. 9. Chu YP. Essays on Toponymys. Beijing: Higher Education Press; 1986. 10. Ma YL. New exploration of toponymy. Nanjing: Nanjing University Press; 1993. 11. Sun DH. History of geographical names in China. Beijing: Chinese Environment Press; 1997. 12. Niu RC. Dictionary of the Origin of Chinese Geographical Names. Beijing: Minzu University of China
Press; 1999. 13. Song JC. Study of Toponyms Culture: Concepts, Ethnic Languages, Toponyms and Others. Beijing:
Law Press; 2010. 14. Hua LF. Research on the History of Chinese toponyms. Jinan: Qilu Publishing House; 2006. 15. Tang GP. Linguistics in Toponyms of Panzhihua. Chengdu: Southwest Jiaotong University Press;
2012. 16. U.S. Department of the Interior. "U.S. Board on Geographic Names", U.S. Board on Geographic Names
homepage, 9 February 2009 [cited 4 March 2019]. Available from: https://geonames.usgs.gov/ 17. Henry G. The origin of certain place names in the United State. USA: Washington government printing
office; 1905. 18. George RS. Names on the land: a historical account of placenaming in USA. Lexikos; 1982. 19. Sousa A, Garcı´a-Murillo P. Can place names be used as indicators of landscape changes? Application
to the Doñana Natural Park (Spain). Landscape Ecology. 2001; 16: 391–406. 20. Sweeney S, Jurek M, Bednar M. Using place names to interpret former floodplain connectivity in the
Morava River, Czech Republic. Landscape Ecology. 2007; 22: 1007–1018. 21. Hu Y. Spatio-temporal relationships among Chinese ancient and modern placenames oriented to Gene-
alogy GIS. M.A. Thesis, Nanjing Normal University. 2008. Writing – original draft: Yingying Wang. Writing – review & editing: Yingying Wang. Writing – review & editing: Yingying Wang. Author Contributions Conceptualization: Yingying Wang. Conceptualization: Yingying Wang. Data curation: Yingying Wang. Formal analysis: Yingying Wang. Funding acquisition: Yingjie Wang. Investigation: Yingying Wang. 21 / 24 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 County-level administrative toponyms cultural landscape Methodology: Yingying Wang. Project administration: Yingying Wang. Resources: Yingying Wang. Supervision: Yingying Wang. Visualization: Yingying Wang, Lei Fang, Shengrui Zhang, Tongyan Zhang, Daichao Li, Daz-
huan Ge. Writing – original draft: Yingying Wang. Writing – review & editing: Yingying Wang. Visualization: Yingying Wang, Lei Fang, Shengrui Zhang, Tongyan Zhang, Daichao Li, Daz-
huan Ge. References Available from: http://cdmd.cnki.com.cn/
Article/CDMD-10319-2008153304.htm 22. Sousa A, Garcı´a-Murillo P, Sahin S, Morales J, Garcı´a-Barro´n L. Wetland place names as indicators of
manifestations of recent climate change in SW Spain (Doñana Natural Park). Climatic Change. 2010;
100: 525–557. 22 / 24 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 County-level administrative toponyms cultural landscape 23. Luo W, Hartmann JF, Wang FH. Terrain characteristics and Tai toponyms: a GIS analysis of Muang,
Chiang and Viang. Geo Journal. 2010; 75: 93–104. 24. Wang FH, Wang GX, Hartmann J, Luo W. Sinification of Zhuang place names in Guangxi, China: a
GIS-based spatial analysis approach. Transactions of the Institute of British Geographers. 2012; 37(2):
317–333. 25. Wang FH, Wang GX, Li XJ. GIS-based spatial analysis of Zhuang place names in Guangxi, China. Geo-
graphical Research. 2013; 32(3): 487–496. 26. Wang CS, Ma QL. Temporal-spatial distribution of villages in Beijing based on toponym change. Journal
of Arid Land Resources and Environment. 2015; 29(7): 18–23. 27. Wang B, Situ SJ. Analysis of spatial characteristics of place names landscape based on GIS technology
in Guangdong Province. Geographical Research. 2007; 26(2): 238–248. 28. Chen C, Xiu CL, Chen W. Spatial distribution characteristics of place-name landscape based on GIS
approach in Beijing and its reasons for the formations. Scientia Geographica Sinica. 2014; 34(4): 420–
429. 29. Sun BS, Guo CE, Yang YT, Ge W. Spatial distribution characteristics of rural place-name cultural land-
scape based on GIS approach in Chengde. Scientia Geographica Sinica. 2017; 37(2): 244–251. 30. Shi WL. Dictionary of Chinese Historical Toponyms. Beijing: China Social Sciences Press; 2005. 31. Niu RC. The New Century Chinese City View. Beijing: People’s Daily Publishing Corporation of China;
2002. 32. National Catalogue Service For Geographic Information. Available from: http://www.webmap.cn/
commres.do?method=result100W 33. Tan QX. The Historical Atlas of China. Beijing: China Cartographic Publishing House; 1982. 34. Resource and Environment Data Cloud Platform. Available from: http://www.resdc.cn/ 35. Population Census Office under the State Council Department of Population and Employment Statistics
National Bureau of Statistics. Tabulation on the 2010 population census of the People’s Republic of
China by township. Beijing: China Statistics Press; 2012. 36. National Earth System Science Data Sharing Infrastructure. Available from: http://www.geodata.cn/ 37. Zhou ZH. The Changes of Administrative Divisions in Chinese History. Beijing: China’s International
Broadcasting Publishing House; 1998. 38. Dai JL. Encyclopedic Dictionary of Ancient and Modern Chinese Geographical Names. Shanghai:
Shanghai Lexicographical Publishing House; 2005. 39. Niu RC. References China’s Toponyms Anecdotes Dictionary. Beijing: China Social Sciences Press; 2016. 40. Liu R, Hu WP, Wang HL, Wu C, He J. The road network evolution of Guangzhou-Foshan metropolitan
area based on Kernel Density Estimation. Scientia Geographica Sinica. 2011; 31(1): 81–86. 41. Yu WH, Ai TH, Yang M, Liu JP. Detecting “Hot Spots” of facility POIs based on kernel density estimation
and spatial autocorrelation technique. Geomatics and Information Science of Wuhan University. 2016;
41(2): 221–226. 42. Chen SP, Yue TX, Li HG. Studies on Geo-Informatic Tupu and its application. Geographical Research. 2000; 19(4): 337–343. 43. Liao K. The discussion and prospect for Geo-Informatic Tupu. Geo-information Science. 2002;(1): 14–
20. 44. Qi QW. The latest development on Geo-Informatic Tupu. Science of Surveying and Mapping. 2004; 29
(6): 15–23. 45. Ye QH, Tian GL, Liu GH, Ye JM, Yao X, Liu QS, Lou WG, Wu SG. Tupu methods of spatial-temporal
pattern on land use changes: A case study in the Yellow River Delta. Journal of Geographical Sciences. 2004; 14(2): 131–142. 46. Zhang RQ. The research progress of Geo-Informatic Tupu. Science of Surveying and Mapping. 2009;
34(1): 14–16, 24. 47. Zheng HR, Luo HX, Zou YQ, Cheng YS, Zhang R. Quantifying vegetation restoration in a karst rocky
desertification area in Chongqing based on Geoinformatic Tupu. Acta Ecologica Sinica. 2016; 36(19):
6295–6307. 48. Liao K. Modern Cartography. Beijing: Science Press; 2003. pp. 294. 49. Hu H, Wang CZ, Liang WJ, Jiang D, Ju B. Design and implementation of a spatio-temporal data model
on account of shapefile and base state with amendments model. Journal of Geo-Information Science. 2012; 14(3): 313–319. 50. Liu JD, Jin RC, Zhou KY. Political Geography of China. Beijing: Science Press; 1999. 23 / 24 PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 County-level administrative toponyms cultural landscape 51. Shi NH. Study on toponyms and the exploration of relevant laws. Collections of Essays on Chinese His-
torical Geography. 1985;(1): 36–41. 52. Zhou ZH. Annals of Chinese Culture Local Administrative System Records. Shanghai: Shanghai Peo-
ple’s Publishing House; 1998. pp. 280–281. 53. Feng ZM, Zhang D, Yang YZ. Relief degree of land surface in China at county level based on GIS and
its correlation between population density and economic development. Jilin University Journal Social
Sciences Edition. 2011; 51(1): 146–151. 54. Zhu KZ. The preliminary study on China’s climate changes in the past 5000 years. Acta Archaeologica
Sinica. 1972;(1): 15–38. 55. PLOS ONE | https://doi.org/10.1371/journal.pone.0217381
May 28, 2019 References Yu WH, Dong XX, Lei M. Climatic change, war and dynasties vicissitude: a review of statistical and
quantitative literature based on Chinese data. Dong Yue Tribune. 2015; 36(9): 81–86. 56. Ka-wai Fan. Climatic change and dynastic cycles in Chinese history: a review essay. Climatic Change,
2010; 101: 565–573. 24 / 24
|
https://openalex.org/W4288429848
|
https://www.nature.com/articles/s41598-022-16781-4.pdf
|
English
| null |
Assessing temporal correlation in environmental risk factors to design efficient area-specific COVID-19 regulations: Delhi based case study
|
Scientific reports
| 2,022
|
cc-by
| 11,671
|
Assessing temporal correlation
in environmental risk factors
to design efficient area‑specific
COVID‑19 regulations: Delhi based
case study
OPEN Vishal Chaudhary1*, Pradeep Bhadola2*, Ajeet Kaushik3,4, Mohammad Khalid5,6,
Hidemitsu Furukawa7 & Ajit Khosla8* Vishal Chaudhary1*, Pradeep Bhadola2*, Ajeet Kaushik3,4, Mohammad Khalid5,6,
Hidemitsu Furukawa7 & Ajit Khosla8* Amid ongoing devastation due to Serve-Acute-Respiratory-Coronavirus2 (SARS-CoV-2), the global
spatial and temporal variation in the pandemic spread has strongly anticipated the requirement
of designing area-specific preventive strategies based on geographic and meteorological state-of-
affairs. Epidemiological and regression models have strongly projected particulate matter (PM) as
leading environmental-risk factor for the COVID-19 outbreak. Understanding the role of secondary
environmental-factors like ammonia (NH3) and relative humidity (RH), latency of missing data
structuring, monotonous correlation remains obstacles to scheme conclusive outcomes. We mapped
hotspots of airborne PM2.5, PM10, NH3, and RH concentrations, and COVID-19 cases and mortalities
for January, 2021-July,2021 from combined data of 17 ground-monitoring stations across Delhi. Spearmen and Pearson coefficient correlation show strong association (p-value < 0.001) of COVID-19
cases and mortalities with PM2.5 (r > 0.60) and PM10 (r > 0.40), respectively. Interestingly, the COVID-
19 spread shows significant dependence on RH (r > 0.5) and NH3 (r = 0.4), anticipating their potential
role in SARS-CoV-2 outbreak. We found systematic lockdown as a successful measure in combatting
SARS-CoV-2 outbreak. These outcomes strongly demonstrate regional and temporal differences in
COVID-19 severity with environmental-risk factors. The study lays the groundwork for designing and
implementing regulatory strategies, and proper urban and transportation planning based on area-
specific environmental conditions to control future infectious public health emergencies. The suddenness and global scope of the Serve Acute Respiratory Coronavirus2 (SARS-CoV-2) outbreak have
stressed the designing pandemic retarding modalities to control the devastation1,2. These contagion restricting
policies includes unprecedented strict lockdown, social distancing, wearing face-masks, massive vaccination, and
disinfecting strategies3–7. Most research is dedicated to modelling advanced point-of-care diagnostic practices
and treatment therapeutics to control pandemic-instigated morbidities and mortalities3,8–13. However, the suc-
cess of employed strategies has been found to vary spatially and influenced by diverse regional environmental
factors14–17. Moreover, some regions worldwide have been identified as SARS-CoV-2 hotspots and are severely 1Research Cell and Department of Physics, Bhagini Nivedita College, University of Delhi, New Delhi 110043,
India. 2Centre for Theoretical Physics and Natural Philosophy, Nakhonsawan Studiorum for Advanced Studies,
Mahidol University, Nakhonsawan 60130, Thailand. 3NanoBioTech Laboratory, Health System Engineering,
Department of Environmental Engineering, Florida Polytechnic University, Lakeland, FL 33805, USA. 4School of
Engineering, University of Petroleum and Energy Studies (UPES) , Dehradun, Uttarakhand, India. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports/ affected by contagion 18,19. The mutable nature of SARS-CoV-2 due to its capricious spread, different transmission
routes, and its interaction with environmental elements, is challenging prevailing preventive strategies, including
massive vaccination. Hence, a critical global public health objective is to identify key tunable regional factors
contributing to variable coronavirus disease (COVID-19) severity.f g
(
)
y
Numerous epidemiological and regression studies have anticipated different regional variables, including
geographical, seasonal, demographic, and environmental factors determining the COVID-19 impact17,20–23. Air
contamination has been identified as the crucial factor governing regional SARS-CoV-2 spread and severity24–28. Air contamination is a complex mixture of particulate and gaseous constituents like particulate matter (PM),
nitrous oxides (NO2), ammonia (NH3), sulfur dioxide (SO2), and carbon oxides (COx), ozone (O3) that varies
temporally and spatially. The air quality index (AQI) depends upon various geographical, economic, and demo-
graphic conditions24–28. Moreover, the urban pollution due to anthropogenic source pollutants or combustion
of traffic-related products induces airway hyper-responsiveness and inflammation resulting in the severity of
cardiovascular and respiratory diseases29–32. p
y
Doremalen et al.33,34 have reported that COVID-19 is also a respiratory disease, and SARS-CoV-2 remains
compelling and infectious in aerosols for a prolonged time. Several In vitro and In vivo studies have demonstrated
that exposure to air contaminants decreases immune response facilitating viral replication and penetration25,27. Moreover, the complex multi-interactions amongst virus and air contaminants through covalent and electrostatic
interactions promote SARS-CoV-2 persistence in the atmosphere and reduce vitamin-D synthesis35. Atmospheric
aerosols resulting from these interactions induce indirect hazards in the human body, alter the immune response,
and are responsible for pro-inflammatory and oxidative mechanisms in the lungs. p
pl
y
g
Numerous outdoor air contaminants, including ammonia, particulate matter (PM), sulfur dioxide, nitrogen
oxides, and carbon oxides, have been argued to possess different roles in COVID-19 transmissibility27,28,32,36. Amongst all, PM is anticipated to be most severe factor, which constitute fine particles possessing a primary (car
exhaust, construction, and road traffic) and a secondary origin (including ammonia, oxides of nitrogen, and
sulfur, which transform into particles through photochemical reactions in the atmosphere)29,37–41. Numerous
scientific findings reviewed by the US Environmental Protection Agency (EPA)42–45 have related fine PM particles
(PM2.5; PM with diameter ≤ 2.5 µm) to diversified adverse health issues, including mortality. The virus’s droplets
are argued to bind with PM, promoting the diffusion of droplets with the SARS-CoV-2 in aerosols36. www.nature.com/scientificreports/ The virus
particles bound to minute PM like PM2.5 are probable to penetrate deeper in the alveolar and tracheobronchial
regions of the susceptible host individual27,36. Additionally, PM plays an indirect role in COVID-19 spread
and impact by weakening the immune system and increasing individual vulnerability towards this disease46,47. It is argued to facilitate SARS-CoV-2 binding with susceptible cells by stimulating overexpression of ACE-2
receptors48–51. The impaired immune ability and chronic inflammation of individuals living in hot -spots of PM
have already been documented38,47,52–54. Additionally, the secondary air contaminants like ammonia, NOx and
tropospheric O3 contribute to PM formation and directly relate to COVID-19 spread and impact54. Moreover,
the nature and concentration of particular air contaminants, meteorological conditions like rainfall, humidity,
temperature, and demographic factors govern their interactions, viral persistence, transmission, and severity.55,56. p
g p
g
p
y
Various regression analyses have suggested a correlation between air contaminants and COVID-19 mortalities
and concluded that the individuals living in polluted areas are more susceptible to COVID-1957. Moreover, the
megacities (like Delhi, Mumbai, London, New York) were severely affected by SARS-CoV-2 during COVID-19
waves18. Furthermore, during the pandemic, India has documented a rise in several emerging and re-emerging
infectious diseases, which has created more burdens on the healthcare system and vaccination drive58–62. The
pollution level in Delhi is reported to decrease by 15% from 2019 to 2020 due to government policies, public
awareness, strict lockdown, and alternative pollution combating solutions60. Nevertheless, according to the World
Air Quality report 2020, released by IQAir, Delhi is the most polluted capital globally, with an average annual
PM2.5 concentration of 84.1 µg per cubic meter58. Various reports in the literature have explored the correlation between COVID-19 severity and AQI dur-
ing the first COVID-19 wave in different megacities25,27,63–65. However, to the best of our knowledge, no report
is dedicated explicitly to evaluating PM and ammonia’s impact on COVID-19 spread and severity during the
second COVID-19 wave in Delhi. Moreover, various noticeable methodological difficulties have been pointed
out in different studies to establish this correlation66–70. Ideally, the study should be comprehended to emphasise
the variability of specific contaminants. The other variables can be mutually highly correlated, which presents
an obstacle for any statistical interference analysis70–73. To address, it we employed two correlation strategies to
analyze the correlation between PM and COVID-19 severity in terms of RH and ammonia during the second
COVID-19 wave in Delhi. www.nature.com/scientificreports/ We aim to provide fundamental reasoning to design regional-based strategies depend-
ing upon environmental variables including PM, RH, and NH3 to prevent and control COVID-19 severity and
future epidemic devastations. Assessing temporal correlation
in environmental risk factors
to design efficient area‑specific
COVID‑19 regulations: Delhi based
case study
OPEN 5Graphene and
Advanced 2D Materials Research Group (GAMRG), School of Engineering and Technology, Sunway University,
No. 5, Jalan University, Bandar Sunway, 47500 Petaling Jaya, Selangor, Malaysia. 6Sunway Materials Smart Science
& Engineering (SMS2E) Research Cluster, Sunway University, No. 5, Jalan Universiti, Bandar Sunway, 47500
Petaling Jaya, Selangor, Malaysia. 7Department of Mechanical Systems Engineering, Graduate School of Science
and Engineering, Yamagata University, Yonezawa, Yamagata 992‑8510, Japan. 8School of Advanced Materials
and Nanotechnology, Xidian University, Xi’an 710126, People’s Republic of China. *email: chaudhary00vishal@
gmail.com; bhadola.pradeep@gmail.com; ajitkhosla@xidian.edu.cn | https://doi.org/10.1038/s41598-022-16781-4 Scientific Reports | (2022) 12:12949 www.nature.com/scientificreports/ Results
l
i Analysis to map hotspots of environmental risk factors. The present trend in the Delhi region
straightaway revealed the potential differences at the regional or sub-station level for reported AQI values. The
contribution of pollutant concentration in Delhi’s atmosphere from 17 monitoring sub-stations varies regionally,
as shown in Fig. 1. Amongst all sub-stations in Delhi, Mundka, Jahangirpuri, and Bawana regions show the highest average
(From Jan to July 2021) contribution for PM10 and PM2.5, except for Chandni Chowk, which has the highest
PM10 is one of the lowest contributors among PM2.5. The average contribution from the Mundka, Jahangirpuri,
and Bawana regions towards atmospheric PM2.5 is estimated to be around 129 µg/cm3. All the three higher
PM2.5 contributors (Mundka, Jahangirpuri, and Bawana) have been identified as industrial areas of Delhi74,75. Scientific Reports | (2022) 12:12949 | https://doi.org/10.1038/s41598-022-16781-4 www.nature.com/scientificreports/ Figure 1. Pie chart showing the regional distribution of percentage contribution of variables observed from all
sub-station locations to the overall AQI of Delhi to locate particular hotspots. stribution of percentage contribution of variables observed from all
Delhi to locate particular hotspots. Figure 1. Pie chart showing the regional distribution of percentage contribution of variables observed from all
sub-station locations to the overall AQI of Delhi to locate particular hotspots. These areas are surrounded by green areas/agricultural regions of Delhi, which contribute to the emission of
fine organic particles into the environment. However, the exceptional presence of the Chandni Chowk region,
identified as the highest PM10 contributor, can be attributed to the recent construction performed in the region
for its ongoing beautification and redevelopment. These areas are surrounded by green areas/agricultural regions of Delhi, which contribute to the emission of
fine organic particles into the environment. However, the exceptional presence of the Chandni Chowk region,
identified as the highest PM10 contributor, can be attributed to the recent construction performed in the region
for its ongoing beautification and redevelopment. i
It is documented that long-term exposure to PM2.5 increases the morbidity and mortality rate due to COVID-
1925,49,50. The pathogenicity of PM is strongly determined by its chemical compositions of particles like trace
elements. The trace elements such as arsenic (As), cadmium (Cd), chromium (Cr), lead (Pb), and mercury (Hg)
represent a minute fraction of net PM mass38,51,52. However, the concentrations are adequate to injure human
health, like cardiopulmonary or lung injuries, neurodegenerative diseases, and low birth weight. Scientific Reports | (2022) 12:12949 | Results
l
i AQI Category Range
PM2.5
PM10
NH3
Good
0-50
0-30
0-200
Satisfactory
51-100
31-60
201-400
Moderately Polluted
101-250
61-90
401-800
Poor
251-350
91-120
801-1200
Very Poor
351-430
121-250
1200-1800
Severe
430+
250+
1800+ Table 1. Category range of AQI for considered pollutants as per National AQI, CPCB (2014) illustrating AQ
categories advisable for human health in India. sectors in Delhi contribute to more significant variable concentrations and are identified as hotspots for respec-
tive variable emissions. It also anticipates the variation of impacts of these pollutants on the atmosphere and
living beings in Delhi. Statistical analysis during the second COVID‑19 impact wave. The statistics of variables (PMs, RH,
COVID-19 mortality, and cases) as average over the period (January 2021 to July 2021) have been estimated. The
average of variables over the considered period has been considered to analyze their correlation for the whole
Delhi region. The mean values of PM2.5 and PM10 for the considered period are moderately and very polluted,
respectively, as per guidelines of the National AQI (Table 1). The maxima of both PMs lie in the abysmal AQI
range, which raises severe public health and environmental concerns. However, the average value of ammonia
over the considered period is in the acceptable range, causing no severe concern. The simultaneous existence
of low ammonia concentration and higher PM concentration can be attributed to the photochemical formation
of PM in the atmosphere at the cost of ammonia41. This COVID-19 period in Delhi has been considered as its
second significant impact wave, costing approximately 74 lives per day in the aforesaid period. The average
confirmed cases are extensive at approximately 4,153 per day, which indicates the rapid spread of COVID-19
in the period described above in Delhi. Hence, the average value of COVID-19 cases and mortalities show the
severe impact of COVID-19 on Delhi in the considered period and justify the choice of the aforesaid period for
analysis. y
Further, the graphical distribution of all the variables for the entire considered period has been represented
in Fig. 2. h
h d
h
b
l
d
h
b
h
d
d f h
d Further, the data has been analyzed month-wise to observe the exact nature and period of the second COVID-
19 impact wave in Delhi. The data for pollutants has also been examined month-wise to perceive the concen-
tration of pollutants during the second COVID-19 impact wave. Results
l
i Recently, Tobler
et al.76 revealed PM’s chemical composition and source attribution in the Delhi-National capital region (NCR). The coarse fraction (PM10) was sourced by organic matter (OM), sulfur- nitrate-ammonium (SNA) ions, and
crustal materials. However, the fine fraction (PM2.5) was dominated by OM, SNA ions, and elemental carbon. Further, Hama et al.77 revealed that the construction and paved road dust are major contributing factors to PM
and are dominated by the most abundant elements, including K, Si, Fe, Al, and Ca. Hence, the analyzed chemical
compositions of PM in various reports suggest mixed sources of PM, including vehicular emission, construction,
biomass burning, and paved road dust. Hence, the prevalence of these PM hotspots is attributed to contributions
from vehicular emission construction work and OM suspended aerosols from agricultural practices78.hf The observed regional variation of ammonia in the Delhi region is different from that for PM. Among all
Delhi sub-stations, DTU, NSIT Dwarka, and Dwarka regions are the topmost contributors to airborne ammonia. These top three regions contribute an average of one-third (over 32%) of airborne ammonia over Delhi. DTU
region has been observed to contribute 14% of total airborne ammonia in overall Delhi. The primary source of
airborne ammonia is the agricultural sector, including livestock, agricultural fields, and biomass degradation79,80. DTU and Dwarka region is surrounded by agricultural fields and is greener than the rest of the Delhi regions. Hence, their observed role as the highest contributor to airborne ammonia is justified.h i
In the present study, the regional variation in relative humidity (RH) has also been analyzed. The RH of all
regions is nearly the same, with not much variance. It can be attributed to similar weather conditions throughout
the Delhi region due to its comparatively smaller area. It reveals the interannual variability of RH depending
enormously on the reference year implying a dominant meteorological influence in Delhi. y
y
p y
g
gl
Hence, the spatial variation of observed variables (PM, NH3, and RH) in Delhi has been observed, and the
appropriate reasoning has been illustrated. It signifies that the regions with higher industrial and agricultural Scientific Reports | (2022) 12:12949 | https://doi.org/10.1038/s41598-022-16781-4 www.nature.com/scientificreports/ Table 1. Category range of AQI for considered pollutants as per National AQI, CPCB (2014) illustrating AQI
categories advisable for human health in India. Results
l
i The month-wise distribution of variables is
illustrated as Line Plot (Fig. 3) and Box plot (Fig. 4) for better understanding.h The mean level of each pollutant within the considered aforesaid period decreased from January 2021 to July
2021. It is attributed to the dependence of PM on meteorological conditions74,76. It is evident from the literature
that the winter season is favourable for the formation of PM78,81. Moreover, paddy stubble burning during the winter in Delhi-NCR regions and adjacent states raises the PM
concentration82. Hence, the concentration of PM in the Delhi atmosphere is comparatively more in January 2021
than in July 2021. A similar pattern of decrease in RH with time was observed for January 2021 to March 2021,
which is the later spring season in Delhi. However, the RH value was found to again increase from April 2021 to
July 2021, which is attributed to the onset of the summer season in Delhi. It has already been reported that the air pollutant concentration varies smoothly with changes in RH83. Hence,
their correlation analysis is of significant importance. The concentration of ammonia in Delhi’s environment was
found to follow the same pattern as RH. It is found to decrease with time from January 2021 to March 2021 and
then increase from March 2021 onwards. During the aforesaid period, the decrease in pollutant concentrations
is attributed to preventive measures like strict lockdown imposed by the government. Scientific Reports | (2022) 12:12949 | https://doi.org/10.1038/s41598-022-16781-4 www.nature.com/scientificreports/ Figure 2. Distributions of the variables for the entire period with variable value at x-axis and frequency of that
value over all locations at the y-axis. Figure 2. Distributions of the variables for the entire period with variable value at x-axis and frequency of that
value over all locations at the y-axis. Figure 2. Distributions of the variables for the entire period with variable value at x-axis and frequency of that
value over all locations at the y-axis. Figure 3. Month-wise distribution of variables including environmental risk factors and COVID-19 severity
through-line plot. Figure 3. Month-wise distribution of variables including environmental risk factors and COVID-19 severity
through-line plot. Scientific Reports | (2022) 12:12949 | https://doi.org/10.1038/s41598-022-16781-4 www.nature.com/scientificreports/ Moreover, it has been observed that there is a sudden increase in COVID-19 cases and mortalities during
Figure 4. Box plot for month-wise distribution of variables including environmental risk factors and COVID-
19 severity. Figure 4. www.nature.com/scientificreports/ ferent variables, as shown in Fig. 5. The diagonal subplots show the distribution of each variable, whereas the
off-diagonal subplots show the pairwise relationship between variables. A linear dependence between PM10 and
PM2.5 has been obtained for the considered period. It can be attributed to the standard climatic requirements
and the chemical composition of PMs. Similarly, a linear relationship between COVID-19 cases and mortalities
has been observed during the pre-lockdown and unlock period. However, there is moderate relative dependence
between the two during the lockdown period. The restriction imposed on residents during the lockdown period
significantly contributed to controlling the COVID-19 spread. gi
y
g
p
Moreover, the imposed restrictions during the lockdown, including restrictions on movement and restrict-
ing construction and industrial activities, have improved the air quality of the Delhi atmosphere60,78. Hence,
the relationship between the COVID-19 and mortalities slightly diverges from linear dependence. However,
the observed dependence of COVID-19 impact on RH and pollutants’ concentration was not linear. Hence, it
indicates that Pearson correlation coefficient analysis may result in discrepancies, which can be considered by
employing Spearman analysis. Analysis of normalized variables with time. All the variables are normalized prior to comparative
analysis. In a typical normalization process, all the variables are rescaled such that they have zero mean and unit
variance such that: New variable: Yi = xi −µx
σ(x) where xi is the old variable, µ is the mean, and σ(x) is the variance.h where xi is the old variable, µ is the mean, and σ(x) is the variance. The evaluated variation of these normalized variables with time for the different considered periods has been
illustrated in Fig. 6. It is done to check and compare the variables during different periods. i
µ
( )
The evaluated variation of these normalized variables with time for the different considered periods has been
illustrated in Fig. 6. It is done to check and compare the variables during different periods. g
gf
It is evident from Fig. 6 that the COVID-19 impact and transmission are higher in the region where RH is
low. Relative humidity (RH) and atmospheric temperature play a crucial role in the progression and persistence
of the virus in the atmosphere. RH affects virus transmission in three significant ways55,56. www.nature.com/scientificreports/ It includes the human
immune system’s enhanced capability in higher RH, the SARS-CoV-2 decays faster when RH is around 60%,
and drier air (low RH) results in further transmission of the virus along with deeper penetration into the human
lungs. Thus, this reasoning further validates our observation of the rise in COVID-19 impact in terms of cases
and mortalities with RH.il During the pre-lockdown period, the pollutant variation shows significant variance with large fluctuations. The pre-lockdown peaks for PMs and NH3 were observed on 14th January 2021. The random fluctuations during
this period can be ascribed to various pollutant contributing factors, including biomass burning, use of agricul-
tural fertilisers, construction and roadside-based dust, industrial emission, and vehicular emission. The more
significant number of contributing factors during the pre-lockdown interval has resulted in more substantial
fluctuations in recorded pollutant concentrations. l
p
A sudden increase in COVID-19 impact cases and mortalities during April 2021 is evident in Fig. 6, which
is considered the commencement of the second wave in Delhi. It has led to a strict lockdown in Delhi as a pre-
cautionary measure to combat the COVID-19 impact. PMs showed the highest concentration values during
the lockdown period on 28–29 April 2021, whereas RH and NH3 showed elevated peaks on 19–20 May 2021. The COVID-19 impact (cases and mortalities) was also highest around similar dates (26–30 April 2021), which
indicates a correlation between PM level and COVID-19 impact. It has also been observed that the COVID-19
cases and mortalities were higher during the initial lockdown phase and started to decline towards the end of
the lockdown phase. COVID-19 is a highly infectious and communicable respiratory disease3. It is observed to
control significantly by imposing lockdown in Delhi by Delhi-NCT and the Central government of India60. The
restrictions on movement, industrial, construction, and commercial activities have helped improve Delhi’s AQI. It restricts the primary and secondary transmission route of the SARS-CoV-2 virus through respiration, aerosols,
and infected surfaces. It also results in combating the immediate impact of pollutant inhalation/absorption in
the human body, such as weakening the immune system and affecting the respiratory system. A similar trend
in variation of PM with time in the lockdown phase has been observed, which suggests the existence of a strong
correlation between PM with COVID-19 impact. Results
l
i Box plot for month-wise distribution of variables including environmental risk factors and COVID-
19 severity. Moreover, it has been observed that there is a sudden increase in COVID-19 cases and mortalities during
April 2021. This period is identified as the second significant COVID-19 impact wave in Delhi. However, the
COVID-19 impact showed a downward trend in the count of COVID-19 cases and mortalities during the lock-
down phase (19th April 2021 to 31st May 2021). It shows that the lockdown was proven efficient in controlling
the COVID-19 impact in Delhi during the second wave. Analysis of pairwise relation amongst the different variables. The entire considered period has
been divided into three-time frames, including pre-lockdown (1st January 2021 to 18th April 2021), lockdown
(19th April 2021 to 31st May 2021), and unlock (1st June 2021 to 31st April) periods based on lockdown imposed
by the government in Delhi. The pairwise relation has been evaluated for all the periods described above for dif- Scientific Reports | (2022) 12:12949 | https://doi.org/10.1038/s41598-022-16781-4 www.nature.com/scientificreports/ RH (− 0.11), PM2.5 (− 0.23), and PM10 (− 0.34). Nevertheless, the correlation is small but cannot be stated as
random as the estimated p-values for these correlations were less than 0.001. However, the observed correlation
between COVID-19 mortalities with NH3 is insignificant, as the p-value for this correlation is 0.274. It anticipates
a probability of a correlation between COVID-19 impact on RH and PM concentration in Delhi’s atmosphere. www.nature.com/scientificreports/ p
However, even after imposing the lockdown on April 19, the COVID-19 cases as well as mortalities still shows
a rise and peaks at around 1st May 2022. The increase in the cases, as well as mortalities even after lockdown, can
be attributed to the incubation period for COVID-19. After 1st May we see a clear downward trend in mortalities
Fig. 6, which is clearly the effect of lockdown.i g
yf
When a significant decrease in deaths and cases during the lockdown period is observed, Delhi and Central
Government implement a phase-wise unlock period. During the unlock Phase, multiple peaks are observed in
the pollutant levels (PMs, NH3). The COVID-19 cases and deaths continue to decline during the unlock Phase
suggesting the proper implementation of government policies regarding step-wise unlock phases to combat
COVID-19 impact in terms of transmission and mortalities. Correlation analysis. The variation from linear dependence relationships amongst the different variables
for the considered period has suggested proper strategies for correlation analysis. It is taken into account by
choosing the Spearmen correlation coefficient strategy for correlation analysis amongst the chosen variables. The obtained Pearson plots for different variables during the aforesaid phases of the chosen period have been
shown in Fig. 7. g
During the pre-lockdown phase, the mortalities due to COVID-19 show a slight negative correlation with
RH and PMs. The Spearmen correlation coefficient for COVID-19 mortalities was estimated to be harmful as Scientific Reports | (2022) 12:12949 | https://doi.org/10.1038/s41598-022-16781-4 www.nature.com/scientificreports/ w.nature.com/scientificreports/
RH (− 0.11), PM2.5 (− 0.23), and PM10 (− 0.34). Nevertheless, the correlation is small but cannot be stated as
d
th
ti
t d
l
f
th
l ti
l
th
0 001 H
th
b
d
l ti
Figure 5. Pairwise relationship of different variables including environmental risk factors and COVID-19
severity for entire period, pre-lockdown period, lockdown period, and unlock period. Figure 5. Pairwise relationship of different variables including environmental risk factors and COVID-19
severity for entire period, pre-lockdown period, lockdown period, and unlock period. Figure 5. Pairwise relationship of different variables including environmental risk factors and COVID-19
severity for entire period, pre-lockdown period, lockdown period, and unlock period. RH (− 0.11), PM2.5 (− 0.23), and PM10 (− 0.34). Nevertheless, the correlation is small but cannot be stated as
random as the estimated p-values for these correlations were less than 0.001. www.nature.com/scientificreports/ www.nature.com/scientificreports/ The slight correlation values can be attributed to various factors contributing to COVID-19 impact, including
free mobility of infected individuals and industrial and commercial activities contributing to more significant
respiratory failures. Moreover, ammonia is the observed common factor linking coronavirus hotspots worldwide
like bat caves, contaminated air in the proximity of slaughterhouses and agricultural fields treated with livestock
farm sewage, and polluted air in megacities with air pollution. The direct role of ammonia is hypothesized to
alkalinize the environment, which is favourable to coronavirus transmission54. It has been argued that the SARS-
CoV-2 S protein undergoes conformational changes in the alkaline pH environment. Such an environment is
required to induce the fusion of the virus with the plasma membrane of target cells34. q
p
g
However, the number of COVID-19 cases shows a high negative correlation with PM2.5 (correlation = − 0.63
with p-value < 0.001) and RH (correlation = − 0.58 with p-value < 0.001) during the pre-lockdown phase. Moreo-
ver, the COVID-19 cases show a moderate negative correlation with PM10 (− 0.41, p-value < 0.001) and NH3
(− 0.33, p-value < 0.001). It anticipates a strong dependence of COVID-19 transmission majorly on PM and RH. The role of RH on COVID-19 transmission has already been discussed in the previous section, which validates
our initial hypothesis of choosing RH as a significant variable in understanding its COVID-19 impact55,56,64,72. It has already been reported that secondary pollutants such as ammonia plays a significant role in COVID-19
transmission. It alkalizes the atmosphere, which is favorable for SARS-CoV-2 transmission and fusion of the virus
with the plasma membrane of target cells54. Hence, there is a moderate correlation between a concentration of
pollutants and RH with COVID-19 impact in the pre-lockdown Phase during the second impact wave in Delhi. h l
kd
d
fi
l h h
l
f
d d
h
h p
p
p
g
p
During the lockdown period, a significantly high positive correlation of COVID-19 caused deaths with PM2.5
(correlation = 0.60 with p-value < 0.001) and PM10 (correlation = 0.56 with p-value < 0.001). However, RH and
ammonia show a high negative correlation with deaths with a correlation value of − 0.59 (p-value < 0.001) and
− 0.35 (p-value = 0.023), respectively. The negative COVID-19 impact on RH again strengthens the hypothesis
that low RH supports higher transmission and penetration of viruses and weakens human immunity56,64. www.nature.com/scientificreports/ How-
ever, the virus droplets bound to PM2.5 are more probable for deeper penetration in susceptible individuals’
alveolar range, causing higher respiratory failures27,36. Hence, the obtained strong correlation amongst PM and
COVID-19 mortalities is justified.h ji
The number of COVID-19 cases during lockdown is also positively correlated with PM2.5 with a correlation
value of 0.56 (p-value < 0.001) and PM10 with a correlation value of 0.51 (p-value < 0.001). It is attributed to the
role of PM as an aerosol-based second route of SARS-CoV-2 transmission. The virus binds to PM particles and
penetrates quickly, making an individual more susceptible to COVID-19 by inhaling infected PM particles36. However, the number of COVID-19 cases shows significantly high anti-correlation with RH (correlation = − 0.55
with p-value < 0.001). It again validates that lower RH supports further virus transmission. Hence, during the
lockdown period, a high correlation exists between PM and RH with COVID-19 cases and mortalities.i p
g
During unlock period, no significant correlation was observed amongst COVID-19 caused mortalities and
PM, RH and NH3. The estimated correlation value of COVID-19 caused deaths with PM2.5, PM10, RH and NH3
are 0.13, 0.04, − 0.10, and 0.23, respectively. However, their more significant p-values (0.32, 0.80, 0.52, and
0.12) predict a random correlation among them, which is not significant. It can be attributed to various factors
contributing to COVID-19 mortalities apart from the chosen variables. There are also reports of delays in updat-
ing data related to COVID-19 mortalities, which has caused this trivial correlation. However, a weak positive
correlation between COVID-19 cases with PM sustains in unlock period. The estimated correlation values of
COVID-19 cases with PM2.5 is 0.33 (p-value of 0.029), and with PM10 is 0.34 (p-value of 0.024). It anticipates the
persistence of dependence of COVID-19 transmission on PM’s concentration in Delhi’s environment. However,
the correlation of COVID-19 cases with other variables is insignificant due to p value > 0.10.fi gi
p
Moreover, Pearson coefficient correlation has also been performed to evaluate correlation amongst the vari-
ables for comparative analysis with Spearmen strategy-based results84–86. The obtained Pearson correlation plots
in the aforesaid period and phases have been shown in Fig. 8.hfi The obtained results from Pearson’s coefficient correlation analysis are the following that obtained from
Spearman’s analysis. www.nature.com/scientificreports/ Moreover, the correlation coefficients between COVID-19 impact and other variables (RH,
PMs, and NH3) obtained from Spearman are more significant than Pearson’s during the lockdown period. It
shows that the COVID-19 impact with pollutants (PMs and NH3) and RH are monotonic and not linear. It
further strengthens the choice of Spearmen analysis over Pearson analysis for current studies and validates the
estimated correlation results. www.nature.com/scientificreports/ However, the observed correlation
between COVID-19 mortalities with NH3 is insignificant, as the p-value for this correlation is 0.274. It anticipates
a probability of a correlation between COVID-19 impact on RH and PM concentration in Delhi’s atmosphere. RH (− 0.11), PM2.5 (− 0.23), and PM10 (− 0.34). Nevertheless, the correlation is small but cannot be stated as
random as the estimated p-values for these correlations were less than 0.001. However, the observed correlation
between COVID-19 mortalities with NH3 is insignificant, as the p-value for this correlation is 0.274. It anticipates
a probability of a correlation between COVID-19 impact on RH and PM concentration in Delhi’s atmosphere. Scientific Reports | (2022) 12:12949 | https://doi.org/10.1038/s41598-022-16781-4 www.nature.com/scientificreports/ Figure 6. Variation of normalized variables with time for full period, pre-lockdown period, lockdown period,
and unlock period. Figure 6. Variation of normalized variables with time for full period, pre-lockdown period, lockdown period,
and unlock period. Figure 7. Spearman correlation plots amongst different variables for pre-lockdown phase, lockdown phase,
and unblock phase. Figure 7. Spearman correlation plots amongst different variables for pre-lockdown phase, lockdown phase,
and unblock phase. Scientific Reports | (2022) 12:12949 | https://doi.org/10.1038/s41598-022-16781-4 Discussions and conclusions In the present study, the impact of various environmental factors, including particulate matter, ammonia and
relative humidity, on COVID-19 spread and mortality has been evaluated using various correlation analyses. It
is evident that these factors vary regionally and play a vital role in COVID-19 progression and severity, which
requires immediate attention. Depending upon the various outcomes of analyses, a scheme of conclusive out-
comes to design area-specific COVID-19 regulations highlighted in Fig. 9 and is as following: 1. Persistence of strong correlation between environmental variables and COVID-19 contagion: The role of
leading environmental risk factors (PM, RH, and NH3) for fatalities, coupled with their emergent evidence of
association with COVID-19 contagion, motivates attention to assessing their concentrations during the sec-
ond impact wave in Delhi. Assessments of ambient concentrations during the pandemic are required better
to understand the relationship between COVID-19 and variable concentrations. However, gaps in ground-
based monitoring, coupled with latency in available data, monotonic dependence of aforesaid variables on
COVID-19 impact, motivate alternative Spearmen correlation coefficient strategy. Statistical correlation
analysis indicated a correlation between COVID-19 impact (cases and mortalities) and environmental factors https://doi.org/10.1038/s41598-022-16781-4 Scientific Reports | (2022) 12:12949 | www.nature.com/scientificreports/ Figure 8. Pearson correlation plots amongst different variables, including environmental risk factors and
COVID-19 severity for pre-lockdown phase, lockdown phase, and unblock phase. Figure 8. Pearson correlation plots amongst different variables, including environmental risk factors and
COVID-19 severity for pre-lockdown phase, lockdown phase, and unblock phase. Figure 9. Schematic of outcomes of determining correlation between environmental risk factors and COVID-
19 spread for designing area-specific regulating modalities for effectively combat the pandemic [created with
BioRender.com]. Figure 9. Schematic of outcomes of determining correlation between environmental risk factors and COVID-
19 spread for designing area-specific regulating modalities for effectively combat the pandemic [created with
BioRender.com]. (PM, NH3, and RH) during the second impact wave in Delhi. The dependence of COVID-19 severity on RH
and NH3 is found to be moderate (>0.4) and significant (p-value < 0.001) during the lockdown phase (19th
April 2021 till 31st May 2021) in Delhi. It strongly predicts the potential role of RH and NH3 in SARS-CoV-2
outbreak through abating the immunity, stimulating receptor cell binding, and formation of aerosols for
secondary transmission. Discussions and conclusions Success of strategic lockdown: Estimations from the regression model during strategic lockdown have clearly
suggested the notable success of unprecedented restrictions in Delhi in the form of economic shutdowns and
limitations on mobility. The estimated decrease in ambient contaminant concentration observed in regres-
sion analyses provides strong justification for the lockdown success rate in Delhi. Nevertheless, the resulting
economic crisis and degrading mental health of confined people are the foremost associated concerns.i g
gi
p
p
4. Prospects: Estimated outcomes from this study, including correlation amongst variables, hotspot identifi-
cation, and lockdown success, must be interrelated to design COVID-19 monitoring and regulating policy
effectively. The assessed temporal and spatial variation in COVID-19 impact based on variation in ambient
concentration of environmental risk factors has strongly suggested the requirement of area-specific regulatory
modalities. The first step is to map the hotspot of COVID-19 severity and environmental risk factors and
estimate their interrelation. The regulatory measures to effectively control the contributing environmental
factors must be employed in furtherance to decrease COVID-19 severity and associated fatalities. The out-
comes from the study also strongly anticipate the requirement of an acceptable ammonia monitoring policy
for agricultural practices, monitoring of indoor RH to slow down viral contagion, and compelling urban
and transportation planning to regulate PM emissions. Delhi being one of the top polluted cities in the world has faced a severe COVID-19 outbreak during the
second wave of the pandemic. In spite of strict lockdown, continuous monitoring and large-scale vaccination,
the severity caused during the second wave was devastating. It strongly suggests the requirement to improve
present regulating policies such as vaccination drive and forming area-specific new policies to control the infec-
tion effectively. This study has revealed the area-specific environmental factors contributing to the severity of
COVID-19. The obtained outcomes strongly anticipate that there exists a strong correlation between environ-
mental factors, especially PM and COVID-19 spread and severity. All the studied environmental factors play a
specific role in amplifying COVID-19 severity either by turning humans susceptible to the infection or acting
as secondary carriers promoting its spread. Moreover, secondary contaminants (RH and NH3) play a significant
role in affecting the immunity, promoting receptor cell binding, and creating a secondary transmission route
by promoting the formation of aerosols. However, there is notable variation in regional concentration of these
environmental factors, which suggests designing the area-specific COVID-19 monitoring and control strategies. Discussions and conclusions The association of COVID-19 cases and mortalities with PM2.5 (r = 0.63 and 0.60)
and PM10 (r = 0.41 and 0.56) is high and significant, anticipating PM as a significant risk factor promoting
COVID-19 outbreak and severity. The associations detected in regression analyses provide strong justifica-
tion for air contamination assisted severe COVID-19 outbreak in Delhi during the second impact wave. (PM, NH3, and RH) during the second impact wave in Delhi. The dependence of COVID-19 severity on RH
and NH3 is found to be moderate (>0.4) and significant (p-value < 0.001) during the lockdown phase (19th
April 2021 till 31st May 2021) in Delhi. It strongly predicts the potential role of RH and NH3 in SARS-CoV-2
outbreak through abating the immunity, stimulating receptor cell binding, and formation of aerosols for
secondary transmission. The association of COVID-19 cases and mortalities with PM2.5 (r = 0.63 and 0.60)
and PM10 (r = 0.41 and 0.56) is high and significant, anticipating PM as a significant risk factor promoting
COVID-19 outbreak and severity. The associations detected in regression analyses provide strong justifica-
tion for air contamination assisted severe COVID-19 outbreak in Delhi during the second impact wave. g
p
2. Hotspot mapping: Moreover, the hotspots of airborne PM2.5, PM10, NH3, and RH in Delhi have been identi-
fied by mapping their ambient concentration. The reasoning for hotspots inferred from the regression model
is based on identifying industrial and agricultural areas in Delhi. Primary PM hotspots are areas with a g
p
2. Hotspot mapping: Moreover, the hotspots of airborne PM2.5, PM10, NH3, and RH in Delhi have been identi-
fied by mapping their ambient concentration. The reasoning for hotspots inferred from the regression model
is based on identifying industrial and agricultural areas in Delhi. Primary PM hotspots are areas with a https://doi.org/10.1038/s41598-022-16781-4 Scientific Reports | (2022) 12:12949 | www.nature.com/scientificreports/ massive construction, vehicular emission, and industrial activities. Hotspots for secondary environmental
factors (NH3 and RH) release are agricultural and green areas. massive construction, vehicular emission, and industrial activities. Hotspots for secondary environmental
factors (NH3 and RH) release are agricultural and green areas. massive construction, vehicular emission, and industrial activities. Hotspots for secondary environmental
factors (NH3 and RH) release are agricultural and green areas. 3. Discussions and conclusions It includes mapping pollution and COVID-19 hotspots, analyzing correlation amongst the two variables, imply-
ing strategic lockdown in hotspots, adopting area-specific measures to control the concentration of supportive
environmental factors towards COVID-19 pandemic, and strategic area-specific vaccination. Hence, a collective
approach from policy-makers, health workers, data scientists, and environmentalists is required to design and
implement region-specific policies to control future COVID-19 waves and public health emergencies due to
virus outbreaks and similar infectious diseases. Materials and methods Study area. In the current study, ’Delhi’, the capital city of India, also one of the highly populated and pol-
luted metropolitan capitals, has been selected for analysis. Delhi is located at 28.61° N 77.23° E and holds the
second position in the world’s leading megacities87. It is India’s most significant urban city with more than 15
million population and a population density of around 11,297 people per km288. The burning of biomass (field
straws) and vehicular emissions are the primary sources of PM in ambient air in Delhi41. However, the agricul-
tural sector majorly contributes to airborne ammonia79,80. The total number of registered vehicles was the high-
est (10.26 million in 2017) compared to other Indian metropolitan cities89. Data collection and processing. The data relating to the concentration of various air contaminants par-
ticulate matter PM2.5, PM10, ammonia (NH3), and relative humidity (RH) has been collected from 17 ground
monitoring stations across various sub-districts in Delhi (https://app.cpcbccr.com/ccr/#/caaqm-dashboard-all/
caaqm-landing). It examines the environmental impact of the COVID-19 cases, deaths, and the government’s
prevention measures before and after the second wave in 2021. The data was collected for a period ranging from
1st January 2021 to 14th July 2021 from the Central Pollution Control Board (CPCB), under the Ministry of
Environment, Forests and Climate Change, Government of India (https://cpcb.nic.in/). The daily covid cases
and deaths of the aforementioned period were procured from the WHO (https://covid19.who.int/). The daily
COVID deaths, cases, and pollutants, particulate matter PM2.5, PM10, NH3, and RH, were analyzed from 1st
January 2021 to 14th July 2021. Missing data processing. The prevalent issue in air quality analysis is the incapability of various statistical
analysis strategies to cater to missing observations90,91. The sources for missing in-situ air pollutant data observa-
tions include power outages, filter changes, malfunctions and errors, pollutant concentrations lower than detec-
tion limits, and computer system crashes91. The problem of missing data is pervasive and can affect the analysis
significantly if not appropriately treated. Previous reports on air quality data analysis have identified three dif-
ferent classes of missing observation scenarios, including missing at random (MAR), missing completely at
random (MCAR), and not missing at random (NMAR)91,92. The data set used in the present studies was the MAR
criterion for 7% of missing data.h g
The most commonly used method for handling missing data is the "last observation carried forward” (LOCF)
method. Materials and methods In the LOCF method, the missing value in the data is replaced with the previous observation of the same Scientific Reports | (2022) 12:12949 | https://doi.org/10.1038/s41598-022-16781-4 www.nature.com/scientificreports/ variable93. The method assumes that the variable’s value remains unchanged for the missing data, and missing
data is replaced with the last observed value.hhi The analysis has been performed in two parts. The first part is the statistical and comparative analysis of the
air quality data of each 17 locations within Delhi from January to July 2021. This part discusses the air quality
index (AQI) across different sub-districts of Delhi and its temporal variation.hf f
The second part is focused on the effect of the COVID-19 spread, mortalities, restrictions, and lockdowns
on the AQI of Delhi. The AQI of Delhi is taken as the average of the AQI over the 17 locations. The period from
January to July 2021 is divided into three stages; the first stage, from 1st January 2021 to 18th April 2021, is
the pre-lockdown period. Due to the high number of COVID-19 cases and mortalities, the Delhi government
imposed a complete lockdown from 19th April 2021 till 31st May 2021, which comprises the second lockdown
stage. The third stage is the unlock period from 1st June 2021 till 14th July 2021. The unlock was done in 5 stages
in a piece-wise manner till the complete services were restored. The entire unlock period has been considered as
a single stage in the current study. The above regression methodology has been employed to fill in the missing
data observations for AQI datasets and COVID-19 data sets utilized in the current study. Methodology for data modeling: Spearmen and Pearson correlation coefficient. Spearman’s
rank correlation coefficient analysis, a non-parametric measure of the rank correlation between two variables
by assessing the monotonic relationship between the variables, has been used in the present analysis86. However,
previous reports on similar correlation analysis have used the Pearson correlation coefficient test. The main dif-
ference between Spearman’s rank correlation and the most commonly used Pearson correlation coefficient is that
the Pearson correlation coefficient can evaluate only the linear relationship between the variables85. In contrast,
Spearman’s rank correlation benchmarks a monotonic relationship. If the Spearman correlation coefficient is
greater than the Pearson correlation coefficient, the correlation between the variables is monotonic but not lin-
ear. Materials and methods Since the collected data for COVID-19 and AQI for Delhi in the present analysis is not normally distributed,
Spearman’s rank correlation coefficient is anticipated to give relevant results with high significance.hfii pfi
p
g
g
gi
The Spearman correlation coefficient is defined as the correlation between the rank variables86. Consider
two variables (Xi, Yi) which are first converted to rank variables (R(Xi), R(Yi)), the Spearman rank coefficient
(ρs) is defined as ρs = cov(R(X), R(Y))
[σR(x)σR(Y)] where Cov(R(X), R(Y)) is the covariance of the rank variables. σR(X) and σR(Y) are the standard deviations of the
rank variables. where Cov(R(X), R(Y)) is the covariance of the rank variables. σR(X) and σR(Y) are the standard deviations of
rank variables. If in the sample of size ‘n’, all n ranks are distinct, then the Spearman rank coefficient (ρs) is given by ρs = 1 −
6 d2
i
n(n2 −1) where di = R(Xi) – R(Yi) is the difference between the rank of each observation between two variables. The cor-
elation coefficients lie in the typical range − 1 (positive correlation) to + 1 (anti-correlation). where di = R(Xi) – R(Yi) is the difference between the rank of each observation between two variables. The cor-
relation coefficients lie in the typical range − 1 (positive correlation) to + 1 (anti-correlation). where di = R(Xi) – R(Yi) is the difference between the rank of each observation between two variables. The cor-
elation coefficients lie in the typical range − 1 (positive correlation) to + 1 (anti-correlation). Significance test. The probability value (p-value) has been calculated to analyze the significance of statis-
tical correlation amongst the mentioned variables94,95. The p-value measures the extent of random correlation
between the given variables and ranges between 0 to 1. For instance, the p-value of 1 implies a 100% probability
of random correlation amongst the given variables. It is stated as a ’null hypothesis’ illustrating no existing sig-
nificant correlation between the given data sets other than randomness. A p-value close to zero signifies that
the null hypothesis is false and a significant pure correlation amongst the given variables. Hence, as the p-value
approaches zero, the significance of obtained correlation amongst the variables is purer and not random. Data availabilityh y
The respective data used in the present analysis are available on CPCB, DPCC, and WHO websites. The respective data used in the present analysis are available on CPCB, DPCC Received: 15 January 2022; Accepted: 15 July 2022 Received: 15 January 2022; Accepted: 15 July 2022 www.nature.com/scientificreports/ www.nature.com/scientificreports/ 8. Pathania, D. et al. Essential oil-mediated biocompatible magnesium nanoparticles with enhanced antibacterial, antifungal, and
photocatalytic efficacies. Sci. Rep. 12, 11431 (2022). p
yfi
p
9. Chaudhary, V. et al. Emergence of MXene–polymer hybrid nanocomposites as high-performance next-generation chemiresistors
for efficient air quality Monitoring. Adv. Funct. Mater. 2112913 (2022). y
g
p y
y
p
fficient air quality Monitoring. Adv. Funct. Mater. 2112913 (2022) fi
q
y
g
(
)
10. Sadique, M. A. et al. High-performance antiviral nano-systems as a shield to inhibit viral infections: SARS-CoV-2 as a model case
study. J. Mater. Chem. B 9, 4620–4642 (2021). q
g
p
study. J. Mater. Chem. B 9, 4620–4642 (2021). y
1. Da Silva Santos, F. et al. Review—recent advances of electrochemical techniques in food, energy, environment, and forensic
applications. ECS Sensors Plus 1, 013603 (2022).fi pp
2. Hyodo, T., Sakata, W., Ueda, T. & Shimizu, Y. Effects of surface modification of platinum electrodes with gold on hydrogen-sensing
properties of diode-type sensors. ECS Sensors Plus 1, 013602 (2022). p
p
yp
(
)
3. Chaudhary, V., Kaushik, A., Furukawa, H. & Khosla, A. Review—towards 5th generation AI and IoT driven sustainable intelligent
sensors based on 2d mxenes and borophene. ECS Sensors Plus 1, 013601 (2022). p
4. Liang, L.-L., Tseng, C.-H., Ho, H. J. & Wu, C.-Y. Covid-19 mortality is negatively associated with test number and government
effectiveness. Sci. Rep. 10, 12567 (2020).f f
p
15. Rahman, M. et al. A global analysis on the effect of temperature, socio-economic and environmental factors on the spread and
mortality rate of the COVID-19 pandemic. Environ. Dev. Sustain. 23, 9352–9366 (2021).h y
16. Liu, X. et al. The role of seasonality in the spread of COVID-19 pandemic. Environ. Res. 195, 110874 (2021). Th k
l
d
l
f
h
d ff
b d
d h
17. Thakur, B. et al. A systematic review and meta-analysis of geographic differences in comorbidities and associated severity and
mortality among individuals with COVID-19. Sci. Rep. 11, 8562 (2021). y
g
p
8. Desai, D. Urban densities and the Covid-19 pandemic: Upending the sustainability myth of global megacities. Observer Research
Foundation vol. 244 (2020). 19. World Health Organization. Archived: WHO Timeline—COVID-19. Wold Health Organization 2020 https://www.who.int/news-
room/detail/27-04-2020-who-timeline---covid-19?gclid=EAIaIQobChMI4MaewOeo6gIVyyMrCh2JRgUIEAAYASAAEgLo3_D_
BwE (2020).l (
)
0. Abu-Hammad, O. et al. Factors influencing global variations in COVID-19 cases and fatalities; A review. Healthcare 8, 216 (2020) 20. Abu-Hammad, O. et al. www.nature.com/scientificreports/ Ten threats to global health in 2019. World Health Organisation (WHO) 1–18 (2019). k
d k
l
ll
d h 32. Suzuki, T., Hidaka, T., Kumagai, Y. & Yamamoto, M. Environmental pollutants and the immune response. Nat. Immunol. 21,
1486–1495 (2020). ty and viability of SARS-CoV-2. N. Engl. J. Med. 382, 1962–1966 ( 33. Stability and viability of SARS-CoV-2. N. Engl. J. Med. 382, 19 4. van Doremalen, N. et al. Aerosol and surface stability of SARS-CoV-2 as compared with SARS-CoV-1. N. Engl. J. Med. 382
1564–1567 (2020).i 5. Crane-Godreau, M. A., Clem, K. J., Payne, P. & Fiering, S. Vitamin D deficiency and air pollution exacerbate COVID-19 through
suppression of antiviral peptide LL37. Front. Public Heal. 8, (2020). pp
p p
36. Nor, N. S. M. et al. Particulate matter (PM2.5) as a potential SARS-CoV-2 carrier. Sci. Rep. 11, 2508 (2021). l
l
l
f
d
d
l
ll 7. Bo, M., Salizzoni, P., Clerico, M. & Buccolieri, R. Assessment of indoor-outdoor particulate matter air pollution: A review. Atmos
phere (Basel). 8, 136 (2017). p
(
)
(
)
38. Das, R. et al. Trace element composition of PM2.5 and PM10 from Kolkata—a heavily polluted Indian metropolis. Atmos. Pollut. Res. 6, 742–750 (2015).fih 39. Griffin, R. J. The sources and impacts of tropospheric particulate matter. Nat. Educ. Knowl. 4, 1 (2013). l
h
h
l
h
h
h
d
l
h 40. Vione, D. et al. Photochemical reactions in the tropospheric aqueous phase and on particulate matter. Chem. Soc. Rev. 35, 441
(2006). 41. Saraswati, Sharma, S. K., Saxena, M. & Mandal, T. K. Characteristics of gaseous and particulate ammonia and their role in the
formation of secondary inorganic particulate matter at Delhi, India. Atmos. Res. 218, 34–49 (2019). 2. Setti, L. et al. SARS-Cov-2RNA found on particulate matter of Bergamo in Northern Italy: First evidence. Environ. Res. 188, 109754
(2020). 43. United States Environmental Protetion Agency (USEPA). Integrated Science Assessment for Particulate Matter. U.S. Environmental
Protection Agency (2009). 44. Wu, X., Nethery, R. C., Sabath, M. B., Braun, D. & Dominici, F. Air pollution and COVID-19 mortality in the United S
Strengths and limitations of an ecological regression analysis. Sci. Adv. 6, 1 (2020). g
g
g
y
45. Brook, R. D. et al. Particulate matter air pollution and cardiovascular disease. Circulation 121, 2331–2378 (2010).h 46. Yang, L., Li, C. & Tang, X. References 1. WHO. Coronavirus (COVID-19) events as they happen. https://www.who.int/emergencies/diseases/novel-coronavirus-2019/
events-as-they-happen. https://www.who.int/emergencies/diseases/novel-cor (2020). y
pp
p
g
2. Yamamoto, V. et al. COVID-19: Review of a 21st century pandemic from etiology to neuro-psychiatric implications. J. Alzheimer’s
Dis. 77, 459–504 (2020). y
pp
p
g
2. Yamamoto, V. et al. COVID-19: Review of a 21st century pandemic from etiology to neuro-psychiatric implications. J. Alzheimer’s
Dis. 77, 459–504 (2020). y
g
2. Yamamoto, V. et al. COVID-19: Review of a 21st century pandemic from etiology to neuro-psychiatric implications. J. Alzheimer’s
Dis. 77, 459–504 (2020). (
)
3. Chaudhary, V., Royal, A., Chavali, M. & Yadav, S. K. Advancements in research and development to combat COVID-19 using
nanotechnology. Nanotechnol. Environ. Eng. 6, 8 (2021). 3. Chaudhary, V., Royal, A., Chavali, M. & Yadav, S. K. Advancements in research and development to combat COVID-19 using
nanotechnology. Nanotechnol. Environ. Eng. 6, 8 (2021). 3. Chaudhary, V., Royal, A., Chavali, M. & Yadav, S. K. Advancements in research and development to combat COVID-19 using
nanotechnology. Nanotechnol. Environ. Eng. 6, 8 (2021). gy
g
4. Sornambikai, S. et al. Review—Systematic review on electrochemical biosensing of breast cancer miRNAs to develop alternative
DCIS diagnostic tool. ECS Sensors Plus 1, 021602 (2022).fi g
5. Chaudhary, V., Mostafavi, E. & Kaushik, A. De-coding Ag as an efficient antimicrobial nano-system for controlling cellular/
biological functions. Matter 5(7), 1995–1998 (2022).i g
5. Chaudhary, V., Mostafavi, E. & Kaushik, A. De-coding Ag as an efficient antimicrobial nano-system for controlling cellular/
biological functions. Matter 5(7), 1995–1998 (2022).i g
5. Chaudhary, V., Mostafavi, E. & Kaushik, A. De-coding Ag as an efficient antimicrobial nano-system for controlling cellular/
biological functions. Matter 5(7), 1995–1998 (2022).i g
6. An emerging weapon in fighting pandemics. Pullangott, G., Kannan, U., S., G., Kiran, D. V. & Maliyekkal, S. M. A comprehensive
review on antimicrobial face masks. RSC Adv. 11, 6544–6576 (2021). 6. An emerging weapon in fighting pandemics. Pullangott, G., Kannan, U., S., G., Kiran, D. V. & Maliyekkal, S. M. A comprehensive
review on antimicrobial face masks. RSC Adv. 11, 6544–6576 (2021). (
)
7. Moonla, C. et al. Review—lab-in-a-mouth and advanced point-of-care sensing systems: detecting bioinformation from the ora
cavity and saliva. ECS Sensors Plus 1, 021603 (2022). https://doi.org/10.1038/s41598-022-16781-4 Scientific Reports | (2022) 12:12949 | www.nature.com/scientificreports/ Factors influencing global variations in COVID-19 cases and fatalities; A review. Healthcare 8, 216 (2020). 21. Middya, A. I. & Roy, S. Geographically varying relationships of COVID-19 mortality with different factors in India. Sci. Rep. 11,
7890 (2021).fi l
g g
21. Middya, A. I. & Roy, S. Geographically varying relationships of COVID-19 mortality with different factors in India. Sci. Rep. 11,
7890 (2021).fi 2. Kontis, V. et al. Magnitude, demographics and dynamics of the effect of the first wave of the COVID-19 pandemic on all-cause
mortality in 21 industrialized countries. Nat. Med. 26, 1919–1928 (2020). y
23. Sorci, G., Faivre, B. & Morand, S. Explaining among-country variation in COVID-19 case fatality rate. Sci. Rep. 10, 18909 (2020). 24 L lli S Ch
Y C W
S H & Vi
G I
t f
t
l
i
l
diti
d i
ll ti
COVID 19
d
i y
23. Sorci, G., Faivre, B. & Morand, S. Explaining among-country variation in COVID-19 case fatality rate. Sci. Rep. 10, 18909 (2020). 24. Lolli, S., Chen, Y.-C., Wang, S.-H. & Vivone, G. Impact of meteorological conditions and air pollution on COVID-19 pandemic y
23. Sorci, G., Faivre, B. & Morand, S. Explaining among-country variation in COVID-19 case fatality rate. Sci. Rep. 10, 18909 (2020). 24. Lolli, S., Chen, Y.-C., Wang, S.-H. & Vivone, G. Impact of meteorological conditions and air pollution on COVID-19 pandemic
transmission in Italy. Sci. Rep. 10, 16213 (2020).h . Sorci, G., Faivre, B. & Morand, S. Explaining among-country va 5. Karan, A., Ali, K., Teelucksingh, S. & Sakhamuri, S. The impact of air pollution on the incidence and mortality of COVID-19. Glob
Heal. Res. Policy 5, 39 (2020). y
26. Brandt, E. B., Beck, A. F. & Mersha, T. B. Air pollution, racial disparities, and COVID-19 mortality. J. Allergy Clin. Immunol. 146,
61–63 (2020).hf 7. Ali, N. & Islam, F. The effects of air pollution on COVID-19 infection and mortality—A review on recent evidence. Front. Public
Heal. 8, (2020).i 28. Becchetti, L. et al. Air quality and COVID-19 adverse outcomes: Divergent views and experimental findings. Environ. Res. 193,
110556 (2021). 29. Manisalidis, I., Stavropoulou, E., Stavropoulos, A. & Bezirtzoglou, E. Environmental and health impacts of air pollution: A review. Front. Public Heal. 8, 14 (2020). 30. Schraufnagel, D. E. et al. Air pollution and noncommunicable diseases. Chest 155, 417–426 (2019). 31. Aranda, S. www.nature.com/scientificreports/ & Rubin, D. B. Bayes and multiple imputation. i 93. Hamer, R. M. & Simpson, P. M. Last observation carried forward versus mixed models in the analysis of psychiatric clinical
Am. J. Psychiatry 166, 639–641 (2009).i 94. Dahiru, T. P-Value, a true test of statistical significance? A cautionary note. Ann. Ibadan Postgrad. Med. 6, 1 (2011).i 94. Dahiru, T. P-Value, a true test of statistical significance? A cautionary note. Ann. Ibadan Postgrad. Med. 6, 1 (2011). h h
l
fi
d
l
l
d
l
f
l
l
l i
95. Shih, W. J. & Aisner, J. Statistical Significance and p-Values. in Statistical Design, Monitoring, and Analysis of Clinical Trials
(Chapman and Hall/CRC, 2021). https://doi.org/10.1201/9781003176527-14. www.nature.com/scientificreports/ Effects of demographic and weather parameters on COVID-19 basic reproduction number. Front. Ecol. Evol. 8, 1
(2021). 3. Gupta, A. & Gharehgozli, A. Developing a machine learning framework to determine the spread of COVID-19. SSRN Electron. J
https://doi.org/10.2139/ssrn.3635211 (2020). 74. Taneja, K., Ahmad, S., Ahmad, K. & Attri, S. D. Spatio-temporal analysis of urban air quality: A comprehensive approach toward
building a smart city. Lecture Notes Civ. Eng. 58, 317–328 (2020).i 75. Goyal, P., Gulia, S. & Goyal, S. K. Identification of air pollution 5. Goyal, P., Gulia, S. & Goyal, S. K. Identification of air pollution hotspots in urban areas—An innovative approach using monitored
concentrations data. Sci. Total Environ. 798, 149143 (2021). i
concentrations data. Sci. Total Environ. 798, 149143 (2021) 6. Tobler, A. et al. Chemical characterization of PM2.5 and source apportionment of organic aerosol in New Delhi, India. Sci. Tota
Environ. 745, 140924 (2020).iiif 7. Hama, S. et al. Chemical source profiles of fine particles for five different sources in Delhi. Chemosphere 274, 129913 (2021). 8 Ni
R P
d
K L i
A J S
R & K h
M P
i i
ff
f COVID 19 l
kd
i
li
f i d
i . Hama, S. et al. Chemical source profiles of fine particles for five iiif
8. Nigam, R., Pandya, K., Luis, A. J., Sengupta, R. & Kotha, M. Positive effects of COVID-19 lockdown on air quality of industria
cities (Ankleshwar and Vapi) of Western India. Sci. Rep. 11, 4285 (2021). iiif
78. Nigam, R., Pandya, K., Luis, A. J., Sengupta, R. & Kotha, M. Positive effects of
cities (Ankleshwar and Vapi) of Western India. Sci. Rep. 11, 4285 (2021). arma, S. K. et al. Measurement of ambient ammonia over the Natio 79. Sharma, S. K. et al. Measurement of ambient ammonia over the National Capital Region of Delhi India. Mapan 29, 165 173 (2014). 80. Chaudhary, V., Gautam, A., Mishra, Y. K. & Kaushik, A. Emerging MXene–polymer hybrid nanocomposites for high-performance
ammonia sensing and monitoring. Nanomaterials 11, 2496 (2021).h Wang, Y. et al. The ion chemistry and the source of PM2.5 aerosol h
82. Beig, G. et al. Objective evaluation of stubble emission of North India and quantifying its impact on air quality of Delhi. Sc
Environ. 709, 136126 (2020).l (
)
83. Jayamurugan, R., Kumaravel, B., Palanivelraja, S. & Chockalingam, M. P. www.nature.com/scientificreports/ www.nature.com/scientificreports/ 54. Manigrasso, M., Protano, C., Guerriero, E., Vitali, M. & Avino, P. May SARS-CoV-2 diffusion be favored by alkaline aerosols and
ammonia emissions? Atmosphere (Basel). 11, 995 (2020).f p
55. Mecenas, P., Bastos, R. T. da R. M., Vallinoto, A. C. R. & Normando, D. Effects of temperature and humidity on the spre
COVID-19: A systematic review. PLoS One 15, e0238339 (2020). y
(
)
56. Božič, A. & Kanduč, M. Relative humidity in droplet and airborne transmission of disease. J. Biol. Phys. 47, 1–29 (2021). 56. Božič, A. & Kanduč, M. Relative humidity in droplet and airborne transmission of disease. J. Biol. Phys. 47, 1–29 (2021). 57 Qin C
t l D
r
l ti n f imm n r p n
in p ti nt
ith
r n
ir
2019 (COVID 19) in W h n Chin
Cli
I f t 57. Qin, C. et al. Dysregulation of immune response in patients with coronavirus 2019 (COVID-19) in Wuhan China. Clin. I
Dis. 71, 762–768 (2020). 58. Greenpeace y IQAir. World’s Most Polluted Cities in 2020 - PM2.5 Ranking | AirVisual. IQAir https://www.iqair.com/world-most-
polluted-cities?continent=&country=&state=&page=1&perPage=50&cities=9vYz9movJKpaMiogG%0Ahttps://www.iqair.com/
world-most-polluted-cities?continent=59af928f3e70001c1bd78e4f&country=&state=&page=2&perPage=50&cities= (2020). Greenpeace y IQAir. World’s Most Polluted Cities in 2020 - PM2.5 Ranking | AirVisual. IQAir https://www.iqair.com/world-most- 8. Greenpeace y IQAir. World’s Most Polluted Cities in 2020 - PM2.5 Ranking | AirVisual. IQAir https://www.iqair.com/world-most-
polluted cities?continent=&country=&state=&page=1&perPage=50&cities=9vYz9movJKpaMiogG%0Ahttps://www iqair com/ 58. Greenpeace y IQAir. World’s Most Polluted Cities in 2020 - PM2.5 Ranking | AirVisual. IQAir https://www.iqair.com/world-m
polluted-cities?continent=&country=&state=&page=1&perPage=50&cities=9vYz9movJKpaMiogG%0Ahttps://www.iqair.c
ld
ll
d
f
f
bd
f
( p
y Q
g |
Q
p //
q
/
olluted-cities?continent=&country=&state=&page=1&perPage=50&cities=9vYz9movJKpaMiogG%0Ahttps://www.iqair.com/
world-most-polluted-cities?continent=59af928f3e70001c1bd78e4f&country=&state=&page=2&perPage=50&cities= (2020). p
y
p g
p
g
9. Dhaka, S. K. et al. PM2.5 diminution and haze events over Delhi during the COVID-19 lockdown period: an interplay between
the baseline pollution and meteorology. Sci. Rep. 10, 13442 (2020).f p
gy
p
0. Mahato, S., Pal, S. & Ghosh, K. G. Effect of lockdown amid COVID-19 pandemic on air quality of the megacity Delhi India. Sci
Total Environ. 730, 139086 (2020).h 61. Rackimuthu, S., Hasan, M. M., Bardhan, M. & Essar, M. Y. COVID-19 vaccination strategies and policies in India: The need for
further re-evaluation is a pressing priority. Int. J. Health Plann. Manage. 37, 1847–1850 (2022). www.nature.com/scientificreports/ 62 Ri
M M A
l Gl b l i
f
i
i
li
d
i
COVID 19
d
i
T
M d H
l h 9 101 (2021) g
y
g
62. Riaz, M. M. A. et al. Global impact of vaccine nationalism during COVID-19 pandemic. Trop. Med. Health 49, 101 (2021). 63. Dutta, A. & Jinsart, W. Air quality, atmospheric variables and spread of COVID-19 in Delhi (India): An analysis. Aerosol Air Qual. Res. 21, 200417 (2021). 4. Tello-Leal, E. & Macías-Hernández, B. A. Association of environmental and meteorological factors on the spread of COVID-19
in Victoria, Mexico, and air quality during the lockdown. Environ. Res. 196, 110442 (2021).f q
y
g
5. Kumar, S. Effect of meteorological parameters on spread of COVID-19 in India and air quality during lockdown. Sci. Total Environ
745, 141021 (2020).h (
)
66. Bontempi, E. The europe second wave of COVID-19 infection and the Italy “strange” situation. Environ. Res. 193, 110476 (2021). 6
B
i E V
lli S & S
i F U d
di
COVID 19 diff
i
i
i
di
i li
l i di
i
l (
)
66. Bontempi, E. The europe second wave of COVID-19 infection and the Italy “strange” situation. Environ. Res. 193, 110476 (2021). 67. Bontempi, E., Vergalli, S. & Squazzoni, F. Understanding COVID-19 diffusion requires an interdisciplinary, multi-dimensional E. The europe second wave of COVID-19 infection and the Italy “ 66. Bontempi, E. The europe second wave of COVID-19 infection p ,h
p
y
g
,
(
)
7. Bontempi, E., Vergalli, S. & Squazzoni, F. Understanding COVID-19 diffusion requires an interdisciplinary, multi-dimensiona
approach. Environ. Res. 188, 109814 (2020). 68. Anand, U. et al. Novel coronavirus disease 2019 (COVID-19) pandemic: From transmission to control with an interdisciplinary
vision. Environ. Res. 197, 111126 (2021). 69. Villeneuve, P. J. & Goldberg, M. S. Methodological considerations for epidemiological studies of air pollution and the SARS and
COVID-19 coronavirus outbreaks. Environ. Health Perspect. 128, 095001 (2020). p
0. Milicevic, O. et al. PM2.5 as a major predictor of COVID-19 basic reproduction number in the USA. Environ. Res. 201, 111526
(2021). 71. Rames, A. COVID-19: reviewing risk factors and breaking transmission. Tost. Unise. Org 8, 45–84 (2021).f ames, A. COVID-19: reviewing risk factors and breaking transmis
l
l
ff
f d
h
d
h 2. Salom, I. et al. www.nature.com/scientificreports/ The Impact of PM2.5 on the Host Defense of Respiratory System. Front. Cell Dev. Biol. 8, (2020).hfl 47. Zhao, J. et al. The biological effects of individual-level PM 2.5 exposure on systemic immunity and inflammatory response in traffic
policemen. Occup. Environ. Med. 70, 426–431 (2013). 48. Borro, M. et al. Evidence-based considerations exploring relations between SARS-CoV-2 pandemic and air pollution: Involvement
of PM2.5-mediated up-regulation of the viral receptor ACE-2. Int. J. Environ. Res. Public Health 17, 5573 (2020). h
d
h
d
f l
d p
g
p
9. Tchicaya, A., Lorentz, N., Omrani, H., de Lanchy, G. & Leduc, K. Impact of long-term exposure to PM2.5 and temperature on
coronavirus disease mortality: observed trends in France. Environ. Heal. 20, 101 (2021).h y
50. Comunian, S., Dongo, D., Milani, C. & Palestini, P. Air pollution and COVID-19: The role of particulate matter in the spread and
increase of COVID-19’s morbidity and mortality. Int. J. Environ. Res. Public Health 17, 4487 (2020).i 51. Sharma, S. K. & Mandal, T. K. Chemical composition of fine mode particulate matter (PM 2.5) in an urban area of Delhi, India
and its source apportionment. Urban Clim. 21, 106–122 (2017).i pp
52. Dominici, F. et al. Chemical composition of fine particulate matter and life expectancy. Epidemiology 26, 556–564 (2015). 53. Mo, Z. et al. Cause analysis of PM2.5 pollution during the COVID-19 lockdown in Nanning, China. Sci. Rep. 11, 11119 (202 https://doi.org/10.1038/s41598-022-16781-4 Scientific Reports | (2022) 12:12949 | www.nature.com/scientificreports/ Influence of temperature, relative humidity and se
variability on ambient air quality in a coastal urban area. Int. J. Atmos. Sci. 2013, 1–7 (2013).fi y
q
y
J
,
(
)
84 Benesty J Chen J Huang Y & Cohen I Pearson correlation coefficient Springer Top Signal Process 2 1 4 (2009) 84. Benesty, J., Chen, J., Huang, Y. & Cohen, I. Pearson correlation coefficient. Springer Top. Signal Process. 2, 1–4 (2009). k
k
f
l
f
’
d
’
l
ffi
h 85. Hauke, J. & Kossowski, T. Comparison of values of Pearson’s and Spearman’s correlation coefficients on the same sets o
QUAGEO 30, 87–93 (2011).fi 86. Sedgwick, P. Spearman’s rank correlation coefficient. BMJ 349, g7327 (2014).h 87. UNDESA Population Division. The World’s Cities in 2018. W
f
d
l
l
l A Population Division. The World’s Cities in 2018. World’s Cities 2 87. UNDESA Population Division. The Worlds Cities in 2018. Worlds Cities 2018—Data Bookl. (ST/ESA/ SER.A/417) 34 (2018
88. Census. Census of India 2011: provisional population totals-India data sheet. Off. Regist. Gen. Census Comm. India. Indian Ce
Bur. (2011).if 89. Kumar, P. et al. Temporary reduction in fine particulate matter due to ‘anthropogenic emissions switch-off’ during COVI
lockdown in Indian cities. Sustain. Cities Soc. 62, 102382 (2020). 90. Hadeed, S. J., O’Rourke, M. K., Burgess, J. L., Harris, R. B. & Canales, R. A. Imputation methods for addressing missing da
short-term monitoring of air pollutants. Sci. Total Environ. 730, 139140 (2020). g
p
91. Khan, A. et al. Variable impact of COVID-19 lockdown on air quality across 91 Indian Cities. Earth Interact. 25, 57–75 (202
92 Littl
R J A & R bi
D B B
d
lti l i
t ti
i 200 220 (2014) htt
//d i
/10 1002/9781119013563 h 91. Khan, A. et al. Variable impact of COVID-19 lockdown on air quality across 91 Indian Cities. Earth Interact. 25, 57–75 (2021). 92. Little, R. J. A. & Rubin, D. B. Bayes and multiple imputation. in 200–220 (2014). https://doi.org/10.1002/9781119013563.ch10. 91. Khan, A. et al. Variable impact of COVID 19 lockdown on air quality across 91 Indian Cities. Earth Interact. 25, 57 75 (2021). 92. Little, R. J. A. & Rubin, D. B. Bayes and multiple imputation. in 200–220 (2014). https://doi.org/10.1002/9781119013563.ch10. y
92. Little, R. J. A. & Rubin, D. B. Bayes and multiple imputation. in 200–220 (2014). https://doi.org/10.1002/9781119013563.ch10. 92. Little, R. J. A. Acknowledgementsh g
The author wishes to thank the Vice-chancellor, University of Delhi, India, Central Pollution control board
(CPCB), Government of India, Delhi Pollution Control Board (DPCC), Government of Delhi, Department
of Science & Technology (DST), Government of India, and World Health Organization (WHO) for provid-
ing e-resources. PB thanks Mahidol University for providing research project grant MRC-MGR 04/2565. This
work was supported in part by JSPS KAKENHI Grant Number JP17H01224, JP18H0547, JP19H01122, JST https://doi.org/10.1038/s41598-022-16781-4 Scientific Reports | (2022) 12:12949 | www.nature.com/scientificreports/ COI Grant Number JPMJCE1314, JST-OPERA Program Grant Number JPMJOP1844, JST -OPERA Program
Grant Number JPMJOP1614, and the Cabinet Office (CAO), Cross-ministerial Strategic Innovation Promotion
Program (SIP), “An intelligent knowledge processing infrastructure, integrating physical and virtual domains”
(funding agency: NEDO) and the Program on Open Innovation Platform with Enterprises, Research Institutes
and Academia (OPERA) from JST. COI Grant Number JPMJCE1314, JST-OPERA Program Grant Number JPMJOP1844, JST -OPERA Program
Grant Number JPMJOP1614, and the Cabinet Office (CAO), Cross-ministerial Strategic Innovation Promotion
Program (SIP), “An intelligent knowledge processing infrastructure, integrating physical and virtual domains”
(funding agency: NEDO) and the Program on Open Innovation Platform with Enterprises, Research Institutes
and Academia (OPERA) from JST. Competing interests h g
The authors declare no competing interests. Author contributions V.C., P.B. and A.K. set up interface of authors and performed correlation analysis. V.C., A.K., A.K., M.K. and
H.F. interpreted results and structured MS. © The Author(s) 2022 Additional information Correspondence and requests for materials should be addressed to V.C., P.B. or A.K. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2022 https://doi.org/10.1038/s41598-022-16781-4 Scientific Reports | (2022) 12:12949 |
|
https://openalex.org/W2323470028
|
https://www.scielo.br/j/bdj/a/qDXkXVZkmJF8mbYqqrGrDZg/?lang=en&format=pdf
|
English
| null |
In Vitro Cytotoxicity Test and Surface Characterization of CoCrW Alloy in Artificial Saliva Solution for Dental Applications
|
Brazilian Dental Journal
| 2,016
|
cc-by
| 4,075
|
Introduction In this work, in vitro cytotoxicity test, surface
characterization and electrochemical behavior of a CoCrW
alloy have been performed in artificial saliva and 0.15
mol.L-1 NaCl medium in order to evaluate its application
as dental prosthesis material. For characterizing the
alloy surface, polarization curves, chronoamperometry,
electrochemical impedance spectroscopy (EIS), scanning
electron microscopy (SEM) , energy dispersive X-ray
spectroscopy (EDS) and X-ray photoelectron spectroscopy
were used. In vitro cytotoxicity test was also performed
to study the biocompatibility of the proposed alloy as
implant material. Metal alloys have been used since the beginning of
the 20th century as material in dentistry due to their
mechanical properties (1,2). Nowadays, due to economic
issues, non-precious alloys are replacing noble alloys in
dentistry. Dental non-precious metal alloys present a thin
passive film of oxide on the surface. This passive film must
have high adhesion, be compact, present high electrical
resistance and absence of defects such as cracks (1,2). The oral environment is perfect to promote the
oxidation of metallic materials. Factors like quantity and
quality of saliva, salivary pH, dental plaque, amount of
protein, chemical and physical properties of food and fluid
intake, and general oral health conditions may influence
the metal corrosion in the oral cavity (3). There are two
concerns about dental materials in oral environment: the
localized effects or systemic damage caused by the release
of corrosion products to the body and the effects on the
physical properties and clinical performance of the alloy
(4-7). Brazilian Dental Journal (2016) 27(1): 181-186
http://dx.doi.org/10.1590/0103-6440201300513 Brazilian Dental Journal (2016) 27(1): 181-186
http://dx.doi.org/10.1590/0103-6440201300513 ISSN 0103-6440 1Institute of Chemistry, USP
- Universidade de São Paulo,
São Paulo, SP, Brazil
2UFABC - Universidade Federal do
ABC, Santo André, SP, Brazil
3IPEN - Instituto de Pesquisas
Energéticas e Nucleares,
São Paulo, SP, Brazil
4Department of Materials Engineering,
UFScar - Universidade Federal de
São Carlos, São Carlos, SP, Brazil Key Words: CoCrW alloy,
surface characterization,
cytotoxity, dental materials. Correspondence: Silvia Maria Leite
Agostinho, Av. Prof. Lineu Prestes,
748, 05508-900 São Paulo, SP,
Brazil. Tel: +55-11-3091-2157.
e-mail: smlagost@iq.usp.br In Vitro Cytotoxicity Test and
Surface Characterization of
CoCrW Alloy in Artificial Saliva
Solution for Dental Applications Klester Santos Souza1, Ruth Flavia Vera Villamil Jaimes2, Sizue Otta Rogero3,
Pedro Augusto de Paula Nascente4, Silvia Maria Leite Agostinho1 Klester Santos Souza1, Ruth Flavia Vera Villamil Jaimes2, Sizue Otta Rogero3,
Pedro Augusto de Paula Nascente4, Silvia Maria Leite Agostinho1 In order to evaluate its application as a dental prosthesis material, a CoCrW alloy was
subjected to in vitro cytotoxicity test, surface characterization and electrochemical
studies performed in artificial saliva and 0.15 mol.L-1 NaCl medium. The used techniques
were: anodic polarization curves, chronoamperometric measurements, electrochemical
impedance spectroscopy (EIS), scanning electron microscopy (SEM), energy dispersive X-ray
spectroscopy (EDS) analysis and X-ray photoelectron spectroscopy (XPS). Cytotoxicity
test was also performed. The electrochemical behavior of CoCrW alloy was compared
in both studied media, from corrosion potential (Ecorr) to a 600 mV anodic overvoltage. From the electrochemical measurements it was observed that the CoCrW alloy in both
media presents only generalized corrosion. SEM and EDS analysis showed that the alloy
presents carbide niobium and silicon and manganese oxides as nonmetallic inclusions. XPS results indicated that cobalt does not significantly contribute to the passivating film
formation. Cytotoxicity test showed no cytotoxic character of CoCrW alloy. These results
suggest that the CoCrW alloy can be used as biomaterial to be applied as prosthesis in
dental implants. Key Words: CoCrW alloy,
surface characterization,
cytotoxity, dental materials. 1Institute of Chemistry, USP
- Universidade de São Paulo,
São Paulo, SP, Brazil
2UFABC - Universidade Federal do
ABC, Santo André, SP, Brazil
3IPEN - Instituto de Pesquisas
Energéticas e Nucleares,
São Paulo, SP, Brazil
4Department of Materials Engineering,
UFScar - Universidade Federal de
São Carlos, São Carlos, SP, Brazil K.S. Souza et al. K.S. Souza et al. Surface Characterization Surface Characterization Scanning Electron Microscopy (SEM) and Energy
Dispersive X-Ray Spectroscopy (EDS) Analysis SEM and EDS analyses were made on a WDX600 Oxford
microscope (Leica Microsystems. Wetzlar, Germany) coupled
to a Stereoscan 440 Leica scanning electron microscope
(Leica Microsystems). The samples were polished using 1 μm
diamond paste, rinsed with water and etanol and air dried. Different surface regions were analyzed using EDS (Table 2). All the experiments were performed in triplicate with
naturally aerated solution at (37.0±0.5) °C and a 6.6 pH
(oral conditions). X-ray Photoelectron Spectroscopy (XPS) Analysis
XPS analyses were made on a spectrometer (model
XSAM HS; Kratos Analytical Ltd, Manchester, UK) under
ultrahigh vacuum (about 10-8 Torr). Non-monochromatic
Mg Kα (hν=1253.6 eV) radiations were used as X-ray source,
with 5 mA emission current at 12 kV voltage. The samples
were polished, rinsed with de-ionized water, cleaned in an
ultrasonic bath with analytical grade acetone for 5 min and
rinsed with de-ionized water. Previous to these experiments,
the samples were immersed in artificial saliva and 0.15
mol.L-1 NaCl solution for 8 h at Ecorr. After 8 h the surface
composition was analyzed to investigate the presence of
cobalt (Co) and chromium (Cr) in the composition of the
passive film. As binding energy reference was used the
value 284.8 eV for the carbon peak corresponding to the
adventitious hydrocarbon (16). The evaluation included
Shirley procedure for background subtraction, mixed
Gaussian/Lorentzian function and a least-square routine
for fitting the peak. Electrochemical Experiments The corrosion resistance was evaluated by
potentiodynamic polarization anodic curves and
chronoamperometric curves. The studied CoCrW alloy was
in artificial saliva and 0.15 mol.L-1 NaCl solutions. These
evaluations were performed after the attainment of a
constant current at each potential application, starting
from the open circuit potential in the anodic direction
from Ecorr till 0.9 V vs. SCE using 1 mV.s-1 scan rate. As
working criterion, it was assumed that the transpassivation
potential is reached when the current density is 10 mA/
cm2. The electrochemical impedance spectroscopy (EIS)
measurements were performed over eight frequency
decades (from 100 KHz to 10 mHz). The potential was
±8 mV, p.p. The experiments were conducted at 37 °C. A
μAutolab type III/FRA2 potentiostat (Metrohm Autolab BV,
the Netherlands) was used coupled to a frequency response
detector and to a microcomputer. Cytotoxicity Tests The cytotoxicity assay was carried out by exposing the
cell culture to the solutions after contact with alloy samples,
immersed for 10 days in culture medium MEM at 37 °C. The NCTC clone 929 cell line was acquired from American
Type Culture Collection (ATCC) bank. The cytotoxicity effect
was evaluated by neutral red uptake (NRU) methodology,
described in previous papers (17,18), according to the
International Organization for Standardization (ISO) (19). Electrodes The CoCrW alloy working electrodes were discs
fabricated from the central part of the bars and had an
area of 0.90 cm2. A cylindrical epoxy-base was fitted
with the steel disk. A concentric brass rod was coupled
to the steel+epoxy-base. The electrodes were prepared by
polishing with successively finer grades of emery papers
of 120, 400, 600 and 2000 mesh and then thoroughly
rinsed with distilled water and ethanol and air dried prior
to the experiments. The auxiliary electrode consisted of a platinum foil
and it was cleaned with acid and flamed just before each
experiment. Saturated calomel electrode (SCE) was used
as reference electrode. Material and Methods
Sample and Solutions The chemical composition of the CoCrW alloy is in
Table 1. The compositions of the artificial saliva were
prepared from three different solutions. The following
reagents/concentrations (mol.L-1) were used: solution A: CoCr-based alloys have been utilized in dental
prosthesis, since they offer good resistance to corrosion
(8,9). Recently, alloys with high content of tungsten instead
molybdenum have been developed aiming a better ceramic-
metal adhesion (10). On the other hand, the CoCrMo alloys
have been more studied (4-12) than CoCrW alloys (13,14). Table 1. Chemical composition of CoCrW alloy (wt%)
Sample
Co
Cr
W
Nb
V
Si
Mo
Fe
CoCrW
59.4
24.5
10.0
2.0
2.0
1.0
1.0
0.1 Braz Dent J 27(1) 2016 NaH2PO4/0.233 + KCl/1.164 + NaCl/0.123 + NH4Cl/0.205
+ sodium citrate/3.74x10-3 + lactic acid/0.039; solution
B- urea/0.167 + uric acid/4.46x10-3 + NaOH/5x10-3 and
solution C: KSCN/0.123. The saliva solution was prepared
daily by mixing A, B and C solutions (1:1:1) and then diluted
50 times in high-purity de-ionized water (15). Electrochemical Results The corrosion potentials (stationary open circuit
potential values) obtained
for artificial saliva and 0.15
mol.L-1 NaCl solutions were
-100±28 mVSCE and -277±23
mVSCE respectively (Fig. 1A). The anodic potentiodynamic
polarization curves were
obtained from corrosion
potential and plotted in
scopy (EDS) for CoCrW
Mo
Ta
W
0.85
-
8.50
0
2.90
-
9.18
8
0.5
6.29
2.45 182
detector and to a microcomputer. The corrosion potentials (stationary open circuit
potential values) obtained
for artificial saliva and 0.15
mol.L-1 NaCl solutions were
-100±28 mVSCE and -277±23
mVSCE respectively (Fig. 1A). The anodic potentiodynamic
polarization curves were
obtained from corrosion
potential and plotted in
Table 2. Different surface regions analyzed using energy dispersive spectroscopy (EDS) for CoCrW
alloy after polishing
Region
C
O
Si
V
Cr
Mn
Co
Sr
Nb
Mo
Ta
W
1
3.98
0.26
0.51
1.76
23.16
-
59.63
1.35
-
0.85
-
8.50
2
6.42
-
1.59
1.36
17.69
-
42.49
2.04
11.30
2.90
-
9.18
3
3.10
30.64
19.40
1.08
16.31
2.15
15.12
-
2.68
0.5
6.29
2.45 182
Table 2. Different surface regions analyzed using energy dispersive spectroscopy (EDS) for CoCrW
alloy after polishing
Region
C
O
Si
V
Cr
Mn
Co
Sr
Nb
Mo
Ta
W
1
3.98
0.26
0.51
1.76
23.16
-
59.63
1.35
-
0.85
-
8.50
2
6.42
-
1.59
1.36
17.69
-
42.49
2.04
11.30
2.90
-
9.18
3
3.10
30.64
19.40
1.08
16.31
2.15
15.12
-
2.68
0.5
6.29
2.45 Table 2. Different surface regions analyzed using energy dispersive spectroscopy (EDS) for CoCrW
alloy after polishing 182 Braz Dent J 27(1) 2016 tests were made in order to verify the presence or not of
pitting corrosion on the passive film. The obtained results
from polarization curves are plotted in Figure 1A with the
chronoamperometric curves at different potential values
(Figs. 1B and 1C) for the alloy, in artificial saliva and NaCl
medium, respectively. Chronoamperometric results are
shown at different potentials and confirm those obtained
by potentiodynamic polarization anodic curves, indicating
the potential range where the metallic surface is passivated. The constant values of the current density observed at
different potential values are an indication that the alloy
does not present pitting corrosion. order to evaluate the effect of potential, to determine the
potential range where the alloy is passivated and the value
of the transpassivation potential. The chronoamperometric Figure 1. Electrochemical Results Potentiodynamic polarizations curves of CoCrW alloy in
artificial saliva and 0.15 mol.L-1 sodium chloride at 37 oC. Panels
A, B and C show the chronoamperograms response at each potential
application. EIS was employed to investigate the passive film/
electrolyte interface and evolution with increasing potential. Figure 2 shows Bode (modulus and phase) diagrams for
CoCrW in artificial saliva (Fig. 2A) and 0.15 mol.L-1 NaCl
solutions (Fig. 2B) at Ecorr and 0.25 VSCE. A similar behavior
may be seen for the alloy/solution interface in both media:
the imaginary and real components of impedance decrease EIS was employed to investigate the passive film/
electrolyte interface and evolution with increasing potential. Figure 2 shows Bode (modulus and phase) diagrams for
CoCrW in artificial saliva (Fig. 2A) and 0.15 mol.L-1 NaCl
solutions (Fig. 2B) at Ecorr and 0.25 VSCE. A similar behavior
may be seen for the alloy/solution interface in both media:
the imaginary and real components of impedance decrease
as the potential becomes more positive than the Ecorr. The proposed equivalent circuit is presented in Figure 2C,
where Rs correspond to the electrolyte resistance, CPEf and
Rf correspond respectively to the constant phase element Figure 1. Potentiodynamic polarizations curves of CoCrW alloy in
artificial saliva and 0.15 mol.L-1 sodium chloride at 37 oC. Panels
A, B and C show the chronoamperograms response at each potential
application. In vitro cytotoxicity test of CoCrWAlloyv 183
Figure 2. Impedance spectra of CoCrW alloy at sodium chloride (A) and artificial saliva. B: Bode plots registered at different potentials (Ecorr
and 0.25 V). (•) experimental values, (___) simulation. C: Electrical equivalent circuits used to fit the experimental data at different potentials. Figure 2. Impedance spectra of CoCrW alloy at sodium chloride (A) and artificial saliva. B: Bode plots registered at different potentials (Ecorr
and 0.25 V). (•) experimental values, (___) simulation. C: Electrical equivalent circuits used to fit the experimental data at different potentials. Figure 2. Impedance spectra of CoCrW alloy at sodium chloride (A) and artificial saliva. B: Bode plots registered at different potentials (Ecorr
and 0.25 V). (•) experimental values, (___) simulation. C: Electrical equivalent circuits used to fit the experimental data at different potentials. 183 Braz Dent J 27(1) 2016 Cytotoxicity test was made and the results are in Figure
4; the cytotoxicity evaluation was performed by neutral red
uptake assay. Surface Characterization SEM and EDS analyses were made to evaluate the
existence of non-metallic inclusions. Different surface
regions were analyzed using EDS after polishing with
1 mm diamond paste. Figure 3 presents three different
regions, 1, 2 and 3. According to EDS analysis the region
1 presents the bulk composition, region 2 corresponds to
niobium carbides and region 3 is due to the presence of
silicon and manganese oxides. Electrochemical Results Positive and negative controls were used to
confirm the adequate performance of the test procedure
and/or to evaluate the results of CoCrW alloy, as well as to
control the cell sensitivity, extraction efficiency and other
test parameters. The sample that presents cell viability
above the IC50(%) line is considered nontoxic and under
the IC50(%) line it is toxic. (pseudo-capacitance) and the polarization resistance for
the oxidation reaction trough the protective film., CPEdc and
Rct correspond, respectively, to the double layer constant
phase element and the charge transfer resistance of the
CoCrW alloy oxidation. The same circuit was proposed for
Ecorr and at 0.25 VSCE in both media. Discussion Figure 1A shows that the Ecorr in saline solution is more
positive, suggesting a better passive film formation when
compared to artificial saliva medium. The CoCrW alloy (Fig. 1A) showed small values of current density (1 to 10 µA/
cm2) for a large analyzed potential range indicating good
characteristics of the passive film formed in both media. The
CoCrW alloy presents lower passivating current densities in
NaCl medium, suggesting more passivating film for sodium
chloride than in artificial saliva. The chronoamperometric
curves (Figs. 1B and 1C) showed constant current density
values even at high potentials, indicating the absence of
pitting corrosion. The results have shown that CoCrW alloy
presents only generalized corrosion at 0.9 V with high
current density values in both media. XPS was used to investigate the presence of Cr and
Co in the composition of the passive film. Table 3 shows
the absolute potentials of the components in the passive
film. The chromium content of the surface is higher than
cobalt in both media and is higher in 0.15 mol.L-1 NaCl
when compared to saliva. The results showed presence of
chromium (III) and cobalt oxides in the passive film. K.S. Souza et al. In vitro cytotoxicity test of CoCrWAlloyv In this study the tested alloy did not present toxic effects
even at 100% extract concentration. All the viability curves
are above the cytotoxicity index line, which means that the
CoCrW alloy showed no cytotoxicity effects in this assay. 7. House K, Sernetz F, Dymock D, Sandy JR, Ireland AJ. Corrosion of
orthodontic appliances – should we care? Am J Orthod Dentfac Orthop
2008:133:584-592. 7. House K, Sernetz F, Dymock D, Sandy JR, Ireland AJ. Corrosion of
orthodontic appliances – should we care? Am J Orthod Dentfac Orthop
2008:133:584-592. 8. Ouerd A, Alemany-Dumont C, Normand B, Szunerits S. Reactivity
of CoCrMo alloy in physiological medium: Electrochemical
characterization of the metal/protein interface. Electrochim Acta
2008;53:4461-4469. In this study the CoCrW alloy presented good behavior
with the formation of a uniform passive film rich in
chromium (III) oxide and a lower amount of cobalt oxides. The film formed in 0.15 mol.L-1 NaCl is more capacitive
and resistive than the one formed in artificial saliva. The chemical complexity of the artificial saliva probably
contributes to greater instability of the passive film. SEM
and EDS show that the CoCrW alloy presents niobium
carbide and silicon and manganese oxides as non-metallic
inclusions. Cytotoxicity tests validated the non-cytotoxic
effect of the alloy according to the used methodology. 9. Contu F, Elsener B, Bohni H. Electrochemical behavior of CoCrMo
alloy in the active state in acidic and alkaline buffered solutions. J
Electrochem Soc 2003:150;419-424. 10. Wang Q, Huang C, Zhang L. Microstruture and tribological properties of
plasma nitriding cast CoCrMo alloy. J Mater Sci Technol.2012;28:60-66. 11. Hedberg YS, Qian B, Shen Z, Virtanen S, Wallinder IO. In vitro
biocompatibility of CoCrMo dental alloys fabricated by selective laser
melting. Dent Mater 2014;30:525-534. 12. Giacchi JV, Morando CN, Fornaro O, Palacio HA. Microstructural
characterization of as-cast biocompatible CoCrMo alloys. Mater
Charact 2011;62:53-61. 13. Karaali A, Mirouh K, Hamamda S, Guiraldenq P. Microstructural study
of tungsten influence on Co-Cr alloys. Materials Mater Sci Eng A. 2005;390:255-259. The results of this study suggest that CoCrW alloys
from the electrochemical and cytotoxicity point of view
may be used as biomaterials to be applied in prosthesis
on dental implants. 14. Karaali A, Mirouh K, Hamamda S, Guiraldenq P. Effect of tungsten 0–8
wt.% on the oxidation of Co–Cr alloys. Comput Mater Sci 2005;33:37-
43. 15. Leung VWH, Darvell BW. Calcium phosphate system in saliva-like
media. J Chem Soc Faraday Trans. 1991;87:1759-1764. 16. K.S. Souza et al. K.S. Souza et al. Cytotoxicity Test
Table 3. X-ray photoelectron spectroscopy (XPS) results for CoCrW
alloy after immersed
Bond Energy (eV)
Cr 2p3/2
Co 2 p3/2
0.15 mol/L NaCl
573.4 (11%)
576.5 (89%)
Cr(III)
Cr0
777.6 (56%)
780.2 (44%)
CoOx
Co0
Artificial Saliva
573.1 (11%)
576.2 (89%)
Cr(III)
Cr0
777.2 (53%)
779.6 (47%)
CoOx
Co0 The characteristics of the passive film were investigated
by EIS at Ecorr and 0.25 VSCE. The results are presented
in Figures 2A and 2B. Bode phase diagrams are typical
of passive systems with high values of both real and
imaginary parts of the impedance at low frequencies. In
the Bode diagrams the values of |Z| were about 105 Ω/
cm² and the angle greater than 75º, showing the high
resistive and capacitive characteristics of the formed film. Bode phase diagrams point out to the same mechanism
in artificial saliva and sodium chloride. Figure 2C also Figure 4. CoCrW alloy viability curves in the cytotoxicity assay by
the neutral red uptake methodology. Figure 3. SEM and EDS images of CoCrW alloy samples polished
through 1 mm diamond paste. Figure 4. CoCrW alloy viability curves in the cytotoxicity assay by
the neutral red uptake methodology. Figure 3. SEM and EDS images of CoCrW alloy samples polished
through 1 mm diamond paste. 184 Braz Dent J 27(1) 2016 que a liga apresenta inclusões não metálicas de carbeto de nióbio, de
óxidos de silício e de manganês. Os resultados de XPS indicaram que o
cobalto não contribui significativamente para a formação do filme passivo. O teste de citotoxicidade mostrou que a liga CoCrW não se apresenta
citotóxica. Estes resultados sugerem que a liga estudada pode ser usada
como biomaterial a ser aplicado como prótese sobre implantes dentários. shows the fitting of the experimental results according to
simulated equivalent circuits. Good agreement between the
experimental results and the equivalent circuits proposed
is observed on the entire range of studied frequencies,
suggesting qualitatively the same mechanism in both media. The film formed in sodium chloride medium (Fig. 2B) has
better quality than that formed in artificial saliva (Fig. 2A). In vitro cytotoxicity test of CoCrWAlloyv Moulder JF, Stickle WF, Sobol PE, Bomben KD. Handbook of X-ray
photoelectron spectroscopy. Chastain, J. Perkin-Elmer Corporation,
Physical Electronics Division, Eden Prairie, MN, USA, 1992. Acknowledgements The authors acknowledge CNPq for the granted scholarship. The composition of the passive film was analyzed by
XPS at Ecorr. The amount of chromium and cobalt in the
passive film formed in sodium chloride is greater than in
artificial saliva. The protective film is formed by chromium
oxide in both media (20). These results were also obtained
by Hodgson et al. (21). In CoCr-based alloys, only at
high potential will be found another state of chromium
oxidation, such as Cr (VI) (21,22).. The chemical shift between
Co2+ and Co3+ is very small (22-24), indicating a difficulty to
characterize the oxidation state on the metallic surface. The
low amount of cobalt in the film suggests that it diffuses
into the solution and it does not significantly contribute
to the oxide phase. References 1. Anusavice KJ. Phillips materiais dentários, Guanabara Koogan, 10 ed.,
Guanabara Koogan, Rio de Janeiro; 1998. 1. Anusavice KJ. Phillips materiais dentários, Guanabara Koogan, 10 ed.,
Guanabara Koogan, Rio de Janeiro; 1998. 2. Balamurugan A, Rajeswari S, Balossier G, Rebelo AHS, Ferreira RJMF. Corrosion aspects of metallic implants – an overview. Mater Corros
2008;59:855-869. 2. Balamurugan A, Rajeswari S, Balossier G, Rebelo AHS, Ferreira RJMF. Corrosion aspects of metallic implants – an overview. Mater Corros
2008;59:855-869. 3. Bumgardner JD, Lucas LC. Cellular response to metallic ions released
from nickel-chromium dental alloys. J Dent Res 1995;74:1521-1527. 3. Bumgardner JD, Lucas LC. Cellular response to metallic ions released
from nickel-chromium dental alloys. J Dent Res 1995;74:1521-1527. 4. Marananche C, Hornberger H. A proposal for the classification of
dental alloys according to their resistance to corrosion. Dent Mater
2007;23:1428-1437. 5. Ameer MA, Khamis E, Al-Motlaq M. Electrochemical behaviour of
recasting Ni-Cr and Co-Cr non-precious dental alloys. Corros Sci
2004;46:2825-2836. 5. Ameer MA, Khamis E, Al-Motlaq M. Electrochemical behaviour of
recasting Ni-Cr and Co-Cr non-precious dental alloys. Corros Sci
2004;46:2825-2836. 6. Upadhyay D, Panchal MA, Dubey RS, Srivastava VK. Corrosion of alloys
used in dentristry: a review. Mater Sci Eng 2006;432:1-11. 6. Upadhyay D, Panchal MA, Dubey RS, Srivastava VK. Corrosion of alloys
used in dentristry: a review. Mater Sci Eng 2006;432:1-11. 21.
International Organization for Standardization (ISO) 10993 - Biological
evaluation of medical devices, Part 5: Tests for In Vitro Cytotoxicity,
2009. 2178.
24. Milosev I, Strehblow HH. The composition of the surface passive
film formed on CoCrMo alloy in simulated physiological solution.
Electrochim Acta 2003;48:2767-2774. 22. Muñoz AI, Mischler S. Interactive effect of albumin and phosphate
ions on the corrosion of CoCrMo implant alloy. J Electrochem Soc
2007;154:562-570. 23. Hodgson AW E, Kurz S, Virtanen S, Fervel V, Olsson COA, Mischler S.
Passive and transpassive behavior of CoCrMo in simulated biological
solutions. Trans Nonferrous Metals Society of China 2004;49:2167- Resumo Estudos eletroquímicos, caracterização de superfície e teste de
citotoxicidade in vitro foram realizados da liga CoCrW em meios de saliva
artificial e NaCl 0,15 mol.L-1, com o objetivo de avaliar a sua aplicação
como material de prótese dentária. Foram usadas como técnicas, curvas
de polarização anódica, medidas cronoamperométricas, espectroscopia de
impedância eletroquímica (EIE), microscopia eletrônica de varredura (MEV),
espectroscopia por energia dispersiva de raios X (EDS) e espectroscopia
fotoeletrônica de raios X (XPS). O teste de citotoxicidade também foi
realizado. O comportamento eletroquímico da liga CoCrW foi comparado
nos dois meios estudados desde o potencial de corrosão (Ecorr) até uma
sobretensão anódica de 600 mV. Foi observado, a partir de medidas
eletroquímicas, que a liga CoCrW se encontra passivada em uma ampla
faixa de potencial e que a sobretensões mais elevadas apresenta apenas
corrosão generalizada nos dois meios. Análises por MEV e EDS mostraram 17. Rogero SO, Hilga OZ, Saiki M, Correia OV, Costa I. Cytotoxicity due to
corrosion of ear piercing studs. Toxicol In Vitro 2000;14:497-504. 18. Jaimes RFVV, Afonso MLCA, Rogero SO, Agostinho SML, Barbosa CA. New material for orthopedic implants: electrochemical study of nickel
free P558 stainless steel in minimum essential medium. Mat Letts
2010;64:1476-1479. 19. Afonso MLCA, Jaimes RFVV, Rogero SO, Nascente PAP, Agostinho SML. Surface characterization, electrochemical behaviour and cytotoxicity
of UNS S31254 stainless steel for orthopedic applications. Mat Letts
Volume 2015;148:71-75 20. Ossa CPO, Rogero SO, Tschiptschin APJ. Cytotoxicity study of plasma-
sprayed hydroxyapatite coating on high nitrogen austenitic stainless
steels. J Mater Med 2006;17:1095–1100. 185 Braz Dent J 27(1) 2016 22. Muñoz AI, Mischler S. Interactive effect of albumin and phosphate
ions on the corrosion of CoCrMo implant alloy. J Electrochem Soc
2007;154:562-570. Received September 5, 2015
Accepted January 7, 2016 23. Hodgson AW E, Kurz S, Virtanen S, Fervel V, Olsson COA, Mischler S. Passive and transpassive behavior of CoCrMo in simulated biological
solutions. Trans Nonferrous Metals Society of China 2004;49:2167- K.S. Souza et al. 186
|
https://openalex.org/W2954545786
|
https://ejournal2.undip.ac.id/index.php/ihis/article/download/5095/2875
|
Indonesian
| null |
Kiai Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara 1942-1972
|
IHiS (Indonesian Historical Studies)
| 2,019
|
cc-by-sa
| 6,733
|
Abstract This study focuses on the life, ideas, and role of Kiai Ahmad Fauzan in developing
Islamic teachings and national values. Islam and nationalism are two things that
interconnected and not contradictory. In Indonesian history, the two of them
caused turmoil, even opposition. The purpose of this study is to prove the return
of the Moslem spirit which is in line with the development of local religious
leaders, primarily through case studies of local scholars in Jepara, such as Kiai
Ahmad Fauzan. This study used a historical method, including heuristics, source
criticism, interpretation, and historiography. Kiai Ahmad Fauzan was a leader of
the Nahdlatul Ulama (NU) who fought through education and politics to uphold
the Ahlussunah wal Jamaah (Aswaja) ideology in Jepara. Fauzan's Islamic and
national ideas can be seen from syair [poems] conveyed to the public. Syair became
a media for propaganda for Kiai Ahmad Fauzan in spreading the religious
understanding of Islam Aswaja. It is delivered to the community as reminder and
awareness of harmonious religious and national values. His role in the religious
and socio-political fields was seen when Japan began occupying Jepara in 1942. He was the target of arrest because of his role as a cleric. Its leadership formed
from religious roles carried out mainly through madrasa and da'wah by traveling
from one village to another. Kiai Ahmad Fauzan was involved in socio-religious
organizations such as the Indonesian Islamic Assembly (MIAI), Indonesian
Muslim Council (Masyumi), and NU, especially during the 1955 elections. Kiai
Ahmad Fauzan was also trusted by the government to be the first leader of the
Ministry of Religion in Jepara after independence revolution. Keywords: Kiai Ahmad Fauzan; Ideas; Religious Role; Local Politics. Disetujui/ Accepted:
4 Juli 2019 Diterima/ Received:
11 Juni 2019 M. Dalhar,* Yety Rochwulaningsih, Dhanang Respati Puguh
Program Studi S2 Ilmu Sejarah, Fakultas Ilmu Budaya, Universitas Diponegoro *Alamat korespondensi: mbahdalhar7@gmail.com Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Abstrak Kajian ini berfokus pada kehidupan dan pemikiran serta peranan Kiai Ahmad
Fauzan dalam mengembangkan ajaran Islam dan nilai-nilai kebangsaan. Islam
dan nasionalisme adalah dua hal yang saling berhubungan dan tidak
bertentangan. Hal tersebut tampak dalam perjalanan bangsa Indonesia. Akan
tetapi, dalam sejarah perjalanan bangsa Indonesia, antara keduanya saling
menimbulkan gejolak, bahkan pertentangan. Tujuan kajian ini adalah untuk
membuktikan kembalinya semangat Islam yang menjadi mayoritas di Indonesia
dan nasionalisme yang seiring sejalan dengan perkembangan tokoh lokal
keagamaan, khususnya melalui studi kasus ulama lokal di Kabupaten Jepara,
yaitu Kiai Ahmad Fauzan. Metode yang digunakan dalam kajian ini adalah
metode sejarah yang meliputi empat tahap, yaitu heuristik, kritik sumber,
interpretasi, dan historiografi. Kiai Ahmad Fauzan adalah tokoh NU Jepara yang
berjuang melalui jalur pendidikan dan politik untuk menegakkan paham Aswaja
di Jepara. Pemikiran Islam dan kebangsaan Fauzan terlihat dari syair-syair yang
disampaikan kepada masyarakat. Syair menjadi media dakwah bagi Kiai Ahmad
Fauzan dalam menyebarkan paham keagamaan Islam Aswaja. Syair-syair yang
disampaikan kepada masyarakat lebih bersifat pengingat dan penyadaran atas
nilai-nilai keagamaan dan kebangsaan yang selaras. Peranannya di bidang Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 | 29 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 keagamaan dan sosial-politik terlihat ketika Jepang mulai menduduki Jepara pada
1942. Ia menjadi target penangkapan karena perannya sebagai ulama. Keulamaan
dan kepemimpinannya terbentuk dari peran-peran keagamaan yang dilakukan
terutama melalui madrasah dan dakwah dengan cara berkelana dari satu desa ke
desa yang lain. Kiai Ahmad Fauzan terlibat dalam organisasi sosial-keagamaan
seperti Majelis Islam A'la Indonesia (MIAI), Majelis Syuro Muslimin Indonesia
(Masyumi), dan NU, terutama saat Pemilu 1955. Kiai Ahmad Fauzan juga
dipercaya pemerintah untuk memimpin Kementerian Agama Kabupaten Jepara
yang pertama setelah berakhirnya revolusi kemerdekaan Indonesia. Kata Kunci: Kiai Ahmad Fauzan; Pemikiran; Kiprah Keagamaan; Politik Lokal. Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Pendahuluan Adanya fenomena pemberontakan yang terjadi di sebuah negara bukan semata-mata
disebabkan oleh faktor keterpurukan ekonomi, ideologi, atau krisis politik. Krisis
kebangsaan memegang peranan penting dalam keruntuhan sebuah bangsa dan
negara. Di tengah berbagai krisis yang ada di sebuah negara, membangun rasa
kebangsaan yang kuat merupakan hal yang sangat penting dilakukan. Atas dasar itu,
menjadi penting bahwa berbagai krisis yang ada diselesaikan agar kepercayaan (trust)
warga terbangun kepada negara. Krisis kebangsaan diawali dari keraguan untuk
hidup bersama sebagai satu negara bangsa yang menjadi alternatif untuk mencapai
kemakmuran lahir dan batin (Sianipar, 2002). Sejarah menunjukkan bahwa Pasca-Proklamasi, terjadi beberapa kali upaya
makar yang dilakukan oleh sekelompok orang yang mengatasnamakan agama
(Santoso, 2014). Pemberontakan Kartosuwiryo melalui Darul Islam dan Tentara Islam
Indonesia (DI-TII) di Jawa Barat menjadi bukti proses berbangsa dan bernegara belum
selesai. Masih ada kelompok-kelompok yang ingin mendirikan negara dengan
ideologi tertentu di dalam sebuah negara yang sudah merdeka. Sikap tersebut
merupakan suatu bentuk resistensi dan separatisme yang dipandang sebagai sikap
tidak loyal dan pengingkaran terhadap “konsensus luhur” dari para pendiri bangsa
(Sulistiyono, 2015). Untuk mengatasi krisis kebangsaan yang terjadi, perlu dilakukan upaya bersama
dan terintegrasi. Dalam permasalahan tersebut, sejarah memiliki peran strategis untuk
membangun kembali rasa kebangsaan sebagai warga negara yang mulai luntur. Sejarah tidak semata memberikan pengetahuan, fakta, dan kronologi, tetapi juga
mampu berkontribusi untuk pembangunan, memperkokoh solidaritas bangsa,
membangkitkan kebanggaan nasional (national pride), dan menumbuhkan aspirasi
terhadap masa depan yang gemilang (Kartodirdjo, 1993). Kontribusi ulama dalam perjalanan sejarah bangsa Indonesia tidak dapat
diabaikan begitu saja. Banyak ulama yang terlahir dari lembaga pendidikan pesantren
bersikap nonkooperatif dan turut berjuang mengusir penjajah. Pesantren sebagai
ruang kaderisasi ulama tersebar luas di daerah-daerah pedesaan di Jawa, terutama
setelah Mataram Islam berdiri di pedalaman Jawa. Sejumlah pesantren berdiri di
sekitar keraton, khususnya pascapemerintahan Sultan Agung pada abad ke-17. Salah
satu pesantren terkemuka adalah pesantren Tegalsari di bawah asuhan Kiai Hasan
Besari di Ponorogo. Pesantren ini berdiri di atas tanah bebas pajak (perdikan) dan diberi | 30 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 tugas oleh keraton untuk mendidik para putra raja dan bangsawan (Suryo, 2009). Pesantren juga turut mengatasi kelangkaan dan diskriminasi pendidikan yang ada
pada masa pemerintah kolonial Belanda. Pesantren mempunyai akar sejarah yang
panjang sekalipun sebagian besar keberadaannya hanya dapat dilacak dari abad ke-
19 hingga ke-20 (Kuntowijoyo, 1991). Kehadiran pendidikan Barat pada awal abad ke-20 menjadikan sekolah sebagai
pendidikan tunggal sempat menggeser peran strategis pesantren. Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Pendahuluan Kondisi tersebut
menjadikan pendidikan Islam masuk dalam isolasi dan malah perlu menyesuaikan
diri dengan syarat-syarat pemerintah untuk mendapatkan bantuan dan pengakuan
resmi (Steenbrink, 1984). Dikotomi pendidikan yang memisahkan antara ilmu agama
dengan ilmu pengetahuan umum tidak menghentikan langkah pesantren sebagai
ruang kaderisasi para ulama. Di dalam kurikulum pesantren, ilmu pengetahuan
umum dan agama telah terintegrasi (Siradj, 2015). Untuk itu, ketika politik diarahkan
pada paradigma Barat, pesantren terus berjalan dengan paradigmanya sendiri. Kemudian lahir banyak tokoh yang bersikap nonkooperatif terhadap pemerintah
kolonial, seperti Pangeran Sambernyowo, Pangeran Diponegoro, Paku Buwono VI,
K.H. Ahmad Rifai, dan K.H. Hasyim Asy’ari (Florida, 2003, p. 71-75). Dalam
sejarahnya, hampir seluruh perlawanan terhadap penjajah dilakukan oleh pimpinan
pesantren. Kalau pun dilakukan oleh keraton atau negara, tentu melibatkan para kiai
atau ulama dan santri dari pesantren (Siradj, 2015). Peran ulama terus berlanjut sampai Indonesia merdeka pada 1945. Momentumnya adalah ketika para ulama yang tergabung dalam wadah NU
mengeluarkan "Resolusi Jihad" pada 22 Oktober 1945 (Wahid, 1999). Resolusi ini
menegaskan sikap para ulama untuk membela kemerdekaan dari upaya kolonialis
yang akan merebut kembali kemerdekaan Indonesia. Isinya berupa seruan bahwa
jihad fi sabilillah, mempertahankan kemerdekaan dari tangan penjajah, adalah wajib
(fardlu 'ain) hukumnya, terutama bagi kaum muslimin yang berada di radius 80 km
dari wilayah pertempuran. Bagi kaum muslimin yang meninggal dalam jihad ini diberi
status sebagai syahid. Fakta tersebut kembali menegaskan bahwa ulama memiliki peranan yang besar
bagi negara ini. Peranan ulama yang mengajarkan keteladanan dan nilai-nilai
kebangsaan perlu dihadirkan kembali untuk meneguhkan dan menggugah semangat
cinta tanah air atau nasionalisme, mengingat pasca-Reformasi 1998, ada banyak
kelompok berambisi ingin mendirikan negara Islam (khilafah) di Indonesia. Keberadaan negara dalam negara merupakan sebuah pengingkaran terhadap
konstitusi dan kesepakatan final para pendiri bangsa. Hal ini sebagaimana sudah
dikukuhkan dalam konstitusi negara yang menyebutkan bahwa bentuk negara yaitu
Negara Kesatuan Republik Indonesia (NKRI) tidak dapat diubah, artinya sudah
menjadi harga mati (Hady, 2010). Ulama sebagai bagian dari komponen bangsa ini memiliki kontribusi strategis
dalam mempersatukan umat. Otoritas ulama yang bersumber dari syariat agama
memfungsikan mereka secara teologis dan sosiologis sebagai penjaga (custodians) dan
penafsir hukum Tuhan (Siregar, et al., 2013). Kedalaman ilmu agama yang dimiliki dan | 31 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 peranan yang dimainkan di tengah masyarakat menjadikan ulama termasuk dalam
kelompok cendekiawan yang berperan sebagai elite kultural dalam masyarakat
muslim. Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Pendahuluan Sebagai kelompok kultural dan kelompok nonpolitis, ulama secara strategis
menduduki posisi sebagai kelompok mediator yang menjembatani jurang pemisah
antara penguasa dan masyarakat (Azra,1989). Penelusuran terhadap peranan ulama sebagaimana Kiai Ahmad Fauzan
merupakan hal yang sangat penting dilakukan mengingat generasi sekarang mulai
kehilangan contoh teladan (uswatun hasanah). Di samping itu, penggalian ini juga
untuk menemukan bagaimana sebenarnya peranannya di bidang keagamaan dan
sosial-politik, karena selama ini kisah-kisah tentang perjuangannya masih berupa
kisah tutur (oral history). Harapannya karya ini dapat berkontribusi untuk
memperkaya khasanah kebangsaan terutama peranan ulama dalam perjuangan
mengisi dan menegakkan kedaulatan NKRI. Kiai Ahmad Fauzan adalah seorang ulama di Jepara yang berjasa dalam mengisi
kemerdekaan baik melalui pikiran, yaitu syair-syair dan peranan di bidang
keagamaan dan sosial-politik. Sebagai bentuk apresiasi, namanya diabadikan sebagai
nama salah satu jalan di Jepara. Akan tetapi, belum banyak hasil penelitian yang
menjelaskan tentang biografi, pemikiran, serta peranan yang dilakukan sepanjang
hidupnya. Biografi, pemikiran, dan kontribusinya dalam memberdayakan masyarakat
di Jepara masih dapat ditemukan melalui ingatan kolektif masyarakat maupun
kesaksian dari keluarga dan para murid Ahmad Fauzan. Sumber-sumber tersebut
tentu saja tidak utuh dan semakin lama akan hilang seiring berjalannya waktu. Metode Sebagai bagian dari kajian sejarah, penelitian ini telah melalui empat tahapan dalam
metode sejarah yaitu heuristik, kritik, interpretasi sumber, dan historiografi. Terdapat
dokumen-dokumen penting yang berhasil ditemukan dan digunakan untuk
menganalisis pemikiran Kiai Ahmad Fauzan, salah satunya adalah dokumen arsip
milik keluarga berupa catatan silisilah dan transkrip ceramah-ceramah yang
disampaikannya. Catatan ini sebagian adalah hasil tulis ulang dari Hj. Lathifah, salah
seorang putri Kiai Ahmad Fauzan yang masih melanjutkan perjuangan dakwahnya. Catatan-catatan syair yang ada masih menggunakan bahasa Jawa dan huruf Arab
(baca: pegon). Di samping catatan-catatan syair, digunakan pula sebagai sumber primer adalah
"Catatan Riwayat Hidup al-Maghfurlah Kiai Ahmad Fauzan" dalam peringatan haul
ke-22 tahun 1994 dan karya (baca: kitab) yang ditulis Kiai Ahmad Fauzan, Allfiyah al-
Ghazali. Dari riwayat hidup tersebut dapat dijadikan sebagai pedoman awal
penyusunan profil Kiai Ahmad Fauzan, sedangkan dari karya syair berbahasa Arab
tersebut didapatkan sebuah sintesis pemikiran dari beberapa pemikiran lain yang
pernah disampaikan. Di dalam karyanya tersebut dijelaskan banyak hal seperti bab
tentang keutamaan ilmu, adab, keadilan, dan sikap hati. Di dalam kitab tersebut juga
diberikan nasihat (baca: pitutur) tentang nilai-nilai dasar bagi seorang sufi agar tidak
terjadi penyimpangan. | 32 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 Sumber-sumber primer lainnya yang digunakan adalah arsip-arsip yang
tersimpan di Kementerian Agama (Kemenag) Kabupaten Jepara, Kantor Arsip Daerah
Kabupaten Jepara, dan Depo Arsip Harian Suara Merdeka tahun 1950-an. Arsip yang
tersedia antara lain foto Ahmad Fauzan sewaktu menjadi Pimpinan Kemenag Jepara
tahun 1950-an, Kartu Anggota NU, Surat Keterangan Pegawai Kemenag tahun 1950,
dan Kartu Partai NU, dan beberapa pemberitaan seputar Pemilu 1955 di Kabupaten
Jepara serta dinamika perpolitikan NU di tingkat nasional. Selain sumber tertulis, wawancara juga dilakukan untuk memperkaya fakta-
fakta sumber dokumen. Wawancara dilakukan dengan putra-putri Ahmad Fauzan
dan beberapa orang yang mengetahui tentang peristiwa yang sedang dikaji. Hal ini
masih sangat dimungkinkan karena para saksi sejarah masih ada sampai saat ini. Meskipun usia mereka tidak lagi muda – hal ini sangat memengaruhi memori mereka
– namun dengan adanya perbandingan dari narasumber yang diwawancara akan
menghasilkan sumber yang kredibel. Wawancara ini digunakan untuk melengkapi
informasi-informasi yang ada di dalam sumber tertulis atau arsip. Dengan adanya
pembanding ini diharapkan informasi yang diperoleh dapat benar-benar valid. 1“Silsilah Keluarga Ahmad Sanwasi Penggung”, ditulis di Balekambang pada 29
Ramadhan 1424 H/ 2003 M oleh Kiai Zaini Dahlan (Sumber tidak diterbitkan). Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Figur Kiai Ahmad Fauzan: Ulama Lokal yang Nasionalis Kontribusi ulama tidak terbatas dalam lingkup politik dan level nasional semata. Terdapat ribuan, bahkan jutaan ulama di berbagai pelosok negeri ini yang belum
tercatat dalam sejarah. Padahal, mereka turut berjasa dalam mengajarkan nilai-nilai
yang secara substansi tidak mempertentangkan agama dan negara. Mereka
menyebarkan paham-paham yang menjadikan negara aman, damai, baldatun
thayyibatun warabbun ghafur. yy
g
f
Kiai Ahmad Fauzan (1905-1972) merupakan satu di antara ulama yang memiliki
jasa besar dalam pemberdayaan masyarakat. Kiai Ahmad Fauzan, begitu ia sering
disebut, merupakan salah seorang ulama kharismatik di Kabupaten Jepara. Ia
dilahirkan di sebuah desa kecil di Distrik Mayong pada 1905. Ia adalah putra keempat
dari lima bersaudara yang terlahir dari pasangan Haji Abdurrasul dengan Ny. Thohiroh. Secara garis keturunan, Kiai Ahmad Fauzan memiliki nasab dari Kasunanan
Surakarta. Kakeknya, Haji Ahmad Sanwasi merupakan seorang ulama keraton yang
meninggalkan istana saat terjadi Perang Diponegoro (1825-1830). Haji Ahmad
Sanwasi adalah sahabat dekat Kiai Umar Al-Samarani yang tidak lain adalah ayah dari
Kiai Sholeh Darat. Hal tersebut dapat dilihat dari periode perjuangan dakwah dari
keduanya dan di tempat yang sama pula, yakni Mayong. Bahkan menurut salah satu
sumber, Haji Ahmad Sanwasi adalah menantu dari Kiai Umar Al-Samarani, yaitu
menikah dengan Ny. Darojah, kakak dari Kiai Soleh Darat.1 Kiai Umar, selain ulama
yang cukup terpandang di kawasan pantai utara Jawa, ia juga adalah orang
kepercayaan Pangeran Diponegoro. Kiai Umar berserta kawan, kolega, dan santri- 1“Silsilah Keluarga Ahmad Sanwasi Penggung”, ditulis di Balekambang pada 29
Ramadhan 1424 H/ 2003 M oleh Kiai Zaini Dahlan (Sumber tidak diterbitkan). hmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 | 33 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 santrinya berjuang gigih mempertahankan kehormatan Tanah Air dari jajahan
Belanda (Ulum dan Arbi, 2016). Melalui ayahnya, yaitu Haji Abdurrasul, Fauzan banyak belajar tentang agama
dan sosial-kemasyarakatan. Akan tetapi, masa belajar dengan ayahnya tidak lama
karena sang ayah wafat saat menunaikan ibadah haji di Tanah Suci. Proses belajar
masa berikutnya berlanjut di pesantren2. Tercatat beberapa pesantren yang pernah
dijadikan sebagai tempat belajarnya, yaitu Pesantren Balekambang, Pesantren
Kasingan Rembang, Pesantren Gemiring, dan Pesantren Tayu. Di samping itu, pada
1924 Kiai Ahmad Fauzan juga berkesempatan melanjutkan perjalanan intelektual ke
Makkah dan Madinah (haramain) selama kurang lebih dua tahun. Berbagai perjalanan
intelektual itu yang di kemudian hari memengaruhi pemikiran Kiai Ahmad Fauzan3. Ibu Ahmad Fauzan, Ny. Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Figur Kiai Ahmad Fauzan: Ulama Lokal yang Nasionalis Thohiroh adalah salah seorang cucu dari Syekh
Muhammad Arif Sendangsari yang merupakan ulama terkemuka yang hidup pada
awal abad ke-19, sezaman dengan Pangeran Diponegoro. Nama Sendangsari yang
melekat di belakang namanya merupakan nama tempat (baca: dukuh/dusun)
perjuangan dakwahnya. Ia adalah seorang ulama dari Hadramaut Yaman dengan
marga Adaniy.4 Perjuangan dakwah Islam dan sikap antipenjajah dapat dilihat
historiografi lokal setempat.5 Salah seorang istri Ahmad Fauzan, Ny. Mukarromah, merupakan cucu dari Haji
Sulaiman dari Jakenan, Pati. Nama aslinya adalah Raden Ngabehi Brontodiwiryo. Ia
juga diyakini sebagai salah seorang ulama Keraton Surakarta yang bersikap
nonkooperatif terhadap Belanda sebagaimana Kiai Umar Al-Samarani dan Haji
Ahmad Sanwasi. Banyak anak cucu dari Kiai Sulaiman yang menetap di Kabupaten
Pati, Jepara, dan sekitarnya untuk meneruskan perjuangan mendakwahkan Islam.6 Terjadinya revivalisme Islam abad ke-19 yang terwujud pada pertumbuhan
pesantren di pedesaan, pendirian tarekat-tarekat, jumlah orang naik haji yang 2Kiai Ahmad Fauzan turut ke Makkah dalam usia belum baligh. Ia diperkirakan berusia
sebelas tahun (1916). Ketika pulang, ia sempat terpisah dari rombongan di Singapura. Keterangan dari putra sulung Kiai Ahmad Fauzan, Kiai Zainurrohman. Kini ia tinggal di
Purwokerto. Informasi yang sama juga disampaikan oleh Hj. Latifah Fauzan. Wawancara
pada 1 Februari 2018. 3Catatan “Riwayat Hidup al-Maghfurlah Kiai Ahmad Fauzan” dalam peringatan Haul
ke-22 bulan September tahun 1994 (sumber tidak diterbitkan). 4Makamnya di Desa Banjaran, Bangsri, Jepara. Silsilahnya Arif bin Hasyim bin Abdul
Muhit bin Hasyim bin Syihabuddin bin Semith bin Arif bin Sholih bin Ahmad bin Ali bin
Ahmad bin Hamah bin ainuddin bin Abdullah al-Adniy (Yaman). Keterangan ini berasal dari
karya Syekh Muhammad Arif yang ditulis pada 8 Januari 1809. Koleksi digitalisasi naskah di
Universitas Nahdlatul Ulama (Unisnu) Jepara. 4Makamnya di Desa Banjaran, Bangsri, Jepara. Silsilahnya Arif bin Hasyim bin Abdul
Muhit bin Hasyim bin Syihabuddin bin Semith bin Arif bin Sholih bin Ahmad bin Ali bin
Ahmad bin Hamah bin ainuddin bin Abdullah al-Adniy (Yaman). Keterangan ini berasal dari
karya Syekh Muhammad Arif yang ditulis pada 8 Januari 1809. Koleksi digitalisasi naskah di
Universitas Nahdlatul Ulama (Unisnu) Jepara. 5Kiai Muhammad Masduri, “Syair Sejarah Simbah Muhammad Arif”, ditulis pada 1987
dengan huruf Arab dan berbahasa Jawa (pegon). Syair ini dibacakan rutin pada peringatan
haulnya setiap 10 Muharram (sumber tidak diterbitkan). 5Kiai Muhammad Masduri, “Syair Sejarah Simbah Muhammad Arif”, ditulis pada 1987
dengan huruf Arab dan berbahasa Jawa (pegon). Syair ini dibacakan rutin pada peringatan
haulnya setiap 10 Muharram (sumber tidak diterbitkan). 2Kiai Ahmad Fauzan turut ke Makkah dalam usia belum baligh. Ia diperkirakan berusia
sebelas tahun (1916). Ketika pulang, ia sempat terpisah dari rombongan di Singapura.
Keterangan dari putra sulung Kiai Ahmad Fauzan, Kiai Zainurrohman. Kini ia tinggal di
Purwokerto. Informasi yang sama juga disampaikan oleh Hj. Latifah Fauzan. Wawancara
pada 1 Februari 2018. 3Catatan “Riwayat Hidup al-Maghfurlah Kiai Ahmad Fauzan” dalam peringatan Haul
ke-22 bulan September tahun 1994 (sumber tidak diterbitkan). 2Kiai Ahmad Fauzan turut ke Makkah dalam usia belum baligh. Ia diperkirakan berusia
sebelas tahun (1916). Ketika pulang, ia sempat terpisah dari rombongan di Singapura.
Keterangan dari putra sulung Kiai Ahmad Fauzan, Kiai Zainurrohman. Kini ia tinggal di
Purwokerto. Informasi yang sama juga disampaikan oleh Hj. Latifah Fauzan. Wawancara
pada 1 Februari 2018.
3Catatan “Riwayat Hidup al-Maghfurlah Kiai Ahmad Fauzan” dalam peringatan Haul
ke-22 bulan September tahun 1994 (sumber tidak diterbitkan).
4Makamnya di Desa Banjaran, Bangsri, Jepara. Silsilahnya Arif bin Hasyim bin Abdul
Muhit bin Hasyim bin Syihabuddin bin Semith bin Arif bin Sholih bin Ahmad bin Ali bin
Ahmad bin Hamah bin ainuddin bin Abdullah al-Adniy (Yaman). Keterangan ini berasal dari
karya Syekh Muhammad Arif yang ditulis pada 8 Januari 1809. Koleksi digitalisasi naskah di
Universitas Nahdlatul Ulama (Unisnu) Jepara.
5Kiai Muhammad Masduri, “Syair Sejarah Simbah Muhammad Arif”, ditulis pada 1987
dengan huruf Arab dan berbahasa Jawa (pegon). Syair ini dibacakan rutin pada peringatan
haulnya setiap 10 Muharram (sumber tidak diterbitkan).
6“Silsilah Raden Ngabehi Brontodiwirjo (Mbah Soleman) Desa Jakenan, Pati tahun 1997”
(sumber tidak diterbitkan). pada 1 Februari 2018.
3Catatan “Riwayat Hidup al-Maghfurlah Kiai Ahmad Fauzan” dalam peringatan Haul
ke-22 bulan September tahun 1994 (sumber tidak diterbitkan).
4Makamnya di Desa Banjaran, Bangsri, Jepara. Silsilahnya Arif bin Hasyim bin Abdul
M hi bi
H
i
bi
S ih b ddi
bi
S
i h bi
A if bi
Sh lih bi
Ah
d bi
Ali bi Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Figur Kiai Ahmad Fauzan: Ulama Lokal yang Nasionalis 6“Silsilah Raden Ngabehi Brontodiwirjo (Mbah Soleman) Desa Jakenan, Pati tahun 1997”
(sumber tidak diterbitkan). ai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 | 34 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 meningkat, memunculkan kepemimpinan para ulama, kiai, dan haji dalam kehidupan
beragama. Dalam konteks tradisional, agama Islam terbukti dapat berfungsi sebagai
lambang pemersatu dan sekaligus sebagai ideologi politik, sehingga menimbulkan
kekuatan politik luar biasa, antara lain seperti yang diwujudkan dalam gerakan
sabilillah, perang jihad, dan sebagainya (Kartodirdjo, 1993). Dari penjelasan di atas patut dicatat bahwa Ahmad Fauzan, selain keturunan
dari ulama, baik dari pihak ayah maupun ibu, juga mengalir semangat perjuangan
melawan penindasan dan penjajahan. Mereka bukanlah ulama yang berdiam diri atas
penjajahan yang ada, melainkan berperan aktif dan berjuang mengusir penjajah. Sikap
tersebut – ingkar terhadap kedhaliman – juga yang tampak dalam pribadi Ahmad
Fauzan dalam perjalanan hidup selanjutnya. Peranan Kiai Ahmad Fauzan dalam Pendidikan Peranan Kiai Ahmad Fauzan dalam Pendidikan
Masa studi mengenai agama yang panjang, ditambah dengan pengabdian kepada
umat (khadim al-ummah) setelah nyantri mendorong terbentuknya keulamaan Kiai
Ahmad Fauzan. Peranan yang dimainkan tidak terbatas dalam bidang agama, tetapi
juga dalam politik pada era 1950-an. Hal ini sudah terlihat sejak ia menyelesaikan
pendidikan pesantren di Tayu dengan mendirikan sebuah sekolah di daerah Bangsri
pada 1930-an awal. Melalui lembaga pendidikan tersebut, Kiai Ahmad Fauzan
semakin memperkuat peranannya sebagai seorang ulama. Banyak masyarakat belajar
di sekolah yang dirintisnya. Kelas antara laki-laki dan perempuan dipisah dan
diterapkan dengan sistem berjenjang (Wawancara dengan Yahya, 16 Januari 2018). Di awal pendirian sampai awal 1970-an, tidak ada seragam khusus yang
dikenakan para siswa di madrasah. Anak laki-laki mengenakan sarung dan baju serta
berpeci, sedangkan anak-anak perempuan mengenakan kebaya dan kerudung. Kelas
dibagi menjadi dua waktu. Pagi sampai siang digunakan untuk murid (siswa) laki-
laki, sedangkan waktu siang sampai sore kelas untuk murid perempuan. Pendirian madrasah dengan model kelas atau klasikal pada 1930-an merupakan
sesuatu yang baru bagi komunitas muslim di Jepara (Wawancara dengan K.H. Noor
Rochman Fauzan, 20 Februari 2018). Model pendidikan kelas memang sudah ada di
Jepara pada awal pemberlakuan Politik Etis tahun 1901. Akan tetapi, sekolah-sekolah
yang didirikan oleh pemerintah Hindia Belanda tidak menekankan pendidikan
agama. Steenbrink membedakan madrasah dan sekolah dengan karakteristik
penekanan kurikulumnya. Meskipun mengajarkan ilmu umum sebagaimana sekolah,
namun kurikulum di madrasah sangat menonjolkan nilai religiusitas masyarakatnya. Sementara itu, sekolah merupakan lembaga pendidikan umum dengan pelajaran
universal dan terpengaruh iklim pencerahan Barat (Streenbrink, 1986). Rata-rata siswa dalam satu ruang berjumlah 20 sampai 30 siswa. Istilah kelas
tidak lazim dipakai, tetapi istilah yang dipakai adalah shifir yang bermakna tingkatan
(Wawancara dengan Zaini, 10 April 2018). Begitu pula dengan sebutan "guru" untuk
para pengajar juga jarang digunakan. Istilah yang digunakan adalah "kang" atau "mas"
untuk yang masih muda, dan ustadz bagi yang lebih tua. | 35 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 Pembentukan sebuah lembaga pendidikan tidak menghentikan gerak langkah
Kiai Ahmad Fauzan dalam pengembangan dakwah Islam di Jepara. Dalam periode
selanjutnya, ia lebih banyak menghabiskan waktu untuk kegiatan berdakwah di luar,
dengan cara berkelana dari satu desa ke desa yang lain. Ia menyerahkan pengelolaan
madrasah kepada para guru yang lebih muda. Meskipun demikian, sesekali Fauzan
tetap memantau proses belajar mengajar di madrasah tersebut. Metode dakwah dengan cara berkelana menjadikannya dikenal oleh masyarakat
luas. Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Peranan Kiai Ahmad Fauzan dalam Pendidikan Di samping itu, ia juga menyederhanakan ilmu-ilmu agama dengan cara
menyajikannya dalam bentuk syair kepada masyarakat. Dari syair-syair yang ditulis
dan disampaikan kepada masyarakat tampak bahwa Kiai Ahmad Fauzan adalah
tipikal ulama substansial yang lebih mengedepankan pada isi daripada kemasan. Pemikiran Fauzan yang tergambar dalam syair-syair yang ditulis menunjukkan
kemahiran dalam berdakwah. Kemampuan menerjemahkan suatu nilai-nilai agama
menjadi syair-syair adalah salah satu ciri khas yang dimilikinya. Di samping dengan
metode ceramah atau bandhongan, masyarakat awam juga dapat belajar agama dengan
melantunkan syair-syair yang diciptakan Fauzan. Terdapat beberapa syair yang
terdokumentasikan dengan baik, antara lain: Syair Asmaul Husna, Syair Isra’ Mikraj,
Syair pribadi Nabi Muhammad Saw, Syair Pitutur, Syair Kemanusiaan, Syair Kenabian, dan
Syair Akhlak Mulia. Di samping syair berbahasa daerah, Fauzan juga menulis sebuah
syair berbahasa Arab, yaitu Alfiyah al-Ghazaliyah. Syair tersebut berisi butir-butir
nasihat yang disarikan dari karya monumental Imam Ghazali, yaitu Ihya Ulumuddin. Gambar 1. Naskah asli Syair Asmaul Husna menggunakan huruf pegon
Sumber: Dokumentasi Keluarga. Gambar 1. Naskah asli Syair Asmaul Husna menggunakan huruf pegon
Sumber: Dokumentasi Keluarga. Suatu yang menarik dari sekian banyak syair yang terdokumentasikan, sebagian
besar menjelaskan posisi manusia baik sebagai hamba Allah, warga negara, maupun
masyarakat. Ketiganya dijelaskan secara bersama-sama dan tidak saling bertentangan. Kiai Ahmad Fauzan menghendaki agar masyarakat dapat menjalankan syariat agama,
bermasyarakat dengan baik, dan menjadi pemimpin dengan baik. Misalnya di dalam
Syair Asmaul Husna yang di dalamnya tidak hanya menjelaskan tentang tauhid atau Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 | 36 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 keimanan semata, tetapi juga menjelaskan etika hidup bersama dengan sesama. Dapat
dipahami bahwa Kiai Ahmad Fauzan adalah ulama yang substansial, yang tidak
menjadikan agama Islam sebagai kekuatan politik yang bersifat formal. Yang menjadi
prinsip adalah ajaran agama Islam dapat terjamin di dalam sebuah negara. Pada saat
itu pula, Islam mampu mewarnai kehidupan berbangsa dan bernegara. 7Catatan “Riwayat Hidup al-Maghfurlah Kiai Ahmad Fauzan” dalam peringatan Haul
ke-22 bulan September tahun 1994 (sumber tidak diterbitkan). Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972
| 37
7Catatan “Riwayat Hidup al-Maghfurlah Kiai Ahmad Fauzan” dalam peringatan Haul
ke-22 bulan September tahun 1994 (sumber tidak diterbitkan). Jaringan Politik Kiai Ahmad Fauzan Gambar 2. Foto Kiai Ahmad
Fauzan saat menjabat
Sumber: Dok. Kemenag, 1994. Nama
Kiai
Ahmad
Fauzan
semakin
banyak
diperbincangkan ketika tentara Jepang tiba di Jepara
pada 1942. Sebagai Ulama, ia mendapatkan perhatian
khusus dari pemerintah militer Jepang. Ia bahkan
sempat
ditangkap
oleh
polisi
militer
Jepang
(kenpeitai) dan dimasukkan ke dalam penjara karena
dituduh sebagai provokator kerusuhan. Akan tetapi,
tuduhan tersebut tidak terbukti dan Kiai Ahmad
Fauzan pun dibebaskan7. Peranan
sebagai
pemimpin
umat
terus
berlanjut. Sekitar
Agustus
1945,
pada
masa
kemerdekaan Indonesia, para ulama di wilayah
Keresidenan Pati berkumpul dan bermusyawarah
untuk menghadapi kekuatan tentara Jepang yang
masih eksis berkuasa. Di dalam musyawarah
tersebut, secara aklamasi para ulama menyepakati
Kiai Ahmad Fauzan sebagai ulama muda untuk
memimpin pelucutan senjata tentara Jepang. Dengan kerja sama semua elemen
masyarakat, serta dukungan penuh dari para ulama yang menggerakkan para santri,
akhirnya pasukan Jepang menyerah kepada massa rakyat di Pati. Gambar 2. Foto Kiai Ahmad
Fauzan saat menjabat
Sumber: Dok. Kemenag, 1994. Gambar 2. Foto Kiai Ahmad
Fauzan saat menjabat
Sumber: Dok. Kemenag, 1994. Momentum proklamasi kemerdekaan juga menjadi jalan bagi Kiai Ahmad
Fauzan untuk aktif pada gerakan-gerakan sosial-keagamaan. NU menjadi organisasi
yang dipimpinnya sejak 1945. Sebelumnya, dia juga aktif di Majelis Syuro Muslimin
Indonesia (Masyumi) sebagai wakil dari organisasi NU. Sekitar 1949, Kiai Ahmad
Fauzan juga dipercaya oleh pemerintah untuk menjadi kepala Kementerian Agama
(Kemenag) Kabupaten Jepara. Jabatan ini diemban sampai akhir 1960-an akhir. Penunjukan tersebut tidak lepas dari keaktifannya dalam kegiatan sosial-keagamaan
di tengah masyarakat, yaitu dakwah dengan cara berkelana dari satu desa ke desa
yang lain. Cara itu telah menjadikan Fauzan lebih mengenal dan dikenal masyarakat
luas serta mengetahui kondisi yang terjadi di masyarakat bawah. Pengalaman-
pengalaman yang ia dapatkan secara langsung dari bawah juga turut mewarnai
pemikiran dan metode dakwah yang digunakan. Salah satu metode yang digunakan | 37 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 adalah dengan model penyampaian nilai-nilai agama melalui syair-syair yang relatif
mudah dipahami masyarakat. Ketika NU menjadi partai politik pada 1952, Kiai Ahmad Fauzan tampil sebagai
pemimpin (suriyah) partai. Benih-benih perpecahan antara NU dan Masyumi sudah
tampak dalam Kongres Masyumi keempat di Yogyakarta pada Desember 1949. Pengaruh kelompok Natsir yang semakin kuat menghasilkan keputusan-keputusan
partai yang memojokkan peran ulama tradisional yang ada di NU. Sementara itu,
Majelis Syuro (Dewan Penasihat) yang diduduki para ulama NU dibatasi peranan
politiknya. Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Jaringan Politik Kiai Ahmad Fauzan Majelis Syuro bukan penentu atas kebijakan partai, tetapi hanya
memberikan pendapat dan fatwa apabila dianggap perlu. Ketidakseimbangan antara
kelompok modernis dan tradisionalis juga tampak terutama dalam distribusi dalam
kursi kabinet pemerintahan (Fealy, 2010: 100-101). Kiai Ahmad Fauzan turut mendukung keputusan pemisahan NU terhadap
Masyumi. Langkah tersebut mudah dipahami mengingat sejak awal perjuangannya,
ia berkiprah dalam organisasi NU Jepara. Ketika menjadi bagian dari Majelis Islam
A'la Indonesia (MIAI) pada masa Hindia Belanda, serta Masyumi pada masa Jepang
dan masa kemerdekaan, Ahmad Fauzan menjadi wakil dari NU atau ulama
tradisional di dalam kedua organisasi federasi tersebut. Kiai Ahmad Fauzan juga merupakan salah seorang ulama generasi pertama yang
memelopori berdirinya NU di Jepara. Meskipun tidak ditemukan bukti tertulis terkait
kapan NU Jepara sebagai sebuah organisasi terbentuk, bukan berarti paham
Ahlussunah wal Jamaah (Aswaja) sebagaimana yang dikembangkan NU tidak
berkembang. Sebelum organisasi NU lahir pada 1926 di Surabaya, praktik
keberagamaan paham Aswaja sudah menjadi bagian dari kehidupan masyarakat
muslim di Jepara. Fakta tersebut dibuktikan dengan kehadiran ulama-ulama beraliran
Aswaja di Jepara seperti Kiai Umar al-Samarani, Kiai Sanwasi, K.H. Hasbullah
pengasuh Pesantren Balekambang, sekaligus salah seorang guru Kiai Ahmad Fauzan. Kepemimpinan Fauzan di Partai NU Cabang Jepara sebagai pimpinan tertinggi
(rais suriyah) 1952 merupakan posisi penting sebagai penentu arah gerak partai. Suriyah di dalam NU baik sebagai organisasi sosial-keagamaan maupun sebagai partai
politik memiliki peranan yang penting dan strategis. Artinya, baik kebijakan maupun
program-program partai terutama dalam Pemilu 1955 harus mendapatkan
pertimbangan dan restu dari suriyah. p
g
y
Barangkali ada pola-pola kampanye Partai NU Cabang Jepara yang tidak
provokatif dan relatif tenang yang membedakannya dengan daerah lain. Kampanye-
kampanye yang dilakukan Partai NU tidak begitu keras dengan hujatan dan cacian
atau makian. Dapat digambarkan bahwa suasana kampanye di Jepara sepanjang 1955 relatif
tenang, meskipun perselisihan di level bawah yang mewakili dari unsur NU dan
Muhammadiyah di Masyumi Jepara sering kali terjadi. Permasalahan yang muncul
terutama terkait dengan keagamaan yang bersifat cabang (furu), bukan pokok (ushul). Ungkapan bahwa Muhammadiyah itu tidak berpaham Aswaja sering kali terlontar di
tingkat bawah (grassroot). Alasannya karena mereka tidak menggunakan mazhab, | 38 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 tetapi secara langsung menggunakan Al-Qur'an dan Hadis. Akan tetapi, ketegangan
itu tidak sampai memunculkan pertikaian atau kerusuhan. Kiai Ahmad Fauzan dengan jaringan politik yang luas, sekurang-kurangnya
dengan massa di pedesaan, telah memberikan dampak yang signifikan kepada partai. Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Jaringan Politik Kiai Ahmad Fauzan Perjalanan sebagai seorang kiai keliling dari satu desa ke desa secara konsisten
menjadikan banyak murid atau pengikutnya mengikuti pilihan politiknya. Tampaknya motif politik bukanlah tujuan utama kegiatan dakwah yang dilakukan
Fauzan sejak 1930-an. Kegiatan politik lebih dimaknai sebagai sebuah sarana untuk
berdakwah. Sikap Kiai Ahmad Fauzan yang santun baik kepada kawan maupun lawan,
tanpa menafikan komponen Partai NU yang lain, adalah tiga dari sekian banyak faktor
yang menjadikan NU Cabang Jepara berhasil dalam agenda politiknya. Kemenangan
Partai NU di Jepara pada Pemilu 1955, serta kondisi daerah yang relatif terjaga tidak
dapat dilepaskan dari kontribusi Kiai Ahmad Fauzan sebagai pimpinan partai. Dari
537 tempat pemungutan suara (TPS) yang tersebar di wilayah Jepara, Partai NU
menang di sebagian besar TPS untuk Dewan Perwakilan Rakyat Daerah (Suara
Merdeka, 1 Oktober 1955). Pada Pemilu DPRD tahap pertama 29 September 1955, terlihat Partai NU di
Jepara unggul dan memenangkan kontestasi pemilihan anggota di parlemen. Kemenangan Partai NU di Jepara jauh lebih unggul di antara tiga partai yang lain,
yaitu: PNI, PKI, dan Masyumi. Para calon legislatif dari Partai NU Jepara tersebar
merata di wilayah Jepara. Mereka adalah para pengurus NU di tingkat kecamatan atau
Majelis Wakil Cabang (MWC) yang memenuhi persyaratan untuk menjadi wakil
rakyat. Komposisinya didasarkan atas jumlah penduduk di daerah pemilihan
(Wawancara dengan K.H. Mubin, 17 April 2018). Pada Pemilu selanjutnya, yaitu 15
Desember 1955 untuk memilih anggota Dewan Konstituante, NU Jepara kembali
mengukir kemenangan mutlak di atas tiga partai besar lainnya. Berikut adalah tabel
hasil perolehan suara Partai NU dengan tiga partai besar yang lain di Kabupaten
Jepara pada Pemilu 1955. Tabel 1. Perolehan Suara Empat Partai Besar di Kabupaten Jepara
Partai Politik
Perolehan suara DPR
Perolehan suara Dewan
Konstituante
NU
110.953
105.221
PNI
45.999
53.043
PKI
19.377
21.705
Masyumi
10.953
10.991
Sumber: Suara Merdeka, 19 Desember 1955. Tabel 1. Perolehan Suara Empat Partai Besar di Kabupaten Jepara Kemenangan Partai NU di Jepara pada Pemilu 1955 tidak dapat dipisahkan dari
kiprah Ahmad Fauzan. Pertama, kegiatan dakwah dan organisasi yang dilakukan sejak
zaman Jepang sampai era kemerdekaan di pedesaan Jepara memiliki kontribusi yang
besar dalam mendulang suara partai. Kemudian di tahun 1950-an, sebagian kegiatan ai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 | 39 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 keagamaan itu terlembagakan di dalam organisasi NU. Pada saat itu wadah organisasi
menjadi semacam pengikat bagi para jamaah. Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Jaringan Politik Kiai Ahmad Fauzan Keadaan tersebut juga didukung dengan faktor kedua, yaitu kondisi sosial-
masyarakat Jepara yang memiliki ikatan yang kuat terhadap ulama. Kharisma dari
Ahmad Fauzan menjadi patron bagi masyarakat, baik dalam mengikuti anjuran,
fatwa, maupun pilihan politik. Hubungan mutualisme ini tidak terjadi dalam waktu
yang pendek, tetapi dalam jangka waktu yang relatif lama. Ketiga, strategi kampanye yang santun dengan tidak bersikap kaku dan
ditambah menggunakan pendekatan nilai-nilai agama. Dengan gaya orasinya yang
khas, yaitu membawakan syair-syair yang mencerminkan suasana zaman menjadikan
konstituen ideologis turut serta dalam gerbong politik Kiai Ahmad Fauzan. Keempat,
kesuksesan Partai NU di Jepara juga ditopang bangunan partai yang mulai mapan
dengan berdirinya badan otonom (Banom) yang banyak didirikan pada 1950-an,
seperti Ansor-Banser, Fatayat, dan Muslimat. Beberapa Banom NU tersebut turut
memainkan peranan penting khususnya di tingkat pedesaan. Atas kemenangan mutlak Partai NU di Jepara, sebagai pimpinan partai, Fauzan
memiliki kesempatan untuk menjadi bupati Jepara. Banyak kalangan yang menilai
Fauzan akan mengambil kesempatan tersebut. Akan tetapi, dugaan tersebut salah. Pilihan yang diambil adalah dengan memberikan kepada kader NU yang lain. Jabatan
tersebut dipercayakan kepada Sahlan. Posisinya sebagai panutan masyarakat
menjadikan faktor penting atas kemenangan Partai NU di Jepara. Di samping itu,
banyaknya masyarakat yang mengenal Fauzan melalui pengajian dan kampanye-
kampanye yang dilakukan cenderung bersikap santun menjadi faktor yang lain. Tidak
sulit bagi Kiai Ahmad Fauzan untuk mendapatkan dukungan politik dari masyarakat
luas. Syair-syair yang disampaikan kepada masyarakat luas lebih bersifat pengingat
dan penyadaran atas penyemaian nilai-nilai keagamaan dan kebangsaan. Manuver
politik yang dilakukan Fauzan bukanlah untuk tujuan kekuasaan, tetapi
diorientasikan untuk kebaikan (kemaslahatan) masyarakat, bangsa, dan negara. Perjalanan dakwah di masyarakat membawa keberhasilan partai dalam
memenangi suara NU di Jepara pada Pemilu 1955. Setelah itu, dalam periode
selanjutnya, Fauzan tidak lagi aktif dalam kegiatan politik praktis. Fauzan lebih
memilih membina umat di Jepara melalui Kemenag Kabupaten Jepara. Komitmen
dalam memegang prinsip agama dibuktikan oleh Fauzan dengan memberikan syarat
kepada pimpinan Kantor Urusan Agama (KUA) di setiap kecamatan yang ada di
Jepara harus dapat membaca kitab para ulama salaf atau yang lazim disebut Kitab
Kuning. Hampir dapat dipastikan bahwa orang-orang yang dapat membaca karya-
karya ulama salaf adalah mereka yang lulus dari pesantren atau orang-orang yang
menempuh pendidikan tradisional yang dikembangkan para ulama-ulama salaf. Keilmuan yang dimiliki dan disertai dengan perilaku (akhlak) yang baik
menjadikannya sebagai seorang ulama sufi yang disegani masyarakat. Berbagai
pemikirannya, baik di bidang keagamaan (akidah, syariat, dan tasawuf) maupun | 40 Indonesian Historical Studies, Vol. Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Jaringan Politik Kiai Ahmad Fauzan 3, No. 1, 29-43 © 2019 kebangsaan terlihat dari syair-syair yang disampaikan kepada masyarakat, terutama
pascakemerdekaan. Kegiatan berdakwah Fauzan masih terus berlanjut meski ia tidak lagi memiliki
jabatan baik politik maupun pemerintahan. Metode dakwah yang khas, yakni
berkelana dan menggunakan syair-syair keagamaan dan kebangsaan masih dilakukan
pada 1960-an. Perjalanan intelektualnya dengan ulama-ulama di Makkah dan
pengalaman belajar dari beberapa pesantren di Jawa Tengah mendorong untuk
menetapkan diri membangun kekuatan Islam di kawasan pedesaan Jepara berbasis
masyarakat tradisional. Nilai-nilai etis yang diajarkan Fauzan baik melalui syair,
karya tulis maupun keteladanan kepada masyarakat masih dapat dirasakan melalui
putra-putrinya. Simpulan Bagi masyarakat Jepara, figur Kiai Ahmad Fauzan berperan dalam pendidikan terlihat
sejak selesainya pendidikan di pesantren dengan mendirikan sebuah madrasah di
daerah Bangsri pada 1930-an awal. Di lembaga pendidikan tersebut Fauzan memulai
merintis peran-peran sebagai kiai. Ketika tentara Jepang tiba di Jepara pada 1942,
Fauzan mendapatkan perhatian khusus dari pemerintah militer Jepang. Dengan
posisinya sebagai ulama, ia sempat ditangkap oleh polisi militer Jepang (kenpeitai) dan
dimasukkan ke dalam penjara atas tuduhan sebagai provokator kerusuhan (baca:
krayahan). Akan tetapi, tuduhan tersebut tidak terbukti dan Fauzan pun dibebaskan. Proses keulamaan Kiai Ahmad Fauzan melalui lembaga pendidikan maupun
berdawah keliling terus dilakukan hingga kemerdekaan. Melalui NU, Kiai Ahmad
Fauzan berjuang untuk menegakkan paham Aswaja. Paham yang dikembangkan
tersebut tampak dalam syair-syair yang disampaikan kepada masyarakat. Syair-
syairnya sebagai salah satu ekspresi pemikiran, yang dibawa hampir sebagian besar
tidak memisahkan kedudukan manusia sebagai hamba dan masyarakat bangsa. Pemikiran Fauzan yang tergambar dalam syair-syair yang ditulis menunjukkan
kemahiran dalam berdakwah. Kemampuan menerjemahkan suatu nilai-nilai agama
menjadi syair-syair adalah salah satu ciri khas yang dimilikinya. Di samping dengan
metode ceramah atau bandhongan, masyarakat awam juga dapat belajar agama dengan
melantunkan syair-syair yang diciptakan Fauzan. Terdapat beberapa syair yang
terdokumentasikan dengan baik, antara lain: Syair Asmaul Husna, Syair Isra Mikraj,
Syair pribadi Nabi Muhammad Saw, Syair Pitutur, Syair Kemanusiaan, Syair Kenabian, dan
Syair Akhlak Mulia. Di samping syair berbahasa daerah, Fauzan juga menulis sebuah
syair berbahasa Arab, yaitu Alfiyah al-Ghazaliyah. Syair tersebut berisi butir-butir
nasihat yang disarikan dari karya monumental Imam Ghazali, Ihya Ulumuddin. Di
dalam syair-syair yang ditulis dan disampaikan oleh Fauzan, terdapat tiga kategori
pemikiran, yaitu akidah, syariat, tasawuf, dan kebangsaan. Syair juga menjadi metode dalam kampanye saat Pemilu 1955. NU yang pada
waktu itu menjadi partai politik memenangi pemilihan secara mutlak di Jepara. Hal
tersebut tidak dapat dipisahkan dari peranan Kiai Ahmad Fauzan sebagai pimpinan | 41 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 partai (suriyah). Di samping itu, pengalaman hidup bersama dengan masyarakat
dalam jangka waktu yang relatif lama menjadi alasan tersendiri. Kegiatan berdakwah Fauzan masih terus berlanjut meski ia tidak lagi memiliki
jabatan baik di politik maupun pemerintahan. Metode dakwah yang khas, yakni
berkelana dan menggunakan syair-syair keagamaan dan kebangsaan masih dilakukan
di tahun 1960-an. Perjalanan intelektual Fauzan dengan ulama-ulama di Makkah dan
pengalaman belajar dari beberapa pesantren di Jawa Tengah mendorongnya untuk
menetapkan diri membangun kekuatan Islam di kawasan pedesaan Jepara berbasis
masyarakat tradisional. Simpulan Nilai-nilai etis yang diajarkan Fauzan baik melalui syair,
karya tulis maupun keteladanan kepada masyarakat masih dapat dirasakan melalui
putra-putrinya. Fauzan wafat pada 12 Maret 1972 di Bangsri, Jepara. Sebagai bentuk
penghorrmatan, nama Fauzan pun diabadikan menjadi salah satu nama jalan di
Kabupaten Jepara. Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Indonesian Historical Studies, Vol. 3, No. 1, 29-43 © 2019 Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972 Ulum, A. Arbi, M. Y. (2016). KH. Sholeh Darat: Maha Guru Para Ulama Nusantara 1820-
1903. Depok: Shahifa.
Wahid, M. (peny.) (1999). Geger di "Republik" NU. Jakarta: Lakpesdam NU dan
Kompas. Ulum, A. Arbi, M. Y. (2016). KH. Sholeh Darat: Maha Guru Para Ulama Nusantara 1820-
1903. Depok: Shahifa.
Wahid, M. (peny.) (1999). Geger di "Republik" NU. Jakarta: Lakpesdam NU dan p
Wahid, M. (peny.) (1999). Geger di "Republik" NU. Jakarta: Lakpesdam NU dan
Kompas. Referensi Azra, Azyumardi (1989). Konteks berteologi masyarakat Indonesia. Jakarta: Wacana Ilmu. Azra, Azyumardi (1989). Konteks berteologi masyarakat Indonesia. Jakarta: Wacana Ilmu. y
(
)
g
y
Fealy, Greg (2003). Ijtihad politik ulama: Sejarah NU 1952-1967, terjemahan Farid Wajdi
dan Mulni Adelina Bachtar. Yogyakarta: LKis. Fealy, Greg (2003). Ijtihad politik ulama: Sejarah NU 1952-1967, terjemahan Farid Wajdi
dan Mulni Adelina Bachtar. Yogyakarta: LKis. Florida, N. (2003). Menyurat yang silam, menggurat yang menjelang: Sejarah sebagai nubuat
di Jawa masa kolonial. Yogyakarta: Bentang Budaya. Florida, N. (2003). Menyurat yang silam, menggurat yang menjelang: Sejarah sebagai nubuat
di Jawa masa kolonial. Yogyakarta: Bentang Budaya. Hady, Nurrudin (2010). Teori konstitusi negara demokrasi: Pemahaman konstitusionalisme
demokrasi pasca amandemen UUD 1945. Malang: Setara Press. Hady, Nurrudin (2010). Teori konstitusi negara demokrasi: Pemahaman konstitusionalisme
demokrasi pasca amandemen UUD 1945. Malang: Setara Press. Kartodirdjo, Sartono (1993). Pendekatan ilmu sosial dalam metodologi sejarah. Jakarta:
Gramedia Pustaka Utama. Kartodirdjo, Sartono (1993). Pendekatan ilmu sosial dalam metodologi sejarah. Jakarta:
Gramedia Pustaka Utama. Kuntowijoyo (1991). Paradigma Islam: Interpretasi untuk aksi. Bandung: Mizan. antoso, L. (2014). Sejarah terlengkap gerakan separatis Islam. Yogyakarta: Palapa. Sianipar (2002). Runtuhnya negara bangsa. Jakarta: Lemhanas Republik Indonesi Siraj, Said Agil (2015). Islam sumber inspirasi budaya nusantara: Menuju masyarakat
mutamaddin. Jakarta: LTN NU. Siregar, F. M., Setiawan, N. K., Setio, R. (2013). Religious leader and charismatic
leadership in Indonesia: The role of kyai in pesantren in Java. Kawistara, 3(2):
117-126. Steenbrink, Karel A. (1984). Beberapa aspek tentang Islam di Indonesia abad ke-19. Jakarta:
PT Bulan Bintang. Steenbrink, Karel A. (1991). Pesantren, madrasah, sekolah, cetakan kedua. Jakarta: LP3ES. Suara Merdeka, 1 Oktober 1955. Suara Merdeka, 19 Desember 1955. Suara Merdeka, 19 Desember 1955. Suhartono (1991). Apanage dan bekel: Perubahan sosial di pedesaan Surakarta 1830-1920. Yogyakarta: Tirta Wacana. Sulistiyono, S. T. (2015). Diaspora maritim dan proses reformasi keindonesiaan. Jakarta:
Badan Informasi Geoospasial. Suryo, Djoko (2009). Transfomasi masyarakat Indonesia dalam historiografi Indonesia
modern. Yogyakarta: STPN Press. | 42 Ulum, A. Arbi, M. Y. (2016). KH. Sholeh Darat: Maha Guru Para Ulama Nusantara 1820-
1903. Depok: Shahifa. Wahid, M. (peny.) (1999). Geger di "Republik" NU. Jakarta: Lakpesdam NU dan
Kompas. p
Wahid, M. (peny.) (1999). Geger di "Republik" NU. Jakarta: Lakpesdam NU dan
Kompas. | 43 Kiai Ahmad Fauzan: Pemikiran dan Peranannya di Kabupaten Jepara, 1942-1972
|
https://openalex.org/W3161401138
|
https://peerj.com/articles/11334v0.2/submission
|
English
| null |
Peer Review #3 of "Variance of vegetation coverage and its sensitivity to climatic factors in the Irtysh River basin (v0.1)"
| null | 2,021
|
cc-by
| 18,956
|
Variance of vegetation coverage and its sensitivity to climatic
factors in Irtysh River basin 1 State Key Laboratory of Desert and Oasis Ecology, Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences (CAS), Urumqi, Xin
China
2 College of Water and Architectural Engineering, Shihezi university, Shihezi, Xinjiang, China
3 Institute of resources and ecology, Yili Normal University, Yining, China
Corresponding Authors: Junjie Yan, Hong-bo Ling
Email address: yan3550@sina.com, linghb@ms.xjb.ac.cn 1 State Key Laboratory of Desert and Oasis Ecology, Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences (CAS), Urumqi, Xinjiang,
China
2 College of Water and Architectural Engineering Shihezi university Shihezi Xinjiang China 1 State Key Laboratory of Desert and Oasis Ecology, Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences (CAS), Urumqi, Xinjiang,
China 1 State Key Laboratory of Desert and Oasis Ecology, Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences (CAS), Urumqi, Xinjiang,
China
2 2 College of Water and Architectural Engineering, Shihezi university, Shihezi, Xinjiang, China 3 Institute of resources and ecology, Yili Normal University, Yining, China Corresponding Authors: Junjie Yan, Hong-bo Ling
Email address: yan3550@sina.com, linghb@ms.xjb.ac.cn Corresponding Authors: Junjie Yan, Hong-bo Ling
Email address: yan3550@sina.com, linghb@ms.xjb.ac.cn Background. Climate change is an important factor driving vegetation changes in arid areas. Identifying
the sensitivity of vegetation to climate variability is crucial for developing sustainable ecosystem
management strategies. Irtysh River is located in the westerly partition of China, and its vegetation cover
is more sensitive to climate change. However, previous studies rarely studied the changes in the
vegetation coverage of the Irtysh River and its sensitivity to climate factors from a spatiotemporal
perspective. Methods. we adopted a vegetation sensitivity index (VSI) based on remote sensing datasets of high
temporal resolution to study the sensitivity of vegetation to climatic factors in Irtysh River basin, then
reveal the driving mechanism of vegetation cover change. Results. The results show that 88.09% of vegetated pixels show an increasing trend in vegetation
coverage, and the sensitivity of vegetation to climate change presents spatial heterogeneity. Sensitivity
of vegetation increases with the increase of coverage. Temperate steppe in the northern mountain and
herbaceous swamp and broadleaf forest in the river valley, where the normalized difference vegetation
index (NDVI) is the highest, show the strongest sensitivity, while the desert steppe in the northern plain,
where the NDVI is the lowest, shows the strongest memory effect (or the strongest resilience). Manuscript to be reviewed 1
Variance of vegetation coverage and its sensitivity to climatic factors in
2
Irtysh River basin 1
Variance of vegetation coverage and its sensitivity to climatic factors in
2
Irtysh River basin
3
Feifei Han1,2, Junjie Yan3, Hongbo Ling1
4
5
1 State Key Laboratory of Desert and Oasis Ecology, Xinjiang Institute of Ecology and
6
Geography, Chinese Academy of Sciences (CAS), Urumqi, Xinjiang, China
7
2 College of Water and Architectural Engineering, Shihezi university, Shihezi, Xinjiang, China
8
3 Institute of resources and ecology, Yili Normal University, Yining, Xinjiang, China
9
10
Corresponding Author:
11
Junjie Yan3, Hongbo Ling1
12
No. 818 South Beijing Road, Urumqi, Xinjiang, 830011, China
13
Email address: yan3550@sina.com (J. Yan); linghb@ms.xjb.ac.cn (H. Ling). 14
15
16
Abstract
17
Background. Climate change is an important factor driving vegetation changes in arid areas. 18
Identifying the sensitivity of vegetation to climate variability is crucial for developing
19
sustainable ecosystem management strategies. Irtysh River is located in the westerly partition of
20
China, and its vegetation cover is more sensitive to climate change. However, previous studies
21
rarely studied the changes in the vegetation coverage of the Irtysh River and its sensitivity to
22
climate factors from a spatiotemporal perspective. 23
24
Methods. we adopted a vegetation sensitivity index (VSI) based on remote sensing datasets of
25
high temporal resolution to study the sensitivity of vegetation to climatic factors in Irtysh River
26
basin, then reveal the driving mechanism of vegetation cover change. 27
28
Results. The results show that 88.09% of vegetated pixels show an increasing trend in vegetation
29
coverage, and the sensitivity of vegetation to climate change presents spatial heterogeneity. 30
Sensitivity of vegetation increases with the increase of coverage. Temperate steppe in the
31
northern mountain and herbaceous swamp and broadleaf forest in the river valley, where the
32
normalized difference vegetation index (NDVI) is the highest, show the strongest sensitivity, while the
33
desert steppe in the northern plain, where the NDVI is the lowest, shows the strongest memory
34
effect (or the strongest resilience). Relatively, the northern part of this area is more affected by a
35
combination of precipitation and temperature, while the southern plains dominated by desert
36
steppe are more sensitive to precipitation. Central river valley dominated by herbaceous swamp
37
is more sensitive to temperature-vegetation dryness index (TVDI). This study underscores that the
38
sensitivity of vegetation cover to climate change is spatially differentiated at the regional scale. 39 24
Methods. Manuscript to be reviewed Manuscript to be reviewed Variance of vegetation coverage and its sensitivity to climatic
factors in Irtysh River basin Relatively,
the northern part of this area is more affected by a combination of precipitation and temperature, while
the southern plains dominated by desert steppe are more sensitive to precipitation. Central river valley
dominated by herbaceous swamp is more sensitive to temperature-vegetation dryness index (TVDI). This
study underscores that the sensitivity of vegetation cover to climate change is spatially differentiated at
the regional scale. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) 40
Introduction 40
Introduction
41
Vegetation covers nearly three-quarters of the land surface(Sarkar & Kafatos 2004), which is the main body of
42
the terrestrial ecosystem(Cramer & Leemans 1993), and plays an important role in linking the soil and atmosphere
43
through energy and mass transport(Deng et al. 2017). In the context of global climate change, vegetation is affected
44
by climate change and becomes the sufferers of climate change, acting as an "indicator" in global climate change
45
research(Piao et al. 2006). The changes vegetation also have feedback effect on climate change (Liu et al. 2006; Sun
46
et al. 2018). Therefore, monitoring and mapping the sensitivity of vegetation to current climate variability is crucial
47
for projecting future vegetation dynamics and developing sustainable ecosystem management strategies. 48
Investigating the responses of vegetation to short-term climate anomalies is of great significance to mitigate the
49
ecological, economic, and social consequences of future climate change (Huete & Alfredo 2016). Much current
50
understanding of vegetation’s respond to climate change is based on changes in mean climate state (Mearns et al. 51
1997; Thomas et al. 2004). We consider this mean state can only represent changes in the equilibrium state (e.g. due
52
to overgrazing or long-term successional cycles (Clifford et al. 2011)) instead of anomalies resulting from short-
53
term climate anomalies. However, the response of vegetation to changing climate is comprehensive and often varies
54
dramatically between regions(Anav & Mariotti 2011). Yet, key knowledge of how to identify and then prioritize
55
those regions that are more sensitive to climatic variability is still lacking. However, a key issue must be addressed
56
before we analyze the sensitivity of vegetation to climate variability and that is how to describe vegetation
57
sensitivity in a quantitative way. There are several ways defining the vegetation’s sensitivity to climate variability. 58
For example, vegetation’s sensitivity refers to the degree and magnitude of vegetation response when the climate
59
anomaly occurs(You et al. 2018) or the degree to which a system changes after a disturbance(Li et al. 2018). In this
60
study, vegetation’s sensitivity is defined as the magnitude of vegetation response at the moment of the climate
61
anomaly(Tilman 1996). Additionally, Moulin S et al. (Moulin et al. 1
Variance of vegetation coverage and its sensitivity to climatic factors in
2
Irtysh River basin we adopted a vegetation sensitivity index (VSI) based on remote sensing datasets of
25
high temporal resolution to study the sensitivity of vegetation to climatic factors in Irtysh River
26
basin, then reveal the driving mechanism of vegetation cover change. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) 40
Introduction 1997) found that due to the relatively slow
62
growth of vegetation, the response of vegetation to climate change often lags, therefore vegetation growth depends
63
both on current disturbances and the residual effects of past climate conditions. We called this phenomenon the
64
vegetation memory effect and it could be described as the persistence of trends in temporal changes of ecosystem
65
properties (Dash et al. 2014; Lhermitte et al. 2010; Simoniello et al. 2008). This ‘memory effect’ should be
66
considered when assessing the immediate response to short-term climate anomalies(De Keersmaecker et al. 2015). 40
Introduction
41
Vegetation covers nearly three-quarters of the land surface(Sarkar & Kafatos 2004), which is the main body of
42
the terrestrial ecosystem(Cramer & Leemans 1993), and plays an important role in linking the soil and atmosphere
43
through energy and mass transport(Deng et al. 2017). In the context of global climate change, vegetation is affected
44
by climate change and becomes the sufferers of climate change, acting as an "indicator" in global climate change
45
research(Piao et al. 2006). The changes vegetation also have feedback effect on climate change (Liu et al. 2006; Sun
46
et al. 2018). Therefore, monitoring and mapping the sensitivity of vegetation to current climate variability is crucial
47
for projecting future vegetation dynamics and developing sustainable ecosystem management strategies. 48
Investigating the responses of vegetation to short-term climate anomalies is of great significance to mitigate the
49
ecological, economic, and social consequences of future climate change (Huete & Alfredo 2016). Much current
50
understanding of vegetation’s respond to climate change is based on changes in mean climate state (Mearns et al. 51
1997; Thomas et al. 2004). We consider this mean state can only represent changes in the equilibrium state (e.g. due
52
to overgrazing or long-term successional cycles (Clifford et al. 2011)) instead of anomalies resulting from short-
53
term climate anomalies. However, the response of vegetation to changing climate is comprehensive and often varies
54
dramatically between regions(Anav & Mariotti 2011). Yet, key knowledge of how to identify and then prioritize
55
those regions that are more sensitive to climatic variability is still lacking. However, a key issue must be addressed
56
before we analyze the sensitivity of vegetation to climate variability and that is how to describe vegetation
57
sensitivity in a quantitative way. Manuscript to be reviewed 67
Since the 1970s, the development of satellite remote sensing technology has made it possible for a human to
68
conduct macro dynamic monitoring of earth vegetation from space (Pettorelli et al. 2014; Weng 2002). In the past
69
decades, there has been an increase in the availability of satellite data measuring climate and other ecologically
70
relevant variables(Kerr & Ostrovsky 2003). These data offer opportunities to characterize ecosystem sensitivity at
71
high spatial resolution. Based on satellite-derived images, Seddon et al. (Seddon et al. 2016) present recently a novel
72
method to identify ecosystem sensitivity and memory effect to short-term climate variability by developing a
73
vegetation sensitivity index (VSI) that explores the linkage between variability in vegetation productivity (defined as
74
enhanced vegetation index, EVI) and three climate variables (namely air temperature, water availability, and cloud-
75
cover) on monthly time scales. The vegetation sensitivity index is a useful metric to quantitatively assess the
76
sensitivity of different ecosystems to climate variability (Huete 2016; Willis et al. 2018) and simultaneously takes
77
into account short-term climate effects and vegetation ‘memory effect’. 67
Since the 1970s, the development of satellite remote sensing technology has made it possible for a human to
68
conduct macro dynamic monitoring of earth vegetation from space (Pettorelli et al. 2014; Weng 2002). In the past
69
decades, there has been an increase in the availability of satellite data measuring climate and other ecologically
70
relevant variables(Kerr & Ostrovsky 2003). These data offer opportunities to characterize ecosystem sensitivity at
71
high spatial resolution. Based on satellite-derived images, Seddon et al. (Seddon et al. 2016) present recently a novel
72
method to identify ecosystem sensitivity and memory effect to short-term climate variability by developing a
73
vegetation sensitivity index (VSI) that explores the linkage between variability in vegetation productivity (defined as
74
enhanced vegetation index, EVI) and three climate variables (namely air temperature, water availability, and cloud-
75
cover) on monthly time scales. The vegetation sensitivity index is a useful metric to quantitatively assess the
76
sensitivity of different ecosystems to climate variability (Huete 2016; Willis et al. 2018) and simultaneously takes
77
into account short-term climate effects and vegetation ‘memory effect’. 78
Drylands (including arid and semi-arid regions) occupy over 41% of the global land surface area and are
79
inhabited by >2 billion people (Safriel et al. 2020). 40
Introduction There are several ways defining the vegetation’s sensitivity to climate variability. 58
For example, vegetation’s sensitivity refers to the degree and magnitude of vegetation response when the climate
59
anomaly occurs(You et al. 2018) or the degree to which a system changes after a disturbance(Li et al. 2018). In this
60
study, vegetation’s sensitivity is defined as the magnitude of vegetation response at the moment of the climate
61
anomaly(Tilman 1996). Additionally, Moulin S et al. (Moulin et al. 1997) found that due to the relatively slow
62
growth of vegetation, the response of vegetation to climate change often lags, therefore vegetation growth depends
63
both on current disturbances and the residual effects of past climate conditions. We called this phenomenon the
64
vegetation memory effect and it could be described as the persistence of trends in temporal changes of ecosystem
65
properties (Dash et al. 2014; Lhermitte et al. 2010; Simoniello et al. 2008). This ‘memory effect’ should be
66
considered when assessing the immediate response to short-term climate anomalies(De Keersmaecker et al. 2015). PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Manuscript to be reviewed 95
& Wang 2011) and predicting yield of crops (Kastens et al. 2005; Mkhabela et al. 2011).Therefore, we chose NDVI
96
to characterize vegetation coverage. Irtysh River basin is located in the arid and semi-arid region in the northwest of
97
China where water resources are scarce, and it is an important water source in this area and plays an important role
98
in regional economic development and ecological protection(Ye & Bai 2014). The Irtysh River is an international
99
river, which has complicated interests with neighboring countries in the international distribution of water resources,
100
prevention and control of water pollution, ecological maintenance and international regional cooperation. The
101
existing research(Huang et al. 2013a; Huang et al. 2012) showed that Irtysh River basin is sensitive to environmental
102
change. Therefore, we selected this area as our study area and analyzed the changes of its natural vegetation and
103
climate, and further identified the spatial differentiation in the sensitivity of the vegetation to climate changes, which
104
are of valuable reference significance for dealing with climate change and maintaining ecological stability of the
105
basin in the future. 95
& Wang 2011) and predicting yield of crops (Kastens et al. 2005; Mkhabela et al. 2011).Therefore, we chose NDVI
96
to characterize vegetation coverage. Irtysh River basin is located in the arid and semi-arid region in the northwest of
97
China where water resources are scarce, and it is an important water source in this area and plays an important role
98
in regional economic development and ecological protection(Ye & Bai 2014). The Irtysh River is an international
99
river, which has complicated interests with neighboring countries in the international distribution of water resources,
100
prevention and control of water pollution, ecological maintenance and international regional cooperation. The
101
existing research(Huang et al. 2013a; Huang et al. 2012) showed that Irtysh River basin is sensitive to environmental
102
change. Therefore, we selected this area as our study area and analyzed the changes of its natural vegetation and
103
climate, and further identified the spatial differentiation in the sensitivity of the vegetation to climate changes, which
104
are of valuable reference significance for dealing with climate change and maintaining ecological stability of the
105
basin in the future. Manuscript to be reviewed Temperature and precipitation are usually the main climatic
80
factors affecting vegetation activities (Li et al. 2015; Seddon et al. 2016; Yang et al. 2015). Arid area is usually
81
characterized by rare precipitation and high temperatures, and both water and heat have important effects on
82
vegetation growth, which makes ecosystems in arid and semi-arid regions are more vulnerable to climatic
83
disturbances(Rotenberg & Yakir 2010). Except for the direct but rare precipitation, soil water supplied by meltwater
84
from ice and snow in mountainous areas is primarily the main water source for vegetation growth in plain areas,
85
especially for forests and wetlands in river valleys. Thus, moisture of the soil is also a profound factor affecting
86
vegetation activities in the arid area. Compared to cloud-cover, soil moisture is far more influential on vegetation in
87
the arid area. TVDI characterizes the changes between vegetation index and land surface temperature, and is an
88
important indicator reflecting the soil moisture status(Grassini et al. 2010; Sandholt et al. 2002). So, we substituted
89
the climatic factor of cloud-cover in Seddon’s model by TVDI in our study. Additionally, among the all kinds of
90
vegetation indexes, NDVI is the most widely used and also is one of the earliest proposed. No other vegetation index
91
is capable to compare with NDVI in its widely use. NDVI has long been applied in monitoring vegetation
92
dynamics(Fensholt et al. 2009; Slayback et al. 2003)(, changes of vegetation phenology(Jeong et al. 2011; Ludeke et
93
al. 1996) and assessing land degradation(Oba et al. 2001; Thiam 2003). NDVI has also been used as basic parameter
94
in modeling of vegetation productivity(Schloss et al. 1999), terrestrial evapotranspiration(Maselli et al. 2014; Yang PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed 106
Based on the remote sensing data, we used the Mann-Kendall non-parametric rank statistical test to analyze the
107
vegetation dynamic and then calculated VSI to discuss the vegetation sensitivity. In this study, our aims are to (1)
108
map the spatial trends of vegetation cover in 2000-2018; (2) investigate the controlling factors of vegetation
109
sensitivity and resilience in Irtysh River basin. 106
Based on the remote sensing data, we used the Mann-Kendall non-parametric rank statistical test to analyze the
107
vegetation dynamic and then calculated VSI to discuss the vegetation sensitivity. In this study, our aims are to (1)
108
map the spatial trends of vegetation cover in 2000-2018; (2) investigate the controlling factors of vegetation
109
sensitivity and resilience in Irtysh River basin. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) 125
Data source and pre-processing 125
Data source and pre processing
126
Meteorological Data
127
Monthly precipitation and air temperature datasets of 72 meteorological stations in the territory of Xinjiang
128
province of China, where the study area located, were collected from the China Meteorological Data Service Center
129
(CMDC). The datasets cover the period of 2000–2018. These 72 meteorological stations are unevenly distributed in
130
space, and interpolation is the reliable and common practice to obtain continuous surface data(Guo et al. 2020). Both
131
precipitation and air temperature are strongly affected by topography. The interpolation of Australian National
132
University Spline or Anusplin, based on the thin plate spline algorithm, is primarily the first choice when
133
interpolation is necessary, as it takes the effect of topography into account (Hutchinson & Xu. 2013). A. Dewi
134
Hartkamp et al.(1999) and Claire H et al.(2001) have long confirmed the superiority of the thin plate spline over
135
other algorithms, such as inverse distance weighting method or kriging. Datasets of Climatic Research Unit (CRU) ,
136
WorldClim 1 and WorldClim 2 are all produced using Anusplin, and the China Meteorological Administration also
137
used the Anusplin in producing the daily gridded dataset(Zhao et al. 2014). Therefore, we also applied the Anusplin
138
to interpolate our raster surface of precipitation and air temperature. The pixel size was set to 250 m in the
139
interpolating to keep in line with that of the NDVI images. Among the 72 meteorological stations, 7 are in or at the
140
border of the study area (Fig.1)(Jarvis & Stuart 2001). 126
Meteorological Data
127
Monthly precipitation and air temperature datasets of 72 meteorological stations in the territory of Xinjiang
128
province of China, where the study area located, were collected from the China Meteorological Data Service Center
129
(CMDC). The datasets cover the period of 2000–2018. These 72 meteorological stations are unevenly distributed in
130
space, and interpolation is the reliable and common practice to obtain continuous surface data(Guo et al. 2020). Both
131
precipitation and air temperature are strongly affected by topography. The interpolation of Australian National
132
University Spline or Anusplin, based on the thin plate spline algorithm, is primarily the first choice when
133
interpolation is necessary, as it takes the effect of topography into account (Hutchinson & Xu. 2013). A. 110
Materials & Methods 110
Materials & Methods
111
Study area
112
Irtysh River basin is in the arid and semi-arid area of northwest China where water resources are scare
113
(45°40′~48°27′N, 85°30′~91°2′E, Fig.1). The region relates to the Altai Mountains in the north and
114
crosses into the northern edge of the fold system of the Junggar Basin in the south and is adjacent to Mongolia,
115
Kazakhstan, and other countries in the East and West. A typical temperate continent cold climate(Ju et al. 2015;
116
QiangJi & Wu 2017) dominates this area, with annual mean temperature ranging from 3.6℃ to 3.9℃, and the cold
117
air activities are frequent in winter and spring in mountainous areas, forming disasters such as blizzards,
118
snowstorms, and avalanches now and then. The annual mean precipitation is about 217.1 mm, the precipitation
119
increases gradually alone the increase of elevation, yet the northwest part is wetter than the southeast part in
120
general(Shen et al. 2007). The vegetation in the study area is mainly desert meadow and grassland, accounting for
121
91.58% of the whole vegetated area. The rest are herbaceous swamp distributed in and around the river valley, and 111
Study area
112
Irtysh River basin is in the arid and semi-arid area of northwest China where water resources are scare
113
(45°40′~48°27′N, 85°30′~91°2′E, Fig.1). The region relates to the Altai Mountains in the north and
114
crosses into the northern edge of the fold system of the Junggar Basin in the south and is adjacent to Mongolia,
115
Kazakhstan, and other countries in the East and West. A typical temperate continent cold climate(Ju et al. 2015;
116
QiangJi & Wu 2017) dominates this area, with annual mean temperature ranging from 3.6℃ to 3.9℃, and the cold
117
air activities are frequent in winter and spring in mountainous areas, forming disasters such as blizzards,
118
snowstorms, and avalanches now and then. The annual mean precipitation is about 217.1 mm, the precipitation
119
increases gradually alone the increase of elevation, yet the northwest part is wetter than the southeast part in
120
general(Shen et al. 2007). The vegetation in the study area is mainly desert meadow and grassland, accounting for
121
91.58% of the whole vegetated area. The rest are herbaceous swamp distributed in and around the river valley, and PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed 122
broad-leaved forest or shrub scattered in or near the swamp area. As growth of the crops are managed by farmers,
123
which disturbs the regulation of climate, we excluded farm and area of non-vegetation in our study. 122
broad-leaved forest or shrub scattered in or near the swamp area. As growth of the crops are managed by farmers,
123
which disturbs the regulation of climate, we excluded farm and area of non-vegetation in our study. 125
Data source and pre-processing Dewi
134
Hartkamp et al.(1999) and Claire H et al.(2001) have long confirmed the superiority of the thin plate spline over
135
other algorithms, such as inverse distance weighting method or kriging. Datasets of Climatic Research Unit (CRU) ,
136
WorldClim 1 and WorldClim 2 are all produced using Anusplin, and the China Meteorological Administration also
137
used the Anusplin in producing the daily gridded dataset(Zhao et al. 2014). Therefore, we also applied the Anusplin
138
to interpolate our raster surface of precipitation and air temperature. The pixel size was set to 250 m in the
139
interpolating to keep in line with that of the NDVI images. Among the 72 meteorological stations, 7 are in or at the
140
border of the study area (Fig.1)(Jarvis & Stuart 2001). 141
Remote sensing Data
142
Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI product (MOD13Q1) covering the period of
143
2000-2018 were used to determine the variation of vegetation cover. This MOD13Q1 product was 16-day NDVI
144
synthetic data using the Maximum Value Composite (MVC) method, and its pixel size is about 250 m×250 m. 145
Time series NDVI images are easily tarnished by noise signal received by the satellite sensors due to effects of the
146
atmosphere, clouds, geometric misregistration or many other uncontrollable factors(Goward et al. 1991). In hopes to
147
eliminate the noise and built high-quality NDVI time series, scholars have developed many kinds of smoothing
148
algorithms, such as the SPLINE-curve fitting, double logistic functions, Savitzky-Golay filter, harmonic analysis of
149
time series and so on(Cai et al. 2017; Pan et al. 2017). Among these smoothing algorithms, both Pan et al. (2017)
150
and Cai et al.(2017) have confirmed the Savitzky-Golay filter’s superiority. We followed the recommendation of 141
Remote sensing Data
142
Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI product (MOD13Q1) covering the period of
143
2000-2018 were used to determine the variation of vegetation cover. This MOD13Q1 product was 16-day NDVI
144
synthetic data using the Maximum Value Composite (MVC) method, and its pixel size is about 250 m×250 m. 145
Time series NDVI images are easily tarnished by noise signal received by the satellite sensors due to effects of the
146
atmosphere, clouds, geometric misregistration or many other uncontrollable factors(Goward et al. 1991). Manuscript to be reviewed We applied this method in our
172
data processing to get a high quality LST dataset and further guarantee our TVDI dataset can reflect the drought of
173
soil more accurately. 165
It is well known that there are ubiquitous data gaps in LST datasets because of non-overlapping satellite orbits,
166
cloud contamination, instrumental malfunction (Chen et al. 2011; Hu et al. 2014) and interpolation methods are
167
usually applied to fill the data gaps (Cai et al. 2017; Garcia 2010). Garcia (2010) have developed a fast and robust
168
smooth regression algorithm that combines the Discrete Cosine Transform (DCT) and the Penalized Least Square
169
approach (PLS) together with the Generalized Cross-Validation (GCV) criterion to fill data gaps. Liu et al. (2010)
170
tested and applied Garcia’s method on the reconstruction of the MODIS LST datasets covering the three continents
171
of South America, Africa and Asia，and confirmed its capability and robustness. We applied this method in our
172
data processing to get a high quality LST dataset and further guarantee our TVDI dataset can reflect the drought of
173
soil more accurately. Manuscript to be reviewed 151
Pan et al. and Cai et al. and performed the Savitzky-Golay filtering on our NDVI images to obtain high-quality
152
NDVI time series. And then we applied MVC to NDVIs covering every year of the period of 2000-2018 to get
153
NDVI that reflects the yearly highest growing level of vegetation, namely NDVI at yearly time scale. We used
154
yearly maximum NDVI to analyze the inter-annual dynamics of vegetation. We also applied MVC to NDVIs
155
covering every month of the period of 2000-2018 to get the monthly NDVI. 156
Temperature Vegetation Dryness Index (TVDI) can be used to characterize the degree of soil drought. This study
157
mainly uses the method proposed by Sanholt I et al. and Yao et al. (Sandholt et al. 2002; Yao et al. 2004) to
158
calculate TVDI. TVDI is a combination of vegetation index (VI) and land surface temperature (LST). Monthly
159
TVDI were calculated using the MODIS NDVI and the 8-day composite MODIS temperature product (MOD11A2). 160
The MOD11A2 product includes LST images of day and night, with a pix size about 1000 m×1000 m. The day LST
161
was used in the study. Every 4 LST images covering the whole corresponding month were averaged to get the
162
monthly LST time series data, and the pixel size was resampled to 250 m×250 m with the nearest-neighbor
163
resampling algorithm(Christman & Rogan 2012; Khan et al. 1995) integrated in toolbox of ArcGIS software to
164
match that the of NDVI images. 165
It is well known that there are ubiquitous data gaps in LST datasets because of non-overlapping satellite orbits,
166
cloud contamination, instrumental malfunction (Chen et al. 2011; Hu et al. 2014) and interpolation methods are
167
usually applied to fill the data gaps (Cai et al. 2017; Garcia 2010). Garcia (2010) have developed a fast and robust
168
smooth regression algorithm that combines the Discrete Cosine Transform (DCT) and the Penalized Least Square
169
approach (PLS) together with the Generalized Cross-Validation (GCV) criterion to fill data gaps. Liu et al. (2010)
170
tested and applied Garcia’s method on the reconstruction of the MODIS LST datasets covering the three continents
171
of South America, Africa and Asia，and confirmed its capability and robustness. 125
Data source and pre-processing In hopes to
147
eliminate the noise and built high-quality NDVI time series, scholars have developed many kinds of smoothing
148
algorithms, such as the SPLINE-curve fitting, double logistic functions, Savitzky-Golay filter, harmonic analysis of
149
time series and so on(Cai et al. 2017; Pan et al. 2017). Among these smoothing algorithms, both Pan et al. (2017)
150
and Cai et al.(2017) have confirmed the Savitzky-Golay filter’s superiority. We followed the recommendation of 141
Remote sensing Data
142
Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI product (MOD13Q1) covering the period of
143
2000-2018 were used to determine the variation of vegetation cover. This MOD13Q1 product was 16-day NDVI
144
synthetic data using the Maximum Value Composite (MVC) method, and its pixel size is about 250 m×250 m. 145
Time series NDVI images are easily tarnished by noise signal received by the satellite sensors due to effects of the
146
atmosphere, clouds, geometric misregistration or many other uncontrollable factors(Goward et al. 1991). In hopes to
147
eliminate the noise and built high-quality NDVI time series, scholars have developed many kinds of smoothing
148
algorithms, such as the SPLINE-curve fitting, double logistic functions, Savitzky-Golay filter, harmonic analysis of
149
time series and so on(Cai et al. 2017; Pan et al. 2017). Among these smoothing algorithms, both Pan et al. (2017)
150
and Cai et al.(2017) have confirmed the Savitzky-Golay filter’s superiority. We followed the recommendation of PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) 174
Mann-Kendall non-parametric rank statistical test 174
Mann Kendall non parametric rank statistical test
175
When using Mann-Kendall non-parametric test(Kendall 1990; Mann 1945) to test the possible trends of climatic
176
elements and time series, we assume that H0 indicates that the time series （x1,x2 ,…, xn） are independent of the
177
data sample, and there is no obvious trend; Assuming that H1 is a bilateral test, the distribution of xi and xj are
178
different for all i, j (i≠j), the calculation formula of the statistical variable S of the test is as follows: 174
Mann Kendall non parametric rank statistical test
175
When using Mann-Kendall non-parametric test(Kendall 1990; Mann 1945) to test the possible trends of climatic
176
elements and time series, we assume that H0 indicates that the time series （x1,x2 ,…, xn） are independent of the
177
data sample, and there is no obvious trend; Assuming that H1 is a bilateral test, the distribution of xi and xj are
178
different for all i, j (i≠j), the calculation formula of the statistical variable S of the test is as follows: 174
Mann Kendall non parametric rank statistical test
175
When using Mann-Kendall non-parametric test(Kendall 1990; Mann 1945) to test the possible trends of climatic
176
elements and time series, we assume that H0 indicates that the time series （x1,x2 ,…, xn） are independent of the
177
data sample, and there is no obvious trend; Assuming that H1 is a bilateral test, the distribution of xi and xj are
178
different for all i, j (i≠j), the calculation formula of the statistical variable S of the test is as follows: PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Manuscript to be reviewed Manuscript to be reviewed 179
𝑆= ∑𝑛‒ 1
𝑖= 1∑𝑛
𝑘= 𝑖+ 1𝑆𝑔𝑛(𝑥𝑘‒ 𝑥𝑖) 181
Among them,
182
𝑆𝑔𝑛（(𝜃)）={
1
𝜃> 0
0
𝜃= 0
‒ 1
𝜃< 0 183
（2） 184
S is normal distribution, the mean is 0, and the variance is as follows: 184
S is normal distribution, the mean is 0, and the variance is as follows: （3）
𝑉𝑎𝑟(𝑠) = [𝑛(𝑛‒ 1)(2𝑛+ 5) ‒ ∑𝑡𝑡(𝑡‒ 1)(2𝑡+ 5)] 18 185
（3）
𝑉𝑎𝑟(𝑠) = [𝑛(𝑛‒ 1)(2𝑛+ 5) ‒ ∑𝑡𝑡(𝑡‒ 1)(2𝑡+ 5)] 18 185
𝑉𝑎𝑟(𝑠) = [𝑛(𝑛‒ 1)(2𝑛+ 5) ‒ ∑𝑡𝑡(𝑡‒ 1)(2𝑡+ 5)] 18 （3） 186
Where t is the width of each unit. 174
Mann-Kendall non-parametric rank statistical test When n> 10, Zc converges to a standard normal distribution and can be 186
Where t is the width of each unit. When n> 10, Zc converges to a standard normal distribution and can be
187
calculated by the following formula 187
calculated by the following formula. 188
0
)
(
1
0
0
0
)
(
1
S
S
Var
S
S
S
S
Var
S
Zc 189
（4） （4） 190
At a given confidence level, when |Zc|>1.96, the changing trend reaches a significant level, |Zc|<1.96, the
𝛼
191
changing trend is not Significant; Zc>0, indicating that the changing trend is increasing, and Zc<0, it is decreasing. 190
At a given confidence level, when |Zc|>1.96, the changing trend reaches a significant level, |Zc|<1.96, the
𝛼 191
changing trend is not Significant; Zc>0, indicating that the changing trend is increasing, and Zc<0, it is decreasing. 192
Theil-Sen median trend analysis(Atta ur & Dawood 2017; Coen et al. 2020) could be used to quantify the trend of
193
time series data, and its calculation formula is as follows: 194
（5）
𝛽= 𝑀𝑒𝑑𝑖𝑎𝑛(
𝑥𝑖‒ 𝑥𝑗
𝑖‒ 𝑗) 194
（5）
𝛽= 𝑀𝑒𝑑𝑖𝑎𝑛(
𝑥𝑖‒ 𝑥𝑗
𝑖‒ 𝑗) （5） （5） 195
In the formula, 1<j<i<n，β means slope, a positive value means “uptrend”, and a negative value means
196
“downtrend”. 195
In the formula, 1<j<i<n，β means slope, a positive value means “uptrend”, and a negative value means
196
“downtrend”. 197
Identifying the sensitivity of vegetation to climate variability 198
The VSI is a novel and empirical metric developed by Seddon et al. (2016) that can quantify the sensitivity of
199
different vegetation areas to climate variability (Huete 2016; Willis et al. 2018). In this study, we tailored the
200
empirical methodology to identify vegetation sensitive to climate variability on the Irtysh River basin. 198
The VSI is a novel and empirical metric developed by Seddon et al. (2016) that can quantify the sensitivity of
199
different vegetation areas to climate variability (Huete 2016; Willis et al. 2018). In this study, we tailored the
200
empirical methodology to identify vegetation sensitive to climate variability on the Irtysh River basin. 198
The VSI is a novel and empirical metric developed by Seddon et al. (2016) that can quantify the sensitivity of
199
different vegetation areas to climate variability (Huete 2016; Willis et al. 2018). In this study, we tailored the
200
empirical methodology to identify vegetation sensitive to climate variability on the Irtysh River basin. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Manuscript to be reviewed 201
Firstly, for the climatic variables, we employed temperature, precipitation, and TVDI, instead of three climate
202
variables as in Seddon et al. (2016). Furthermore, we included the one-month-lagged NDVI monthly data as a fourth
203
variable in the regression to investigate the potential influence of memory effects driving vegetation dynamics. 204
Secondly, any month with a mean NDVI of <0.1 were excluded to reduce the potential impact of noisy data at
205
low NDVI values, which are attributed to areas with extremely sparse or inexistent vegetation cover(Zhang 2015;
206
Zhu et al. 2019). And to remove seasonal component underlying monthly time series, we de-trended the monthly
207
data and then we standardized the de-trended data utilizing the Z-score standardization formula: 208
(6)
𝑍𝑖,𝑗 =
𝑥𝑖,𝑗‒ 𝑥𝑗
𝜎𝑗 208
(6)
𝑍𝑖,𝑗 =
𝑥𝑖,𝑗‒ 𝑥𝑗
𝜎𝑗 (6)
𝑍𝑖,𝑗 =
𝑥𝑖,𝑗‒ 𝑥𝑗
𝜎𝑗 (6) (6) 209
Where xi,j is the detrended data in the jth month of the ith year, xj and
are the mean and standard deviation of
𝜎𝑗
210
the variable x in the jth month of all years, respectively. 209
Where xi,j is the detrended data in the jth month of the ith year, xj and
are the mean and standard deviation of
𝜎𝑗
210
the variable x in the jth month of all years, respectively. 209
Where xi,j is the detrended data in the jth month of the ith year, xj and
are the mean and standard deviation of
𝜎𝑗
210
the variable x in the jth month of all years, respectively. 210
the variable x in the jth month of all years, respectively. Manuscript to be reviewed 211
Thirdly, in this study, the sensitivity of vegetation to climate variability on the Irtysh River basin was primarily
212
calculated using AR1 multiple linear regression approach in each pixel, as follows: 211
Thirdly, in this study, the sensitivity of vegetation to climate variability on the Irtysh River basin was primarily
212
l
l t d
i
AR1
lti l li
i
h i
h i
l
f ll 212
calculated using AR1 multiple linear regression approach in each pixel, as follows: 213
(7)
𝑁𝐷𝑉𝐼𝑡= 𝛼× 𝑁𝐷𝑉𝐼𝑡‒ 1 + 𝛽× 𝑇𝑒𝑚𝑡+ 𝛾× 𝑃𝑟𝑒𝑡+ 𝛿× 𝑇𝑉𝐷𝐼𝑡+ 𝜀𝑡 213
(7)
𝑁𝐷𝑉𝐼𝑡= 𝛼× 𝑁𝐷𝑉𝐼𝑡‒ 1 + 𝛽× 𝑇𝑒𝑚𝑡+ 𝛾× 𝑃𝑟𝑒𝑡+ 𝛿× 𝑇𝑉𝐷𝐼𝑡+ 𝜀𝑡 213
𝑁𝐷𝑉𝐼𝑡= 𝛼× 𝑁𝐷𝑉𝐼𝑡‒ 1 + 𝛽× 𝑇𝑒𝑚𝑡+ 𝛾× 𝑃𝑟𝑒𝑡+ 𝛿× 𝑇𝑉𝐷𝐼𝑡+ 𝜀𝑡 (7) 214
Where
is the standardized NDVI at time t,
is the standardized NDVI anomaly at time t-1,
,
𝑁𝐷𝑉𝐼𝑡
𝑁𝐷𝑉𝐼𝑡‒ 1
𝑇𝑒𝑚𝑡
215
and
are the standardized temperature, precipitation, and TVDI at time t, respectively. is the residual
𝑃𝑟𝑒𝑡
𝑇𝑉𝐷𝐼𝑡
𝜀𝑡
216
term at time t, and
are coefficients for temperature, precipitation, TVDI and
of each pixel,
𝛼, 𝛽, 𝛾, 𝑎𝑛𝑑 𝛿
𝑁𝐷𝑉𝐼𝑡‒ 1
217
respectively. Each of
is a metric of ecosystem stability (Table 1; (De Keersmaecker et al. 2015)). 𝛼, 𝛽, 𝛾, 𝑎𝑛𝑑 𝛿
218
Compared to the correlation coefficient which can only indicate whether the ecosystem responds to climate
219
variability, the regressive coefficient can further reflect the response magnitude. 214
Where
is the standardized NDVI at time t,
is the standardized NDVI anomaly at time t-1,
,
𝑁𝐷𝑉𝐼𝑡
𝑁𝐷𝑉𝐼𝑡‒ 1
𝑇𝑒𝑚𝑡
215
and
are the standardized temperature, precipitation, and TVDI at time t, respectively. is the residual
𝑃𝑟𝑒𝑡
𝑇𝑉𝐷𝐼𝑡
𝜀𝑡
216
term at time t, and
are coefficients for temperature, precipitation, TVDI and
of each pixel,
𝛼, 𝛽, 𝛾, 𝑎𝑛𝑑 𝛿
𝑁𝐷𝑉𝐼𝑡‒ 1
217
respectively. Each of
is a metric of ecosystem stability (Table 1; (De Keersmaecker et al. 2015)). 𝛼, 𝛽, 𝛾, 𝑎𝑛𝑑 𝛿
218
Compared to the correlation coefficient which can only indicate whether the ecosystem responds to climate
219
variability, the regressive coefficient can further reflect the response magnitude. 220
Fourthly, to eliminate the effects of co-linearity between four climate variables, the principal components
221
regression (PCR) was also applied within each pixel to quantify the relative importance of each variable driving
222
variations in the monthly NDVI (Seddon et al. 2016). Manuscript to be reviewed The principal components that had significant relationships
223
with climate (p < 0.1) were selected, and we subsequently multiplied the loading scores of each variable by the PCR
224
coefficients. The product scores were summed to estimate the relative importance of each variable in driving
225
monthly changes in NDVI, which provided an empirical approach for mapping the relative importance of climate on
226
vegetation change (climate weights). 220
Fourthly, to eliminate the effects of co-linearity between four climate variables, the principal components
221
regression (PCR) was also applied within each pixel to quantify the relative importance of each variable driving
222
variations in the monthly NDVI (Seddon et al. 2016). The principal components that had significant relationships
223
with climate (p < 0.1) were selected, and we subsequently multiplied the loading scores of each variable by the PCR
224
coefficients. The product scores were summed to estimate the relative importance of each variable in driving
225
monthly changes in NDVI, which provided an empirical approach for mapping the relative importance of climate on
226
vegetation change (climate weights). PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Manuscript to be reviewed 227
The climate weights from each variable were rescaled between 0 and 1 (using the minimum and maximum value
228
of any of the climate coefficient values), to be used for calculations of vegetation sensitivity. To estimate th
229
variations of both the climate variables and NDVI on these time series, we used the residuals of a linear model fitte
230
to the mean-variance relationship of both the NDVI and climate variables for each pixel. We standardized thes
231
residuals between 0 and 100 for each variable. Our sensitivity metrics are the log10-transformed ratios of NDV
232
variability and each of the climate variables. Each ratio was then weighted according to the importance of th
233
climate variable to EVI variability by multiplying it by the value of the regression coefficient (climate weights). 234
Finally, we summed the sensitivity scores for each of our variables to identify areas of enhanced variability for th
235
period of study. 236
(8)
𝑉𝑆𝐼= 𝑇𝑒𝑚𝑤𝑒𝑖× 𝑇𝑒𝑚𝑠𝑒𝑛𝑠+ 𝑃𝑟𝑒𝑤𝑒𝑖× 𝑃𝑟𝑒𝑠𝑒𝑛𝑠+ 𝑇𝑉𝐷𝐼𝑤𝑒𝑖× 𝑇𝑉𝐷𝐼𝑠𝑒𝑛𝑠
237
Where VSI is vegetation sensitivity index, Temwei, Prewei and TVDIwei are the relative importance o
238
temperature, precipitation and TVDI on vegetation change (climate weights), respectively, and Temsens, Presen
239
and TVDIsens are the sensitivity of NDVI to temperature, precipitation and TVDI, respectively. The VSI has n
240
units and therefore provides relative information, wherein a high VSI value is associated with a high response rate o
241
vegetation productivity to climate variability. The detailed algorithm for calculating VSI and the R script can b
242
found in Seddon et al. (2016). 243
244
Results
245
Variances in vegetation cover and climatic factors
246
Based on the annual NDVI data of the study area in 2000-2018, we calculated the annual mean NDVI on a pixe
247
scale and divided it into 6 levels (Fig.2a) to analyze its spatial pattern. Vegetation coverage in the study are
248
increased from the southern plain to the northern mountain area. Vegetation in the plain area, except for the rive 227
The climate weights from each variable were rescaled between 0 and 1 (using the minimum and maximum values
228
of any of the climate coefficient values), to be used for calculations of vegetation sensitivity. Manuscript to be reviewed To estimate the
229
variations of both the climate variables and NDVI on these time series, we used the residuals of a linear model fitted
230
to the mean-variance relationship of both the NDVI and climate variables for each pixel. We standardized these
231
residuals between 0 and 100 for each variable. Our sensitivity metrics are the log10-transformed ratios of NDVI
232
variability and each of the climate variables. Each ratio was then weighted according to the importance of the
233
climate variable to EVI variability by multiplying it by the value of the regression coefficient (climate weights). 227
The climate weights from each variable were rescaled between 0 and 1 (using the minimum and maximum values
228
of any of the climate coefficient values), to be used for calculations of vegetation sensitivity. To estimate the
229
variations of both the climate variables and NDVI on these time series, we used the residuals of a linear model fitted
230
to the mean-variance relationship of both the NDVI and climate variables for each pixel. We standardized these
231
residuals between 0 and 100 for each variable. Our sensitivity metrics are the log10-transformed ratios of NDVI
232
variability and each of the climate variables. Each ratio was then weighted according to the importance of the
233
climate variable to EVI variability by multiplying it by the value of the regression coefficient (climate weights). 234
Finally we summed the sensitivity scores for each of our variables to identify areas of enhanced variability for the 233
climate variable to EVI variability by multiplying it by the value of the regression coefficient (climate weights). 234
Finally, we summed the sensitivity scores for each of our variables to identify areas of enhanced variability for the
235
period of study. 236
(8)
𝑉𝑆𝐼= 𝑇𝑒𝑚𝑤𝑒𝑖× 𝑇𝑒𝑚𝑠𝑒𝑛𝑠+ 𝑃𝑟𝑒𝑤𝑒𝑖× 𝑃𝑟𝑒𝑠𝑒𝑛𝑠+ 𝑇𝑉𝐷𝐼𝑤𝑒𝑖× 𝑇𝑉𝐷𝐼𝑠𝑒𝑛𝑠
237
Where VSI is vegetation sensitivity index, Temwei, Prewei and TVDIwei are the relative importance of
238
temperature, precipitation and TVDI on vegetation change (climate weights), respectively, and Temsens, Presens
239
and TVDIsens are the sensitivity of NDVI to temperature, precipitation and TVDI, respectively. The VSI has no
240
units and therefore provides relative information, wherein a high VSI value is associated with a high response rate of
241
vegetation productivity to climate variability. The detailed algorithm for calculating VSI and the R script can be
242
found in Seddon et al. (2016). Manuscript to be reviewed 243
244
Results
245
Variances in vegetation cover and climatic factors
246
Based on the annual NDVI data of the study area in 2000-2018, we calculated the annual mean NDVI on a pixel
247
scale and divided it into 6 levels (Fig.2a) to analyze its spatial pattern. Vegetation coverage in the study area
248
increased from the southern plain to the northern mountain area. Vegetation in the plain area, except for the river
249
valley, had low coverage (NDVI<0.2) and occupied 68.03% of the vegetated pixels. NDVI of the mountain areas in
250
the north and some plain areas in the west is about 0.2-0.6, accounting for 24.08% of the vegetated pixels. The
251
central valley area where the wetland and broad-leaved forest are distributed showed high coverage (NDVI>0.6),
252
accounting for 7.89% of the vegetated pixels. 253
In the period of 2000-2018, the annually averaged NDVI of the whole study area shows a significant increase
254
trend (statistics Zc=2.17, P<0.05), the changing rateβis 0.0017 (Fig.3). Spatially, the NDVI of 70.28% of the
255
vegetated pixels showed non-significant increase trend and they are mainly located in the low coverage region 234
Finally, we summed the sensitivity scores for each of our variables to identify areas of enhanced variability for the
235
period of study. 234
Finally, we summed the sensitivity scores for each of our variables to identify areas of enhanced variability for the
235
period of study. 236
(8)
𝑉𝑆𝐼= 𝑇𝑒𝑚𝑤𝑒𝑖× 𝑇𝑒𝑚𝑠𝑒𝑛𝑠+ 𝑃𝑟𝑒𝑤𝑒𝑖× 𝑃𝑟𝑒𝑠𝑒𝑛𝑠+ 𝑇𝑉𝐷𝐼𝑤𝑒𝑖× 𝑇𝑉𝐷𝐼𝑠𝑒𝑛𝑠 236
(8)
𝑉𝑆𝐼= 𝑇𝑒𝑚𝑤𝑒𝑖× 𝑇𝑒𝑚𝑠𝑒𝑛𝑠+ 𝑃𝑟𝑒𝑤𝑒𝑖× 𝑃𝑟𝑒𝑠𝑒𝑛𝑠+ 𝑇𝑉𝐷𝐼𝑤𝑒𝑖× 𝑇𝑉𝐷𝐼𝑠𝑒𝑛𝑠 (8) (8) 237
Where VSI is vegetation sensitivity index, Temwei, Prewei and TVDIwei are the relative importance of
238
temperature, precipitation and TVDI on vegetation change (climate weights), respectively, and Temsens, Presens
239
and TVDIsens are the sensitivity of NDVI to temperature, precipitation and TVDI, respectively. The VSI has no
240
units and therefore provides relative information, wherein a high VSI value is associated with a high response rate of
241
vegetation productivity to climate variability. The detailed algorithm for calculating VSI and the R script can be
242
found in Seddon et al. (2016). 245
Variances in vegetation cover and climatic factors
246
Based on the annual NDVI data of the study area in 2000-2018, we calculated the annual mean NDVI on a pixel
247
scale and divided it into 6 levels (Fig.2a) to analyze its spatial pattern. Manuscript to be reviewed Vegetation coverage in the study area
248
increased from the southern plain to the northern mountain area. Vegetation in the plain area, except for the river
249
valley, had low coverage (NDVI<0.2) and occupied 68.03% of the vegetated pixels. NDVI of the mountain areas in
250
the north and some plain areas in the west is about 0.2-0.6, accounting for 24.08% of the vegetated pixels. The
251
central valley area where the wetland and broad-leaved forest are distributed showed high coverage (NDVI>0.6),
252
accounting for 7.89% of the vegetated pixels. 253
In the period of 2000-2018, the annually averaged NDVI of the whole study area shows a significant increase
254
trend (statistics Zc=2.17, P<0.05), the changing rateβis 0.0017 (Fig.3). Spatially, the NDVI of 70.28% of the
255
vegetated pixels showed non-significant increase trend and they are mainly located in the low coverage region 253
In the period of 2000-2018, the annually averaged NDVI of the whole study area shows a significant increase
254
trend (statistics Zc=2.17, P<0.05), the changing rateβis 0.0017 (Fig.3). Spatially, the NDVI of 70.28% of the
255
vegetated pixels showed non-significant increase trend and they are mainly located in the low coverage region PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed 263
Climatic factors are the main driving forces for variation of vegetation. Therefore we performed Mann-Kendall
264
tests on precipitation, temperature and TVDI data to reveal their changing trends separately. The results (Fig.4)
265
showed that the precipitation and temperature in the study area showed a non-significant increasing trend from 2000
266
to 2018, meaning whether condition in the study area were getting warmer and wetter. Relatively, TVDI in the study
267
area showed a non-significant decreasing trend from 2000 to 2018, indicating that the soil moisture in the study area
268
is gradually increasing. The increased precipitation and temperature and decreased TVDI indicate that the
269
hydrothermal conditions and soils required for vegetation growth in the study area have been greatly improved in the
270
past 19 years, which promoted the increasing of the NDVI for the whole study area. 263
Climatic factors are the main driving forces for variation of vegetation. Therefore we performed Mann-Kendall
264
tests on precipitation, temperature and TVDI data to reveal their changing trends separately. The results (Fig.4)
265
showed that the precipitation and temperature in the study area showed a non-significant increasing trend from 2000
266
to 2018, meaning whether condition in the study area were getting warmer and wetter. Relatively, TVDI in the study
267
area showed a non-significant decreasing trend from 2000 to 2018, indicating that the soil moisture in the study area
268
is gradually increasing. The increased precipitation and temperature and decreased TVDI indicate that the
269
hydrothermal conditions and soils required for vegetation growth in the study area have been greatly improved in the
270
past 19 years, which promoted the increasing of the NDVI for the whole study area. g
y
272
Vegetation memory effects have been widely reported in water-limited ecosystems at various time-scales (Los et
273
al. 2006; Schwinning et al. 2004). Seddon et al. found that a one-month lag provided the best explanatory power for
274
vegetation responses to variability on short-term timescales. Therefore, we included the one-month-lagged NDVI
275
monthly data as a fourth variable in the regression to investigate the potential influence of memory effects driving
276
vegetation dynamics. The areas with high variance explained by the t−1 variable in the AR1 model, indicating
277
systems where memory effects play a more important role than contemporary climate conditions in determining
278
vegetation cover (Fig.5). Manuscript to be reviewed 256
dominated by desert meadow and grassland (Fig.2b). 17.81% of the vegetated pixels showed a significant increasing
257
trend, and are mainly located in the western part and central part in the south. Areas with NDVI showing decreasing
258
trend are sparsely distributed in the northern piedmont area and part of the west end. For the changing rate β，the
259
proportion of pixels with β>0 reached 89.04%. NDVI of areas in the central valley, mountains in the east and central
260
plains of the south showed the most rapid increase(β>0.002), the proportion is 26.79%. The increasing rate was
261
relatively low(0<β<0.001) for the plain of the east and southwest, and the central mountain of the south, which
262
occupied 62.26% of the vegetated pixels. 256
dominated by desert meadow and grassland (Fig.2b). 17.81% of the vegetated pixels showed a significant increasing
257
trend, and are mainly located in the western part and central part in the south. Areas with NDVI showing decreasing
258
trend are sparsely distributed in the northern piedmont area and part of the west end. For the changing rate β，the
259
proportion of pixels with β>0 reached 89.04%. NDVI of areas in the central valley, mountains in the east and central
260
plains of the south showed the most rapid increase(β>0.002), the proportion is 26.79%. The increasing rate was
261
relatively low(0<β<0.001) for the plain of the east and southwest, and the central mountain of the south, which
262
occupied 62.26% of the vegetated pixels. 263
Climatic factors are the main driving forces for variation of vegetation. Therefore we performed Mann-Kendall
264
tests on precipitation, temperature and TVDI data to reveal their changing trends separately. The results (Fig.4)
265
showed that the precipitation and temperature in the study area showed a non-significant increasing trend from 2000
266
to 2018, meaning whether condition in the study area were getting warmer and wetter. Relatively, TVDI in the study
267
area showed a non-significant decreasing trend from 2000 to 2018, indicating that the soil moisture in the study area
268
is gradually increasing. The increased precipitation and temperature and decreased TVDI indicate that the
269
hydrothermal conditions and soils required for vegetation growth in the study area have been greatly improved in the
270
past 19 years, which promoted the increasing of the NDVI for the whole study area. Manuscript to be reviewed 295
Specifically, vegetation memory effects (α) presented a quadratic parabola relationship with both precipitation
296
(R2=0.202, P<0.05, Fig.6b) and temperature (R2=0.155, P<0.05, Fig.6c). Manuscript to be reviewed The larger the t-1 coefficient weight (that is, coefficient α), the stronger the memory effect,
279
and the weaker the sensitivity (with lower VSI). g
y
272
Vegetation memory effects have been widely reported in water-limited ecosystems at various time-scales (Los et
273
al. 2006; Schwinning et al. 2004). Seddon et al. found that a one-month lag provided the best explanatory power for
274
vegetation responses to variability on short-term timescales. Therefore, we included the one-month-lagged NDVI
275
monthly data as a fourth variable in the regression to investigate the potential influence of memory effects driving
276
vegetation dynamics. The areas with high variance explained by the t−1 variable in the AR1 model, indicating
277
systems where memory effects play a more important role than contemporary climate conditions in determining
278
vegetation cover (Fig.5). The larger the t-1 coefficient weight (that is, coefficient α), the stronger the memory effect,
279
and the weaker the sensitivity (with lower VSI). 280
Vegetation showed strong memory effects (α>0.4) across almost the whole study area (Fig.5), especially in some
281
parts of the eastern plain where the coefficient reached more than 0.6, and the area with α>0.6 accounted for 27.27%
282
of the vegetated pixels in the study area. Coefficient α of areas in the northwest, the border of the east and river
283
valley in the middle is relatively small (α<0.4), indicating weaker memory effects, and the areas proportion is 280
Vegetation showed strong memory effects (α>0.4) across almost the whole study area (Fig.5), especially in some
281
parts of the eastern plain where the coefficient reached more than 0.6, and the area with α>0.6 accounted for 27.27%
282
of the vegetated pixels in the study area. Coefficient α of areas in the northwest, the border of the east and river
283
valley in the middle is relatively small (α<0.4), indicating weaker memory effects, and the areas proportion is PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed 284
20.60%. Yet for most parts of the study area, the coefficient α is about 0.4-0.6, and the area proportion reached
285
52.13%. Notably, vegetation with big NDVI showed weak memory effect in general, such as the herbaceous swamp
286
and broadleaf forest in the river valley and grassland in the mountain area of the north border. In contrast, the desert
287
grass in the plain area showed strong memory effect. 284
20.60%. Yet for most parts of the study area, the coefficient α is about 0.4-0.6, and the area proportion reached
285
52.13%. Notably, vegetation with big NDVI showed weak memory effect in general, such as the herbaceous swamp
286
and broadleaf forest in the river valley and grassland in the mountain area of the north border. In contrast, the desert
287
grass in the plain area showed strong memory effect. 288
As shown in Fig.6, the memory effect tends to change along the gradient of NDVI and climatic factors. To
289
identify the controlling factors for vegetation memory effects, we regressed α (t-1 coefficient weight) against three
290
climatic factors (precipitation, temperature, and TVDI) and vegetation cover (defined as NDVI). And considering
291
the hydrothermal conditions required for vegetation growth, we selected the mean NDVI, mean precipitation, mean
292
temperature, and mean TVDI of the growing season (GS) to participate in the regression. The results showed that
293
vegetation memory effects (α) decreased logarithmically as NDVI increased (R2=0.67, P<0.05, Fig.6a). Relatively,
294
vegetation memory effects (α) increased logarithmically as TVDI increased (R2=0.407, P<0.05, Fig.6d). 295
Specifically, vegetation memory effects (α) presented a quadratic parabola relationship with both precipitation
296
(R2=0.202, P<0.05, Fig.6b) and temperature (R2=0.155, P<0.05, Fig.6c). 288
As shown in Fig.6, the memory effect tends to change along the gradient of NDVI and climatic factors. To
289
identify the controlling factors for vegetation memory effects, we regressed α (t-1 coefficient weight) against three
290
climatic factors (precipitation, temperature, and TVDI) and vegetation cover (defined as NDVI). And considering
291
the hydrothermal conditions required for vegetation growth, we selected the mean NDVI, mean precipitation, mean
292
temperature, and mean TVDI of the growing season (GS) to participate in the regression. The results showed that
293
vegetation memory effects (α) decreased logarithmically as NDVI increased (R2=0.67, P<0.05, Fig.6a). Relatively,
294
vegetation memory effects (α) increased logarithmically as TVDI increased (R2=0.407, P<0.05, Fig.6d). 297
Vegetation sensitivity to climatic variables 297
Vegetation sensitivity to climatic variables
298
Compared to strong memory effects, the VSI in the study area is rather low, with 60.69% of the vegetated pixels
299
where VSI is less than 30 (Fig.7b). Spatially, areas of big VSI (VSI>30) generally overlap that of weak memory
300
effects (α<0.4), such as the grassland in the north border and herbaceous swamp and broadleaf forest in the river
301
valley, indicating that areas with higher NDVI usually shows weaker memory effects and higher sensitivity to
302
climate variability over the past 19 years. Areas of the desert plain show low sensitivity (VSI>30) to climate
303
variability, and overlap the areas of strong memory effects (α>0.4). 298
Compared to strong memory effects, the VSI in the study area is rather low, with 60.6 304
The relative importance of three climate variables (temperature, precipitation, and TVDI) to vegetation sensitivity
305
also displayed clear spatially heterogeneity across the study area (Fig.8). Most areas are more sensitive to
306
precipitation, mainly distributed in the southern and central plains dominated by desert meadow. While variation in
307
vegetation cover (defined as NDVI) of the southeastern areas were mainly affected by a combination of precipitation
308
and temperature, and the northern part of this area is affected by a combination of TVDI and temperature. 309
Additionally, vegetation cover in the northwest areas was mainly driven by precipitation and TVDI. Remarkably,
310
the central river valley dominated by herbaceous swamp was more sensitive to TVDI. And the mountain areas with
311
higher elevations in the north are more sensitive to both temperature and precipitation. 304
The relative importance of three climate variables (temperature, precipitation, and TVDI) to vegetation sensitivity
305
also displayed clear spatially heterogeneity across the study area (Fig.8). Most areas are more sensitive to
306
precipitation, mainly distributed in the southern and central plains dominated by desert meadow. While variation in
307
vegetation cover (defined as NDVI) of the southeastern areas were mainly affected by a combination of precipitation
308
and temperature, and the northern part of this area is affected by a combination of TVDI and temperature. 309
Additionally, vegetation cover in the northwest areas was mainly driven by precipitation and TVDI. Remarkably,
310
the central river valley dominated by herbaceous swamp was more sensitive to TVDI. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed 2012; Ye & Bai 2014),
335
which can timely recharge the groundwater aquifer in the valley area, thus increasing the soil moisture in this area,
336
and TVDI presents a decreasing trend. In addition to the supply of rainfall, the vegetation cover in this region
337
showed a significant trend of increase from 2000 to 2018, and the increase rate was also very fast. Studies (Jiang et
338
al. 2017) have shown that the impact of air temperature on vegetation growth is topographically different. In the
339
central river valley where it is relatively wet, elevated temperature can promote plant photosynthetic activity and 313
Discussion 313
Discussion
314
Disentangling the driving factors for variations in vegetation cover
315
Irtysh River basin is located in arid and semi-arid region. Scarce precipitation and high temperature lead to large
316
evapotranspiration and low soil water storage in this area, which is not conducive to the growth of vegetation,
317
especially in the low land of the southern plain. Therefore, most part of the study area is dominated by desert
318
vegetation. However, as the altitude increases, the precipitation increases and temperature decreases (Navarro et al. 319
2020), and this relieves the severe climatic restrictions. So, grassland in areas of high altitude, mainly the mountain
320
areas in the north, is well developed and the vegetation coverage is also high (0.4<NDVI<0.6). The river valley can
321
rely on the rivers to supply ample water required for vegetation growth, so the herbaceous swamp and broadleaf
322
forest with the highest NDVI (NDVI>0.6) are well developed in this area. 314
Disentangling the driving factors for variations in vegetation cover
315
Irtysh River basin is located in arid and semi-arid region. Scarce precipitation and high temperature lead to large
316
evapotranspiration and low soil water storage in this area, which is not conducive to the growth of vegetation,
317
especially in the low land of the southern plain. Therefore, most part of the study area is dominated by desert
318
vegetation. However, as the altitude increases, the precipitation increases and temperature decreases (Navarro et al. 319
2020), and this relieves the severe climatic restrictions. So, grassland in areas of high altitude, mainly the mountain
320
areas in the north, is well developed and the vegetation coverage is also high (0.4<NDVI<0.6). Manuscript to be reviewed The river valley can
321
rely on the rivers to supply ample water required for vegetation growth, so the herbaceous swamp and broadleaf
322
forest with the highest NDVI (NDVI>0.6) are well developed in this area. 314
Disentangling the driving factors for variations in vegetation cover 323
The results of the Mann-Kendall trend test of climatic factors show that the temperature and precipitation in the
324
study area showed an increasing trend from 2000 to 2018, and the TVDI showed a decreasing trend. This is
325
consistent with the findings of Huang et al (Huang et al. 2013b). The analysis of the results indicate that the
326
hydrothermal and soil conditions required for vegetation growth in the Irtysh River basin have been greatly
327
improved. However, with the change of altitude gradient and climatic factors, the variation trend (Zc) and rate (β) of
328
vegetation cover show obvious spatial heterogeneity. The increase of rainfall brings abundant water resources to the
329
desert plain area in the southern study area, improves the available moisture content of the local soil, thus promoting
330
the absorption of plant nutrients, which is conducive to the improvement of water utilization rate of plants, and
331
greatly promotes the growth of vegetation and the increasing rate of local vegetation cover. The vegetation growth
332
in the central valley, which is dominated by herbaceous marshes, mainly depends on rivers to supply groundwater to
333
meet the requirements of soil moisture. In recent years, with the continuous strengthening of national ecological
334
protection, the ecological water volume of Irtysh River has been well guaranteed (Yang et al. 2012; Ye & Bai 2014),
335
which can timely recharge the groundwater aquifer in the valley area, thus increasing the soil moisture in this area,
336
and TVDI presents a decreasing trend. In addition to the supply of rainfall, the vegetation cover in this region
337
showed a significant trend of increase from 2000 to 2018, and the increase rate was also very fast. Studies (Jiang et
338
al. 2017) have shown that the impact of air temperature on vegetation growth is topographically different. In the
339
central river valley where it is relatively wet, elevated temperature can promote plant photosynthetic activity and
340
thus lead to a positive response of vegetation growth. 297
Vegetation sensitivity to climatic variables And the mountain areas with
311
higher elevations in the north are more sensitive to both temperature and precipitation. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Manuscript to be reviewed 312
313
Discussion
314
Disentangling the driving factors for variations in vegetation cover
315
Irtysh River basin is located in arid and semi-arid region. Scarce precipitation and high temperature lead to large
316
evapotranspiration and low soil water storage in this area, which is not conducive to the growth of vegetation,
317
especially in the low land of the southern plain. Therefore, most part of the study area is dominated by desert
318
vegetation. However, as the altitude increases, the precipitation increases and temperature decreases (Navarro et al. 319
2020), and this relieves the severe climatic restrictions. So, grassland in areas of high altitude, mainly the mountain
320
areas in the north, is well developed and the vegetation coverage is also high (0.4<NDVI<0.6). The river valley can
321
rely on the rivers to supply ample water required for vegetation growth, so the herbaceous swamp and broadleaf
322
forest with the highest NDVI (NDVI>0.6) are well developed in this area. 323
The results of the Mann-Kendall trend test of climatic factors show that the temperature and precipitation in the
324
study area showed an increasing trend from 2000 to 2018, and the TVDI showed a decreasing trend. This is
325
consistent with the findings of Huang et al (Huang et al. 2013b). The analysis of the results indicate that the
326
hydrothermal and soil conditions required for vegetation growth in the Irtysh River basin have been greatly
327
improved. However, with the change of altitude gradient and climatic factors, the variation trend (Zc) and rate (β) of
328
vegetation cover show obvious spatial heterogeneity. The increase of rainfall brings abundant water resources to the
329
desert plain area in the southern study area, improves the available moisture content of the local soil, thus promoting
330
the absorption of plant nutrients, which is conducive to the improvement of water utilization rate of plants, and
331
greatly promotes the growth of vegetation and the increasing rate of local vegetation cover. The vegetation growth
332
in the central valley, which is dominated by herbaceous marshes, mainly depends on rivers to supply groundwater to
333
meet the requirements of soil moisture. In recent years, with the continuous strengthening of national ecological
334
protection, the ecological water volume of Irtysh River has been well guaranteed (Yang et al. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Environmental impacts on vegetation memory effect during growth season 348
Environmental impacts on vegetation memory effect during growth season
349
We can see in Fig.5 that areas with the strongest memory effects are generally located in the desert pain of the
350
southeast, where the NDVI is the smallest and the drought the strongest. And areas with weakest memory effects is
351
gathered in the mountain areas of the north and river valleys in the middle, where the NDVI is the biggest and the
352
drought the weakest. This character in the matching between the memory effects and both of NDVI and drought can
353
also be seen in the clean decreasing trend of metric α along with the increasing NDVI and also the increasing trend
354
with the increasing TVDI (Fig.6a, 6d). Vegetation in the arid area or desert are usually characterized by their strong
355
capability to coping with disturbances in climatic factors, this can be seen in the constant and largely stable low
356
productivity conditions despite large climate variability and also strong cyclical variability with periods of very low
357
and stable NDVI. So vegetation of these areas usually show strong memory effects. This contrasts to areas with high
358
NDVI, such as the river valley and the mountain area, where the river water and the more precipitation can moderate
359
the severe drought and provide better conditions for vegetation growth, yet the growth of vegetation is restricted by
360
the variation in water supply. 361
In addition, vegetation memory effects in the study area does not show a clear linear relationship with temperature
362
and precipitation, but we can see from the figure (Fig.6b, 6c) that there is an obvious inflection point in the image,
363
which means there a threshold in both the precipitation and temperature effects on memory effects. This might be
364
related to the co-effect of temperature and precipitation on vegetation. For areas in the arid region, altitude usually
365
controls the spatial differentiation of precipitation and temperature, therefore relationship between the memory
366
effects and climatic factors are branded with the influences of altitude on temperature and precipitation. Plain of low
367
altitude is usually characterized by high temperature and scare precipitation, whereas mountain areas are usually
368
characterized by low temperature and abundant precipitation. Manuscript to be reviewed In the northern study region where it is relatively dry, increase 323
The results of the Mann-Kendall trend test of climatic factors show that the temperature and precipitation in the
324
study area showed an increasing trend from 2000 to 2018, and the TVDI showed a decreasing trend. This is
325
consistent with the findings of Huang et al (Huang et al. 2013b). The analysis of the results indicate that the
326
hydrothermal and soil conditions required for vegetation growth in the Irtysh River basin have been greatly
327
improved. However, with the change of altitude gradient and climatic factors, the variation trend (Zc) and rate (β) of
328
vegetation cover show obvious spatial heterogeneity. The increase of rainfall brings abundant water resources to the
329
desert plain area in the southern study area, improves the available moisture content of the local soil, thus promoting
330
the absorption of plant nutrients, which is conducive to the improvement of water utilization rate of plants, and
331
greatly promotes the growth of vegetation and the increasing rate of local vegetation cover. The vegetation growth
332
in the central valley, which is dominated by herbaceous marshes, mainly depends on rivers to supply groundwater to
333
meet the requirements of soil moisture. In recent years, with the continuous strengthening of national ecological
334
protection, the ecological water volume of Irtysh River has been well guaranteed (Yang et al. 2012; Ye & Bai 2014),
335
which can timely recharge the groundwater aquifer in the valley area, thus increasing the soil moisture in this area,
336
and TVDI presents a decreasing trend. In addition to the supply of rainfall, the vegetation cover in this region
337
showed a significant trend of increase from 2000 to 2018, and the increase rate was also very fast. Studies (Jiang et
338
al. 2017) have shown that the impact of air temperature on vegetation growth is topographically different. In the
339
central river valley where it is relatively wet, elevated temperature can promote plant photosynthetic activity and
340
thus lead to a positive response of vegetation growth. In the northern study region where it is relatively dry, increase PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed 341
of temperature can intensify the water deficit through elevated evaporation and thus causes a negative response of
342
NDVI. In particular, although the climatic conditions in the study area showed a pattern of improvement, the impact
343
of human activities on vegetation cover could not be ignored. The research results of Yang et al.(Han et al. 2013)
344
showed that from 1990 to 2010, the overall landscape pattern of the Irtysh River Basin tended to be fragmented,
345
with serious spatial heterogeneity, which was increasingly affected by human activities over time. The research
346
results of this paper also reflect this phenomenon. The regions with different elevations have different responses to
347
the same climatic factors. 348
Environmental impacts on vegetation memory effect during growth season
349
We can see in Fig.5 that areas with the strongest memory effects are generally located in the desert pain of the
350
southeast, where the NDVI is the smallest and the drought the strongest. And areas with weakest memory effects is
351
gathered in the mountain areas of the north and river valleys in the middle, where the NDVI is the biggest and the
352
drought the weakest. This character in the matching between the memory effects and both of NDVI and drought can
353
also be seen in the clean decreasing trend of metric α along with the increasing NDVI and also the increasing trend
354
with the increasing TVDI (Fig.6a, 6d). Vegetation in the arid area or desert are usually characterized by their strong
355
capability to coping with disturbances in climatic factors, this can be seen in the constant and largely stable low
356
productivity conditions despite large climate variability and also strong cyclical variability with periods of very low
357
and stable NDVI. So vegetation of these areas usually show strong memory effects. This contrasts to areas with high
358
NDVI, such as the river valley and the mountain area, where the river water and the more precipitation can moderate
359
the severe drought and provide better conditions for vegetation growth, yet the growth of vegetation is restricted by
360
the variation in water supply. Environmental impacts on vegetation memory effect during growth season So, the limiting factor on the growth of vegetation
369
changes gradually from precipitation to temperature along the variation in altitude, which results in vegetation 361
In addition, vegetation memory effects in the study area does not show a clear linear relationship with temperature
362
and precipitation, but we can see from the figure (Fig.6b, 6c) that there is an obvious inflection point in the image,
363
which means there a threshold in both the precipitation and temperature effects on memory effects. This might be
364
related to the co-effect of temperature and precipitation on vegetation. For areas in the arid region, altitude usually
365
controls the spatial differentiation of precipitation and temperature, therefore relationship between the memory
366
effects and climatic factors are branded with the influences of altitude on temperature and precipitation. Plain of low
367
altitude is usually characterized by high temperature and scare precipitation, whereas mountain areas are usually
368
characterized by low temperature and abundant precipitation. So, the limiting factor on the growth of vegetation
369
changes gradually from precipitation to temperature along the variation in altitude, which results in vegetation 372
Spatial heterogeneity of VSI distribution 372
Spatial heterogeneity of VSI distribution
373
VSI reflects the sensitivity of vegetation cover to climate change, and we can identify regions that exhibits
374
amplified responses to climate variability through VSI. While Memory effect measures the capability of vegetation
375
returning to its normal state after suffering the disturbance. Specially, areas with low VSI values showed the largest
376
memory effect (Seddon et al. 2016), which is consistent with our study results（Fig.5 and Fig.7a）.The desert plain
377
area in the south of the study area has low sensitivity to climate change and strong vegetation memory effect. When
378
the adverse conditions for vegetation growth are generated due to the vicious climate development or other
379
disturbances in the region, the vegetation will make a hysteresis response to such changes, so that the ecosystem can
380
make timely adjustments to the environmental deterioration. Different types of vegetation respond differently. The
381
main vegetation type is desert meadow, which is a short-lived plant that survives in arid areas by escaping drought
382
(Guo et al. 2004; Lu et al. 2019). The ephemeral plants germinate and grow quickly in spring when the elevated
383
temperature melts the frozen soil or the covering snow and complete their life cycle before the coming of the hot and
384
dry summer. In addition, studies (J.Z. & H.J. 2015) have shown that the water content of soil at different depths is
385
affected differently by precipitation. Among them, the water content of shallow soil (0-20cm) is most affected by
386
precipitation. Desert meadow in arid and semi-arid areas have relatively short rooting system that mainly absorb
387
moisture from shallow soil, so the vegetation changes in this area are more sensitive to precipitation. 388 372
Spatial heterogeneity of VSI distribution
373
VSI reflects the sensitivity of vegetation cover to climate change, and we can identify regions that exhibits
374
amplified responses to climate variability through VSI. While Memory effect measures the capability of vegetation
375
returning to its normal state after suffering the disturbance. Specially, areas with low VSI values showed the largest
376
memory effect (Seddon et al. 2016), which is consistent with our study results（Fig.5 and Fig.7a）.The desert plain
377
area in the south of the study area has low sensitivity to climate change and strong vegetation memory effect. Manuscript to be reviewed 370
changes in certain areas are co-affected of temperature and precipitation and the threshold in both the precipitation
371
and temperature effects on memory effects. 370
changes in certain areas are co-affected of temperature and precipitation and the threshold in both the precipitation
371
and temperature effects on memory effects. 404
Conclusions 405
This study applied a new method and remote sensing datasets of high temporal resolution to quantify the
406
sensitivity and memory effects of vegetation in the Irtysh River, and further reveal the mechanism of vegetation
407
response to climate change at the regional scale. We found that the variation trend of precipitation, temperature and
408
TVDI all indicated that the climate condition of the Irtysh River basin had been greatly improved, and the vegetation
409
coverage also showed overall increasing trend. From the south to the north of the study area, with the change of
410
topographic and geomorphic features, the memory effect of vegetation and its sensitivity to different climatic factors
411
showed obvious spatial heterogeneity. It is mainly manifested in the following aspects, the memory effect of
412
vegetation in the southern desert plain was stronger and the plants there are more sensitive to precipitation, while the
413
herbaceous swamp and broad-leaf forest in the central valley showed weaker memory effect and were more
414
sensitive to TVDI. The temperate steppe in the northern mountain is highly sensitive to climate change and were
415
more affected by the combination of both precipitation and temperature. These results will help us locate different
416
ecological protection environment types more accurately in the future basin management process, and develop
417
optimal adaptive ecological protection strategies to protect this vulnerable ecosystem. 405
This study applied a new method and remote sensing datasets of high temporal resolution to quantify the
406
sensitivity and memory effects of vegetation in the Irtysh River, and further reveal the mechanism of vegetation
407
response to climate change at the regional scale. We found that the variation trend of precipitation, temperature and
408
TVDI all indicated that the climate condition of the Irtysh River basin had been greatly improved, and the vegetation
409
coverage also showed overall increasing trend. From the south to the north of the study area, with the change of
410
topographic and geomorphic features, the memory effect of vegetation and its sensitivity to different climatic factors
411
showed obvious spatial heterogeneity. 372
Spatial heterogeneity of VSI distribution When
378
the adverse conditions for vegetation growth are generated due to the vicious climate development or other
379
disturbances in the region, the vegetation will make a hysteresis response to such changes, so that the ecosystem can
380
make timely adjustments to the environmental deterioration. Different types of vegetation respond differently. The
381
main vegetation type is desert meadow, which is a short-lived plant that survives in arid areas by escaping drought
382
(Guo et al. 2004; Lu et al. 2019). The ephemeral plants germinate and grow quickly in spring when the elevated
383
temperature melts the frozen soil or the covering snow and complete their life cycle before the coming of the hot and
384
dry summer. In addition, studies (J.Z. & H.J. 2015) have shown that the water content of soil at different depths is
385
affected differently by precipitation. Among them, the water content of shallow soil (0-20cm) is most affected by
386
precipitation. Desert meadow in arid and semi-arid areas have relatively short rooting system that mainly absorb
387
moisture from shallow soil, so the vegetation changes in this area are more sensitive to precipitation. 388
In contrast to the desert plain area in the south, the central Irtysh River valley showed higher sensitivity and
389
weaker memory effect. The central river valley area is dominated by herbaceous swamps, which are typical low-
390
level swamps, with year-round accumulation of water or drenched soil (H.Y. et al. 2020), and the water supply
391
mainly depends on the Irtysh River. The local vegetation is dominated by perennial plants such as caress and
392
gramineous plants. Many plants have short root systems and mainly absorb shallow soil water, which are highly
393
dependent on soil moisture conditions. Therefore, when the amount of river water decreases, the soil water content
394
of the swamp will also decrease, exerting a hard impact on the growth of herbs. Additionally, TVDI is an index
395
reflecting soil moisture, smaller TVDI indicates that the wetter soil (Sandholt et al. 2002), so the vegetation on the
396
site is more sensitive to TVDI. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Manuscript to be reviewed Compared to the powerful ability of the desert plants in coping the severe drought, temperate steppe in
400
the mountain area is well developed because the elevated altitude relieves both the restrictions of scarce
401
precipitation and high temperature and provides hydrothermal conditions suitable for the grass plants, which makes
402
variations in the coverage of temperate steppe are sensitive to both the precipitation and temperature. 403 397
It is worth noting that the northern mountain regions show strong sensitivity to climate change (VSI>50 in some
398
areas). And the vegetation variation in this area is mainly affected by the combined effects of temperature and
399
precipitation. Compared to the powerful ability of the desert plants in coping the severe drought, temperate steppe in
400
the mountain area is well developed because the elevated altitude relieves both the restrictions of scarce
401
precipitation and high temperature and provides hydrothermal conditions suitable for the grass plants, which makes
402
variations in the coverage of temperate steppe are sensitive to both the precipitation and temperature. Manuscript to be reviewed 397
It is worth noting that the northern mountain regions show strong sensitivity to climate change (VSI>50 in some
398
areas). And the vegetation variation in this area is mainly affected by the combined effects of temperature and
399
precipitation. Compared to the powerful ability of the desert plants in coping the severe drought, temperate steppe in
400
the mountain area is well developed because the elevated altitude relieves both the restrictions of scarce
401
precipitation and high temperature and provides hydrothermal conditions suitable for the grass plants, which makes
402
variations in the coverage of temperate steppe are sensitive to both the precipitation and temperature. 403
404
Conclusions
405
This study applied a new method and remote sensing datasets of high temporal resolution to quantify the
406
sensitivity and memory effects of vegetation in the Irtysh River, and further reveal the mechanism of vegetation
407
response to climate change at the regional scale. We found that the variation trend of precipitation, temperature and
408
TVDI all indicated that the climate condition of the Irtysh River basin had been greatly improved, and the vegetation
409
coverage also showed overall increasing trend. From the south to the north of the study area, with the change of
410
topographic and geomorphic features, the memory effect of vegetation and its sensitivity to different climatic factors
411
showed obvious spatial heterogeneity. It is mainly manifested in the following aspects, the memory effect of
412
vegetation in the southern desert plain was stronger and the plants there are more sensitive to precipitation, while the
413
herbaceous swamp and broad-leaf forest in the central valley showed weaker memory effect and were more
414
sensitive to TVDI. The temperate steppe in the northern mountain is highly sensitive to climate change and were
415
more affected by the combination of both precipitation and temperature. These results will help us locate different
416
ecological protection environment types more accurately in the future basin management process, and develop
417
optimal adaptive ecological protection strategies to protect this vulnerable ecosystem. 418
419
Acknowledgements
420
This work was supported by the Xinjiang Tianshan Youth Program (2019Q006)，West Light Foundation of 397
It is worth noting that the northern mountain regions show strong sensitivity to climate change (VSI>50 in some
398
areas). And the vegetation variation in this area is mainly affected by the combined effects of temperature and
399
precipitation. 425
Anav A, and Mariotti A. 2011. Sensitivity of natural vegetation to climate change in the Euro-Mediterranean area.
426
Climate Research 46:277-292. 10.3354/cr00993 Manuscript to be reviewed 427
Atta ur R, and Dawood M. 2017. Spatio-statistical analysis of temperature fluctuation using Mann-Kendall and Sen's
428
slope approach. Climate Dynamics 48:783-797. 10.1007/s00382-016-3110-y 427
Atta ur R, and Dawood M. 2017. Spatio-statistical analysis of temperature fluctuation using Mann-Kendall and Sen's
428
slope approach. Climate Dynamics 48:783-797. 10.1007/s00382-016-3110-y 429
Cai Z, Jonsson P, Jin H, and Eklundh L. 2017. Performance of Smoothing Methods for Reconstructing NDVI Time-
430
Series and Estimating Vegetation Phenology from MODIS Data. Remote Sensing 9. 10.3390/rs9121271
431 431
Chen S, Chen X, Chen W, Su Y, and Li D. 2011. A simple retrieval method of land surface temperature from AMSR-E
432
passive microwave data-A case study over Southern China during the strong snow disaster of 2008. 433
International Journal of Applied Earth Observation and Geoinformation 13:140-151. 10.1016/j.jag.2010.09.007 434
Christman ZJ, and Rogan J. 2012. Error Propagation in Raster Data Integration: Impacts on Landscape Composition
435
and Configuration. Photogrammetric Engineering and Remote Sensing 78:617-624. 10.14358/pers.78.6.617 436
Clifford MJ, Cobb NS, and Buenemann M. 2011. Long-Term Tree Cover Dynamics in a Pinyon-Juniper Woodland:
437
Climate-Change-Type Drought Resets Successional Clock. Ecosystems
438
14:949-962. 439
Coen MC, Andrews E, Bigi A, Martucci G, Romanens G, Vogt FPA, and Vuilleumier L. 2020. Effects of the
440
prewhitening method, the time granularity, and the time segmentation on the Mann-Kendall trend detection
441
and the associated Sen's slope. Atmospheric Measurement Techniques 13:6945-6964. 10.5194/amt-13-6945-2020 442
Cramer WP, and Leemans R. 1993. Assessing Impacts of Climate Change on Vegetation Using Climate Classification
443
Systems. Vegetation Dynamics Global Change. 444
Dash J, Carr AS, and Harris A. 2014. Remote sensing of vegetation cover dynamics and
445
Africa. International journal of applied earth observation geoinformation. 446
De Keersmaecker W, Lhermitte S, Tits L, Honnay O, Somers B, Coppin PJGE, and Biogeography. 2015. A model
447
quantifying global vegetation resistance and resilience to short-term climate anomalies and their
448
relationship with vegetation cover. 24:539-548. p
g
449
Deng ZJ, Guan HD, Hutson J, Forster MA, Wang YQ, and Simmons CT. 2017. A vegetation-focused soil-plant-
450
atmospheric continuum model to study hydrodynamic soil-plant water relations. Water Resources Research
451
53:4965-4983. 10.1002/2017wr020467 452
Fensholt R, Rasmussen K, Nielsen TT, and Mbow C. 2009. Evaluation of earth observation based long term
453
vegetation trends - Intercomparing NDVI time series trend analysis consistency of Sahel from AVHRR
454
GIMMS, Terra MODIS and SPOT VGT data. Remote Sensing of Environment 113:1886-1898. 404
Conclusions It is mainly manifested in the following aspects, the memory effect of
412
vegetation in the southern desert plain was stronger and the plants there are more sensitive to precipitation, while the
413
herbaceous swamp and broad-leaf forest in the central valley showed weaker memory effect and were more
414
sensitive to TVDI. The temperate steppe in the northern mountain is highly sensitive to climate change and were
415
more affected by the combination of both precipitation and temperature. These results will help us locate different
416
ecological protection environment types more accurately in the future basin management process, and develop
417
optimal adaptive ecological protection strategies to protect this vulnerable ecosystem. 419
Acknowledgements g
420
This work was supported by the Xinjiang Tianshan Youth Program (2019Q006)，West Light Foundation of
421
Chinese Academy of Sciences (2019-XBQNXZ-A-001) and Science and Technology Service Network Project of
422
Chinese Academy of Sciences (KFJ-STSQYZD-114). 420
This work was supported by the Xinjiang Tianshan Youth Program (2019Q006)，West Light Foundation of 421
Chinese Academy of Sciences (2019-XBQNXZ-A-001) and Science and Technology Service Network Project of
422
Chinese Academy of Sciences (KFJ STSQYZD 114) 421
Chinese Academy of Sciences (2019-XBQNXZ-A-001) and Science and Technology Service Network Project of 422
Chinese Academy of Sciences (KFJ-STSQYZD-114). 424
References 425
Anav A, and Mariotti A. 2011. Sensitivity of natural vegetation to climate change in the Euro-Mediterranean area. 426
Climate Research 46:277-292. 10.3354/cr00993 425
Anav A, and Mariotti A. 2011. Sensitivity of natural vegetation to climate change in the Euro-Mediterranean area. 426
Climate Research 46:277-292. 10.3354/cr00993 425
Anav A, and Mariotti A. 2011. Sensitivity of natural vegetation to climate change in the Euro-Mediterranean area. 426
Climate Research 46:277-292. 10.3354/cr00993 PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed 455
10.1016/j.rse.2009.04.004 456
Garcia D. 2010. Robust smoothing of gridded data in one and higher dimensions with missing values. Computational
457
Statistics & Data Analysis 54:1167-1178. 10.1016/j.csda.2009.09.020 458
Goward SN, Markham B, Dye DG, Dulaney W, and Yang JL. 1991. NORMALIZED DIFFERENCE VEGETATION
459
INDEX MEASUREMENTS FROM THE ADVANCED VERY HIGH-RESOLUTION RADIOMETER. Remote
460
Sensing of Environment 35:257-277. 10.1016/0034-4257(91)90017-z g f
(
)
461
Grassini P, You J, Hubbard KG, and Cassman KG. 2010. Soil water recharge in a semi-arid temperate climate of the
462
Central U.S. Great Plains. Agricultural Water Management 97:0-1069. 463
Guo B, Zhang J, Meng X, Xu T, and Song Y. 2020. Long-term spatio-temporal precipitation variations in China with
464
precipitation surface interpolated by ANUSPLIN. Scientific Reports 10. 10.1038/s41598-019-57078-3 p
p
p
y
f
p
465
Guo Z, Liang T, and Liu X. 2004. Features of grassland resources and their classified management in Alatai region of
466
Xi ji
Yi
h
t i
b
Th j
l f
li d
l
15 1594 1598 465
Guo Z, Liang T, and Liu X. 2004. Features of grassland resources and their classified management in Alatai region of
466
Xinjiang. Ying yong sheng tai xue bao = The journal of applied ecology 15:1594-1598. 467
H.Y. Z, J. Z, D. L, X.X. Q, and H.J. X. 2020. Characteristics and determining factors for ecological stoichiometry of soil
468
carbon, nitrogen, and phosphorus in different marsh wetlands. ARID ZONE RESEARCH 37:618-626. 469
Han Y, Xuegang C, and Yaqi W. 2013. Dynamic Change of Wetland Landscapes in the Ertix River Basin, Xinjiang
470
during the Period of 1990-2010. ARID ZONE RESEARCH 30:211-218. 471
Hu L, Brunsell NA, Monaghan AJ, Barlage M, and Wilhelmi OV. 2014. How can we use MODIS land surface
472
temperature to validate long-term urban model simulations? Journal of Geophysical Research-Atmospheres
473
119:3185-3201. 10.1002/2013jd021101 j
474
Huang F, Xia Z, Guo L, and Yang F. 2013a. Climate change detection and annual extreme temperature analysis of the
475
Irtysh Basin. Theoretical and Applied Climatology 111:465-470. 10.1007/s00704-012-0676-0 y
pp
gy
476
Huang F, Xia Z, Li F, Guo L, and Yang F. 2012. Hydrological Changes of the Irtysh River and the Possible Causes. 477
Water Resources Management 26:3195-3208. 10.1007/s11269-012-0067-4 478
Huang F, Xia ZQ, Guo LD, and Yang FC. 2013b. Climate change detection and annual extreme temperature analysis
479
of the Irtysh Basin. Theoretical and Applied Climatology 111:465-470. Manuscript to be reviewed f
j
492
Ju B, Ye W, and Hu d. 2015. Variation characteristics and trend of precipitation in Irtysh Ｒiver basin of Xinjiang. 493
Journal of Water Ｒesources ＆ Water Engineering:115-119. 494
Kastens JH, Kastens TL, Kastens DLA, Price KP, Martinko EA, and Lee RY. 2005. Image masking for crop yield
495
forecasting using AVHRR NDVI time series imagery. Remote Sensing of Environment 99:341-356. 496
10.1016/j.rse.2005.09.010 497
Kendall MG. 1990. Rank correlation methods. British Journal of Psychology 25:86–91. 498
Kerr JT, and Ostrovsky M. 2003. From space to species: ecological applications for remote sensing. Trends in Ecology &
499
Evolution 18:299-305. 10.1016/s0169-5347(03)00071-5 498
Kerr JT, and Ostrovsky M. 2003. From space to species: ecological applications for remote s
499
E olutio
18 299 305 10 1016/ 0169 5347(03)00071 5 498
Kerr JT, and Ostrovsky M. 2003. From space to species: ecological applications for
499
Evolution 18:299-305. 10.1016/s0169-5347(03)00071-5 500
Khan B, Hayes LWB, and Cracknell AP. 1995. THE EFFECTS OF HIGHER-ORDER RESAMPLING ON AVHRR
501
DATA. International Journal of Remote Sensing 16:147-163. 10.1080/01431169508954377 502
Lhermitte S, Verbesselt J, Verstraeten WW, Veraverbeke S, and Coppin P. 2010. Assessing intra-annual vegetation
503
regrowth after fire using the pixel based regeneration index %J ISPRS Journal of Photogrammetry and
504
Remote Sensing. 66. g
505
Li D, Wu S, Liu L, Zhang Y, and Li S. 2018. Vulnerability of the global terrestrial ecosystems to climate change. Global
506
Change Biology 24:4095-4106. 10.1111/gcb.14327 507
Li Z, Ma W, Liang C, Liu Z, Wang W, and Wang L. 2015. - Long-term vegetation dynamics driven by climatic
508
variations in the Inner Mongolia grassland: findings from 30-year monitoring. - 30:- 1711. 509
Liu ZY, Notaro M, Kutzbach J, and Liu N. 2006. Assessing global vegetation-climate feedbacks from observations. 510
Journal of Climate 19:787-814. 10.1175/jcli3658.1 511
Los SO, Weedon GP, North PRJ, Kaduk JD, Taylor CM, and Cox PM. 2006. An observation-based estimate of the
512
strength of rainfall-vegetation interactions in the Sahel. Geophysical Research Letters 33. 10.1029/2006gl027065 513
Lu K-Q, Li M, Wang G-H, Xu L-S, Ferguson DK, Trivedi A, Xuan J, Feng Y, Li J-F, Xie G, Yao Y-F, and Wang Y-F. 514
2019. New pollen classification of Chenopodiaceae for exploring and tracing desert vegetation evolution in
515
eastern arid central Asia. Journal of Systematics and Evolution 57:190-199. 10.1111/jse.12462 516
Ludeke MKB, Ramge PH, and Kohlmaier GH. 1996. Manuscript to be reviewed 10.1007/s00704-012-0676-0 478
Huang F, Xia ZQ, Guo LD, and Yang FC. 2013b. Climate change detection and annual extreme temperature analysis
479
of the Irtysh Basin. Theoretical and Applied Climatology 111:465-470. 10.1007/s00704-012-0676-0 PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed 480
Huete, and Alfredo. 2016. Ecology: Vegetation\"s responses to climate variability. Nature. 481
Huete A. 2016. ECOLOGY Vegetation's responses to climate variability. Nature 531:181-182. 10.1038/nature17301 481
Huete A. 2016. ECOLOGY Vegetation's responses to climate variability. Nature 531:181-182. 10.1038/nature17301 481
Huete A. 2016. ECOLOGY Vegetation's responses to climate variability. Nature 531:181-182. 10.1038/nature17301
482 482
J.Z. L, and H.J. M. 2015. Response of vegetation and soil to precipitation in the desert area in the middle reaches of
483
Heihe River Practical Forestry Technology:8-10. 482
J.Z. L, and H.J. M. 2015. Response of vegetation and soil to precipitation in the desert area in the middle reaches of
483
Heihe River Practical Forestry Technology:8-10. 482
J.Z. L, and H.J. M. 2015. Response of vegetation and soil to precipitation in the desert area in the middle reaches of
483
Heihe River Practical Forestry Technology:8-10. 484
Jarvis CH, and Stuart N. 2001. A comparison among strategies for interpolating maximum and minimum daily air
485
temperatures. Part I: the selection of "guiding" topographic and land cover variables. Journal of Applied
486
Meteorology 40:1060-1074. 10.1175/1520-0450(2001)040<1060:Acasfi>2.0.Co;2 484
Jarvis CH, and Stuart N. 2001. A comparison among strategies for interpolating maximum and minimum daily air
485
temperatures. Part I: the selection of "guiding" topographic and land cover variables. Journal of Applied
486
Meteorology 40:1060-1074. 10.1175/1520-0450(2001)040<1060:Acasfi>2.0.Co;2 487
Jeong S-J, Ho C-H, Gim H-J, and Brown ME. 2011. Phenology shifts at start vs. end of growing season in temperate
488
vegetation over the Northern Hemisphere for the period 1982-2008. Global Change Biology 17:2385-2399. 489
10.1111/j.1365-2486.2011.02397.x 487
Jeong S-J, Ho C-H, Gim H-J, and Brown ME. 2011. Phenology shifts at start vs. end of growing season in temperate
488
vegetation over the Northern Hemisphere for the period 1982-2008. Global Change Biology 17:2385-2399. 489
10.1111/j.1365-2486.2011.02397.x j
490
Jiang L, Jiapaer G, Bao A, Guo H, and Ndayisaba F. 2017. Vegetation dynamics and responses to climate change and
491
human activities in Central Asia. Science of the Total Environment 599:967-980. 10.1016/j.scitotenv.2017.05.012 490
Jiang L, Jiapaer G, Bao A, Guo H, and Ndayisaba F. 2017. Vegetation dynamics and responses to climate change and
491
human activities in Central Asia. Science of the Total Environment 599:967-980. 10.1016/j.scitotenv.2017.05.012 f
j
492
Ju B, Ye W, and Hu d. 2015. Variation characteristics and trend of precipitation in Irtysh Ｒiver basin of Xinjiang. 493
Journal of Water Ｒesources ＆ Water Engineering:115-119. Manuscript to be reviewed 533
Oba G, Post E, and Stenseth NC. 2001. Sub-saharan desertification and productivity are linked to hemispheric climate
534
variability. Global Change Biology 7:241-246. 10.1046/j.1365-2486.2001.00405.x 535
Pan Z, Hu Y, and Cao B. 2017. - Construction of smooth daily remote sensing time series data: a higher
536
spatiotemporal resolution perspective. - 2. 537
Pettorelli N, Laurance WF, O'Brien TG, Wegmann M, Nagendra H, and Turner W. 2014. Satellite remote sensing for
538
applied ecologists: opportunities and challenges. Journal of Applied Ecology 51:839-848. 10.1111/1365-
539
2664.12261 540
Piao S, Mohammat A, Fang J, Cai Q, and Feng J. 2006. NDVI-based increase in growth of temperate grasslands and its
541
responses to climate changes in China. Global Environmental Change-Human and Policy Dimensions 16:340-348. 542
10.1016/j.gloenvcha.2006.02.002 j g
543
QiangJi L, and Wu S. 2017. Variation Characteristics of Diurnal Temperature and Influence Factors of Irtysh River in
544
Xinjiang. Journal of Soil and Water Conservation 031:351-356. 545
Rotenberg E, and Yakir D. 2010. Contribution of Semi-Arid Forests to the Climate System. Science 327:451-454. 546
10.1126/science.1179998 547
Safriel U, Adeel Z, Niemeijer D, Puigdefabregas J, White R, Lal R, Winslow M, Ziedler J, Prince S, Archer E, King C,
548
Shapiro B, Wessels K, Nielsen T, Portnov B, Becker-Reshef I, Thonnell J, Lachman E, and McNab D. 2020.-
549
662. 550
Sandholt I, Rasmussen K, and Andersen J. 2002. A simple interpretation of the surface temperature/vegetation index
551
space for assessment of surface moisture status. Remote Sensing of Environment 79:213-224. 10.1016/s0034-
552
4257(01)00274-7 553
Sarkar S, and Kafatos M. 2004. Interannual variability of vegetation over the Indian sub-continent and its relation to
554
the different meteorological parameters. Remote Sensing of Environment 90:268-280. 10.1016/j.rse.2004.01.003 555
Schloss AL, Kicklighter DW, Kaduk J, Wittenberg U, and Participants Potsdam NPPMI. 1999. Comparing global
556
models of terrestrial net primary productivity (NPP): comparison of NPP to climate and the Normalized
557
Difference Vegetation Index (NDVI). Global Change Biology 5:25-34. 10.1046/j.1365-2486.1999.00004.x g
g
gy
j
558
Schwinning S, Sala OE, Loik ME, and Ehleringer JR. 2004. Thresholds, memory, and seasonality: understanding pulse
559
dynamics in arid/semi-arid ecosystems. Oecologia 141:191-193. 10.1007/s00442-004-1683-3 560
Seddon AWR, Macias-Fauria M, Long PR, Benz D, and Willis KJ. 2016. Sensitivity of global terrestrial ecosystems to
561
climate variability. Nature 531:229-+. 10.1038/nature16986 562
Shen Y, Wang G, Su H, Han M, Gao Q, and Wang S. 2007. Manuscript to be reviewed The use of satellite NDVI data for the validation of global
517
vegetation phenology models: Application to the Frankfurt Biosphere Model. Ecological Modelling 91:255-270. 518
10.1016/0304-3800(95)00192-1 519
Mann HB. 1945. Non-Parametric Test Against Trend. Econometrica 13:245-259. 520
Maselli F, Papale D, Chiesi M, Matteucci G, Angeli L, Raschi A, and Seufert G. 2014. Operational monitoring of daily
521
evapotranspiration by the combination of MODIS NDVI and ground meteorological data: Application and
522
evaluation in Central Italy. Remote Sensing of Environment 152:279-290. 10.1016/j.rse.2014.06.021 Mearns LO, Rosenzweig C, and Goldberg R. 1997. Mean and variance change in climate scen
agricultural applications, and measures of uncertainty. Climatic Change 35:367-396. 10.1023/a 525
Mkhabela MS, Bullock P, Raj S, Wang S, and Yang Y. 2011. Crop yield forecasting on the Canadian Prairies using
526
MODIS NDVI data. Agricultural and Forest Meteorology 151:385-393. 10.1016/j.agrformet.2010.11.012 527
Moulin S, Kergoat L, Viovy N, and Dedieu G. 1997. Global-scale assessment of vegetation phenology using
528
NOAA/AVHRR
satellite
measurements. Journal
of
Climate
10:1154-1170. 10.1175/1520-
529
0442(1997)010<1154:Gsaovp>2.0.Co;2 530
Navarro A, Garcia-Ortega E, Merino A, Luis Sanchez J, and Tapiador FJ. 2020. Orographic biases in IMERG
531
precipitation estimates in the Ebro River basin (Spain): The effects of rain gauge density and altitude. 532
Atmospheric Research 244. 10.1016/j.atmosres.2020.105068 PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Hydrological Processes Responding to Climate Warming
563
in the Upper Reaches of Kelan River Basin with Snow-dominated of the Altay Mountains Region, Xinjiang,
564
China. JOURNAL OF GLACIOLOGY AND GEOCRYOLOGY 029:845-854. 565
Simoniello T, Lanfredi M, Liberti M, Coppola R, and Macchiato M. 2008. Estimation of vegetation cover resilience
566
from satellite time series. Hydrology and Earth System Sciences 12:1053-1064. 10.5194/hess-12-1053-2008 567
Slayback DA, Pinzon JE, Los SO, and Tucker CJ. 2003. Northern hemisphere photosynthetic trends 1982-99. Global
568
Change Biology 9:1-15. 10.1046/j.1365-2486.2003.00507.x 567
Slayback DA, Pinzon JE, Los SO, and Tucker CJ. 2003. Northern hemisphere photosynthetic trends 1982-99. Global
568
Change Biology 9:1-15. 10.1046/j.1365-2486.2003.00507.x g
gy
j
569
Sun G, Wang C, Chang L, Wu Y, Li L, and Jin Z. 2018. Effects of feedback regulation on vegetation patterns in semi-
570
arid environments. Applied Mathematical Modelling 61:200-215. 10.1016/j.apm.2018.04.010 g
gy
j
569
Sun G, Wang C, Chang L, Wu Y, Li L, and Jin Z. 2018. Effects of feedback regulation on vegetation patterns in semi-
570
arid environments. Applied Mathematical Modelling 61:200-215. 10.1016/j.apm.2018.04.010 pp
g
j p
571
Thiam AK. 2003. The causes and spatial pattern of land degradation risk in southern Mauritania using multitemporal
572
AVHRR-NDVI imagery and field data. Land Degradation & Development 14:133-142. 10.1002/ldr.533 571
Thiam AK. 2003. The causes and spatial pattern of land degradation risk in southern Mauritania using multitemporal
572 571
Thiam AK. 2003. The causes and spatial pattern of land degradation risk in southern Mauritania using multitem
572
AVHRR NDVI i
d fi ld d
L
d D
d ti
& D
l
t 14 133 142 10 1002/ld 533 AK. 2003. The causes and spatial pattern of land degradation risk in southern Mauritania using mul
AVHRR-NDVI imagery and field data. Land Degradation & Development 14:133-142. 10.1002/ldr.533 573
Thomas CD, Cameron A, Green RE, Bakkenes M, Beaumont LJ, Collingham YC, Erasmus BFN, de Siqueira MF,
574
Grainger A, Hannah L, Hughes L, Huntley B, van Jaarsveld AS, Midgley GF, Miles L, Ortega-Huerta MA,
575
Peterson AT, Phillips OL, and Williams SE. 2004. Extinction risk from climate change. Nature 427:145-148. 576
10.1038/nature02121
577 D. 1996. Biodiversity: Population versus ecosystem stability. Ecology 77:350-363. 10.2307/2265614 577
Tilman D. 1996. Biodiversity: Population versus ecosystem stability. Ecology 77:350-363. 10. QH. 2002. Land use change analysis in the Zhujiang Delta of China using satellite remote sensin
stochastic modelling. Journal of Environmental Management 64:273-284. Manuscript to be reviewed 584
Yang J, and Wang Y. 2011. Estimating evapotranspiration fraction by modeling two-dimensional space of
585
NDVI/albedo and day-night land surface temperature difference: A comparative study. Advances in Water
586
Resources 34:512-518. 10.1016/j.advwatres.2011.01.006 584
Yang J, and Wang Y. 2011. Estimating evapotranspiration fraction by modeling two-dimensional space of
585
NDVI/albedo and day-night land surface temperature difference: A comparative study. Advances in Water
586
Resources 34:512-518 10 1016/j advwatres 2011 01 006 /
y
g
p
586
Resources 34:512-518. 10.1016/j.advwatres.2011.01.006 j
587
Yang Y, Zhao C, Han M, Li Y, and Yang R. 2015. Temporal Patterns of Shrub Vegetation and Variation with
588
Precipitation in Gurbantunggut Desert, Central Asia. Advances in Meteorology 2015. 10.1155/2015/157245 587
Yang Y, Zhao C, Han M, Li Y, and Yang R. 2015. Temporal Patterns of Shrub Vegetation and Variation with
588
Precipitation in Gurbantunggut Desert, Central Asia. Advances in Meteorology 2015. 10.1155/2015/157245 589
Yao C, Zhang Z, and Wang X. 2004. Evaluating Soil Moisture Status in XinJiang Using the Temperature Vegetation
590
Dryness Index (TVDI). REMOTE SENSING TECHNOLOGY AND APPLICATION 19:473-479. 589
Yao C, Zhang Z, and Wang X. 2004. Evaluating Soil Moisture Status in XinJiang Using the Temperature Vegetation
590
Dryness Index (TVDI). REMOTE SENSING TECHNOLOGY AND APPLICATION 19:473-479. 591
Ye G, and Bai Y. 2014. Study On the Ecological Flow in the Middle Reaches of Irtysh Rivers. Third International
592
Conference on Agro-Geoinformatics, 128-131. 591
Ye G, and Bai Y. 2014. Study On the Ecological Flow in the Middle Reaches of Irtysh Rivers. Third International
592
Conference on Agro-Geoinformatics, 128-131. 593
You N, Meng J, and Zhu L. 2018. Sensitivity and resilience of ecosystems to climate variability in the semi-arid to
594
hyper-arid areas of Northern China: a case study in the Heihe River Basin. Ecological Research 33:161-174. 595
10.1007/s11284-017-1543-3 593
You N, Meng J, and Zhu L. 2018. Sensitivity and resilience of ecosystems to climate variability in the semi-arid to
594
hyper-arid areas of Northern China: a case study in the Heihe River Basin. Ecological Research 33:161-174. 595
10.1007/s11284-017-1543-3 596
Zhang X. 2015. Reconstruction of a complete global time series of daily vegetation index trajectory from long-term
597
AVHRR data. Remote Sensing of Environment 156:457-472. 10.1016/j.rse.2014.10.012 596
Zhang X. 2015. Reconstruction of a complete global time series of daily vegetation index trajectory from long-term
597
AVHRR data. Remote Sensing of Environment 156:457-472. Manuscript to be reviewed 10.1006/jema.2001.0509 580
Willis KJ, Jeffers ES, and Tovar C. 2018. What makes a terrestrial ecosystem resilient? A complex set of biotic and
581
abiotic factors determines the resilience of an ecosystem. Science 359:988-989. 10.1126/science.aar5439 580
Willis KJ, Jeffers ES, and Tovar C. 2018. What makes a terrestrial ecosystem resilient? A complex set of biotic and
581
abiotic factors determines the resilience of an ecosystem. Science 359:988-989. 10.1126/science.aar5439 y
582
Yang F, Xia Z, Yu L, and Guo L. 2012. Calculation and Analysis of the Instream Ecological Flow for the Irtysh River. 583
In: Zhou B ed 2012 International Conference on Modern Hydraulic Engineering 438-441 y
582
Yang F, Xia Z, Yu L, and Guo L. 2012. Calculation and Analysis of the Instream Ecological Flow for the Irtysh River. 583
In: Zhou B, ed. 2012 International Conference on Modern Hydraulic Engineering, 438-441. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Manuscript to be reviewed Table 1(on next page) Manuscript to be reviewed 10.1016/j.rse.2014.10.012 g f
j
598
Zhao Y, Zhu J, and Xu Y. 2014. Establishment and assessment of the grid precipitation datasets in China for recent 50
599
years. Journal of the Meteorological Sciences 34:414-420. 598
Zhao Y, Zhu J, and Xu Y. 2014. Establishment and assessment of the grid precipitation datasets in China for recent 50
599
years. Journal of the Meteorological Sciences 34:414-420. 600
Zhu Y, Zhang J, Zhang Y, Qin S, Shao Y, and Gao YJC. 2019. Responses of vegetation to climatic variations in the
601
desert region of northern China. 175:27-36. 600
Zhu Y, Zhang J, Zhang Y, Qin S, Shao Y, and Gao YJC. 2019. Responses of vegetation to climatic variations in the
601
desert region of northern China. 175:27-36. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Table 1(on next page) Interpretation of the coefficients in the AR1 multiple linear regression approach
Interpretation of the coefficients in the AR1 multiple linear regression approach PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Manuscript to be reviewed Manuscript to be reviewed Figure 1 Overview of the Irtysh River basin Overview of the Irtysh River basin 1
Table 1 1
Table 1
2
Interpretation of the coefficients in the AR1 multiple linear regression approach 1
Table 1
2
Interpretation of the coefficients in the AR1 multiple linear regression approach
3
Coefficient
Implication
Meaning of absolute value
Meaning of sign
𝛼
Revealing
the
potential influence of
memory
effects
driving
vegetation
dynamics. A large absolute value indicates low
resilience, which means that vegetation
slowly
recovers
from
previous
disturbance. Positive value of shows NDVI is similar
𝛿
to the previous anomaly. Negative value of shows NDVI is similar
𝛿
to the previous anomaly but with the
opposite trend. 𝛽/ 𝛾/𝛿
Climatic
sensitivity
index denoting the
magnitude
of
immediate
response
of vegetation to the
contemporary
variation in climate
variable. Large
absolute
values
indicate
low
resistance
to
temperature/precipitation/TVDI. Positive
Higher
temperature/precipitation/TVDI
than
average induces a positive NDVI response
(higher NDVI). Negative Lower
temperature/precipitation/TVDI than
average induces a negative NDVI
response (lower NDVI). 2
Interpretation of the coefficients in the AR1 multiple linear regression approach Meaning of sign 3 PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Overview of the Irtysh River basin PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Figure 2 Distribution map of annual average NDVI levels, NDVI change trend Zc value and
change rate β value (A) Distribution map of annual average NDVI levels in the study area from 2000 to 2018 and
Spatial distribution of variance in vegetation cover; (B) NDVI change trend Zc value and (C)
change rate β value spatial distribution map. Characterizing change trend of each vegetated
pixel in the study area. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Manuscript to be reviewed Figure 3 Interannual variation curve of overall NDVI in the study area from 2000 to 2018. Interannual variation curve of overall NDVI in the study area from 2000 to 2018. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Figure 4 Interannual variation curves of overall precipitation, temperature, and TVDI in the study
area from the period 2000-2018. Interannual variation curves of overall precipitation (A), temperature (B), and TVDI (C) in the
study area from the period 2000-2018. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Figure 5 Spatial distribution of t-1 (AR1) coefficient weight Spatial distribution of t-1 (AR1) coefficient weight (that is, coefficient α) from monthly
multiple regression between vegetation cover (defined as NDVI), vegetation cover at t-1, and
three climatic variables. Characterizing the memory effects of vegetation cover in the Irtysh
River basin during 2000-2018. Manuscript to be reviewed PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed Manuscript to be reviewed Figure 6 Scatter plots of α (t-1 coefficient weight) along mean growing season climatic factors Scatter plots of α (t-1 coefficient weight) along mean growing season climatic factors
Correlations between α (t-1 coefficient weight) and mean growing season (A)NDVI,
(B)precipitation, (C)temperature, (D)TVDI in the Irtysh River basin during 2000-2018. The Scatter plots of α (t-1 coefficient weight) along mean growing season climatic factors Correlations between α (t-1 coefficient weight) and mean growing season (A)NDVI,
(B)precipitation, (C)temperature, (D)TVDI in the Irtysh River basin during 2000-2018. The
green curves indicate the fitted regression lines. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Figure 7 Distribution of Vegetation sensitivity index (VSI) in Irtysh River basin during 2000-2018 Figure 8 RGB composite of climate weights Manuscript to be reviewed Distribution of Vegetation sensitivity index (VSI) in Irtysh River basin during 2000-2018 (A) Spatial distribution of Vegetation sensitivity index (VSI) in Irtysh River basin during
2000-2018. The index ranges from 0(low sensitivity) to 100(high sensitivity). (B) VSI
distribution histogram, insert panel of violin plot shows the frequency distribution of pixel VSI
values. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) Manuscript to be reviewed PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) RGB composite of climate weights RGB composite of climate weights from monthly multiple regression between vegetation
cover (defined as NDVI), vegetation cover at t-1, and three climatic variables. Notably,
temperature, red; TVDI, green; and precipitation, blue. PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021) PeerJ reviewing PDF | (2020:11:55104:1:2:NEW 18 Mar 2021)
|
https://openalex.org/W4321603482
|
https://www.shs-conferences.org/articles/shsconf/pdf/2023/08/shsconf_iclcc2023_02009.pdf
|
English
| null |
The Influence of Perceived Value of Online Game Users on their Participation in Value Co-creation Behavior
|
SHS web of conferences
| 2,023
|
cc-by
| 8,046
|
The Influence of Perceived Value of Online Game Users on their
Participation in Value Co-creation Behavior Haowen Xu 1,*
1 Social Media Communication, Shenzhen University, Shenzhen, 518000, China Haowen Xu 1,* 1 Social Media Communication, Shenzhen University, Shenzhen, 518000, China Abstract: Online games have gradually become a form of entertainment and a pastime for many people. In
recent years, the growth rate of China’s online game industry has slowed, game users have entered the
storage area, and the game industry has entered a stable development stage. More and more game
enterprises need to rely on multiple parties to participate in value co-creation to integrate and build game
brand value and gain advantages in the industry competition. Based on Perceived Value theory and Value
Co-creation theory, this paper constructs a theoretical model of online game users’ Value Co-creation
Behavior, analyzes the mediating role of Brand Relationship Quality between Perceived Value and Value
Co-creation Behavior, and the moderating role of Game Literacy. The study found that Self Fulfillment and
Social-emotional Development had significant effects on Autonomous Value Co-creation Behavior and
Sponsored Value Co-creation Behavior. In contrast, Character Experience and Recreational Release
significantly affected Autonomous Value Co-creation Behavior and Sponsored Value Co-creation Behavior. Brand Relationship Quality mediates between Perceived Value and Value Co-creation Behavior; Game
Literacy mediates between Perceived Value and Value Co-creation Behavior. The moderating effect of
Game Literacy in the relationship between Perceived Value and Value Co-creation Behavior is not
significant. The above study results have important reference significance for the development of online
games. more and more game enterprises need to rely on multiple
parties to participate in value co-creation to integrate and
build game brand value. In this phenomenon, the main
position of game users is very obvious. They actively
participate in the game development and innovation
process and are willing to share and spread their game
experience,
gradually
transforming
from
passive
receivers to passive recipients to active participants,
disseminators, and creators [3]. What is the catalyst for
user participation in game brand value co-creation? There have been many academic discussions on this
issue, such as taking online game virtual communities as
the research object and exploring the path of consumers’
participation in brand value co-creation in communities
from the perspective of social networks [4]; or taking
game brands [5] or game digital distribution platforms [6]
as the core to study the mechanism of game value co-
creation. The Influence of Perceived Value of Online Game Users on their
Participation in Value Co-creation Behavior However, in previous studies, the influence of
users’ perceptions and attitudes towards the game
product itself on value co-creation has been neglected. Such as the users’ feelings about playing, the audio and
visual stimuli they feel in the game, and the related links
they have with the game brand after playing deeply. Does this paper pose the question: can game companies
increase the quality of the relationship between users and
game brands by improving the quality of their perception
of the game product itself? Can such emotional attitudes * Corresponding author: 21035315@saultcollege.ca
© The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0
(http://creativecommons.org/licenses/by/4.0/). SHS Web of Conferences 159, 02009 (2023) SHS Web of Conferences 159, 02009 (2023) SHS Web of Conferences 159, 02009 (2023)
ICLCC 2023 https://doi.org/10.1051/shsconf/202315902009 EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0
icenses/by/4.0/). 1. Introduction Benefiting from technological advances and the
expansion of the size of Internet users, China’s online
game industry has maintained a high growth rate every
year in the past few years. However, compared to the
37.7% year-on-year growth in 2014 and 22.9% year-on-
year growth in 2015, the market size of the online game
industry only grew by 7.7% in 2019 [1]. With the
disappearance of the demographic dividend, Chinese
online game users have entered the stock interval, and
the game industry has entered a stable development stage. The epidemic has provided new opportunities for the
development of online games. In the first quarter of 2020,
leading Chinese game companies such as Tencent,
Netease, and San7 Interactive Entertainment all achieved
growth in online game revenue, with an average growth
rate of 30% [2]. During this period, new features also
emerged in the online game industry: demand for games
increased, and user behaviour was mainly driven by
social. And after three years of the epidemic, players’
overall enthusiasm for entertainment returned to normal,
with gaming habits still present but with a higher
demand for game quality. In this period, the competition
of game platforms will also return to the refined
operation of game IP. In order to retain users, create high-quality game
products and gain advantages in the industry competition, SHS Web of Conferences 159, 02009 (2023)
ICLCC 2023 SHS Web of Conferences 159, 02009 (2023) https://doi.org/10.1051/shsconf/202315902009 Character Experience Development, and Recreational
Release into four dimensions [9]. Ouyang Banghong and
Liu Chen et al. divided the Perceived Value of game
players into four dimensions: fun value, character ability
value, visual management value, and monetary value
[10]. Based on Yu Ping and Sang Wentian’s view, this
paper divides consumers’ Perceived Value of online
games into four dimensions Self Fulfillment, Social-
emotional
Development,
Character
Experience
Development, and Recreational Release; the study was
conducted. further promote users’ participation in the Value Co-
creation Behaviour? Character Experience Development, and Recreational
Release into four dimensions [9]. Ouyang Banghong and
Liu Chen et al. divided the Perceived Value of game
players into four dimensions: fun value, character ability
value, visual management value, and monetary value
[10]. Based on Yu Ping and Sang Wentian’s view, this
paper divides consumers’ Perceived Value of online
games into four dimensions Self Fulfillment, Social-
emotional
Development,
Character
Experience
Development, and Recreational Release; the study was
conducted. 1. Introduction Based on this, this paper combines Perceived Value
Theory and Value Co-creation Theory to explore the
influence mechanism and path of users’ perceived
quality on Value Co-creation Behavior from the
perspective of users’ perception of online game products. Through literature collation and empirical research, we
hope to achieve the following objectives: from the
perspective of users’ Perceived Value, clarifying the
process of Value Co-creation Behavior can enable game
brands to recognize the influencing factors behind users’
participation in Value Co-creation Behavior, and better
find the focus points for game products to improve
quality and stimulate users’ value perception, which is
important for This is important for game companies to
retain users, build a good brand relationship and promote
better user participation in brand value co-creation,
which can create more business value for the companies. At the same time, in the context of value co-creation
theory, game users are not only passive recipients of the
brand’s value output but can also create value together
with the company. Hence, value co-creation is a win-win
initiative for both users and brands. This paper’s results
can provide some suggestions to game enterprises
regarding game production upgrading and game IP
refinement operation, which is also conducive to
improving users’ gaming experience. In addition, this
paper constructs a theoretical model of the influence of
game users’ Perceived Value on Value Co-creation
Behavior, explores the influence of different dimensions
of users’ Perceived Value of message products on their
participation in Value Co-creation Behavior, and argues
for the mediating role of Brand Relationship Quality in it. The study’s results will also be an important element of
value co-creation theory in online games. In the most research literature, Perceived Value has
been studied as an independent or mediating variable. Many scholars have demonstrated that Perceived Value
positively impacts customers’ emotions and behaviors. Zhang Qiyao and Zheng Aicheng used night market
customers as the research object. They introduced
customer emotion as a mediating variable to argue that
Perceived Value has a significant positive impact on the
willingness to consume in night markets [11]. Users
perceive
Self
Fulfillment,
Social-emotional
Development, Character Experience Development, and
Recreational Release values in online games as
important factors in brand relationship building, and in
the process of deeper game user experience Brand
Relationship Quality gets a boost. 2.2. Brand Relationship Quality The primary purpose of the game is to gain a gaming
experience. As the player’s gameplay deepens, i.e., the
interaction between the consumer and the brand deepens,
further emotional connections are made to the game
brand, and a brand relationship is formed. As for game
brands, since the content of this study is conducted with
the online game players’ perception of the game product
itself as the core, the game brands in the study refer to
the brands of the game products and services themselves
rather than the brands of the developers, operators or
publishers of the games. For example, ‘Food Language’
is a game brand, and the developer of ‘Food Language’,
Guangzhou Tianti Network Technology Company
Limited, and the publisher, Tencent Games, are not
game brands. 1. Introduction Based on their love for
the game brand, customers are willing to participate in
activities initiated by the game brand and share the game
experience
with
others,
maintaining
the
brand
relationship over time and creating value together. In this
paper, we explore the effect of different dimensions of
Perceived Value on Brand Relationship Quality and
different types of Value Co-creation Behaviour. 2.1. Perceived Value The idea of Customer Perceived Value can be traced
back to 1954. Drucker proposed in Management Practice
that what customers buy and consume is not a product
but a value [7]. With the expansion of the concept of
customer value in the marketing field, its connotation
and dimensional theoretical models have also been
developing. There are commonly four-dimensional and
five-dimensional models, etc. Perceived value is the
subjective feeling of customers towards the product,
which is obviously influenced by subjective factors. Different customers perceive the value of the same
commodity differently. And for online games, due to the
unique virtual nature and Park and Lee focus on virtual
goods in online games and propose four dimensions of
measurement: fun value, character ability value, visual
dominance value, and monetary value [8]. Based on the
characteristics
of
customer
value
hierarchy
and
experience economy, Yu, Ping, and Sang classified the
Perceived Value of online games into four dimensions:
Self
Fulfillment,
Social-emotional
Development, Blackston was the first to propose the concept of
brand relationships, pointing out that the quality of
interaction between consumers and the brand mainly
influences brand relationships. Through many brand
analyses, he concluded that successful interaction
consists of customer trust and brand satisfaction [12]. Fournier first identified Brand Relationship Quality as a
measure of brand relationships and proposed that Brand
Relationship Quality consists of six components: love
and passion, self-connection, interdependence, personal 2 2 SHS Web of Conferences 159, 02009 (2023) SHS Web of Conferences 159, 02009 (2023)
ICLCC 2023 https://doi.org/10.1051/shsconf/202315902009 commitment, intimacy, and the companionship quality of
the brand [13]. Through empirical research, Zhou
Zhimin and Lu Taihong found that Brand Relationship
Quality
can
be
divided
into
five
dimensions:
commitment relevance, attribution concern, familiarity
and understanding, trust and respect, and association re-
identification [14]. Zhao Jinglin and Zhao Hong and
Zhang Tuixi have pointed out that although there is no
consensus among academics on the Brand Relationship
Quality dimensions [15,16]. Most studies include the
three dimensions of brand satisfaction, brand trust, and
brand commitment, with brand trust and commitment
being a deeper level of brand relationship than customer
satisfaction. Trust implies that consumers have reliability
and honesty in the brand Trust means that consumers
have confidence in the reliability and honesty of the
brand. In contrast, commitment means consumers are
willing to maintain a long-term relationship with the
brand. 2.4. Value Co-creation Prahalad and Ramaswany were the first to propose the
concept of value co-creation, suggesting that firms can
co-create value with consumers through interactive
behaviours, which can help firms gain an edge over their
competitors [20]. Wu and Chen divide value co-creation
theory into value co-creation based on producer logic
and value co-creation based on consumer logic. This
refers to firms investing and integrating consumer
resources to improve their value output, and consumers
are integrating resources to create value or solve
problems for themselves, respectively [21]. Pinho,
Beirao et al. point out that the mobile Internet’s
development has made value co-creation more diverse
and complex, with consumers, companies, suppliers, and
collaborators all being able to participate in value co-
creation [22]. Ultimately, value co-creation is an
interactive
and
cooperative
process
in
which
stakeholders integrate multiple resources to create value
for multiple parties. Specifically, in the game field, He
Yuanshuo believes that value co-creation theory has
strong applicability in the game field. Multiple resources
in the upstream and downstream of the game industry
chain can achieve resource integration through the
interaction of multiple actors, in which game users play a
very important role [5]. Empirical studies have demonstrated that consumer
perceptions
significantly
positively
impact
Brand
Relationship Quality and that Brand Relationship
Quality positively impacts customer behaviour. In a
study on tourism brands, Zhang Hui and Chen Ye
demonstrated that brand fit significantly impacts all
dimensions of Brand Relationship Quality [17]. This
paper combines the characteristics of online games. It
explores the influence of the Perceived Value of game
users on Brand Relationship Quality and the effect of
Brand Relationship Quality on Value Co-creation
Behavior from the user’s game experience. 2.1. Perceived Value This paper uses brand satisfaction, trust, and
commitment as dimensions of Brand Relationship
Quality to test the subsequent research hypotheses. play the game successfully and their experience and
ability to co-create value. Suppose users have a low level
of knowledge and experience in games. In that case, they
can easily become frustrated in games, resulting in their
inability to perceive game content properly, develop a
positive feeling towards game brands, and communicate
about game products or suggest corrections. 3.1. The relationship between Perceived Value
and Value Co-creation Behaviour Perception is the basis of all mental activity. Numerous
studies have proven that Perceived Value is an important
influencing factor for consumers’ participation in Value
Co-creation Behavior. Tang Jing and Xu Yahan [29]
have demonstrated that the perceived economic impact
of tourism and the perceived environmental impact of
tourism significantly and positively influence the Value
Co-creation Behavior of tourism destination residents, as
well as the mediating role of subjective well-being. For
online games, when game players form subjective
evaluations of various aspects of the game after
understanding the game content and experiencing the
game fun in the game process, good brand perceptions
can improve customer satisfaction at a certain level,
which in turn promotes their participation in brand value
co-creation activities. Based on this, this paper proposes the following
hypothesis: Based on this, this paper proposes the following
hypothesis: H2a: Self Fulfillment positively influences Brand
Relationship Quality. H2b:
Social-emotional
Development
positively
influences Brand Relationship Quality. H2c: Character Experience Development positively
influences Brand Relationship Quality. H2d: Recreational Release positively influences
Brand Relationship Quality. 3. RESEARCH MODEL AND RESEARCH
HYPOTHESIS H1h: Recreational Release development positively
influences Sponsored Value Co-creation Behaviour. Based on the “Awareness-Attitude-Behaviour” theory,
this study takes the Perceived Value of game users in the
game process as the basis of awareness and through the
Brand Relationship Quality. This represents the attitude,
and under the moderating effect of Game Literacy,
affects the participation of game users in the game brand. The concept model of the relationship between
Perceived Value, Brand Relationship Quality, and Value
Co-creation Behavior of game users is thus constructed,
as shown in Figure 1. 3.2. The relationship between Perceived Value
and Brand Relationship Quality Customer Perceived Value is an important antecedent for
customers to build goodwill toward a brand. A study by
Yang Shuang and Zhou Xing on SNS users concluded
that
multiple
Perceived
Values
of
SNS
users
significantly affect relationship quality dimensions such
as satisfaction, trust, and commitment [30,31]. Zhenquan
Sha and Meixian He et al. empirically demonstrated that
customer Perceived Value and positive customer
sentiment significantly impacted Brand Relationship
Quality [32]. When online game users play games, they
gain multi-dimensional perceptual experiences in the
virtual world of the Internet, including various aspects of
the game, such as story content, audio, visual stimulation,
and friendship fun. In this process, game users can feel
the cultural connotation of the game brand from the
worldview built by the game and feel the production
technology of the game brand from the production of the
game graphics, etc. Good Perceived Value can trigger
game users’ good feelings towards the game brand, thus
establishing a good brand relationship. However, no
research has yet provided a detailed answer to how each
perceived dimension of online games affects Brand
Relationship Quality. Fig.1. Concept model (Photo credit: Original) Fig.1. Concept model (Photo credit: Original) 2.3. Game Literacy Game Literacy may affect their ability to 3 3 SHS Web of Conferences 159, 02009 (2023) SHS Web of Conferences 159, 02009 (2023)
ICLCC 2023 https://doi.org/10.1051/shsconf/202315902009 positive brand relationships and gives game users
positive attitudes toward game brands. Based on this
good feeling, users are willing to participate in the value
co-creation of game brands. In this process, users’ good
or bad Game Literacy may affect their Value Co-creation
Behavior, and users’ Game Literacy may affect their
Value Co-creation Behaviour. Based on this, this study
constructs a model based on the “awareness-attitude-
behavior” relationship theory. The relationship between
Perceived Value, Brand Relationship Quality, Game
Literacy, and Value Co-creation Behavior of online
game users are studied. H1a:
Self
Fulfillment
positively
influences
Autonomous Value Co-creation Behaviour. H1a:
Self
Fulfillment
positively
influences
Autonomous Value Co-creation Behaviour. H1b:
Social-emotional
Development
positively
influences Autonomous Value Co-creation Behaviour. H1c: Character Experience Development positively
influences Autonomous Value Co-creation Behaviour. H1d: Recreational Release development positively
influences Autonomous Value Co-creation Behaviour. H1e:
Self
Fulfillment
positively
influences
Sponsored Value Co-creation Behaviour. H1f:
Social-emotional
Development
positively
influences Sponsored Value Co-creation Behaviour. H1g: Character experience development positively
influences Sponsored Value Co-creation Behaviour. 2.3. Game Literacy Scholars have also researched the dimensional
division of Value Co-creation Behavior. Most scholars
divide it into Customer Engagement Behaviour and
Customer Citizenship Behaviour (Yi and Gong [23],
Chen Ying and Ren Lejing et al. [24]) based on previous
studies, which further explain Customer Engagement
Behaviour
and
Citizenship
Behaviour
from
the
perspective of Value Co-creation. Zwass and Chaohui Li
and Yongsheng Jin et al. argue that Value Co-creation
Behavior can be divided into two types [25,26]. Namely,
Sponsored Value Co-creation Behavior and Autonomous
Value Co-creation Behavior, depending on the initiator. Ranjan and Read distinguish three types of value-
creation behavior: co-production, use of value, and
knowledge sharing [27]. Based on Zwass’ view, this
study distinguishes the Value Co-creation Behavior of
game users according to the subject. It divides it into
corporate Sponsored Value Co-creation Behavior and
consumer Autonomous Value Co-creation Behavior. Game Literacy is a broad concept that refers to the skills
of knowing, using, experiencing, and effectively
constructing
games
correctly
and
appropriately. According to Zhang Qianwei, Game Literacy consists of
five aspects: game knowledge, game rules, game skills,
game motivation, and game emotion [18]. Gou Chaoqun
pointed out that Game Literacy mainly consists of three
parts: game emotion, game rules, game awareness, and
game skills and creative ability [19]. Although scholars
have different dimensions of Game Literacy, it can be
concluded that Game Literacy combines players’ game
emotions, consciousness, and skills. This requires
players to have a good and healthy state of mind to feel
the fun of the game, to be able to appreciate and
comprehend the music, plot, and graphics, and to
understand and abide by the game’s rules. In addition,
players need to have certain operating skills and
cognitive levels, reflected in the operation level of
players in competitive games or the ability to make plans
and strategies in strategic games. Players with certain
gaming skills can think creatively and critically about the
game, recreate it, or propose changes. Empirical studies have shown that consumers’
characteristics and Perceived Value influence Value Co-
creation
Behavior. Wei
Qinggangdemonstrated
a
positive
correlation
between
consumer-perceived
fairness and perceived risk and consumer participation in
Value Co-creation Behavior[28]. This paper notes that
consumers’ Perceived Value of online games helps build This paper focuses on the impact of users’ Perceived
Value on their participation in Value Co-creation
Behaviour. 4.1. Questionnaire design This study used the questionnaire method to conduct an
empirical study. The questionnaire design was divided
into two parts: the first part consisted of basic
information about the respondents, including gender, age,
education, and the frequency of participation in online
games; the second part measured the variables involved
in the model. The variables were measured on a five-
point Richter scale, ranging from “very unlikely” to
“very likely”. The variables used were based on scales
published in refereed journals and adapted to the purpose
of the study to form the initial questionnaire. To ensure
that the questionnaire was more appropriate to the
context of the respondents, the English and Chinese
scales were repeatedly revised. Then 10 in-depth online
game players were invited to fill in the questionnaire and
propose changes to form the final scale. Specifically: (1) In terms of game user perceptions,
this study adopted Yu Ping and Sang Wentian’s [9] scale,
which divides consumers’ Perceived Value of online
games
into
Self
Fulfillment,
Social-emotional
Development, Character Experience Development, and
(2) In terms of Brand Relationship Quality, this study
draws on the scales of Hu Yinhua and He Yan [35] and
Xiong Aihua and Chen Xiaoyun [34] to measure brand
satisfaction, brand trust and brand commitment as the
dimensions of Brand Relationship Quality; (3) In terms
of game In terms of Literacy, this study refers to Gou
Chaoqun’s [19] study and Shi Rui’s and Han Donglin’s
[36] scales to measure Game Literacy in three
dimensions: Game Emotion, Game Awareness, and
Game Skills; (4) Value Co-creation Behavior, this study
refers to Swass’ [25] and Jiang Ying’s [37], the Value
Co-creation Behavior of game users was divided into
corporate Sponsored Value Co-creation Behavior and
game users’ Autonomous Value Co-creation Behavior
for measurement. Based on this, this paper proposes the following
hypothesis: H3a:
Brand
Relationship
Quality
positively
influences Autonomous Value Co-creation Behavior
behaviour. H3a:
Brand
Relationship
Quality
positively
influences Autonomous Value Co-creation Behavior
behaviour. H3b:
Brand
Relationship
Quality
positively
influences Sponsored Value Co-creation Behavior. H3b:
Brand
Relationship
Quality
positively
influences Sponsored Value Co-creation Behavior. H3b:
Brand
Relationship
Quality
positively
influences Sponsored Value Co-creation Behavior. 3.4. The Mediating Role of Brand Relationship
Quality In their study on virtual brand communities, Xiong
Aihua and Chen Xiaoyun confirmed that Brand
Relationship Quality partially mediates the influence of
Perceived Value on Value Co-creation Behavior [34]. Based on Hypothesis 2 and 3, this paper proposes the
following hypothesis to test this mediating role. H4: Brand Relationship Quality mediates the
relationship between Perceived Value and Value Co-
creation Behavior. H4: Brand Relationship Quality mediates the
relationship between Perceived Value and Value Co-
creation Behavior. 4.2. Data collection This study took online game users as the research object
and
distributed
questionnaires
through
the
“Questionnaire
Star”
platform
and
diffused
questionnaires through various online game communities. A total of 305 questionnaires were collected online, and
197 were valid after excluding invalid questionnaires
that were not online game players, had too short a
response time, and had too many common options—the
sample description for analysis through SPSS23.0
software. From the sample structure, 140 respondents
had participated in online games for more than two years,
accounting for 72.2% of the survey population,
indicating that most of the sample were deep online
game
players. Their
perceived
experiences
and
behaviours better reflected the overall characteristics of
online game players, and the overall representativeness
of the sample was good. 3.3.
The
relationship
between
Brand
Relationship Quality and Value Co-creation
Behaviour Numerous studies have shown that the brand relationship
is a two-way loyalty between customers and the brand
and that consumers’ post-purchase behaviour results Based on this, this paper proposes the following
hypothesis: 4 4 SHS Web of Conferences 159, 02009 (2023) SHS Web of Conferences 159, 02009 (2023)
ICLCC 2023 https://doi.org/10.1051/shsconf/202315902009 from high levels of Brand Relationship Quality. Zhang
Xinsheng and Li Xianguo empirically demonstrated that
Brand
Relationship
Quality,
which
focuses
on
satisfaction and trust, has a positive impact on
consumers’ willingness to participate in value co-
creation and that when consumers have a higher level of
satisfaction and trust in a brand’s virtual community. In
addition, they are more likely to favour the brand and
thus develop positive post-purchase behaviors such as
participating in the brand’s product improvement
activities [33]. Playing a game is also a process of deeper
communication between the user and the game brand,
and as the game progresses, the user builds a positive
brand relationship with the game brand. According to
reciprocity theory, in return for a quality brand
relationship, game users will also reciprocate to game
users, as shown by their willingness to give feedback to
the game brand on their own gaming experience and
suggestions for improvement or share their gaming
experience with friends and other Value Co-creation
Behaviour, expecting their behaviour to bring positive
effects to the game brand. 5.1. Reliability and validity tests This study used SPSS 23.0 and AMOS 21.0 software to
test the reliability and validity of the scale. The
Cronbach’s alpha coefficient was used to analyze the
questionnaire data, and the results showed that
Cronbach’s alpha value of each scale variable was
greater than 0.75 (Table 1), indicating that the question
items of this questionnaire were reliable. The reliability Table 1. Results of confirmatory factor analysis
Construct
Survey Items
Std. Estimate
Self Fulfillment
Cronbach’s Alpha=0.848
AVE=0.593
CR=0.853
SF1
I can achieve self-fulfillment by accomplishing
various goals in the game. 0.823
SF2
My efforts could be easily recognized by friends
around me in the game. 0.842
SF3
I can experience the thrill of competing with others
through my game characters. 0.705
SF4
I use the game as a resource to improve my
strengths and hobbies. 0.699
Social-emotional
Development
Cronbach’s Alpha=0.892
AVE=0.623
CR=0.892
SED1
I enjoy interacting with other players and having
fun in the game. 0.787
SED2
I can meet many friends in the game and they bring
me happiness. 0.831
SED3
I love being involved in different types of activities
in the game. 0.783
SED4
I love teamwork and fighting with my partners in
the game. 0.731
SED5
Friends can give me care, assistance and comfort in
the game. 0.81
Character Experience
Development
Cronbach’s Alpha=0.874
AVE=0.649
CR=0.88
CED1
I can experience a different life from the real one
through my role in the game. 0.84
CED2
I run the characters in the game and experience the
satisfaction of character growth. 0.785
CED3
I always discover new things and enjoy trying them
out through the game. 0.884
CED4
Through the game, I can improve my intelligence,
emotion and courage. 0.703
Recreational Release
Cronbach’s Alpha=0.870
AVE=0.626
CR=0.869
RR1
I think playing online games is a form of
entertainment. 0.723
RR2
Through the game, I can pass the boring time and
make my life more fulfilling. 0.855
RR3
Through the game, I can release my stress and
frustration. 0.759
RR4
Through the game, I can feel the excitement and
the pleasure. 0.82
Brand Relationship Quality
Cronbach’s Alpha=0.885
AVE=0.667
CR=0.889
BRQ1
I am satisfied with the service provided by the
online game. 0.815
BRQ2
The actual performance of the online game meets
my expectations. 0.833
BRQ3
I am proud to be a member of the online game
players. 0.881
BRQ4
I would like to continue playing online games. 3.5. The Moderating Role of Game Literacy Game Literacy is an important factor influencing how
players experience and perceive game content. Players
with better Game Literacy can quickly acquire game
skills, have a sharper perception of game content, and
share more feelings and experiences with other players. This provides more effective suggestions for game
improvement and even can recreate game content. Based on this, this paper proposes the following
hypothesis: H5: Game Literacy positively moderates the
relationship between online game users’ Perceived Value
and their participation in Value Co-creation Behavior. 5 5 https://doi.org/10.1051/shsconf/202315902009 SHS Web of Conferences 159, 02009 (2023) SHS Web of Conferences 159, 02009 (2023)
ICLCC 2023 ICLCC 2023 5. DATA ANALYSIS AND HYPOTHESIS
TESTS of the scale was good. The results of the KMO and
Bartlett’s spherical tests conducted on the scale showed
that the KMO value for the overall sample of this study
was 0.931, with a p-value infinitely close to 0 (Table 2),
which reached a significant level. The KMO value was
greater than 0.6, indicating that this study was suitable
for factor analysis. The validation factor analysis of the
seven variables involved in the study yielded Std. Estimate greater than 0.65 for each variable of the scale,
indicating good aggregation of the overall model. Ave
values were greater than 0.5, and combined reliability
CR was greater than 0.75, indicating that the scale has
good internal consistency and convergent validity. 5.2. Correlation test This study conducted Pearson correlation analysis on
each dimension of Perceived Value, Brand Relationship
Quality, and Value Co-creation Behavior of online game
users using SPSS23.0 software. As shown in Table 3, the
four dimensions of the Perceived Value of online game
users were significantly and positively correlated with
Brand
Relationship
Quality
(p<0.01),
the
four Table 3. Correlations
SF
SED
CED
RR
BRQ
GL
AVCB
SVCB
SF
1
SED
.582**
1
CED
.644**
.624**
1
RR
.516**
.447**
.661**
1
BRQ
.569**
.473**
.620**
.641**
1
GL
.572**
.551**
.628**
.658**
.685**
1
AVCB
.515**
.539**
.523**
.434**
.655**
.591**
1
SVCB
.417**
.466**
.334**
.208**
.527**
.429**
.773**
1
** Correlation is significant at the 0.01 level (2-tailed). hypotheses H1b, H1d, H1g, and H1h were invalid. Significance and therefore hypothesis H2b does not hold. The rest of the paths were significant at the 0.05 level. Self Fulfillment (p=0.012<0.05; p=0.004<0.05) and
Social-emotional Development (p=0<0.05; p=0<0.05)
significantly affected Autonomous Value Co-creation
Behavior and Sponsored Value Co-creation Behavior,
hypotheses H1a, H1b, H1e, and H1f hold; Self
Fulfillment
(p=0.002<0.05),
Character
Experience
Development (p=0.015<0.05), and Recreational Release
(p=0.0< 0.05) significantly affect Brand Relationship
Quality, hypotheses H2a, H2c and H2d hold; Brand
Relationship
Quality
(p=0.0<0.05;
p=0.0<0.05)
significantly affects Autonomous Value Co-creation
Behavior and Sponsored Value Co-creation Behavior,
hypotheses H3a and H3b hold. 5.1. Reliability and validity tests 0.732
Game Literacy
Cronbach’s Alpha=0.751
AVE=0.521
CR=0.764
GL1
I often stay positive in the game. 0.77
GL2
I can understand and follow the rules of the game. 0.737
GL3
I have good handling skills during the game. 0.654
Autonomous
Value Co-creation Behavior
AVCB1
I often take the initiative to share and exchange my
game experience with others. 0.759 6 6 https://doi.org/10.1051/shsconf/202315902009 SHS Web of Conferences 159, 02009 (2023)
ICLCC 2023 Cronbach’s Alpha=0.880
AVE=0.65
CR=0.881
AVCB2
I often post about gaming topics on the Internet. 0.81
AVCB3
I actively like and comment on internet topics
about online games. 0.85
AVCB4
I am willing to help others with problems related to
online games. 0.803
Sponsored Value Co-creation
Behavior
Cronbach’s Alpha=0.917
AVE=0.745
CR=0.921
SVCB1
I often volunteer to fill in feedback questionnaires
published by game brands. 0.821
SVCB2
I am actively involved in the creative competition
organized by game brands. 0.949
SVCB3
I actively participate in game promotion activities,
such as participating in official online prize draw
activities or offline events. 0.887
SVCB4
I often watch the official tournaments of the game
0.786 Table 2. KMO and Bartlett’s Test
Kaiser-Meyer-Olkin Measure of Sampling Adequacy
0.931
Bartlett’s Test of Sphericity
Approx. Chi-Square
4848.888
df
496
Sig. .000 dimensions of Perceived Value of online game users
were significantly and positively correlated with two
dimensions of Value Co-creation Behavior (p<0.01), and
the four dimensions of Brand Relationship Quality was
significantly
and
positively
correlated
with
two
dimensions of Value Co-creation Behavior (p<0.01). The
correlations of the above variables are in line with the
research model and provide a basis for further analysis in
this study. 5.3.1. Main effects analysis From
the table above, it can be seen that the interaction term
between Perceived Value and Game Literacy does not
show significance (t=-0.037, p=0.970>0.05), as well as
from model 1, X produces an impact relationship for Y,
implying that Perceived Value has an impact on Value
Co-creation Behavior when moderating variable Game
Literacy remains consistent across levels, suggesting that
the moderating effect of Game Literacy is not significant. Hypothesis H5 does not hold online game participation per day; Model 2 adds the
moderator Game Literacy to Model 1, and Model 3 adds
the interaction term (the product of the independent and
moderator variables) to Model 2. Model 1 is to
investigate the effect of the independent variable,
Perceived Value, on the dependent variable, Value Co-
creation Behavior, when the moderating variable, Game
Literacy, is not taken into account. As seen from Table 6,
the independent variable Perceived Value showed
significance (t=8.401, p=0.000<0.05), which means that
Perceived Value has a significant relationship with
Value Co-creation Behavior. The moderating effect can
be viewed in two ways and the first is to view the
significance of the change in F-value when going from
model 2 to model 3; the second is to view the
significance of the interaction term in model 3, this time,
the moderating effect is analyzed in a second way. From
the table above, it can be seen that the interaction term
between Perceived Value and Game Literacy does not
show significance (t=-0.037, p=0.970>0.05), as well as
from model 1, X produces an impact relationship for Y,
implying that Perceived Value has an impact on Value
Co-creation Behavior when moderating variable Game
Literacy remains consistent across levels, suggesting that
the moderating effect of Game Literacy is not significant. Hypothesis H5 does not hold 5.3.1. Main effects analysis This study used AMOS 21.0 software to construct a
structural equation model to test the main effects. The
structural equation model CMIN/DF == 2.105, RMSEA
= 0.071, GFI = 0.799, CFI, NFI, IFI are all greater than
0.9, and the model fit is good. Analysis of the test data
for each pathway (see Table 4) revealed that Character
Experience Development (p=0.112>0.05; p=0.933>0.05)
and Recreational Release (p=0.164>0.05; p=0.317>0.05)
had a positive effect on Autonomous Social-emotional
Development (p=0.402>0.05) on Brand Relationship
Quality did not show a significant path. Therefore 7 SHS Web of Conferences 159, 02009 (2023) SHS Web of Conferences 159, 02009 (2023)
ICLCC 2023 https://doi.org/10.1051/shsconf/202315902009 Table 4. Hypotheses test results
Impact pathway
SE
CR
p
Standardized
Coefficients Beta
H1a
SF→AVCB
0.087
2.57
0.012
0.2
valid
H1b
SED→AVCB
0.069
3.761
0
0.284
valid
H1c
CED→AVCB
0.099
1.619
0.112
0.146
not valid
H1d
RR→AVCB
0.102
1.413
0.164
0.107
not valid
H1e
SF→SVCB
0.102
2.918
0.004
0.249
valid
H1f
SED→SVCB
0.08
4.262
0
0.352
valid
H1g
CED→SVCB
0.127
0.084
0.933
0.008
not valid
H1h
RR→SVCB
0.117
-1.001
0.317
-0.083
not valid
H2a
SF→BRQ
0.064
3.114
0.002
0.215
valid
H2b
SED→BRQ
0.05
0.851
0.402
0.057
not valid
H2c
CED→BRQ
0.08
2.481
0.015
0.199
valid
H2d
RR→BRQ
0.074
5.558
0
0.373
valid
H3a
BRQ→AVCB
0.064
12.183
0
0.655
valid
H3b
BRQ→SVCB
0.077
8.711
0
0.527
valid online game participation per day; Model 2 adds the
moderator Game Literacy to Model 1, and Model 3 adds
the interaction term (the product of the independent and
moderator variables) to Model 2. Model 1 is to
investigate the effect of the independent variable,
Perceived Value, on the dependent variable, Value Co-
creation Behavior, when the moderating variable, Game
Literacy, is not taken into account. As seen from Table 6,
the independent variable Perceived Value showed
significance (t=8.401, p=0.000<0.05), which means that
Perceived Value has a significant relationship with
Value Co-creation Behavior. The moderating effect can
be viewed in two ways and the first is to view the
significance of the change in F-value when going from
model 2 to model 3; the second is to view the
significance of the interaction term in model 3, this time,
the moderating effect is analyzed in a second way. 5.3.2. Mediating effect test As both Perceived Value and Value Co-creation
Behavior of online game users are multi-dimensional
variables, here, Perceived Value and Value Co-creation
Behavior of online game users are used as overall
variables to test the mediating nature of Brand
Relationship Quality. This study used the process plug-in
in SPSS software to conduct a mediation effect test using
Bootstrap hair, selecting a sample size of 5000 and
model 4, with a confidence interval set at 95%, and
analyzing the mediation effect based on whether the
upper and lower bounds of the confidence interval
contained 0 between them. The test results showed
(Table 5) that the total effect of the Perceived Value of
online game users on Value Co-creation Behavior was
0.733, with a 95% confidence interval of [0.192, 0.141]
excluding 0. The comprehensive analysis concluded that
Brand Relationship Quality significantly affected the
relationship between Perceived Value and Value Co-
creation Behavior. Co-creation Behavior has a partially
mediating role between Perceived Value and Value Co-
creation Behavior, and hypothesis H4 holds. Table 5. Mediating affect test results
Effect
Boot
SE
p
95% Boot
CI
Perceived
Value=>Brand
Relationship
Quality=>Value
Co-creation
Behavior
0.733**
0.057
0.00
0.192 ~
0.414
5.3.3. Reconciliation effect test Table 5. Mediating affect test results Table 6. Reconciliation affects test results
Model1
Model2
Model3
Perceived
Value
0.677**
(8.401)
0.427**
(3.837)
0.427**
(3.823)
Game
Literacy
0.467**
(3.189)
0.466**
(3.168)
Perceived
Value*Game
Literacy
-0.001
(-0.037)
R 2
0.371
0.404
0.404 5.3.3. Reconciliation effect test Model 1 includes the independent variable Perceived
value and five control variables, including Gender, Age,
Educational qualifications, Length of participation in
online games, and the Average number of hours of 8 https://doi.org/10.1051/shsconf/202315902009 SHS Web of Conferences 159, 02009 (2023) SHS Web of Conferences 159, 02009 (2023)
ICLCC 2023 Model1
Model2
Model3
Adjusted R 2
0.352
0.381
0.378
F
F(6,190)=18.7
13,p=0.000
F(7,189)=18.2
66,p=0.000
F(8,188)=15.8
99,p=0.000
△R 2
0.371
0.032
0.000
△F
F(6,190)=18.7
13,p=0.000
F(1,189)=10.1
70,p=0.002
F(1,188)=0.00
1,p=0.970
* p<0.05 ** p<0.01；t values are in parentheses game brand, such as sharing This positive attitude
influences the users’ own behaviour, prompting them to
take actions that create value for the game brand, such as
sharing their gaming experience or actively engaging
with the game brand. Third,
Self
Fulfillment,
Character
Experience
Development, and Recreational Release significantly
impact Brand Relationship Quality. In contrast, Social-
emotional Development does not significantly impact
Brand Relationship Quality. 6.1. Main conclusions Based on consumer perceived value theory and value co-
creation theory, this paper constructs a theoretical model
of the influence of game users’ Perceived Value on
Value Co-creation Behavior, analyzes the mediating role
of Brand Relationship Quality and the moderating role of
Game Literacy, and draws the following conclusions
after empirical research. p
First, the Perceived Value of game users significantly
influences Value Co-creation Behavior. Regarding the
specifics of each dimension, Self Fulfillment and Social-
emotional Development have a significant impact on
Autonomous Value Co-creation Behavior and Sponsored
Value Co-creation Behavior. The effects of Character
Experience and Recreational Release on Autonomous
Value Co-creation Behavior and Sponsored Value Co-
creation Behavior were insignificant. In contrast, the
effects of Character Experience and Recreational
Release on Autonomous Value Co-creation Behavior
and Sponsored Value Co-creation Behavior were
insignificant. This suggests that the different aspects of
perceptions that game users form during the game have
different degrees of influence on their participation in
brand value co-creation behaviour. Game users place
more importance on the sense of achievement and self-
esteem felt in the game and the good interpersonal
relationships and sense of belonging built up during the
game, both of which bring more obvious pleasurable
stimuli. In contrast, character development is the basic
content of online games and the basic value experience
of the game for entertainment and leisure players, both
of which bring less stimulation to players and thus have
a limited effect on user behaviour. 6. CONCLUSION Finally, The moderating effect of Game Literacy in
the relationship between Perceived Value and Value Co-
creation Behavior is not significant, which is contrary to
the original hypothesis of this study. Considering the
actual situation, the reason may be that online games are
a popular and universal entertainment item that does not
require high operating skills, and even people who lack
gaming experience Even those who lack experience in
the game can have fun in the process of exploring new
games. 5.3.2. Mediating effect test This suggests that the sense
of achievement, character immersion, and recreational
release that players perceive in the game can increase
their satisfaction with the game brand. As they progress
through the game, this satisfaction will further deepen
into trust and commitment, enhancing the quality of the
brand relationship to a higher degree. In contrast,
players’ experience of social interaction in games is
mainly reflected in emotional interaction with others and
has no significant impact on players’ good feelings
towards the game brand. * p<0.05 ** p<0.01；t values are in parentheses References 16. Zhang, Tuohi. A study on the interaction between
brand knowledge, brand relationship quality and
word-of-mouth
recommendation
behavior[J]. Business Economics Research,2022(14):75-78. 1. Huatai Securities, High growth in the industry,
outbound + cloud gaming is a bright spot, 2020.8.10,
2022.11.27, https://wk.askci.com/Search/网络游戏/ 17. Zhang Hui, Chen Ye. The impact of brand fit on
brand
relationship
quality
and
repurchase
intention[J]. Journal of Tourism,2017,32(04):43-53. 2. Ai Media Consulting, 2020 China Mobile Game
Industry Development Special Research Report,
2021.1.4,
2022.11.27,
https://www.iimedia.cn/c400/76223.html 18. Zhang Qianwei. Game literacy and education in the
information age[J]. Journal of Electrochemical
Education,2009(11):14-18. 3. Wu Xiaoling. Internet, games and online games [J]. Contemporary Communication,2006(01):54-55+58. 19. Gou Chaoqun. A brief discussion on game
literacy[J]. Agricultural
Network
Information,2012(09):89-92. 4. Dou Congying. Research on brand value co-creation
of virtual communities in online games under social
network environment[D]. Huazhong University of
Science and Technology, 2018. 20. Prahalad C K, Ramaswamy V. Co-creating unique
value with customers[J]. Strategy & leadership,
2004. 5. He Yuanshuo. Research on the model of game brand
value co-creation[D]. Wuhan University, 2020. DOI:10.27379/d.cnki.gwhdu.2020.000265. 21. Wu WJ, Chen QJ. The impact of customer
engagement behavior on their satisfaction and
behavioral intention based on co-creating value
perspective[J]. Management
Review,2017,29(09):167-
180.DOI:10.14120/j.cnki.cn11-5057/f.2017.09.015. 6. Peng Yuhong. Research on value co-creation model
of digital distribution platform for video games[D]. Wuhan University,2019. 7. Drucker P. The practice of management [M]. Routledge, 2012. 22. Pinho N, Beirão G, Patrício L, et al. Understanding
value co-creation in complex services with many
actors [J]. Journal of Service Management, 2014. 8. Park B W, Lee K C. Exploring the value of
purchasing online game items[J]. Computers in
human behavior, 2011, 27(6): 2178-2185. 23. Yi Y, Gong T, Lee H. The impact of other
customers on customer citizenship behavior[J]. Psychology & Marketing, 2013, 30(4): 341-356. 9. Yu P, Sang W T. Exploration of customer value
components in Chinese online game market[J]. Journal of Beijing University of Technology and
Business (Social Science Edition), 2011,26(04):76-
82. DOI:10.16299/j.1009-6116.2011.04.018. 24. Chen E, Ren LJ, Yang Y. The impact of virtual
brand community support on customer value co-
creation: the mediating role of customer readiness[J]. Times Business & Economics,2022,19(07):102-
108.DOI:10.19463/j.cnki.sdjm.2022.07.034. 10. Ouyang Banghong, Liu Chen, Wu Silan, Wei Hua,
Zhou Zongkui. A study on the relationship between
satisfaction, perceived value and purchase behavior
of online games [C]//. Abstracts of the 17th National
Psychology
Conference. [publisher
unknown],2014:94-96. 25. Zwass V. Co-creation: Toward a taxonomy and an
integrated research perspective [J]. 6.2. Countermeasures Game brands integrate
players’ opinions to continuously improve the formation
of game products, which can give consumers a better
gaming experience. 13. Fournier S. Consumers and their brands: Developing
relationship theory in consumer research[J]. Journal
of consumer research, 1998, 24(4): 343-373. 14. Lu, T.H., Zhou, C.M.. Brand theory based on brand
relationships: research models and perspectives[J]. Business Economics and Management, 2003(02):4-
9. 15. Zhao Jinglin, Zhao Hong. Research on the
relationship
between
social
capital,
brand
relationship quality and consumer innovativeness in
virtual
brand
communities[J]. Science
and
Technology Management,2019,40(08):71-86. 6.2. Countermeasures From the perspective of value co-creation theory,
consumers are the value co-creators of a company or
brand. Therefore, the findings of this paper will not only
help game companies to create better game products but
also help game users to build up a positive perception of
the game brand and to participate in value co-creation
behaviour, as well as to facilitate game users to
experience the game and to participate in value co-
creation activities of the game brand. Increasing the perceived value of the game to the
game user. The positive perceived value of a game
product can lead to a positive feeling of loyalty and
engagement with the game brand. Game companies can
consider adding more diverse value realization paths so
different players can feel self-fulfillment in the game. At
the same time, game enterprises need to make the game
have richer interactive gameplay so that players can
more obviously feel the care and recognition among
partners in the game. At the same time, the quality of the
game product itself is also very important. Excellent
game works can give players a better sense of character
immersion and immersive experience, and better feel the
pleasure of the game. Second, Brand Relationship Quality significantly
influences Autonomous Value Co-creation Behavior and
Sponsored Value Co-creation Behavior, suggesting that
in the process of playing games, users build positive
brand relationships with game brands and that this
positive attitude influences their own behaviour,
prompting them to do things that create value for the When gamers find common interests and points of
interest with other gamers, they are more likely to
initiate value co-creation behaviour on their own. Therefore, when promoting their products, game
companies can also seek channels to build bridges 9 https://doi.org/10.1051/shsconf/202315902009 SHS Web of Conferences 159, 02009 (2023)
ICLCC 2023 SHS Web of Conferences 159, 02009 (2023) Advertising’s role in building strong brands, 1993:
113-124. between consumers, such as running a good online
community for games or a branded community app, so
that there is a good environment and atmosphere for
players to communicate. In addition, game companies
also need to launch more different types of co-creation
activities so that players have more channels to contact
the game brand and make the activities more interesting
and fun to increase the willingness of game users to
participate. In this process, the power of players gives
game brands a stronger influence. References International
journal of electronic commerce, 2010, 15(1): 11-48. 26. Li ZHAOhui, Jin Yongsheng, Bu Qingjuan. Study
on the impact of customer participation in virtual
brand community value co-creation on brand equity
- the mediating role of brand experience[J]. Journal
of Marketing Science,2014,10(04):109-124. 11. Zhang Qiyao,Zheng Aicheng. The effect of
promotional stimuli on consumers’ willingness to
co-create green brand values[J]. Journal of Hunan
University
of
Technology
(Social
Science
Edition),2022,27(05):32-41. 27. Ranjan K R, Read S. Value co-creation: concept and
measurement[J]. Journal of the academy of
marketing science, 2016, 44(3): 290-315. 12. Blackston M. Beyond brand personality: building
brand relationships [J]. Brand equity and advertising: 10 SHS Web of Conferences 159, 02009 (2023)
ICLCC 2023 https://doi.org/10.1051/shsconf/202315902009 28. Wei QG. Research and empirical analysis on the
influence mechanism of value co-creation based on
customer experience[J]. Hebei Industrial Science
and
Technology,2013,30(6):407-413. doi:10.7535/hbgykj.2013yx0602. 29. Tang J, Xu YH, Tu QJ. Research on tourism impact
perception and value co-creation behavior of
tourism destination residents 30. Based on the mediating role of subjective well-being
[J]. Journal of Fujian Agriculture and Forestry
University
(Philosophy
and
Social
Science
Edition),2021,24(5):52-59. DOI:10.13322/j.cnki.fjsk.2021.05.007. 31. Yang Shuang, Zhou Xing, Zou Junyi. Study on the
influence of the composition dimensions of SNS
users’ perceived value on relationship quality [J]. Consumer Economics,2010,26(06):63-67. 32. Sha Zhenquan, He Meixian, Jiang Yuwei. Research
on customer engagement affecting the quality of
brand
relationships[J]. Business
Research,2013(10):60-66+103. DOI:10.13902/j.cnki.syyj.2013.10.018. 33. Zhang Xinsheng,Li Xianguo. The impact of virtual
brand community characteristics on consumers’
willingness to co-create value - an explanation based
on the satisfaction and trust mediation model[J]. China
Circulation
Economy,2017,31(07):70-
82.DOI:10.14089/j.cnki.cn11-3664/f.2017.07.009. 34. Xiong Aihua, Chen Xiaoyun, Zhang Qibin. Research on the relationship between perceived
value and value co-creation behavior of virtual
brand
communities[J]. Journal
of
Shandong
University
of
Finance
and
Economics,2022,34(02):92-102. 35. Hu Yinhua,He Yan,Zhong Lei. Research on the
impact of virtual brand community interaction on
brand relationship quality from the perspective of
value co-creation[J]. Journal of Nanchang College
of Engineering,2019,38(02):103-110. 36. Shi R, Han Donglin. Research on the influence of
online game experience and platform services on
user satisfaction in digital culture industry[J]. Journal of Qiqihar University (Philosophy and
Social
Science
Edition),
2021(06):82-88. doi:10.13971/j.cnki.cn23-1435/c.2021.06.020. 37. Jiang Ying. Research on the influence of perceived
value of Hanfu consumers on the willingness to
participate in value co-creation [D]. South China
University
of
Technology,
2021. doi:10.27151/d.cnki.ghnlu.2021.004858. 11
|
https://openalex.org/W3204211274
|
https://zenodo.org/records/5706626/files/BDJ_article_70912.pdf
|
English
| null |
Geographical and temporal distribution of hawkmoth (Lepidoptera: Sphingidae) species in Africa
|
Biodiversity data jurnal
| 2,021
|
cc-by
| 3,863
|
© Kioko E et al. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY
4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are
credited. Geographical and temporal distribution of
hawkmoth (Lepidoptera: Sphingidae) species in
Africa Esther N. Kioko , Alex Mutinda Musyoki , Augustine E. Luanga , Mwinzi Duncan Kioko ,
Esther W. Mwangi , Lawrence Monda
‡ N ti
l M
f K
N i
bi K
‡
‡
‡
‡
‡
‡ ‡ National Museums of Kenya, Nairobi, Kenya Corresponding author: Esther N. Kioko (kiokoesther08@gmail.com)
Academic editor: Rodolphe Rougerie
Received: 01 Jul 2021 | Accepted: 16 Sep 2021 | Published: 09 Nov 2021
Citation: Kioko EN, Musyoki AM, Luanga AE, Kioko MD, Mwangi EW, Monda L (2021) Geographical and
temporal distribution of hawkmoth (Lepidoptera: Sphingidae) species in Africa. Biodiversity Data Journal 9:
e70912. https://doi.org/10.3897/BDJ.9.e70912 Biodiversity Data Journal 9: e70912
doi: 10.3897/BDJ.9.e70912
Data Paper Biodiversity Data Journal 9: e70912
doi: 10.3897/BDJ.9.e70912
Data Paper Biodiversity Data Journal 9: e70912
doi: 10.3897/BDJ.9.e70912
Data Paper Data Paper Kioko E et al 2 Kioko E et al 2 can be easily sampled and identified. However, hawkmoths have rarely been surveyed
over the long term for this purpose. Long-term datasets are of unquestionable significance
for understanding and monitoring temporal changes in biodiversity. These hawkmoth data
have addressed one of the most significant challenges to insect conservation, the lack of
baseline information concerning species diversity and distribution and have provided key
historic hawkmoth species diversity and distribution data that can be used to monitor their
populations in the face of climate change and other environmental degradation issues that
are facing the world today. The publication of the hawkmoth species occurrence data
records in GBIF has enhanced data visibility to a wider audience promoting availability for
use. Keywords hawkmoths, Sphingidae, species, diversity, distribution, Africa, National Museums of
Kenya. New information The hawkmoth (Lepidoptera: Sphingidae) collection at the National Museums of Kenya
was digitised from 2017 – 2020 and this paper presents details of species occurrence
records as in the insect collection at the NMK, Nairobi, Kenya. The collection holds 5,095 voucher specimens consisting of 88 genera and 208 species. The collection covers the period between 1904 and 2020. The geographical distribution of the hawkmoths housed at the NMK covers East Africa at
81.41%, West Africa at 7.20%, Southern Africa at 6.89%, Central Africa at 4.02% and North
Africa at 0.2%. Background Hawkmoths consist of species where most adults are nocturnal, but there are some day-
flying genera. Hawkmoth species have a wide variety of life-history traits, comprising
species with adults (mostly nectarivorous though with some exceptions, honey-feeding),
but there are also species that do not feed at all. The nectarivorous species are an
important component of tropical ecosystems, with significant roles as major pollinators of
both crops and wild flora with the pollination done by the adult stage. Pollinators are in
decline world-wide and there is need for baseline data to provide information about their
conservation strategies. Species occurrence data from Museum collections have been
shown to be of great value as a tool for prioritising conservation actions in Africa. The
National Museums of Kenya (NMK) have a large and active entomology collection that is in
continuous growth. The NMK’s collection of hawkmoths had not been digitised prior to
2017. This moth family Sphingidae includes about 1,602 species and 205 genera
worldwide (Kitching et al. 2018) with the majority of these species occurring in Africa. These moth species can also be used as indicators in biodiversity assessments as they General description Purpose: To create an online freely accessible, openly licensed resource for users. Introduction Insect pollinators have been undergoing a decline in abundance, occurrence and diversity
in many parts of the world (Biesmeijer et al. 2006, Ollerton et al. 2014, Macgregor et al. 2016). Using existing records of hawkmoths in museum and private collections over a 112-
year period, Young et al. (2017) detected declines in eight species of north-eastern U.S. hawkmoth pollinators. Hawkmoth declines may have ecological effects on both the plants
pollinated by these species and vertebrate predators of the moths (Young et al. 2017). Some of the plants, pollinated by hawkmoths including orchids, are rare (Sheviak and
Bowles 1986, Martins and Johnson 2007) highlighting the potential conservation
consequences of hawkmoth population declines especially in Africa where 70% of the
species occur (Kawahara et al. 2009, Ballesteros-Mejia et al. 2013).Hawkmoths are strong
flying Lepidoptera which include many pollinating species that typically feed nocturnally,
but a few feed diurnally on pale-coloured flowers with long corollas and a sweet scent
(Miller 1997, Johnson and Raguso 2015, Johnson et al. 2016). Increasing attention to Geographical and temporal distribution of hawkmoth (Lepidoptera: Sphingidae) ... 3 pollinators and their role in providing ecosystem services has revealed a paucity of studies
on long-term population trends of most insect pollinators in many parts of the world (Young
et al. 2017, Chiquetto-Machado et al. 2018). Museums have long-term datasets of
unquestionable significance for understanding and monitoring temporal changes in
biodiversity and that can address one of the most significant challenges to insect
conservation, the lack of baseline information concerning species diversity and distribution
(Summerville and Crist 2003, Lampe and Striebing 2005). Though hawkmoths play
important roles in the ecosystem in pollination and as indicator species, data on their
diversity, temporal and geographic distribution in Africa is limited. The National Museums of
Kenya have a large active entomology collection that is in continuous growth (Kioko et al. 2020). The NMK’s collection of hawkmoths had not been digitised prior to 2017 (Kioko et
al. 2021). This project was undertaken to excavate data from the NMK collection to avail
data on the spatial and geographical coverage of hawkmoth species in Africa. Project description Title: Digitising the hawkmoth voucher specimens housed at the National Museums of
Kenya. Personnel: Data mining from the National Museums of Kenya collection and additional
field data from the Taita Hills ecosystem that forms the northernmost Eastern Arc
Mountains was done by Esther N. Kioko, Alex M. Musyoki, Augustine Luanga, Duncan
Mwinzi and others. Bioinformatics support for online publication of the data was provided
by Esther W. Mwangi and Lawrence Monda. Funding: The project is supported by the JRS Biodiversity Foundation, USA, with co-
funding provided by National Museums of Kenya. Sampling methods Verification of the taxonomic names was done by checking
against various references (Carcasson 1967, D'Abrera 1986, Kitching and Cadiou 2000,
Kitching 2017).However, Sphingidae is a diverse family of moths and the taxonomy of the
species is far from complete. There are some uncertainties in some of the identifications
considering the existence of species complexes and also considering the dynamic nature
of taxonomic treatments and changes in species concepts associated with the names
used. Sampling methods Study extent: The digitised hawkmoth voucher specimens are all from Africa with several
regions: East Africa at 81.41% with 4,148 records, West Africa at 7.20% with 367 records,
Southern Africa at 6.89% with 351 records, Central Africa at 4.02% with 205 records and
North Africa at 0.2% with one record.The spatial coverage within the five African regions is
as shown in Fig. 1. The leading records for East Africa consists of 2,566 from Kenya, 829
from Uganda and 427 from Tanzania. Sampling description: The hawkmoth specimens, housed at the NMK Invertebrate
collection, are the result of multiple field expeditions and research projects. Most of the
specimens lack information on the sampling protocol and, in case a certain method was Kioko E et al 4 used, then it was not indicated on the specimen label. The specimens were first catalogued
and pinned; they were then preserved by drying in an oven. Figure 1. Geographic coverage of the hawkmoth voucher specimens housed at the National Museums
of Kenya. g
Geographic coverage of the hawkmoth voucher specimens housed at the National Museums
of Kenya. Quality control: Once the specimens are brought to the invertebrate collection, taxa
experts revise the associated metadata i.e. species name (taxonomy) and locality. The
geographical coordinates that were lacking, as is the case with old museum specimens,
were obtained using a geo-referencing web service GEOLocate (Rios 2014) by use of
available textual locality data. Verification of the taxonomic names was done by checking
against various references (Carcasson 1967, D'Abrera 1986, Kitching and Cadiou 2000,
Kitching 2017).However, Sphingidae is a diverse family of moths and the taxonomy of the
species is far from complete. There are some uncertainties in some of the identifications
considering the existence of species complexes and also considering the dynamic nature
of taxonomic treatments and changes in species concepts associated with the names
used. Quality control: Once the specimens are brought to the invertebrate collection, taxa
experts revise the associated metadata i.e. species name (taxonomy) and locality. The
geographical coordinates that were lacking, as is the case with old museum specimens,
were obtained using a geo-referencing web service GEOLocate (Rios 2014) by use of
available textual locality data. Geographic coverage Description: The digitised hawkmoth voucher specimens are all from different regions
within Africa as follows: East Africa at 81.41% with 4,148 records, West Africa at 7.20%
with 367 records, Southern Africa at 6.89% with 351 records, Central Africa at 4.02% with Geographical and temporal distribution of hawkmoth (Lepidoptera: Sphingidae) ... 5 205 records, North Africa at 0.2% with one record and records not with assigned region at
0.02% with the spatial coverage as shown in Fig. 1. 205 records, North Africa at 0.2% with one record and records not with assigned region at
0.02% with the spatial coverage as shown in Fig. 1. Coordinates: -35.174 and 37.44 Latitude; -17.578 and 52.383 Longitude. Coordinates: -35.174 and 37.44 Latitude; -17.578 and 52.383 Longitude. oordinates: -35.174 and 37.44 Latitude; -17.578 and 52.383 Longitude. Kioko E et al Kioko E et al 6 Temporal coverage Notes: The digitised hawkmoth collections date from 1904 to 2020. The years 1960-1964
recorded the highest values at 1,689 followed by 2010 – 2020 with 857 records, while the
period 1900-1909 recorded the least at four records (Fig. 2).The voucher specimens were
collected throughout the year with the highest month of collection being April with 696
records, followed by December with 570, while the months with fewest collection records
were November with 235 and September with 234 (Fig. 3). Collection data
Collection name: Invertebrate Zoology Section Collection, National Museums of Kenya
Specimen preservation method: Pinned
Curatorial unit: Species collecting event. Usage licence
Usage licence: Creative Commons Public Domain Waiver (CC-Zero)
IP rights notes: This work is licensed under a Creative Commons Attribution Non
Commercial (CC-BY-NC) 4.0 Licence. Figure 3. Monthly abundance of hawkmoth collection at the National Museums of Kenya. Figure 3. Monthly abundance of hawkmoth collection at the National Museums of Kenya. Figure 3. Monthly abundance of hawkmoth collection at the National Museums of Kenya. Figure 3. Monthly abundance of hawkmoth collection at the National Museums of Kenya. Collection data Collection name: Invertebrate Zoology Section Collection, National Museums of Kenya
Specimen preservation method: Pinned
Curatorial unit: Species collecting event. Taxonomic coverage Description: At the National Museums of Kenya, there are 5,095 hawkmoth voucher
specimens that have been digitised and published in GBIF through the Integrated
Publishing Tool (Kioko et al. 2021). The specimens comprise 243 species belonging to 88
genera, with the leading genus in occurrence records being Temnora at 699. Amongst the
species, Hippotion celerio is the most abundant with 403 records, Agrius convolvuli at 210,
Leucophlebia afra at 137, Euchloron megaera at 121, Nephele comma at 107, Hippotion
eson at 103 and Acherontia atropos at 100. Taxa included:
Rank
Scientific Name
Common Name
family
Sphingidae
Hawkmoths
Figure 2. Temporal abundance of hawkmoth collection at the National Museums of Kenya. Temporal abundance of hawkmoth collection at the National Museums of Kenya. Common Name Data resources Data package title: Occurrence data of hawkmoths (Lepidoptera: Sphingidae) in the
National Museums of Kenya Zoological Collection
Resource link: https://www.gbif.org/dataset/302155f9-49a6-4ee2-a01b-293067ddeeed
Alternative identifiers: 302155f9-49a6-4ee2-a01b-293067ddeeed, http://ipt.museums. or.ke/ipt/resource?r=hawkmoth_nmk_i Data package title: Occurrence data of hawkmoths (Lepidoptera: Sphingidae) in the
National Museums of Kenya Zoological Collection
Resource link: https://www.gbif.org/dataset/302155f9-49a6-4ee2-a01b-293067ddeeed
Alternative identifiers: 302155f9-49a6-4ee2-a01b-293067ddeeed, http://ipt.museums. or.ke/ipt/resource?r=hawkmoth_nmk_i Usage licence Usage licence: Creative Commons Public Domain Waiver (CC-Zero) IP rights notes: This work is licensed under a Creative Commons Attribution Non
Commercial (CC-BY-NC) 4.0 Licence. Geographical and temporal distribution of hawkmoth (Lepidoptera: Sphingidae) ... 7 Number of data sets: 1 Data set name: Occurrence data of hawkmoths (Lepidoptera: Sphingidae) in the
National Museums of Kenya Zoological Collection Download URL: https://www.gbif.org/dataset/302155f9-49a6-4ee2-a01b-293067
ddeeed Data format: omma-separated values (CSV) Description: This resource is a digitised format of data on the occurrence of hawkmoth
species housed in the Zoology Department, National Museums of Kenya insect
collection. The data provide baseline information on the distribution of different
hawkmoth species and can be used for future ecology studies on hawkmoths, as well
as monitoring of population trends in various habitats. Column label
Column description
occurrenceID
An identifier for the Occurrence. basisOfRecord
The specific nature of the data record. eventDate
The date-time when the event was recorded. year
The four-digit year in which the Event occurred, according to the Common Era
Calendar. month
The integer month in which the Event occurred. day
The integer day of the month on which the Event occurred. scientificName
The full scientific name, with authorship and date information, if known. higherClassification
A list (concatenated and separated) of taxa names terminating at the rank immediately
superior to the taxon referenced in the taxon record. kingdom
The full scientific name of the kingdom in which the taxon is classified. phylum
The full scientific name of the phylum or division in which the taxon is classified. class
The full scientific name of the class in which the taxon is classified. order
The full scientific name of the order in which the taxon is classified. family
The full scientific name of the family in which the taxon is classified. Kioko E et al Kioko E et al 8 8 genus
The full scientific name of the genus in which the taxon is classified. specificEpithet
The name of the first or species epithet of the scientificName. taxonRank
The taxonomic rank of the most specific name in the scientificName. nomenclaturalCode
The nomenclatural code (or codes in the case of an ambiregnal name) under which
the scientificName is constructed. decimalLatitude
The geographic latitude (in decimal degrees, using the spatial reference system given
in geodeticDatum) of the geographic centre of a Location. decimalLongitude
The geographic longitude (in decimal degrees, using the spatial reference system
given in geodeticDatum) of the geographic centre of a Location. geodeticDatum
The ellipsoid, geodetic datum or spatial reference system (SRS) upon which the
geographic coordinates given in decimalLatitude and decimalLongitude are based. Additional information Kioko E, Musyoki A, Luanga A, Sese J, Nyangena L, Mwinzi D (2021): Occurrence data of
hawkmoths (Lepidoptera: Sphingidae) in the National Museums of Kenya Zoological
Collection. v.1.7. National Museums of Kenya. Dataset/Occurrence. http://ipt. museums.or.ke/ipt/resource?r=hawkmoth_nmk_i&v=1.7 Acknowledgements We acknowledge and thank the JRS Biodiversity Foundation for the financial support that
has facilitated this work. We thank the NMK management for the support that was given to
this work. Stephen W. Waweru is acknowledged for the assistance in mapping the species
occurrence data in Africa. Author contributions Esther Kioko conceived the study, collected field and collection data and wrote the
manuscript, Alex Mutinda collected field and collection data, analysed data, reviewed and
edited the manuscript, A. Luanga, and D. Mwinzi collected field and collection data and
reviewed manuscript, Esther Mwangi and Lawrence Monda provided bioinformatics skills,
reviewed and edited the manuscript. Geographical and temporal distribution of hawkmoth (Lepidoptera: Sphingidae) ... 9 Number of data sets: 1 verbatimCoordinateSystem The coordinate format for the verbatimLatitude and verbatimLongitude or the
verbatimCoordinates of the Location. georeferencedBy
A list (concatenated and separated) of names of people, groups or organisations who
determined the georeference (spatial representation) for the Location. georeferencedDate
The date on which the Location was georeferenced. higherGeography
A list (concatenated and separated) of geographic names less specific than the
information captured in the locality term. continent
The name of the continent in which the Location occurs. country
The name of the country or major administrative unit in which the Location occurs. countryCode
The standard code for the country in which the Location occurs. locality
The specific description of the place. type
The set of classes specified by the Darwin Core Type Vocabulary, used to categorise
the nature or genre of the resource. language
The language in which the resource is written. institutionID
An identifier for the institution having custody of the object(s) or information referred to
in the record. institutionCode
The name (or acronym) in use by the institution having custody of the object(s) or
information referred to in the record. collectionID
An identifier for the collection or dataset from which the record was derived. collectionCode
The name, acronym, coden or initialism identifying the collection or dataset from which
the record was derived. catalogNumber
An identifier (preferably unique) for the record within the dataset or collection. IndividualCount
The number of individuals represented present at the time of the Occurrence. organismQuantity
A number or enumeration value for the quantity of organisms. References •
Ballesteros-Mejia L, Kitching IJ, Jetz W, Nagel P, Beck J (2013) Mapping the
biodiversity of tropical insects: species richness and inventory completeness of African
sphingid moths. Global Ecology Biogeography 22 (586595): 586‑59. https://doi.org/
10.1111/geb.12039 •
Ballesteros-Mejia L, Kitching IJ, Jetz W, Nagel P, Beck J (2013) Mapping the
biodiversity of tropical insects: species richness and inventory completeness of African
sphingid moths. Global Ecology Biogeography 22 (586595): 586‑59. https://doi.org/
10.1111/geb.12039 10.1111/geb.12039
•
Biesmeijer JC, Roberts SPM, Reemer M, Ohlemüller R, Edwards M, Peeters T,
Schaffers AP, Potts SG, Kleukers R, Thomas CD, Settele J, Kunin WE, et al. (2006)
Parallel declines in pollinators and insect-pollinated plants in Britain and the
Netherlands. Science 313 (5785): 351‑4. https://doi.org/10.1126/science.1127863
•
Carcasson R (1967) Revised catalogue of the African Sphingidae (Lepidoptera) with
Descriptions of the East African Species. Journal of The East Africa Natural History
Society and National Museum XXVI (3): 1 148 URL: https://www biodiversitylibrary org/ •
Biesmeijer JC, Roberts SPM, Reemer M, Ohlemüller R, Edwards M, Peeters T,
Schaffers AP, Potts SG, Kleukers R, Thomas CD, Settele J, Kunin WE, et al. (2006)
Parallel declines in pollinators and insect-pollinated plants in Britain and the
Netherlands. Science 313 (5785): 351‑4. https://doi.org/10.1126/science.1127863 •
Carcasson R (1967) Revised catalogue of the African Sphingidae (Lepidoptera) with
Descriptions of the East African Species. Journal of The East Africa Natural History
Society and National Museum XXVI (3): 1‑148. URL: https://www.biodiversitylibrary.org/
content/part/EANHS/XXVI_No.3__115__1_1967_Carcasson.pdf •
Chiquetto-Machado P, Amorim F, Duarte M (2018) Long-term stability of the hawkmoth
fauna (Lepidoptera, Sphingidae) in a protected area of Brazilian Atlantic Rain Forest. Journal of Insect Conservation 22 (2): 277‑286. https://doi.org/10.1007/s10841-
018-0061-0 10 Kioko E et al •
D'Abrera B (1986) Sphingidae Mundi: hawk moths of the world: based on a checklist by
Alan Hayes and the collection he curated in the British Museum (Natural History). EW
Classey Ltd, 226 pp. [ISBN 9780860960225] •
Johnson S, Raguso R (2015) The long-tongued hawkmoth pollinator niche for native
and invasive plants in Africa. Annals of Botany 117 (1): 25‑36. https://doi.org/10.1093/
aob/mcv137 •
Johnson S, Moré M, Amorim F, Haber W, Frankie G, Stanley D, Cocucci A, Raguso R
(2016) The long and the short of it: a global analysis of hawkmoth pollination niches and
interaction networks. Functional Ecology 31 (1): 101‑115. https://doi.org/10.1111/
1365-2435.12753 •
Kawahara A, Mignault A, Regier J, Kitching I, Mitter C (2009) Phylogeny and
biogeography of hawkmoths (Lepidoptera: Sphingidae): Evidence from five nuclear
genes. References PLOS One 4 (5). https://doi.org/10.1371/journal.pone.0005719 Kioko E, Musyoki A, Luanga A, Kioko M, Mwangi E, Monda L (2020) Swallowtail
butterflies (Lepidoptera: Papilionidae) species diversity and distribution in Africa: The •
Kioko E, Musyoki A, Luanga A, Kioko M, Mwangi E, Monda L (2020) Swallowtail
butterflies (Lepidoptera: Papilionidae) species diversity and distribution in Africa: The
Papilionidae collection at the National Museums of Kenya, Nairobi, Kenya. Biodiversity
Data Journal 8 https://doi org/10 3897/bdj 8 e50664 Papilionidae collection at the National Museums of Kenya, Nairobi, Kenya. Biodiversity •
Kioko E, Musyoki A, Luanga A, Sese J, Nyangena L, Mwinzi D (2021) Occurrence data
of hawkmoths (Lepidoptera: Sphingidae) in the National Museums of Kenya Zoological
Collection. Version 1.7. National Museums of Kenya. 1.7. GBIF. Release
date: 2021-5-25. URL: https://www.gbif.org/dataset/302155f9-49a6-4ee2-a01b-
293067ddeeed •
Kitching I, Rougerie R, Zwick A, Hamilton C, St Laurent R, Naumann S, Ballesteros
Mejia L, Kawahara A (2018) A global checklist of the Bombycoidea (Insecta:
Lepidoptera). Biodiversity Data Journal 6 https://doi.org/10.3897/bdj.6.e22236 •
Kitching I, Rougerie R, Zwick A, Hamilton C, St Laurent R, Naumann S, Ballesteros
Mejia L, Kawahara A (2018) A global checklist of the Bombycoidea (Insecta:
Lepidoptera). Biodiversity Data Journal 6 https://doi.org/10.3897/bdj.6.e22236 p
p
)
y
p
g
j
•
Kitching IJ, Cadiou J (2000) Hawkmoths of the world. An annotated and illustrated
revisionary checklist (Lepidoptera: Sphingidae). Cornell University Press
URL: http://zoobank.org/32574dd1-5433-4241-8539-70b1e2633402
•
Kitching IJ (2017) Sphingidae taxonomic inventory. URL: http://sphingidae. myspecies.info/ •
Kitching IJ, Cadiou J (2000) Hawkmoths of the world. An annotated and illustrated
revisionary checklist (Lepidoptera: Sphingidae). Cornell University Press
URL: http://zoobank.org/32574dd1-5433-4241-8539-70b1e2633402 •
Kitching IJ (2017) Sphingidae taxonomic inventory. URL: http://sphingidae. myspecies.info/ •
Lampe K, Striebing D (2005) How to digitize large insect collections — preliminary
results of the DIG project. African Biodiversity385‑393. https://doi.org/10.1007/
0-387-24320-8_38 p
g
(
)
g
g
p
y
results of the DIG project. African Biodiversity385‑393. https://doi.org/10.1007/
0-387-24320-8_38
•
Macgregor CJ, Evans DM, Fox R, Pocock MJO (2016) The dark side of street lighting:
impacts on moths and evidence for the disruption of nocturnal pollen transport. Global
Change Biology 23 (2): 697‑707. https://doi.org/10.1111/gcb.13371
•
Martins DJ Johnson SD (2007) Hawkmoth pollination of aerangoid orchids in Kenya Macgregor CJ, Evans DM, Fox R, Pocock MJO (2016) The dark side of street lighting:
impacts on moths and evidence for the disruption of nocturnal pollen transport. Global
Change Biology 23 (2): 697‑707. •
Sheviak CJ, Bowles ML (1986) The prairie fringed orchids: a pollinator-isolated species
pair. Rhodora 88: 267‑290.
•
Summerville K, Crist T (2003) Determinants of lepidopteran community composition and
species diversity in eastern deciduous forests: roles of season, eco-region and patch
size. Oikos 100 (1): 134‑148. https://doi.org/10.1034/j.1600-0706.2003.11992.x
•
Young B, Auer S, Ormes M, Rapacciuolo G, Schweitzer D, Sears N (2017) Are
pollinating hawk moths declining in the Northeastern United States? An analysis of
collection records. PLOS One 12 (10). https://doi.org/10.1371/journal.pone.0185683 References https://doi.org/10.1111/gcb.13371 •
Martins DJ, Johnson SD (2007) Hawkmoth pollination of aerangoid orchids in Kenya,
with special reference to nectar sugar concentration gradients in the floral spurs. American Journal of Botany 94 (4): 650‑9. https://doi.org/10.3732/ajb.94.4.650 •
Miller WE (1997) Diversity and evolution of tongue length in hawkmoths (Sphingidae). Journal of the Lepidopterists’ Society 51: 9‑31. URL: https://images.peabody.yale.edu/
lepsoc/jls/1990s/1997/1997-51(1)9-Miller.pdf •
Ollerton J, Erenler H, Edwards M, Crockett R (2014) Extinctions of aculeate pollinators
in Britain and the role of large-scale agricultural changes. Science 346 (6215):
1360‑1362. https://doi.org/10.1126/science.1257259 •
Rios NE (2014) Georeferencing natural history collections data: The GEOLocate
Project. Digitization Workshop. Geographical and temporal distribution of hawkmoth (Lepidoptera: Sphingidae) ... 11 •
Young B, Auer S, Ormes M, Rapacciuolo G, Schweitzer D, Sears N (2017) Are
pollinating hawk moths declining in the Northeastern United States? An analysis of
collection records. PLOS One 12 (10). https://doi.org/10.1371/journal.pone.0185683
|
https://openalex.org/W4226166273
|
http://www.ajnr.org/content/ajnr/42/12/2130.full.pdf
|
English
| null |
Deep Learning Enables 60% Accelerated Volumetric Brain MRI While Preserving Quantitative Performance: A Prospective, Multicenter, Multireader Trial
|
American journal of neuroradiology
| 2,021
|
cc-by
| 5,355
|
ABSTRACT BACKGROUND AND PURPOSE: In this prospective, multicenter, multireader study, we evaluated the impact on both image quality
and quantitative image-analysis consistency of 60% accelerated volumetric MR imaging sequences processed with a commercially
available, vendor-agnostic, DICOM-based, deep learning tool (SubtleMR) compared with that of standard of care. MATERIALS AND METHODS: Forty subjects underwent brain MR imaging examinations on 6 scanners from 5 institutions. Standard of care
and accelerated datasets were acquired for each subject, and the accelerated scans were enhanced with deep learning processing. Standard
of care, accelerated scans, and accelerated–deep learning were subjected to NeuroQuant quantitative analysis and classified by a neuroradi-
ologist into clinical disease categories. Concordance of standard of care and accelerated–deep learning biomarker measurements were
assessed. Randomized, side-by-side, multiplanar datasets (360 series) were presented blinded to 2 neuroradiologists and rated for apparent
SNR, image sharpness, artifacts, anatomic/lesion conspicuity, image contrast, and gray-white differentiation to evaluate image quality. RESULTS: Accelerated–deep learning was statistically superior to standard of care for perceived quality across imaging features de-
spite a 60% sequence scan-time reduction. Both accelerated–deep learning and standard of care were superior to accelerated
scans for all features. There was no difference in quantitative volumetric biomarkers or clinical classification for standard of care
and accelerated–deep learning datasets. CONCLUSIONS: Deep learning reconstruction allows 60% sequence scan-time reduction while maintaining high volumetric quantifi-
cation accuracy, consistent clinical classification, and what radiologists perceive as superior image quality compared with standard
of care. This trial supports the reliability, efficiency, and utility of deep learning–based enhancement for quantitative imaging. Shorter scan times may heighten the use of volumetric quantitative MR imaging in routine clinical settings. D D
eep learning (DL) is a subset of machine learning that uses
convolutional neural networks to process large volumes of
data.1-6 While traditional reconstruction techniques can be lim-
ited by long scan times, SNR constraints, and motion artifacts, the recent application of DL to image reconstruction can enable
faster image acquisitions with equal or enhanced image quality.2,3
While DL can boost SNR among other advantages over conven-
tional methods,2-6 concerns exist over whether postprocessing
can mask or alter pathology and whether the quantitative values
derived are consistent with those obtained from routine standard
of care (SOC) scans over the gamut of scanner vendors and
across field strengths. of October 23, 2024.
This information is current as
Multicenter, Multireader Trial
Quantitative Performance: A Prospective,
Volumetric Brain MRI While Preserving
Deep Learning Enables 60% Accelerated
Shankaranarayanan and L.N. Tanenbaum
S. Bash, L. Wang, C. Airriess, G. Zaharchuk, E. Gong, A.
http://www.ajnr.org/content/42/12/2130
https://doi.org/10.3174/ajnr.A7358
doi:
2021, 42 (12) 2130-2137
AJNR Am J Neuroradiol of October 23, 2024. This information is current as of October 23, 2024. This information is current as ORIGINAL RESEARCH
ADULT BRAIN ORIGINAL RESEARCH
ADULT BRAIN Received October 25, 2020; accepted after revision August 17, 2021.
From RadNet Inc (S.B., L.N.T.), Los Angeles, California; Subtle Medical (L.W., E.G.,
A.S.), Menlo Park, California; Cortechs.ai. (C.A.), San Diego, California; Stanford
University Medical Center (G.Z.), Stanford, California; and Lenox Hill Radiology
(L.N.T.), New York, New York.; and
Paper previously presented at: Annual Meetings and Exhibitions of the
International Society for Magnetic Resonance in Medicine and the Society for
Magnetic Resonance Technologists, August 8–14, 2020; Virtual; and May 14–20,
2021, Virtual; Annual Meetings of the American Society of Neuroradiology, May 30
to June 4, 2020; Virtual; and May 15–20, 2021; Virtual; and Annual Meeting of the
European Congress of Radiology, May 3–7, 2021, Porto Alegre, Brazil.
Please address correspondence to Suzie Bash, MD, RadNet, 1510 Cotner Ave,
Los Angeles, CA 90025; e-mail: suzie.bash@radnet.com; @suzie_bash
Indicates open access to non-subscribers at www.ajnr.org
http://dx.doi.org/10.3174/ajnr.A7358
2130
Bash
Dec 2021
www.ajnr.org ABBREVIATIONS: DL ¼ deep learning; FAST ¼ accelerated scan; HOC ¼ hippocampal occupancy score; HV ¼ hippocampal volumes; ILV ¼ inferior lateral
ventricles; MCI ¼ mild cognitive impairment; SLV ¼ superior lateral ventricles; SOC ¼ standard of care Participants With Western Institutional Review Board approval and patient
consent, 40 consecutive subjects (mean age, 69 [SD, 17]years; 21
men, 19 women) undergoing clinically indicated brain MR imag-
ing examinations for subjective memory loss were prospectively
recruited during an 8-month period. ABSTRACT MR imaging depicts brain anatomy with high spatial and con-
trast resolution, qualities crucial when applying anatomic seg-
mentation and quantitative volumetric analysis. Quantitative
volumetric analysis requires 3D radiofrequency spoiled gradient-
echo T1-weighted scans, and k-space acceleration opportunities
have limits. Reduced excitations and undersampling techniques
like compressed sensing accelerate scans at a cost of increased
image noise (reduced SNR). Acceleration via decreases in the
imaging matrix can improve both SNR and contrast resolution 2130 Hitachi, and so forth), scanner models, field strengths, and clini-
cal sites, as well as a variety of disease states/clinical indications,
thus experiencing a range of tissue contrasts, acquisition parame-
ters, patient anatomies, and variable image quality. but reduce image sharpness. For this study, both undersampling
and reduced imaging matrices were used for the accelerated scans
(FAST), which were then processed with a commercially available
DL tool (SubtleMR; Subtle Medical) that provides both denoising
and sharpness enhancement (FAST-DL). Our goal was to match
or exceed SOC image SNR and spatial resolution while maintain-
ing clinical and quantitative integrity.7 The DL network was trained on paired low-/high-resolution
images to impart structure-preserving noise reduction and sharp-
ness enhancement to newly acquired images.7 Processing does
not use proprietary raw k-space input (DICOM-based) and is,
thus, vendor-agnostic. For the study, DL processing required
,1 minute per series on a scanner-connected GPU server and
finished before the next sequence acquisition was completed,
thus not impacting overall examination time. Images were gath-
ered from different sites and presented to the reviewers on a com-
mercial DICOM viewer. In this prospective, multireader, multicenter study, we
explored the impact of DL-based enhancement of 60% acceler-
ated 3D T1-weighted brain MR image acquisitions. We found
that the DL-processed images demonstrated high volumetric
quantification accuracy and matched clinical disease status pre-
dictability and provided what readers perceived as superior image
quality compared with the longer SOC examinations, suggesting
good generalizability, accuracy, and potential utility of DL
enhancement in routine clinical settings. The SOC, FAST, and FAST-DL image sets were processed
with a machine learning–based FDA-cleared quantitative volu-
metric software product, NeuroQuant (Cortechs.ai). The hippo-
campal occupancy score (HOC), a biomarker to predict the
progression of neurodegenerative diseases, as well as the volumes
of the hippocampi (HV), superior lateral ventricles (SLV), and in-
ferior lateral ventricles (ILV) were analyzed for this study. Image Acquisition Imaging was obtained on 6 scanners (3T Skyra, Siemens; 3T
Discovery 750 and 3T Discovery 750w, 3T Signa Premier, and
1.5T HDxt, GE Healthcare; 3T Verio, Siemens) at 3T (n ¼ 32)
and 1.5T (n ¼ 8) from 5 different institutions in New York
(n ¼ 8) and California (n ¼ 32). The image acquisitions consisted of paired 3D T1-weighted
sagittally acquired datasets: 1 SOC image set (mean scan time, 6
minutes, 56 seconds for 3T and 4minutes for 1.5T), and 1 FAST
image set (mean scan time, 2 minutes, 44 seconds for 3T and 2
minutes, 40 seconds for 1.5T). Native sagittal images were acquired
and quantitatively postprocessed at 1-mm section thickness (for
3T) and 1.2-mm section thickness (for 1.5T). The spatial-resolu-
tion matrix was 256 256 for SOC and 128 256 for FAST. Acquisition parameters included MPRAGE (flip angle ¼ 9°, TI ¼
1100ms, section thickness ¼ 1.0–1.2mm); fast-spoiled gradient
recalled (flip angle ¼ 8°, TI ¼ 900ms, section thickness ¼ 1.0–
1.2mm); and BRAVO (GE Healthcare; flip angle ¼ 12°, TI ¼
450ms, section thickness ¼ 1.0–1.2 mm). The average pooled (3T
+ 1.5T) 3D T1WI scan times for all 40 patients (n ¼ 32/3T and
n ¼ 8/1.5T) were 6 minutes, 1 second for SOC (range, 4–7
minutes) and 2 minutes, 43 seconds for the FAST dataset (range,
2–2 minutes, 50 seconds), representing a 60% sequence scan-time
reduction for the FAST acquisition. To assess clinical classification performance, we categorized
the quantitative biomarkers obtained from 80 datasets (40 SOC
and 40 FAST-DL) in a blinded, randomized fashion. Each dataset
was rated using a binary predictive classification system (healthy/
mild cognitive impairment [MCI] versus dementia) with ground
truth established according to the statistical significance of 3 bio-
markers falling .2 SDs from the mean: HOC (,5%), HV
(,5%), and ILV (.95%) based on an (n . 4000) age- and sex-
matched normative data base, with 0/3 and 1/3 statistically signif-
icant biomarkers categorized as healthy/MCI; and 2/3 and 3/3
categorized as likely dementia. Radiologic Assessment For the image-quality assessment, 2 experienced board-certified
neuroradiologists (.17years’ experience each) were presented
with 40 paired side-by-side multiplanar 3D T1-weighted series
datasets (360 series). The blinded datasets (SOC versus FAST,
SOC versus FAST-DL, FAST-DL versus FAST) were randomized
in disease classification, image plane, and left-right display order. The readers evaluated 2 images side-by-side and provided a single
Likert scale ranking between 1 and 5 that described whether the
left or right image was superior (3 ¼ both images were preferred
equally; 2 or 4 ¼ right/left mildly preferred; 1 or 5 ¼ right/left
strongly preferred) for the following: 1) perceived SNR; 2) per-
ceived spatial resolution (sharpness); 3) imaging artifacts; 4) ana-
tomic/lesion conspicuity; 5) image contrast; and 6) gray-white
matter differentiation. Anatomic conspicuity of brain structures
such as the deep gray nuclei was used in cases in which a lesion
was not conspicuous on the 3D T1-weighted images. Sample
lesions in our datasets included infarcts and prominent white
matter ischemic disease. AJNR Am J Neuroradiol 42:2130–37
Dec 2021
www.ajnr.org RESULTS Qualitative and Quantitative Performance
FAST-DL was statistically superior to SOC in subjective image
quality
for
perceived
SNR,
sharpness,
artifact
reduction,
anatomic/lesion conspicuity, and image contrast (all P values
, .008), despite a 60% reduction in sequence scan time. Both
FAST-DL and SOC were statistically superior to FAST for all ana-
lyzed features (all P values , .001). Wilcoxon rank sum statistical
results are collectively summarized in Table 1. Table 3: Paired t test for SOC versus FASTa
SOC
FAST
Paired
t Test
HOC (mean)
0.68 (SD, 0.16)
0.68 (SD, 0.17)
0.63
HV (mean) (cm3)
6.45 (SD, 1.70)
6.56 (SD, 1.88)
0.60
SLV volume (mean)
(cm3)
44.30 (SD, 20.60)
43.44 (SD, 20.01)
,0.05
ILV volume (mean)
(cm3)
3.07 (SD, 1.78)
3.17 (SD, 1.85)
0.93
a There is less optimal agreement between SOC versus FAST (compared with SOC
versus FAST-DL) for quantitative assessment of HOC, HV, SLV volume, and ILV
volume. Table 3: Paired t test for SOC versus FASTa
SOC
FAST
Paired
t Test
HOC (mean)
0.68 (SD, 0.16)
0.68 (SD, 0.17)
0.63
HV (mean) (cm3)
6.45 (SD, 1.70)
6.56 (SD, 1.88)
0.60
SLV volume (mean)
(cm3)
44.30 (SD, 20.60)
43.44 (SD, 20.01)
,0.05
ILV volume (mean)
(cm3)
3.07 (SD, 1.78)
3.17 (SD, 1.85)
0.93
a There is less optimal agreement between SOC versus FAST (compared with SOC
versus FAST-DL) for quantitative assessment of HOC, HV, SLV volume, and ILV
volume. y
Paired t test analysis demonstrated excellent agreement of
quantitative data on both the SOC and FAST-DL images (Table
2). As expected, there was less agreement between the SOC and
FAST datasets (Table 3) due to the lower spatial resolution of the
FAST images. There was no statistically significant difference
between mean HOC values in the SOC (0.68 [SD, 0.16]) and
the mean FAST-DL (0.68 [SD, 0.16]) datasets. The difference of
the HV, SLV volumes, and ILV volumes was also negligible
(,2%) for the SOC and FAST-DL datasets. The linear regression
graphs (Fig 1) and Bland-Altman plot graph analysis (Fig 3) fur-
ther demonstrated strong agreement between quantitative values
in each dataset across the range of conditions (normal, MCI,
Alzheimer disease) with the HOC ranging from 0.32 to 0.95 mL. There was 100% agreement in clinical disease classification of
both the SOC and FAST-DL datasets (n = 29 healthy/MCI and
n = 11 dementia). Statistical Analysis Wilcoxon rank sum tests were performed to assess the equiva-
lence or superiority of the image quality for each feature (Table
1). Statistically significant superiority for a feature was deter-
mined by a P value , .05. There was excellent interreader agreement between the 2 neu-
roradiologists on the Spearman rank correlation test applied to
the Likert image quality ratings, with a Spearman r value of
0.725 (P , .01). The k value of 0.62 (P , .001) also confirms sub-
stantial interrater agreement on the Likert scale rankings. Adjustment for significance tests for multiple comparisons
was made using the Bonferroni correction, which adjusts the sig-
nificance level to P , .05/.06 (0.00833). Paired t test analysis was performed to test the equivalence
of quantitative data on both the SOC versus FAST-DL images
(Table 2) and SOC versus FAST images (Table 3). Linear regres-
sion graphs (Figs 1 and 2) and Bland-Altman analysis (Figs 3
and 4) were performed to assess quantitative volumetric bio-
marker equivalence of the datasets. The Spearman rank correla-
tion test was applied to assess interreader agreement between
the 2 neuroradiologists on image-quality ratings. Additionally, RESULTS The cross-correlation factor and degree of scat-
ter was consistently worse for the SOC and FAST images com-
pared with the SOC and FAST-DL images as demonstrated on
the linear regression graphs (Fig 2) and Bland Altman plot graph
analysis (Fig 4). a There is less optimal agreement between SOC versus FAST (compared with SOC
versus FAST-DL) for quantitative assessment of HOC, HV, SLV volume, and ILV
volume. a There is less optimal agreement between SOC versus FAST (compared with SOC
versus FAST-DL) for quantitative assessment of HOC, HV, SLV volume, and ILV
volume. again after postprocessing with NeuroQuant, with the goal of vis-
ually assessing the quality of matched color-coded segmentation
of the latter. The FAST scans were excluded from this analysis
because they would not be typically used as input to quantitative
segmentation software given their lower spatial resolution. Image Processing FAST-DL was performed off-line using an FDA-cleared, vendor-
agnostic, DICOM-based, convolutional neural networks–de-
pendent deep learning artificial intelligence image-enhancement
software product, SubtleMR (Version 1.2). The training set
included hundreds of thousands of MR imaging datasets from a
variety of vendors (GE Healthcare, Philips Healthcare, Siemens, Following qualitative feature ranking and quantitative analy-
sis, both readers were presented the SOC and FAST-DL datasets
in a side-by-side fashion, randomized in right-left orientation to
qualitatively assess the overall diagnostic quality of the 3D T1-
weighted images before postprocessing with NeuroQuant and 2131 Table 1: Wilcoxon rank sum test results—both readers combineda Table 1: Wilcoxon rank sum test results—both readers combined
Feature
SOC vs FAST
FAST-DL vs SOC
FAST-DL vs FAST
Mean
P Value
Mean
P Value
Mean
P Value
Perceived SNR
4.1 (SD, 0.8)
,.001
3.5 (SD, 1.3)
,.001
4.3 (SD, 1.0)
,.001
Sharpness
4.5 (SD, 0.7)
,.001
3.5 (SD, 1.5)
.005
4.7 (SD, 0.9)
,.001
Artifacts
3.9 (SD, 0.8)
,.001
3.5 (SD, 1.1)
,.001
4.1 (SD, 0.9)
,.001
Anatomic/lesion conspicuity
4.2 (SD, 0.7)
,.001
3.3 (SD, 1.1)
.006
4.3 (SD, 0.7)
,.001
Image contrast
4.0 (SD, 0.7)
,.001
3.4 (SD, 1.1)
.004
4.1 (SD, 0.8)
,.001
GM/WM differentiation
4.3 (SD, 0.7)
,.001
3.4 (SD, 1.2)
.009
4.5 (SD, 0.8)
,.001
a SOC is superior to FAST for all criteria (P values ,.001). Numbers higher than 3 represent preference for the first of the 2 sequences listed in the upper row. FAST-DL is
superior to SOC for all criteria (P values , .008), except for GM/WM differentiation. While this metric trended to be superior for FAST-DL versus SOC, it did not reach
statistical significance after Bonferroni correction (P ¼ .009). FAST-DL is superior to FAST for all criteria (P values ,.001). Table 2: Paired t test for SOC versus FAST-DLa
SOC
FAST-DL
Paired
t Test
HOC (mean)
0.68 (SD, 0.16)
0.68 (SD, 0.16)
0.58
HV (mean) (cm3)
6.45 (SD, 1.70)
6.47 (SD, 1.69)
0.77
SLV volume (mean)
(cm3)
44.30 (SD, 20.60)
43.63 (SD, 20.35)
,0.05
ILV volume (mean)
(cm3)
3.07 (SD, 1.78)
3.04 (SD, 1.69)
0.27
a There is excellent agreement between SOC and FAST-DL for quantitative assess-
ment of HOC, HV, SLV volume, and ILV volume. Table 2: Paired t test for SOC versus FAST-DLa interrater reliability analysis was performed using an equal-
spacing weighted Cohen k statistic to measure the consistency
of the 2 readers’ evaluation of image quality. Overall Diagnostic Quality All SOC and FAST-DL datasets were rated of diagnostic quality by
both interpreting neuroradiologists. Both readers determined that
there was similar quality of segmentation for both the SOC and
FAST-DL datasets. Representative imaging examples for image- 2132 2132
Bash
Dec 2021
www.ajnr.org FIG 1. Linear regression results for SOC versus FAST-DL. The plot graphs demonstrate linear distribution without scatter, indicating consistent
concordance between SOC (x-axis) and FAST-DL (y-axis) in quantitative assessment of HOC (A), HV (B), SLV volume (C), and ILV volume (A). FIG 1. Linear regression results for SOC versus FAST-DL. The plot graphs demonstrate linear distribution without scatter, indicating consistent
concordance between SOC (x-axis) and FAST-DL (y-axis) in quantitative assessment of HOC (A), HV (B), SLV volume (C), and ILV volume (A). FIG 2. Linear regression results for SOC versus FAST. The plot graphs demonstrate a modestly linear distribution though some scatter is present,
indicating less optimal concordance of the cross-correlation factor between SOC (x-axis) and FAST (y-axis) (compared with SOC versus FAST-
DL) in a quantitative assessment of HOC (A), HV (B), SLV volume (C), and ILV volume (A). AJNR Am J Neuroradiol 42:2130–37
Dec 2021
www.ajnr.org
2133 FIG 2. Linear regression results for SOC versus FAST. The plot graphs demonstrate a modestly linear distribution though some scatter is present,
indicating less optimal concordance of the cross-correlation factor between SOC (x-axis) and FAST (y-axis) (compared with SOC versus FAST-
DL) in a quantitative assessment of HOC (A), HV (B), SLV volume (C), and ILV volume (A). AJNR Am J Neuroradiol 42:2130–37
Dec 2021
www.ajnr.org
2133 2133 FIG 3. Bland-Altman results for SOC versus FAST-DL. The plot graphs demonstrate a linear distribution without significant scatter, indicating
consistent concordance between SOC and FAST-DL in the quantitative assessment of HOC, HV, SLV volume, and ILV volume. FIG 3. Bland-Altman results for SOC versus FAST-DL. The plot graphs demonstrate a linear distribution without significant scatter, indicating
consistent concordance between SOC and FAST-DL in the quantitative assessment of HOC, HV, SLV volume, and ILV volume. FIG 4. Bland-Altman results for SOC versus FAST. The plot graphs demonstrate a modestly linear distribution though some scatter is present,
indicating less optimal concordance of the cross-correlation factor between SOC versus FAST (compared with SOC versus FAST-DL) in the
quantitative assessment of HOC, HV, SLV volume, and ILV volume. 2134
Bash
Dec 2021
www.ajnr.org FIG 4. Bland-Altman results for SOC versus FAST. Overall Diagnostic Quality The plot graphs demonstrate a modestly linear distribution though some scatter is present,
indicating less optimal concordance of the cross-correlation factor between SOC versus FAST (compared with SOC versus FAST-DL) in the
quantitative assessment of HOC, HV, SLV volume, and ILV volume. 2134
Bash
Dec 2021
www.ajnr.org 2134 pediatric brain disorders. The software
segments, labels, and calculates the vol-
umes
of
substructures
(including
lesions) in the brain. The derived quan-
titative values are compared with a
large normative age and sex-matched
data base, aiding in the diagnosis and
longitudinal follow-up of clinical con-
ditions such as Alzheimer’s disease. Quantitative assessment reduces reader
subjectivity.9-12 FIG 5. Representative axial 3D T1-weighted images on a 3T scanner. Left to right, SOC (scan time, 4
minutes, 55 seconds), FAST (scan time, 2 minutes, 10 seconds), FAST-DL (scan time, 2 minutes, 10
seconds). MR imaging provides excellent ana-
tomic detail and superb contrast reso-
lution but involves trade-offs in SNR,
spatial resolution, and scan duration.7
While DL-based augmentation is a rec-
ognized solution for accelerated MR
imaging, it is important to validate the
reliability and generalizability of its
enhancement capabilities with quanti-
tative biomarker accuracy.7 FIG 5. Representative axial 3D T1-weighted images on a 3T scanner. Left to right, SOC (scan time, 4
minutes, 55 seconds), FAST (scan time, 2 minutes, 10 seconds), FAST-DL (scan time, 2 minutes, 10
seconds). FIG 6. Representative 3D T1-weighted multiplanar images with volumetric segmentation on a 3T
scanner. Left to right, Axial, coronal, sagittal T1-weighted images with SOC (scan time, 5 minutes,
01 second) on the upper row (A) and FAST-DL (scan time, 2 minutes, 37 seconds) on lower row (B). The MR imaging experience is
uncomfortable
and
associated
with
frank anxiety reactions in up to 30% of
patients.13,14 Faster MR image acquisi-
tion can thus increase patient satisfac-
tion and may reduce motion artifacts. Motion is a significant challenge in MR
imaging, occurring in 29% of inpatient/
emergency department examinations
and 7% of outpatient studies and can
lead to repeat portions of or even com-
plete examinations.15 Andre et al16
found that 19.8% of all MR imaging
sequences need to be repeated due to
motion artifacts, a $592 revenue loss per
hour and $115,000 loss annually per
scanner due to motion artifacts. DL-
based reconstruction solutions promise
to enable shorter examinations with
decreased patient motion and improved
patient comfort.14 FIG 6. Overall Diagnostic Quality Representative 3D T1-weighted multiplanar images with volumetric segmentation on a 3T
scanner. Left to right, Axial, coronal, sagittal T1-weighted images with SOC (scan time, 5 minutes,
01 second) on the upper row (A) and FAST-DL (scan time, 2 minutes, 37 seconds) on lower row (B). quality analysis for SOC, FAST, and FAST-DL are shown in Fig 5. Sample volumetric segmentation of the SOC and FAST-DL images
are shown in Fig 6. In our study, we achieved a scan-time reduction of 60% while
exceeding perceived routine 3D T1-weighted image quality. If DL-
enhanced fast protocols were used on all pulse sequences across ev-
ery study, one could anticipate a proportional increase in examina-
tion-based workflow efficiency for an imaging facility. One recent
trial explored DL enhancement across all pulse sequences in clinical
spine MR imaging, with preservation of quantitative features using
a structural similarity index measure as well as gains in perceived
SNR and artifact reduction, despite a 40% scan-time reduction.2
Future research could explore whether scan-time reduction of this
scale results in a true-positive impact on workflow, eg, the ability to
scan more patients per day. AJNR Am J Neuroradiol 42:2130–37
Dec 2021
www.ajnr.org
2 CONCLUSIONS DL can enable 60% faster brain MR image acquisitions with
matched clinical disease status predictability and statistically
superior perceived image quality, while maintaining high quanti-
tative accuracy compared with the longer SOC examinations. This trial supports the reliability, efficiency, and utility of DL-
based enhancement for quantitative imaging. Shorter scan times
may boost the use of volumetric quantitative MR imaging in rou-
tine clinical settings. Blinded subjective assessment by the neuroradiologists found
that the 60% accelerated, DL-enhanced 3D T1-weighted brain MR
images delivered consistent clinical classification and were superior
to standard of care MR imaging across essentially all analyzed qual-
ity features (perceived SNR, perceived spatial resolution, artifact
reduction, anatomic/lesion conspicuity, and image contrast). These findings offer confidence that DL processing can add value
and efficiency to clinical MR imaging brain examinations. Disclosures: Suzie Bash—UNRELATED: Consultancy: Cortechs.ai, icometrix, Subtle
Medical, Guerbet, GE Pharmaceuticals, Comments: 1) honoraria in the past for con-
sultancy work for each entity listed, and 2)
Medical Advisory Board member for
Cortechs.ai (stipend and stock options) and icometrix (stipend only); Payment for
Lectures Including Service on Speakers Bureaus: icometrix, Subtle Medical, Guerbet,
Comments: Honoraria for webinars/talks for the entities listed; Stock/Stock
Options: Cortechs.ai, Comments: stock options as member of MAB; Travel/
Accommodations/Meeting Expenses Unrelated To Activities Listed: Cortechs.ai, GE
Pharmaceuticals, Comments: travel/accommodations for consultancy work. Greg
Zaharchuk—UNRELATED: Board Membership: Subtle Medical*; Grants/Grants
Pending: various National Institutes of Health projects, GE Healthcare, Bayer
Healthcare*; Patents (Planned, Pending or Issued): assorted patents on deep learn-
ing; Royalties: Cambridge University Press; Stock/Stock Options: equity in Subtle
Medical. Ajit Shankaranarayanan—UNRELATED: Employment: Subtle Medical; Stock/
Stock Options: Subtle Medical, Comments: equity in the company. Lawrence
Tanenbaum—UNRELATED: Consultancy: Subtle Medical, GE Healthcare, Philips
Healthcare, Hitachi, Siemens, Fuji Healthcare, Bracco Diagnostics; Payment for
Lectures Including Service on Speakers Bureaus: GE Healthcare, Fuji, Subtle Medical,
Guerbet, Bracco, Philips Healthcare, Money paid to the individual; Stock Options:
Subtle Medical; Other: Aidoc, Enlitic, Agamon, iMedis, Comments: consultant. Long
Wang—UNRELATED: Employment: Subtle Medical. Enhao Gong—UNRELATED:
Board Membership: Subtle Medical; Employment: Subtle Medical; Stock/Stock
Options: Subtle Medical. *Money paid to the institution. The quantitative biomarkers of HV, HOC, SLV volume, and
ILV volume were statistically equivalent for the FAST-DL
sequences and the SOC, supporting the absence of corruption by
DL processing and demonstrating the robustness of the DL tool
in maintaining quantitative integrity and enhancing image qual-
ity despite significant scan-time acceleration. CONCLUSIONS Not unexpectedly,
the cross-correlation factor was inferior for the SOC versus the
lower resolution FAST dataset. The strengths of this study include the use of a prospective,
randomized, multicenter, multireader study design with images
obtained on magnets from multiple vendors and of variable ages
and field strengths, with preserved accuracy of quantitative volu-
metric measures and clinical predictive categories. The results of
this trial support the use of DL enhancement to shorten clinical
MR imaging brain examinations, even when additional quantita-
tive tools such as volumetric analysis are applied. DISCUSSION DL enhancement of MR images is known to provide multiple
benefits, including increased SNR,2,3 but questions remain about
reliability in general clinical use.8 The approach used in our mul-
ticenter, multivendor study explored the impact on the image
quality and consistency of quantitative volumetric analysis results
obtained with FAST-DL compared with that obtained with SOC
scans. Quantitative volumetric MR imaging analytical tools are in
widespread clinical use for the evaluation of patients with demen-
tia, seizures, multiple sclerosis, traumatic brain injury, and In this trial, the SOC images serve as the standard for image
preference. Our randomized blinded assessment of the imaging
features is meant to reflect human subjective perception 2135 of comparative image quality. A radiologist’s qualitative assess-
ment of noninferiority is critical before a DL-enhanced alterna-
tive would be considered acceptable for clinical use. On the other
hand, processed images should satisfy both qualitative and quan-
titative measures to ensure that diagnostically relevant features
are not altered and the integrity of the processed image informa-
tion is maintained. though preliminary data suggest observed gains in acceleration
and perceived quality with other DL-enhancement tools as
well.17,18 However, this is the first study that the authors are
aware of that specifically confirms the quantitative volumetric ac-
curacy and consistent clinical disease categorization of the DL-
enhanced dataset. Future investigations might explore different methods and
tools. Another area of future research might include a similar
methodology applied to different clinical scenarios that demand
accurate segmentation but where scan time acceleration would be
desirable, such as in patients with multiple sclerosis, intracranial
metastases, epilepsy, and traumatic brain injury. Follow-up stud-
ies could also assess whether the difference between the FAST
and FAST-DL datasets was significant enough to impact the cor-
rect clinical diagnosis or alter the reader’s ability to detect a
lesion. Concerns exist about DL postprocessing introducing instabil-
ities in an image, in which tiny perturbations in the sampling do-
main have been shown to be capable of translating into noticeable
artifacts on the reconstructed image.8 This issue has been shown
for highly-contrived noise additions to k-space data, but it is
unclear whether such effects occur under normal operating condi-
tions. The current method starts from image-based DICOM data
rather than within the k-space and is likely less susceptible to this
effect. DISCUSSION To our knowledge, this is the first prospective, randomized,
multicenter study of DL reconstruction capabilities assessing the
impact on the integrity of quantitative volumetric analysis of clin-
ical brain MR imaging examinations. The DL tool applied in this
study shifts the usual MR imaging trade-off equation by imprint-
ing a boost in spatial resolution on the target FAST series, which,
due to inherently larger native voxel sizes, can have a higher SNR
and contrast-to-noise ratio than even the basis SOC series.7
Along with sharpness enhancement, DL offers structure-preserv-
ing denoising, contributing to statistically significant gains in per-
ceived SNR compared with SOC. 6
Bash
Dec 2021
www.ajnr.org REFERENCES 1. Radlak K, Malinski L, Smolka B. Deep learning-based switching filter
for impulsive noise removal in color images. Sensors 2020;20:2782
CrossRef 1. Radlak K, Malinski L, Smolka B. Deep learning-based switching filter
for impulsive noise removal in color images. Sensors 2020;20:2782
CrossRef 2. Tanenbaum LN, Bash S, Johnson B, et al. Deep learning recon-
structed, 40% faster spine MR examinations match or exceed the
quality of standard of care exams.. In: Proceedings of the Annual
Meeting of the Radiological Society of North America; November 29 to
December 5, 2020; Virtual 2. Tanenbaum LN, Bash S, Johnson B, et al. Deep learning recon-
structed, 40% faster spine MR examinations match or exceed the
quality of standard of care exams.. In: Proceedings of the Annual
Meeting of the Radiological Society of North America; November 29 to
December 5, 2020; Virtual Weaknesses include the small number of imaging subjects
and the use of only a single DL and quantitative brain-analysis
tool. The DL-enhancement tool used in this study is a vendor-
agnostic, DICOM-based, commercially available solution. It is
possible that alternative vendor-specific, k-space-based, commer-
cially available DL enhancement tools may perform differently, 3. Tanenbaum LN, Bash S, Gibbs W, et al. CNN based deep learning
enhances brain 3D FLAIR perceived quality, SNR and resolution
at 30% less scan time. In: Proceedings of the Annual Meeting of the
American Society of Neuroradiology, May 30 to June 4, 2020; Virtual 2136 2136
Bash
Dec 2021
www.ajnr.org communities/Artificial-Intelligence/bringing-ai-and-genetics-together-
to-support-clinical-decisions. Accessed July 1, 2020 communities/Artificial-Intelligence/bringing-ai-and-genetics-together-
to-support-clinical-decisions. Accessed July 1, 2020 4. Tian C, Xu Y, Li Z, et al. Attention-guided CNN for image denois-
ing. Neural Netw 2020;124:117–29 CrossRef Medline 5. Zhang K, Zuo W, Chen Y, et al. Beyond a Gaussian denoiser: resid-
ual learning of deep CNN for image denoising. IEEE Trans Image
Process 2017;26:3142–55 CrossRef Medline 13. Melendez JC, McCrank E. Anxiety-related reactions associated with
magnetic resonance imaging examinations. JAMA 1993;270:745–47
CrossRef Medline 6. Lunderwold AS, Lunderwold A. An overview of deep learning in
medical imaging focusing on MRI. Z Med Phys 2019;29:102–27
CrossRef Medline 14. Tanenbaum L. Quality, efficiency and survival with patient centric
imaging. In: Proceedings of the 7th Snowmass 2019: Hot Topics in
Radiology: Advanced Applications and Artificial Intelligence, Snowmass,
Colorado; February 10–15, 2019 7. Chaudhari A, Zhongnan F, Lee J, et al. Deep learning super-resolu-
tion enables rapid simultaneous morphological and quantitative
magnetic resonance imaging. AJNR Am J Neuroradiol 42:2130–37
Dec 2021
www.ajnr.org
2 REFERENCES In: Proceeding of Machine Learning for
Medical Image Reconstruction Workshop at MICCA, Granada, Spain;
September 16, 2018 15. Zaitsev M, Maclaren J, Herbst M. Motion artifacts in MRI: a com-
plex problem with many partial solutions. J Magn Reson Imaging
2015;42:887–901 CrossRef Medline 16. Andre J, Bresnahan B, Mossa-Basha M, et al. Toward quantifying
the prevalence, severity, and cost associated with patient motion
during clinical MR examinations. J Am Coll Radiol 2015;12:689–95
CrossRef Medline 8. Antun V, Renna F, Poon C, et al. On instabilities of deep learning
in image reconstruction and the potential costs of AI. Proc Natl
Acad Sci U S A 2020;117:30088–95 CrossRef Medline 9. Bash S. Eye on AI: enhancing neuroimaging with artificial intelli-
gence. Appl Radiol 2020;49:20–21 17. Bash S, Thomas M, Fung M, et al. K-space based deep learning
reconstruction empowers 60-70% acceleration of MR imaging
of the spine. In: Proceedings of the Annual Meeting of the
Radiological
Society
of
North
America,
Chicago,
Illinois;
December 1–6, 2019 10. Bryant M. Eye on AI: the potential and reality of AI in clinical
application. Appl Radiol 2020;49:10–11 11. Tanenbaum LN, Bash S, Davis M. Appl Radiol (AR Connect Expert
Discussions); AI: clinical applications. In: Proceedings of the Annual
Meeting of the Radiological Society of North America, Chicago,
Illinois; December 1–6, 2019 18. Tanenbaum LN, Bash S, Thomas M, et al. K-space based deep learning
reconstruction empowers 50% acceleration of MR spine imaging: a
prospective, multicenter, multireader trial. In: Proceedings of the
European Congress of Radiology, February 26 to March 1, 2020; Virtual 12. Bryant M. Bringing AI and genetics together to support clinical
decisions. Appl Radiol July 1, 2020. https://appliedradiology.com/ AJNR Am J Neuroradiol 42:2130–37
Dec 2021
www.ajnr.org
2137 2137
|
https://openalex.org/W2784641949
|
https://jyx.jyu.fi/bitstream/123456789/57150/1/aaaj0320162449.pdf
|
English
| null |
Blame game or dialogue? Financial, professional and democratic accountabilities in a complex health care setting
|
Accounting, auditing & accountability/Accounting auditing & accountability
| 2,018
|
public-domain
| 16,465
|
This is an electronic reprint of the original article.
This reprint may differ from the original in pagination and typographic detail.
Author(s):
Title:
Year:
Version:
Please cite the original version:
Blame game or dialogue? Financial, professional and democratic accountabilities in a
complex health care setting
Pellinen, Jukka; Mättö, Toni; Sippola, Kari; Rautiainen, Antti
Pellinen, J., Mättö, T., Sippola, K., & Rautiainen, A. (2018). Blame game or dialogue?
Financial, professional and democratic accountabilities in a complex health care
setting. Accounting, Auditing and Accountability Journal, 31(2), 626-650.
https://doi.org/10.1108/AAAJ-03-2016-2449
2018 This is an electronic reprint of the original article. This reprint may differ from the original in pagination and typographic detail. Author(s):
Title:
Year:
Version:
Please cite the original version:
Blame game or dialogue? Financial, professional and democratic accountabilities in a
complex health care setting
Pellinen, Jukka; Mättö, Toni; Sippola, Kari; Rautiainen, Antti
Pellinen, J., Mättö, T., Sippola, K., & Rautiainen, A. (2018). Blame game or dialogue? Financial, professional and democratic accountabilities in a complex health care
setting. Accounting, Auditing and Accountability Journal, 31(2), 626-650. https://doi.org/10.1108/AAAJ-03-2016-2449
2018 Blame game or dialogue? Financial, professional and democratic accountabilities in a complex health
care setting Jukka O Pellinen, Toni Mättö, Kari Sippola, Antti Ilmari Rautiainen, This is an electronic reprint of the original article.
This reprint may differ from the original in pagination and typographic detail. This is an electronic reprint of the original article. This reprint may differ from the original in pagination and typographic detail. Author(s): Pellinen, Jukka; Mättö, Toni; Sippola, Kari; Rautiainen, Antti Please cite the original version:
Pellinen, J., Mättö, T., Sippola, K., & Rautiainen, A. (2018). Blame game or dialogue? Financial, professional and democratic accountabilities in a complex health care
setting. Accounting, Auditing and Accountability Journal, 31(2), 626-650. https://doi.org/10.1108/AAAJ-03-2016-2449 the original version:
Pellinen, J., Mättö, T., Sippola, K., & Rautiainen, A. (2018). Blame game or dialogue? Financial, professional and democratic accountabilities in a complex health care
setting. Accounting, Auditing and Accountability Journal, 31(2), 626-650. https://doi.org/10.1108/AAAJ-03-2016-2449 All material supplied via JYX is protected by copyright and other intellectual property rights, and
duplication or sale of all or part of any of the repository collections is not permitted, except that
material may be duplicated by you for your research use or educational purposes in electronic or
print form. You must obtain permission for any other use. Electronic or print copies may not be
offered, whether for sale or otherwise to anyone who is not an authorised user. For Authors If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
information about how to choose which publication to write for and submission guidelines are available for all. Please
visit www.emeraldinsight.com/authors for more information. Accounting, Auditing & Accountability Journal Accounting, Auditing & Accountability Journal
Blame game or dialogue? Financial, professional and democratic accountabilities in a complex health
care setting Accounting, Auditing & Accountability Journal Article information: To cite this document: To cite this document:
Jukka O Pellinen, Toni Mättö, Kari Sippola, Antti Ilmari Rautiainen, "Blame game or dialogue? Financial, professional
and democratic accountabilities in a complex health care setting", Accounting, Auditing & Accountability Journal, https://
doi.org/10.1108/AAAJ-03-2016-2449
Permanent link to this document: References: this document contains references to 0 other documents. To copy this document: permissions@emeraldinsight.com To copy this document: permissions@emeraldinsight.com Downloaded by UNIVERSITY OF JYVASKYLA At 02:16 19 January 2018 (PT) Access to this document was granted through an Emerald subscription provided by emerald-srm:306933 []
F
A th About Emerald www.emeraldinsight.com Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services. Downloaded by UNIVERSITY OF JYVASKYLA At 02 Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication
Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation. *Related content and download information correct at time of download. 1. Introduction In recent decades, public governance reforms have largely been analyzed through the
concepts of the public financial initiative and new public management, where the central idea
is to extend the use of management accounting practices applied by firms to public services
and governance (Lapsley, 2008; Lapsley et al., 2003; Kurunmäki and Miller, 2011). Skepticism about the effects of these changes on accountability practices is clearly evident in
the public-sector accounting research (Chua, 1995; Cochrane, 1993; Cordery et al., 2010;
Ezzamel and Willmott, 1993; Gray and Jenkins, 1993; Humphrey et al., 1993). While this has
led to a widespread recognition of the shortcomings of performance measures in portraying
accountability (e.g., Messner, 2009; Roberts, 2009), there is little research that examines the
interplay of the different forms of accountability, such as financial, professional, and
democratic forms of accountability, in complex public organizations. This study aims to narrow this research gap by examining the evolving nature of the
accountability relations in a municipal health care organization in Finland. There has been a
trend since the 1990s in the organization of public services to move from vertically governed
and hierarchical state entities toward the co-ordination of networks, and this trend has been
widely analyzed in policy studies (Papadopoulos, 2003; Pierre, 2000; Rhodes, 1997, 2007). However, network organization forms that are integral to new public governance affect the
practices of financial, professional, and democratic accountability (Almquist et al., 2013;
Hodges, 2012). According to Hodges (2012, p. 42) “Changing political and economic
structures of the public services have led to increased complexity in defining the scope and
nature of both vertical and horizontal accountabilities.” There are only a few studies in
accounting of such network governance contexts (Cordery et al. 2010; Kurunmäki and Miller,
2011). Earlier literature suggests that changes in governance, such as those from bureaucratic
models to network co-ordination approaches, can weaken democratic accountability (Pierre,
2000; Rhodes, 1997). Rhodes (1997; also 2007) predicts that the complexity of current
public-sector organizations erodes accountability because the institutional complexity of
network structures obscures who is accountable to whom and for what. At its worst, this leads
to growing governance problems that limit the ability of public-sector managers to comply
with multiple accountability demands, and thus act effectively. Further, complex
organizational forms may promote sub-optimization, individually varied understandings of
responsibility or democratic accountability, encourage avoiding responsibility, and present
ample opportunities to shift blame between stakeholders (Hood, 2002, 2011). Abstract Purpose: This article investigates how the complexity of the network governance setting
affects accountability practices. We pay particular attention to the organizational
characteristics that may enable a common understanding of multiple accountability
relationships, or lead to problems in reconciling competing forms of accountability, thereby
appearing as blame game type behavior. Design: We conducted a case study with 31 semi-structured interviews in a Finnish health
care organization (FHC) that offers basic public health care services. The organization
represents a co-operative arrangement with the main city and three smaller municipalities. The FHC has faced difficulties in balancing budget constraints with the provision of statutory
care to citizens. This case is analyzed with the help of theories relating to accountability, the
blame game, and dialogue. Findings: We found that in the FHC operating under austerity constraints, attempts to
reconcile financial, professional, and democratic accountability were made but, instead of
dialogue and consensus, the different stakeholder groups resorted to defensive tactics in order
to protect their resources, position, or sense of professional obligation. We suggest that in a
context of network governance, accompanied by an increasing emphasis on financial
accountability, organizational practices are susceptible to conflicting accountabilities and
behavior characterized in this paper as a blame game. Originality/value: The study contributes to the empirical studies on accountability in the
new public governance context by analyzing the complex accountability relations between
stakeholder groups with different agendas. We suggest organizational characteristics that may
exacerbate conflicts between different stakeholder groups and prevent constructive dialogue. Furthermore, the study analyzes the composition of democratic accountability within the
studied organization. Keywords – Case study, accountability, responsibility, health care, blame game, network governance 1 1. Introduction However, a reform involving moving from bureaucratic governmental entities to network
organizations may also involve an attempt to move beyond narrowly understood (e.g.,
financial or managerial) accountability relations toward widely understood and balanced
accountability (or holistic accountability, see Cordery et al., 2010) to multiple stakeholders. Kurunmäki and Miller (2011) suggest that management accounting can facilitate integration
between network partners. Typically, however, the related requirements for creating
commonly shared values, open dialogue, an estimation of the impact on the environment, and
a consideration of a range of stakeholders other than hierarchical superiors have not been 2 realized (O’Dwyer and Unerman, 2007; Cordery et al., 2010). Although previous studies in
accounting have conceptualized different forms of accountability (Sinclair, 1995; Ahrens,
1997; Roberts, 2009;) or noted potential conflicts between them (Ahrens and Chapman, 2002;
Kurunmäki, 2004; Messner, 2009; Mutiganda, 2013), there remains a need to understand how
an increase in organizational complexity affects accountability relations, and how the
different forms of accountability can be reconciled. A distinctive feature of the healthcare sector is that the services are in the hands of strong
professional bodies. Doctors and nurses have considerable autonomy in terms of the work
processes governing caring and healing. Increasing demands for financial accountability in
the context of health care could give rise to conflicts with professional accountability. The
possibility of such conflicts has led to social practices that can sometimes seem irrational, and
apparently motivated by the protection of the status and autonomy of the ruling profession
within the organization (Chua, 1995; Kurunmäki, 2004; Llewellyn, 1998; Mutiganda, 2013). In addition, the possibility of ethical problems concerning the use of performance measures in
enforcing managerialism and financial accountability in public-sector organizations has been
recognized (Messner, 2009; Roberts, 2009). We conducted our case study in a Finnish health care organization (FHC) where four
municipalities (main city and three smaller municipalities) co-operate to deliver basic health
services in the region. The FHC operating context has elements of both hierarchy and
network, the network elements including the four controlling municipalities, special health
care units in the area, a city hospital, and other city health care units (see Appendix 1). The
governance of the FHC involves the city taking the lead over the three smaller municipalities
in the region. The governance structure is justified by the promise of better services and
lower costs. 1. Introduction The city population and service needs have been growing for many years but, at
the same time, the city has become indebted and now struggles to balance its budgets. Therefore, the management context of the health center is defined by the economic austerity
policy of the city and the contradictions between financial, professional, and political
objectives and the related accountabilities coming to a head. The case analysis focuses particularly on the differences between financial, professional, and
democratic accountability relations and also on the potential outcomes of emphasizing these
relations differently within the case organization. We examine the different forms of
accountability in the FHC operating under network governance. We aim to analyze the
multitude of accountability practices in the field, the points of conflict between them, as well
as how they might be reconciled (e.g., through a blame game or with dialogue). Our research
questions are: What are the different accountabilities and how are they reconciled in a public health care
organization? Why does a given form of reconciliation take place? 3 How does the complexity of a network health care setting affect accountability practices? How does the complexity of a network health care setting affect accountability practices? The paper is structured as follows. The following section provides an overview of the
concepts of responsibility, accountability, and the different strategies for conducting the
blame game as a result of reacting to the interplay of different accountabilities in a complex
health care setting. The next section presents the case and the methods of data collection. The
paper then presents the complex situation between the different forms of accountability and
seeks to analyze different practices of reconciliation through the empirical data, before
discussing the findings and presenting the conclusions. Multiple accountabilities and responsibility Although accountability expectations are
often presented through performance indicators, several researchers have warned about over-
emphasizing accountability through a limited set of indicators as doing so obscures the actual
complexity of organizational, socially constructed reality (Cooper and Johnston, 2012;
Roberts, 2009). There is a large body of literature addressing problems generated by an excessive focus on
managerial and measurable forms of accountability. Cooper and Johnston (2012) note that the
mere provision of information is not accountability, and how accountability also encompasses
decisions with an impact on oneself and others. Roberts (2003) differentiates between the
hierarchical and socializing forms of accountability. Hierarchical forms of accountability
involve achieving certain norms and standards provided by, for example, accounting. In
contrast, the socializing form of accountability is more concerned with the recognition of the
self and others, particularly how somebody appears to be responsible. In his later work,
Roberts (2009) argues that the increased focus on transparency, and the resulting focus on
performance based accountability, could make blame avoidance (see Hood 2011) a rational
choice on the individual level. In Roberts’ (2009) view, blame gaming stems from the
recognition of the negative organizational consequence of over-zealous pursuit of
transparency. Another, psychological reason to resort to shifting the blame is presented by
Cooper and Johnston (2012): When faced with bad results or failure to meet a set ideal, a
person could turn his/her self-criticism into a need to find others on whom to project all that
is inadequate and which violates the sense of personal perfection. In this paper, the term professional accountability refers to work-related self-control and
professional expertise as a foundation for both individual decisions and organizational
effectiveness (Romzek and Dubnick, 1987; Sinclair, 1995). Professional accountability also
touches upon internal aspects of responsibility, in that it involves the professional’s ethical
working principles and personal feeling of liability and responsibility toward those he or she
works for, or with. For example, in the highly specialized and independent medical
profession, this sense of responsibility might be directed toward patients, peers, or clients and
there could be conflict between the accountability to the profession and the lay
representatives of the public, for example, regarding resource consumption over time. Because of the potential conflicts in external accountabilities, many such professionals feel
primarily accountable to themselves, to their personal values, and to their conscience (Day
and Klein, 1987). However, Romzek and Dubnick (1987, p. Multiple accountabilities and responsibility Responsibility and accountability are ethically charged concepts that have often been used
interchangeably. According to Mulgan (2000), however, responsibility mainly refers to the
internal and personal aspects of moral obligation whereas accountability focuses on external
justifications, sanctions, and liabilities to administrative and political superiors. According to Butler (2005), an individual’s perception of responsibility and the inner self is
constructed through dialogue with other individuals: “by virtue of the relation to the Other
that is established at the level of my primary and irreversible susceptibility” (Butler 2005, p. 88). Further, the different perceptions of responsibility among individuals and groups lead a
society to be influenced by several types of ethos. If one group claims their ethos is
universally right, social intercourse between the groups can lead to moral conflicts and
“ethical violence” (Butler 2005, p. 5). moral questions arise only when the collective ethos has ceased to hold sway. …
although the collective ethos is no longer shared … it can impose its claim to
commonality only through violent means. (Butler 2005, p. 4) Bovens (1998) raises the notion of the “active form of responsibility” where an individual
acts as both a manager or employee and a citizen in parallel. Accordingly, the individual
might be placed in a conflicting situation if a professional accountability for outcomes
conflicts with the personal notion of responsibility toward conscience, peers, and citizens
(pp.149–164). In such a situation, choosing a legitimate course of action can be difficult. However, loyalty to one’s superiors could be considered a primary virtue since without it
organizations would falter in accomplishing their goals. According to Bovens (1998), an
individual could actively respond to such conflicting situations by resigning, refusing to carry
out instructions, whistleblowing or “tinkering with the structure” (pp. 176–220), thereby
relieving personal feelings of guilt or shifting blame for poor outcomes (see also Cooper and
Johnston 2012; Hood, 2002). 4 Bovens (1998, 2005) characterizes accountability as “a social relationship in which an actor
feels an obligation to explain and to justify his or her conduct to some significant other” and
Borowiak (2011) uses the term punishability to refer to the extent that the blame and
sanctions can be placed for one’s behavior and its effects. Therefore, punishability is
associated with a particular form of accountability if sanctions are triggered by a failure to
meet the expectations of the accountability holder. Multiple accountabilities and responsibility 229) note: …public officials must rely on skilled and expert employees to provide appropriate
solutions. Those employees expect to be held fully accountable for their actions and 5 insist that agency leaders trust them to do the best job possible. If they fail to meet job
performance expectations, it is assumed that they can be reprimanded or fired. insist that agency leaders trust them to do the best job possible. If they fail to meet job
performance expectations, it is assumed that they can be reprimanded or fired. Accountability is also context and organization specific (Roberts and Scapens, 1985; Romzek
and Dubnick, 1987). In the public sector, accountability essentially concerns the relationship
between citizens and politicians, and, within the hierarchy of a public organization, between
politicians and public managers (Barberis, 1998; Mulgan, 2000). According to Borowiak
(2011), the primary function of democratic accountability is to provide a way for citizens “to
exert discipline and control over governing bodies” (p.4). Prerequisites for democratic
accountability are “the ultimate authority of a bounded political community of citizens” and
institutions through which to operate. Thus, deficits in democratic accountability are caused
either by the lack of citizen empowerment or poorly designed accountability mechanisms (see
Borowiak, 2011, p. 4). Smyth (2012) notes that in such situations, citizens sometimes seek to
challenge the predominant form of accountability by active campaigning against
institutionalized practices. Brown (2009) argues that if democracy is understood as ultimately
empowering citizens, accountability designs based on dialogue and participatory forms of
control have the capability to further the goal of democracy. If accountability is separated
from the power of decision, it becomes empty (Borowiak, 2011; Cooper and Johnston, 2012;
Lowe et al., 2012). Accountability is also context and organization specific (Roberts and Scapens, 1985; Romzek
and Dubnick, 1987). In the public sector, accountability essentially concerns the relationship
between citizens and politicians, and, within the hierarchy of a public organization, between
politicians and public managers (Barberis, 1998; Mulgan, 2000). According to Borowiak
(2011), the primary function of democratic accountability is to provide a way for citizens “to
exert discipline and control over governing bodies” (p.4). Prerequisites for democratic
accountability are “the ultimate authority of a bounded political community of citizens” and
institutions through which to operate. Thus, deficits in democratic accountability are caused
either by the lack of citizen empowerment or poorly designed accountability mechanisms (see
Borowiak, 2011, p. 4). Multiple accountabilities and responsibility Smyth (2012) notes that in such situations, citizens sometimes seek to
challenge the predominant form of accountability by active campaigning against
institutionalized practices. Brown (2009) argues that if democracy is understood as ultimately
empowering citizens, accountability designs based on dialogue and participatory forms of
control have the capability to further the goal of democracy. If accountability is separated
from the power of decision, it becomes empty (Borowiak, 2011; Cooper and Johnston, 2012;
Lowe et al., 2012). Sinclair (1995) analyzed top municipal managers and found that accountability is
subjectively constructed, situation-specific, flexible, and a continuously redefined
phenomenon comprising five different types of accountability: the political, public,
managerial, professional, and personal. Recently, however, financial accountability has been
emphasized as a key managerial issue in public management (e.g., Lapsley, 2008). Previous
studies in accounting have largely concentrated on the contradictions between the financial
and professional forms of accountability. For example, according to Llewellyn (1998),
professionals in social care protect their autonomy at work by not delegating responsibility
for budgets. Further, Broadbent et al. (2001) noticed a tendency for professional
organizations to resist accounting and finance based changes. Hyvönen and Järvinen (2006)
found in their case that budgetary bias and old budgeting practices prevailed in hospital
budgeting, that is, they were incorporated into the new budgeting model. Mutiganda (2013)
noted that financial accountability clearly related to budgetary limits, while professional
accountability is spread among the medical top managers, doctors, and nurses, indicating a
decoupling of these two accountabilities. Kurunmäki (2004), however, found the integration
of accounting knowledge and medical knowledge among clinical professionals to have
positive effects. In order to achieve this integration, Goddard (2005) sees it as important to
change the governance mentality of public-sector officials from being conformance centric to
performance centric, and to develop budgeting practices (and associated accountability
mechanisms) accordingly. The contradiction between professional and financial accountability has been well established
in public-sector studies. Democratic accountability, however, deserves more research
attention especially when studying complex health care organizations under austerity 6 conditions. In one of the few studies to address democratic accountability at the municipality
level, Ahrens and Ferry (2015) studied the Newcastle City Council which delegated some
responsibility for the focus of budget cuts to stakeholder groups, thereby dispersing
accountability and reducing the perceived negative effects on service clients. 1 Hood (2011, p. 95) refers to institutional partnerships and other arrangements that make several, separate
organizations jointly responsible for a service as a “joined-up government”. These types of arrangements relate
to public organizations, although in some cases they extend to the business sector or other independent bodies.
Our case involves a health care organization working in a network governance context that Hood (2011) might
label a “joined-up” arrangement. Multiple accountabilities and responsibility Dispersing
accountability can also be seen as dividing, shifting, or sharing the potential blame for the
budget cuts among a wider range of actors (cf. Hood, 2011). Blame game and dialogue Policy strategies rely on procedures and office guidelines. They involve a “selection of
policies and operating routines that minimize institutional or individual blame” (Hood 2011,
p. 31). Policy or operational strategies include obscure program targets and processes as well
as the protocolization of practices so that exact, measurable results and individual blame can
be avoided, or hidden behind bureaucratic practices, ideals, and protocols. Examples include
formula-driven budget allocations or rigid protocols in place of independent professional
judgment. Referring to public-sector organizations or governmental bodies, Hood (2011) distinguishes
four types of individuals that either utilize blame avoidance strategies, or are the target of
those strategies. These are the top managers in public-sector organizations, public-facing
case-handling professionals, people hidden from the public eye such as accountants, middle
managers, and resource allocators and finally, the civil society consisting of those who
scrutinize the public services such as clients, voters, and patients. Hood (2011) does not state
explicitly where politicians fit into this four-level classification of actors, although he seems
to imply that they are part of the group of top managers. Owing in part to national differences
(i.e., in Finland, the city mayor is not a politician), in this paper, municipal politicians are
considered a special group that overlaps with civil society and the top management. While
politicians oversee the public organizations and demand value for money in their public
speeches, they also act in the bodies that provide the public organization with its budget and
operating guidelines. The four groups of actors mentioned have various opportunities to utilize the three different
strategies of blame avoidance (Hood, 2011). The top managers are able to utilize all three;
they can appoint the right people to positions where blame is likely to be attributed if things
go wrong. They make use of different forms of rhetoric when presenting themselves to the
public. Finally, they are also able to shape the organizational structures to allow blame to be
dispersed into complex organizational structures. Professionals, on the other hand, often
utilize the presentational blame avoidance strategy in a way that shifts the blame back
upwards; professionals may blame bad orders, flawed policies, time-consuming operating
systems, or inadequate resource allocations, for example. The third group, consisting of
middle level managers, financial administrators, and other supportive personnel, may find the
blame placed on them from both the top management and the professional level. Blame game and dialogue While public organizations are exposed to different initiatives targeting reform, transparency,
and clearer accountabilities (Lapsley, 2008; Kurunmäki, 2004), Hood (2011) notes that
pinpointing blame and finding the right people responsible in complex public organizations
can be difficult. Further, Hood (2011) states that people are, by nature, averse to blame and
often seek to avoid it to retain their chances of re-election, collecting bonuses, retaining a job,
or safeguarding operating conditions, or their reputation. This raises the issue of a blame
game, which involves a range of different actions and strategies utilized to avoid (public)
blame, often with relatively little concern for organizational efficiency (see Hood, 2002;
2011). Hood (2011) differentiates between three strategies that individuals, groups or organizations
use to shift blame: presentational strategies, agency strategies, and policy strategies. Together
these three strategies of blame avoidance constitute a set of actions that is called the blame
game, which involves various actors attempting to place the blame on some other person,
entity, or circumstance. Hood (2011) associates presentational blame avoidance strategies
with the rhetorical dimension of politics and management, and also to the correct framing of
arguments. One presentational strategy to avoid blame is to diminish the problem for which
one is potentially blamed by presenting it as a minor setback or short-term issue that can
deliver benefits in the long term. One might also present credible claims as to why a
particular officeholder is not to blame for the incident. Other presentational strategies include
deliberately inaccurate reporting, and making accounting disclosure opaque so that it is
difficult to make precise claims against any specific person. In contrast, agency strategies aim to promote blame avoidance by utilizing the responsibility
guidelines of the public organization. This involves the delegation of activities that will
attract blame and retaining activities likely to earn credit. Other examples of agency strategy
include working in groups in a way that diminishes the opportunity to assign blame to an
individual, or creating multi-agency arrangements in public services1 that make “shuffling
blame about” possible or make it “disappear” between the different organizational entities 7 7 (Hood 2011, p. 32). In addition, Hood (2011) considers the constant transfer of officeholders
or change in organizational arrangements to be a type of agency strategy designed to avoid
blame: By the time blame can be assigned to an entity or an individual, the subject in
question has already been replaced. Blame game and dialogue This,
however, also gives them the chance to shift blame in multiple directions, including sideways
if the organizational structure permits. When shifting blame upwards, they tend to utilize
agency strategy to associate top management with every decision that might carry the risk of
blame. When blame is to be shifted downwards to professionals, they utilize a policy strategy
and refer to operating guidelines and organizational rules to point out mistakes in the
professional’s work (Hood 2011). 8 Blame gaming behavior utilizes the inaccuracies of accountability design in a complex
organization (see Roberts, 2009). However, Gårseth-Nesbakk and Kjærland (2016) note that
resorting to blame game tactics may have unforeseen and counter-productive effects. They
suggest that once blame game is initiated, subsequent developments cannot be controlled due
to complexity of the context in which it is used. For example, there are limits to how much
blame can be shifted to others: even if responsibility were to be formally delegated, the voters
might still hold a key politician, for example, responsible for the outcome (Hood, 2011, 21). Shared understanding of accountability and blame sharing (delegating some but still
accepting part of the possible blame) can be prerequisites for constructive debate and
organizational development. Instead of actors utilizing different blame game strategies, an
organization might be able to concentrate on solving the problematic issues. Building shared
language and exchanging ideas across the parties to accountability relationships is called
intelligent accountability (Roberts, 2009) or dialogue (Messner, 2009). In this paper, we use
the latter term. Explaining and reasoning are central to dialogue, and to creating common
understanding of the opportunities and constraints affecting accountability relations. Dialogue
fosters aspects like a common vision of service production, encouraging individuals to
overcome situations characterized by competing accountabilities (Goretzki and Messner,
2016). This common vision also suggests co-operation, thereby sharing possible political
blame among, and possibly dispersing accountability to, a wider selection of actors (see
Ahrens and Ferry, 2015; Hood, 2008). In this paper, we combine the analysis of different
forms of accountability with the notions of the blame game and dialogue. 3. Methodology Within the context of a health care case organization working within a network setting, a field
study of accountability and accounting was conducted between February 2013 and March
2014 in the FHC providing basic health care services (Appendix 1) for about 150 000 people
living in the municipal area of a main city and three other smaller municipalities in Finland. The FHC has 12 health care units (clinics) within its area, most of which operate from 8 am
to 4 pm on weekdays, and it currently has a net budget (net expenses after fees from
customers) of about EUR 80 million and about 1,000 employees. The four municipalities
involved finance the FHC jointly according to a cost-sharing contract, but in practice the
FHC operates under the control of the main city organization. Elected political decision
makers are involved in the budgeting process through defining the budgetary limits for the
organization. We conducted 31 interviews between 2013 and 2014 (see Appendix 2). Further, we analyzed
accounting reports, budgets, internet reports, and local newspaper articles. We mainly used
qualitative data to obtain a broad appreciation of the case events, but also included some
quantitative data in accordance with case study principles (see Yin, 1984). The interviews
were semi-structured and conducted by the authors. Some of the interviews had two 9 interviewers or interviewees present and they often resembled an active discussion more than
a recording of precisely formulated questions and responses. These features permitted the
subject matter to be viewed from different angles and the avoidance of any single type of
vocabulary (Alvesson, 2003). Our research project started in early 2013 with the first set of interviews in the FHC
organization. The first nine interviewees in seven interviews were mostly managers and
people responsible for the financial administration at the FHC. These interviews provided an
understanding of the financial administration of the city’s health care services. Another
important aim was to focus our research questions on the most promising research
phenomena. We asked the interviewees questions on different control systems in use with the
financial administration, the performance measures used, the role of budgeting in health care,
and on the development needs of the financial administration of the health care system. Each
interview was transcribed and analyzed. Following the transcription, each researcher
reviewed the interview material and sorted the collection of remarks into themes. 3. Methodology These
themes were then compared to enhance the validity of the interpretations. Therefore, the
analysis of interview data included features of content analysis, such as categorization, and
ethnographic analysis that included trying to understand the socially constructed
organizational reality (see e.g., Silverman, 2001, pp. 122–129). A major theme raised in the nine interviews was the role of budgeting and the various issues
arising from it (e.g., difficulties with getting professional groups to take responsibility for the
budget, under budgeting resulting in part from ambiguous input from the health professionals,
and a lack of awareness of budgetary limits among some of the deputy chief physicians). The
analysis revealed the different accountabilities and the processes creating them to be a very
interesting research topic. The second set of interviews was conducted in the autumn of 2013 in order to extend our
view of the professional accountability perceptions and the discrepancies between them. In
research project meetings with the FHC representatives, three health clinics were selected
from the pool of 12 clinics in the area. Six people from each health clinic were interviewed
separately, including two nurses, two doctors, a head nurse, and the deputy chief physician
responsible for the health clinic. The interviews covered several issues, including views on
accountability and responsibility, the clinic’s targets, the measurement of those targets,
budgeting control, budgetary reporting, and the main problems and development needs of the
health clinic. Finally, in order to obtain a view of the financial and democratic accountability
developments, in a third round of interviews, six people from the city management, were
interviewed early in 2014. In addition to the interviews, the research group collected material such as reports, cost
estimates, the cost driver and cost allocation data used in the FHC, population data for the
city area, the budgets and economic forecasts for 2013, city strategy papers, and the number
of patient visits handled by doctors and nurses. The research group also collected local
newspaper articles on aspects of health care to provide internal background information, and 10 also contacted the Regional State Administrative Agency to obtain reports on the service
levels of the FHC. This article pursues data-driven theory refinement relating to the accountability literature
through case research findings. The validity of the research is increased through multiple
rounds of interviews focusing the research questions in addition to independent analysis
conducted by four researchers. 3. Methodology Further, the rich case description included aims to convey the
“convincingness” (Golden-Biddle and Locke, 1993) of the research. Convincingness is
established when research is able to convey that the researchers have been in the case
organization, understand the behavior of the actors, and can make plausible arguments based
on case data interpretations. Further, critical reflection upon the assumptions made in earlier
literature or by case actors helps avoid bias and to deliver contributions (Golden-Biddle and
Locke, 1993; McKinnon, 1988). Downloaded by UNIVERSITY OF JYVASKYLA At 02:16 19 January 2018 (PT) Case setting Finland’s 313 municipalities are responsible for organizing basic health care for the 5.5
million citizens of the country. The Constitution of Finland together with the Health Care Act
mandates equal quality and availability of basic health care regardless of the region where the
citizen lives. The regional state authorities monitor that the health services are meeting the
required standards of service delivery, and can fine municipalities failing to do so. However,
the exact organization of local health care provision varies in Finland; it may, for example, be
organized as a type of network organization together with other municipalities, or through
private corporations, or the municipality may decide to provide its health care by itself. Regarding the financial constraints on health services, the Finnish Municipality Act requires
that municipalities balance their finances or, in the worst case, a municipality can be deprived
of its autonomy and be compulsorily merged with another municipality. In an attempt to improve efficiency and deliver budget savings, the FHC was founded in 2010
with the agreement of four municipalities to create a health care organization for the
provision of basic health care services to the citizens of the region. The agreement required
the three municipalities to transfer the responsibilities relating to the administration of health
care services to the main city. As a result, the organizational structure became a complex
combination of 11 hierarchical levels of the city and a network of four municipalities. From
2011 onwards, employees from the different health care areas were transferred to the main
city’s payroll and absorbed into its administrative processes. The FHC organization is directed by the basic welfare board, which is an administrative
entity of the main city. A basic welfare board is a typical Finnish administrative arrangement,
where municipal politicians control the budgetary limits of a health care organization, make
decisions concerning the provision of health care services (such as changing the number of 11 clinics in the municipality’s area), and set the service guidelines for social and health care in
the area. In Finland, a basic welfare board is used as an administrative entity irrespective of
the governance structure of health care in the area (i.e., network or single municipality). The
basic welfare board in our case comprises 13 members from different political parties, all of
whom are from the main city. Case setting The members of the basic welfare board are chosen from
among the elected politicians on the municipal council; in Finland, municipal politicians
often have several roles on different boards in the city. In addition, an advisory board referred
to as a health care division operates under the basic welfare board, but has no real jurisdiction
over administrative matters. The purpose of the health care division is to gather the budget
proposals from the member municipalities and present them to the basic welfare board, which
then decides on the final budgetary proposal to the main city council. The health care division
comprises nine members, six of whom are political representatives of the main city, and three
of whom come from the other member municipalities. The health care division liaises
between the political administration of a health care organization, the city management, and
the clinical professionals. As the influence of the surrounding municipalities in the FHC is
limited to an advisory role in the health care division, this type of arrangement effectively
transfers both the decision-making power and the responsibility for service provision to the
main city. Although the main city retains the administrative rights for service provision and the
members on the basic welfare board, the joint agreement for health care service provision
initially contained a stipulation that no single municipality could change the co-ordination of
services, quality standards, the organization of activities, or regional health care service
requirements without the consent of the other member municipalities. This, however, slowed
down the administrative processes and ultimately resulted in a supplemental contract that
increased the role of the basic welfare board and the administrative management, thereby
strengthening the main city’s mandate for action. In Finland, a city mayor and deputy city mayor differ from politicians in that they are
appointed by the municipal council as the officials in charge of the city administration. As
such, they are not directly elected by voters, nor are they typically members of any political
boards in the municipality. The deputy city mayor responsible for the field of health care
supervises the implementation of the budgets and guidelines set by the basic welfare board. The city mayor is the highest-ranking official of the city management who directs the
municipality, supervises the finances, and has the power of decision granted by the municipal
council. Case setting The Finnish Municipality Act states the municipal council is responsible for setting an annual
budget for the next calendar year. The municipal council is also required to make and accept
a financial plan for a minimum three-year planning period. The budget sets the financial and
functional targets for the municipality. The municipality’s budget must be balanced (or be in
surplus) within a planning period (Finnish Municipality Act, ch.8, para. 65). The Finnish
Health Care Act includes some supplementary stipulations on health care responsibilities 12 (e.g., maximum patient waiting times). Regional authorities monitor the municipalities’
service provision accordingly and ensure that legislation is complied with. In the aftermath of the recent global financial crisis, marked by a growth in legislative
requirements for municipal services and simultaneous governmental savings, our case city
has had financial difficulties. In 2011, it recorded a deficit of about EUR 24 million, which
amounted to a little over three percent of all expenditure. In 2012, the city deficit almost
doubled to close to EUR 46 million, amounting to over six percent of all expenditure. In
2013, the deficit was about EUR 38 million, and during both 2014 and 2015, deficits of about
EUR 13 million were recorded. Cuts to the basic health care budget have also been made
during this period, creating further pressure on FHC employees. Multiple accountabilities in health care The city mayor emphasizes financial accountability. For example, the deputy city mayor in
charge of social and health affairs states: My boss measures my success only based on the fact of whether I stay on budget or not. If that budget overruns, it is a very bad situation. This is the fact. (Deputy city mayor) My boss measures my success only based on the fact of whether I stay on budget or not.
If that budget overruns, it is a very bad situation. This is the fact. (Deputy city mayor) In the FHC, as in many other health centers in Finland, the management of clinical
professions has been separated. This means that doctors and nurses each have their own
supervisors: for doctors, that means deputy chief physicians, and head nurses supervise
nurses. One of the results of this arrangement is that the management of personnel is
somewhat uncoordinated. Deputy chief physicians cannot directly manage nurses because
they report to head nurses. Head nurses are not accountable for the budget, while deputy chief
physicians ought to monitor the spending of their own clinic. However, the formal
accountability for the budget and the service process rests with the director of basic health
care services, who has not delegated budget accountability formally to the deputy chief
physicians of the 12 clinics. One interpretation is that the budget responsibility is deliberately
focused on one individual to enable the other members of the professional group to
concentrate on their clinical work. The responsibilities of the FHC’s nurses include taking phone calls from potential patients,
deciding on the right course of action (i.e., whether a patient sees a doctor, a nurse, or is
treated at home), seeing patients themselves, and recording their actions on various databases. These include the national database for health care services that monitors the number of
patient visits per clinic, the type of diagnosis made, treatment decisions and suchlike. The
responsibilities of the doctors include seeing patients, recording their actions on the same
databases as the nurses, being on call for emergencies, managing the ward, and performing
special functions assigned to them. These special functions vary, and examples include visits
to local schools, or providing an out-call service to designated nursing homes. 13 Doctors and politicians are trying to keep the service level high; while the doctors do not
usually concern themselves with costs of the services, politicians generally demand high
levels of service at minimum costs. The city management, on the other hand, tries to keep the
costs down. This generates a situation where accountabilities of the city management and the
clinical professionals conflict. Finally, another institution that affects the situation is the
Regional State Administrative Agency (RSAA) that monitors the legality of services and also
directs, licenses, and oversees healthcare. It enforces a required quality level for treatments
and monitors whether the various legislative metrics of the services are met. My boss measures my success only based on the fact of whether I stay on budget or not.
If that budget overruns, it is a very bad situation. This is the fact. (Deputy city mayor) For example, it
requires that non-urgent treatment is available for the patient within 30 days of contact. A
director commented on the situation: Many times it is us [clinical professionals] that are worried about legality of the services
and that patients get the care that is required. But then, it seems that city management
doesn’t care about that. They only care about services being provided cheaply. Also, it
seems that the agency here [RSAA] is much stricter than elsewhere. These regional
agencies seem to interpret the law differently; this further limits our options. (Director of
basic health care) While the day-to-day management of the city remains in the hands of non-politicians like the
mayor and the deputy city mayors, the situation is further complicated by many municipal
politicians having several different roles on the administrative boards of the city. These roles
may have differing responsibilities and different forms of accountability; for example, a
politician may drive budget cuts and cost savings in the city council, but as a member of the
basic welfare board, he/she may try to argue for budget increases for health care. While the day-to-day management of the city remains in the hands of non-politicians like the
mayor and the deputy city mayors, the situation is further complicated by many municipal
politicians having several different roles on the administrative boards of the city. These roles
may have differing responsibilities and different forms of accountability; for example, a
politician may drive budget cuts and cost savings in the city council, but as a member of the
basic welfare board, he/she may try to argue for budget increases for health care. Politicians in the city council and on the basic welfare board tend to think in terms of the
meeting they are participating in. This results in contradictory behavior. And there are
so many different agendas, I mean, it’s not only the party politics that determine
politicians’ opinions, it is also regional politics, so politicians may be inclined to prevent
the closure of a clinic in their constituency. (General health manager) Clinical work involves several responsibilities and associated reporting processes. Deputy
chief physicians hoped that the city administration would take political responsibility by
issuing a statement to clinics confirming the budget cuts and demands for savings. My boss measures my success only based on the fact of whether I stay on budget or not.
If that budget overruns, it is a very bad situation. This is the fact. (Deputy city mayor) This
would shift the blame away from the deputy chief physicians to the city administration and
politicians, thus constituting a type of blame sharing. Clinical work involves several responsibilities and associated reporting processes. Deputy
chief physicians hoped that the city administration would take political responsibility by
issuing a statement to clinics confirming the budget cuts and demands for savings. This
would shift the blame away from the deputy chief physicians to the city administration and
politicians, thus constituting a type of blame sharing. Personnel here seem to think that management of the clinic wants to pressure them as
tightly as possible. They keep saying that we’ve always had the extra funding when we
needed it, because we can’t stop the public service. So, they want to see some
representation from the city administration to come down here and confirm the crisis and
the budget cuts. (Head nurse) Clinical professionals reported an increase in performance measurement, mandatory
statistical reporting, and financial accountability at the cost of good patient care. Examples 14 include a new access control system with working time metrics, increased reporting
responsibility to outside organizations, administrative responsibilities, demands for cost-
cutting, and an increased emphasis on efficiency measurement. As a doctor, I find it hard to accept that reporting our work to outside organizations is
more important than finding the right treatment for a patient. It seems that nowadays,
everything is more important than the patient. (Deputy chief physician) A statutory care guarantee determines the content of the work of nurses, with the most
important performance metric being telephone answer times. These are monitored and if the
required performance level is not met, those responsible for not meeting the metric are
required to explain themselves to administration. It started from a simple thing: access control. Then came the phone call measurement. They want to know how quickly we answer the phone calls. Of course, we’ve always been
measured to some extent, but now it’s become so easy with computers and technical
systems. (Nurse) Modern patient work involves input into different clinical databases. This, in turn has enabled
the city management to push financial accountability in the form of various performance
measures to the clinical professionals irrespective of the budgetary control of individual
clinics. Conflicting accountabilities and reasons for blame games Conflicting accountabilities and reasons for blame games Clinical professionals see their main goal as being the welfare of their patients. They would
rather concentrate on the best possible care and seem to think that given the right volume of
resources, savings will appear eventually. The belief is based on the assumption that by
devoting sufficient resources to preventive care by treating the minor problems of today, and
reserving the time for lifestyle instruction sessions, major problems could be avoided in the
future. This highlights that financial or democratic accountability considerations can differ if
considered from a short-term or from a long-term perspective. However, with the current
pressure to deliver cost savings, preventive health services have been scaled back, prompting
arguments like these: If we could treat these patients like we want, if we had the time to really concentrate on
their situation, I know that in 10 or 15 years we would see the results and there would be
a real reduction in strokes and heart attacks. And treating a stroke: that really costs a
lot. But now, we are forced to downsize all the time, even to a foolish degree and I tell
you, in the future, we will see even higher costs. (Director of basic health care) If we could treat these patients like we want, if we had the time to really concentrate on
their situation, I know that in 10 or 15 years we would see the results and there would be
a real reduction in strokes and heart attacks. And treating a stroke: that really costs a
lot. But now, we are forced to downsize all the time, even to a foolish degree and I tell
you, in the future, we will see even higher costs. (Director of basic health care) A former general health manager who had been relocated before the interviews following
several budgetary issues seemed to have a similar view: 15 We are ignoring some stuff that has to be taken care of in the future. By saving in the
short term we are piling up stuff for the future. In a way, we are taking a loan that has to
be paid off some time. (Relocated general health manager) The situation between the city administration and clinical management is highlighted by the
fact that the 2013 regional basic health care budget was cut by 10 percent by the city
administration. Conflicting accountabilities and reasons for blame games The action created a gulf between the expected level of service and the
expected cost cuts. A head nurse commented: The budget is way too tight. The budget is even lower than last year’s. This really does
not seem good, our patient queues are increasing, and are now close to the regulatory
boundary and that could be crossed. (Head nurse) However, accountability for the budget is unclear; for many years, the case organization has
under-budgeted its activities, and acquiring extra funding near the end of the year has become
the way of doing things. Also, it may be that in the FHC, the steering influence of the
budgetary control has diminished due to sanctions from the alleged overspending being
actualized in the form of tighter budgets. Further, as budget responsibility has not been
delegated formally to the deputy chief physicians in charge of the clinics, this creates an
ambiguous situation concerning the meaning of financial accountability and may further
promote emphasis on professional accountability at the operational level. This under budgeting has been our way for years now. We just ask for extra funding
every year when we realize that the budget is not going to hold. And I understand it now,
the budget doesn’t mean a thing; extra funding comes anyway. (Head nurse) This under budgeting has been our way for years now. We just ask for extra funding
every year when we realize that the budget is not going to hold. And I understand it now,
the budget doesn’t mean a thing; extra funding comes anyway. (Head nurse) This under budgeting has been our way for years now. We just ask for extra funding
every year when we realize that the budget is not going to hold. And I understand it now,
the budget doesn’t mean a thing; extra funding comes anyway. (Head nurse) The budget is considered unrealistic by the personnel and the financial reporting is not
trusted. The overhead costs designated to the clinic in particular are seen as distorted. Management at the clinics seems to be confused by the overheads, as these are not within the
control of the clinic and are apparently hard to anticipate. The budget is considered unrealistic by the personnel and the financial reporting is not
trusted. The overhead costs designated to the clinic in particular are seen as distorted. Conflicting accountabilities and reasons for blame games Management at the clinics seems to be confused by the overheads, as these are not within the
control of the clinic and are apparently hard to anticipate. I do not even understand why these costs are put to us. What I would like to see is the
salaries, equipment, and materials we have used so far. This would be realistic for me. These overheads, they seem to come at an unsystematic pace. When I think that we are
on budget, suddenly there comes a sudden addition of overheads designated to our clinic
and all of a sudden, we’re way off. (Head nurse) Well they do send the reports to us on a regular basis. I even try to follow them, but you
cannot trust the reports coming from the financial administration. You can’t believe the
numbers. It is a constant pain to try to follow the budget. On the one hand, we’re
required to be cost-conscious, but on the other, these reports do not give us any way of
doing it. (Deputy chief physician) There is clear indication of the clinical professionals’ distrust of the financial information
provided. Previous budgeting mistakes, ever increasing cuts to the budget and a lack of 16 understanding about the principles behind the calculation of the overhead figures add to the
ambiguous situation the clinical professionals perceive. This has led to the clinical profession
to emphasize professional accountability in spite of the city administration’s aim to increase
the financial accountability in the organization. Further, clinical professionals direct blame
toward the financial administration and city management for previous mistakes. We had this error in the budget a while back. One cost item was left out of the budget,
which amounts to over one million euros. You can think for yourself what this means to
us. We have to stay on budget, while those making the budget leave out some numbers. We still have to pay the salary, even for those that are not included in the budget. (Director of basic health care) There is a lot of collation and analysis of statistics in a health care organization. However,
clinical professionals are loyal to each other and are careful not to disclose information that
could potentially be harmful. The financial administration is deliberately excluded from the
decision making. In addition, clinical management seems to continue to question the
reliability of the financial administration’s reports. Conflicting accountabilities and reasons for blame games This is
because then the patient has his/her problems taken care of. (Head nurse) There are doctors who take 4000–5000 patients per year, and then there are doctors who
take only 2000 patients per year. But then, the measure doesn’t take into account that
some doctors like to have patients come three times for one thing at a time, and other
doctors might take the time to handle everything in a single visit. While one doctor
handles everything in 30 minutes, the other takes 45 minutes to handle these things in
three separate visits. Although our system indicates three visits for the one guy and only
one for the other, there’s no question in my mind who was the most efficient. (Director of
basic health services) The financial manager views the deputy chief physicians in charge of individual clinics as
largely ignorant of the budgeting and monitoring process. The budget accountability centers
on the director of basic health care, while, according to the financial manager, deputy chief
physicians mostly try to organize individual clinics using non-financial performance
measurement metrics, such as waiting times, and the number of patient visits to a clinic. The
financial manager seems to argue that doctors are resistant to change and that they should
start to learn accounting. It is also an indication of a lack of dialogue, signaling that
professional groups do not understand each other. Deputy chief physicians…they know little about accounting. It is actually a bit sad that
we in the financial management have to fill them in and tell them where we are, on a
monthly basis. We’d really need them to start learning about accounting; perhaps then
they would understand what the numbers mean. The problem is, the power of the
profession is really high, doctors think that because we’ve always done like this, there is
no reason to change. They are always trying to resist different development ideas
(Financial manager) The clinical administration of the FHC is of the opinion that in this situation delivering cost
savings in health services would require shutting down some of the health care clinics. However, any proposal to close some of the clinics results in political debates and, ultimately
ends up being blocked by the politicians on the various boards of the city administration. Conflicting accountabilities and reasons for blame games This is evident in the reworking of
budgetary reports into calculations by the clinical management. Hood (2011) characterizes
such purpose-oriented reporting as a presentational strategy to initiate blame avoidance, while
professional grouping and internal dialogue suggests an agency related strategy for
diminishing individual implication. These reworked calculations are displayed without the
approval of the financial administration in the health care division and ultimately, to the basic
welfare board. Following Roberts (2009), we note that in complex organizations where
managerial (or financial) accountability is opaque but over-emphasized, blame game type
behavior might well be a rational choice; an actor is reconciling the conflicting situation in
order to preserve self-integrity and the necessary resources for operational work (See also
Cooper and Johnston, 2012; Hood, 2011; Messner, 2009). However, as a side effect, this kind
of behavior contains an obvious risk of sub-optimization and even manipulation of numbers
and reports: Our current general health manager is very enthusiastic about this (re-calculation of
financial reports), but the problem with doctors is that they do not want to show these
calculations to outside parties. Our relocated manager guarded these clinical
calculations very carefully and I can already see that his successor is also somewhat
reluctant to talk about them. (Deputy city mayor) Our current general health manager is very enthusiastic about this (re-calculation of
financial reports), but the problem with doctors is that they do not want to show these
calculations to outside parties. Our relocated manager guarded these clinical
calculations very carefully and I can already see that his successor is also somewhat
reluctant to talk about them. (Deputy city mayor) Overhead costs are distributed to individual clinics on the basis of full capacity. In our case
organization, one of the problems resulting in part from a restriction on hiring substitute
personnel is that most of the clinics are understaffed. This results in disproportionate
overhead costs being attributed to a single patient visit, thus causing sub-optimization and
distorted cost calculations. In addition to the budget, the efficiency of the different clinics is evaluated through the metric
of patient visits. This encourages the clinics to try to increase the number of patient visits to 17 both nurses and doctors. This measure is seen to promote sub-optimization as the content of
visits is not taken into account. In our work, it may well be that the fewer the visits, the more efficient it is. Conflicting accountabilities and reasons for blame games Politicians seem to be reluctant to take the responsibility for a decision to weaken services,
since that could weaken their chances of being re-elected. Whenever we try to make a big decision about service coverage, like shutting down a
couple of clinics to cut costs, we encounter political resistance and they keep saying that
we must leave the service coverage untouched. It is difficult for a health manager to
make any tough decisions when you don’t get political support. (General health manager) 18 Health care managers face the difficult situation of being accountable for the health care
budget while trying to make cost savings in the contradictory context of municipal politics. The case of the former general health manager mentioned above who was relocated can be
viewed as an example of politicians acting to secure their own position in terms of being
accountable to the public. The relocation of officeholders is one example of the agency
strategy for avoiding blame in Hood’s (2011) typology. This strategy can be utilized in the
form of dismissal (to shift blame to the person being dismissed), and by resigning, thereby
avoiding a blame falling on oneself. At the time, they (politicians) saw the manager’s responsibilities in the organization as
simply to be a cost-cutter. They didn’t appreciate cost-effectiveness, as in doing an
effective job or thinking about long-term cost effects, rather they wanted immediate cost
efficiency. And now that I think about what has happened since (the relocation), I can
already see that in one year, they (the FHC) are already overrunning the budget more
than ever. (Relocated general health manager) The general health manager reports to both the deputy chief mayor and the basic welfare
board. Trust between these actors defines the shape of the accountability relations. The
general health manager had a problematic relationship with the members of the basic welfare
board, which eventually led to his relocation. The situation at the time was also noted in an
internal auditing report of the organization: There is no mutual point of view about development needs or activity targets. This
observation includes both the trustees [political representatives of the municipalities] as
well as the operative management. (Internal audit report 4/2013) A local newspaper commented on the apparent conflict between administration and the
operative management: A local newspaper commented on the apparent conflict between administration and the
operative management: The position of the general health manager has been quite unstable for the last few
years. The shop steward of the city’s doctors’ association commented that they [general
health managers] haven’t been given the industrial peace in their work, their authority
has been constantly undermined and their decisions have been overturned by municipal
politicians. (Local newspaper article 1.7.2015) The situation at the FHC seems to be vulnerable to conflicts between forms of accountability
and prone to blame shifting, an example being the conflict between the members of the basic
welfare board and the general health manager. Between 2011 and early 2016, there were
eight general health managers, two of whom were substitutes. This issue has also been
discussed many times in the local newspaper in articles concerning professional cliques,
which report that doctors and their professional association resist candidates who lack a
clinical background. The situation is further complicated by the expectations of the city
administration and the politicians; consequently, the general health managers have found
their position to be quite unstable during the last years. 19 19 The conflict between the financial accountability advocated by the mayor and clinical
professional accountability is also evident. The mayor’s means are mostly limited to
rhetorical speeches and budgetary control (resulting in even more emphasis on financial
accountability) in a situation where most of the city’s clinics are already understaffed, and the
work in clinics is described in the local newspaper as strenuous ‘salt-mine work.’ The vice manager of the doctor’s union commented on the situation of the FHC clinics as
being like working in a salt-mine… especially young doctors do not want to work there. The ratio of pay to volume of work is just so bad in the FHC…In the FHC area,
understaffing and difficulties in attracting doctors for vacant positions have been a
constant issue. (Local newspaper article 20.1.2014) The views of the city management and the clinical profession concerning responsibilities are
different, and politicians seem to condone tighter budgets in the basic welfare board. The
media has started to disseminate the apparent difficulties inside the organization to the wider
public. A need for dialogue But we have
agreed things, and that’s how things work (Relocated general health manager) It might seem strange that a relocated manager still considers it possible to influence the
decision-making processes, and perhaps suggests some presence of unofficial power
structures and reluctance to change. Moreover, the complex multi-agency arrangement of the
FHC structure seems to allow for the relocation of a manager into a different position in the
organization while retaining some influence and a close relationship with the city’s doctors. The complexity of the context in which the FHC operates is highlighted by the point that in
addition to the clinical professionals and administration, stakeholders in the FHC include the
media and the political decision makers who represent the citizens. It might seem strange that a relocated manager still considers it possible to influence the
decision-making processes, and perhaps suggests some presence of unofficial power
structures and reluctance to change. Moreover, the complex multi-agency arrangement of the
FHC structure seems to allow for the relocation of a manager into a different position in the
organization while retaining some influence and a close relationship with the city’s doctors. The complexity of the context in which the FHC operates is highlighted by the point that in
addition to the clinical professionals and administration, stakeholders in the FHC include the
media and the political decision makers who represent the citizens. All the time, we have to try to find compromises and think about the process…if we can
ignore some things, or move some tasks to be done by somebody else…But as I
mentioned, there are not only the citizens, and politicians, but there is also the law. It is a
very big pattern, and there is also the media, which is a big opinion leader. (Relocated
general health manager) The autonomy of the clinical profession is so strong that it does not need to take on
development ideas from outside of the profession; the only time that outsiders are included in
the talks is when they involve determining resources for the operations. Even in these
situations, clinical professionals tend to strive for control over the discussion and terms. Well, at least in my personal opinion, whenever I have tried to suggest some course of
action or development idea, they (doctors) have been very cold, to the point of throwing
me out of their meetings. A need for dialogue … A shop steward for the city’s doctors continues; “Doctors in the city do not trust the
administration. They feel that the decisions made by the administration are in a constant
flux. Also, it is clear that there is no willingness [on their part] to really get together and
solve these issues [conflicts between administration and doctors] with mutual
discussions” (Local newspaper article 1.7.2015) I think it was about one and a half years ago that the mayor was talking to the clinical
management team and said that the situation is dire, the situation needs to be improved,
both in terms of mutual relationships and of budget control. He also said that if someone
thinks that they’re not accountable, or they don’t want to be, they don’t have to be here
(in the organization). And I was there, too. And I looked at the expression on their faces
[clinical professionals] and I already knew they weren’t going to change a thing. (Deputy chief mayor) This lack of dialogue between city administration and clinical professionals suggests that
when two strong accountabilities conflict and common goals and blame sharing are absent,
different parties reconcile the situation by utilizing the blame avoidance strategies. One aim
of such strategies seen in our case is to downplay financial accountability, action that comes
close to resembling civil disobedience in terms of the vertical hierarchy. Another implication
in our case is that loyalty to the profession and an individual sense of responsibility is
stronger than loyalty to the organization, at least when the organization is both vertically and
horizontally complex. Demands for cost efficiency manifest as an internal debate between clinical professionals
with the aim of reconciling conflicting accountabilities. In this situation, the essential goal is 20 to keep different accountabilities aligned and make decisions that have the minimum effect
on clinical leeway, while at the same time satisfying the city administration. Further, the
relocated general health manager seems to aim to retain some influence over the processes. I will have discussions with the current general health manager, and also discussions
with the people in the main processes, in which we will agree what template we will
have. I mean, it’s the personnel cost, that is the biggest cost for us, so there will not be a
problem, even here the service manager has the budget accountability. A need for dialogue For example, I tried to offer a development idea about co-
ordination of closing times between clinics but nobody listened. (Development officer) In the citywide organization (Figure 1 – Appendix 1), the city mayor and deputy mayor act as
superiors to the general health manager, a position to date always occupied by a medical
doctor. The city administration expects the general health manager to reduce costs, while at
the same time the clinical professionals expect to be provided with reasonable resources to
maintain the expected level of service for social and health services. In the citywide organization (Figure 1 – Appendix 1), the city mayor and deputy mayor act as
superiors to the general health manager, a position to date always occupied by a medical
doctor. The city administration expects the general health manager to reduce costs, while at
the same time the clinical professionals expect to be provided with reasonable resources to
maintain the expected level of service for social and health services. Municipal politicians expect the organization to provide good quality services at minimum
cost. The municipalities expect to be provided with timely financial reports on the different
areas of the FHC to assist their oversight responsibilities. However, some of the 21 municipalities in the FHC were unhappy about the financial reporting provided by the main
city: … member municipalities need to be informed better of the developments in the FHC
during the fiscal period. Timely and precise reports are needed to ensure that member
municipalities can keep track of the expenses. … a representative [of one of the member
municipalities] stated “we shouldn’t be forced to dig out the information from databases
ourselves and try to calculate the different cost units from pooled data. For example, last
year [2014] our expenses [from the FHC] were bigger than the agreed budget” (Local
newspaper article 24.2.2015) The health care organization has sought savings by closing health clinics for short periods,
stopping the hiring of substitute staff, using group sessions for patients (when permitted by
privacy law), and instructing patients on self-care whenever possible. In addition, in some
health clinics, elements of the workload have been transferred from doctors to nurses and in
some cases, a nurse has been hired in place of a doctor. Health care professionals see the
situation as critical in terms of service level. A need for dialogue The cost-saving measures initiated by the health
care organization do not seem to have produced the savings expected by the city
administration. Hence, the point where the service level starts to degrade is near. One of the
directors of nursing services called for the politicians to take responsibility and initiate open
dialogue with the citizens: We do understand that the city is running out of money. But then, we would like the
decision makers to tell the public that there is no more money, that we cannot get the
same service anymore. Of course, we have tried to come up with different solutions to
save where possible, but these are minor things. Like group sessions, temporary closures
of clinics, and then there are restrictions on hiring substitutes. (Director of nursing
services) An important way for the city administration to enforce cost savings is through the human
resource management function of the city. All recruitment, arrangement of temporary
substitute personnel, and other decisions concerning the number of personnel and salary
issues, have to be accepted there. The high salary cost of using substitute doctors combined
with the scarcity of resources has forced the FHC to operate without a sufficient number of
substitutes. Accordingly, the FHC has defined the critical functions for each main process,
while the remaining functions are not given priority. We have the staffing plans that we will make together…we have agreed for example how
we take the substituting personnel. It means that we have to take care of certain critical
functions, whatever our financial situation, because of patient security. (Relocated
general health manager) We have the staffing plans that we will make together…we have agreed for example how
we take the substituting personnel. It means that we have to take care of certain critical
functions, whatever our financial situation, because of patient security. (Relocated
general health manager) A new manager, responsible for health care services, tried to find solutions to the situation by
holding both personal and small group discussions over three months with the staff of the
FHC. In an effort to reconcile conflicting accountabilities, the manager also had a great deal 22 of dialogue with the decision makers on the basic welfare board, aiming to teach them to
recognize the health care issues affected by the decision-making processes. This implies the
(as yet unrealized) potential of dialogue between professional groups to reconcile conflicting
accountabilities. A need for dialogue I had some good discussions with the basic welfare board, and I tried to bring the
information to the decision makers from the viewpoints which are important in health
care…also there was some understanding, at least partly, about the financial aspects. But then again, when we went to the budget negotiations in the autumn, it was never
realized in practice, because the decision makers saw that now we have the budget
frames here and we cannot go over these frames, even though we know what should be
done. (General health manager) Although the strict budgetary process did not seem to allow for any leeway at the time,
attempts at dialogue were initiated, and the new general health manager seemed somewhat
optimistic at the time that future decisions would acknowledge the health care point of view
to a greater extent than had happened in the past. Further, the new general health manager
seemed to have the trust of the city administration as well as an opportunity to build a bridge
between clinical professionals and the city administration. There is still some of that old hatred of the upper classes, so to speak. They (old
management) were able to build a kind of idea among the clinical employees that the city
administration is trying to thwart every effort and constrict all the resources to the bare
minimum. Now we have this new (general health) manager, he’s different. He seems to
be open-minded, talks with us as well as the clinical professionals…I do think that we
will get development efforts moving forward now. (Deputy chief mayor) However, the deputy chief mayor and the general health manager interviewed for this study
have since resigned. Only the latter position has been filled (by a temporary substitute) which
suggests the difficult and contradictory conditions in the organization remain. 5. Discussion The FHC was constructed as a network type of provider of health care services by the
surrounding municipalities, but the power of decision was ultimately granted to the main city. One purpose of this arrangement was to seek financial savings. In order to realize these well-
intended financial objectives, the city mayor focused on financial accountability and tried to
push that accountability through the budgets and performance measurement into the health
care organization with an authoritative approach. However, the FHC’s budget was constantly
exceeded and the role of the budget as a steering mechanism in the FHC seems weak in light
of our empirical findings. At the organizational level, we found different blame game
strategies as an outcome of conflicting accountabilities (see Hood, 2011). Clinical
professionals defend themselves against accusations of overspending by employing various
rhetorical strategies, combined with the non-delegation of budget accountability to the clinic 23 level. Constant replacements in the position of the general health manager and the difficulties
in recruiting doctors can be seen as a possible use of agency strategies in the organization. Understaffing in the health clinics has led to increases in treatment waiting lists and
weakened the reputation of the FHC among clinical professionals and the citizens of the
region. The problems in the FHC covered by the media have contributed to this bad
reputation being circulated nationwide. Messner (2009) warns about ethical conflicts on the individual level that result from
increased emphasis on measurable aspects of accountability. In a similar way, Roberts (2009)
sees the increased emphasis on transparency and the associated performance measures as
problematic. These authors base their views on Butler (2005), who determines the
responsibility of the individuals by way of the groups that they choose to associate
themselves with. This individual responsibility may manifest through decisions, such as
resignations or difficulties in fulfilling positions as evidenced in our case FHC. Possible
causes for ethical contradictions on the individual level in the FHC include the mismatch
between individual or professional views and the financial accountability driven by the city
management. However, our findings indicate that clinical professionals are not passively
submitting to this ethos but utilize different defensive tactics, characterized in this study as
behavior similar to the blame game described by Hood (2002, 2011). 5. Discussion The purpose of such
blame game tactics is, in line with the clinical ethos, to preserve self-integrity and the
necessary resources to provide good quality care for the patients. These actions
demonstrating medical professionals’ resistance can also be seen as a form of democratic
accountability. Bovens (1998) considers active resistance to management to be a wholly legitimate course of
action in the face of serious conflict between organizational goals and public morale. Some
examples in our case include demands by the health professionals for the city management’s
active responsibility, prioritizing quality patient care over financial considerations as well as
the critical views expressed in the media. Previous accounting research has analyzed the
democratic accountability and resistance by the service users to managerial forms of
accountability (e.g., Cooper and Johnston, 2012; Smyth 2012). In the FHC, however, we
found that clinical professionals utilized democratic accountability to actively resist financial
accountability. This finding contributes to the different forms of expressing democratic
accountability within a complex public-sector organization, and by providing an example of
one leading profession acting on the basis of citizenship (cf. Bovens, 1998; Smyth, 2012). Problems associated with democratic accountability in the organization appeared as political
failures of the politicians to agree on issues, including the closing of small clinics. This is
explained by the main city politicians’ reluctance to make decisions that would reduce the
service level provided to the local electorate. Moreover, in the FHC administration, the
surrounding municipalities are limited to an advisory role through the health care division,
while decisions concerning the FHC are made at the level of the basic welfare board,
comprising members exclusively drawn from the main city’s municipal council. Accounting
could have an important intermediary role between network partners (Kurunmäki and Miller 24 2011), but in our case, the surrounding municipalities felt that the reporting of financial
information between municipalities was insufficient. Borowiak (2011) sees punishability as a
central aspect of democratic accountability. However, the surrounding municipalities had
only limited means to control the main city’ decisions concerning the FHC. Owing to this
inadequate democratic accountability design, the only option for the surrounding
municipalities would be to disengage from the FHC organization. However, this would make
it difficult for the smaller municipalities to provide the statutory level of services. Moreover,
it would lead to difficult negotiations with the other municipalities on the terms of the
dismantling of the organization. 5. Discussion Earlier studies have indicated that a move from hierarchical state entities to co-ordination of
networks incorporates demands for increased dialogue to be expanded to citizen groups and
other active stakeholders (Ahrens and Ferry, 2015; O’Dwyer and Unerman, 2007; Cordery et
al., 2010). It is within the influence of the city mayor to choose to either exercise an
authoritarian style of leadership or to engage the different stakeholders through dialogue. Within the democratic idea of providing services to the people sustainably, it might be in the
long-term interests of different stakeholders to co-operate. Engaging different stakeholders
through dialogue might make it possible to reconcile different views on accountability within
an organization (Ahrens and Ferry, 2015; Goretzki and Messner, 2016). However, our
findings confirm earlier results (Cordery et al., 2010; Lowe et al., 2012) that hierarchical
practices and a focus on financial accountability are difficult to change into more interactive
(or holistic) forms of accountability; particularly if the management has little will to engage
in such practices. In our case, we observed only limited dialogue between the different
stakeholder groups. Instead, we observed organizationally counter-productive behavior,
characterized in this study as behavior characteristic of a blame game (Hood, 2002; 2011). It
is noteworthy that what seems at the organizational level to be avoidance of responsibility or
blame shifting, may be the rational and justified choices of individuals acting in a very
complex situation. However, blame games are often detrimental to the organization in that
they produce unproductive contradictions between different stakeholder groups. Our empirical results indicate the need for dialogue between the parties to accountability
relationships to tie the financial, health service, and political targets more closely together. This, as noted by Cordery et al. (2010), seems to demand strong leadership efforts supporting
co-operation, both personally and among the professions to understand the views of the other
parties affected. Our findings add to those of Cordery et al. (2010), who found that dialogue-
based reconciliation between accountabilities is complicated by an unclear division of
different accountabilities and the stability of old hierarchical practices. Our findings indicate
that other obstacles to open dialogue are: a strong emphasis on financial accountability, a
complex organizational structure, an unwillingness to provide relevant information, and a
silo-style operating culture. According to Roberts (2009), if there is no common vision and information is not exchanged
between parties, the essential prerequisites for dialogue are missing. 5. Discussion We elaborate on these
characteristics by noting that although information might be exchanged, it can be flawed. 25 Examples from our case include an unwillingness to provide financial information to other
parties, errors in budgetary information, and deliberate reworking of financial reports for
people’s own purposes. Further, in a complex network organization, mutual dialogue
allowing for a common vision can be difficult to instill even at the organization level. In a
complex network context with a focus on financial pressures, the blame for any situation can
be shifted between different bodies, and calculations can be deliberately made to seem more
favorable to the interest group in question. In addition, the complexity of the democratic
accountability concept offers opportunities to shift blame among stakeholders and
organizational units. When accountabilities are lost among the complex organizational
structures of a network, making informed decisions can become increasingly difficult. In the FHC organization, the prominent accountabilities include financial accountability,
professional accountability, and also democratic accountability. We see these different
accountabilities as necessary parts of the organization, although in our case, the complexity
of the organization coupled with accountabilities being weighted differently by different
stakeholder groups has led to the problems in reconciling these multiple accountabilities. The
interplay of these different accountability relations is complex, and results in a self-sustaining
vicious circle of distrust, ethical issues, and a blame game that is hard to dismantle. 6. Conclusions One important professional
group that is unwilling to engage in information exchange hampers the creation of dialogue
and change in current practices. Further, we agree that if the drive for financial accountability
is to be amended with a dialogical approach aiming to involve a larger group of stakeholders,
an accountability system must be designed to support it (Brown, 2009; Ahrens and Ferry,
2015), citizens need to be empowered in decision making (Borowiak, 2011; Cooper and
Johnston, 2012) and the senior management needs to have the will to engage in such dialogue
(O’Dwyer and Unerman, 2007; Cordery et al., 2010; Lowe et al., 2012). This study also
provides empirical evidence for the detrimental effects that the complex networked structure
can have on democratic accountability. Second, we amend the obstacles to consensus seeking dialogue discussed by Cordery et al. (2010) who found that dialogue-based reconciliation between accountabilities is complicated
by an unclear division of different accountabilities and the stability of old hierarchical
practices. Our findings indicate that other obstacles to open dialogue include a strong
emphasis on financial accountability, a complex organizational structure, an unwillingness to
provide relevant information, and a silo-style operating culture. One important professional
group that is unwilling to engage in information exchange hampers the creation of dialogue
and change in current practices. Further, we agree that if the drive for financial accountability
is to be amended with a dialogical approach aiming to involve a larger group of stakeholders,
an accountability system must be designed to support it (Brown, 2009; Ahrens and Ferry,
2015), citizens need to be empowered in decision making (Borowiak, 2011; Cooper and
Johnston, 2012) and the senior management needs to have the will to engage in such dialogue
(O’Dwyer and Unerman, 2007; Cordery et al., 2010; Lowe et al., 2012). This study also
provides empirical evidence for the detrimental effects that the complex networked structure
can have on democratic accountability. Third, our results empirically contribute to the discussion on the blame game and the
accountability-related rationales behind it. At an individual level, this study illustrates the
blame game can be a rational reaction to situations characterized by moral contradictions due
to an increasing emphasis on financial (and managerial) forms of accountability (Roberts,
2009; Messner, 2009). 6. Conclusions This study considered different instances of the blame game (Hood, 2002; 2011) in relation to
financial, professional, and democratic accountabilities. We discussed the responsibility of
individuals and the different reasons behind blame game type behavior. Further, we examined
democratic accountability and citizenship in complex organizations under economic austerity. The study contributes to the literature by highlighting the multiple ways in which different
forms of accountability, particularly democratic accountability, can manifest and be
understood in a municipal organization with vertical and horizontal features. This study set out to investigate the interplay of different accountabilities in a complex
healthcare organization with the help of literature-based concepts of the blame game and
dialogue. We identified failures in finding mutually acceptable solutions to organizational
development needs (see, Cordery et al. 2010; O’Neill, 2002; Roberts, 2009), and instead
found a situation characterized by separate and conflicting accountability views and the
utilization of defensive tactics characterized in this study as blame game strategies. These
problems were evident among the different organizational levels, units, and stakeholder
groups involved in service provision (see, Hood, 2008, 2011). The findings from this study make several contributions. First, we contribute to existing
literature by empirically noting the different, conflicting accountabilities as a potential root of
organizationally detrimental behavior (cf. Messner, 2009; Hood, 2008; 2011; Roberts, 2009)
and illustrate the apparent rationality of Hood’s (2011) different strategies in shifting blame
in a networked public health care case organization. In our case, the conflict between 26 professional and financial accountability seems to be key: the city management utilized
policy-related strategies combined with agency strategies in directing blame toward
individuals and the clinical professional group. In contrast, the clinical professionals seemed
to utilize a presentational strategy to avoid blame. In addition, the inability of municipal
politicians to agree on cost-saving measures initiated by the clinical management suggests a
short-term view of democratic accountability and a reluctance to reduce the service level to
the local electorate, and also indicate the utilization of policy and agency strategies. Second, we amend the obstacles to consensus seeking dialogue discussed by Cordery et al. (2010) who found that dialogue-based reconciliation between accountabilities is complicated
by an unclear division of different accountabilities and the stability of old hierarchical
practices. Our findings indicate that other obstacles to open dialogue include a strong
emphasis on financial accountability, a complex organizational structure, an unwillingness to
provide relevant information, and a silo-style operating culture. References Ahrens, T. (1997) Styles of accountability. Accounting, Organizations and Society, Vol. 21 No 2/3 pp. 139-173. Ahrens, T. and Chapman, C. (2002) The structuration of legitimate performance measures and
management: day-to-day contests of accountability in a UK restaurant chain. Management
Accounting Research, Vol. 13 No. 2 pp. 151-171. Ahrens, T. and Ferry, L. (2015) Newcastle City Council and the grassroots: accountability and
budgeting under austerity. Accounting, Auditing & Accountability Journal, Vol. 28 No. 6, pp. 909-933. Almquist, R., Grossi, G., van Helden, G. J. and Reichard, C. (2013) Public sector governance and
accountability. Critical Perspectives on Accounting, Vol. 24 No. 7/8, pp. 479-487. Alvesson, M. (2003) Beyond neopositivists, romantics and localists: a reflexive approach to
interviews in organizational research. The Academy of Management Review, Vol. 28 No. 1, pp. 13–33. Barberis, P. (1998) The New Public Management and a New Accountability. Public Administration,
Vol. 76: 451-470. Borowiak, C. T. (2011) Accountability & Democracy – The Pitfalls and Promise of Popular Cotnrol. Oxford University Press. New York. Bovens, M. (1998) The quest for responsibility – Accountability and citizenship in complex
organisations. Cambridge University Press. Bovens, M. (2005) Public accountability. In Ewan F., Lawrence Jr L., Pollitt, C. (eds.). The Oxford
Handbook of Public Management. Oxford Handbooks. Broadbent, J., Jacobs, K. and Laughlin, R. (2001) Organisational resistance strategies to unwanted
accounting and finance changes: The case of general medical practice in the UK. Accounting,
Auditing & Accountability Journal, Vol. 14 No. 5, p.565-586. Brown, J. (2009) Democracy, sustainability and dialogic accounting technologies: Taking pluralism
seriously. Critical Perspectives on Accounting, Vol. 20, pp. 313-342. Butler J (2005) Giving an Account of Oneself Fodham University Press New York ,
(
)
y,
y
g
g
g
g p
seriously. Critical Perspectives on Accounting, Vol. 20, pp. 313-342. Butler, J. (2005) Giving an Account of Oneself. Fodham University Press. New York. y
p
g,
, pp
Butler, J. (2005) Giving an Account of Oneself. Fodham University Press. New York. Chua, W. F. (1995) Experts, networks and inscriptions in the fabrication of accounting images: a story
of the representation of three public hospitals. Accounting, Organizations and Society, Vol. 20
No. 2-3, pp. 111-145. Cochrane, A. (1993) From financial control to strategic management: the changing faces of
accountability in British central government. Accounting, Auditing and Accountability Journal,
Vol. 6, No. 3, pp. 30-51. Cooper, C. and Johnston, J. (2012) Vulgate accountability: instights from the field of football. 6. Conclusions Further, a type of punishability (Borowiak, 2011) was evidenced in
our case in clinical professionals resisting cost-saving demands based on financial
accountability, thus providing an example of one leading profession acting on the basis of
citizenship (Bovens, 2005; Cooper and Johnston, 2012; Smyth, 2012). Our study explains
how individuals manage an uncertain situation involving contradictions between
management’s focus on financial savings, a professional ethos, and the wellbeing of citizens. As this study is the first that we know of to empirically illustrate Hood’s (2011) theory of the
blame game in a case study, we would welcome further research on the subject. We would
particularly welcome studies that uncover practices supporting co-operation and dialogue in
network organizations. We also call for further research addressing the multifaceted nature of
democratic accountability, particularly in network organizations, where democratic
accountability can be seen in many different ways over time and is constantly redefined. 27 References Accounting, Auditing & Accountability Journal, Vol. 25, No. 4, pp. 602-634. Cordery, C., Baskerville, R. and Porter, B. (2010) Control or collaboration? Contrasting
accountability relationships in the primary health sector. Accounting, Auditing and
Accountability Journal, Vol. 23, No. 6, pp. 793-813. Day, P. and Klein, R. (1987) Accountabilities: Five public services. Tavistock, London. Ezzamel, M. and Willmott, H. (1993) Corporate governance and financial accountability: recent
reforms in the UK public sector. Accounting, Auditing and Accountability Journal, Vol. 6, No. 3, pp. 109-32. Golden-Biddle, K. and Locke, K. (1993) Appealing work: an investigation of how ethnographic texts
convince. Organization Science, Vol. 4, No. 4, pp. 595-616. 28 Goddard, A. (2005) Accounting and NPM in UK local government – contributions towards
governance and accountability. Financial Accountability and Management, Vol. 21 No. 2, pp. 191-218. Goretzki, L. and Messner, M. (2016) Coordination under uncertainty – A sensemaking perspective on
cross-functional planning meetings. Qualitative Research in Accounting & Management, Vol. 13 No. 1, pp. 92-126. Gray, A. and Jenkins, B. (1993) Codes of accountability in the new public sector. Accounting,
Auditing and Accountability Journal, Vol. 6, No. 3, pp. 75-99. Gårseth-Nesbakk, L. and Kjærland, F. (2016) Precarious Investments and Blame Gaming -
Adverse Effects and the Inherent Danger of Simplification. Financial Accountability &
Management, Vol. 32 No. 3, pp. 281-308. Hodges, R. (2012) Joined-up government and the challenge to accounting and accountability
researchers. Financial Accountability and Management, Vol. 28 No. 1, pp. 26-51. Hood, C. (2002) The risk game and the blame game. Government and Opposition, Vol. 37 No. 1, pp. 15–37. Hood, C. (2008) The risk game and the blame game. In Boin, A. (Ed.) Crisis Management, Volume
III. London: Sage Publications, pp. 171–186. Hood, C. (2011) The Blame Game. Spin, Bureaucracy, and Self-Preservation at Government. Princeton: Princeton University Press. Humphrey, C., Miller, P. and Scapens, R. (1993) Accountability and accountable management in the
UK public sector. Accounting, Auditing and Accountability Journal, Vol. 6, No. 3, pp. 7-29. Hyvönen, T. and Järvinen, J. (2006) Contract-based budgeting in health care: a study of the
institutional processes of accounting change. European Accounting Review, Vol. 15, No. 1, pp. 3–36. Kurunmäki, L. (2004) A hybrid profession – the acquisition of management accounting expertise by
medical professionals. Accounting, Organizations and Society, Vol. 29 No., pp. 327-47. Kurunmäki, L. and Miller, P. (2011) Regulatory hybrids: partnerships, budgeting and modernizing
government. Management Accounting Research, Vol. 22, No. 4, pp. References 220-41. Lapsley, I., Kurunmäki, L. and Melia, K. (2003) Accountingization v. legitimation: a comparative
study of the use of accounting information in intensive care. Management Accounting
Research, Vol. 14, No. 2, pp. 112-39. Lapsley, I. (2008) The NPM agenda: back to the future. Financial Accountability & Management,
Vol. 24, No. 1, pp. 77-96. Llewellyn, S. (1998) Pushing budgets down the line: ascribing financial responsibility in the UK
social services. Accounting, Auditing & Accountability Journal. Volume 11, No. 3, pp. 292-
308. Lowe, A., Locke, J. and Lymer, A. (2012) The SEC’s retail investor 2.0: Interactive data and the rise
of calculative accountability. Critical Perspectives on Accounting, Vol. 23, No. pp. 183-200. McKinnon, J. (1988) Reliability and validity in field research: some strategies and tactics. Accounting,
Auditing & Accountability Journal, Volume 1, No. 1, pp. 34–54. Messner, M. (2009) The limits of accountability. Accounting, Organizations and Society, Vol. 34, No. 8, pp. 918-938. Mulgan, R. (2000) ‘Accountability’: An ever-expanding concept? Public Administration, Vol. 78 No. 3, pp. 555-573. Mutiganda, J. C. (2013) Budgetary governance and accountability in public sector organisations: An
institutional and critical realism approach. Critical Perspectives on Accounting, Vol. 24, No. 7/8, pp. 518-31. 29 O’Dwyer, B. and Unerman, J. (2007) From functional to social accountability: Transforming
the accountability relationship between funders and non‐governmental development
organisations. Accounting, Auditing & Accountability Journal. Vol. 20, Iss: 3, pp.446 - 471. g
O’Neill, O. (2002) A Question of Trust. Cambridge University Press. Neill, O. (2002) A Question of Trust. Cambridge U Papadopoulos, Y. (2003) Cooperative forms of governance: Problems of democratic accountability in
complex environments. European Journal of Political Research, Vol. 42, pp. 473-501. Pierre, J. (2000) Introduction: Understanding governance. In J. Pierre (ed.) Debating Governance. Oxford University Press, pp. 1-10. Rhodes, R. A. W. (1997) Understanding Governance. Public networks, governance, reflexivity and
accountability. Open University Press. Rhodes, R. A. W. (2007) Understanding Governance: Ten Years On. Organization Studies, Vol. 28
No. 8, pp. 1243-1264. Roberts, J. (2003) The manufacture of corporate social responsibility; constructing corporate
sensibility. Organization, Vol. 10 No. 2, pp. 249–265 Roberts, J. (2009) No one is perfect: the limits of transparency and an ethic for intelligent
accountability. Accounting, Organizations and Society, Vol. 34, No. 8, pp. 957-70. Roberts, J. and Scapens, R. (1985) Accounting Systems and Systems of Accountability –
Understanding
Accounting
Practices
in
their
Organizational
Context. Accounting,
Organizations and Society, Vol. 19, No. 2, pp. 443-56. Yin, R. K. (1984). Case study research: Design and methods. Newbury Park, CA: Sage. References Romzek, B. S. and Dubnick, M. J. (1987) Accountability in the public sector: Lessons from the
Challenger tragedy. Public Administration Review, Vol. 47, pp. 227-38. Silverman, D. (2001) Interpreting Qualitative Data. Methods for Analysing Talk, Text and
Interaction. 2nd Ed. (first published 1993). London: Sage Publications. Sinclair, A. (1995) The chameleon of accountability: forms and discourses. Accounting
Organizations and Society, Vol. 20, Nos. 2/3, pp. 219-237. Smyth, S. (2012) Contesting public accountability: A dialogic exploration of accountability and socia
housing. Critical Perspectives on Accounting, Vol. 23, No. 3, pp. 230-43. Yin, R. K. (1984). Case study research: Design and methods. Newbury Park, CA: Sage. 30 Appendix 1
Figure 1 – Main city structure and the FHC Figure 1 – Main city structure and the FHC 1 2 Appendix 2 – List of interviews First round of interviews
Internal auditing chief and external auditor 6.2.2013
Accounting chief 11.2.2013
Financial manager of the social and health services 11.2.2013
Director of nursing profession 19.2.2013
Director of basic health care 19.2.2013
Accounting service chief and accounting secretary 17.4.2013
Financial manager of the social and health services 24.4.2013 Health clinic 3
Nurse1, 9.10.2013
Nurse2, 15.10.2013
Doctor1, 18.10.2013
Doctor2, 18.10.2013
Deputy chief physician, 18.10.2013
Head nurse, 18.10.2013 Third round of interviews
Internal auditing chief and city financial chief 7.2.2014
Deputy city mayor 18.2.2014
General health manager 18.2.2014
Development officer 4.3.2014
Director of nursing services 7.3.2014
Relocated general health manager 7.3.2014 1
|
https://openalex.org/W1988583290
|
https://zenodo.org/record/896799/files/article.pdf
|
English
| null |
Science-based decision-making on complex issues: Marcellus shale gas hydrofracking and New York City water supply
|
Science of the total environment
| 2,013
|
cc-by-sa
| 23,342
|
*Manuscript Manuscript
Click here to download Manuscript: 4threvSTOTEN-D-13-00220R3.doc Click here to view linked References 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
1
Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2
Timothy T. Eaton
3
School of Earth and Environmental Sciences
4
Queens College - City University of New York
5
65-30 Kissena Blvd. 6
Flushing, NY 11367 U.S.A. 7
Phone: 718 997-3327
8
Fax: 718 997-3299
9
Timothy.Eaton@qc.cuny.edu
10
Abstract
11
Complex scientific and non-scientific considerations are central to the
12
pending decisions about "hydrofracking" or high volume hydraulic
13
fracturing (HVHF) to exploit unconventional natural gas resources
14
worldwide. While incipient plans are being made internationally for major
15
shale reservoirs, production and technology are most advanced in the
16
United States, particularly in Texas and Pennsylvania, with a pending
17
decision in New York state whether to proceed. In contrast to the narrow
18
scientific and technical debate to date, focused on either greenhouse gas
19
emissions or water resources, toxicology and land use in the watersheds
20
that supply drinking water to New York City (NYC), I review the scientific
21
and technical aspects in combination with global climate change and other
22
critical issues in energy tradeoffs, economics and political regulation to
23
evaluate the major liabilities and benefits. Although potential benefits of
24
Marcellus natural gas exploitation are large for transition to a clean energy
25
economy, at present the regulatory framework in New York State is
26
inadequate to prevent potentially irreversible threats to the local
27
environment and New York City water supply. Major investments in state
28 1
Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2
Timothy T. Eaton
3
School of Earth and Environmental Sciences
4
Queens College - City University of New York
5
65-30 Kissena Blvd. 6
Flushing, NY 11367 U.S.A. *Manuscript 7
Phone: 718 997-3327
8
Fax: 718 997-3299
9
Timothy.Eaton@qc.cuny.edu
10
Abstract
11
Complex scientific and non-scientific considerations are central to the
12
pending decisions about "hydrofracking" or high volume hydraulic
13
fracturing (HVHF) to exploit unconventional natural gas resources
14
worldwide. While incipient plans are being made internationally for major
15
shale reservoirs, production and technology are most advanced in the
16
United States, particularly in Texas and Pennsylvania, with a pending
17
decision in New York state whether to proceed. In contrast to the narrow
18
scientific and technical debate to date, focused on either greenhouse gas
19
emissions or water resources, toxicology and land use in the watersheds
20
that supply drinking water to New York City (NYC), I review the scientific
21
and technical aspects in combination with global climate change and other
22
critical issues in energy tradeoffs, economics and political regulation to
23
evaluate the major liabilities and benefits. Although potential benefits of
24
Marcellus natural gas exploitation are large for transition to a clean energy
25
economy, at present the regulatory framework in New York State is
26
inadequate to prevent potentially irreversible threats to the local
27
environment and New York City water supply. Major investments in state
28 Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 Major investments in state
28 1 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
and federal regulatory enforcement will be required to avoid th
29
environmental consequences, and a ban on drilling within the N
30
supply watersheds is appropriate, even if more highly regulated
31
gas production is eventually permitted elsewhere in New York
32
Keywords: hydrofracking, energy, fossil fuel; unconventiona
33
gas; shale; water resources; environment; economics; regulatio
34
Key Points:
35
x Previous analyses have taken too narrow a perspective on
36
benefits of hydrofracking or HVHF in New York state, w
37
for unconventional natural gas production elsewhere arou
38
x Benefits of HVHF natural gas production for reducing dep
39
more damaging coal-fired electrical power generation are
40
environmental and public health liabilities
41
x Protecting watersheds for NYC and other major municipa
42
is paramount but strengthening of federal and state regula
43
needed for reducing potential adverse impacts of HVHF e
44
state
45
46
1. Introduction
47
There is an urgent need to reduce the current global energ
48
fossil fuels, because of the risks of rising greenhouse gas (GHG)
49
driving global climate change (IPCC, 2007), and because most co
50
and gas reserves may no longer be reliably supplied due to politic
51
New unconventional discoveries have dramatically expanded esti
52 2
and federal regulatory enforcement will be required to avoid these
29
environmental consequences, and a ban on drilling within the NYC water
30
supply watersheds is appropriate, even if more highly regulated Marcellus
31
gas production is eventually permitted elsewhere in New York state. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 Abstract
11
Complex scientific and non-scientific considerations are central to the
12
pending decisions about "hydrofracking" or high volume hydraulic
13
fracturing (HVHF) to exploit unconventional natural gas resources
14
worldwide. While incipient plans are being made internationally for major
15
shale reservoirs, production and technology are most advanced in the
16
United States, particularly in Texas and Pennsylvania, with a pending
17
decision in New York state whether to proceed. In contrast to the narrow
18
scientific and technical debate to date, focused on either greenhouse gas
19
emissions or water resources, toxicology and land use in the watersheds
20
that supply drinking water to New York City (NYC), I review the scientific
21
and technical aspects in combination with global climate change and other
22
critical issues in energy tradeoffs, economics and political regulation to
23
evaluate the major liabilities and benefits. Although potential benefits of
24
Marcellus natural gas exploitation are large for transition to a clean energy
25
economy, at present the regulatory framework in New York State is
26
inadequate to prevent potentially irreversible threats to the local
27
environment and New York City water supply. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 32
Keywords: hydrofracking, energy, fossil fuel; unconventional; natural
33
gas; shale; water resources; environment; economics; regulation
34
Key Points:
35
x Previous analyses have taken too narrow a perspective on the risks and
36
benefits of hydrofracking or HVHF in New York state, with implications
37
for unconventional natural gas production elsewhere around the world
38
x Benefits of HVHF natural gas production for reducing dependence on even
39
more damaging coal-fired electrical power generation are great but so are
40
environmental and public health liabilities
41
x Protecting watersheds for NYC and other major municipality water supply
42
is paramount but strengthening of federal and state regulatory oversight is
43
needed for reducing potential adverse impacts of HVHF elsewhere in NY
44
state
45
46
1. Introduction
47
There is an urgent need to reduce the current global energy dependence on
48
fossil fuels, because of the risks of rising greenhouse gas (GHG) emissions
49
driving global climate change (IPCC, 2007), and because most conventional oil
50
and gas reserves may no longer be reliably supplied due to political instability. 51
New unconventional discoveries have dramatically expanded estimates of natural
52 Formatted: Font: Not Bold 2 2 gas reserves (US DOE, 2009; IEA, 2011; US EIA, 2012) and natural gas is a
53
preferred fuel for energy-efficient electricity production because it is cleaner-
54
burning compared to coal (Hultman et al. 2011). Recent discovery of
55
unconventional gas in the Marcellus shale in the northern Appalachian mountains
56
of the US provides a potential new energy source close to major mid-Atlantic
57
urban centers. Therefore, many have advocated a greater use of natural gas, as a
58
“bridge” fuel towards a renewable energy future (e.g. IPCC, 2007; Moniz et al.,
59
2010; Jenner and Lamadrid, 2012) despite controversy over the economics
60
(Hughes, 2011; Brooks, 2012) and life-cycle GHG costs (Burnham et al., 2012;
61
O'Sullivan and Paltsev, 2012) of unconventional gas compared to other energy
62
options. 63 Unconventional gas production from shales like the Marcellus Formation
64
in the eastern United States (Soeder and Kappel, 2009; Kerr, 2010; Kargbo et al.,
65
2010; Lee et al., 2011) raises important questions about scientific decisionmaking,
66
environmental protection, public health and water resources (US GAO, 2012). Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 67
For this reason, in New York State , the governor has imposed a de-facto
68
moratorium on the method for gas production: "hydrofracking" or high-volume
69
hydraulic fracturing (HVHF), pending completion of further environmental and
70
public health studies. An ongoing state regulatory process has resulted in a public
71
document, the draft supplemental generic environmental impact study (dSGEIS
72
available at http://www.dec.ny.gov/energy/58440.html), which is currently
73
undergoing review by the New York State Departmental of Environmental
74
Conservation (DEC). In contrast to a point-by-point evaluation of that lengthy
75
draft dSGEIS, this paper focuses on the interaction among scientific and technical
76
issues of local environmental protection and other relevant spheres of concern to
77 3 humankind such as energy policy, land use, economics, regulation, politics and
78
ultimately global climate change. These interactions really determine how to
79
prioritize risks for health and wellbeing of affected populations. Formal risk
80
assessment is premature without analysis of such interactions and initial
81
screening of risk (AEA Technology 2012). This work therefore involves more the
82
problem formulation of risk as opposed to formal risk assessment (US EPA, 2003;
83
US NRC, 2008). 84 16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
4
Public controversy over the hydraulic fracturing methods necessary for
85
unconventional gas production has stimulated numerous highly focused and
86
conflicting contributions in the literature on narrow technical issues (Schon 2011;
87
Pyron 2011; Osborn et al. 2011; Howarth et al., 2011; Warner et al., 2012;
88
O'Sullivan and Paltsev, 2012). While clarity on narrow issues is important, a sole
89
focus on scientific and technical aspects is unlikely to have prevented such recent
90
environmental catastrophes as the Fukushima Daiichi nuclear plant
91
explosions/tsunami disaster in Japan or the Deep Horizon oil-well blowout in the
92
Gulf of Mexico. A better analogy to potential unforeseen impacts of Marcellus
93
natural gas production might be the slower-developing but even more disastrous
94
epidemic of arsenic poisoning due to widespread consumption of contaminated
95
groundwater in Bangladesh (Dhar et al., 1997). Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 These unforeseen catastrophes
96
result not just from scientific uncertainty but more importantly from an avoidable
97
reactive cascade of events driven by economic and political choices. 98
While some have touched on broader considerations concerning Marcellus
99
shale gas production (Howarth and Ingraffea, 2011; Engelder, 2011),
100
scientifically-based decisionmaking needs to explicitly account for non-scientific
101
issues related to human activities. Despite general studies of the intersection
102 Public controversy over the hydraulic fracturing methods necessary for
85
unconventional gas production has stimulated numerous highly focused and
86
conflicting contributions in the literature on narrow technical issues (Schon 2011;
87
Pyron 2011; Osborn et al. 2011; Howarth et al., 2011; Warner et al., 2012;
88
O'Sullivan and Paltsev, 2012). While clarity on narrow issues is important, a sole
89
focus on scientific and technical aspects is unlikely to have prevented such recent
90
environmental catastrophes as the Fukushima Daiichi nuclear plant
91
explosions/tsunami disaster in Japan or the Deep Horizon oil-well blowout in the
92
Gulf of Mexico. A better analogy to potential unforeseen impacts of Marcellus
93
natural gas production might be the slower-developing but even more disastrous
94
epidemic of arsenic poisoning due to widespread consumption of contaminated
95
groundwater in Bangladesh (Dhar et al., 1997). These unforeseen catastrophes
96
result not just from scientific uncertainty but more importantly from an avoidable
97
reactive cascade of events driven by economic and political choices. 98 While some have touched on broader considerations concerning Marcellus
99
shale gas production (Howarth and Ingraffea, 2011; Engelder, 2011),
100
scientifically-based decisionmaking needs to explicitly account for non-scientific
101
issues related to human activities. Despite general studies of the intersection
102 4 between energy use and water resources (Harte and El-Gasseir 1978; Harte 1983;
103
Gleick 1994, Jenner and Lamadrid, 2012), there are few pertinent tradeoff
104
analyses in specific situations (Rahm and Riha 2012; Stephenson et al., 2012). 105
Scientists have particular responsibilities (Hansen 2007, Maxim and van der
106
Sluijs, 2011) to help develop timely, systematic approaches that consider
107
overlapping scientific, technical, environmental, sociological, economic and
108
political considerations, evaluate their relative importance for the issue at hand,
109
and thereby formulate recommendations for policy decisionmaking. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 110
The novel aspect of this review is that it combines an analysis of the
111
scientific and technical aspects of hydraulic fracturing to produce natural gas from
112
the Marcellus formation in New York state compared to the risks of endangering
113
the New York City (NYC) water supply (Figure 1), while also considering the
114
broader impacts on global climate change and even more critical issues regarding
115
energy tradeoffs, economics and political regulation. Because of its similarity to
116
larger and equally time-sensitive natural resource issues that require policy
117
responses, like global climate change, decisions about Marcellus shale gas drilling
118
have much larger implications beyond the U.S. Mid-Atlantic region. Although
119
the United States is currently the only country with widespread unconventional
120
natural gas production using hydraulic fracturing, many countries are thought to
121
have significant unconventional natural gas potential (Rogers 2011), and concerns
122
have been expressed by the European Union about the potential environmental
123
and human health risks of such production (AEA Technology 2012). 124
Starting with a narrow perspective on the scientific (hydrological and
125
chemical) and technical aspects of water resources and natural gas drilling in the
126
M
ll
h l
th
d l i
l
d di t
b
d
l
i
l f
t
127 Starting with a narrow perspective on the scientific (hydrological and
125
chemical) and technical aspects of water resources and natural gas drilling in the
126
Marcellus shale, the underlying land-disturbance and geological factors are
127 5 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
analyzed, then the broader energy and economic aspects, followed by the
128
regulatory and ultimately political foundations of this issue. The intent is to
129
develop a proactive framework for rational, timely decision-making by weighing
130
relative merits in the face of incomplete information, and seek a broader
131
perspective on common ground for consensus in the case of Marcellus shale
132
drilling in and near the watersheds that provide New York City water supply. 133
2. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 Hydrological and chemical aspects
134
2.1 Protection of the New York City water supply
135
The almost nine million residents of New York City have been supplied
136
since 1915 with drinking water from the Catskill-Delaware watersheds west of the
137
Hudson River, which directly overlie the northeastern corner (about 4100 km2 or
138
8.5%) of the Marcellus shale subcrop that extends from the Appalachians north
139
across the southern tier of New York State. Land surface runoff from these
140
watersheds drains into several reservoirs from which water is transported via
141
aqueducts and tunnels to the city (Figure 1). New York City has the largest
142
unfiltered water supply (NYC DEP, 2010) among United States large cities, most
143
of which operate expensive water filtration and treatment facilities. 144
Over the last couple decades, New York City Department of
145
Environmental Protection (NYC DEP) has demonstrated every five years that the
146
system meets strict criteria according to the US Environmental Protection Agency
147
(US EPA) Surface Water Treatment Rule, thereby enabling a federal Filtration
148
Avoidance Determination (FAD) (NYC DEP, 2010). To accomplish this, the
149
NYC DEP is pursuing an aggressive preventive campaign of subsidies for
150
i
lt
l b
t
t
ti
(BMP ) i
ll b
ti
ith l
d
151 6
Over the last couple decades, New York City Department of
145
Environmental Protection (NYC DEP) has demonstrated every five years that the
146
system meets strict criteria according to the US Environmental Protection Agency
147
(US EPA) Surface Water Treatment Rule, thereby enabling a federal Filtration
148
Avoidance Determination (FAD) (NYC DEP, 2010). To accomplish this, the
149
NYC DEP is pursuing an aggressive preventive campaign of subsidies for
150
agricultural best management practices (BMPs), in collaboration with large and
151 6 small landowners, to maintain contaminants such as excess coliforms, pathogens,
152
turbidity and nutrients considerably below federally-mandated levels (NYC DEP,
153
2010). Water quality is carefully managed in all watersheds, and protection
154
efforts in the largest and westernmost Cannonsville watershed have been topics of
155
research and modeling (Bryant et al., 2008, Rao et al. 2009). 156
157
2.2 New water resources threats
158
In these watersheds supplying New York City potable water, drilling for
159
Marcellus shale gas development presents additional threats to water resources,
160
which the current BMPs cannot mitigate. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 186
Although methane, the major component of natural gas, occurs naturally in
187
shallow groundwater (Molofsky et al., 2011) in northern Pennsylvania and
188
southern New York, controversy surrounds the role of gas drilling in shallow
189
aquifer contamination. Careful geochemical analysis, and methane and strontium
190
isotopic signatures (Osborn et al., 2011; Warner et al. 2012) have now shown that
191
some methane and groundwater salinization in shallow drinking water wells is
192
attributable to natural upward seepage of natural gas and brine, respectively, from
193
reservoirs like the Marcellus formation. This has important implications for large-
194
scale development of natural gas drilling using thousands of boreholes reaching
195
depths of over a mile below land surface, as described later. 196
2.4 Surface water impacts
197
The greater threat appears to be to surface water because of local water
198 synthetic-based fluids (SBF) are often preferable in shales because they are more
176
stable, less reactive and support the borehole better. OBF-saturated rock
177
fragments, or cuttings, removed from the borehole during drilling can be a
178
significant source of contaminants to the environment because they do not
179
degrade readily (Sadiq et al., 2003). Few studies have evaluated relative merits
180
and risks of these various fluids and rock cuttings, however the high levels of
181
natural radioactivity in cuttings from some Marcellus shale well borings have
182
raised concern (Kargbo et al., 2010, Lee et al., 2011). 183 Clear evidence for past contamination by drilling fluids is slim (Kargbo et
184
al., 2010; US EPA 2011), but natural gas that seeps into shallow groundwater
185
presents an explosion risk if it degasses into confined areas such as basements. 186
Although methane, the major component of natural gas, occurs naturally in
187
shallow groundwater (Molofsky et al., 2011) in northern Pennsylvania and
188
southern New York, controversy surrounds the role of gas drilling in shallow
189
aquifer contamination. Careful geochemical analysis, and methane and strontium
190
isotopic signatures (Osborn et al., 2011; Warner et al. 2012) have now shown that
191
some methane and groundwater salinization in shallow drinking water wells is
192
attributable to natural upward seepage of natural gas and brine, respectively, from
193
reservoirs like the Marcellus formation. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 Extracting natural gas from shale (Kerr,
161
2010; Lee et al., 2011) involves the latest drilling industry techniques of
162
horizontal directional drilling and high-volume hydraulic fracturing (HVHF,
163
hydrofracturing or "hydrofracking") (US GAO, 2012; Kargbo et al., 2010; Bybee,
164
2007). These techniques use 7500-38,000 m3 of water per well (Kargbo et al. 165
2010; US DOE-NETL 2010) and various chemical additives (Waxman et al. 166
2011) injected at high pressures to open and force sand into fractures in the rock,
167
enabling release of gas. Although the Marcellus shale lies hundreds to thousands
168
of meters below land surface, these drilling activities present a threat to both
169
groundwater and surface water resources. 170
2.3 Well drilling process
171
The well-drilling process itself, developed from oil and mineral
172
exploration, uses a clay slurry as lubricant. The base fluid can be water, oil, or a
173
synthetic such as vegetable esters or olefins (Sadiq et al., 2003). Although water-
174
based fluids (WBF) are more environmentally benign, oil-based fluids (OBF) or
175 7
The well-drilling process itself, developed from oil and mineral
172
exploration, uses a clay slurry as lubricant. The base fluid can be water, oil, or a
173
synthetic such as vegetable esters or olefins (Sadiq et al., 2003). Although water-
174
based fluids (WBF) are more environmentally benign, oil-based fluids (OBF) or
175 7 synthetic-based fluids (SBF) are often preferable in shales because they are more
176
stable, less reactive and support the borehole better. OBF-saturated rock
177
fragments, or cuttings, removed from the borehole during drilling can be a
178
significant source of contaminants to the environment because they do not
179
degrade readily (Sadiq et al., 2003). Few studies have evaluated relative merits
180
and risks of these various fluids and rock cuttings, however the high levels of
181
natural radioactivity in cuttings from some Marcellus shale well borings have
182
raised concern (Kargbo et al., 2010, Lee et al., 2011). 183
Clear evidence for past contamination by drilling fluids is slim (Kargbo et
184
al., 2010; US EPA 2011), but natural gas that seeps into shallow groundwater
185
presents an explosion risk if it degasses into confined areas such as basements. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 232 16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
10
sustains current streamflow, also adversely affecting watershed storage. 232
Water quality impacts from natural gas exploitation depend on the
233
constituents of produced wastewater from drilling operations (Fakhru’l-Razi et al.,
234
2009), which include both additives (Waxman et al., 2011; Aminto and Olson
235
2012) and natural contaminants such as minerals and radionuclides in the
236
Marcellus (Kargbo et al., 2010; Lee et al., 2011). As with coalbed methane
237
(CBM) extraction (Clarke 1996; Healy et al., 2008), produced waste-water poses
238
the most important environmental risks, often having total dissolved solids (TDS)
239
concentrations in the tens to hundreds of thousands of mg/L (US GAO, 2012). Of
240
the total volumes of water needed for hydrofracturing the Marcellus shale, gas
241
production causes 10-40% return flow up the borehole (Gregory et al., 2011;
242
Hazen and Sawyer, 2009) although an increasing proportion of this produced
243
water is now recycled (US GAO, 2012). Recycling or disposal of the remaining
244
waste brines will likely require dilution and treatment because deep reinjection, a
245
common method in oilfields, is more expensive. 246
Industry accounts of hydrofracturing de-emphasize the amount of chemical
247
additives, many of which are carcinogenic, as a proportion of hydrofracturing
248
water (1-2% by volume). However, over the 20 year timeframe projected for the
249 Water quality impacts from natural gas exploitation depend on the
233
constituents of produced wastewater from drilling operations (Fakhru’l-Razi et al.,
234
2009), which include both additives (Waxman et al., 2011; Aminto and Olson
235
2012) and natural contaminants such as minerals and radionuclides in the
236
Marcellus (Kargbo et al., 2010; Lee et al., 2011). As with coalbed methane
237
(CBM) extraction (Clarke 1996; Healy et al., 2008), produced waste-water poses
238
the most important environmental risks, often having total dissolved solids (TDS)
239
concentrations in the tens to hundreds of thousands of mg/L (US GAO, 2012). Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 This has important implications for large-
194
scale development of natural gas drilling using thousands of boreholes reaching
195
depths of over a mile below land surface, as described later. 196 2.4 Surface water impacts
197
The greater threat appears to be to surface water because of local water
198
demand and wastewater disposal, although groundwater concerns are revisited
199 8 later in conjunction with bedrock integrity and well abandonment issues. There is
200
also a need at land surface for effective management and safe disposal of the tens
201
of thousands of cubic meters of water and additives (sand and chemicals) needed
202
per well for hydrofracturing. The final environmental impact assessment
203
commissioned by the New York City Department of Environmental Protection
204
(Hazen and Sawyer, 2009) assumes a full build-out of up to 6000 hydrofractured
205
Marcellus shale gas wells in the NYC water supply watersheds and presents a
206
comprehensive evaluation of the environmental risks, the most important of which
207
are addressed here. 208 Although few have examined water usage impacts of shale-gas drilling
209
(O'Shea, 2011; Rahm and Riha 2012), natural gas production from shales
210
elsewhere, using similar drilling methods, provides examples of likely impacts on
211
surface water availability. Permitting requests of up to 2.5 x 105 m3 of water per
212
year over 10 years have been reported for the Barnett shale in Texas (Rahm,
213
2011). After initial hydrofracturing, wells decline rapidly in gas output after the
214
first year or two, and repeated hydrofracturing and infill drilling is used to
215
maximize ultimate recovery (as is now happening in Texas), which could increase
216
water demand (US DOE-NETL 2010). More comprehensive analyses (Elcock,
217
2010) indicate that increased exploitation of unconventional energy resources like
218
gas, and their use for electrical power generation will require significant growth in
219
U.S. freshwater use. 220 9
For example, in the Susquehanna River basin, overlying the Marcellus in
221
western NY state (Figure 1), a recent analysis (Rahm and Riha 2012) suggested
222
that surface water availability in all but the largest rivers, and effective treatment
223
capacity is inadequate to support the drilling of hundreds of gas production wells
224 9 per year. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 In the NYC supply watersheds, projected diversions of water needed for
225
hydrofracturing a maximum build-out of wells could range from 0.8 to up to 1.5
226
x107 m3 per year of additional demand (Hazen and Sawyer, 2009). The higher
227
level of diversion represents 1000x the amount anticipated to require significant
228
expansion of NYC water supply storage for maintaining supply safety (Flexible
229
Flow Management Program, 2008). Alternatively, groundwater withdrawals to
230
supply such hydrofracturing would deplete shallow aquifers and baseflow that
231
sustains current streamflow, also adversely affecting watershed storage. 232
Water quality impacts from natural gas exploitation depend on the
233
constituents of produced wastewater from drilling operations (Fakhru’l-Razi et al.,
234
2009), which include both additives (Waxman et al., 2011; Aminto and Olson
235
2012) and natural contaminants such as minerals and radionuclides in the
236
Marcellus (Kargbo et al., 2010; Lee et al., 2011). As with coalbed methane
237
(CBM) extraction (Clarke 1996; Healy et al., 2008), produced waste-water poses
238
the most important environmental risks, often having total dissolved solids (TDS)
239
concentrations in the tens to hundreds of thousands of mg/L (US GAO, 2012). Of
240
the total volumes of water needed for hydrofracturing the Marcellus shale, gas
241
production causes 10-40% return flow up the borehole (Gregory et al., 2011;
242
Hazen and Sawyer, 2009) although an increasing proportion of this produced
243
water is now recycled (US GAO, 2012). Recycling or disposal of the remaining
244
waste brines will likely require dilution and treatment because deep reinjection, a
245
common method in oilfields, is more expensive. 246
Industry accounts of hydrofracturing de-emphasize the amount of chemical
247
additives, many of which are carcinogenic, as a proportion of hydrofracturing
248
water (1-2% by volume). However, over the 20 year timeframe projected for the
249 per year. In the NYC supply watersheds, projected diversions of water needed for
225
hydrofracturing a maximum build-out of wells could range from 0.8 to up to 1.5
226
x107 m3 per year of additional demand (Hazen and Sawyer, 2009). The higher
227
level of diversion represents 1000x the amount anticipated to require significant
228
expansion of NYC water supply storage for maintaining supply safety (Flexible
229
Flow Management Program, 2008). Alternatively, groundwater withdrawals to
230
supply such hydrofracturing would deplete shallow aquifers and baseflow that
231
sustains current streamflow, also adversely affecting watershed storage. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 Of
240
the total volumes of water needed for hydrofracturing the Marcellus shale, gas
241
production causes 10-40% return flow up the borehole (Gregory et al., 2011;
242
Hazen and Sawyer, 2009) although an increasing proportion of this produced
243
water is now recycled (US GAO, 2012). Recycling or disposal of the remaining
244
waste brines will likely require dilution and treatment because deep reinjection, a
245
common method in oilfields, is more expensive. 246 Industry accounts of hydrofracturing de-emphasize the amount of chemical
247
additives, many of which are carcinogenic, as a proportion of hydrofracturing
248
water (1-2% by volume). However, over the 20 year timeframe projected for the
249 10 11
development of Marcellus shale gas drilling, the total mass of chemical additives
250
(not including sand proppant) amounts to several hundred tons per day, and over
251
500 tons per day if repeated hydrofracturing is used to delay inevitable well
252
production declines (Hazen and Sawyer, 2009). In addition to diesel fuel, until
253
recently used in hydrofracturing (Kargbo et al., 2010), other less-well-known
254
hydrocarbon additives are hazardous to human and environmental health
255
(Waxman et al. 2011; Aminto and Olson 2012). These include biocides
256
(Struchtemeyer et al. 2012), endocrine-disrupting compounds, mutagens,
257
teratogens and other toxins that present human health risks at very low dosages
258
with long-term exposure (Hazen and Sawyer, 2009). The mere introduction and
259
usage of hundreds of tons per day, over decades, of such toxic chemical additives
260
in watersheds that provide drinking water to millions of New York City residents,
261
is a significant cause of concern. 262
Although hydrofracturing fluids can be highly variable in their
263
composition depending on the geology and fracturing outcome desired, most of
264
these additives are unregulated with regard to drinking water standards and do not
265
have maximum contaminant levels (MCLs) established by federal (US EPA) or
266
state (NYS Dept of Health) authorities. A recent modeling study (Aminto and
267
Olson 2012) of a hypothetical spill into air, water and soil of additives used in
268
Pennsylvania Marcellus hydrofracturing has shown that resulting concentrations
269
in a receiving surface water body exceed the 5 μg/L MCL standard for many
270
organic compounds in New York state. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 Furthermore, the environmental impact
271
study commissioned by the NYC Dept of Environmental Protection (Hazen and
272
Sawyer, 2009) presents two dilution scenarios in which acute spills of
273
hydrofracturing chemicals from a dozen wells or less could threaten the volumes
274 development of Marcellus shale gas drilling, the total mass of chemical additives
250
(not including sand proppant) amounts to several hundred tons per day, and over
251
500 tons per day if repeated hydrofracturing is used to delay inevitable well
252
production declines (Hazen and Sawyer, 2009). In addition to diesel fuel, until
253
recently used in hydrofracturing (Kargbo et al., 2010), other less-well-known
254
hydrocarbon additives are hazardous to human and environmental health
255
(Waxman et al. 2011; Aminto and Olson 2012). These include biocides
256
(Struchtemeyer et al. 2012), endocrine-disrupting compounds, mutagens,
257
teratogens and other toxins that present human health risks at very low dosages
258
with long-term exposure (Hazen and Sawyer, 2009). The mere introduction and
259
usage of hundreds of tons per day, over decades, of such toxic chemical additives
260
in watersheds that provide drinking water to millions of New York City residents,
261
is a significant cause of concern. 262 Although hydrofracturing fluids can be highly variable in their
263
composition depending on the geology and fracturing outcome desired, most of
264
these additives are unregulated with regard to drinking water standards and do not
265
have maximum contaminant levels (MCLs) established by federal (US EPA) or
266
state (NYS Dept of Health) authorities. A recent modeling study (Aminto and
267
Olson 2012) of a hypothetical spill into air, water and soil of additives used in
268
Pennsylvania Marcellus hydrofracturing has shown that resulting concentrations
269
in a receiving surface water body exceed the 5 μg/L MCL standard for many
270
organic compounds in New York state. Furthermore, the environmental impact
271
study commissioned by the NYC Dept of Environmental Protection (Hazen and
272
Sawyer, 2009) presents two dilution scenarios in which acute spills of
273
hydrofracturing chemicals from a dozen wells or less could threaten the volumes
274 11 12
of water contained in several major reservoirs (assumed partial mixing, reservoirs
275
at low levels). Resulting exceedances of the US EPA MCL in those reservoirs
276
highlights the severe risk posed by large-scale Marcellus shale gas exploitation. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 277
The risk is likely even greater, and more insidious, of numerous small site spills
278
which go undetected and eventually enter drinking water reservoirs, with
279
irreversible consequences. Aggressive enforcement of BMPs for pollution
280
prevention, stormwater control, waste minimization and handling could reduce,
281
but never eliminate such risk. 282
283
3. Land-disturbance, and geologic factors
284
The expansion of similar unconventional natural gas drilling in other areas
285
of the United States has been dramatic (Figure 2) in the last decade. While future
286
growth is difficult to predict (recent production drilling in the Barnett shale has
287
since lagged due to declines in natural gas prices (Rogers, 2011)), projected
288
expansion of the Marcellus shale drilling from neighboring Pennsylvania into
289
New York is likely to follow similar trends, starting from the date that HVHF
290
permits are issued. To compensate for the different regional extents of the shales
291
in the locations illustrated, the data (adapted from Hazen and Sawyer 2009) have
292
been normalized for well density per 2600 km2 (1000 mi2). However, even this
293
data reduction cannot fully account for evolution in well densities as natural gas
294
fields are developed, because future well siting and infill development depends on
295
production records of existing wells. Nevertheless, since exponential trends
296
cannot be sustained, applying a logistic function fitted to existing data for the
297
much smaller Barnett Formation in Texas, but offset in time, suggests that
298 of water contained in several major reservoirs (assumed partial mixing, reservoirs
275
at low levels). Resulting exceedances of the US EPA MCL in those reservoirs
276
highlights the severe risk posed by large-scale Marcellus shale gas exploitation. 277
The risk is likely even greater, and more insidious, of numerous small site spills
278
which go undetected and eventually enter drinking water reservoirs, with
279
irreversible consequences. Aggressive enforcement of BMPs for pollution
280
prevention, stormwater control, waste minimization and handling could reduce,
281
but never eliminate such risk. 282 The expansion of similar unconventional natural gas drilling in other areas
285
of the United States has been dramatic (Figure 2) in the last decade. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 While future
286
growth is difficult to predict (recent production drilling in the Barnett shale has
287
since lagged due to declines in natural gas prices (Rogers, 2011)), projected
288
expansion of the Marcellus shale drilling from neighboring Pennsylvania into
289
New York is likely to follow similar trends, starting from the date that HVHF
290
permits are issued. To compensate for the different regional extents of the shales
291
in the locations illustrated, the data (adapted from Hazen and Sawyer 2009) have
292
been normalized for well density per 2600 km2 (1000 mi2). However, even this
293
data reduction cannot fully account for evolution in well densities as natural gas
294
fields are developed, because future well siting and infill development depends on
295
production records of existing wells. Nevertheless, since exponential trends
296
cannot be sustained, applying a logistic function fitted to existing data for the
297
much smaller Barnett Formation in Texas, but offset in time, suggests that
298 12 3.1
Land use changes
301
Such large-scale exploitation of natural gas resources from the underlying
302
Marcellus shale would necessarily fragment the largely rural, forested and
303
agricultural landscape near the NYC water supply watersheds. Of immediate
304
concern are land use changes such as the construction of roads, well-pads and
305
pipelines that accompany intensive natural gas drilling. While the impact of each
306
individual well drilling operation is relatively minor, the cumulative impact of
307
thousands of wells scattered across the watershed threatens the quality of runoff to
308
streams and water supply reservoirs over time (Mitchell and Casman, 2011). 309
Experience in other shale-gas-producing areas shows that a density of 3.5 or more
310
wells per km2 can be anticipated in highly productive areas for fully developed
311
gas fields (Hazen and Sawyer, 2009; US DOE NETL 2010), although these
312
densities have not been reached to date in most unconventional fields (Figure 2). 313
Multiple directional wells are expected to be drilled from each wellpad for natural
314
gas production in the Marcellus shale. Each wellpad is likely to have a footprint
315
of about 2.8 ha., part of which will remain in operation for the well’s productive
316
life of up to 20 years. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 317
Industrial operation involving heavy truck traffic requires a compacted
318 Industrial operation involving heavy truck traffic requires a compacted
318
gravel substrate, leading to increased stormwater runoff and erosion potential
319
(Hazen and Sawyer, 2009). Each well is estimated to require 900 to 1300 truck
320
access trips, up to 6600 for multiple horizontal well pads (NTC, 2009), resulting
321
in tens to hundreds of thousands of additional truck trips for many wells over a
322 13 large area. In Wyoming, Huntington and Ksaibati (2009) showed that county
323
roads suffered severe damage from heavy truck traffic associated with well
324
drilling, with one half of road repairs concentrated on only 15% of the roads. 325
These impacts in a semiarid environment likely underestimate the damage and
326
repair costs necessary for a more humid climate like New York. 327 3.2
Aquifer, well and bedrock integrity
328
Contamination of shallow aquifers, used for individual home water supply
329
in rural areas, could result from either infiltration of wastewater (Healy et al.,
330
2008) or subsurface leakage of drilling fluids or natural gas through or along drill
331
casings (ODNR, 2008; US EPA 2011). Standard techniques in well drilling, such
332
as cementing casing pipe, are increasingly scrutinized (Ladva et al., 2005) since
333
effective seals may not be achieved in many cases (Harrison, 1985; US EPA,
334
2011). Although casing defects and the subsurface migration of natural gas
335
through fractures are rare, the consequences can be catastrophic, resulting in
336
surface explosions over 11 km away from a leaking deep gas storage well in
337
Kansas in 2001 (Nissen et al. 2004a, 2004b; Watney et al. 2003). Regulatory
338
oversight of natural gas and oil-well seals has lagged the proliferation of well
339
borings in the 20th century, such that even in a highly regulated operation in
340
Alberta, Canada, up to 10% of existing wells have been found to have inadequate
341
seals, though more recent failure rates are down to 2% (Watson and Bachu,
342
2009). Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 343
The geology underlying the NYC water supply watersheds (Hazen and
344
Sawyer 2009; US DOE NETL 2010) consists of thin surficial deposits and
345 3.2
Aquifer, well and bedrock integrity
328
Contamination of shallow aquifers, used for individual home water supply
329
in rural areas, could result from either infiltration of wastewater (Healy et al.,
330
2008) or subsurface leakage of drilling fluids or natural gas through or along drill
331
casings (ODNR, 2008; US EPA 2011). Standard techniques in well drilling, such
332
as cementing casing pipe, are increasingly scrutinized (Ladva et al., 2005) since
333
effective seals may not be achieved in many cases (Harrison, 1985; US EPA,
334
2011). Although casing defects and the subsurface migration of natural gas
335
through fractures are rare, the consequences can be catastrophic, resulting in
336
surface explosions over 11 km away from a leaking deep gas storage well in
337 The geology underlying the NYC water supply watersheds (Hazen and
344
Sawyer, 2009; US DOE NETL 2010), consists of thin surficial deposits and
345
Devonian-age sedimentary rocks (sandstone, shale, siltstone and limestone) that
346 14 include the Marcellus shale. Although conventional hydrogeologic analyses
347
assume extremely slow flow rates through these rocks based on equivalent porous
348
medium assumptions (ICF International, 2009), more sophisticated detailed
349
studies (e.g., Runkel et al., 2006) show that even non-karstic sedimentary rocks
350
(sandstone and shale) contain significant brittle fractures, which allow faster
351
preferential flowpaths along bedding-planes over distances of kilometers. While
352
solute transport is typically slower than pressurized gas flow in such fractures,
353
both are highly unpredictable and essentially undetectable, barring major events. 354
The limited flow between deep and shallow formations that has already been
355
shown using isotope tracers (Warner at al., 2012) can be locally enhanced by
356
interconnection of existing preferential flowpaths by well drilling and
357
hydrofracturing, which destabilize existing hydraulic gradients by changing
358
pressure regimes. 359
Understanding of flow and leakage through such heterogeneous connected
360
fracture networks even in sedimentary rock requires new paradigms (Eaton, 2006,
361
and references therein), and is at the forefront of hydrogeologic research
362
especially for purposes of geological carbon sequestration (DOE NETL, 2010). Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 363
Considerable information on brittle fault structures and linear features extending
364
laterally for kilometers and vertically for thousands of meters has been
365
documented by engineering studies for the emplacement of the current water
366
supply tunnels (Figure 1) that transport water outside the watersheds to New York
367
City (Hazen and Sawyer, 2009). The NYS DEC dSGEIS study anticipates buffer
368
zones between sensitive resources or infrastructure and permitted natural gas
369
drilling locations, but the widths of those buffer zones (100s of m) are well short
370
of the known lengths of many mapped linear fault features. 371 include the Marcellus shale. Although conventional hydrogeologic analyses
347
assume extremely slow flow rates through these rocks based on equivalent porous
348
medium assumptions (ICF International, 2009), more sophisticated detailed
349
studies (e.g., Runkel et al., 2006) show that even non-karstic sedimentary rocks
350
(sandstone and shale) contain significant brittle fractures, which allow faster
351
preferential flowpaths along bedding-planes over distances of kilometers. While
352
solute transport is typically slower than pressurized gas flow in such fractures,
353
both are highly unpredictable and essentially undetectable, barring major events. 354
The limited flow between deep and shallow formations that has already been
355
shown using isotope tracers (Warner at al., 2012) can be locally enhanced by
356
interconnection of existing preferential flowpaths by well drilling and
357
hydrofracturing, which destabilize existing hydraulic gradients by changing
358
pressure regimes. 359 Understanding of flow and leakage through such heterogeneous connected
360
fracture networks even in sedimentary rock requires new paradigms (Eaton, 2006,
361
and references therein), and is at the forefront of hydrogeologic research
362
especially for purposes of geological carbon sequestration (DOE NETL, 2010). 363
Considerable information on brittle fault structures and linear features extending
364
laterally for kilometers and vertically for thousands of meters has been
365
documented by engineering studies for the emplacement of the current water
366
supply tunnels (Figure 1) that transport water outside the watersheds to New York
367
City (Hazen and Sawyer, 2009). The NYS DEC dSGEIS study anticipates buffer
368
zones between sensitive resources or infrastructure and permitted natural gas
369
drilling locations, but the widths of those buffer zones (100s of m) are well short
370
of the known lengths of many mapped linear fault features. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 371 15 16
The New York City DEP has expressed concern about possible impacts on
372
this tunnel infrastructure from extensive hydrofracturing in close proximity
373
(Hazen and Sawyer, 2009). A major issue is that the tunnels extend up to 8 km
374
outside the hydrographic boundaries of the watersheds (Figure 1) and therefore
375
are not entirely included in the currently proposed protected area described in the
376
NYS DEC dSGEIS. These bedrock water tunnels lie 100-300 m below grade
377
(well below the water table), are concrete-lined, and have served well for decades,
378
but are designed to retain transported potable water, not resist external
379
overpressures. In fact, existing chronic leakage through tunnel liner cracks
380
indicates they would be vulnerable to additional damage from changing external
381
stresses, accumulation of explosive natural gas in access and maintenance
382
infrastructure and even fracture-flow contamination at occasional low operating
383
pressures (atmospheric) due to groundwater inflow (Hazen and Sawyer 2009). 384
385
4. Energy and economic aspects
386
Analysis of possible impacts of Marcellus shale gas drilling must consider
387
the resulting tradeoffs in the larger context of global climate change driven by
388
fossil fuel GHG emissions. Specifically, compared to current domination of coal-
389
fired electrical generation in the United States, what are the environmental
390
consequences that may be avoided by a potential substitution with natural gas? In
391
fact, a recent study (Lu et al. 2012) has shown that such substitution has already
392
contributed to a reduction in CO2 emissions from US electrical generation from
393
2008 to 2009. Potential economic and environmental benefits of natural gas
394
drilling, representing a desirable transition to a cleaner energy economy, need to
395
be weighed against the economic and environmental costs, that are not necessarily
396 16 limited to New York state. 397
398
4.1
Global greenhouse gas (GHG) emissions
399
Closure of coal-fired power plants and their substitution by higher
400
efficiency, lower-emission electrical generation, like using natural gas, has been
401
identified as one of the principal options to reduce greenhouse gas (GHG)
402
emissions (Pacala and Socolow, 2004). Other than natural gas, there is no other
403
readily deployable energy generation technology that provides the necessary
404
replacement base load to balance the intermittency of renewable energy
405
generation like wind. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 Therefore, increased natural gas production from the
406
Marcellus would be beneficial in this regard. But economic and cleaner energy
407
benefits of natural gas may be illusory if only GHG emissions at the point of
408
combustion are considered (Hughes 2011). Furthermore, natural gas consists of
409
mostly methane, a more powerful driver of global climate change than carbon
410
dioxide (Howarth et al., 2011, Shindell et al. 2009). 411 limited to New York state. 397 Due to poor regulation, production and pipeline transportation of natural
412
gas causes numerous unaccounted-for sources of GHG emissions to the
413
atmosphere, the magnitude of which is under debate (Howarth et al. 2011;
414
O'Sullivan and Paltsev 2012). Unconventional gas exploitation causes methane
415
emissions from the wellhead during and after the drilling is completed, and while
416
the gas is processed and transported. These fugitive methane emissions are poorly
417
constrained (US EPA, 2010), but could conservatively amount to up to 7.9% of
418
lifecycle well production (Howarth et al., 2011). While modeling studies of GHG
419
emissions from shale gas production with differing assumptions are proliferating
420 17 (e.g. Jiang et al. 2011; Weber and Clavin 2012), there is a shortage of actual field
421
studies. However, recent work (Petron et al., 2012), focusing on VOCs and
422
methane from a natural gas field in Colorado, showed that the uncertainty of and
423
actual GHG emissions percentages are higher than many assumed values in the
424
models, and closer to those of Howarth et al. (2011). 425
426
4.2 Substitution for coal, and public health impacts
427
Many existing "lifecycle" analyses in the debate over the environmental
428
impact of natural gas do not take into account the default (current) GHG
429
emissions of coal-fired generation (Weber and Clavin, 2012), and others do not
430
evaluate environmental impact other than GHG emissions (Howarth et al., 2011;
431
Hultman et al. 2011; Burnham et al., 2012) in their assessment of different fuels
432
for electrical generation. However, non-GHG impacts dominate current U.S. 433
electricity production, almost half of which is generated using coal, and 34% of
434
that capacity is from plants more than 40 years old, with little to no modern
435
pollution controls, such as scrubbers or other technology (Hughes, 2011). Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 Many
436
existing "lifecycle" analyses focusing on GHG impacts are therefore too narrow
437
for comparison of tradeoffs related to increased Marcellus shale gas production. 438
In fact, in contrast to natural gas production, most "externalities" or
439
environmental damages from coal-fired electricity are not related to climate
440
change (Levy et al., 2009; US NRC, 2009; Epstein et al., 2011). These impacts
441
are largely due to air pollution from sulfates and other particulates and related
442
cumulative public health consequences. Others are related to water
443
contamination due to coal sludge storage accidents, and ecological and economic
444 (e.g. Jiang et al. 2011; Weber and Clavin 2012), there is a shortage of actual field
421
studies. However, recent work (Petron et al., 2012), focusing on VOCs and
422
methane from a natural gas field in Colorado, showed that the uncertainty of and
423
actual GHG emissions percentages are higher than many assumed values in the
424
models, and closer to those of Howarth et al. (2011). 425 18
4.2 Substitution for coal, and public health impacts
427
Many existing "lifecycle" analyses in the debate over the environmental
428
impact of natural gas do not take into account the default (current) GHG
429
emissions of coal-fired generation (Weber and Clavin, 2012), and others do not
430
evaluate environmental impact other than GHG emissions (Howarth et al., 2011;
431
Hultman et al. 2011; Burnham et al., 2012) in their assessment of different fuels
432
for electrical generation. However, non-GHG impacts dominate current U.S. 433
electricity production, almost half of which is generated using coal, and 34% of
434
that capacity is from plants more than 40 years old, with little to no modern
435
pollution controls, such as scrubbers or other technology (Hughes, 2011). Many
436
existing "lifecycle" analyses focusing on GHG impacts are therefore too narrow
437
for comparison of tradeoffs related to increased Marcellus shale gas production. 438
In fact, in contrast to natural gas production, most "externalities" or
439
environmental damages from coal-fired electricity are not related to climate
440
change (Levy et al., 2009; US NRC, 2009; Epstein et al., 2011). These impacts
441
are largely due to air pollution from sulfates and other particulates and related
442
cumulative public health consequences. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 Others are related to water
443
contamination due to coal sludge storage accidents, and ecological and economic
444 18 costs (including opportunity costs) of land transformation due to mountain-top
445
removal (MTR) in Appalachia (Epstein et al., 2011). Estimates of non-climate-
446
related total hidden annual costs of coal for electrical generation range from $62
447
billion (US NRC, 2009) to $281 billion (Epstein et al., 2009). The total hidden
448
cost (environmental and health) damage from the most polluting coal-fired
449
electrical generation plants is estimated to be seven times as much as the damage
450
from the most polluting natural gas-fired electrical generation plants (US NRC,
451
2009). US national net impacts from substitution of natural gas for coal in
452
electrical production are likely to be positive due to reduction of coal-related
453
externalities along with GHG emissions, however the full economics of the global
454
climate change problem (Goodstein, 2011) are beyond the scope of this work. 455
Nevertheless, a more-straightforward assessment of economic costs of
456
shale gas exploitation is possible based simply on potential health impacts, and
457
effects on local populations in New York. Air quality has deteriorated in the
458
United States where natural gas resources are currently being exploited (Kargbo et
459
al., 2010, Petron et al. 2012). Public health impacts in Colorado and Pennsylvania
460
have been estimated (Colorado School of Public Health, 2011; Lauver, 2011;
461
McKenzie et al. 2012). The leading air-quality risk to public health in Colorado is
462
increased subchronic exposures to airborne hydrocarbon carcinogens and
463
increased cumulative cancer risks for those residing within 0.8 km (0.5 mile) of
464
gas-producing wells compared to those living farther away (McKenzie et al.,
465
2012). Other impacts considered in the Colorado public health study (Colorado
466
School of Public Health, 2011) and elsewhere (Lauver, 2012) involve particulates,
467
degradation of water quality, light pollution, and industrial noise from drilling and
468
compressor stations. These health impacts are clearly potentially severe for
469 19 residents of New York state where Marcellus drilling may be permitted, and
470
would need to be substantially mitigated. 471
472
4.3. Economic impacts and tradeoffs in New York State
473
The tradeoffs between who benefits and who is adversely affected by
474
natural gas drilling, what size these populations are and where they are located,
475
are relevant here. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 Conventional environmental economics methods (willingness
476
to pay, choice experiments, contingent valuation or analytical hierarchy processes)
477
for assessing risks and costs of natural resources degradation (Martin-Ortega and
478
Berbel, 2010) suffer from incomplete information (Konishi and Coggins, 2008)
479
and depend on polling the inhabitants of the landscape affected. Further issues
480
for such local market-based, cost-benefit analysis are that widely accepted
481
watershed protection methods such as the US EPA total maximum daily load
482
(TMDL) approach simply do not produce positive benefit-cost ratios (e.g. 483
Borisova et al., 2008), and most studies do not consider spatial heterogeneity
484
(populations not inhabiting the areas in question) (Brouwer et al., 2010). Such
485
environmental economics methods are impractical in the case of Marcellus shale
486
drilling and the New York City water supply because the benefits and costs accrue
487
at least in part to different populations. 488
Even so, natural gas drilling has been promoted, in industry-sponsored
489
economic impact reports, as either an economic boon to the state and struggling
490
communities in rural areas (see Kinnaman, 2012 and references therein), or
491 4.3. Economic impacts and tradeoffs in New York State
473
The tradeoffs between who benefits and who is adversely affected by
474
natural gas drilling, what size these populations are and where they are located,
475
are relevant here. Conventional environmental economics methods (willingness
476
to pay, choice experiments, contingent valuation or analytical hierarchy processes)
477
for assessing risks and costs of natural resources degradation (Martin-Ortega and
478
Berbel, 2010) suffer from incomplete information (Konishi and Coggins, 2008)
479
and depend on polling the inhabitants of the landscape affected. Further issues
480
for such local market-based, cost-benefit analysis are that widely accepted
481
watershed protection methods such as the US EPA total maximum daily load
482
(TMDL) approach simply do not produce positive benefit-cost ratios (e.g. 483
Borisova et al., 2008), and most studies do not consider spatial heterogeneity
484
(populations not inhabiting the areas in question) (Brouwer et al., 2010). Such
485
environmental economics methods are impractical in the case of Marcellus shale
486
drilling and the New York City water supply because the benefits and costs accrue
487
at least in part to different populations. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 488 Even so, natural gas drilling has been promoted, in industry-sponsored
489
economic impact reports, as either an economic boon to the state and struggling
490
communities in rural areas (see Kinnaman, 2012 and references therein), or
491
alternatively a threat to the current tourism-based economy in rural counties
492
(Rumbach, 2011). Detailed analysis of these dueling economic impact
493 20 perspectives is beyond the scope of this work, but many of their assumptions and
494
economic modeling procedures have been questioned (Kinnaman, 2012). What is
495
clear is that lower natural gas prices due to increased production from the
496
Marcellus will benefit electricity consumers and many others by reducing the
497
percentage of power produced from coal-fired plants and associated externalities. 498
The difference between the shorter term (10-20 years) of the "boom" type
499
economic development benefits associated with natural gas production and the
500
eventual long-term costs of the permanent transformation of the landscape would
501
depend on the economic discount rate used (Goodstein 2011). 502
However, considering the relative populations concerned who stand to
503
benefit or suffer adverse consequences provides a baseline for comparison. The
504
southern tier New York state counties along the Pennsylvania border (Figure 1)
505
have a population of less than 700,000, many of whom would benefit from
506
royalties due to leasing of mineral rights to natural gas producers (Kargbo et al.,
507
2010). This population could suffer health-related impacts from proximity to the
508
gas-producing wells and economic impacts from landscape transformation. 509
Compared to this are the costs and public health consequences of potential
510
degradation of a public water supply for a much larger population (almost 9
511
million) in New York City. Currently, New York City taxpayers invest a modest
512
US$3 million/yr to support conservation practices in the Cannonsville watershed
513
(Bryant et al., 2008), and about US$50 million/yr for the combined watershed
514
protection program. However, if the US EPA rescinds its filtration avoidance
515
determination (FAD), the alternative costs of water filtration and treatment for
516
New York City have been estimated in the billions of US$ (Bryant et al., 2008),
517
largely for the construction and operation of the facilities needed. 518 perspectives is beyond the scope of this work, but many of their assumptions and
494
economic modeling procedures have been questioned (Kinnaman, 2012). Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 Despite solidifying scientific consensus (IPCC, 2007)
525
on the need for GHG emissions reduction, and widespread international
526
ratification of the 1997 Kyoto Protocol, the unwillingness of the United States to
527
ratify and the collapse of the former Soviet Union and Eastern European industrial
528
production have probably had more impact on GHG emission trends over the last
529
20 years. Recent attempts in the field of uncertainty analysis to address this
530
dilemma have called for scientific knowledge that is used in political
531
decisionmaking to be placed in the proper socio-political context to be relevant
532
(Maxim and van der Sluijs, 2011), and such is the intent of this work. 533
5.1 Federal regulatory gaps
534
There are numerous exemptions and limitations in federal environmental
535
legislation and US EPA authority to regulate unconventional natural gas drilling
536
(Wiseman, 2012; US GAO, 2012). Of the eight major pieces of legislation (Safe
537
Drinking Water Act - SDWA; Clean Water Act - CWA; Clean Air Act - CAA;
538
Resource Conservation and Recovery Act - RCRA; Comprehensive
539
Environmental Response, Compensation and Liability Act - CERCLA;
540
Emergency Planning and Community Right-to-Know Act - EPCRA; Toxic
541
Substances Control Act - TSCA; and Federal Insecticide, Fungicide and
542 5. Regulatory and political aspects
520
Finally, the ultimate consideration when assessing a scientific and
521
technical issue with major public policy implications is the political and
522
regulatory landscape. Experts tend to view the scientific and technical aspects in
523
isolation, whereas the success of public policy decisions about these issues can
524
depend more on politics. Despite solidifying scientific consensus (IPCC, 2007)
525
on the need for GHG emissions reduction, and widespread international
526
ratification of the 1997 Kyoto Protocol, the unwillingness of the United States to
527
ratify and the collapse of the former Soviet Union and Eastern European industrial
528
production have probably had more impact on GHG emission trends over the last
529
20 years. Recent attempts in the field of uncertainty analysis to address this
530
dilemma have called for scientific knowledge that is used in political
531
decisionmaking to be placed in the proper socio-political context to be relevant
532
(Maxim and van der Sluijs, 2011), and such is the intent of this work. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 What is
495
clear is that lower natural gas prices due to increased production from the
496
Marcellus will benefit electricity consumers and many others by reducing the
497
percentage of power produced from coal-fired plants and associated externalities. 498
The difference between the shorter term (10-20 years) of the "boom" type
499
economic development benefits associated with natural gas production and the
500
eventual long-term costs of the permanent transformation of the landscape would
501
depend on the economic discount rate used (Goodstein 2011). 502 However, considering the relative populations concerned who stand to
503
benefit or suffer adverse consequences provides a baseline for comparison. The
504
southern tier New York state counties along the Pennsylvania border (Figure 1)
505
have a population of less than 700,000, many of whom would benefit from
506
royalties due to leasing of mineral rights to natural gas producers (Kargbo et al.,
507
2010). This population could suffer health-related impacts from proximity to the
508
gas-producing wells and economic impacts from landscape transformation. 509
Compared to this are the costs and public health consequences of potential
510
degradation of a public water supply for a much larger population (almost 9
511
million) in New York City. Currently, New York City taxpayers invest a modest
512
US$3 million/yr to support conservation practices in the Cannonsville watershed
513
(Bryant et al., 2008), and about US$50 million/yr for the combined watershed
514
protection program. However, if the US EPA rescinds its filtration avoidance
515
determination (FAD), the alternative costs of water filtration and treatment for
516
New York City have been estimated in the billions of US$ (Bryant et al., 2008),
517
largely for the construction and operation of the facilities needed. 518 21 21 519 5. Regulatory and political aspects
520
Finally, the ultimate consideration when assessing a scientific and
521
technical issue with major public policy implications is the political and
522
regulatory landscape. Experts tend to view the scientific and technical aspects in
523
isolation, whereas the success of public policy decisions about these issues can
524
depend more on politics. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 533 22
5.1 Federal regulatory gaps
534
There are numerous exemptions and limitations in federal environmental
535
legislation and US EPA authority to regulate unconventional natural gas drilling
536
(Wiseman, 2012; US GAO, 2012). Of the eight major pieces of legislation (Safe
537
Drinking Water Act - SDWA; Clean Water Act - CWA; Clean Air Act - CAA;
538
Resource Conservation and Recovery Act - RCRA; Comprehensive
539
Environmental Response, Compensation and Liability Act - CERCLA;
540
Emergency Planning and Community Right-to-Know Act - EPCRA; Toxic
541
Substances Control Act - TSCA; and Federal Insecticide, Fungicide and
542 22 Rodenticide Act - FIFRA), the first six have important exemptions related to oil
543
and natural gas development. 544 Rodenticide Act - FIFRA), the first six have important exemptions related to oil
543
and natural gas development. 544 The most important exemptions to these laws were created by the 2005
545
Energy Policy Act, by which hydraulic fracturing is specifically exempted from
546
regulation under the SDWA Underground Injection Control program, except if
547
diesel fuel is injected (US GAO, 2012; Wiseman, 2012). Under the CWA,
548
pollutant discharges from industrial sites and wastewater treatment facilities are
549
regulated, however oil and gas production well sites are exempted from National
550
Pollutant Discharge Elimination System (NPDES) permitting. The CAA exempts
551
certain naturally-occurring hydrocarbon mixtures from air quality regulation and
552
prohibits aggregating emissions from multiple well sites, pipelines or pumping
553
stations, hence no oil and gas wells have been regulated as air pollutant sources to
554
date (US GAO, 2012). 555 The US EPA issued a controversial determination in 1988 that oil and gas
556
development waste are not covered under RCRA, governing hazardous solid
557
waste, but the agency retains "imminent and substantial endangerment"
558
authorization to intervene. It is clear that these major gaps and exemptions
559
hinder federal oversight of environmental protection in the case of hydraulic
560
fracturing, and legislation (the FRAC Act) to close the CWA exemptions has been
561
introduced in Congress (Rahm, 2011), but not yet passed. Due to the limitations
562
in federal legislation, the primary responsibility for enforcement of environmental
563
regulation of oil and gas production has rested at the state level, in particular
564
where states have been delegated responsibility ("primacy") to enforce federal
565
law. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 566 23 24
5.2 State regulatory authority and experience
567
The considerable differences in state authority, regulatory structure and
568
history of oil and gas exploration make comparisons among states' levels of
569
regulatory effectiveness very difficult. Several recent studies have analyzed
570
various aspects of regulatory experience, focusing on Pennsylvania (Mitchell and
571
Casman, 2011), Texas (Rahm, 2011) and a comparison of these and several other
572
states (Wiseman, 2012). Over the period 2008-2011, Pennsylvania and Texas,
573
both with long histories of oil and gas exploration, provide a comparison between
574
a state with fairly aggressive enforcement (PA) leading to the largest number of
575
violations of state environmental or oil and gas laws (Wiseman, 2012); and a state
576
(TX) with a very "oil and gas-friendly" regulatory environment and looser
577
enforcement (Rahm, 2011). 578
One concern that emerges is that oil and gas production operational
579
methods developed under less restrictive state regulatory structures (Texas) are
580
inconsistent with prevailing regulation (Pennsylvania) that is being established by
581
states with higher environmental protection standards. Another is that
582
technological innovations and economic incentives that do not currently include
583
the costs of environmental protection are now driving the boom in unconventional
584
natural gas production. Gas well productivity is declining, and unconventional
585
natural gas is now being produced at a loss, given that the marginal cost of
586
production is much higher than world gas prices (Rogers, 2011; Hughes, 2011). 587
Current regulatory authority and environmental protection have not been able to
588
keep up with the economic drivers of unconventional gas production where it is
589
now occurring in the United States. These regulatory shortfalls are manifest
590
especially in the areas of well plugging or sealing and in the ability of states to
591 5.2 State regulatory authority and experience
567
The considerable differences in state authority, regulatory structure and
568
history of oil and gas exploration make comparisons among states' levels of
569
regulatory effectiveness very difficult. Several recent studies have analyzed
570
various aspects of regulatory experience, focusing on Pennsylvania (Mitchell and
571
Casman, 2011), Texas (Rahm, 2011) and a comparison of these and several other
572
states (Wiseman, 2012). Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 Over the period 2008-2011, Pennsylvania and Texas,
573
both with long histories of oil and gas exploration, provide a comparison between
574
a state with fairly aggressive enforcement (PA) leading to the largest number of
575
violations of state environmental or oil and gas laws (Wiseman, 2012); and a state
576
(TX) with a very "oil and gas-friendly" regulatory environment and looser
577
enforcement (Rahm, 2011). 578 24
One concern that emerges is that oil and gas production operational
579
methods developed under less restrictive state regulatory structures (Texas) are
580
inconsistent with prevailing regulation (Pennsylvania) that is being established by
581
states with higher environmental protection standards. Another is that
582
technological innovations and economic incentives that do not currently include
583
the costs of environmental protection are now driving the boom in unconventional
584
natural gas production. Gas well productivity is declining, and unconventional
585
natural gas is now being produced at a loss, given that the marginal cost of
586
production is much higher than world gas prices (Rogers, 2011; Hughes, 2011). 587
Current regulatory authority and environmental protection have not been able to
588
keep up with the economic drivers of unconventional gas production where it is
589
now occurring in the United States. These regulatory shortfalls are manifest
590
especially in the areas of well plugging or sealing and in the ability of states to
591 24 field sufficent inspectors for oversight. 592 5.3 State regulatory shortfalls
593
At the end of their economically useful life, wells in oil and natural gas
594
fields must be properly sealed with cement to close direct pathways between the
595
reservoir and the surface, or even shallow groundwater, to avoid environmental
596
damage as detailed previously. Many of the hundreds of thousands of wells
597
estimated to have been drilled over the last century in Pennsylvania and New
598
York have not been adequately plugged (Crain 1969), in part because modern
599
record-keeping and verification of well sealing only began in the 1980s. Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 632
6. Discussion
633
To evaluate these major scientific and technical issues, the related
634
geological, land use, economic and energy aspects, as well as the regulatory and
635
liti
l
t
t f th M
ll
h l
l it ti
d th N
Y
k Cit
636 minimum federal and state bonding levels are generally inadequate and
616 26
The other major concern is the need for adequate field-based monitoring
618
and inspections by regulatory personnel to ensure compliance with environmental
619
standards. Beyond the gaps in federal legislation and the historical development
620
of the oil and gas industry, the states' regulatory framework has often been
621
reactive and inadequate to prevent violations, notably in Pennsylvania (Rahm,
622
2011). Some of the reasons for this ineffectiveness include lack of state funding
623
(Mitchell and Casman, 2012), inefficient organization or incompleteness of
624
records (Wiseman, 2012), fragmentation of environmental regulation authority
625
and an anti-regulatory political climate (Rahm, 2011). For example, prior to the
626
growth in unconventional gas drilling in Pennsylvania, it was estimated that due
627
to inadequate funding rates, it would require 160 years to plug known existing
628
"orphan" wells (Mitchell and Casman, 2012). Furthermore, in Texas, for natural
629
gas extraction from the Barnett shale, violations recorded declined in 2009-2011
630
from the rates in 2008, apparently because the Texas Railroad Commission
631
regulator suffered personnel losses due to a hiring freeze (Wiseman, 2012). 632
6. Discussion
633
To evaluate these major scientific and technical issues, the related
634
geological, land use, economic and energy aspects, as well as the regulatory and
635
political context of the Marcellus shale gas exploitation and the New York City
636
water supply, a framework for decisionmaking is needed. Consider the
637
semiquantitative interplay of three general domains originally identified by
638
Rogers (2011): geology; technology; and regulatory and public acceptance. A
639 Science-based decision-making on complex issues: Marcellus shale
1
gas hydrofracking and New York City water supply
2 While
600
more modern oil and gas fields in Alberta (Watson and Bachu 2009) have such
601
records numbering in the hundreds of thousands, older oil and gas provinces in
602
Pennsylvania and New York where wells were drilled to 2000 ft depth or more
603
(Hartnagel and Russell 1925; Van Tyne 1998) only have records of tens of
604
thousands of wells (NYSDEC regulates approximately 40,000 known wells, most
605
of which are not sealed, but operational or on "inactive status"). The mismatch in
606
numbers corresponds to numerous leaking "legacy wells" whose location is often
607
unknown (Watson and Bachu, 2009; Mitchell and Casman, 2011). 608 While enforcement has undoubtedly improved, the economic incentives
609
that led to this poor regulatory compliance still exist. As natural gas wells decline
610
in production after the first couple years, they are generally transferred from the
611
original gas producers to smaller entities, either other gas producers or even
612
landowners. Furthermore, many wells are put on "inactive status" or otherwise
613
escape regulatory oversight when operators default. Costs of proper well sealing,
614
which ranges from $100,000 to $700,000 per well, are thereby avoided because
615 25 minimum federal and state bonding levels are generally inadequate and
616
reclamation costs are often deferred for decades (Mitchell and Casman, 2011). 617
The other major concern is the need for adequate field-based monitoring
618
and inspections by regulatory personnel to ensure compliance with environmenta
619
standards. Beyond the gaps in federal legislation and the historical development
620
of the oil and gas industry, the states' regulatory framework has often been
621
reactive and inadequate to prevent violations, notably in Pennsylvania (Rahm,
622
2011). Some of the reasons for this ineffectiveness include lack of state funding
623
(Mitchell and Casman, 2012), inefficient organization or incompleteness of
624
records (Wiseman, 2012), fragmentation of environmental regulation authority
625
and an anti-regulatory political climate (Rahm, 2011). For example, prior to the
626
growth in unconventional gas drilling in Pennsylvania, it was estimated that due
627
to inadequate funding rates, it would require 160 years to plug known existing
628
"orphan" wells (Mitchell and Casman, 2012). Furthermore, in Texas, for natural
629
gas extraction from the Barnett shale, violations recorded declined in 2009-2011
630
from the rates in 2008, apparently because the Texas Railroad Commission
631
regulator suffered personnel losses due to a hiring freeze (Wiseman, 2012). 6.
Discussion
633 To evaluate these major scientific and technical issues, the related
634
geological, land use, economic and energy aspects, as well as the regulatory and
635
political context of the Marcellus shale gas exploitation and the New York City
636
water supply, a framework for decisionmaking is needed. Consider the
637
semiquantitative interplay of three general domains originally identified by
638
Rogers (2011): geology; technology; and regulatory and public acceptance. A
639 26 Venn-type graphical approach used in pharmacology and bioinformatics (Ruskey
640
et al., 2006; Chen and Boutros, 2011) allows plotting the overlap of these three
641
areas to analyze common ground for decisionmaking. It is useful to consider
642
Venn diagram circles overlaid on a triaxial plot (Figure 3) to constrain relative
643
size, corresponding to domain possibility, and proximity to origin, corresponding
644
to how well the possibility is put into practice for that domain. An arbitrary scale
645
on the axes represents increasing implementation toward the origin in an abstract
646
"decision-space". Since perfect implementation (concentric circles at origin) is
647
unattainable, the centers of the circles, for which the relative size needs to be
648
determined, plot at different locations on the three axes. 649
650
6.1 Triaxial Venn diagram logic and analysis
651
A convenient starting point is the known geographical setting of the
652
Marcellus shale, the northeastern end of which underlies the southern tier of New
653
York state (Figure 1). From a geological perspective, all of the Marcellus shale is
654
potentially a source of natural gas, but only a subset (unknown until sufficient
655
wells have been drilled) of that area will be the most highly productive. 656
Extraction of natural gas from such shales was not even technologically viable
657
until recent decades with the advent of hydraulic fracturing, and depth, thickness
658
and organic content are still limiting factors. Therefore, the size of the circle
659
representing technology is necessarily smaller than the circle representing
660
geology, analogous to the difference between reserve and resource of any fossil
661
fuel. 662
The development of hydraulic fracturing has enlarged the circle of
663 Venn-type graphical approach used in pharmacology and bioinformatics (Ruskey
640
et al., 2006; Chen and Boutros, 2011) allows plotting the overlap of these three
641
areas to analyze common ground for decisionmaking. 6.
Discussion
633 It is useful to consider
642
Venn diagram circles overlaid on a triaxial plot (Figure 3) to constrain relative
643
size, corresponding to domain possibility, and proximity to origin, corresponding
644
to how well the possibility is put into practice for that domain. An arbitrary scale
645
on the axes represents increasing implementation toward the origin in an abstract
646
"decision-space". Since perfect implementation (concentric circles at origin) is
647
unattainable, the centers of the circles, for which the relative size needs to be
648
determined, plot at different locations on the three axes. 649 6.1 Triaxial Venn diagram logic and analysis
651
A convenient starting point is the known geographical setting of the
652
Marcellus shale, the northeastern end of which underlies the southern tier of New
653
York state (Figure 1). From a geological perspective, all of the Marcellus shale is
654
potentially a source of natural gas, but only a subset (unknown until sufficient
655
wells have been drilled) of that area will be the most highly productive. 656
Extraction of natural gas from such shales was not even technologically viable
657
until recent decades with the advent of hydraulic fracturing, and depth, thickness
658
and organic content are still limiting factors. Therefore, the size of the circle
659
representing technology is necessarily smaller than the circle representing
660
geology, analogous to the difference between reserve and resource of any fossil
661
fuel. 662
The development of hydraulic fracturing has enlarged the circle of
663 27 technological viability and moved it closer to the center, enabling considerable
664
overlap with the area of geological resource. Similarly, United States' energy
665
needs and growth of natural gas production have enlarged the circle of regulatory
666
and public acceptance and moved it closer to the center, enabling overlap with the
667
other two circles. It is clear, however, that the circle of regulatory and public
668
acceptance is the smallest of the three, and the challenge is to identify what is the
669
overlap of the three circles, what lies inside and outside, and what might be
670
necessary to maximize the size of the intersecting common ground for publicly
671
acceptable Marcellus shale gas drilling. 6.
Discussion
633 672
673
6.2 Application to decisionmaking about hydraulic fracturing in
674
New York state
675
The dramatic expansion of drilling for natural gas in the Marcellus shale
676
indicates that there is considerable growing overlap of technology and geology
677
due to rapidly advancing technological innovation in the drilling industry,
678
however this may be counterbalanced by economic considerations such as low
679
natural gas prices. The trend of the much smaller overlap for regulatory and
680
public acceptance in New York state is not as evident due to the continuing
681
controversy in the scientific literature and public media over the environmental
682
impact. In any event, a careful weighing of the different merits and liabilities
683
described earlier is necessary for expansion of the area of regulatory and public
684
acceptance. The major pros and cons are summarized in Table 1, taking into
685
account a broader range of issues than are commonly discussed. 686
A key consideration must be that the geographical area occupied by the
687 technological viability and moved it closer to the center, enabling considerable
664
overlap with the area of geological resource. Similarly, United States' energy
665
needs and growth of natural gas production have enlarged the circle of regulatory
666
and public acceptance and moved it closer to the center, enabling overlap with the
667
other two circles. It is clear, however, that the circle of regulatory and public
668
acceptance is the smallest of the three, and the challenge is to identify what is the
669
overlap of the three circles, what lies inside and outside, and what might be
670
necessary to maximize the size of the intersecting common ground for publicly
671
acceptable Marcellus shale gas drilling. 672
673 6.2 Application to decisionmaking about hydraulic fracturing in
674
New York state
675
The dramatic expansion of drilling for natural gas in the Marcellus shale
676
indicates that there is considerable growing overlap of technology and geology
677
due to rapidly advancing technological innovation in the drilling industry,
678
however this may be counterbalanced by economic considerations such as low
679
natural gas prices. The trend of the much smaller overlap for regulatory and
680
public acceptance in New York state is not as evident due to the continuing
681
controversy in the scientific literature and public media over the environmental
682
impact. 6.
Discussion
633 In any event, a careful weighing of the different merits and liabilities
683
described earlier is necessary for expansion of the area of regulatory and public
684
acceptance. The major pros and cons are summarized in Table 1, taking into
685
account a broader range of issues than are commonly discussed. 686 28
A key consideration must be that the geographical area occupied by the
687
watersheds supplying municipal water to New York City is less than 10% of the
688 28 area of the Marcellus underlying New York state. A reasonable first step in
689
decisionmaking would therefore be to recognize that for the NYC watersheds
690
overlying the Marcellus, the risks clearly outweigh any merits, and that the de-
691
facto moratorium on drilling within a generous buffer setback of those watersheds
692
and associated infrastructure (water supply tunnels), be formalized into a ban. 693
This would include directional drilling from areas outside those buffer zones, and
694
accept that the NYC watersheds lie permanently outside the circle of regulatory
695
and public acceptance for Marcellus drilling (Figure 3). 696 29
Beyond the boundaries of the buffer setbacks around the NYC water
697
supply watersheds and infrastructure, the situation is less clear-cut. The benefits
698
of increased natural gas use from the Marcellus shale (Table 1) are potentially
699
great for transition to cleaner energy if the liabilities could be substantially
700
reduced. Much of the environmental impact from Marcellus shale drilling is due
701
to regulatory lapses, either at the federal or state level. This suggests that
702
considerable strengthening of state and federal regulatory oversight is a possible
703
avenue for reducing the environmental liability. In other words, enhancing
704
oversight could enlarge the circle of regulatory and public acceptance (Figure 3)
705
in New York state and move it down the axis of increasing implementation to
706
expand the area of overlap or common ground for publicly acceptable HVHF. 707
708
6.3 Regulatory enhancement recommendations
709
Federal oversight is increasing with the US EPA promulgation in 2012 of
710
the final rules on GHG reporting by operators of petroleum and natural gas
711
systems, with a minimum facility threshold of 25,000 metric tons carbon dioxide
712 29 equivalent (CO2e) per year
713
(http://www.epa.gov/ghgreporting/reporters/subpart/w.html). 6.
Discussion
633 743
While it is technically feasible to mitigate much of this environmental
744
impact of expanded Marcellus gas drilling in New York State, market forces will
745
cause natural gas companies to avoid the responsibility and costs of such
746
mitigation (Mitchell and Casman, 2011). Many technologies exist for treatment
747
of produced water (Fakhru’l-Razi et al., 2009; Gregory et al. 2011), and New
748
York state regulations should mandate such treatment and a 90% minimum for
749
recycling instead of simply encouraging such practices. Recently reported
750
alternatives to massive water usage and wastewater generation include
751
hydrofracturing using liquid petroleum gas and liquid CO2 (Kargbo et al., 2010),
752
which could be strongly encouraged. While limits on gas venting and flaring,
753
reduced emissions completions (REC) to minimize gas well GHG emissions, and
754
restrictions on diesel engines are proposed as part of the revised NY state
755
dSGEIS, more is needed to ensure necessary environmental protection. For
756
example, requiring gas distribution lines to be in place upon well completion
757
would allow a ban on all venting and flaring except in case of emergency, and
758
onsite engines could be required to operate with a 20% minimum percentage of
759
biodiesel or compressed natural gas (CNG), limits that would increase in 5 year
760
increments. 761 reported number of inspectors less than 20 in 2012, will be able to enforce
738
effective regulations, no matter how rigorous they are. The major step to build
739
public confidence that regulation will be able to reduce the environmental
740
liabilities (Table 1) is to hire and train skilled regulatory inspection staff in the
741
thousands, as is currently the case in major gas-producing states like Michigan
742
and Texas (Wiseman, 2011). 743 While it is technically feasible to mitigate much of this environmental
744
impact of expanded Marcellus gas drilling in New York State, market forces will
745
cause natural gas companies to avoid the responsibility and costs of such
746
mitigation (Mitchell and Casman, 2011). Many technologies exist for treatment
747
of produced water (Fakhru’l-Razi et al., 2009; Gregory et al. 2011), and New
748
York state regulations should mandate such treatment and a 90% minimum for
749
recycling instead of simply encouraging such practices. A serious concern remains about proper well construction and sealing or
762 6.
Discussion
633 Furthermore, the
714
US EPA is tightening New Source Review (Bushnell and Wolfram 2012)
715
requirements for airborne emission standards for the oil and gas sector
716
(http://www.epa.gov/airquality/oilandgas/pdfs/20120418rtc.pdf), regulations due
717
to be in place by 2015, barring potential litigation. An obvious next step would
718
be for the U.S. Congress to close the gaps in federal law (CWA, SDWA, CAA,
719
RCRA, CERCLA, EPCRA) that have long exempted the oil and gas industry from
720
national environmental protection legislation (US GAO, 2012; Wiseman, 2012). 721
However, such an outcome is uncertain due to the current ideological deadlock in
722
the U.S. Congress. 723 equivalent (CO2e) per year
713 State regulatory oversight in New York, as proposed in the draft
724
supplemental generic environmental impact study (dSGEIS) currently undergoing
725
review by the NYS Dept of Environmental Conservation, is more rigorous than in
726
many other states, particularly with respect to mandated buffer zones for natural
727
gas infrastructure from natural resources (Wiseman, 2012) and requirements for
728
closed-tank systems for wastewater capture in most cases. However, budget
729
reductions at both the U.S. state and federal levels have thinned out regulatory
730
personnel ranks below the minimum needed for current enforcement
731
responsibilities. In 2010, the head of New York State DEC was dismissed
732
following the release of an internal memo that documented a 21% reduction in
733
workforce since 2008, and warned that fewer staff will be available for oversight
734
of Marcellus natural gas drilling. Over the last 3 years, Pennsylvania, with several
735
hundred inspectors, has been unable to effectively prevent serious violations
736
(Wiseman, 2011; Rahm, 2011), so it is unlikely that New York, with a widely
737 30 reported number of inspectors less than 20 in 2012, will be able to enforce
738
effective regulations, no matter how rigorous they are. The major step to build
739
public confidence that regulation will be able to reduce the environmental
740
liabilities (Table 1) is to hire and train skilled regulatory inspection staff in the
741
thousands, as is currently the case in major gas-producing states like Michigan
742
and Texas (Wiseman, 2011). 6.
Discussion
633 Recently reported
750
alternatives to massive water usage and wastewater generation include
751
hydrofracturing using liquid petroleum gas and liquid CO2 (Kargbo et al., 2010),
752
which could be strongly encouraged. While limits on gas venting and flaring,
753
reduced emissions completions (REC) to minimize gas well GHG emissions, and
754
restrictions on diesel engines are proposed as part of the revised NY state
755
dSGEIS, more is needed to ensure necessary environmental protection. For
756
example, requiring gas distribution lines to be in place upon well completion
757
would allow a ban on all venting and flaring except in case of emergency, and
758
onsite engines could be required to operate with a 20% minimum percentage of
759
biodiesel or compressed natural gas (CNG), limits that would increase in 5 year
760
increments. 761 31 32
plugging. Outside the NYC watersheds to the west (Figure 1), the counties along
763
the Pennsylvania state line (the "Southern Tier") account for over 77% of oil and
764
natural gas wells registered in New York State, and are likely to be centers of
765
Marcellus shale gas production. According to a recent survey of public records,
766
most (89%) depleted oil and gas wells in New York state have not been
767
adequately plugged or sealed over the last 25 years, leaving tens of thousands as
768
potential conduits for contamination, most of which have unknown locations. 769
Current proposed regulations (revised dSGEIS) require Marcellus well drilling
770
permit applicants to identify non-producing or abandoned (sealed) wells within
771
one mile of the proposed drilling location for HVHF, however it is not clear if this
772
entails locating all previously unknown or improperly sealed wells that may pose
773
a catastrophic environmental protection risk during new well drilling. Mitigating
774
the rare but very real hazards posed by such historical wells would require
775
substantially strengthened regulatory oversight. 776
Fortunately, New York State has a strong tradition of environmental
777
protection of part of the NYC watersheds in question, which is off-limits to
778
development in perpetuity in the Catskill State Forest Preserve (Figure 1). The
779
Catskill State Forest includes private land within a boundary known as the “blue
780
line”, which encompasses an area occupying effectively the southeastern half of
781
the New York City water supply watersheds. 6.
Discussion
633 Avoiding
798
this outcome can be accomplished by maintaining the present moratorium while
799
addressing the broader issues described in this work. 800
It may never be possible to quantify all of the costs and benefits associated
801
with the prospect of natural gas drilling in and near the watersheds that supply
802
New York City with drinking water. However, while potential benefits may be
803
great in the context of future energy policy towards a low-carbon economy,
804
liabilities are also very significant and could outweigh benefits (Table 1). 805
Whether benefits of Marcellus shale gas drilling exceed liabilities thus depends on
806
enforcement of strong environmental regulations to minimize liabilities and
807
achieve greater public acceptance of HVHF in New York State (Figure 3). In the
808
current political and regulatory climate, it remains unclear whether this can be
809
accomplished. Specific recommendations from this analysis for New York State
810
to obtain greater public acceptance and environmental protection include:
811 necessary setbacks to that protected area and associated water supply
788
infrastructure (currently proposed to be only ~1200 m). 789
790
7. Conclusions
791
Even if consensus can be achieved on the scientific and technical issues
792
outlined earlier, gas production from the Marcellus shale is so economically
793
important that the New York State governor and legislature will ultimately decide
794
whether and how to allow such natural resource development to proceed. The
795
nature of political decision-making is primarily non-scientific, and as in recent
796
integrated water management (Kragt et al., 2009), other interdisciplinary factors
797
enter into consideration and may even trump science-based analysis. Avoiding
798
this outcome can be accomplished by maintaining the present moratorium while
799
addressing the broader issues described in this work. 800 It may never be possible to quantify all of the costs and benefits associated
801
with the prospect of natural gas drilling in and near the watersheds that supply
802
New York City with drinking water. However, while potential benefits may be
803
great in the context of future energy policy towards a low-carbon economy,
804
liabilities are also very significant and could outweigh benefits (Table 1). 805
Whether benefits of Marcellus shale gas drilling exceed liabilities thus depends on
806
enforcement of strong environmental regulations to minimize liabilities and
807
achieve greater public acceptance of HVHF in New York State (Figure 3). 6.
Discussion
633 Current regulatory proposals
782
anticipate including all watersheds or major aquifer areas that supply major
783
municipalities like NYC, which are already partially protected by other City-
784
owned property or conservation easements, in the area prohibited to HVHF
785
drilling (Figure 1). It would also be important to prohibit deep directional
786
drilling underneath those areas (from surface locations outside) and to increase
787 plugging. Outside the NYC watersheds to the west (Figure 1), the counties along
763
the Pennsylvania state line (the "Southern Tier") account for over 77% of oil and
764
natural gas wells registered in New York State, and are likely to be centers of
765
Marcellus shale gas production. According to a recent survey of public records,
766
most (89%) depleted oil and gas wells in New York state have not been
767
adequately plugged or sealed over the last 25 years, leaving tens of thousands as
768
potential conduits for contamination, most of which have unknown locations. 769 Current proposed regulations (revised dSGEIS) require Marcellus well drilling
770
permit applicants to identify non-producing or abandoned (sealed) wells within
771
one mile of the proposed drilling location for HVHF, however it is not clear if this
772
entails locating all previously unknown or improperly sealed wells that may pose
773
a catastrophic environmental protection risk during new well drilling. Mitigating
774
the rare but very real hazards posed by such historical wells would require
775
substantially strengthened regulatory oversight. 776 32
Fortunately, New York State has a strong tradition of environmental
777
protection of part of the NYC watersheds in question, which is off-limits to
778
development in perpetuity in the Catskill State Forest Preserve (Figure 1). The
779
Catskill State Forest includes private land within a boundary known as the “blue
780
line”, which encompasses an area occupying effectively the southeastern half of
781
the New York City water supply watersheds. Current regulatory proposals
782
anticipate including all watersheds or major aquifer areas that supply major
783
municipalities like NYC, which are already partially protected by other City-
784
owned property or conservation easements, in the area prohibited to HVHF
785
drilling (Figure 1). 6.
Discussion
633 It would also be important to prohibit deep directional
786
drilling underneath those areas (from surface locations outside) and to increase
787 32 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
necessary setbacks to that protected area and associated water supply
788
infrastructure (currently proposed to be only ~1200 m). 789
790
7. Conclusions
791
Even if consensus can be achieved on the scientific and technical issues
792
outlined earlier, gas production from the Marcellus shale is so economically
793
important that the New York State governor and legislature will ultimately decide
794
whether and how to allow such natural resource development to proceed. The
795
nature of political decision-making is primarily non-scientific, and as in recent
796
integrated water management (Kragt et al., 2009), other interdisciplinary factors
797
enter into consideration and may even trump science-based analysis. Avoiding
798
this outcome can be accomplished by maintaining the present moratorium while
799
addressing the broader issues described in this work. 800
It may never be possible to quantify all of the costs and benefits associated
801
with the prospect of natural gas drilling in and near the watersheds that supply
802
New York City with drinking water. However, while potential benefits may be
803
great in the context of future energy policy towards a low-carbon economy,
804
liabilities are also very significant and could outweigh benefits (Table 1). 805
Whether benefits of Marcellus shale gas drilling exceed liabilities thus depends on
806
enforcement of strong environmental regulations to minimize liabilities and
807 33
necessary setbacks to that protected area and associated water supply
788
infrastructure (currently proposed to be only ~1200 m). 789
790
7. Conclusions
791
Even if consensus can be achieved on the scientific and technical issues
792
outlined earlier, gas production from the Marcellus shale is so economically
793
important that the New York State governor and legislature will ultimately decide
794
whether and how to allow such natural resource development to proceed. The
795
nature of political decision-making is primarily non-scientific, and as in recent
796
integrated water management (Kragt et al., 2009), other interdisciplinary factors
797
enter into consideration and may even trump science-based analysis. 6.
Discussion
633 In the
808
current political and regulatory climate, it remains unclear whether this can be
809
accomplished. Specific recommendations from this analysis for New York State
810
to obtain greater public acceptance and environmental protection include:
811 33 x Immediate hiring and training of sufficient NYS DEC
812
inspectors (1000+) and increasing agency funding for
813
monitoring eventual HVHF and gas pipeline operations in
814
NY state
815
x Permanent banning of HVHF within NYC and other major
816
municipal water supply watersheds, including a 5000 m
817
buffer zone from associated infrastructure (water supply
818
tunnels) and watershed perimeter, and prohibiting deep
819
directional drilling underneath such watersheds or major
820
aquifers. 821
x Mandating a minimum of 90% produced water recycling,
822
minimum 20% biodiesel or CNG for all drilling site and
823
truck operations, banning gas venting or flaring except in
824
case of emergency, mandating immediate connection to gas
825
distribution lines. 826
Without significant investment in state and federal regulatory enforcement,
827
the intense scrutiny of thousands of gas wells necessary to avoid incremental
828
degradation of watershed protection or shallow groundwater will not be possible. 829
However, recent developments provide grounds for guarded optimism. The
830
history of mineral exploitation in the United States is one of dramatic reduction of
831
adverse environmental consequences of mining operations. 19th century
832
practices resulted in acid mine drainage and numerous Superfund site designations
833
in Colorado and other western states, but 21st century mine reclamation with
834
sufficient regulatory oversight has avoided environmental degradation in several
835
U.S. states and Canada. 836 x Immediate hiring and training of sufficient NYS DEC
inspectors (1000+) and increasing agency funding for
monitoring eventual HVHF and gas pipeline operations in
NY state x Permanent banning of HVHF within NYC and other major
municipal water supply watersheds, including a 5000 m
buffer zone from associated infrastructure (water supply
tunnels) and watershed perimeter, and prohibiting deep
directional drilling underneath such watersheds or major
aquifers. x Mandating a minimum of 90% produced water recycling,
minimum 20% biodiesel or CNG for all drilling site and
truck operations, banning gas venting or flaring except in
case of emergency, mandating immediate connection to gas
distribution lines. 6.
Discussion
633 Without significant investment in state and federal regulatory enforcement,
827
the intense scrutiny of thousands of gas wells necessary to avoid incremental
828
degradation of watershed protection or shallow groundwater will not be possible. 829
However, recent developments provide grounds for guarded optimism. The
830
history of mineral exploitation in the United States is one of dramatic reduction of
831
adverse environmental consequences of mining operations. 19th century
832
practices resulted in acid mine drainage and numerous Superfund site designations
833
in Colorado and other western states, but 21st century mine reclamation with
834
sufficient regulatory oversight has avoided environmental degradation in several
835
U.S. states and Canada. 836 34 In this situation involving politics, economics, geology, hydrology and
837
water quality, the usual methods of risk assessment may not provide useful
838
answers (Hattis and Goble, 2003) and the “precautionary principle” may be the
839
best benchmark for decision-making (Stayner et al., 2002). Maintaining the
840
current New York State moratorium on hydrofracturing of horizontal wells seems
841
a prudent step in light of the legislative uncertainty. It is likely that HVHF will be
842
highly regulated and eventually permitted in the remaining 90%+ of the Marcellus
843
shale subcrop across the southern tier of New York. Time will tell whether the
844
potential liabilities or benefits (Table 1) of natural gas exploitation from the
845
Marcellus shale will be realized. 846
847
Acknowledgements: The assistance of Alan Mason in helping to draft an earlier version
848
of this work is greatly appreciated. The comments of four anonymous reviewers also
849
were helpful in improving the manuscript. 850
851
852
853
854
855
856
857
858
859
860
861
862 In this situation involving politics, economics, geology, hydrology and
837
water quality, the usual methods of risk assessment may not provide useful
838
answers (Hattis and Goble, 2003) and the “precautionary principle” may be the
839
best benchmark for decision-making (Stayner et al., 2002). Maintaining the
840
current New York State moratorium on hydrofracturing of horizontal wells seems
841
a prudent step in light of the legislative uncertainty. It is likely that HVHF will be
842
highly regulated and eventually permitted in the remaining 90%+ of the Marcellus
843
shale subcrop across the southern tier of New York. 6.
Discussion
633 Time will tell whether the
844
potential liabilities or benefits (Table 1) of natural gas exploitation from the
845
Marcellus shale will be realized. 846 35 863 Table 1. Considerations for assessing environmental impact of Marcellus shale
864
gas drilling in New York State
865 gas drilling in New York State
865
Major Liabilities
Major Benefits
1. Degradation of watershed and
groundwater protection
x
US EPA FAD may not be
renewed, requiring NYC
water supply filtration
x
Long-term increase in chronic
contamination of NYC and
other water supply, requiring
increased storage, health
degradation of NY residents
2. Industrial infrastructure
degradation of rural landscape
x
Loss of recreational tourism
and resulting local and state
revenue
x
Fragmentation of wildlife
habitat by roads and pipelines
3. Uncertainty of effect on global
GHG emissions reduction
efforts
x
Fugitive emissions may result
in increased global warming
potential
4. Local pollution of air by
VOCs/ozone precursors
x
Long-term health effects on
local populations
x
Reduction of regional air
quality
5. Local pollution and depletion of
streams
x
Long-term health effects on
local populations
x
Ecological impact of increased
pollution on local wildlife
1. Substitution for coal in
electricity generation
x
Less particulate air emissions,
lower health impacts
x
Enables phase-out of coal to
combat climate change
x
Reduction of mountaintop
removal for coal mining
2. Transition fuel to carbon-
constrained economy
x
Lower natural gas prices
speed rather than slow
closing obsolete coal power
plants
x
Enhance baseload electrical
generation capability
substituting for intermittent
sources (e.g. wind, solar)
3. Proximity of energy source for
electrical generation to the
urban centers of the NE United
States
x
Avoids natural gas supply
bottleneck due to imports and
long-distance transportation
costs
x
Lower costs for locally-
produced gas, electricity
4. Local economic investment in
upstate NY
x
Employment year-round in
contrast to current seasonal
tourism
x Per-capita revenue for local
landowners from mineral
leasing x
Lower natural gas prices
speed rather than slow
closing obsolete coal power
plants x
Loss of recreational tourism
and resulting local and state
revenue x
Enhance baseload electrical
generation capability
substituting for intermittent
sources (e.g. wind, solar) x
Fragmentation of wildlife
habitat by roads and pipelines 3. Uncertainty of effect on global
GHG emissions reduction
efforts 3. 6.
Discussion
633 Proximity of energy source for
electrical generation to the
urban centers of the NE United
States x
Fugitive emissions may result
in increased global warming
potential x
Avoids natural gas supply
bottleneck due to imports and
long-distance transportation
costs x
Lower costs for locally-
produced gas, electricity 4. Local economic investment in
upstate NY 4. Local economic investment in
upstate NY x
Employment year-round in
contrast to current seasonal
tourism x Per-capita revenue for local
landowners from mineral
leasing 36 869 Figure Captions
870
871
Figure 1. Location of the watersheds (inset) that supply drinking water to New
872
York City (adapted from NYC DEP) in relation to the subcrop of the Marcellus
873
Formation (USGS); heavy line indicates eastern boundary of Marcellus shale
874
subcrop (Milici and Swezey, 2006). Contours in regional map indicate shale
875
thickness; dashed lighter line on inset indicates boundary of Catskill Forest
876
Reserve (NYS DEC). 877
Figure 2. History and projected growth of U.S. unconventional natural gas
878
development in different shale formations. Curves are a logistic function fitted to
879
data from the Barnett Formation in Texas, showing expected trends for other
880
production areas, including the Marcellus shale in New York State (data from
881
Hazen and Sawyer (2009))
882
883
Figure 3. Conceptual triaxial Venn diagram showing common ground among
884
domains of geology, technology, and public and regulatory acceptance for HVHF
885
in New York State. Axes show increased implementation towards origin. 886
Vertical hachures show present limited acceptance of Marcellus HVHF, angled
887
hachures show possible future consensus if regulatory and public acceptance
888
domain can be expanded and moved inward along implementation axis. 889
890
References
891
AEA Technology. Support to the identification of potential risks for the
892
environment and human health arising from hydrocarbons operations
893
involving hydraulic fracturing in Europe. Report AEA/R/ED57281 for
894
European Commission DG Environment 2012; Available at
895
http://ec.europa.eu/environment/integration/energy/pdf/fracking%20study. 896
pdf [Accessed March 2013]
897 871
Figure 1. Location of the watersheds (inset) that supply drinking water to New
872
York City (adapted from NYC DEP) in relation to the subcrop of the Marcellus
873
Formation (USGS); heavy line indicates eastern boundary of Marcellus shale
874
subcrop (Milici and Swezey, 2006). Contours in regional map indicate shale
875
thickness; dashed lighter line on inset indicates boundary of Catskill Forest
876
Reserve (NYS DEC). 877 Figure 2. History and projected growth of U.S. unconventional natural gas
878
development in different shale formations. Curves are a logistic function fitted to
879
data from the Barnett Formation in Texas, showing expected trends for other
880
production areas, including the Marcellus shale in New York State (data from
881
Hazen and Sawyer (2009))
882
883 883
Figure 3. 4. Local economic investment in
upstate NY Conceptual triaxial Venn diagram showing common ground among
884
domains of geology, technology, and public and regulatory acceptance for HVHF
885
in New York State. Axes show increased implementation towards origin. 886
Vertical hachures show present limited acceptance of Marcellus HVHF, angled
887
hachures show possible future consensus if regulatory and public acceptance
888
domain can be expanded and moved inward along implementation axis. 889
890
References
891
AEA Technology. Support to the identification of potential risks for the
892
environment and human health arising from hydrocarbons operations
893
involving hydraulic fracturing in Europe. Report AEA/R/ED57281 for
894 37 Aminto A, Olson MS. Four-compartment partition model of hazardout
898
components in hydraulic fractureing fluid additives. Journal of Natural
899
Gas Science and Engineering 2012; 7:16-21,
900
doi:10.1016/j.jngse.2012.03.0006
901
Borisova T, Collins A, D'Souza G, Benson M, Wolfe ML, Benham B. A benefit-
902
cost analysis of total maximum daily load implementation. Journal of the
903
American Water Resources Association (JAWRA) 2008; 44(4):1009-
904
1023, doi:10.1111/j.1752-1688.2008.00216.x
905
Brooks A. Optimistic NPC report could point US energy strategy in wrong
906
direction, Energy Strategy Reviews 2012; 1:57-61
907
doi:10.1016/j.esr.2011.11.002
908
Brouwer R, Martin-Ortega J, Berbel J. Spatial preference heterogeneity: a choice
909
experiment, Land Economics 2010; 86(3): 552-568
910
Bryant RB, Veith TL, Kleinman PJA, Gburek WJ. Cannonsville Reservoir and
911
Town Brook watersheds: documenting conservation efforts to protect New
912
York City's drinking water. Journal of Soil and Water Conservation 2008;
913
63(6): 339-344
914
Burnham A, Han J, Clark CE, Wang M, Dunn JB, Palou-Rivera I. Life-cycle
915
greenhouse gas emissions of shale gas, natural gas, coal and petroleum,
916
Environmental Science and Technology 2012; 46:619-627
917
dx.doi.org/10.1021/es201942m
918
Bushnell JB, Wolfram CD. Enforcement of vintage differentiated regulations: the
919
case of new source review. Journal of Environmental Economics and
920
Management 2012; 64: 137-152,
921
http://dx.doi.org/10.1016/j.jeem.2012.01.006
922
Bybee K. Optimizing completion strategies for fracture initiation in Barnett Shale
923
horizontal wells, Highlights of Ketter AA, Daniels JL, Heinze JR, Waters
924
G, SPE paper 103232. Journal of Petroleum Technology 2007; 59(3): 45-
925
46
926
Chen H, Boutros PC. VennDiagram: a package for the generation of highly-
927
customizable Venn and Euler diagrams in R. BMC Bioinformatics 2011;
928
12:35, http://www.biomedcentral.com/1471-2105/12/35
929 Aminto A, Olson MS. Four-compartment partition model of hazardout
898
components in hydraulic fractureing fluid additives. 4. Local economic investment in
upstate NY Journal of Natural
899
Gas Science and Engineering 2012; 7:16-21,
900 doi:10.1016/j.jngse.2012.03.0006
901 Borisova T, Collins A, D'Souza G, Benson M, Wolfe ML, Benham B. A benefit-
902
cost analysis of total maximum daily load implementation. Journal of the
903 Borisova T, Collins A, D Souza G, Benson M, Wolfe ML, Benham B. A benefit-
902
cost analysis of total maximum daily load implementation. Journal of the
903
American Water Resources Association (JAWRA) 2008; 44(4):1009-
904 American Water Resources Association (JAWRA) 2008; 44(4):1009-
904 1023, doi:10.1111/j.1752-1688.2008.00216.x
905 1023, doi:10.1111/j.1752-1688.2008.00216.x
905
Brooks A. Optimistic NPC report could point US energy strategy in wrong
906
direction, Energy Strategy Reviews 2012; 1:57-61
907
doi:10.1016/j.esr.2011.11.002
908
Brouwer R, Martin-Ortega J, Berbel J. Spatial preference heterogeneity: a choice
909
experiment, Land Economics 2010; 86(3): 552-568
910
Bryant RB, Veith TL, Kleinman PJA, Gburek WJ. Cannonsville Reservoir and
911
Town Brook watersheds: documenting conservation efforts to protect New
912
York City's drinking water. Journal of Soil and Water Conservation 2008;
913
63(6): 339-344
914
Burnham A, Han J, Clark CE, Wang M, Dunn JB, Palou-Rivera I. Life-cycle
915
greenhouse gas emissions of shale gas, natural gas, coal and petroleum,
916
Environmental Science and Technology 2012; 46:619-627
917
dx.doi.org/10.1021/es201942m
918
Bushnell JB, Wolfram CD. Enforcement of vintage differentiated regulations: the
919
case of new source review. Journal of Environmental Economics and
920
Management 2012; 64: 137-152,
921
http://dx.doi.org/10.1016/j.jeem.2012.01.006
922
Bybee K. Optimizing completion strategies for fracture initiation in Barnett Shale
923
horizontal wells, Highlights of Ketter AA, Daniels JL, Heinze JR, Waters
924
G, SPE paper 103232. Journal of Petroleum Technology 2007; 59(3): 45-
925
46
926 Brooks A. Optimistic NPC report could point US energy strategy in wrong
906 direction, Energy Strategy Reviews 2012; 1:57-61
907 doi:10.1016/j.esr.2011.11.002
908 doi:10.1016/j.esr.2011.11.002
908 Brouwer R, Martin-Ortega J, Berbel J. Spatial preference heterogeneity: a choice
909 Brouwer R, Martin-Ortega J, Berbel J. Spatial preference heterogeneity: a choice
909
experiment, Land Economics 2010; 86(3): 552-568
910 experiment, Land Economics 2010; 86(3): 552-568
910 Bryant RB, Veith TL, Kleinman PJA, Gburek WJ. Cannonsville Reservoir and
911
Town Brook watersheds: documenting conservation efforts to protect New
912 York City's drinking water. Journal of Soil and Water Conservation 2008;
913
63(6): 339-344
914 63(6): 339-344
914
Burnham A, Han J, Clark CE, Wang M, Dunn JB, Palou-Rivera I. Life-cycle
915
greenhouse gas emissions of shale gas, natural gas, coal and petroleum,
916
Environmental Science and Technology 2012; 46:619-627
917
dx.doi.org/10.1021/es201942m
918
Bushnell JB, Wolfram CD. Enforcement of vintage differentiated regulations: the
919
case of new source review. Journal of Environmental Economics and
920
Management 2012; 64: 137-152,
921
http://dx.doi.org/10.1016/j.jeem.2012.01.006
922
Bybee K. Optimizing completion strategies for fracture initiation in Barnett Shale
923
horizontal wells, Highlights of Ketter AA, Daniels JL, Heinze JR, Waters
924
G, SPE paper 103232. Journal of Petroleum Technology 2007; 59(3): 45-
925
46
926
Chen H, Boutros PC. VennDiagram: a package for the generation of highly-
927
customizable Venn and Euler diagrams in R. BMC Bioinformatics 2011;
928
12:35, http://www.biomedcentral.com/1471-2105/12/35
929 38 Clarke LB. Environmental aspects of coalbed methane extraction, with emphasis
930
on water treatment and disposal. Transactions of the Institution of Mining
931
and Metallurgy Section A Mining industry 1996; 105: A105-A113
932
Colorado School of Public Health. Battlement Mesa Health Impact Assessment
933
(2nd and final draft) and Environmental Health and Monitoring Study,
934
Garfield Co. Colorado 2011; Available at http://www.garfield-
935
county.com/environmental-health/battlement-mesa-health-impact-
936
assessment-draft2.aspx. [Accessed January 2013]
937
Crain LJ. Groundwater pollution from natural gas and oil production in New
938
York, New York Water Resources Commission Report of Investigation
939
RI-5, 1969; prepared by USGS in cooperation with NY State Health Dept. 940
15 p. 941 RI-5, 1969; prepared by USGS in cooperation with NY State Health Dept. 940
15 p. 941 Dhar RK, Biswas BK, Samanta G, Mandal BK, Chakraborti D, Roy S, Khan AW,
942 Ahmed SA, Hadi SA . Groundwater arsenic calamity in Bangladesh. 943 Current Science 1997; 73(1): 48-59
944 Eaton TT. On the importance of geological heterogeneity for flow simulation. 945 Sedimentary Geology 2006; 184(3-4): 187-201
946 Elcock D. Future U.S. Water Consumption: The Role of Energy Production. doi:10.1016/j.esr.2011.11.002
908 947 Journal of the American Water Resources Association (JAWRA) 2010;
948 46(3): 447-460, doi:10.1111 ⁄ j.1752-1688.2009.00413.x
949 Engelder T. Comment: Should fracking stop? - No, it's too valuable. Nature 2011;
950
477: 271-75
951 477: 271-75
951
Epstein PR, Buonocore JJ, Eckerle K, Hendryx M, Stout III BM, Heinberg R et
952
al., Full cost accounting for the life cycle of coal. Annals of the New York
953
Academy of Sciences 2011; 1219:73-98. doi: 10.1111/j.1749-
954
6632.2010.05890.x
955
Fakhru'l-Razi A, Pendashteh A, Abdullah LC, Biak DRA, Madaeni SS, Abidin
956
ZZ. Review of technologies for oil and gas produced water treatment. 957
Journal of Hazardous Materials 2009; 170 : 530-551,
958
doi:10.1016/j.jhazmat.2009.05.044
959
Flexible Flow Management Program. Agreement of the Parties to the 1954 U.S. 960
Supreme Court Decree, Effective June 1, 2012. Available at
961
http://water.usgs.gov/osw/odrm/documents/FFMP_FINAL.pdf. [Accessed
962
January 2013]
963 Epstein PR, Buonocore JJ, Eckerle K, Hendryx M, Stout III BM, Heinberg R et
952 al., Full cost accounting for the life cycle of coal. Annals of the New York
953
Academy of Sciences 2011; 1219:73-98. doi: 10.1111/j.1749-
954 Academy of Sciences 2011; 1219:73-98. doi: 10.1111/j.1749-
954 Fakhru'l-Razi A, Pendashteh A, Abdullah LC, Biak DRA, Madaeni SS, Abidin
956
ZZ. Review of technologies for oil and gas produced water treatment. 957 Journal of Hazardous Materials 2009; 170 : 530-551,
958 doi:10.1016/j.jhazmat.2009.05.044
959 39 Goodstein E. Reconciling the science and economics of climate change, Climatic
964
Change 2011; 106: 661-5, doi:10.1007/s10584-011-0039-3
965 Gleick PH. Water and Energy. Annual Reviews of Energy and the Environment
966
1994; 19: 267-299
967 Gregory KB, Vidic RD, Dzombak DA. Water management challenges associated
968
with the production of shale gas by hydraulic fracturing. Elements 2011;
969
7 181 1886 DOI 10 2113/
l
7 3 181
970 Gregory KB, Vidic RD, Dzombak DA. Water management challenges associated
968
with the production of shale gas by hydraulic fracturing. Elements 2011;
969 7:181-1886, DOI: 10.2113/gselements.7.3.181
970 Hansen JE. Scientific reticence and sea-level rise. Environmental Research
971
Letters 2007; 2 doi:10.1088/1748-9326/2/2/024002
972 Hansen JE. Scientific reticence and sea-level rise. Environmental Research
971 Letters 2007; 2 doi:10.1088/1748-9326/2/2/024002
972 Harrison SS. Contamination of aquifers by overpressuring the annulus of oil and
973
ll
G
d W
1985 23(3) 317 324
974 Harrison SS. Contamination of aquifers by overpressuring the annulus of oil and
973
gas wells. Ground Water 1985; 23(3): 317-324
974 Harrison SS. doi:10.1016/j.esr.2011.11.002
908 Contamination of aquifers by overpressuring the annulus of oil and
973
gas wells. Ground Water 1985; 23(3): 317-324
974 Hartnagel CA, Russell WL. The oil fields of New York State. AAPG Bulletin
975
1925; 9(4): 798
976 Hartnagel CA, Russell WL. The oil fields of New York State. AAPG Bulletin
975
1925; 9(4): 798
976
Harte J. Water constraints on energy development: a framework for analysis. 977
Water Resources Bulletin 1983; 19(1): 51-57
978
Harte J, El-Gasseir M. Energy and water. Science 1978; 199: 623-634
979
Hattis D, Goble R. The Red Book, risk assessment and policy analysis: the road
980
not taken. Human and Ecological Risk Assessment 2003; 9: 1297-1306. 981
doi: 10.1080/10807030390240319
982
Hazen and Sawyer. Final impact assessment report: Impact assessment of natural
983
gas production in the New York City water supply watershed. report
984
commissioned by the New York City Dept of Environmental Protection
985
2009; Available at http://www.nyc.gov/html/dep/html/press_releases/09-
986
15pr.shtml [Accessed January 2013]
987
Healy RW, Rice CA, Bartos TT, McKinley MP. Infiltration from an
988
impoundment for coal-bed natural gas, Powder River Basin, Wyoming:
989
Evolution of water and sediment chemistry. Water Resources Research
990
2008; 44(6): Article Number W06424
991
Howarth RW, Santoro R, Ingraffea A . Methane and the greenhouse-gas
992
footprint of natural gas from shale formations - a letter. Climatic Change
993
2011; 106:679-690. doi:10.1007/s10584-011-0061-5
994
Howarth RW, Ingraffea A. Comment: Should fracking stop? - Yes, it's too high
995 Harte J. Water constraints on energy development: a framework for analysis. 977 Harte J. Water constraints on energy development: a framework for analysis. 977 Harte J, El-Gasseir M. Energy and water. Science 1978; 199: 623-634
979 Hattis D, Goble R. The Red Book, risk assessment and policy analysis: the road
980
not taken. Human and Ecological Risk Assessment 2003; 9: 1297-1306. 981
doi: 10.1080/10807030390240319
982 Hazen and Sawyer. Final impact assessment report: Impact assessment of natural
983
gas production in the New York City water supply watershed. report
984
commissioned by the New York City Dept of Environmental Protection
985
2009; Available at http://www.nyc.gov/html/dep/html/press_releases/09-
986
15pr.shtml [Accessed January 2013]
987 Healy RW, Rice CA, Bartos TT, McKinley MP. Infiltration from an
988
impoundment for coal-bed natural gas, Powder River Basin, Wyoming:
989
Evolution of water and sediment chemistry. Water Resources Research
990
2008; 44(6): Article Number W06424
991 Howarth RW, Ingraffea A. Comment: Should fracking stop? doi:10.1016/j.esr.2011.11.002
908 - Yes, it's too high
995
risk. Nature 2011; 477: 271-73
996 40 Hughes JD. Will Natural Gas Fuel America in the 21st Century? Post Carbon
997
Institute; 2011. Available at http://www.postcarbon.org/report/331901-
998
report-will-natural-gas-fuel-america [Accessed January 2013]
999
Hultman NE, Rebois D, Scholten M, Ramig C. The greenhouse gas impact of
1000
unconventional gas for electricity generation. Environmental Research
1001
Letters 2011; 6 (4):1-9, doi:10.1088/1748-9326/6/4/044008
1002
Huntington G, Ksaibati K. Method for assessing heavy traffic impacts on gravel
1003
roads serving oil and gas-drilling operations. Transportation Research
1004
Record: Journal of the Transportation Research Board 2009; No.2101: 17-
1005
24
1006
ICF International. Technical assistance for the Draft Supplemental Generic EIS:
1007
Oil, Gas and Solution Mining Regulatory Program - Well Permit Issuance
1008
for Horizontal Drilling and High-Volume Hydraulic Fracturing to Develop
1009
the Marcellus Shale and Other Low Permeability Gas Reservoirs -
1010
Agreement No. 9679, 2009. report to the New York State Energy Energy
1011
Research and Development Authority. Available at
1012
www.nyserda.ny.gov/~/Publications/Research-and-Development-
1013
Technical-Reports/Other-Technical-
1014
Reports/~/media/Files/Publications/NYSERDA/ng/icf-task-1.ashx. 1015
[Accessed January 2013]
1016
IEA. World Energy Outlook 2011. Paris: International Energy
1017
Agency; 2011. Available at http://www.worldenergyoutlook.org/
1018
[Accessed January 2013]. 1019
IPCC. Climate Change 2007: Synthesis Report. Contribution of Working Groups
1020
I, II, and III to the Fourth Assessment Report of the Intergovernmental
1021
Panel on Climate Change (Core Writing Team: Pachauri RK, Reisinger A
1022
(eds.)). IPCC, Geneva, Switzerland; 2007. 1023
Jenner S, Lamadrid AJ. Shale gas vs. coal: Policy implications from
1024
environmental impact comparisons of shale gas, conventional gas, and
1025
coal on air, water and land in the United States. Energy Policy 2012;
1026
http://dx.doi.org/10.1016/j.enpol.2012.11.010
1027
Jiang M, Griffin WM, Hendrickson G, Jaramillo P, VanBriesen J, Venkatesh A. 1028
Life cycle greenhouse gas emissions of Marcellus shale gas. 1029
Environmental Research Letters 2011; 6 (3), 1-9, doi:10.1088/1748-
1030 41 9326/6/3/034014Kargbo DM, Wilhelm RG, Campbell DJ. Natural gas
1031
plays in the Marcellus Shale: challenges and potential opportunities. 1032
Environmental Science and Technology 2010; 44: 5679-5684
1033
Kerr RA. Natural gas from shale bursts onto the scene. Science 2010; 328: 1624-
1034
1626
1035
Kinnaman TC. The economic impact of shale gas extraction: a review of existing
1036
studies. Ecological Economics 2011; 70: 1243-1249. 1037
doi:10.1016/j.ecolecon.2011.02.005
1038
Konishi Y, Coggins JS. Environmental risk and welfare valuation under imperfect
1039
information. doi:10.1016/j.esr.2011.11.002
908 Resource and Energy Economics 2008; 30: 150-169,
1040
doi:10.1016/j.reseneeco.2007.05.002
1041
Kragt ME, Newham LTH, Bennett J, Jakeman AJ. An integrated approach to
1042
linking economic valuation and catchment modeling. Environmental
1043
Modeling and Software 2011; 26: 92-102. 1044
doi:10.1016/j.envsoft.2010.04.002
1045
Ladva HKJ, Craster B, Jones TGJ, Goldsmith G, Scott D. The cement-to-
1046
formation interface in zonal isolation. SPE Drilling and Completion 2005;
1047
20(3): 186-197
1048
Lauver LS. Environmental health advocacy: an overview of natural gas drilling in
1049
Northeast Pennsylvania and implications for pediatric nursing. Journal of
1050
Pediatric Nursing 2012; 27: 383-389, doi:10.1016/j.pedn.2011.07.012
1051
Lee DS, Herman JD, Elsworth D, Kim HT, Lee HS. A critical evaluation of
1052
unconventional gas recovery from the Marcellus shale, northeastern
1053
United States. KSCE Journal of Civil Engineering 2011; 15(4): 679-687
1054
Levy JI, Baxter LK, Schwartz J. Uncertainty and variability in health-related
1055
damages from coal-fired power plants in the United States. Risk Analysis
1056
2009; 29: 7. DOI: 10.1111/j.1539-6924.2009.01227.x
1057
Lu X, Salovaara J, McElroy MB. Implications of the Recent Reductions in
1058
Natural Gas Prices for Emissions of CO2 from the US Power Sector. 1059
Environmental Science and Technology 2012; 46(5): 3014-3021,
1060
dx.doi.org/10.1021/es203750k |
1061
Martin-Ortega J, Berbel J . Using multi-criteria analysis to explore non-market
1062
monetary values of water quality changes in the context of the Water
1063 9326/6/3/034014Kargbo DM, Wilhelm RG, Campbell DJ. Natural gas
1031 plays in the Marcellus Shale: challenges and potential opportunities. 1032 Environmental Science and Technology 2010; 44: 5679-5684
1033 Kerr RA. Natural gas from shale bursts onto the scene. Science 2010; 328: 1624-
1034
1626
1035 Kinnaman TC. The economic impact of shale gas extraction: a review of existing
1036
studies. Ecological Economics 2011; 70: 1243-1249. 1037 doi:10.1016/j.ecolecon.2011.02.005
1038 Konishi Y, Coggins JS. Environmental risk and welfare valuation under imperfect
1039
information. Resource and Energy Economics 2008; 30: 150-169,
1040 Kragt ME, Newham LTH, Bennett J, Jakeman AJ. An integrated approach to
1042
linking economic valuation and catchment modeling. Environmental
1043
Modeling and Software 2011; 26: 92-102. 1044 doi:10.1016/j.envsoft.2010.04.002
1045 42 Framework Directive. Science of the Total Environment 2010; 408: 3990-
1064
3997, doi:10.1016/j.scitotenv.2010.03.048
1065
Maxim L, van der Sluijs JP. Quality in environmental science for policy:
1066
assessing uncertainty as a component of policy analysis. Environmental
1067
Science 2011; 14: 482-492, doi:10.1016/j.envsci.2011.01.003
1068
McKenzie LM, Witter RZ, Newman LS, Adgate JL. doi:10.1016/j.esr.2011.11.002
908 Human health risk
1069
assessment of air emissions from development of unconventional natural
1070
gas resources. Science of the Total Environment 2012; 79-87. 1071
doi:10.1016/j.scitotenv.2012.02.018
1072
Milici R, Swezey C. Assessment of Appalachian Basin oil and gas resources:
1073
Devonian shale - Middle and Upper Paleozoic Total Petroleum System. 1074
2006. U.S. Geological Survey Open-File Report 2006-1237. Available at
1075
http://pubs.usgs.gov/of/2006/1237/. [Accessed August 2011]
1076
Mitchell AL, Casman EA. Economic incentives and regulatory framework for
1077
shale gas well site reclamation in Pennsylvania. Environmental Science
1078
and Technology 2011; 45: 9506-9514, dx.doi.org/10.1021/es2021796
1079
Molofsky, L., Connor, J., Wylie, A., Wagner, T. Methane in Pennsylvania water
1080
wells unrelated to Marcellus shale fracturing. Oil and Gas Journal 2011;
1081
109(49):54
1082
Moniz EJ, Jacoby HD, Meggs AJM. The Future of Natural Gas - Interim Report. 1083
Massachusetts Institute of Technology, Cambridge, MA: MIT Energy
1084
Initiative. Available at http://mitei.mit.edu/publications/reports-studies. 1085
[Accessed January 2013]
1086
Nissen SE, Watney WL, Xia J. High-resolution seismic detection of shallow
1087
natural gas beneath Hutchinson, Kansas. Environmental Geosciences
1088
2004a; 11(3): 129-142
1089
Nissen SE, Watney WL, Bhattacharya S, Byrnes AP, Young D. Geologic factors
1090
controlling natural gas distribution related to the January 2001 gas
1091
explosions in Hutchinson, Kansas. Kansas Geological Survey 2004b;
1092
Open-File Report 2004-21, Available at
1093
http://www.kgs.ku.edu/PRS/publication/2004/AAPG/NG_Migration/P1-
1094
02.html [Accessed January 2013]
1095
NTC. Impacts on Community Character of Horizontal Drilling and High Volume
1096
Hydraulic Fracturing in Marcellus Shale and Other Low--Permeability Gas
1097 Molofsky, L., Connor, J., Wylie, A., Wagner, T. Methane in Pennsylvania water
1080
wells unrelated to Marcellus shale fracturing. Oil and Gas Journal 2011;
1081
109(49):54
1082 Moniz EJ, Jacoby HD, Meggs AJM. The Future of Natural Gas - Interim Report. 1083
Massachusetts Institute of Technology, Cambridge, MA: MIT Energy
1084
Initiative. Available at http://mitei.mit.edu/publications/reports-studies. 1085
[Accessed January 2013]
1086 Nissen SE, Watney WL, Xia J. High-resolution seismic detection of shallow
1087
natural gas beneath Hutchinson, Kansas. Environmental Geosciences
1088
2004a; 11(3): 129-142
1089 Nissen SE, Watney WL, Bhattacharya S, Byrnes AP, Young D. Geologic factors
1090
controlling natural gas distribution related to the January 2001 gas
1091
explosions in Hutchinson, Kansas. Kansas Geological Survey 2004b;
1092
Open-File Report 2004-21, Available at
1093 http://www.kgs.ku.edu/PRS/publication/2004/AAPG/NG_Migration/P1-
1094
02.html [Accessed January 2013]
1095 NTC. Impacts on Community Character of Horizontal Drilling and High Volume
1096
Hydraulic Fracturing in Marcellus Shale and Other Low--Permeability Gas
1097 43 Reservoirs. doi:10.1016/j.esr.2011.11.002
908 revised 1/2011, report commissioned by New York State
1098
Energy Research and Development Authority. 2009. Available at
1099
http://www.nyserda.ny.gov/en/Publications/NYSERDA-General-
1100
Reports/~/media/Files/Publications/NYSERDA/ng/ntc.ashx [Accessed
1101
January 2013]
1102
NYC DEP. Filtration avoidance annual report for the period January 1 through
1103
December 31, 2009. New York City Department of Environmental
1104
Protection; 2010
1105
http://www.nyc.gov/html/dep/pdf/2009_bws_fad_annual.pdf. [Accessed
1106
January 2013]
1107
ODNR. Report on the investigation of the natural gas invasion of aquifers in
1108
Bainbridge Township of Geauga County, Ohio. Ohio Department of
1109
Natural Resources Division of Mineral Resources Management 2008. 1110
Available at http://www.dnr.state.oh.us/Portals/11/bainbridge/report.pdf. 1111
[Accessed January 2013]
1112
Osborn SG, Vengosh A, Warner NR, Jackson RB. Methane contamination of
1113
drinking water accompanying gas-well drilling and hydraulic fracturing. 1114
Proceedings of the National Academies of Science 2011; 108 (20): 8172-
1115
8176 - www.pnas.org/cgi/doi/10.1073/pnas.1100682108. 1116
O'Shea, KJ. A conceptual review of water extraction requirements associated
1117
with shale gas activities in New Brunswick. Atlantic Geology 2011; 47:34
1118
O'Sullivan F, Paltsev S. Shale gas production: potential versus actual greenhouse
1119
gas emissions, Environmental Research Letters 2012; 7: doi:10.1088/1748-
1120
9326/7/4/044030
1121
Pacala S, Socolow R. Stabilization wedges: Solving the climate problem for the
1122
next 50 years with current technologies. Science 2004; 305: 968-972
1123
Petron G, Frost G, Miller BR, Hirsch AI, Montzka SA, Karion A, et al. 1124
Hydrocarbon emissions characterization in the Colorado front range: a
1125
pilot study. Journal of Geophysical Research 2012; 117: D04304,
1126
http://dx.doi.org/ 10.1029/2011JD016360. 1127
Pyron AJ. Marcellus shale gas, hydrogeology, and the truth. Oil and Gas Journal
1128
2011; 109(13): 60-7
1129
Rahm D. Regulating hydraulic fracturing in shale gas plays: the case of Texas. 1130
Energy Policy 2011; 39: 2974-2981. doi:10.1016/j.enpol.2011.03.009
1131 44 Rahm BG, Riha SJ. Toward strategic management of shale gas development:
1132
Regional, collective impacts on water resources. Environmental Science &
1133
Policy 2012; 17: 12-23
1134
Rao NS, Easton ZM, Schneiderman EM, Zion MS, Steenhuis DTS. Modeling
1135
watershed-scale effectiveness of agricultural best management practices to
1136
reduce phosphorus loading. Journal of Environmental Management 2009;:
1137
1385-1395
1138
Rogers H. Shale gas - the unfolding story, Oxford Review of Economic Policy
1139
2011; 27(1):117-143, doi: 10.1093/oxrep/grr004
1140
Rumbach A. Natural Gas Drilling in the Marcellus Shale: Potential Impacts
1141
on the Tourism Economy of the Southern Tier. Technical Report. Southern
1142
Tier Central Regional Planning & Development Board. 2011. Available at
1143
http://www.stcplanning.org/usr/Program_Areas/Energy/Naturalgas_Resou
1144
rces/STC_RumbachMarcellusTourismFinal.pdf. doi:10.1016/j.esr.2011.11.002
908 [Accessed January 2013]
1145
Runkel AC, Tipping RG, Alexander Jr EC, Alexander SC. Hydrostratigraphic
1146
characterization of intergranular and secondary porosity in part of the
1147
Cambrian sandstone aquifer system of the cratonic interior of North
1148
America: improving predictability of hydrogeologic properties. 1149
Sedimentary Geology 2006; 184(3-4): 281-304
1150
Ruskey F, Savage CD, Wagon S. The search for simple symmetric Venn
1151
diagrams. Notices of the AMS 2006; 53(11): 1304-1311. 1152
Sadiq R, Husain T, Veitch B, Bose N. Evaluation of generic types of drilling fluid
1153
using a risk-based analytic hierarchy process. Environmental
1154
Management 2003; 32(6): 778-787
1155
Schon S. Letter: Hydraulic fracturing not responsible for methane migration. 1156
Proceedings of the National Academy of Sciences 2011; 108:37
1157
www.pnas.org/cgi/doi/10.1073/pnas.1107960108
1158
Shindell DT, Faluvegi G, Koch DM, Schmidt GA, Unger N, Bauer SE. Improved
1159
attribution of climate forcing to emissions. Science 2009; 326: 716-718
1160
Soeder DJ, Kappel WM. Water resources and natural gas production from the
1161
Marcellus Shale, U.S. Geological Survey Fact Sheet 2009–303; 2009
1162
Stayner L, Toraason M, Hattis D. Risk assessment at the crossroads of the 21st
1163
century: opportunities and challenges for research. Human and Ecological
1164
Risk Assessment 2002; 8(6): 1195-1202
1165 Rahm BG, Riha SJ. Toward strategic management of shale gas development:
1132
Regional, collective impacts on water resources. Environmental Science &
1133
Policy 2012; 17: 12-23
1134 Rao NS, Easton ZM, Schneiderman EM, Zion MS, Steenhuis DTS. Modeling
1135
watershed-scale effectiveness of agricultural best management practices to
1136
reduce phosphorus loading. Journal of Environmental Management 2009;:
1137
1385-1395
1138 Rao NS, Easton ZM, Schneiderman EM, Zion MS, Steenhuis DTS. Modeling
1135
watershed-scale effectiveness of agricultural best management practices to
1136
reduce phosphorus loading. Journal of Environmental Management 2009;:
1137
1385-1395
1138
Rogers H. Shale gas - the unfolding story, Oxford Review of Economic Policy
1139
2011; 27(1):117-143, doi: 10.1093/oxrep/grr004
1140
Rumbach A. Natural Gas Drilling in the Marcellus Shale: Potential Impacts
1141
on the Tourism Economy of the Southern Tier. Technical Report. Southern
1142
Tier Central Regional Planning & Development Board. 2011. Available at
1143
http://www.stcplanning.org/usr/Program_Areas/Energy/Naturalgas_Resou
1144
rces/STC_RumbachMarcellusTourismFinal.pdf. [Accessed January 2013]
1145
Runkel AC, Tipping RG, Alexander Jr EC, Alexander SC. Hydrostratigraphic
1146
characterization of intergranular and secondary porosity in part of the
1147
Cambrian sandstone aquifer system of the cratonic interior of North
1148
America: improving predictability of hydrogeologic properties. doi:10.1016/j.esr.2011.11.002
908 1149
Sedimentary Geology 2006; 184(3 4): 281 304
1150 Rogers H. Shale gas - the unfolding story, Oxford Review of Economic Policy
1139
2011; 27(1):117-143, doi: 10.1093/oxrep/grr004
1140 2011; 27(1):117-143, doi: 10.1093/oxrep/grr004
1140
Rumbach A. Natural Gas Drilling in the Marcellus Shale: Potential Impacts
1141
on the Tourism Economy of the Southern Tier. Technical Report. Southern
1142
Tier Central Regional Planning & Development Board. 2011. Available at
1143
http://www.stcplanning.org/usr/Program_Areas/Energy/Naturalgas_Resou
1144
rces/STC_RumbachMarcellusTourismFinal.pdf. [Accessed January 2013]
1145
Runkel AC, Tipping RG, Alexander Jr EC, Alexander SC. Hydrostratigraphic
1146
characterization of intergranular and secondary porosity in part of the
1147
Cambrian sandstone aquifer system of the cratonic interior of North
1148
America: improving predictability of hydrogeologic properties. 1149
Sedimentary Geology 2006; 184(3-4): 281-304
1150 Ruskey F, Savage CD, Wagon S. The search for simple symmetric Venn
1151 diagrams. Notices of the AMS 2006; 53(11): 1304-1311. 1152 Sadiq R, Husain T, Veitch B, Bose N. Evaluation of generic types of drilling fluid
1153 Sadiq R, Husain T, Veitch B, Bose N. Evaluation of generic types of drilling fluid
1153
using a risk-based analytic hierarchy process. Environmental
1154
Management 2003; 32(6): 778-787
1155 Sadiq R, Husain T, Veitch B, Bose N. Evaluation of generic types of drilling fluid
1153
using a risk-based analytic hierarchy process. Environmental
1154
Management 2003; 32(6): 778-787
1155 Schon S. Letter: Hydraulic fracturing not responsible for methane migration. 1156 Proceedings of the National Academy of Sciences 2011; 108:37
1157 www.pnas.org/cgi/doi/10.1073/pnas.1107960108
1158 45 Stephenson E, Doukas A, Shaw K. "Greenwashing gas: might a 'transition fuel'
1166
label legitimize carbon-intensive natural gas development?" Energy Policy
1167
2012; 452-459 http://dx.doi.org/10.1016/j.enpol.2012.04.010
1168
Struchtemeyer CG, Morrison MD, Elshahed MS. A critical assessment of the
1169
efficacy of biocides used during the hydraulic fracturing process in shale
1170
natural gas wells. International Biodeterioration & Biodegradation 2012;
1171
71: 15-21
1172
US DOE. Modern shale gas development in the United States: A Primer, Office of
1173
Fossil Energy and National Energy Technology Laboratory Report, United
1174
States Department of Energy; 2009. Available at:
1175
http://www.fossil.energy.gov/programs/oilgas/publications/naturalgas_gen
1176
eral/Shale_Gas_Primer_2009.pdf [Accessed January 2013]. 1177
US DOE-NETL. Impact of the Marcellus Shale gas play on current and future
1178
CCS activities. doi:10.1016/j.esr.2011.11.002
908 United States Dept of Energy National Energy
1179
Technology Laboratory report, 2010; Available at
1180
www.netl.doe.gov/technologies/carbon_seq/refshelf/Marcellus_CCS.pdf
1181
[Accessed January 2013]
1182
US EIA, Annual Energy Outlook 2012 Early Release Overview, Washington, DC:
1183
Energy Information Administration, United States Department of Energy;
1184
2012, Available at: http://www.eia.gov/forecasts/aeo/er/index.cfm
1185
[Accessed January 2013]. 1186
US EPA. Investigation of ground water contamination near Pavillion, Wyoming,
1187
U.S. Environmental Protection Agency draft report EPA 600/R-00/000,
1188
2011, Available at
1189
http://www.epa.gov/region8/superfund/wy/pavillion/index.html
1190
[Accessed January 2013]
1191
US EPA. Greenhouse Gas Emissions Reporting from the Petroleum and Natural
1192
Gas Industry. Background Technical Support Document. United States
1193
Environmental Protection Agency 2010, Available at
1194
http://www.epa.gov/ghgreporting/documents/pdf/2010/Subpart-
1195
W_TSD.pdf. [Accessed January 2013]
1196
US EPA. Framework for cumulative risk assessment, National Center for
1197
Environmental Assessment. U.S. Environmental Protection Agency. 2003. 1198
EPA/600/P-02/001F. Washington DC. 1199 46 US GAO. Unconventional oil and gas development: Key environmental and
1200
public health requirements. U.S. Government Accounting Office; 2012. 1201
Available at http://www.gao.gov/products/GAO-12-874 [Accessed
1202
January 2013]
1203
US NRC. Hidden Costs of Energy - Unpriced Consequences of Energy
1204
Production and Use. Expert consensus Report In Brief by the Committee
1205
on Health, Environment and other External Costs and Benefits of Energy
1206
Production and Consumption. National Academy of Sciences; 2009,
1207
Available at http://dels.nas.edu/resources/static-assets/materials-based-on-
1208
reports/reports-in-brief/hidden_costs_of_energy_Final.pdf. [Accessed
1209
January 2013]
1210
US NRC. Science and Decisions: Advancing Risk Assessment. report of the
1211
National Research Council. National Academies Press. 2008. 1212
Washington, D.C. 421 p. 1213
Van Tyne AM. History of New York State oil fields. AAPG Bulletin 1998; 82(9):
1214
1775. 1215
Warner NR, Jackson RB, Darrah TH, Osborn SG, Down A, Zhao K, White A,
1216
Vengosh A. Geochemical evidence for possible natural migration of
1217
Marcellus Formation brine to shallow aquifers in Pennsylvania. 1218
Proceedings of the National Academies of Science 2012; 109(30): 11961-
1219
11966, www.pnas.org/cgi/doi/10.1073/pnas.1121181109
1220
Watney WL, Nissen SE, Bhattacharya S, Young D. Evaluation of the Role of
1221
Evaporite Karst in the Hutchinson, Kansas Gas Explosions, January 17
1222
and 18, 2001. In Johnson, KS, Neal J T, editors. Evaporite
1223
karst and engineering/environmental problems in the United States:
1224
Oklahoma Geological Survey Circular 109; 2003 p.119-147. 1225
Watson TL, Bachu S. Evaluation of the potential for gas and CO2 leakage along
1226
wellbores. SPE Paper 106817. SPE Drilling and Completion 2009; 24(1). Figure 1
Click here to download Figure: EatonSTEFig1.eps Figure 2
Click here to download Figure: EatonSTEFig2.eps doi:10.1016/j.esr.2011.11.002
908 1227
115-126
1228
Waxman HA, Markey EJ, DeGette D. Chemicals used in hydraulic fracturing. 1229
United States House of Representatives Committee on Energy and
1230
Commerce 2011; Available at
1231
http://democrats.energycommerce.house.gov [Accessed March 2013]
1232 US NRC. Science and Decisions: Advancing Risk Assessment. report of the
1211 National Research Council. National Academies Press. 2008. 1212 Washington, D.C. 421 p. 1213 Van Tyne AM. History of New York State oil fields. AAPG Bulletin 1998; 82(9):
1214
1775. 1215 Warner NR, Jackson RB, Darrah TH, Osborn SG, Down A, Zhao K, White A,
1216 ,
,
,
,
,
,
,
Vengosh A. Geochemical evidence for possible natural migration of
1217
Marcellus Formation brine to shallow aquifers in Pennsylvania. 1218
Proceedings of the National Academies of Science 2012; 109(30): 11961-
1219
11966, www.pnas.org/cgi/doi/10.1073/pnas.1121181109
1220
Watney WL, Nissen SE, Bhattacharya S, Young D. Evaluation of the Role of
1221
Evaporite Karst in the Hutchinson, Kansas Gas Explosions, January 17
1222
and 18, 2001. In Johnson, KS, Neal J T, editors. Evaporite
1223
karst and engineering/environmental problems in the United States:
1224
Oklahoma Geological Survey Circular 109; 2003 p.119-147. 1225
Watson TL, Bachu S. Evaluation of the potential for gas and CO2 leakage along
1226
wellbores. SPE Paper 106817. SPE Drilling and Completion 2009; 24(1). 1227
115-126
1228
Waxman HA, Markey EJ, DeGette D. Chemicals used in hydraulic fracturing. 1229
United States House of Representatives Committee on Energy and
1230
Commerce 2011; Available at
1231
http://democrats.energycommerce.house.gov [Accessed March 2013]
1232 Vengosh A. Geochemical evidence for possible natural migration of
1217 Marcellus Formation brine to shallow aquifers in Pennsylvania. 1218 Proceedings of the National Academies of Science 2012; 109(30): 11961-
1219 47 Weber CL, Clavin C. Life cycle carbon footprint of shale gas: review of evidence
1233
and implications. Environmental Science and Technology 2012; 46: 5688-
1234
5695. dx.doi.org/10.1021/es300375n
1235
Wiseman HJ, State Enforcement of Shale Gas Development Regulations,
1236
Including Hydraulic Fracturing. Figure 3
Click here to download Figure: EatonSTEFig3.eps doi:10.1016/j.esr.2011.11.002
908 FSU College of Law 2012; Public Law
1237
Research Paper Forthcoming; Available at SSRN:
1238
http://ssrn.com/abstract=1992064 or
1239
http://dx.doi.org/10.2139/ssrn.1992064 [Accessed January 2013]
1240
1241
1242
1243 48 Figure 1
Click here to download Figure: EatonSTEFig1.eps Figure 2
Click here to download Figure: EatonSTEFig2.eps Figure 3
Click here to download Figure: EatonSTEFig3.eps
|
https://openalex.org/W2028799066
|
https://www.hal.inserm.fr/inserm-00915813/file/1471-2431-13-202.pdf
|
English
| null |
Is early detection of abused children possible?: a systematic review of the diagnostic accuracy of the identification of abused children
|
BMC pediatrics
| 2,013
|
cc-by
| 10,041
|
To cite this version: Marion Bailhache, Valériane Leroy, Pascal Pillet, Louis-Rachid Salmi. Is early detection of abused
children possible?: a systematic review of the diagnostic accuracy of the identification of abused
children.. BMC Pediatrics, 2013, 13 (1), pp.202. 10.1186/1471-2431-13-202. inserm-00915813 Is early detection of abused children possible?: a
systematic review of the diagnostic accuracy of the
identification of abused children. Marion Bailhache, Valériane Leroy, Pascal Pillet, Louis-Rachid Salmi Is early detection of abused children possible?: a
systematic review of the diagnostic accuracy of the
identification of abused children. * Correspondence: marion.bailhache@free.fr
1CHU de Bordeaux, Pole de pediatrie, F-33000 Bordeaux, France
2Centre INSERM U897-Epidemiologie-Biostatistique, University Bordeaux,
ISPED, F-33000 Bordeaux, France
Full list of author information is available at the end of the article Marion Bailhache1,2,3*, Valériane Leroy2,3, Pascal Pillet1 and Louis-Rachid Salmi2,3,4 Marion Bailhache1,2,3*, Valériane Leroy2,3, Pascal Pillet1 and Louis-Rachid Salmi2,3,4 Marion Bailhache1,2,3*, Valériane Leroy2,3, Pascal Pillet1 and Louis-Rachid Salmi2,3,4 HAL Id: inserm-00915813
https://inserm.hal.science/inserm-00915813v1
Submitted on 9 Dec 2013 L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés. HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entific research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers. Bailhache et al. BMC Pediatrics 2013, 13:202
http://www.biomedcentral.com/1471-2431/13/202 Open Access Abstract Background: Early detection of abused children could help decrease mortality and morbidity related to this major
public health problem. Several authors have proposed tools to screen for child maltreatment. The aim of this
systematic review was to examine the evidence on accuracy of tools proposed to identify abused children before
their death and assess if any were adapted to screening. Methods: We searched in PUBMED, PsycINFO, SCOPUS, FRANCIS and PASCAL for studies estimating diagnostic
accuracy of tools identifying neglect, or physical, psychological or sexual abuse of children, published in English or
French from 1961 to April 2012. We extracted selected information about study design, patient populations,
assessment methods, and the accuracy parameters. Study quality was assessed using QUADAS criteria. Results: A total of 2 280 articles were identified. Thirteen studies were selected, of which seven dealt with physical
abuse, four with sexual abuse, one with emotional abuse, and one with any abuse and physical neglect. Study
quality was low, even when not considering the lack of gold standard for detection of abused children. In 11
studies, instruments identified abused children only when they had clinical symptoms. Sensitivity of tests varied
between 0.26 (95% confidence interval [0.17-0.36]) and 0.97 [0.84-1], and specificity between 0.51 [0.39-0.63] and
1 [0.95-1]. The sensitivity was greater than 90% only for three tests: the absence of scalp swelling to identify
children victims of inflicted head injury; a decision tool to identify physically-abused children among those
hospitalized in a Pediatric Intensive Care Unit; and a parental interview integrating twelve child symptoms to
identify sexually-abused children. When the sensitivity was high, the specificity was always smaller than 90%. Conclusions: In 2012, there is low-quality evidence on the accuracy of instruments for identifying abused children. Identified tools were not adapted to screening because of low sensitivity and late identification of abused children
when they have already serious consequences of maltreatment. Development of valid screening instruments is a
pre-requisite before considering screening programs. Methods: We searched in PUBMED, PsycINFO, SCOPUS, FRANCIS and PASCAL for studies estimating diagnostic
accuracy of tools identifying neglect, or physical, psychological or sexual abuse of children, published in English or
French from 1961 to April 2012. We extracted selected information about study design, patient populations,
assessment methods, and the accuracy parameters. Study quality was assessed using QUADAS criteria. Keywords: Child abuse, Child neglect, Systematic review, Diagnostic accuracy © 2013 Bailhache et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited. © 2013 Bailhache et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited. Background issue worldwide. Gilbert et al. estimated that every year
in high-income countries about 4 to 16% of children
were physically abused, one in ten was neglected or psy-
chologically abused, and between 5 and 10% of girls and
up to 5% of boys were exposed to penetrative sexual
abuse during childhood [2]. Child maltreatment can
cause death of the child or major consequences on men-
tal and physical health, such as post-traumatic stress dis-
order and depression, in childhood or adulthood [2]. WHO estimated that 155 000 deaths in children younger The World Health Organization (WHO) defines child
maltreatment as “all forms of physical and/or emotional
ill-treatment, sexual abuse, neglect or negligent treat-
ment or commercial or other exploitation, resulting in
actual or potential harm to the child’s health, survival,
development or dignity” [1]. It is a major public health * Correspondence: marion.bailhache@free.fr
1CHU de Bordeaux, Pole de pediatrie, F-33000 Bordeaux, France
2Centre INSERM U897-Epidemiologie-Biostatistique, University Bordeaux,
ISPED, F-33000 Bordeaux, France
Full list of author information is available at the end of the article Bailhache et al. BMC Pediatrics 2013, 13:202
http://www.biomedcentral.com/1471-2431/13/202 Page 2 of 11 age. However two of these three studies were limited to
a subset of children admitted with severe injuries. Besides, assessments by the medical team were rarely
based on standardized criteria, and therefore not re-
producible and usable in practice [13]. The same team
published another study about the same markers (age,
repeated attendance, and type of injury) to identify chil-
dren victims of physical abuse or neglect among injured
children attending Emergency departments [14]. They
found no evidence that any of the markers were
sufficiently accurate. Thus these two large studies only
reviewed the accuracy of tests for two types of child
abuse among children who attended Emergency depart-
ments and already had injuries. A last study had initially
the aim of evaluating the accuracy of tools identifying
early abused children, but only reported an accuracy
assessment of tools identifying high-risk parents before
occurrence of child maltreatment [15]. than 15 years occurred worldwide in 2000 as a result of
abuse or neglect [3]. In France, a retrospective study carried out in three
regions from 1996 to 2000 showed that many children
who died from abuse were not identified as abused be-
fore their deaths. Background After excluding clear neonaticides, 25
of 53 (47%) infants who died from suspicious or violent
death had signs of prior abuse, such as fractures of
different ages, discovered during post-mortem investiga-
tions. Only eight of these children were already known
to be victims of abuse [4]. Similarly, only 33% of children
who were born in California between 1999 and 2006 and
died from intentional injury during the first five years of
life had been previously reported to Child Protection
Services [5]. Consequently, children who died from child
maltreatment can be victims of chronic child abuse while
they were not diagnosed before their death. Systematic
early detection of abused children could help prevent
these deaths and lessen child maltreatment-related mor-
bidity. However, as in usual screening programs, it is
important to balance potential positive and negative
effects and to determine the conditions for a screening
program of child maltreatment to be effective. A first
necessary condition is the availability of a test identifying
correctly abused children before they have serious or irre-
versible consequences of maltreatment. The aim of our study was to review the evidence on
the accuracy of instruments for identifying abused
children during any stage of child maltreatment evo-
lution before their death, and to assess if any might
be adapted to screening, that is if accurate screening
instruments were available. We define as instruments
any reproducible assessment used in any types of
setting. Diagnostic accuracy of ocular signs in abusive head
trauma and clinical and neuroradiological features asso-
ciated with abusive head trauma have been already syn-
thesized [6-9]. In the reviewed studies, however, markers
identified children when they had already serious conse-
quences of child maltreatment. Sometimes the diagnosis
had been done when the child was dead. Furthermore,
the diagnostic accuracy of markers was not always esti-
mated, the analysis being limited to estimating the asso-
ciation between a marker and maltreatment. Similarly,
diagnostic accuracy of genital examination for identify-
ing sexually abused prepubertal girls was reviewed [10],
but tools only identified children who were victims of a
severe form of sexual abuse (genital contact with pene-
tration). Furthermore, the sensitivity for several potential
markers, such as hymeneal transections, deep notches or
perforations, was never reported. Information sources and search terms Electronic searches were carried using PUBMED data-
base from 1966 to April 2012, PsycINFO database from
1970 to April 2012, SCOPUS database from 1978 to April
2012, PASCAL and FRANCIS databases from 1961 to
April 2012, to identify articles published in French or
English. Search terms used were child abuse, child mal-
treatment, battered child syndrome, child neglect, Munch-
ausen syndrome, shaken baby syndrome, child sexual
abuse, combined with sensitivity, specificity, diagnostic ac-
curacy, likelihood ratio, predictive value, false positive,
false negative, validity, test validation, and diagnosis,
measurement, psychodiagnosis, medical diagnosis, screen-
ing, diagnosis imaging, physical examination, diagnostic
procedure, scoring system, diagnostic, scoring system, score,
assessment (Table 1). Several authors have already considered screening in
emergency departments [11-13]. A large study in the
United Kingdom evaluated the accuracy of potential
makers: child age, type of injuries, incidence of repeat
attendance, and the accuracy of clinical screening as-
sessments for detecting physical abuse in injured chil-
dren attending Accident and Emergency departments
[13]. They found no relevant comparative studies for in-
cidence of repeat attendance, only one study which re-
ported a direct comparison of type of injury in abused
and non-abused children, and three studies for child Data collection process, data items and analysis
Th f
f
l
d
d The first assessment of selected papers was done by MB,
and results were discussed in regular meetings by both ep-
idemiologists MB and LRS. To reduce the likelihood that
potentially relevant articles were missed, reference lists
from relevant articles were checked. From each included
study, we abstracted information about study design,
population characteristics, number of participants, screen-
ing instrument or procedure, abuse or neglect outcome,
and estimates of diagnostic accuracy. Results were not
mathematically pooled due to varying methods and types
of child abuse identified. the information and method to carry the assessment,
and not only the result of this assessment. As there is no
gold standard for detecting child maltreatment, we de-
fined acceptable reference standards as: expert assess-
ments, such as child’s court disposition; substantiation
by the child protection services or other social services;
diagnosis by a medical, social or judicial team using one
or several information sources (caregivers or child inter-
view, child symptoms, child physical examination, and
other medical record review). The assessment made only
by the caregiver was not accepted because 80% or more
of maltreatment, other than sexual abuse, has been esti-
mated to be perpetrated by parents or parental guardians
[2]. Thus, the caregiver likely would not want to reveal
that his child is maltreated. Comparative studies of any
design examining the results of tools identifying abused
children in two population groups (abused children and
not abused children) were accepted (case control, cohort, the information and method to carry the assessment,
and not only the result of this assessment. As there is no
gold standard for detecting child maltreatment, we de-
fined acceptable reference standards as: expert assess-
ments, such as child’s court disposition; substantiation
by the child protection services or other social services;
diagnosis by a medical, social or judicial team using one
or several information sources (caregivers or child inter-
view, child symptoms, child physical examination, and
other medical record review). The assessment made only
by the caregiver was not accepted because 80% or more
of maltreatment, other than sexual abuse, has been esti-
mated to be perpetrated by parents or parental guardians
[2]. Thus, the caregiver likely would not want to reveal
that his child is maltreated. Comparative studies of any
design examining the results of tools identifying abused
children in two population groups (abused children and
not abused children) were accepted (case control, cohort, Eligibility criteria To be included in this analysis, articles had to 1) state as
an objective to estimate at least one accuracy parameter
(sensitivity, specificity, predictive value or likelihood ra-
tio) of a test identifying abused children (persons under
age 18); 2) include a reference standard to determine
whether a child had actually been abused; and 3) de-
scribe the assessed test, e. g. when the authors presented Page 3 of 11 Bailhache et al. BMC Pediatrics 2013, 13:202
http://www.biomedcentral.com/1471-2431/13/202 Table 1 Search terms used to identify potentially eligible
articles
Database
Search terms
PUBMED
(“child abuse” [Mesh] or “child maltreatment”)
AND
(“sensitivity and specificity” [Mesh] OR “sensitivity” OR
“specificity” OR “diagnostic accuracy” OR “likelihood ratio”
OR “predictive value” OR “false positive” OR “false
negative”)
PsycINFO
(“battered child syndrome” OR “child abuse”)
AND
(“diagnosis” OR “measurement” OR “psychodiagnosis” OR
“medical diagnosis” OR “screening”)
SCOPUS
(“child abuse” OR “child maltreatment” OR “child neglect”
OR “battered child syndrome” OR “munchausen
syndrome” OR “shaken baby syndrome”)
AND
(“diagnosis” OR “measurement” OR “screening” OR
“diagnostic imaging” OR “physical examination” OR
“diagnostic procedure” OR “scoring system”)
AND
(“predictive value” OR “diagnostic accuracy” OR
“likelihood ratio” OR “sensitivity” OR “specificity”)
FRANCIS/
PASCAL
(“child abuse” OR “child maltreatment” OR “child neglect”
OR “child sexual abuse” OR “battered child syndrome” OR
“munchausen syndrome” OR “shaken baby syndrome”)
AND
(“diagnosis” OR “measurement” OR “screening” OR
“physical examination” OR “diagnostic” OR “scoring
system” OR “score” OR “assessment”)
AND
(“test validation” OR “validity” OR “sensitivity” OR
“specificity” OR “predictive value” OR “diagnostic
accuracy” OR “likelihood ratio”) and cross-sectional studies). Descriptive studies with only
one group of abused or not abused children, of which the
aim was to estimate one accuracy parameter, were also ac-
cepted. To avoid missing any potentially relevant tool, no
particular setting nor category of patients were used as in-
clusion or exclusion criteria. We did not consider tests to identify abusive caregivers,
abused children after their death or children victims of
intimate-partner violence. Articles were also excluded
when they did not provide original data. Tests that identi-
fied abused children after their death were excluded as
they are by definition not relevant for early detection. Intimate-partner violence, regarded as a separate form of
child maltreatment by several authors, was excluded be-
cause the main victim is not the child [2]. Study selection
Eli ibilit
f t Eligibility of studies was checked by a junior epidemiolo-
gist and pediatrician (MB), from April, 2012 to May, 2012,
and the resulting selection checked by a senior medical
epidemiologist (LRS). Articles were first screened by titles. They were excluded when the title showed that the article
did not address accuracy of tools identifying abused chil-
dren. If the title did not clearly indicate the article’s sub-
ject, the summary was read. Abstracts were retained for
full review when they met the inclusion criteria or when
more information was required from the full text to ascer-
tain eligibility. Three criteria related to the index test: with a high enough specificity to avoid stigmatization of
caretakers who were not abusers. 3) the index test was described in sufficient details to
permit replication;
4) when the index test was a score, the cutoff was
determined before results were available;
5) the index test was interpreted without knowledge of
the results of the reference standard. 3) the index test was described in sufficient details to
permit replication; Quality of studies
Th
i 14) uninterpretable test results were well-balanced be-
tween the reference standard and the index test. y
The maximum number of quality criteria met was eight
of fourteen, and five studies met four or less criteria
(Table 2). The accuracy of the reference standard was
never determined because no gold standard to identify
abused children is available. We could not judge patients
representativeness, by lack of sufficient information about
methods of patient recruitment [21,24,26,28,30-33], or re-
fusal by many families, for undocumented reasons [22,23]. In three studies, details on the imaging technique or
assessment of impact trauma were not sufficiently de-
scribed to replicate the index test [25,27,28]. The reference
standard was different in the three case–control studies
[21,22,31]. In one study, the result of the index test was
used to establish the final diagnosis [23]. The time period
between the two tests was rarely available; in one study, it
was on average 36.4 weeks, so that the situation about
child abuse could have changed [33]. We could not judge
if the circumstances of test evaluation were the same than
in routine practice, by lack of information about the kind
of practice considered [22,25-29,31,33]. One criterion related to applicability: One criterion related to applicability: 15) same clinical data available when test results were
interpreted as would be available when the test is
used in practice. 15) same clinical data available when test results were
interpreted as would be available when the test is
used in practice. Quality of studies was summarized by counting the
number of criteria that were respected. Results of the final
selection and analysis where reviewed by another senior
medical epidemiologist (VL) and a senior pediatrician (PP). Study selection 4) when the index test was a score, the cutoff was
determined before results were available; Of 2 280 references identified in the databases, 524 were
selected from their title, of which 137 abstracts were
read; after exclusion of duplicates, 92 full articles were
assessed (Figure 1). Studies excluded for lack of refer-
ence standard were case–control studies with control
groups recruited in the general population without verify-
ing if children were abused or not. Studies were excluded
when the reference standard was only the opinion of care-
givers who had been asked whether their children were
abused or not. One study was excluded because the
method of the index text, an assessment by primary care
clinicians, was not described [19]. Finally, one study was
excluded because an unknown number of children less
than fifteen years old examined in a medical center, who
should have been tested during the study period, had not
received the index test but were not registered [20]. This
limit was noticed because several abused children identi-
fied by the reference standard and who had inclusion cri-
teria, had not received the index test by the medical team
and were not reported. Thirteen articles met the inclusion
criteria. The outcome of interest was sexual abuse in four
studies [21-24], physical abuse in seven [25-31], psycho-
logical abuse in one [32], and several forms of child mal-
treatment (physical abuse, psychological abuse, sexual
abuse, and physical neglect) in one [33]. Eight studies were
prospective [21-26,32,33], and five retrospective assess-
ment of the diagnostic accuracy [27-31]. 5) the index test was interpreted without knowledge of
the results of the reference standard. Three criteria related to the reference standard: Three criteria related to the reference standard: 7) the reference standard was described in sufficient
details to permit replication; 8) the reference standard was interpreted without
knowledge of the results of the index test. One criterion related to both the index test and refer-
ence standard: 9) the reference standard and the index test were
independent. Five criteria related to flow and timing: 10) the whole population or a random selection
received the reference standard; 11) the study population received the same reference
standard; 12) the time period between the reference standard
and the index test was short enough so the
situation of the child did not change; 13) uninterpretable test results were reported; Quality assessment The selected studies were assessed by MB and reviewed
by LRS, using the QUADAS-1 criteria to assess quality
of studies of diagnostic accuracy [16]. The standardized
checklist included 15 criteria, grouped according to the
domains defined by QUADAS-2 [17]. Two criteria related to patient selection: 1) patients were representative of a spectrum of
population including all stages of maltreatment
before the death of the child; 2) selection criteria were well described. Bailhache et al. BMC Pediatrics 2013, 13:202
http://www.biomedcentral.com/1471-2431/13/202 Page 4 of 11 Identification of physical abuse Four studies were about children with inflicted head in-
jury (Table 3) [25-28]. One test identified abused chil-
dren among those admitted to a tertiary care pediatric
hospital for acute traumatic intracranial injury, when
caregivers reported no history of trauma or a history of
low-impact trauma, i.e. with a fall from ≤3 feet or with
other low-impact non-fall mechanisms [27]. The other
tests identified abused children by using findings of phys-
ical examination or Computer Tomographic among chil-
dren hospitalized in Pediatric Intensive Care Units [25,26],
Neurosurgical [25,26] or Emergency departments [25,26]
or a regional pediatric medical center [28] for head trauma. A prediction rule combining four variables (hygroma; con-
vexity subdural hematoma without hygroma; no fracture;
and interhemispheric subdural hematoma in Computer
Tomographic images at clinical presentation) could iden-
tify 84% of abused children [28]. Diagnostic accuracy possible child abuse to a specialized team [30]. Finally, a
score was developed to identify physical abused children
14 years old or younger, with at least one diagnosis of in-
jury as defined by the International Classification of Dis-
ease (ICD-9), 9the revision (codes 800 to 959), in 1961
hospitals in 17 states of the United States. The 26-point
score based on presence of fracture of base or vault of
skull (1 point), eye contusion (3 points), rib fracture (3
points), intracranial bleeding (4 points), multiple burns
(3 points), and age of the child (3 points for age group
1-3 y, 12 points for age group 0-1 y) identified 87% of
physical abused child when the score was ≥3 [29]. Assessment of tools adaptation to screening Tools were considered adapted to screening, according
to the WHO criteria on the adequacy of tests used in
screening programs [18], if they fulfilled the following
criteria: 1) identify abused children before they have serious
consequences of child maltreatment; 2) identify abused
children with a high sensitivity; 3) identify abused children Bailhache et al. BMC Pediatrics 2013, 13:202
http://www.biomedcentral.com/1471-2431/13/202 Page 5 of 11 Figure 1 Diagram illustrating the study selection process, April 2012. Figure 1 Diagram illustrating the study selection process, April 2012. Identification of sexual abuse The sensitivity of tests using the results of children anal
and genital examination were estimated at best at 56%
(95% CI: 33-77), and the specificity at 98% (95% CI: 91-
100) [22,23] (Table 4). The frequency of a variety of sex-
ual behaviors of the child over the previous six months
prior to assessment was not associated with sexual abuse
[24]. A list of 12 symptoms expressed by the child, such
as difficulty getting to sleep, change to poor school per-
formance, or unusually interest about sex matters, iden-
tified sexual abused children when caretakers reported
at least three symptoms, with a sensitivity of 91% and a
specificity of 88% [21]. The setting in which the studies
took place were consultations with specialized team in
child abuse, or when a control group was chosen, con-
sultations at pediatric clinics for well-child examination
or others complaints. Three studies estimated accuracy of tests identifying
physical abuse and were not limited to intentional head
trauma [29-31]. A decision tool based on three questions
(age of child; localization of bruise during the initial 72
hours of patient’s admission; and confirmation of acci-
dent in public setting) identified abused children among
children aged 0 to 4 y admitted to a Pediatric Intensive-
Care Unit, with a sensitivity of 97% (95% CI: 84-100)
[31]. In another study, presence of bruises in the same
body site than a fracture identified 26% of abused chil-
dren among children with acute fractures referred for Table 2 Quality of studies of the diagnostic accuracy of tests identifying child neglect or abuse
Criteria of quality
Studies
Berenson
et al, 2002
[22]
Bernstein
et al,
1997 [33]
Chang
et al,
2005 [29]
Cheung
et al,
2004 [23]
Drach et al,
2001 [24]
Fernando-
pulle et al,
2003 [32]
Hettler et al,
2003 [27]
Pierce et al,
2010 [31]
Valvano
et al,
2009 [30]
Vinchon
et al,
2010 [25]
Vinchon
et al,
2005 [26]
Wells
et al,
2002 [28]
Wells
et al,
1997 [21]
1. Representative spectrum
of patients
Unclear
Unclear
Yes
Unclear
Unclear
Unclear
Yes
Unclear
Unclear
No
Unclear
Unclear
Unclear
2. Description of selection
criteria
Yes
No
Yes
No
No
No
Yes
No
No
Yes
No
No
No
3. Replication of the index
test
Yes
Yes
Unclear
Yes
Yes
Yes
No
Yes
Yes
No
Unclear
No
Yes
4. Identification of sexual abuse Cutoff determined before
results were available
Yes
No
No
NA*
Yes
No
NA*
No
NA*
NA*
NA*
No
No
5. Interpretation without
knowledge of the results of
reference standard
Unclear
Yes
Unclear
Unclear
Unclear
Yes
Unclear
No
Unclear
Unclear
Unclear
Yes
Unclear
6. Classification by reference
standard
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
7. Replication of the reference
standard
No
No
No
Yes
No
No
No
No
No
No
No
No
No
8. Interpretation without
knowledge of the results
of index test
Unclear
Yes
Unclear
Unclear
Yes
Yes
Unclear
Yes
Yes
Unclear
Unclear
Yes
Unclear
9. Independence of
reference and index tests
Yes
Unclear
Unclear
No
Yes
Yes
Yes
Unclear
Yes
Unclear
Unclear
Unclear
Unclear
10. Systematic reference
standard
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
11. Same reference standard
No
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
No
12. Short enough time
period between reference and
index tests
Yes
No
Yes
Unclear
Unclear
Unclear
Yes
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
13. Uninterpretable results
reported
Yes
No
No
No
No
No
Unclear
No
No
No
No
No
No
14. Uninterpretable results
balanced
Yes
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
15. Same clinical data
available as in routine
Unclear
Unclear
Unclear
Yes
Yes
Yes
Unclear
Unclear
Unclear
Unclear
Unclear
Unclear
No
*NA Not Applicable Bailhache et al. Identification of sexual abuse †† Seven body sites: four extremities, torso, pelvis and head/neck. ‡‡ ICD International Classification of Diseases, Ninth Revision. §§ SIPCA Screening Index for Physical Child Abuse. *PICU Pediatric Intensive Care Unit. † ND Neurosurgical Department. ‡ HT Head Trauma. § RH Retinal Hemorrhage. ‖ SDH Subdural Hematoma. ¶ CT Computed Tomographic. **CPS Child Protection Service. †† Seven body sites: four extremities, torso, pelvis and head/neck. ‡‡ ICD International Classification of Diseases, Ninth Revision. §§ SIPCA Screening Index for Physical Child Abuse. Accuracy was estimated for each form of child maltreat-
ment in an adolescent psychiatric population. Physical neg-
lect was defined as the failure of caretakers to provide for a
child’s basic physical needs like food or clothing. The esti-
mated sensitivity and specificity were the best for sexual
abuse. The sensitivity were estimated at 86% (95% CI: 71-
94), and the specificity at 76% (95% CI: 67-83) [33]. Accuracy was estimated for each form of child maltreat-
ment in an adolescent psychiatric population. Physical neg-
lect was defined as the failure of caretakers to provide for a
child’s basic physical needs like food or clothing. The esti-
mated sensitivity and specificity were the best for sexual
abuse. The sensitivity were estimated at 86% (95% CI: 71-
94), and the specificity at 76% (95% CI: 67-83) [33]. Identification of psychological abuse In a self-administered questionnaire, children were ex-
pected to indicate how often they experienced a given
parental/caregiver behavior (Table 4). The scale was ad-
ministered to children aged 13-15 years without spe-
cific complaints attending a school within the city of
Colombo. At a cutoff of 95 and greater, 20 of 26 abused
children were identified [32]. Adaptation to screening Identification of several forms of child maltreatment
The Childhood Trauma Questionnaire is a 70-item screen-
ing inventory that assesses self-reported experiences of
abuse and neglect in childhood and adolescence (Table 4). Identification of sexual abuse BMC Pediatrics 2013, 13:202
http://www.biomedcentral.com/1471-2431/13/202 Page 7 of 11 Table 3 Description of selected studies estimating diagnostic accuracy of tests identifying physical abused children
Source
Inclusion criteria
Form of
child abuse
Index test
Sample
size
Reference standard
Sensitivity Specificity
%
(95% CI)
%
(95% CI)
Vinchon et al,
2010 [25]
Children <2 y referred
alive to Emergency, PICU*
or ND† for HT‡ with
cerebral scan
Inflicted
head injury
Severe RH§
84
Assessment by forensic
neurosurgeon, pediatrician,
psychologist, social worker
57
97
Brain ischemia
27
97
SDH‖
27
97
No scalp swelling
98
77
Vinchon et al,
2005 [26]
Children <2 y referred
alive to Emergency, PICU*
or ND† for HT‡ with
cerebral scan
Inflicted
head injury
RH § Grade 1, 2 or 3
207
Assessment by forensic
neurosurgeon, pediatrician,
psychologist,
ophthalmologist, social
worker
75(62-86)
93(85-78)
RH § Grade 2 or 3
66(52-78)
100(95-100)
Hettler et al,
2003 [27]
Children < 3 y
hospitalized for HT‡ with
intracranial hemorrhage
Inflicted
head injury
No history of trauma
or low-impact trauma
163
Assessment by medical team
integrating witnessed or
confessed abuse, predefined
specific findings during
physical child examination
69(55-82)
97(83-100)
Wells et al,
2002 [28]
Children <3 y hospitalized
for HT‡ with intracranial
hemorrhage
Inflicted
head injury
Score integrating CT¶
imaging patterns
257
Assessment by medical team,
integrating history, age and
sex of child, results of official
investigation, medical records
excluding CT¶
84(78-90)
83(74-90)
Pierce et al,
2010 [31]
Newborn to 4 y
hospitalized in PICU* for
trauma
Physical
abuse
Decision tool
integrating bruise
region, age of child,
trauma history
95
Assessment by medical,
juridical team, and CPS**
97(84-100)
84(69-94)
Valvano et al,
2009 [30]
Children <18 y referred to
specialized team with
fracture, excluded head
Physical
abuse
Bruise in the same
body sites†† than
fracture
150
Expert assessment integrating
history, type of injuries and
familial characteristics
26(17-36)
75(62-86)
Chang et al,
2005 [29]
children ≤14 y with at
least one trauma
diagnostic with ICD-9‡‡
Physical
abuse
SIPCA§§, score
integrating age of
child, physical
examination and
results of imaging
58 558
E codes and certain ICD-9
codes‡‡
87(84-90)
81(81-81) ting diagnostic accuracy of tests identifying physical abused children selected studies estimating diagnostic accuracy of tests identifying physical abused children *PICU Pediatric Intensive Care Unit. † ND Neurosurgical Department. ‡ HT Head Trauma. § RH Retinal Hemorrhage. ‖ SDH Subdural Hematoma. ¶ CT Computed Tomographic. **CPS Child Protection Service. Identification of several forms of child maltreatment Assessment by nurse
integrating D/P vulvar
Penetration Rating Scale§
29 (22-36)
86 (81-91)
Drach et al,
2001 [24]
Children 2-12 y
referred to SCAP
team‖
Sexual abuse
209
CSBI† parental
interview about child
sexual behavior
Expert assessment
integrating child
interview, history and
physical examination
50 (37-63)
50 (42-58)
Wells et al,
1997 [21]
Boy < 18 y referred
to CPS or consulting
for well-child
examination
Sexual abuse
74
SASA¶, parental
interview integrating
12 child symptoms
Assessment by CPS or by
a series of screening
techniques
91 (71-99)
88 (77-96)
Fernan-dopulle
et al, 2003 [32]
Children
Emotional abuse
98
Self-report
questionnaire directed
to children
Psychiatrist’s assessment
during child interview
77 (56-91)
51 (39-63)
13-15 y in school
Bernstein et al,
1997 [33]
Children
Physical abuse
190
CTQ**, self-report
questionnaire directed
to children
Assessment by therapists
integrating structured
child interview, follow-up
information and assess-
ment of CPS†
82 (70-90)
73 (63-81)
12-17 y hospitalized
in psychiatry
Emotional abuse
79 (66-88)
72 (62-80)
Sexual abuse
86 (71-94)
76 (67-83)
Physical neglect
78 (62-89)
61 (53-70)
*Team evaluating children during reporting to Child Protection Services Table 4 Description of selected studies estimating diagnostic accuracy of test identifying abused children, excluding
physical abuse
Source
Inclusion Criteria
Form of child
Sample
Index Test
Reference Standard
Sensitivity
Specificity Table 4 Description of selected studies estimating diagnostic accuracy of test identifying abused children, excluding
physical abuse selected studies estimating diagnostic accuracy of test identifying abused children, excluding timating diagnostic accuracy of test identifying abused children, excluding *Team evaluating children during reporting to Child Protection Services. † CSBI Child Sexual Behavior Inventory. ‡ CPS Child Protection Services. § Score evaluation the probability of sexual penetration. ‖ Spurwink Child Abuse Program for identifying abused children in Oregon. ¶ SASA Signs Associated with Sexual Abuse. **CTQ Childhood Trauma Questionnaire. g
g
p
† CSBI Child Sexual Behavior Inventory. sensitivity was low: among the 43 cases determined to be
abused by the Child Abuse Pediatrician, four were mis-
coded as accidents, two as injuries of undetermined cause,
and four did not receive any injury code [34]. In 1991-
1992 in California, the sensitivity of hospital E-coded data
in identifying child victims of intentional injuries had been
estimated at 75% (95% CI: 64-84) [35]. This classification
underestimates the number of abused children, therefore
does not seem to be a good reference test. Identification of several forms of child maltreatment Cases of child
physical abuse are considered as accidents and cases clas-
sified as physical abuse are not representative of all the
cases of physical abuse, because some cases did not re-
ceive any injury code. Identification of several forms of child maltreatment Identified tools were not adapted to screening because
of low sensitivity and late identification of abused chil-
dren when they have already serious consequences of
maltreatment. The Childhood Trauma Questionnaire is a 70-item screen-
ing inventory that assesses self-reported experiences of
abuse and neglect in childhood and adolescence (Table 4). Page 8 of 11 Bailhache et al. BMC Pediatrics 2013, 13:202
http://www.biomedcentral.com/1471-2431/13/202 Table 4 Description of selected studies estimating diagnostic accuracy of test identifying abused children, excluding
physical abuse
Source
Inclusion Criteria
Form of child
abuse
Sample
size
Index Test
Reference Standard
Sensitivity
Specificity
%
(95% CI)
%
(95% CI)
Cheung et al,
2004 [23]
Children <18 y,
referred to
specialized team*
Sexual abuse
77
Classification of anal
and genital
examination findings
Assessment by medical
team integrating medical
history, children behavior,
laboratory results,
anogenital findings
56 (33-77)
98 (91-100)
Berenson et al,
2002 [22]
Girls 3-8 y referred
to specialized team*
or consulting at the
pediatric clinics
Sexual abuse with
penetration
386
Horizontal diameter
of the hymen > or ≤
6.5 mm in knee-chest
position
Assessment by nurse,
psychologist or social
worker integrating
children interview, CSBI†
and assessment by CPS‡. Discussion Assessment of the accuracy of instruments is difficult,
because there is no gold standard for identifying abused
children. To optimize the reference standard, opinion of
experts or medical, social or judicial teams are usually
used [21,24-28,30-33], but the accuracy of these assess-
ments is not known. Furthermore, the information used
for this assessment was rarely specified so that it was diffi-
cult to verify the independence between the index test and
the reference standard. The incorporation of index test re-
sults in the reference standard would overestimate accur-
acy of the test [21,25,26,28,29,31,33]. Chang et al used the
International Classification of Diseases (ICD), 9th Revision,
and E-codes (External cause), used to categorize intent
and mechanism of an injury, for reference standard [29]. In a recent study in the Yale-New Haven Children’s hos-
pital from 2007 to 2010, the specificity of coding injuries
as physical abuse was 100% (95% CI: 96-100). But the In this systematic review, the quality of selected stud-
ies was low, even when not considering the criterion re-
lated to the reference standard. Available information
was often insufficient to make a judgment for many cri-
teria. Some of the limitations, for instance the utilization Bailhache et al. BMC Pediatrics 2013, 13:202
http://www.biomedcentral.com/1471-2431/13/202 Page 9 of 11 of the index test to establish the final diagnostic, are par-
ticularly worrisome as they reflect an important miscon-
ception of what is good diagnostic research. This overall
poor quality likely limits the validity of the selection of
studies, as many could have been excluded on the basis
of quality alone. Clearly, the quality of reporting of stud-
ies of diagnostic accuracy on child maltreatment needs
to improve. Furthermore in five studies, the retrospect-
ive evaluation based on a review of records could have
introduced bias [27-31]. And in the three case–control
studies, the performance of index test could have been
overestimated because of the increase of differences be-
tween both groups by excluding children for whom mal-
treatment is difficult to diagnose [21,22,31]. cannot help reduce deaths in the most vulnerable groups. Indeed, fatal child maltreatment occurs most frequently
when children are younger [2,37-39]. Over a half of the
600 victims of child maltreatment under five years reported
to the National Violent Death Reporting System of the
United States of America from 2003 to 2006 were under
one-year-old [40]. Discussion The WHO definition of child maltreatment is prob-
lematic as it is defined by consequences of neglectful or
abusive behaviors that, themselves, are not defined [1,3]. Similarly, the Article 19 of the United Nations convention
on the rights of the child, stating “all forms of physical or
mental violence, injury and abuse, neglect or negligent
treatment, maltreatment or exploitation, including sexual
abuse” does not define these behaviors. Moreover, pro-
posed definitions based only on abusive behaviors can vary
widely. For example, physical contact or penetration are
applied before defining reported experiences as sexual abu-
sive by some authors and not others [41-44]. Instruments
designed to diagnose abusive caregivers such as the Child
Abuse Potential Inventory [45], the International Society
for the Prevention of Child Abuse and Neglect (IPSCAN)
Child Abuse Screening Tool-Parent [46] measure these po-
tential abusive behaviors of caregiver. Consequently, what
they measure is not well known and defined. Furthermore
they can identify only child maltreatment which is directly
due to the questioned parent. These problems might ex-
plain why child maltreatment is usually recognized only
when the child has consequences of abusive behaviors. We were interested in tools identifying abused children
as early as possible in the evolution of child maltreatment. Existing instruments reported to diagnose child maltreat-
ment were not designed for screening. Many tools identify
abused children when they have already clinical conse-
quences of child maltreatment, such as head injury, frac-
ture, or behavior problems [21,24-31]. The identification
of abused children already at the clinical stage comes too
late. The performance of tests was also not adapted to
screening. Screening instruments require high sensitivity
for missing very few abused children. In our synthesis,
most sensitivity estimations were low [22-27,30,32,33]. Furthermore, the specificity of tests is also important
because of the negative effects of a misidentification, in
particular the psychological impact and the effect of a po-
tential stigmatization on the child and his parents [36]. As
usual, when the sensitivity of the test was high, the specifi-
city was often low [25]. The sensitivity was greater than
90% and the specificity greater than 80% only for two tests
[21,31]. However, one was a decision tool to identify
physically abused children among those hospitalized in a
Pediatric Intensive Care Unit, so that children had severe
injuries [31]. The other test was based on twelve child
symptoms to identify sexually-abused children [21]. Competing interests criteria was that the aim of the study was clearly to esti-
mate the diagnostic accuracy of a test identifying abused
children. This might have disqualified some studies in
which some parameters of diagnostic accuracy could be
estimated. Finally, we were interested in all forms of child
maltreatment and all types of tools and we have not speci-
fied a particular such as emergency departments. Depend-
ing on the context, some tools could not be applied: for
example a test requiring a specific laboratory result if the
laboratory exam cannot be performed routinely. Besides,
we reviewed the evidence on the accuracy of instruments
for identifying abused children during any stage of child
maltreatment evolution before their death. Thus both diag-
nostic and screening studies could be included in our re-
view. We evaluated among the selected studies if accurate
screening instruments were available. However the fact
that screening test is sensitive and specific is not enough. The side effects, the reliability and the cost of the test
should be also considered. Indeed before considering a
screening program of child maltreatment, several other
criteria need to be respected [18]. A screening program
should also be acceptable to families and professionals. Negative effects for the family are consequences of false
negatives (children identified wrongly as not abused) and
of false positives (children identified wrongly as abused and
parents identified wrongly as abusers). The stigmatization
of families is an important ethical issue. Furthermore, con-
firming the relevance of screening of child maltreatment is
not enough, as the modalities of the program should also
be specified, including the site; the relevant target popula-
tion group if screening is not mass screening, the child age
at the time of screening, and the frequency if screening is
repeated. At last, a screening program could become use-
less because of effective primary prevention program of
child abuse. Several primary prevention programs, such as
the Nurse Family Partnership [47] and the Early Start [48],
have been proposed, but the evidence is currently insuffi-
cient to assess the balance between benefits and harms of
primary care interventions [49]. The authors declare that they have no competing interests. Received: 26 April 2013 Accepted: 20 November 2013
Published: 5 December 2013 Received: 26 April 2013 Accepted: 20 November 2013
Published: 5 December 2013 References Maguire S, Pickerd N, Farewell D, Mann M, Tempest V, Kemp AM: Which
clinical features distinguish inflicted from non-inflicted brain injury? A
systematic review. Arch Dis Child 2009, 94(11):860–867. 8. Piteau SJ, Ward MGK, Barrowman NJ, Plint AC: Clinical and radiographic
characteristics associated with abusive and nonabusive head trauma: a p
p
y
,
( )
6. Bhardwaj G, Chowdhury V, Jacobs MB, Moran KT, Martin FJ, Coroneo MT: A
systematic review of the diagnostic accuracy of ocular signs in pediatric
abusive head trauma. Ophthalmology 2010, 117(5):983–992. e17. 7. Maguire S, Pickerd N, Farewell D, Mann M, Tempest V, Kemp AM: Which
clinical features distinguish inflicted from non-inflicted brain injury? A
systematic review. Arch Dis Child 2009, 94(11):860–867. 8. Piteau SJ, Ward MGK, Barrowman NJ, Plint AC: Clinical and radiographic
characteristics associated with abusive and nonabusive head trauma: a
systematic review. Pediatrics 2012, 130(2):315–323. 9. Kemp AM, Jaspan T, Griffiths J, Stoodley N, Mann MK, Tempest V, et al:
Neuroimaging: what neuroradiological features distinguish abusive from
non-abusive head trauma? A systematic review. Arch Dis Child 2011,
96(12):1103–1112. 9. Kemp AM, Jaspan T, Griffiths J, Stoodley N, Mann MK, Tempest V, et al:
Neuroimaging: what neuroradiological features distinguish abusive from
non-abusive head trauma? A systematic review. Arch Dis Child 2011,
96(12):1103–1112. 10. Berkoff MC, Zolotor AJ, Makoroff KL, Thackeray JD, Shapiro RA, Runyan DK: Has
this prepubertal girl been sexually abused? JAMA 2008, 300(23):2779–2792. 10. Berkoff MC, Zolotor AJ, Makoroff KL, Thackeray JD, Shapiro RA, Runyan DK: Has
this prepubertal girl been sexually abused? JAMA 2008, 300(23):2779–2792. 11. Louwers ECFM, Korfage IJ, Affourtit MJ, Scheewe DJH, Van de Merwe MH,
Vooijs-Moulaert A-FSR, et al: Effects of systematic screening and detection
of child abuse in emergency departments. Pediatrics 2012, 130(3):457–464. 11. Louwers ECFM, Korfage IJ, Affourtit MJ, Scheewe DJH, Van de Merwe MH,
Vooijs-Moulaert A-FSR, et al: Effects of systematic screening and detection
of child abuse in emergency departments. Pediatrics 2012, 130(3):457–464. Authors’ contributions MB conceptualized and designed the study, participated in the acquisition,
analysis and interpretation of data, drafted the initial manuscript. VL
participated in the analysis and interpretation of data, critically reviewed the
manuscript. PP participated in the interpretation of data, critically reviewed
the manuscript. LRS conceptualized and designed the study, participated in
analysis and interpretation of data, drafted the initial manuscript. All authors
read and approved the final manuscript. Discussion These
symptoms could be severe psychological consequences
as depression: sudden emotional and behavior changes,
changes to poor school performance, frequent stomach-
aches, difficulty getting to sleep or sleeping more than
usual. Due to the lack of knowledge of the evolution of child
maltreatment, studying the accuracy of diagnostic instru-
ments identifying abused children early remains challen-
ging. Research is required to define what subclinical and
clinical abusive behaviors are and when the child maltreat-
ment begins. A multidisciplinary approach might be ne-
cessary to correctly identify child maltreatment because of
its multiple targets, the child and the caregiver. Input from
adult psychiatry is necessary to be able to assess the
potential abusive behaviors of caregivers. One might rea-
sonably hypothesize that tools based on simultaneous as-
sessment of potential abusive behaviors and health and
development of the child could allow earlier identification
of abused child or abusive caregiver than tools based only
on separate assessments of the child or caregiver. How-
ever, if a combined approach is likely to be more sensitive,
it might also be less specific. Furthermore, because of the
several types of child maltreatment and the varied conse-
quences to children, several tests might be necessary to
screen all types of child maltreatment. The final value of
features used for screening will also depend on the preva-
lence of these features. Child maltreatment is the “disease” of both the child and
his caregiver. Obviously, an abusive caregiver is defined by
his abusive behavior and child maltreatment begins by abu-
sive behavior of caregiver. This abusive behavior is respon-
sible for poor health and development of the child. Thus,
identification of child maltreatment could consider the
identification of both the abused child and his abusive
caregiver. Two self-report questionnaires were directed to
children who had to indicate if they had experienced given
behaviors of parents or caregivers [32,33]. As only children
old enough for reading could answer, these questionnaires We reviewed studies only in French and English and
only published studies in databases, and might have ex-
cluded interesting research. Also, one of our inclusion Bailhache et al. BMC Pediatrics 2013, 13:202
http://www.biomedcentral.com/1471-2431/13/202 Page 10 of 11 Page 10 of 11 criteria was that the aim of the study was clearly to esti-
mate the diagnostic accuracy of a test identifying abused
children. Abbreviations
E-code: External causes-code; ICD: International classification of diseases;
WHO: World Health Organization. Conclusions
h There is very scarce and low-quality evidence on the ac-
curacy of instruments for identifying abused children. Child maltreatment is mostly identified when children
have already serious consequences and the sensitivities
and specificities of tools are inadequate. Before consider-
ing a screening program of child maltreatment, better
knowledge on the beginning of child maltreatment and
development of valid screening instruments at subclin-
ical stages remain necessary. 12. Louwers EC, Korfage IJ, Affourtit MJ, De Koning HJ, Moll HA: Facilitators and
barriers to screening for child abuse in the emergency department. BMC Pediatr 2012, 12:167. 12. Louwers EC, Korfage IJ, Affourtit MJ, De Koning HJ, Moll HA: Facilitators and
barriers to screening for child abuse in the emergency department. BMC Pediatr 2012, 12:167. 13. Woodman J, Pitt M, Wentz R, Taylor B, Hodes D, Gilbert RE: Performance of
screening tests for child physical abuse in accident and emergency
departments. Health Technol Assess 2008, 12(33):iii, xi-xiii 1–iii, xi-xiii 95. 13. Woodman J, Pitt M, Wentz R, Taylor B, Hodes D, Gilbert RE: Performance of
screening tests for child physical abuse in accident and emergency
departments. Health Technol Assess 2008, 12(33):iii, xi-xiii 1–iii, xi-xiii 95. 14. Woodman J, Lecky F, Hodes D, Pitt M, Taylor B, Gilbert R: Screening injured
children for physical abuse or neglect in emergency departments: a
systematic review. Child Care Health Dev 2010, 36(2):153–164. 14. Woodman J, Lecky F, Hodes D, Pitt M, Taylor B, Gilbert R: Screening injured
children for physical abuse or neglect in emergency departments: a
systematic review. Child Care Health Dev 2010, 36(2):153–164. y
15. Nygren P, Nelson HD, Klein J: Screening children for family violence: a
review of the evidence for the US preventive services task force. Ann Fam Med 2004, 2(2):161–169. 15. Nygren P, Nelson HD, Klein J: Screening children for family violence: a
review of the evidence for the US preventive services task force. Ann Fam Med 2004, 2(2):161–169. 16. Whiting P, Rutjes AWS, Reitsma JB, Bossuyt PMM, Kleijnen J: The
development of QUADAS: a tool for the quality assessment of studies of
diagnostic accuracy included in systematic reviews. BMC Med Res
Methodol 2003, 3:25. 16. Whiting P, Rutjes AWS, Reitsma JB, Bossuyt PMM, Kleijnen J: The
development of QUADAS: a tool for the quality assessment of studies of
diagnostic accuracy included in systematic reviews. BMC Med Res
Methodol 2003, 3:25. Discussion This might have disqualified some studies in
which some parameters of diagnostic accuracy could be
estimated. Finally, we were interested in all forms of child
maltreatment and all types of tools and we have not speci-
fied a particular such as emergency departments. Depend-
ing on the context, some tools could not be applied: for
example a test requiring a specific laboratory result if the
laboratory exam cannot be performed routinely. Besides,
we reviewed the evidence on the accuracy of instruments
for identifying abused children during any stage of child
maltreatment evolution before their death. Thus both diag-
nostic and screening studies could be included in our re-
view. We evaluated among the selected studies if accurate
screening instruments were available. However the fact
that screening test is sensitive and specific is not enough. The side effects, the reliability and the cost of the test
should be also considered. Indeed before considering a
screening program of child maltreatment, several other
criteria need to be respected [18]. A screening program
should also be acceptable to families and professionals. Negative effects for the family are consequences of false
negatives (children identified wrongly as not abused) and
of false positives (children identified wrongly as abused and
parents identified wrongly as abusers). The stigmatization
of families is an important ethical issue. Furthermore, con-
firming the relevance of screening of child maltreatment is
not enough, as the modalities of the program should also
be specified, including the site; the relevant target popula-
tion group if screening is not mass screening, the child age
at the time of screening, and the frequency if screening is
repeated. At last, a screening program could become use-
less because of effective primary prevention program of
child abuse. Several primary prevention programs, such as
the Nurse Family Partnership [47] and the Early Start [48],
have been proposed, but the evidence is currently insuffi-
cient to assess the balance between benefits and harms of
primary care interventions [49]. Competing interests Author details
1 1CHU de Bordeaux, Pole de pediatrie, F-33000 Bordeaux, France. 2Centre
INSERM U897-Epidemiologie-Biostatistique, University Bordeaux, ISPED,
F-33000 Bordeaux, France. 3Centre INSERM U897-Epidemiologie-Biostatistique,
INSERM, ISPED, F-33000 Bordeaux, France. 4CHU de Bordeaux, Pole de sante
publique, Service d’information medicale, F-33000 Bordeaux, France. References 1. World Health Organization: Report of the consultation on child abuse
prevention. Geneva: World Health Organization, 1999; 1999. Document
WHO/HSC/PVI/99.1. 1. World Health Organization: Report of the consultation on child abuse
prevention. Geneva: World Health Organization, 1999; 1999. Document
WHO/HSC/PVI/99.1. 2. Gilbert R, Widom CS, Browne K, Fergusson D, Webb E, Janson S: Burden
and consequences of child maltreatment in high-income countries. Lancet 2009, 373(9657):68–81. 2. Gilbert R, Widom CS, Browne K, Fergusson D, Webb E, Janson S: Burden
and consequences of child maltreatment in high-income countries. Lancet 2009, 373(9657):68–81. 3. Krug EG, Dahlberg LL, Mercy JA, Zwi AB, Lozano R: World report on violence
and health. Geneva: World Health Organization; 2002. 4. Tursz A, Crost M, Gerbouin-Rérolle P, Cook JM: Underascertainment of
child abuse fatalities in France: retrospective analysis of judicial data to
assess underreporting of infant homicides in mortality statistics. Child Abuse Negl 2010, 34(7):534–544. 5. Putnam-Hornstein E: Report of maltreatment as a risk factor for injury death:
a prospective birth cohort study. Child Maltreat 2011, 16(3):163–174. 6. Bhardwaj G, Chowdhury V, Jacobs MB, Moran KT, Martin FJ, Coroneo MT: A
systematic review of the diagnostic accuracy of ocular signs in pediatric
abusive head trauma. Ophthalmology 2010, 117(5):983–992. e17. 7. Maguire S, Pickerd N, Farewell D, Mann M, Tempest V, Kemp AM: Which
clinical features distinguish inflicted from non-inflicted brain injury? A
systematic review. Arch Dis Child 2009, 94(11):860–867. 8. Piteau SJ, Ward MGK, Barrowman NJ, Plint AC: Clinical and radiographic
h
d
h b
d
b
h
d 5. Putnam-Hornstein E: Report of maltreatment as a risk factor for injury death:
a prospective birth cohort study. Child Maltreat 2011, 16(3):163–174. 6. Bhardwaj G, Chowdhury V, Jacobs MB, Moran KT, Martin FJ, Coroneo MT: A
systematic review of the diagnostic accuracy of ocular signs in pediatric
abusive head trauma. Ophthalmology 2010, 117(5):983–992. e17. 5. Putnam-Hornstein E: Report of maltreatment as a risk factor for injury death:
a prospective birth cohort study. Child Maltreat 2011, 16(3):163–174. 5. Putnam-Hornstein E: Report of maltreatment as a risk factor for injury death:
a prospective birth cohort study. Child Maltreat 2011, 16(3):163–174. 6. Bhardwaj G, Chowdhury V, Jacobs MB, Moran KT, Martin FJ, Coroneo MT: A
systematic review of the diagnostic accuracy of ocular signs in pediatric
abusive head trauma. Ophthalmology 2010, 117(5):983–992. e17. 7. Abbreviations Page 11 of 11 Page 11 of 11 Bailhache et al. BMC Pediatrics 2013, 13:202
http://www.biomedcentral.com/1471-2431/13/202 Bailhache et al. BMC Pediatrics 2013, 13:202
http://www.biomedcentral.com/1471-2431/13/202 41. Holmes WC, Slap GB: Sexual abuse of boys: definition, prevalence,
correlates, sequelae, and management. JAMA 1998, 280(21):1855–1862 17. Whiting P, Rutjes AWS, Westwood ME, Mallet S, Deeks JJ, Reitsma JB, et al:
QUADAS-2: A revised tool for the quality assessment of diagnostic
accuracy studies. Ann Intern Med 2011, 155(8):529–536. 42. Bassani DG, Palazzo LS, Béria JU, Gigante LP, Figueiredo ACL, Aerts DRGC, et
al: Child sexual abuse in southern Brazil and associated factors: a
population-based study. BMC Public Health 2009, 9:133. 18. Wilson JMG, Jungner G: Principles and practice of screening for disease. Geneva: World Health Organization; 1968. 43. Priebe G, Svedin CG: Prevalence, characteristics, and associations of
sexual abuse with sociodemographics and consensual sex in a
population-based sample of Swedish adolescents. J Child Sex Abus 2009,
18(1):19–39. 19. Sege R, Flaherty E, Jones R, Price LL, Harris D, Slora E, et al: To report or not
to report: examination of the initial primary care management of
suspicious childhood injuries. Acad Pediatr 2011, 11(6):460–466. 20. Hammond J, Perez-Stable A, Ward CG: Predictive value of historical and
physical characteristics for the diagnosis of child abuse. South Med J
1991, 84(2):166–168. 44. Halpérin DS, Bouvier P, Jaffé PD, Mounoud RL, Pawlak CH, Laederach J, et al:
Prevalence of child sexual abuse among adolescents in Geneva: results
of a cross sectional survey. BMJ 1996, 312(7042):1326–1329. 21. Wells R, McCann J, Adams J, Voris J, Dahl B: A validational study of the
structured interview of symptoms associated with sexual abuse (SASA)
using three samples of sexually abused, allegedly abused, and
nonabused boys. Child Abuse Negl 1997, 21(12):1159–1167. 45. Robertson KR, Milner JS: Convergent and discriminant validity of the child
abuse potential inventory. J Pers Assess 1985, 49(1):86–88. 46. Runyan DK, Dunne MP, Zolotor AJ, Madrid B, Jain D, Gerbaka B, et al: The
development and piloting of the ISPCAN child abuse screening tool-
parent version (ICAST-P). Child Abuse Negl 2009, 33(11):826–832. 22. Berenson AB, Chacko MR, Wiemann CM, Mishaw CO, Friedrich WN, Grady JJ:
Use of hymenal measurements in the diagnosis of previous penetration. Pediatrics 2002, 109(2):228–235. 47. Olds DL, Eckenrode J, Henderson CR, Kitzman H, Powers J, Cole R, et al:
Long-term effects of home visitation on maternal life course and child
abuse and neglect. Abbreviations Fifteen-year follow-up of a randomized trial. JAMA 1997, 278(8):637–643. 23. Cheung PCH, Ko CH, Lee HYM, Ho LMC, To WWK, Ip PLS: Correlation of
colposcopic anogenital findings and overall assessment of child sexual
abuse: prospective study. Hong Kong Med J 2004, 10(6):378–383. 48. Fergusson DM, Boden JM, Horwood LJ: Nine-year follow-up of a home-
visitation program: a randomized trial. Pediatrics 2013, 131(2):297–303. 24. Drach KM, Wientzen J, Ricci LR: The diagnostic utility of sexual behavior
problems in diagnosing sexual abuse in a forensic child abuse
evaluation clinic. Child Abuse Negl 2001, 25(4):489–503. 49. Moyer VA, on behalf of the U.S. Preventive Services Task Force: Primary
care interventions to prevent child maltreatment: U.S. preventive
services task force recommendation statement. Ann Intern Med 2013,
159(4):289–295. 25. Vinchon M, De Foort-Dhellemmes S, Desurmont M, Delestret I: Confessed
abuse versus witnessed accidents in infants: comparison of clinical,
radiological, and ophthalmological data in corroborated cases. Childs Nerv Syst 2010, 26(5):637–645. doi:10.1186/1471-2431-13-202
Cite this article as: Bailhache et al.: Is early detection of abused children
possible?: a systematic review of the diagnostic accuracy of the
identification of abused children. BMC Pediatrics 2013 13:202. 26. Vinchon M, Defoort-Dhellemmes S, Desurmont M, Dhellemmes P: Acciden-
tal and nonaccidental head injuries in infants: a prospective study. J Neu
rosurg 2005, 102(4 Suppl):380–384. 27. Hettler J, Greenes DS: Can the initial history predict whether a child with
a head injury has been abused? Pediatrics 2003, 111(3):602–607. 28. Wells RG, Vetter C, Laud P: Intracranial hemorrhage in children younger
than 3 years: prediction of intent. Arch Pediatr Adolesc Med 2002,
156(3):252–257. 29. Chang DC, Knight VM, Ziegfeld S, Haider A, Paidas C: The multi-
institutional validation of the new screening index for physical child
abuse. J Pediatr Surg 2005, 40(1):114–119. 30. Valvano TJ, Binns HJ, Flaherty EG, Leonhardt DE: Does bruising help
determine which fractures are caused by abuse? Child Maltreat 2009,
14(4):376–381. 31. Pierce MC, Kaczor K, Aldridge S, O’Flynn J, Lorenz DJ: Bruising
characteristics discriminating physical child abuse from accidental
trauma. Pediatrics 2010, 125(1):67–74. 32. Fernandopulle S, Fernando D: Development and initial validation of a
scale to measure emotional abuse among school children aged 13-15
years in Sri Lanka. Child Abuse Negl 2003, 27(10):1087–1099. 33. Bernstein DP, Ahluvalia T, Pogge D, Handelsman L: Validity of the
childhood trauma questionnaire in an adolescent psychiatric population. J Am Acad Child Adolesc Psychiatry 1997, 36(3):340–348. 34. Abbreviations Hooft A, Ronda J, Schaeffer P, Asnes AG, Leventhal JM: Identification of
physical abuse cases in hospitalized children: accuracy of international
classification of diseases codes. J Pediatr 2013, 162(1):80–85. 35. Winn DG, Agran PF, Anderson CL: Sensitivity of hospitals’ E-coded data in
identifying causes of children’s violence-related injuries. Public Health Rep
1995, 110(3):277–281. Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit Submit your next manuscript to BioMed Central
and take full advantage of: 36. Teeuw AH, Derkx BHF, Koster WA, Van Rijn RR: Educational paper:
detection of child abuse and neglect at the emergency room. Eur J
Pediatr 2012, 171(6):877–885. 37. Sidebotham P, Bailey S, Belderson P, Brandon M: Fatal child maltreatment
in England, 2005-2009. Child Abuse Negl 2011, 35(4):299–306. 38. Herman-Giddens ME, Brown G, Verbiest S, Carlson PJ, Hooten EG, Howell E,
et al: Underascertainment of child abuse mortality in the United States. JAMA 1999, 282(5):463–467. 39. Bennett MD Jr, Hall J, Frazier L Jr, Patel N, Barker L, Shaw K: Homicide of
children aged 0-4 years, 2003-04: results from the national violent death
reporting system. Inj Prev 2006, 12(Suppl 2):ii39–ii43. 40. Klevens J, Leeb RT: Child maltreatment fatalities in children under 5:
findings from the national violence death reporting system. Child Abuse
Negl 2010, 34(4):262–266. 40. Klevens J, Leeb RT: Child maltreatment fatalities in children under 5:
findings from the national violence death reporting system. Child Abuse
Negl 2010, 34(4):262–266.
|
https://openalex.org/W2160998780
|
https://zenodo.org/record/2292666/files/article.pdf
|
English
| null |
A Direct-Reading Decremeter and Wave Meter
|
Proceedings of the IRE
| 1,915
|
public-domain
| 6,482
|
'Delivered before The Institute of Radio Engineers, Washington Sec-
tion, February 5, 1914.
This article is based on a publication in the Bulletin
of the Bureau of Standards, Vol. 11 (Scientific Paper Number 235). INTRODUCTION The laws of the United States governing radio communica-
tion specify, among other things, that at all stations the log-
arithmic decrement per complete oscillation in the wave trains
emitted by the transmitter shall not exceed two-tenths, except
when sending distress signals or messages relating thereto. The importance of the regulation lies in the fact that when
persistent oscillations of single frequency are emitted from a
radio transmitting station much more selective receiving ap-
paratus may be employed with advantage at receiving stations,
permitting sharp tuning with consequent minimizing of inter-
ference caused by stations other than those with which com-
munication is desired. When full advantage is taken of the rapid scientific and
technical progress which has been made in the methods of trans-
mission of electromagnetic waves it is not at all difficult to
comply with this requirement. In fact, it is practicable as well
as desirable to keep well within this limiting value of two-
tenths. The purpose of this paper is to describe a new direct-reading
instrument for measuring the logarithmic decrement and wave
length, especially designed about two years ago for the radio
inspection service of the Bureau of Navigation, Department of
Commerce, and since adopted by the War and Navy Depart-
ments. A DIRECT-READING DECREMETER AND WAVE
METER' BY
FREDERICK A. KOLSTER 2 In practice it is found permissible to make the change in capacity from
Cr to Cl such that
I2 becomes Y2Ir2, thus making the expression under the
radical sign equal to unity. GENERAL THEORY It is supposed that in the primary of two coupled circuits
there exist damped electrical oscillations of decrement di, the
natural decrement of the primary. The natural decrement of
the secondary circuit is 82. When the secondary circuit is tuned 29 so that maximum current is induced in it by the primary, it is
assumed that its capacity is Cr and the current in it Ir. If the
secondary capacity is altered to a new value, C, the current will
drop to a value
I. Bjerknes
has shown
that
the sum of
primary and secondary decrements is given by the equation: Oi + (2 =
r
C
J
\'r2 The conditions under which this formula may be applied
with sufficient accuracy are, 1. That °' + °2 be small as compared with 2,.. 2. That Cr C C be small as compared with unity. C 2. That Cr C C be small as compared with unity. C 3. That the degree of coupling between the circuits be small. i
i
i
l
i h i Let us assume that it is desired to determine the logarithmic
decrement of the oscillations in the aerial circuit of a radio trans-
mitter as shown in Figure 1. A circuit containing inductance H
A
M
FIGURE I-Decremeter Circuit Coupled Loosely to the Antenna
Circuit of a Transmitter H
A
M
FIGURE I-Decremeter Circuit Coupled Loosely to the Antenna
Circuit of a Transmitter A M FIGURE I-Decremeter Circuit Coupled Loosely to the Antenna
Circuit of a Transmitter L, a calibrated variable condenser C, and a sensitive low re-
sistance hot-wire instrument H, is very loosely coupled to the
aerial or antenna circuit A. Readings of the hot-wire instrument L, a calibrated variable condenser C, and a sensitive low re-
sistance hot-wire instrument H, is very loosely coupled to the
aerial or antenna circuit A. Readings of the hot-wire instrument 30 H, which are proportional to the square of the current in the
circuit, are taken for several values of capacity C on both sides
of the resonant value
Cr. Plotting
these readings
against
capacity, a resonance curve such as Figure 2 is obtained, and
from one of the following formulas,2 the sum of the logarithmic
decrements 8i and 82 may be obtained. GENERAL THEORY ~
±02 =
Cr
iCl\/r
J12
81+ ° 2 =7
C-4pI
31+82
;:
C,
41 2 _ 12
()l1+d2=2-Cr<
12
12r2-12
oi+02 =7WC2-C<l
J2
C2 + Cl
Ir2-I2 If the decrement d2 of the measuring circuit has been previ-
ously determined, the decrement 8j of the aerial circuit under
test is at once obtained. Altho' th
measurement of the logarithmic decrement as out-
-lined above appears to be comparatively simple, nevertheless,
to obtain reasonably consistent and accurate results, the observa-
tions must be taken with considerable care, the resonance curve
must be plotted on a large scale, and calculations must be made
from several points on the curve. In fact, it is only with labora-
tory
conveniences
that
satisfactory
measurements
can
be
obtained. The instrument, which it is the purpose of this paper to
describe, was designed for the purpose of facilitating the work
involved in making measurements of decrement and yet permit-
ting as great accuracy as c-an be expected in the ordinary labora-
tory method. These requirements are particularly desirable for
the purposes of the inspection service of the Bureau of Navigation
in the enforcement
of the radio communication laws. The
inspection of a radio station on board ship, for example, has to
be done quickly and in many cases under very unfavorable
circunistances. In Figure 2, a typical resonance curve is shown, and the
task of obtaining the logarithmic decrement by the ordinary
method is indicated. Identical results are obtained in a very
much shorter time by means of the direct reading instrument. 31 A5 A5
Resonanico
Curve
for experimental determination
AO
of logarithmic decrement. ,61- 62 = A
C,
= a 2GC
C_A
I
.30
25
20
0
0
x
.15
r10
.05
Using
Decremeoter
6,+6,= 0.121 on scale. By *xpermment
using usual laboratory method
with
ordinary wavenmeter. .122
.124
'4
.120
.119
0.121
DEGREES
ON CONDENSER
93
94
95
96
97
98
99
100
101
102
103
104
105
100
107
108
FIGIURE 2-Resonance Curve for Experimental Determination of
Logarithmic Decrement .30
25
20
0
0
x r10 .05 THEORY OF THE INSTRUMENT The shape of the moving plates or surfaces of the ordinary
variable condenser in common use is such that for equal angular
displacements of these surfaces from the position of minimum
capacity
to
that
of maximum
capacity,
an approximately
straight line variation of capacity is obtained. It is evident,
therefore, that for any given displacement of the plates, the
i
dc
i
l
h percentage change of capacity, dc
will not be equal to that percentage change of capacity, dc
will not be equal to that 32 obtained for an equal displacement of the plates at any other
point. In order to make it possible to attach to a variable condenser
a single predetermined scale giving values of logarithmic decre-
ment corresponding to various percentage changes of capacity
thruout the range of capacity of the condenser as defined by the
Bjerknes formula, Cr
C
for 11 =
2I the capacity variation with equal displacements of the moving
plates must be such that for any given displacement of the plates,
taken at any point from the original position of the plates to
their final position, Cr-C
IlC = a constant. C
C The problem, therefore,
of constructing a direct reading
instrument
for
decrement measurements
is
largely
that
of
determining the proper shape for the plates or surfaces of the
condenser to produce a variable capacity such that for any given
displacement the value of C will be constant thruout the range
f
i
f
h
i
f of motion of the moving surfaces. of motion of the moving surfaces. THEORY OF THE INSTRUMENT In other words, for a displacement of iX, Figure 3, C2 - Cl
C3- C2
C4- C3
Cn- Cn-1
Cl
C2
C3
Cn-I
c2-c1 _C3-c2
if,
= but if, C22 = C1 C3, or C2 =-/Cl C3
C3 = VC2 C4
Cn = \/Cn -1 Cn+1 then similarly, or, or, It is seen, therefore, that the capacity of the variable con-
denser must vary in accordance with the law of geometric
progression, and it is now easy to formulate the equation giving
the connection between the value of capacity and the position 33 Cit-I
/\X
AX
A
X,E
3
FIGURE 3-Capacity Varying in Accordance with the Law of
Geometric Progression FIGURE 3-Capacity Varying in Accordance with the Law of
Geometric Progression of the moving plates, for since the curve of capacity must obey
the law of geometric progression, we have the following, Figure 4. of the moving plates, for since the curve of capacity must obey
the law of geometric progression, we have the following, Figure 4. at x = o let C0 = aK0 = a
then at x = 1
C, = aK'
x = 2
C2=aK2
x = 3
C3 = aK3
x = nn
Cn = aK0,
, in general,
C = aKx
(1) C = aKx
(1) or, in general,
C = aKx
(1) (1) or, in general, This is equivalent to equation (1) derived above.' This law might have been deduced more directly as follows: The funda-
mental requirement of the condenser may be written: dC -ndx
(2)
logC = nx+h
C =
nx+h = aEnx (2) logC = nx+h
C =
nx+h = aEnx and For the case of a rotary condenser where a is the displacement angle in
degrees C = asmo
(3) (3) 34 34 a
C~~~~~~~~~~~
0
1
2
3
4
5
n-2
n-i
ni
FIGURE 4-Geometric Progression FIGURE 4-Geometric Progression DESIGN OF CONDENSER
i
i
i Since the capacity of a condenser is directly proportional to
the active area of the movable surface, neglecting edge effects
we may write, A = b EmO A being the area of the active surface of the moving plate, and 6
the angle of displacement. If we now consider Figure 5, the actual shape of the moving
plate can be determined. FIGURE 5-Shape of Rotary Plate of Condenser. 35 FIGURE 5-Shape of Rotary Plate of Condenser. 35 By analogy with equation (2) By analogy with equation (2) By analogy with equation (2) dA = mde/
A
or,
d A =b mmo d0
but,
dA =(P2 -r2)d 9 dA = mde/
A
d A =b mmo d0
dA =(P2 -r2)d 9 or, but, P being the distance from the center 0 to the enveloping curve
of the plate, or the radius vector, and r being the radius of the
small circular space occupied by the separating washers between
the plates. p2 - r2=bm mo
2
2
p=
2 b m mo+ r2 p2 - r2=bm mo
2
2 Then p=
2 b m mo+ r2 and p=
2 b m mo+ r2 b and m are constants which determine the minimum and
maximum values of capacity of the coiAdenser to be used, and
having c-hosen these constants to suit our particular requirements,
we can immediately determine the size and shape of our plates
and construct a condenser, the capacity of which will vary in
accordance with the law of geometric progression. In Figure 6 are shown the stationary and rotary plates of the
condenser. The stationary plates are made semicircular for
convenience. Stationary Surface
movable
5urface
/,
FIGURE 6-Stationary Surface, Movable Surface
36 FIGURE 6-Stationary Surface, Movable Surface
36 Equations (1) and (3) are identical and we may write Equations (1) and (3) are identical and we may write Kx = Smo
x-logK = m()
x log K
m
- Kx = Smo
or
x-logK = m()
x log K
therefore
m
- or or If we assume that when If we assume that when
x= 1, 0
1800
then
x
180
log K
and
180
therefore,
C = a.~180
for
0=o, CO=a for
0 1800
C180 =aK 0 1800
C180 =aK 0 1800
C180 =aK hence, the ratio between inaximum and minimum capacity will be hence, the ratio between inaximum and minimum capacity will be C18=K
= K
Co In order to obtain the ratio K which has been chosen to suit
our particular requirements, a fixed condenser is connected in
parallel with the rotary condenser. The capacity of this fixed
condenser is determined experimentally. i A rotary condenser constructed in accordance with the theory
just given, with a fixed capacity connected in parallel with it, so
chosen as to give the desired ratio between the maximum and
minimum capacity of the combined condensers, will give a cali-
bration curve in exact agreement with theoretical values. and again, the displacement angle J 0 = O2-Or may be read-
ily calculated for various values of 3 = dl + a2. DETERMINATION OF DECREMENT SCALE It has been shown that since the capacity of the condenser
to be used in the instrument varies in accordance with the law of geometric progression, the term
Cr
in the formula,
C -
r -C112
_
r-
l +82
4
-I
=C C,C
J2
12
J2=
r2 when will remain constant for any given angular displacement of the
rotary plates thruout the range of motion from 0° to 1800. 37 37 In order, therefore, to predetermine a scale which can be
attached to the rotary condenser and which will indicate directly
the value of
di + d2 for various displacements of the rotary
plates, the following calculations are made. al+,2
Cr -Cl
Case I 81 +02=
C, where Cr is the value of capacity at resonance and C1 is a smaller
capacity of such a value that the current squared is reduced to
one-half of its value at complete resonance. p
Since Cr is proportional to
smor and C1 is proportional to
sm6l, we may write 1rr_ rmi
)\
3=
+ (62 for convenience,
+m(6r-61) =
1
+
1 J
r
Or- o=
log
m Let then, and The displacement angle JO = Or -01 may therefore be imme-
diately calculated for various values of o = d1 + A2. m is as before
a constant dependent upon the ratio of maximum to minimum
capacity of the condenser and is equal to log-K
where K repre-
180
sents this ratio. capacity of the condenser and is equal to log-K
where K repre-
180
sents this ratio. C2 - Cr
Case II: J+ 2=z
J
C
r
C2 where Cr is again the value of capacity at complete resonance and
C2 is a larger capacity of such a value that the current squared is
reduced to one-half of the value at complete resonance. Since C2 is proportional to smO2 and Cr is proportional to
EMor, m02 _ mOrmor
8~ ~~~~B
= '\
+
:
se
mO2/
- =4
12-Orlog
m
4r-o then then and and again, the displacement angle J 0 = O2-Or may be read-
ily calculated for various values of 3 = dl + a2. and again, the displacement angle J 0 = O2-Or may be read-
ily calculated for various values of 3 = dl + a2. 38 38 For the particular case in question, the following angles and
corresponding decrements as calculated are tabulated: Case I. DETERMINATION OF DECREMENT SCALE Reducing
Case II. Increas-
81 + a2
capacityfromres-
ing capacity from
onance: or -61
resonance: 02-r
0.05
10.292
10.313
0.10
2 .564
2 .650
0.15
3 .821
4 .008
0.20
5 .055
5 .389
0.25
6 .272
6 .793
0.30
7 .472
8 .222 It should be emphasized that the formula (1 + 42
Cr
C
-Is does not strictly apply in cases where
N1 + (2 is great in com-
parison with 27r and Cr C
is great in comparison with unity. does not strictly apply in cases where
N1 + (2 is great in com-
parison with 27r and Cr C
is great in comparison with unity. For 4, + 42 = 0.2 the formula may still be applied for practi-
cal purposes with reasonable accuracy. In the foregoing tabula-
tion calculations have been made for
1 + 42 as high as 0.3,
but the method and formula should preferably not be applied
at values of 4, + 32 greater than 0.2. It will be noted that the angles tabulated above are very small,
and if the decrement scale were attached directly to the shaft of
the condenser it would be extremely short and difficult to read. y
In order to open out the scale, it is geared to the condenser
shaft at a 6-to-1 ratio, as shown in Figure 7. Furthermore, the decrement readings are taken in such a way
as to simultaneously include both measurements as defined by
cases I and II. The displacement angle is then the sum of the
angles tabulated under these two cases. DETERMINATION OF DECREMENT SCALE JO = 02 - Or + Or -01 = 02 - 01 JO = 02 - Or + Or -01 = 02 - 01
The value of this angle JO = 02 - 01 could have been calculated
directly from the formula JO = 02 - Or + Or -01 = 02 - 01
The value of this angle JO = 02 - 01 could have been calculated
directly from the formula 8~~C
+
aC2Cl
02 ,-
02
010
for since C2 is proportional to Sm.2 and C, is proportional to
el,
then carrying out the method used in cases I and II we get
directly, 1
w+Ja
02 - 01 =-log 1
w+Ja
02 - 01 =-log
39 0f
FIGURE 7-Variable Condenser, Showing Gears and Scales
Mechanically Attached FIGURE 7-Variable Condenser, Showing Gears and Scales
Mechanically Attached The final graduations for the decrement scale are obtained by
multiplying 02-01 by the gear ratio of 6, as in the following
table: 11+12
0. 0.05
0.10
0.15
0.20
0.25
0.30
02- 01
0
20.605
5 .214
7 .830
10 .444
13 .065
15 .694
(92- 01) X6
0
150.63
31 .28
46 .98
62 .70
78 .39
94 .20 The decrement scale is marked to the left and to the right of
zero in accordance with this table. 40 THE MEASUREMENT OF LOGARITHMIC DECREMENT THE MEASUREMENT OF LOGARITHMIC DECREMENT Considering now Figure 8, the operation for measuring the
logarithmic decrement is as follows: f\
A~~~~A
FIGURE 8-Diagram Showing Relation Between Decrement Scale
and Resonance Curve FIGURE 8-Diagram Showing Relation Between Decrement Scale
and Resonance Curve The rotary condenser is first set at the position of complete
resonance as indicated by the maximum deflection of the sensitive
hot-wire instrument, the scale readings of which are proportional
to the current squared. This maximum deflection is now reduced
to one-half its value by decreasing or increasing the capacity of
the rotary condenser. The decrement scale, which may be
rotated independently, is now set at zero, then clamped so that
when the condenser is again varied it will rotate with it. 41 Starting at the zero setting with the hot-wire instrument
reading one-half the maximum deflection, the condenser is varied
continuously in one direction until the needle of the hot-wire
instrument makes a complete excursion from one-half deflection
to maximum deflection and back again to one-half deflection. The scale reading now opposite the index mark 0 is the value of
a, + d2, d1 being the decrement of the circuit under test and 82
the kriown decrement of the instrument. It will be noted by referring to Figure 7, that it is desirable to
make the zero setting of the decrement scale at the point of half
deflection and also to take the final reading at the point of half
deflection, because at these points the resonance curve is steep,
and consequently the settings are sharply defined and easily
made. In this connection it will be noted that the formula -PIC2-Cl
dl d2dl~C2 + Cl does not involve the resonant value of capacity Cr, but only those
at the points of half deflection where the slope of the resonance
curve is steep. This formula is therefore the most desirable one
to use, and the decremeter is consequently operated in accordance
with it. In Figure 9 a schematic diagram of the circuit is shown. I is a single-turn coil which may be connected in the circuit
under test, as, for example, the aerial circuit of a radio trans-
mitter. The inductance of this single turn is, in the majority of
practical cases, small as compared with the total inductance of
the circuit under test, and therefore will not affect the tuning
adjustment. THE MEASUREMENT OF LOGARITHMIC DECREMENT The coil L is the inductance of the decremeter circuit and is so
arranged that the mutual inductance between it and coil I can be
easily varied. It is very essential that the degree of coupling be-
tween the circuit under test and the decremeter circuit be small. CV is the variable condenser to which the decrement scale is
attached thru gears. In parallel with Cv is a small condenser
Cf which remains fixed in value after proper adjustment. p
p
j
H represents the hot-wire instrument or indicating device, the
scale- of which is so marked that the readings are proportional to
the square of the current passing thru it. A crystal dector D is provided and the wave length of distant
stations may be measured by using telephone receivers T. y
g
p
By means of a switch, the buzzer circuit R B E may be con-
nected to the instrument for calibration purposes. 42 H
IL=)
FIGURE 9-Diagram of Connections
Figure 10 shows a top view of the instrument in detail. FIGURE 10-Top View of Decremeter
W=Wave length scale
C =Condenser scale in degrees
D =Decrement scale
H =Scale of hot wire instrument
A =Knob operating variable condenser
43
B=Clamping screw
S =Short circuit release
Z=Zero adjusting screw
a b= Terminals of condensers
c d= Terminals for series conniection H
IL=)
FIGURE 9-Diagram of Connections H
IL=)
FIGURE 9-Diagram of Connections FIGURE 9-Diagram of Connections FIGURE 9-Diagram of Connections FIGURE 9-Diagram of Connections Figure 10 shows a top view of the instrument in detail. FIGURE 10-Top View of Decremeter
W=Wave length scale
C =Condenser scale in degrees
D =Decrement scale
H =Scale of hot wire instrument
A =Knob operating variable condenser
B=Clamping screw
S =Short circuit release
Z=Zero adjusting screw
a b= Terminals of condensers
c d= Terminals for series conniection Figure 10 shows a top view of the instrument in detail. FIGURE 10-Top View of Decremeter
W Wave l
th
l
B Cl
i Figure 10 shows a top view of the instrument in detail. Figure 10 shows a top view of the instrument in detail. FIGURE 10-Top View of Decremeter W=Wave length scale
C =Condenser scale in degrees
D =Decrement scale
H =Scale of hot wire instrument
A =Knob operating variable condenser 43 EXPERIMENTAL DATA In order to determine the accuracy of the instrument for
measurement of N1 + 42 the following experiments were made: Experiment 1.-The decremeter was used as an ordinary wave
meter, loosely coupled to the secondary of a quenched spark
transmitter. A resonance curve was obtained similar to that
shown in Figure 2, and for several ratios I2, th + d2 was calcu-
lated from the formula Experiment 1.-The decremeter was used as an ordinary wave
meter, loosely coupled to the secondary of a quenched spark
transmitter. A resonance curve was obtained similar to that ormula
x2 +
Ir2_J2 ormula
x2 +
Ir2_J2 and the following values obtained: g
Ir2
81 + S2
12
1.180
0.0970
1.475
0.0893
1.735
0.0911
2.000
0.0903
Average
0.0919 A single measurement obtained by means of the decremeter
used as a direct reading instrument gave at once a value of 0.091
for 81 + d2 A further check was obtained by calculating the required value
of 02- 01 for ON + 'N = 0.091 from the formula 02-O =-log
forI2 = 2Ir2
m
J-.d
2r 02-O =-log
forI2 = 2Ir2
m
J-.d
2r
02-01 = 4°.75 by calculation
02
0
40 68 as determined fro 02-01 = 4°.75 by calculation
i 02 - 0 = 40.68 as determined from the experi-
mentally obtained resonance curve. 02
0
40.68 as determined from the experi
mentally obtained resonance curve. mentally obtained resonance curve. mentally obtained resonance curve. Experiment 2.-In this experiment the decremeter was again
loosely coupled to the secondary circuit of a quenched spark
transmitter and a direct measurement made of
Al + '2. A
resistance, in the form of a short straight piece of about No. 40
manganin wire, was then inserted in the circuit of the instrument
and a direct measurement made of 'N + 'N + J82, J'2 being the
additional decrement due to the inserted resistance. The capacity of the condenser and the frequency of the
oscillations being known, the value of the inserted resistance R
was calculated from the formula nC(o and the following results obtained. 44 TEST No. 1
s1 + s2 read from
81 + 82+J82 read
decrement scale
from decrement
of instrument
scale of instrument
0.132
0.168
0.130
0.169
0.130
0.163
0.131
0.172
0.130
0.167
Average
0.131
Average= 0.168
71+ `2+ J(2 = 0.168
C)1+4-2=0.131
.2J
0.037 TEST No. EXPERIMENTAL DATA 1
s1 + s2 read from
81 + 82+J82 read
decrement scale
from decrement
of instrument
scale of instrument
0.132
0.168
0.130
0.169
0.130
0.163
0.131
0.172
0.130
0.167
Average
0.131
Average= 0.168 71+ `2+ J(2 = 0.168
C)1+4-2=0.131
.2J
0.037 Capacity of condenser at resonance = 334 .f. O= 2 n
3.66 X 106. _(_ =963I7l
irCwt By measurement on D. C. bridge, R = 9.51fl Another test was made at a different frequency and conse-
quently with a different value of capacity. TEST No. 2
i + c72 + J(72 = 0.155
OS1 + *S2
= 0.099
JX2 = 0.056 Capacity of condenser at resonance = 764 k.pf. = 2-,rn = 2.47 X 10'
R = '(2 = 9.45Q
'rC(O Capacity of condenser at resonance = 764 k.pf. = 2-,rn = 2.47 X 10'
R = '(2 = 9.45Q
'rC(O Value of R measured on D. C. bridge = 9.51. Value of R measured on D. C. bridge = 9.51. Test No. 1: R= 9.63Q
Test No. 2: R = 9.45
Average = 9.54Q Experiment 3.-In this case resistance was inserted in the
secondary circuit of the transmitter and the value of this resist-
ance calculated in the same manner as in experiment 2. 45 TEST No. 1
d1+ Jd
2
+ 42= 0.141
01
+ 82 = 0.089
Jd1
= 0.052 Capacity of condenser at resonance = 3900 .,u.f. Capacity of condenser at resonance = 3900 .,u.f. Capacity of condenser at resonance = 3900 .,u.f. ( = 2.n = 3.35 X 106
R=
=1.27Q
;rC(I, ( = 2.n = 3.35 X 106
R=
=1.27Q
;rC(I, Value of R measured on D. C. bridge = 1.242Q Value of R measured on D. C. bridge = 1.242Q Value of R measured on D. C. bridge = 1.242Q TEST No. 2
31 + J81 + 82 = 0.111
, +
822= 0.074
JI1
= 0.037 Capacity of condenser at resonance = 3900 .f. Capacity of condenser at resonance = 3900 .f. wo = 2.43 X 106
R=
= 1.24Q
Test No. 1: R = 1.27Q
Test No. 2: R = 1.24Q
Average = 1.2550 wo = 2.43 X 106
R=
= 1.24Q DETERMINATION OF WAVE-LENGTH SCALE Since the capacity of the variable condenser in the instrument
varies according to a definitely known law, it is possible to attach
to this condenser a predetermined scale indicating wave lengths
directly. The graduations of the wave length scale are deter-
mined by calculation in the following manner: y
g
It has been shown that the capacity of the condenser may be
expressed as C = acme. p
The wave length A is proportional to \/C p
The wave length A is proportional to \/C
therefore,
i is proportional to V/¢mO p
The wave length A is proportional to \/C
therefore,
i is proportional to V/¢mO
mO
or
A is proportional to r 2 = nO
m
where
n= 2
46 h A is proportional to \/C
i is proportional to V/¢mO
mO
A is proportional to r 2 = nO
m
n= 2
46 \/
i is proportional to V/¢mO therefore,
i is proportional to V/¢mO mO
A is proportional to r 2 = nO or m
n= 2 where where 46 Now let A, be any wave length within the range of the instru-
ment, and A2 any other wave length desired, then A2
n6l2
0-01
SR.nOl
2_
__
- sn(O2-61)
log
= n(02 - 01)
A1
02-U1
log2 =
log-
n
Ai
m
Al
02 =,1i4m log A
m
A1 A2
n6l2
0-01
SR.nOl
2_
__
- sn(O2-61)
and
log
= n(02 - 01)
A1
or
02-U1
log2 =
log-
n
Ai
m
Al
therefore
02 =,1i4m log A
m
A1 and and or therefore therefore For the purpose of determining the graduations for the
scale, A, may be any wave length, as, for example, 300 meters. Furthermore, 6, may be taken as zero, for convenience, then 62 =
2- log
)2
m
300 From this equation 02 may be calculated for any wave length
22. The scale is arranged so that it can rotate about the shaft of
the variable condenser independently but remains stationary
when the condenser is rotated. A pointer is attached to the
shaft of the condenser and travels over the scale as the condenser
rotates. In order to cover a wide range of wave lengths, several coils
are supplied with the instrument, each coil covering a part of the
range. It is necessary, therefore, to adjust the position of the
wave length scale to correspond with the particular coil in circuit. DETERMINATION OF WAVE-LENGTH SCALE Red marks on the wave length scale indicate the maximum wave
length obtainable with the coils 1, 2, and 3, respectively. The
position of these red marks is determined experimentally. THE MEASUREMENT OF WAVE LENGTH The variable condenser is first set at 1800, the wave length
scale is then adjusted so that the red mark on the scale cor-
responding to the coil in circuit is directly under the pointer
attached-to the condenser shaft. The wave length scale remains
in this position and, as the condenser is varied, the pointer
travels over the scale, indicating the wave length when resonance
is obtained, as shown by the maximum reading of the hot-wire
instrument needle. WAThen the instrument is used as a receiver with detector and
telephones, or as a transmitter using the buzzer, the wave length 47 scale does not strictly apply for the wave length range below the
900 position of the condenser. In these cases it is necessary to
refer to calibration curves for the small correction. The instrument may be used in an interesting manner for
receiving audible signals from a source of undamped oscillations,
such as an arc circuit or a high frequency alternator. To ac-
complish this, oscillations are set up in the wave meter circuit
by means of the buzzer and the telephone receivers are con-
nected to the detector at T and to the binding post a instead of
b
(see Figure 10). Under these conditions, when undamped
oscillations are induced in the circuit, a heterodyne effect is
produced and the wave meter becomes a comparatively sensitive
receiver of undamped oscillations. Very weak harmonics existing
in arc circuits may be readily measured by using the instrument
in this manner. DETERMINATION OF THE DECREMENT OF THE INSTRUMENT To obtain the logarithmic decrement, 81 of the circuit under
test, it is necessary to know (J2, the decrement of the instrument,
in order that it may be subtracted from the scale reading which
is ,%+ An ideal method for determining d2 would be to charge the
condenser of the instrument at a given potential and allow it to
discharge thru the circuit, first without inserted resistance and
then with a known resistance inserted in the circuit, noting in
each case the reading of the hot-wire instrument. The energy in the circuit in both cases would then be equal
and, 112R = 122 (R + JR) where R is the resistance of the circuit and JR is the resistance
inserted. Then, R = JR
122
12- 122 where I22 and I12 represent, respectively, the readings of the hot-
wire instrument with and without inserted resistance, these read-
ings being proportional to the square of the current flowing in
the circuit. If the inductance L or capacity C of the circuit are known,
°2 is then determined for any value of (e), since, 02 =
=
RRC(o=;
L(o A method used in practice which approaches this ideal case
very closely is as follows: Energy is supplied to the instrument by means of impact
excitation, in which case nearly free oscillations exist in the
circuit of the instrument. These oscillations, therefore, have a
frequency and damping determined by the constants of the
circuit. To determine the resistance of the circuit, readings of
the hot-wire instrument are taken with and without inserted
resistance. The energy in the circuit, however, would not in
practice be strictly equal in the two cases and I12R = K I22 (R + JR)
or
R=JR
K=
_
I2- K122 or It has been shown in previous works
on this subject that Jo
K = 1 + d +
2 where di is the decrement of the exciting circuit, 42 the decrement
of the instrument circuit, and Jo the additional decrement due
to the insertion of a small resistance JR. DETERMINATION OF THE DECREMENT OF THE INSTRUMENT It is seen that for the case of impact excitation where 'i is very
large as compared to JIJQ K will be very nearly unity and for
practical purposes we may write 1221
R=R
2
J=JR
112 -I22
J12
1
122 Where it is desired to make JR, the inserted resistance equal
to R, the resistance of the instrument circuit, which corresponds
to making JJ equal to 42, then for the case of impact excitation, 112
122 On the other hand, if di = 0 as in the case of undamped oscilla-
tions, I2
K =2 and
= 4
122 In general, therefore, when it is desired to make the inserted In general, therefore, when it is desired to make the inserted
'Lehrbuch der Drahtlosen Telegraphic. Zenneck, 1913, p. 142. 49 resistance JR equal to the resistance of the instrument R, the
J.2
amount by which L12 must be reduced or the ratio of
£
depends
122
upon the ratio of °
to 82, for when J
= N~~~~~
01
+
02
01+
02
=0,K = 2 and
= 4
122
1 =C, K = 1 and I=2
122 and for and for For intermediate values of
l between 0 and c, K will vary from
J2
2 to 1 and the ratio
correspondingly from 4 to 2. 122 The most direct and simple method, however, for obtaining
'2 is to excite the instrument by means of undamped oscillations. Then C2 -(C1112
09 =
C2+ C
Ir212 as shown in the earlier part of this paper. If suitable means for producing undamped oscillation are not
available the method of impact excitation is very satisfactory,
provided that d1 is very large as compared with Jd. The curves in Figure 11 give the values of 02 for coils 1, 2, and
3, at various settings of the variable condenser. These values
were obtained by the method of impact excitation. On curve 3 are shown values of 82 obtained by using undamped
oscillations from a Poulsen arc as a source of excitation. In Figure 12, the assembled decremeter is shown, with the
cover of the carrying case turned back. As will be seen, the
coupling and decremeter coils and their supports are mounted
for transportation in the cover. The function of the other parts
shown is made clear by reference to the key of Figure 10. DETERMINATION OF THE DECREMENT OF THE INSTRUMENT Figure
13 is a photograph of the interior of the instrument. To the
right is the special variable condenser, and to the left the variable
condenser which is set once for all in the initial calibration of the
instrument. This latter condenser is of the rectangular sliding
interwoven plate type. 50 d2
03
0
200
40
60
80
100
11 001o 0
c
100
1800
\COL 3 *=IMPACT
EXCITATION
.02
____C
=UNDAMPED
.01XnO
200
400
600
800
1000
1200
1400
160
1800
FIGURE 11
Decrement of Instrument d2
03
0
200
40
60
80
100
11 001o 0
c
100
1800
\COL 3 *=IMPACT
EXCITATION
.02
____C
=UNDAMPED
.01XnO
200
400
600
800
1000
1200
1400
160
1800
FIGURE 11
Decrement of Instrument
SUMMARY: A new type of direct-reading decremeter and wave meter is
described. It is based on the Bjerknes method, but much simplifies the
measurement and increases the accuracy thereof. The variable condenser of
this wave meter is of a special type, the movable plates being cut to an ap-
propriate outline. By placing a fixed capacity in parallel with this variable
condenser, which fixed capacity is adjusted once for all in each instrument,
a standard and predetermined calibration curve can be used with all instru-
ments. The design and construction of the apparatus is given in detail, and
a number of measurements with it are considered critically. Several meth-
ods of determining the decrement of the instrument itself are shown. The
instrument can also be used as an exciting circuit of known wave length and
decrement. An interesting example of heterodyne reception of sustained
oscillations with this instrument is described. FIGURE 11
Decrement of Instrument SUMMARY: A new type of direct-reading decremeter and wave meter is
described. It is based on the Bjerknes method, but much simplifies the
measurement and increases the accuracy thereof. The variable condenser of
this wave meter is of a special type, the movable plates being cut to an ap-
propriate outline. By placing a fixed capacity in parallel with this variable
condenser, which fixed capacity is adjusted once for all in each instrument,
a standard and predetermined calibration curve can be used with all instru-
ments. The design and construction of the apparatus is given in detail, and
a number of measurements with it are considered critically. Several meth-
ods of determining the decrement of the instrument itself are shown. DETERMINATION OF THE DECREMENT OF THE INSTRUMENT The
instrument can also be used as an exciting circuit of known wave length and
decrement. An interesting example of heterodyne reception of sustained
oscillations with this instrument is described. 51 FIGURE 12-Decremeter Mounted in Leather Carrying Case FIGURE 12-Decremeter Mounted in Leather Carrying Case A
-i
o
-
t
,
o
gf
o
XS
9x,D
5
CD
;X
'I
'0
53 -i
o
t
o
o
9x,
5
CD
'I -i
o
t
o
o
9x,
5
CD
'I -i
o
t
o
o
9x,
5
CD
'I o
t
o
o
9x
5
CD
'I 53 53
|
https://openalex.org/W2625588031
|
https://link.springer.com/content/pdf/10.1007/978-3-319-60774-0.pdf
|
English
| null |
Security of Networks and Services in an All-Connected World
|
Lecture notes in computer science
| 2,017
|
cc-by
| 79,425
|
Daphne Tuncer · Robert Koch
Rémi Badonnel · Burkhard Stiller (Eds.)
11th IFIP WG 6.6 International Conference
on Autonomous Infrastructure, Management, and Security, AIMS 2017
Zurich, Switzerland, July 10–13, 2017, Proceedings
Security of Networks
and Services in an
All-Connected World Daphne Tuncer · Robert Koch
Rémi Badonnel · Burkhard Stiller (Eds.) Security of Networks
and Services in an
All-Connected World LNCS 10356 11th IFIP WG 6.6 International Conference
on Autonomous Infrastructure, Management, and Security, AIMS 2017
Zurich, Switzerland, July 10–13, 2017, Proceedings 11th IFIP WG 6.6 International Conference
on Autonomous Infrastructure, Management, and Security, AIMS 2017
Zurich, Switzerland, July 10–13, 2017, Proceedings 11th IFIP WG 6.6 International Conference
on Autonomous Infrastructure, Management, and Security, AIMS 2017
Zurich, Switzerland, July 10–13, 2017, Proceedings Lecture Notes in Computer Science
Commenced Publication in 1973
Founding and Former Series Editors:
Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen Lecture Notes in Computer Science
Commenced Publication in 1973
Founding and Former Series Editors:
Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen 10356 Editorial Board Editorial Board
David Hutchison
Lancaster University, Lancaster, UK
Takeo Kanade
Carnegie Mellon University, Pittsburgh, PA, USA
Josef Kittler
University of Surrey, Guildford, UK
Jon M. Kleinberg
Cornell University, Ithaca, NY, USA
Friedemann Mattern
ETH Zurich, Zurich, Switzerland
John C. Mitchell
Stanford University, Stanford, CA, USA
Moni Naor
Weizmann Institute of Science, Rehovot, Israel
C. Pandu Rangan
Indian Institute of Technology, Madras, India
Bernhard Steffen
TU Dortmund University, Dortmund, Germany
Demetri Terzopoulos
University of California, Los Angeles, CA, USA
Doug Tygar
University of California, Berkeley, CA, USA
Gerhard Weikum
Max Planck Institute for Informatics, Saarbrücken, Germany Jon M. Kleinberg
Cornell University, Ithaca, NY, USA Friedemann Mattern
ETH Zurich, Zurich, Switzerland John C. Mitchell
Stanford University, Stanford, CA, USA Moni Naor
Weizmann Institute of Science, Rehovot, Israel C. Pandu Rangan
Indian Institute of Technology, Madras, India Bernhard Steffen
TU Dortmund University, Dortmund, Germany Demetri Terzopoulos
University of California, Los Angeles, CA, USA Doug Tygar
University of California, Berkeley, CA, USA Gerhard Weikum
Max Planck Institute for Informatics, Saarbrücken, Germany More information about this series at http://www.springer.com/series/7411 Daphne Tuncer
• Robert Koch
Rémi Badonnel
• Burkhard Stiller (Eds.)
Security of Networks
and Services in an
All-Connected World
11th IFIP WG 6.6 International Conference
on Autonomous Infrastructure, Management, and Security, AIMS 2017
Zurich, Switzerland, July 10–13, 2017
Proceedings Daphne Tuncer
• Robert Koch
Rémi Badonnel
• Burkhard Stiller (Eds.) 11th IFIP WG 6.6 International Conference
on Autonomous Infrastructure, Management, and Security, AIMS 2017
Zurich, Switzerland, July 10–13, 2017
Proceedings ISSN 0302-9743
ISSN 1611-3349
(electronic)
Lecture Notes in Computer Science
ISBN 978-3-319-60773-3
ISBN 978-3-319-60774-0
(eBook)
DOI 10.1007/978-3-319-60774-0 ISSN 0302-9743
ISSN 1611-3349
(electronic)
Lecture Notes in Computer Science
ISBN 978-3-319-60773-3
ISBN 978-3-319-60774-0
(eBook)
DOI 10.1007/978-3-319-60774-0 Library of Congress Control Number: 2017943842 LNCS Sublibrary: SL5 – Computer Communication Networks and Telecommunications LNCS Sublibrary: SL5 – Computer Communication Networks and Telecommunications The Editor(s) (if applicable) and The Author(s) 2017. This book is an open access publication. A
Thi b
k i li
d
d
h
f h C
i
C
A
ib i
4 0 I © The Editor(s) (if applicable) and The Author(s) 2017. This book is an open access publication. Open Access This book is licensed under the terms of the Creative Commons Attribution 4.0 International
License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution
and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this book are included in the book’s Creative Commons license,
unless indicated otherwise in a credit line to the material. If material is not included in the book’s Creative
Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use,
you will need to obtain permission directly from the copyright holder. y
p
y
py g
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are
believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors
give a warranty, express or implied, with respect to the material contained herein or for any errors or
omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. Printed on acid-free paper Editors
Daphne Tuncer
University College London
London
UK
Robert Koch
Universität der Bundeswehr München
Neubiberg
Germany Rémi Badonnel
LORIA - Inria
Villers-lès-Nancy
France
Burkhard Stiller
University of Zurich
Zurich
Switzerland Preface The International Conference on Autonomous Infrastructure, Management, and
Security (AIMS 2017) is a single-track event targeted at junior researchers and PhD
students in network and service management and security. It features a range of ses-
sions including conference paper presentations, hands-on lab courses, and educational
keynotes. One of the key goals of AIMS is to offer junior researchers and PhD students
a dedicated place where they can discuss their research work and experience, receive
constructive feedback from senior scientists, and benefit from practical hands-on ses-
sions on emerging technologies. By putting the focus on junior researchers and PhD
students, AIMS acts as a complementary piece in the set of international conferences in
the network and service management community, providing an optimal environment
for in-depth discussions and networking. AIMS 2017 — which took place during July 10-13, 2017, in Zürich, Switzerland,
and was hosted by the University of Zürich — was the 11th edition of a conference
series on management and security aspects of distributed and autonomous systems. It
followed the already established tradition of an unusually vivid and interactive con-
ference series, after successful events in Munich, Germany, in 2016, Ghent, Belgium, in
2015, Brno, Czech Republic, in 2014, Barcelona, Spain, in 2013, Luxembourg, Lux-
embourg, in 2012, Nancy, France, in 2011, Zürich, Switzerland, in 2010, Enschede, The
Netherlands, in 2009, Bremen, Germany, in 2008, and Oslo, Norway, in 2007. AIMS 2017 focused on security of networks and services in an all-connected world. To address these challenges, solutions for the design, monitoring, configuration, and
protection of the next generation of networked systems in an efficient, secure, and smart
manner are investigated. The theme is reflected in the technical program with papers
presenting novel approaches and evaluation studies for the security management of rich
network services and environments. AIMS 2017 was organized as a 4-day program to
encourage the active participation of and interaction with the audience. The program
consisted of technical sessions for the main track and PhD sessions, interleaved with
three lab sessions and two keynotes. The lab sessions offered hands-on experience in the topics of security and advanced
network management techniques, and were organized in on-site labs preceded by short
tutorial-style teaching sessions. The first lab session was run by Martin Drašar
(Masaryk University, Czech Republic) and focused on an introduction to security
games. Printed on acid-free paper This Springer imprint is published by Springer Nature This Springer imprint is published by Springer Nature The registered company is Springer International Publishing AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland Preface Preface Preface The technical program consisted of six sessions, divided into three full-paper sessions
and three short-paper sessions. The three full-paper sessions covered technical pre-
sentations on the themes of: (1) Security Management, (2) Management of Cloud
Environments and Services, and (3) Evaluation and Experimental Study of Rich
Network Services. They included a total of eight full papers, which were selected after a
thorough reviewing process out of 24 submissions. Each paper received at least three
independent reviews. The three short-paper sessions included 11 short papers. These
covered PhD research papers on the themes of “Methods for the Protection of Infras-
tructure and Services,” and “Autonomic and Self-Management Solutions” as well as six
short presentations on the topic of “Security, Intrusion Detection, and Configuration.” During all the PhD research presentations, doctoral students had the opportunity to
present and discuss their research ideas, and more importantly to obtain valuable
feedback from the AIMS audience about their PhD research work. All PhD research
proposals included in this volume describe the current state of these investigations,
including well-defined research problem statements, proposed approaches, and an
outline of emerging and promising results achieved to date. The present volume of the Lecture Notes in Computer Science series includes all
papers presented at AIMS 2017 as defined within the overall final program. It
demonstrates again the European scope of this conference series, since most of the
accepted papers originate from European research groups. In addition, by hosting two
tracks specifically dedicated to research proposals, AIMS 2017 stayed true to its
defined DNA of a conference with a strong educational goal, focusing especially on
issues and challenges associated with the security of networks and services. The editors would like to thank the many people who helped to make AIMS 2017
such a high-quality and successful event. Firstly, many thanks are extended to all
authors who submitted their contributions to AIMS 2017, and to the lab session
speakers as well as the keynote speakers. The great review work performed by the
members of the AIMS Technical Program Committee as well as additional reviewers is
greatly acknowledged. Thanks also to Thomas Bocek and Martin Drašar for organizing
the lab sessions. Additionally, many thanks are extended to the local organizers for
handling logistics and hosting the AIMS 2017 event. Preface The second lab session was supervised by Thomas Bocek and Moritz Schneider
(University of Zürich, Switzerland) and presented how to program smart contracts. Finally, the last session was held by Salvatore Signorello (University of Luxembourg,
Luxembourg) and Jérôme François (Inria, France) and explored P4, the emerging
high-level data plane programing language and its applicability to packet processors. The keynotes were presented by two experts in their domain: Marcel Waldvogel
(University of Konstanz, Germany), who discussed “Getting Rid of IoT Insecurity,”
and Matthias Bossardt (KPMG, Switzerland), who shared his view with the audience
on “Cyber Security Challenges – A Business Perspective.” VI Preface Finally, the editors would like to express their thanks to Springer, especially Anna
Kramer, for the smooth cooperation in finalizing these proceedings. Additionally, special
thanks go to the AIMS 2017 supporters, University of Zürich UZH, Communication
Systems Group CSG, Research Institute for Cyber Defense and Smart Data CODE,
München, Germany, and the European FP7 NoE FLAMINGO under Grant No. 318488. May 2017
Daphne Tuncer
Robert Koch
Rémi Badonnel
Burkhard Stiller
NoE
FLAMINGO May 2017 May 2017 Daphne Tuncer
Robert Koch
Rémi Badonnel
Burkhard Stiller Daphne Tuncer
Robert Koch
Rémi Badonnel
Burkhard Stiller NoE
FLAMINGO General Chair AIMS 2017
Burkhard Stiller
University of Zürich, Switzerland
Technical Program Committee Co-chairs
Daphne Tuncer
University College London, UK
Robert Koch
Universität der Bundeswehr München, Germany
Labs Co-chairs
Martin Drašar
Masaryk University, Czech Republic
Thomas Bocek
University of Zürich, Switzerland
Publications Co-chairs
Rémi Badonnel
LORIA, Inria, France
Burkhard Stiller
University of Zürich, Switzerland
Local Chair
Barbara Jost
University of Zürich, Switzerland
Publicity Chair and Web Master
Corinna Schmitt
University of Zürich, Switzerland
AIMS Steering Committee
Anna Sperotto
University of Twente, The Netherlands
Pavel Čeleda
Masaryk University, Czech Republic
Filip De Turck
Ghent University - iMinds, Belgium
Rémi Badonnel
LORIA, Inria, France
Aiko Pras
University of Twente, The Netherlands
Burkhard Stiller
University of Zürich, Switzerland
Robert Koch
Universität der Bundeswehr München, Germany
Technical Program Committee
Alexander Clemm
Huawai, USA
Alexander Keller
IBM Global Technology Services, USA General Chair AIMS 2017
Burkhard Stiller
University of Zürich, Switzerland
Technical Program Committee Co-chairs
Daphne Tuncer
University College London, UK
Robert Koch
Universität der Bundeswehr München, Germany
Labs Co-chairs
Martin Drašar
Masaryk University, Czech Republic
Thomas Bocek
University of Zürich, Switzerland
Publications Co-chairs
Rémi Badonnel
LORIA, Inria, France
Burkhard Stiller
University of Zürich, Switzerland
Local Chair
Barbara Jost
University of Zürich, Switzerland
Publicity Chair and Web Master
Corinna Schmitt
University of Zürich, Switzerland
AIMS Steering Committee
Anna Sperotto
University of Twente, The Netherlands
Pavel Čeleda
Masaryk University, Czech Republic
Filip De Turck
Ghent University - iMinds, Belgium
Rémi Badonnel
LORIA, Inria, France
Aiko Pras
University of Twente, The Netherlands
Burkhard Stiller
University of Zürich, Switzerland
Robert Koch
Universität der Bundeswehr München, Germany
Technical Program Committee
Alexander Clemm
Huawai, USA
Alexander Keller
IBM Global Technology Services, USA Publications Co-chairs VIII Organization Tufts University, USA
Imperial College London, UK
University of Twente, The Netherlands
Orange Labs, France
Université de Mons, Belgium
University of Zürich, Switzerland
Imperial College London, UK
Otto-von-Guericke Universität Magdeburg, Germany
Ghent University - iMinds, Belgium
Universität der Bundeswehr München, Germany
Troyes University of Technology, France
TELECOM Nancy, Université de Lorraine, France
Brno University of Technology, Czech Republic
Inria Nancy Grand Est, France
Universitat Politècnica de Catalunya, Spain
Jacobs University Bremen, Germany
Blekinge Institute of Technology, Sweden
UFRGS, Brazil Alva L. Couch
Anandha Gopalan
Anna Sperotto
Bertrand Mathieu
Bruno Quoitin
Burkhard Stiller
Daniele Sgandurra
David Hausheer
Filip De Turck
Gabi Dreo Rodosek
Guillaume Doyen
Isabelle Chrisment
Jan Kořenek
Jérôme François
Joan Serrat
Jürgen Schönwälder
Kurt Tutschku
Lisandro Zambenedetti
Granville
Mario Golling
Martin Barrère
Mauro Tortonesi
Michelle Sibilla
Olivier Festor
Pavel Čeleda
Philippe Owezarski
Rashid Mijumbi
Rémi Badonnel
Ricardo Schmidt
Roberto Riggio
Steven Latré
Thomas Bocek Alva L. Couch
Anandha Gopalan
Anna Sperotto
Bertrand Mathieu
Bruno Quoitin
Burkhard Stiller
Daniele Sgandurra
David Hausheer
Filip De Turck
Gabi Dreo Rodosek
Guillaume Doyen
Isabelle Chrisment
Jan Kořenek
Jérôme François
Joan Serrat
Jürgen Schönwälder
Kurt Tutschku
Lisandro Zambenedetti
Granville
Mario Golling
Martin Barrère
Mauro Tortonesi
Michelle Sibilla
Olivier Festor
Pavel Čeleda
Philippe Owezarski
Rashid Mijumbi
Rémi Badonnel
Ricardo Schmidt
Roberto Riggio
Steven Latré
Thomas Bocek Universität der Bundeswehr München, Germany
Imperial College London, UK
University of Ferrara, Italy
Paul Sabatier University, France
INRIA Nancy Grand Est, France
Masaryk University, Czech Republic
LAAS-CNRS, France
Waterford Institute of Technology, Ireland
Telecom Nancy, Université de Lorraine, France
University of Twente, The Netherlands
CREATE-NET, Italy
University of Antwerp, iMinds, Belgium
University of Zürich, Switzerland Universität der Bundeswehr München, Germany
Imperial College London, UK
University of Ferrara, Italy
Paul Sabatier University, France
INRIA Nancy Grand Est, France
Masaryk University, Czech Republic
LAAS-CNRS, France
Waterford Institute of Technology, Ireland
Telecom Nancy, Université de Lorraine, France
University of Twente, The Netherlands
CREATE-NET, Italy
University of Antwerp, iMinds, Belgium
University of Zürich, Switzerland Additional Reviewers Detailed reviews for papers submitted to AIMS 2017 were undertaken by the Technical
Program Committee as listed above and additionally by the following reviewers: Messaoud Aouadj, Jeremias Blendin, Remi Cogranne, Ariel Dalla-Costa, Muriel Franco,
Borislava Gajic, Christian Jacquenet, Christian Koch, Radek Krejci, Genaro Longoria,
Christian Mannweiler, Hassnaa Moustafa, Tan Nguyen, Leonhard Nobach, Vinícius
Schaurich, and Eder John Scheid. Getting Rid of IoT Insecurity Marcel Waldvogel University of Konstanz, Distributed Systems Group, Universitätsstr. 10/229,
78457 Konstanz, Germany
Marcel.Waldvogel@uni-konstanz.de University of Konstanz, Distributed Systems Group, Universitätsstr. 10/229,
78457 Konstanz, Germany
Marcel.Waldvogel@uni-konstanz.de Abstract. The Internet-of-Things (IoT) is already everywhere, but even then,
there is still much more to come. Right now, IoT security is a mess, chaotic,
unsustainable, and unmanageable. To prevent this is going to remain like this,
and that these devices will continue to risk or endanger increasing amounts of
our and everybody's lives, we need coordinated actions by manufacturers,
vendors, integrators, ISPs, and customers. But it is the researchers, you, who need to make a long-term difference: how
to create blueprints, on which new products may be based, which may include
design for privacy, security, manageability, while not overwhelming the users is
probably the biggest challenge of them all. This talk will present three examples, which clearly outlines the challenges,
describes open problems, and proposes a coherent framework, into which your
next solutions hopefully will fit. Cyber Security Challenges –
A Business Perspective Matthias Bossardt Lead Partner for Cyber Security, KPMG Switzerland, Zürich, Switzerland
mbossardt@kpmg.com Lead Partner for Cyber Security, KPMG Switzerland, Zürich, Switzerland
mbossardt@kpmg.com Abstract. This keynote will shed light on real world challenges that companies
face when dealing with cyber threats on a global scale. In global organizations
and where cyber security has to scale to hundred thousands of employees,
contractors, suppliers, and clients as well as thousands of business processes and
applications, understanding the organization’s risk exposure and implementing
effective protection measures is very complex. And the plethora of challenges related to the (Industrial) Internet-of-Things
and managing cyber security becomes a daunting task. To secure an organiza-
tion, understanding human behavior and mastering organizational change is as
important as implementing security technology. This talk will discuss those
security capabilities needed in an organization and it will highlight those topics
that can benefit greatly from additional research. CSIRT-MU, Masaryk University, Brno, Czech Republic
drasar@ics.muni.cz CSIRT-MU, Masaryk University, Brno, Czech Republic
drasar@ics.muni.cz Abstract. Maintaining infrastructure security or hardening a system is never a
simple task. Nor it is a one-click operation. Often it requires the adoption of
attacker’s mindset to identify correctly weak spots or to even understand that a
threat is imminent. This, however, is not possible without acquiring a large body
of knowledge, which is usually dispersed around the Internet or available only
as dry technical reports. While the process of assembling these bits of infor-
mation may appeal to somebody, a majority will prefer something more
entertaining. Security games are one such approach. This lab is aimed at beginners and will serve as a brief introduction to
hacking as a way to better understand computer security. It will discuss available
learning resources and focus mostly on security games: why, which, where, and
how to play them for maximum benefit? It will also give participants an
opportunity to try out some of these games in a guided manner. These games
will be executed both locally as virtual machines on attendees' laptops and
remotely in a virtual sandbox environment [1]. Attendees will also be asked to
participate in a survey regarding skill self-assessment and effectiveness of
knowledge transfer, which fosters further research as presented in [2]. Hacking your Way to Safety – A Beginner’s
Guide to Security Games Martin Drašar
CSIRT-MU, Masaryk University, Brno, Czech Republic
drasar@ics.muni.cz Martin Drašar CSIRT-MU, Masaryk University, Brno, Czech Republic
drasar@ics.muni.cz 2. Ykopal, J., Bartak, M.: On the Design of Security Games: From Frustrating to Engaging
Learning, In: USENIX Workshop on Advances in Security Education. ASE 2016, Austin,
Texas, USA, August 2016 1. Kourill, D., Rebok, T., Jirsik, T., Čegan, J., Drasar, M., Vizvary, M., Vykopal, J.:
Cloud-based Testbed for Simulation of Cyber Attacks. In: IFIP/IEEE Network Operations and
Management Symposium. NOMS 2014, Krakow, Poland, May 2016
2. Ykopal, J., Bartak, M.: On the Design of Security Games: From Frustrating to Engaging
Learning, In: USENIX Workshop on Advances in Security Education. ASE 2016, Austin,
Texas, USA, August 2016 Programming Smart Contracts Thomas Bocek and Moritz Schneider University of Zürich UZH, Department of Informatics IfI, Communication
Systems Group CSG, Binzmühlestrasse 14, 8050 Zürich, Switzerland
bocek@ifi.uzh.ch, moritz.schneider3@uzh.ch University of Zürich UZH, Department of Informatics IfI, Communication
Systems Group CSG, Binzmühlestrasse 14, 8050 Zürich, Switzerland
bocek@ifi.uzh.ch, moritz.schneider3@uzh.ch Abstract. Blockchains and smart contracts have gained a lot of attention. Public
blockchains are considered secure and exist without centralized control. As one
of the most prominent blockchain examples, Bitcoin has the potential to disrupt
financial services. However, the blockchain technology is applicable to a wider
range of application domains, such as smart contracts, public registries, registry
of deeds, or virtual organizations. Another prominent blockchain example, Ethereum, which is considered a
general approach for smart contracts, is the second biggest public blockchain
with respect to market capitalization. A smart contract in Ethereum [1] is written
in the language Solidity [2]. These contracts allow not only sending and
receiving funds, but since Solidity its a Turing-complete language, it allows for
the definition of any kind of rules. The introduction of this lab session will address the history and an overview
of blockchains as well as their categorization. Blockchain basics are explained in
terms of basic building blocks and how they work, including the essential
consensus mechanisms. Thus, the Solidity language is introduced in terms of
syntax and main constructs, combined with simple code snippets and examples
[3]. The audience will compile and deploy a simple smart contract with the goal
to familiarize itself with the language and the development environment. Fur-
thermore, the lab shows on the basis of Ethereum smart contracts how to create
your own tokens or cryptocurrency [4]. The tokens or cryptocurrency initiator
can create initial tokens that can be transferred to any address. References 1. Kourill, D., Rebok, T., Jirsik, T., Čegan, J., Drasar, M., Vizvary, M., Vykopal, J.:
Cloud-based Testbed for Simulation of Cyber Attacks. In: IFIP/IEEE Network Operations and
Management Symposium. NOMS 2014, Krakow, Poland, May 2016 2. Ykopal, J., Bartak, M.: On the Design of Security Games: From Frustrating to Engaging
Learning, In: USENIX Workshop on Advances in Security Education. ASE 2016, Austin,
Texas, USA, August 2016 3. Contract examples for Ethereum. https://github.com/fivedogit/solidity-baby-steps. Accessed
May 1, 2017 1. Homestead Release: ethereum. https://www.ethereum.org/. Accessed May 1, 2017 4. Create your own crypto-currency with Ethereum. https://www.ethereum.org/token. Accessed
May 1, 2017 2. Solidity. http://solidity.readthedocs.io. Accessed May 1, 2017 2. Solidity. http://solidity.readthedocs.io. Accessed May 1, 2017 1. Homestead Release: ethereum. https://www.ethereum.org/. Accessed May 1, 2017
2. Solidity. http://solidity.readthedocs.io. Accessed May 1, 2017
3. Contract examples for Ethereum. https://github.com/fivedogit/solidity-baby-steps. Accessed
May 1, 2017
4. Create your own crypto-currency with Ethereum. https://www.ethereum.org/token. Accessed
May 1, 2017 References 1. Homestead Release: ethereum. https://www.ethereum.org/. Accessed May 1, 2017 2. Solidity. http://solidity.readthedocs.io. Accessed May 1, 2017 1. Bosshart, P., Daly, D., Gibb, G., Izzard, M., McKeown, N., Rexford, J., Schlesinger, C.,
Talayco, D., Vahdat, A., Varghese, G., Walker, D.: P4: Programming Protocol-independent
Packet Processors. Comput. Commun. Rev. 44(3), 87–95
2. P4. http://p4.org/join-us Programming Data Planes in P4 – A High-level
Language for Packet Processors Salvatore Signorello1 and Jérôme François2 Salvatore Signorello1 and Jérôme François2
1 SnT, University of Luxembourg, Luxembourg, and LORIA,
University of Nancy, Nancy, France
2 MADYNES Team at INRIA, Nancy Grand-Est, France
salvatore.signorello@uni.lu, jerome.francois@inria.fr 1 SnT, University of Luxembourg, Luxembourg, and LORIA,
University of Nancy, Nancy, France
2 MADYNES Team at INRIA, Nancy Grand-Est, France
alvatore.signorello@uni.lu, jerome.francois@inria.fr Abstract. This lab will introduce the audience to the P4 language [1], providing
them with the knowledge necessary to develop and prototype their own research
ideas in P4. The lab starts by providing an overview of the research that led to
the emergence of the language and by illustrating the P4 language consortium
objectives and related ongoing activities. Additionally, the lab explains the P4
language programming model and introduces an open source development
environment [2], which can be used to write and test P4 programs on a single
machine. The presented software toolset includes a P4 front-end compiler, a P4
software target, and the Command Line Interface (CLI) used to program this
target at run-time. Finally, the lab interactively presents the language’s syntax
and main constructs. Throughout the entire lab, simple P4 code snippets and examples are written,
compiled, and executed by the participants. Furthermore, full assignments of
increasing complexity are proposed to strengthen the understanding of the
programming model and of the main language constructs. More in detail, simple
tasks, like the definition of a custom encapsulation protocol and the imple-
mentation of an access control list, help the audience to familiarize itself with the
definition and the parsing of new protocols and with the definition of the control
flow of a P4 program. While more complex assignments, like the implemen-
tation of a port-knock firewall, are meant to explore advanced language con-
structs, which can be used to implement stateful network functions. Contents Security Management
Making Flow-Based Security Detection Parallel . . . . . . . . . . . . . . . . . . . . . 3
Marek Švepeš and Tomáš Čejka
A Blockchain-Based Architecture for Collaborative DDoS Mitigation
with Smart Contracts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Bruno Rodrigues, Thomas Bocek, Andri Lareida, David Hausheer,
Sina Rafati, and Burkhard Stiller
Achieving Reproducible Network Environments with INSALATA . . . . . . . . 30
Nadine Herold, Matthias Wachs, Marko Dorfhuber, Christoph Rudolf,
Stefan Liebald, and Georg Carle
Management of Cloud Environments and Services
Towards a Software-Defined Security Framework for Supporting
Distributed Cloud . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Maxime Compastié, Rémi Badonnel, Olivier Festor, Ruan He,
and Mohamed Kassi-Lahlou
Optimal Service Function Chain Composition in Network
Functions Virtualization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Andrés F. Ocampo, Juliver Gil-Herrera, Pedro H. Isolani,
Miguel C. Neves, Juan F. Botero, Steven Latré, Lisandro Zambenedetti,
Marinho P. Barcellos, and Luciano P. Gaspary
Evaluation and Experimental Study of Rich Network Services
An Optimized Resilient Advance Bandwidth Scheduling for Media
Delivery Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Maryam Barshan, Hendrik Moens, Bruno Volckaert, and Filip De Turck
The Evaluation of the V2VUNet Concept to Improve
Inter-vehicle Communications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Lisa Kristiana, Corinna Schmitt, and Burkhard Stiller
Towards Internet Scale Quality-of-Experience Measurement with Twitter. . . . References 1. Bosshart, P., Daly, D., Gibb, G., Izzard, M., McKeown, N., Rexford, J., Schlesinger, C.,
Talayco, D., Vahdat, A., Varghese, G., Walker, D.: P4: Programming Protocol-independent
Packet Processors. Comput. Commun. Rev. 44(3), 87–95
2. P4. http://p4.org/join-us Contents 108
Dennis Kergl, Robert Roedler, and Gabi Dreo Rodosek XX Contents Short Papers: Security, Intrusion Detection, and Configuration
Hunting SIP Authentication Attacks Efficiently. . . . . . . . . . . . . . . . . . . . . . 125
Tomáš Jansky, Tomáš Čejka, and Václav Bartoš
MoDeNA: Enhancing User Security for Devices in Wireless Personal
and Local Area Networks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Robert Müller, Marcel Waldvogel, and Corinna Schmitt
Flow-Based Detection of IPv6-specific Network Layer Attacks. . . . . . . . . . . 137
Luuk Hendriks, Petr Velan, Ricardo de O. Schmidt,
Pieter-Tjerk de Boer, and Aiko Pras
Towards a Hybrid Cloud Platform Using Apache Mesos . . . . . . . . . . . . . . . 143
Noha Xue, Hårek Haugerud, and Anis Yazidi
Visual Analytics for Network Security and Critical Infrastructures. . . . . . . . . 149
Karolína Burská and Radek Ošlejšek
Preserving Relations in Parallel Flow Data Processing . . . . . . . . . . . . . . . . . 153
Tomáš Čejka and Martin Žádník
Ph.D. Track: Autonomic and Self-Management Solutions
SmartDEMAP: A Smart Contract Deployment and Management Platform . . . 159
Markus Knecht and Burkhard Stiller
Optimizing the Integration of Agent-Based Cloud Orchestrators
and Higher-Level Workloads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Merlijn Sebrechts, Gregory Van Seghbroeck, and Filip De Turck
Ph.D. Track: Methods for the Protection of Infrastructure and Services
Situational Awareness: Detecting Critical Dependencies and Devices
in a Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Martin Laštovička and Pavel Čeleda
A Framework for SFC Integrity in NFV Environments . . . . . . . . . . . . . . . . Contents 179
Lucas Bondan, Tim Wauters, Bruno Volckaert, Filip De Turck,
and Lisandro Zambenedetti Granville
Multi-domain DDoS Mitigation Based on Blockchains . . . . . . . . . . . . . . . . 185
Bruno Rodrigues, Thomas Bocek, and Burkhard Stiller
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191 Short Papers: Security, Intrusion Detection, and Configuration c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 3–15, 2017.
DOI: 10.1007/978-3-319-60774-0 1 1 FIT, CTU in Prague, Thakurova 9, 160 00 Prague 6, Czech Republic
svepemar@fit.cvut.cz
2 CESNET, a.l.e., Zikova 4, 160 00 Prague 6, Czech Republic
cejkat@cesnet.cz 1 FIT, CTU in Prague, Thakurova 9, 160 00 Prague 6, Czech Republic
svepemar@fit.cvut.cz
2 CESNET, a.l.e., Zikova 4, 160 00 Prague 6, Czech Republic
cejkat@cesnet.cz Abstract. Flow based monitoring is currently a standard approach suit-
able for large networks of ISP size. The main advantage of flow process-
ing is a smaller amount of data due to aggregation. There are many
reasons (such as huge volume of transferred data, attacks represented
by many flow records) to develop scalable systems that can process flow
data in parallel. This paper deals with splitting a stream of flow data
in order to perform parallel anomaly detection on distributed computa-
tional nodes. Flow data distribution is focused not only on uniformity
but mainly on successful detection. The results of an experimental analy-
sis show that the proposed approach does not break important semantic
relations between individual flow records and therefore it preserves detec-
tion results. All experiments were performed using real data traces from
Czech National Education and Research Network. Making Flow-Based Security Detection Parallel Marek ˇSvepeˇs1 and Tom´aˇs ˇCejka2(B) 1
Introduction Flow-based monitoring plays a key role in network management. Not only it
provides an overview of the traffic mix, it greatly helps with network security
issues such as malicious traffic detection. There are many types of malicious traffic that should be detected in real
networks. As the speed and size of computer networks grow, it is necessary
for network operators to process more and more data to be informed about
the status of their network. However, with the increasing traffic volume, it is
difficult to run lots of detection algorithms at once using just a single machine. The more data, the more computing resources are needed and the longer time
the processing takes. In order to overcome resource limits of a single machine, parallelism plays
an important role. Various types of scalable architecture have been invented to
process data in parallel. Generally, to be able to process more data, analyzer
has to either run parts of its algorithms in parallel or split data for separate
processing units. Since the parallelization of individual detection algorithms is very dependent
on the nature of the algorithm and, additionally, according to Amdahl’s law, M. ˇSvepeˇs and T. ˇCejka 4 there are parts of algorithms that can’t be run in parallel, we have decided to
focus on data distribution for independent processing units. Our aim is to split a
continuous stream of network data (more specifically flow records, i.e. aggregated
packet headers) into much smaller subsets that are being processed separately
in parallel. We also focus on evaluation of the impact of data splitting on the
security analysis results. The contribution of this paper is to present our experiments with processing
data traces from the real backbone network. The aim is to use existing algorithms
from a single machine processing and deploy them in a distributed environment. This paper shows, that data splitting for such purpose is complicated due to
semantic relations in data which should be preserved. Breaking the relations can
cause that the obtained detection results are significantly worse than using a
single processing machine. The paper also shows a feasible way how to split flow
data with respect to semantic relations. Proposed approach preserves detection
results and allows a scalable deployment. This paper is organized as follows. Sect. 2 describes existing related work, i.e. systems for anomaly detection, traffic sampling and network traffic processing
in parallel. Sect. 1
Introduction 3 describes scattering methods that can be used to split flow
records into a separate groups for parallel processing by independent compu-
tational nodes. Sect. 4 describes our testing environment that was created for
our experiments. The section also presents results of measurement of described
scattering methods. Sect. 5 concludes the paper. 2
Related Work This section describes related approaches of parallel network traffic analysis
and anomaly detection usually done using Network Intrusion Detection System
(NIDS) or Network Intrusion Prevention System (NIPS). There are many existing systems for network traffic analysis and anomaly
detection that are modular by design. For instance, TOPAS [1] and NEMEA [2]
are flow-based systems that consists of modules that process data. When there
is a big volume of flow data, running the systems on a single machine may reach
resource limits of the machine. The systems do not support data distribution
natively as it is available for various big data frameworks. However, NEMEA
modules can be easily run and connected in a distributed environment. Xinidis et al. in [3] presented an architecture with Active Splitter for distrib-
uted NIDS aiming for performance optimization of the detection sensors running
Snort [4] (packet-based system performing deep packet inspection). The split-
ter uses hash functions for packet distribution and three techniques to optimize
the performance of the sensors. Cumulative Acknowledgements reduce redun-
dant sending of packets between splitter and sensors, Early Filtering in splitter
applies Snort rule subset on packet headers (no payload inspection) and finally
Locality Buffering reorders packets in a way that improves the locality of sensors
memory accesses. Scalable Approach of Detection 5 Sallay et al. in [5] made the network traffic analysis distributed using
switch/router. The architecture contains dedicated sensors for individual ser-
vices (e.g. FTP) and the incoming traffic is forwarded to them according to
switching table of the switch/router. Sensors are running Snort but only with
needed rule subset for their service. Since the volume of traffic of individual
services can differ significantly, the load of computational nodes wouldn’t be
uniform. Therefore, this approach is not suitable for us. Kim et al. in [6] compare static and dynamic hash-based load balancing
schemes and propose dynamic (i.e. adaptive) load-balancing scheme for NIDS. It uses a lookup table which is periodically reorganized according to historical
packet distribution and current load of individual nodes. If needed, flows with
the smallest volume are reorganized. Proposed method distributes packets in a
way that does not break the flow stream, however, they don’t take into account
relations between individual flow records and the impact of splitting on detection
results is not evaluated. Valentin et al. in [7] presented a NIDS cluster for scalable intrusion detec-
tion. 2
Related Work It consists of frontend nodes that distribute packets between backend nodes
running Bro [8] for intrusion detection. Moreover, there are proxy nodes prop-
agating state information of backend nodes and also one central manager node
for collecting and aggregating results. Each frontend node distributes data from
one monitored line and uses a hashing distribution scheme with a single hash
function. The architecture requires backend nodes and proxy nodes to exchange
data with detection subresults. In our approach, we are dealing with splitting
a stream of flow records instead of packets. Our hashing distribution scheme is
adjusted to provide all needed data to the detection methods for correct intru-
sion detection. Therefore, our computational nodes running intrusion detection
are independent and don’t communicate with each other. Finally, our proposed
hashing distribution scheme represents a general way, how to split flow data with
respect to detection results. Big data frameworks such as Hadoop [9] or Spark [10] are distributed by
nature. They are based on storage of data onto some distributed file system. A special designed parallel algorithm can be used to run on many distributed
nodes and process all data. A universal and the most popular approach of dis-
tributed processing is MapReduce. However, the overall result of this kind of
computation usually depends on the Reduce phase that merges local results
from all nodes. Therefore, only low attention is paid to any relations or seman-
tics during the data distribution and storage. An improved data distribution
in Hadoop was presented as Hashdoop in [11]. Contrary, our approach is more
general and it is applicable even on stream-wise processing with multiple differ-
ent algorithms. Even though the main focus of ours is to make non-distributed
system working in parallel, the principle described in our paper can be used for
improvement of data distribution in big data frameworks as well. Sampling has a common goal with parallel processing – capability to handle
more data at the same time. Mai in [12] shows impact of the packet sampling on
detection of portscanning and Bartos in [13] deals with flow sampling techniques
for anomaly detection. However none of these approaches can be applied on data
splitting. M. ˇSvepeˇs and T. ˇCejka 6 3
Flow Distribution Scheme When designing a distribution scheme, several aspects have to be taken into
account: (i) the data should be distributed uniformly between all computational
nodes, (ii) the distribution algorithm should be as fast as possible in order to
process as much data as possible, (iii) the impact of splitting the data on detec-
tion results should be minimized. In general, there are two ways how to distribute the data, statically or dynam-
ically (also called static and dynamic load-balancing) and both have some pros
and cons when applied in parallel NIDS. Static distribution has immutable rules
for splitting the data e.g. a packet with source IP address 1.2.3.4 goes to node 1
and a packet with source IP address 5.6.7.8 goes to node 2. This preserves the
data stream with possible security incident. However, it cannot affect the load
of individual computational nodes when the distribution is not uniform. On the
other hand dynamic distribution can perform some actions in order to make the
load uniform (e.g. redirect some packets to less loaded computational nodes). Unfortunately, this behaviour can make the security incident invisible. There-
fore, we have decided to use static distribution and focus on uniformity. Figure 1 shows high level view of the infrastructure of scalable and distributed
network flows analysis using NIDS. The following subsections describe several
splitting mechanisms used in the flow scatter. Fig. 1. A high level view of the infrastructure of scalable and distributed network flows
analysis using NIDS. Fig. 1. A high level view of the infrastructure of scalable and distributed network flows
analysis using NIDS. 3.1
Random Scattering Lets assume the detection results are not dependent on any semantic relations
between flow data, i.e. scattering mechanism can distribute the data regardless
of the information from flow records. In that case, the flow scatter can distribute Scalable Approach of Detection 7 the flow records between nodes using statistical uniform distribution, which is
optimal for load-balancing. Received records by flow scatter are forwarded to
computational nodes according to random number generator. It is clear that
every random distribution splits the flow records into different subsets. However,
as we discuss in Evaluation (Sect. 4), breaking semantic relations in flow data
using random distribution affects the detection results. 3.2
Scattering Based on Network Topology Scattering based on network topology is another logical way of distributing the
flow data. Since the computer networks are designed using hierarchical model
that usually respects geographical and logical division into subnets and network
lines. Figure 2 shows a high level topology of CESNET2 National Research and
Education Network (NREN), which is a backbone academic network and it is
also used as a transit network. It is inter-connected with other networks via
several lines that are being monitored. The data taken from the monitoring
probes contain a line identification—the line number. Flow scatter can easily
distribute the data using these line numbers. Standard monitoring infrastructure collects flow records from monitoring
probes onto one central collector. In case of scattering based on network topol-
ogy, this concept can be changed and it would be more efficient to send exported
flow records directly to computational nodes. Fig. 2. Topology of Czech national research and education network (NREN) CES-
NET2, network traffic on the perimeter is analyzed. Fig. 2. Topology of Czech national research and education network (NREN) CES-
NET2, network traffic on the perimeter is analyzed. M. ˇSvepeˇs and T. ˇCejka M. ˇSvepeˇs and T. ˇCejka 8 3.3
Hash-Based Scattering Hash functions are used to transform an input data into an output form with
a fixed length. Cryptography expects that the output of an ideal hash function
meets requirements such as uniform distribution and missing relation between
output and input. In our case, the hash function can be used in the flow scatter to
select an appropriate computational node number uniformly. Information from
the incoming flow records can be used as an input for the hash function. The dependency of selected node number on the input data of the hash func-
tion leads to divison of flow records into subsets with the same characteristics. The subsets with the same characteristics are then processed together on the
same computational node and this can be used to preserve the detection results. For instance, if we use only the source IP address for hashing, all flow records
having the same source IP address ends up on the same node. Meanwhile, flow
records with different source IP addresses have a high probability to be processed
with different nodes. Let some set of flow records contain a security incident that can be detected
using some detection method. Then, there exists a minimal subset of flow records
with semantic relations that must be processed by this detection method together
to get a correct result. In order to find the semantic relations in flow data, a set
of detection methods was studied. The aim is to find a suitable set of information
that is used as an input for hash function. Studied Detection Methods – Vertical SYN scanning can be detected using a threshold-based method pub-
lished in [14]. To successfully detect this type of scanning, the method needs
to receive all flow records of the same source IP address which is a possi-
ble attacker (scanner). Similar method can be used to detect horizontal SYN
scanning. Source IP address is used for hashing. – Brute-force password guessing against remote management services (SSH,
TELNET, RDP etc.) can be detected using a method which needs to inspect
all flow records between two IP addresses in both directions. An ordered pair
of source and destination IP addresses (i.e. bi-flow) is used for hashing. (
)
– There are many public lists of malicious addresses (black-lists). These
addresses were abused due to various reasons like sharing malware, controlling
botnets or acting in some anomalous evil way. Communication with a black-
listed IP address can indicate some malware infection and thus it should be
reported. The detection is quite easy—every time any blacklisted address
appears in a flow record, an alert can be sent. This type of detection is very
efficient with a scattered data, because just a single flow record is needed to
trigger an alert. Source IP address is used for hashing. – More complex method based on statistics about IP addresses and matching
the rules describing malicious traffic is able to detect DoS, DNS amplifica-
tion, SSH brute-force password guessing and horizontal scanning. It needs to
receive all flow records with the same IP address regardless of whether it is a
source or a destination IP. Therefore, hashing both source and destination IP – More complex method based on statistics about IP addresses and matching
the rules describing malicious traffic is able to detect DoS, DNS amplifica-
tion, SSH brute-force password guessing and horizontal scanning. It needs to
receive all flow records with the same IP address regardless of whether it is a
source or a destination IP. Therefore, hashing both source and destination IP Scalable Approach of Detection 9 addresses separately is needed in this case, which can result in duplication. The flow record can be forwarded to two different computational nodes. The
duplication effect will be discussed later in this section. – One of the detection methods based on application layer can detect brute-
force attacks and scanning of user accounts on a Session Initiation Protocol
(SIP) device. Studied Detection Methods The detection method analyzes SIP responses from the server
so all flows with the same source IP address must be delivered to the same
node. Source IP address is used for hashing. In general, we have recognized three groups of detection algorithms, whereas
each group has to process all flow data with the same characteristic (e.g. same
source IP address) on a single computation node. Therefore, we have a group of
detection algorithms expecting all flow records with the same source address,
a group expecting flow records with the same destination address and a group
expecting flow records with the same ordered pair of source and destina-
tion addresses. Figure 3 shows all three hash functions of the flow scatter where
each hash function has the same color as the corresponding group of detection
algorithms. Fig. 3. Flow scatter contains three hash functions, each uses a specific information
from flow records. The result of a hash function determines the computational node
that processes the flow record with corresponding group of detection methods. (Color
figure online) Fig. 3. Flow scatter contains three hash functions, each uses a specific information
from flow records. The result of a hash function determines the computational node
that processes the flow record with corresponding group of detection methods. (Color
figure online) Since we want to run all detection algorithms in parallel, all three hash
functions must be computed for every flow record. Naturally, results of the three
hash functions can be different. Therefore, one flow record can be sent to at least 10 M. ˇSvepeˇs and T. ˇCejka one and, in the worst case, up to three computational nodes. This duplication
is caused by the number of different groups of algorithms and it is needed to
provide all flow records that should be processed together to the algorithms (to
preserve correct detection results). )
In fact, the number of duplicates does not affect overall scaling of the paral-
lel processing i.e. higher number of computational nodes does not increase the
duplicates. Moreover, each hashing function determines a computational node,
which processes the flow record with corresponding group of detection meth-
ods. Therefore, each group processes the flow record only on one computational
node and every flow record is processed by all groups of detection methods. Studied Detection Methods For example, if the selected nodes are 2 (for the SRC IP red hash) and 5 (for
the DST IP yellow hash and for the IP pair green hash), it is processed by red
group on node 2, yellow and green group on node 5. It means, that flow record
may be duplicated, but only on a communication level between flow scatter and
computational nodes. To compare our approach with a single hashing function e.g. NIDS cluster [7]
uses hash = md5(srcIP + dstIP), we can show, that it would not work for us. Let’s take methods for detection of horizontal port scanning and brute-force
password guessing discussed in Sect. 3.3. The method for brute-force password
guessing needs to see all flow records between source and destination IP addresses
in both directions, so this hash function would work (md5(A + B) is equal md5(B
+ A)). On the other hand, horizontal scanning has the same source IP address
but different destination addresses, so it is possible that two flow records with
the same source IP but different destination IP could end up on a different
computational node. Our approach with multiple hash functions can be applied on arbitrary detec-
tion method. To do so, it is necessary to determine characteristics of needed flow
data for correct detection result, as it was done in Sect. 3.3. 4
Experiments and Evaluation In order to evaluate all important aspects of the scattering methods (unifor-
mity, speed, impact on detection), the NEMEA system was chosen as a platform
for our experiments and evaluation. The system itself has already implemented
detection methods, which were studied and discussed in Sect. 3.3 and its efficient
libraries allow us to process traffic from high speed backbone network. Overall,
we have processed over 5 billions of flow records of real data traces in 10 dif-
ferent (pseudonymised) data sets captured in CESNET2 NREN during August
and September 2016 with on average of 60,000 flows/s. 4.1
Testing Environment For our experiments we used a virtual machine with 64b Scientific Linux 7 OS,
with the following hardware specification: 16 CPU cores, 24 GB RAM, 2 TB free
disk capacity. Scalable Approach of Detection 11 Figure 4 shows the configuration of our testing environment. IPFIXsend and
IPFIXcol [15] were used for replaying the IPFIX data in real-time. The flow
data were received by the flow scatter and also directly by the node 0 which
was used as a reference single instance (it processes all flow data without any
splitting). The node 0 was a ground truth for us to evaluate an impact of data
splitting on detection results. The flow scatter distributes flow data between
nodes 1–8 as it was described earlier. All nodes contain exactly the same set
of detection methods. During the experiments, we have collected data from 8
exporting probes that monitor different lines. Fig. 4. Testing environment for experiments and evaluation of various methods of flow
data distribution between computational nodes running flow-based NIDS NEMEA. Fig. 4. Testing environment for experiments and evaluation of various methods of flow
data distribution between computational nodes running flow-based NIDS NEMEA. 1 We expect that such unbalanced distribution based on observation points can be
observed in every network with lines of different bandwidth. 4.2
Results The detection results from all nodes were stored and the analysis is described in
this section. Figure 5 shows a comparison of an average distribution of flow records based
on various scattering methods. The optimal value (red dashed line) is 12.5% for 8
nodes. Random scattering achieves optimal results because of the used statistical
uniform distribution. However, hash-based scattering is not significantly worse
than the random (i.e. optimal) one. On the other hand, link-wise scattering is
unbalanced because of different speeds of the monitored lines and the volume
of traffic1. Node 1 even processed no data because there were no data exported 1 We expect that such unbalanced distribution based on observation points can be
observed in every network with lines of different bandwidth. M. ˇSvepeˇs and T. ˇCejka 12 from the first line. For hash-based scattering, we have compared data distribu-
tion using two different hashing algorithms—CRC32 and Jenkins. On average,
CRC32 had better results and therefore it was chosen as a final solution. Fig. 5. Comparison of average flow records distributions using various scattering meth-
ods. (Color figure online) Fig. 5. Comparison of average flow records distributions using various scattering meth-
ods. (Color figure online) To analyze the reported alerts, we needed to compare the set of unique events
from the reference node 0 with the set of unique events from all distributed
nodes. To achieve this, the reported events of each detection method and each
node were analyzed separately at first. Subsequently, the unique events were
merged together. For example, in the case of horizontal scanning, if an attacker
probes 50 or more computers in two different subnets, where 50 is a threshold
of the detection algorithm, 2 events should be reported. Hash-based scattering
delivers all flow records representing this traffic to the same node due to the
source IP address hashing. Using link-wise scattering, the flow records could end
up on different nodes because the traffic can go through different lines. Random
scattering will split the flow records randomly. Figure 6 shows a comparison of the detection results after applying various
scattering methods. Note, that Hoststatsnemea in the figure legend stands for the
method based on statistics about IP addresses, which was discussed in Sect. 3.3. 4.2
Results The first column represents the reference instance with 100% reported events,
whereas each type of events has a different color and it is normalized so that
the number of different event types are represented equally. Random distribu-
tion (the second column) has a huge impact on the detection results because
of breaking the semantic relations between flow records. This was an expected
result, however, the random distribution is a reference of optimal flow data dis-
tribution. Scattering based on the network topology (the third column) caused Scalable Approach of Detection 13 Fig. 6. Comparison of the detection results after applying various scattering methods. Each part of column with different color stands for normalized number of unique events
reported by different detection method. (Color figure online) Fig. 6. Comparison of the detection results after applying various scattering methods. Each part of column with different color stands for normalized number of unique events
reported by different detection method. (Color figure online) that some of distributed attacks and, in general, N:1 or 1:N attacks (DDoS,
horizontal scanning etc.) were not detected. The last column shows that scat-
tering based on hashing specific information from flow data has the best results. The reason of undetected events is probably a periodic clean-up of structures
containing information and timing of stream-wise detection algorithms. After the evaluation of the uniformity and the impact on the detection, we
tested a maximal throughput of the hash-based flow scatter as the best method
for distribution. A simple NEMEA module was created to generate and send
100 million flow records at full speed to the flow scatter. Measured computation
time was focused on the main cycle receiving the flow record, hashing, making
decision about number of computational nodes the flow belongs to according to
the computed hashes and sending the flow record. The maximal throughput was
on average 1.8 million flow records per second. 5
Conclusion This paper presented the results of practical experiments with different
approaches of splitting a stream of network flow data for the purposes of parallel
anomaly detection. The aim of our work was to compare not only a uniformity
of distribution but also an impact of data splitting on the detection results. Our experiments were performed using real traffic traces from Czech national
research and education network (NREN). For simulation of parallel processing,
we used an open source detection system NEMEA, however, the analysis results M. ˇSvepeˇs and T. ˇCejka 14 are general enough and we believe that the proposed distribution approach can
be used with any other detection system. We have recognized three groups of detection algorithms with different
requirements on data. Therefore, we have designed a flow scatter that uses three
different hashing specific information from flow records (source address, desti-
nation address, ordered pair of source and destination address) to provide all
needed data to independent computational nodes. The results of our experi-
ment show that our approach preserves semantic relations in flow data that are
important for different groups of detection algorithms and therefore the results
of parallel detection are similar to reference results without splitting the data. With the proposed approach of flow data distribution, it is possible to use
detection methods that are deployed on a single machine and run them in parallel
without changes. As a future work, we want to make more experiments with
scaling beyond the measured throughput of the flow scatter by using multiple
flow scatters in parallel and distribute incoming flow records between the flow
scatters with e.g. round robin. Acknowledgment. This work was supported by the Technology Agency of the Czech
Republic under No. TA04010062 Technology for processing and analysis of network
data in big data concept, grant No. SGS17/212/OHK3/3T/18 funded by MEYS and
the project Reg. No. CZ.02.1.01/0.0/0.0/16 013/0001797 co-funded by the MEYS and
ERDF. References 1. Munz, G., Carle, G.: Real-time analysis of flow data for network attack detec-
tion. In: 2007 10th IFIP/IEEE International Symposium on Integrated Network
Management, pp. 100–108, May 2007. doi:10.1109/INM.2007.374774 /
2. Cejka, T., Bartos, V., Svepes, M., Rosa, Z., Kubatova, H.: NEMEA: a framework
for network traffic analysis. In: 2016 12th International Conference on Network and
Service Management (CNSM), pp. 195–201, October 2016. doi:10.1109/CNSM. 2016.7818417 3. Xinidis, K., Charitakis, I., Antonatos, S., Anagnostakis, K.G., Markatos, E.P.: An
active splitter architecture for intrusion detection and prevention. IEEE Trans. Dependable Secure Comput. 3(1), 31–44 (2006). doi:10.1109/TDSC.2006.6 ( )
(
)
/
4. Roesch, M.: Snort - lightweight intrusion detection for networks. In: Proceedings
of the 13th USENIX Conference on System Administration, LISA 1999, Berkeley,
CA, USA, pp. 229–238. USENIX Association (1999) 5. Sallay, H., Alshalfan, K.A., Fred, O.B., Words, K.: A scalable distributed IDS
architecture for high speed networks. IJCSNS Int. J. Comput. SciNetw. Secur. 9(8), 9–16 (2009) 6. Kim, N.-U., Jung, S.-M., Chung, T.-M.: An efficient hash-based load balancing
scheme to support parallel NIDS. In: Murgante, B., Gervasi, O., Iglesias, A., Taniar,
D., Apduhan, B.O. (eds.) ICCSA 2011. LNCS, vol. 6782, pp. 537–549. Springer,
Heidelberg (2011). doi:10.1007/978-3-642-21928-3 39 15 Scalable Approach of Detection 7. Vallentin, M., Sommer, R., Lee, J., Leres, C., Paxson, V., Tierney, B.: The NIDS
cluster: scalable, stateful network intrusion detection on commodity hardware. In:
Kruegel, C., Lippmann, R., Clark, A. (eds.) RAID 2007. LNCS, vol. 4637, pp. 107–126. Springer, Heidelberg (2007). doi:10.1007/978-3-540-74320-0 6 8. Paxson, V.: Bro: a system for detecting network intruders in real-time. Comput. Netw. 31(23–24), 2435–2463 (1999). doi:10.1016/S1389-1286(99)00112-7 (
)
(
)
9. Apache: Hadoop. http://hadoop.apache.org 10. Apache: Spark. http://spark.apache.org 11. Fontugne, R., Mazel, J., Fukuda, K.: Hashdoop: a MapReduce framework for
network anomaly detection. In: IEEE Conference on Computer Communications
Workshops (INFOCOM) (2014). doi:10.1109/INFCOMW.2014.6849281 (
) (
)
/
12. Mai, J., Sridharan, A., Chuah, C.N., Zang, H., Ye, T.: Impact of packet sampling
on portscan detection. IEEE J. Sel. Areas Commun. 24(12), 2285–2298 (2006). doi:10.1109/JSAC.2006.884027 13. Bartos, K., Rehak, M.: Towards efficient flow sampling technique for anom-
aly detection. In: Pescap`e, A., Salgarelli, L., Dimitropoulos, X. (eds.) TMA
2012. LNCS, vol. 7189, pp. 93–106. Springer, Heidelberg (2012). doi:10.1007/
978-3-642-28534-9 11 14. Cejka, T., Svepes, M.: Analysis of vertical scans discovered by naive detection. In:
Badonnel, R., Koch, R., Pras, A., Draˇsar, M., Stiller, B. (eds.) AIMS 2016. LNCS,
vol. 9701, pp. References 165–169. Springer, Cham (2016). doi:10.1007/978-3-319-39814-3 19 15. Velan, P., Krejˇc´ı, R.: Flow information storage assessment using IPFIXcol. In:
Sadre, R., Novotn´y, J., ˇCeleda, P., Waldburger, M., Stiller, B. (eds.) AIMS
2012. LNCS, vol. 7279, pp. 155–158. Springer, Heidelberg (2012). doi:10.1007/
978-3-642-30633-4 21 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. A Blockchain-Based Architecture
for Collaborative DDoS Mitigation
with Smart Contracts Bruno Rodrigues1(B), Thomas Bocek1, Andri Lareida1, David Hausheer2,
Sina Rafati1, and Burkhard Stiller1 1 Communication Systems Group (CSG), Department of Informatics (IfI),
University of Z¨urich (UZH), Z¨urich, Switzerland
{rodrigues,bocek,lareida,rafati,stiller}@ifi.uzh.ch
2 P2P Systems Engineering Lab,
Department of Electrical Engineering and Information Technology,
TU Darmstadt, Darmstadt, Germany
hausheer@ps.tu-darmstadt.de 1 Communication Systems Group (CSG), Department of Informatics (IfI),
University of Z¨urich (UZH), Z¨urich, Switzerland
{rodrigues,bocek,lareida,rafati,stiller}@ifi.uzh.ch
2 P2P Systems Engineering Lab,
Department of Electrical Engineering and Information Technology,
TU Darmstadt, Darmstadt, Germany
hausheer@ps.tu-darmstadt.de 1 Communication Systems Group (CSG), Department of Informatics (IfI),
University of Z¨urich (UZH), Z¨urich, Switzerland
{rodrigues,bocek,lareida,rafati,stiller}@ifi.uzh.ch
2 P2P Systems Engineering Lab, Department of Electrical Engineering and Information Technology,
TU Darmstadt, Darmstadt, Germany
hausheer@ps.tu-darmstadt.de Abstract. The rapid growth in the number of insecure portable and
stationary devices and the exponential increase of traffic volume makes
Distributed Denial-of-Service (DDoS) attacks a top security threat to
services provisioning. Existing defense mechanisms lack resources and
flexibility to cope with attacks by themselves, and by utilizing other’s
companies resources, the burden of the mitigation can be shared. Emerg-
ing technologies such as blockchain and smart contracts allows for the
sharing of attack information in a fully distributed and automated fash-
ion. In this paper, the design of a novel architecture is proposed by
combining these technologies introducing new opportunities for flexible
and efficient DDoS mitigation solutions across multiple domains. Main
advantages are the deployment of an already existing public and distrib-
uted infrastructure to advertise white or blacklisted IP addresses, and
the usage of such infrastructure as an additional security mechanism to
existing DDoS defense systems, without the need to build specialized reg-
istries or other distribution mechanisms, which enables the enforcement
of rules across multiple domains. Keywords: Distributed
Denial-of-Service
(DDoS) ·
Security ·
Blockchain · Software-defined Networks (SDN) · Network management c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 16–29, 2017.
DOI: 10.1007/978-3-319-60774-0 2 1
Introduction In the past years, a rise in DDoS attacks could be observed [1]. DDoS attacks
have the simple goal of interrupting or suspending services available on the Inter-
net and its motivations range from personal grudges over blackmail to political
reasons [10]. A recent example is an attack conducted against Domain Name
System (DNS) servers responsible for domains such as Twitter, PayPal, and Blockchain-Based Architecture for Collaborative DDoS Mitigation 17 Spotify [20] in October 2016. As a consequence, those services became unavail-
able to many US (United States) users for several hours. Besides the frequency,
also the strength and duration of DDoS attacks are growing making them more
efficient and dangerous. One reason for the increasing size of attacks is the avail-
ability of many reflectors, and i.e., weakly secured or configured IoT (Internet
of Things) devices or home gateways [20]. By exploiting legal services on those devices, e.g., the Simple Service Dis-
covery Protocol, the power of a DDoS attack is amplified, and the problem
of defense is made more complicated. Thus, the impact of DDoS varies from
minor inconvenience to severe financial losses for enterprises that rely on their
online availability [14]. Various mitigation techniques have been proposed. How-
ever, only a few have been considered for widespread deployment because of
their effectiveness and implementation complexities. An ongoing IETF (Internet
Engineering Task Force) proposal discusses the development of a collaborative
protocol called DOTS (DDoS Open Threat Signaling) to advertise DDoS attacks
[13]. However, this paper proposes an infrastructure of blockchains and smart
contracts, which provide the required instrumentation without the need to main-
tain design and development complexities of such a new protocol. As with a different direction, the adoption of DDoS protection services,
offered by companies such as Akamai [1] or CloudFlare [3], is increasing [7]. Those cloud-based solutions can absorb DDoS attacks by increasing capacity and
taking the burden of detection away from the device under attack by exporting
flow records from edge routers and switches. Additional analysis is performed
in the cloud and packet filtering is used to balance, reroute, or drop the traffic
inside the cloud. However, those solutions requires a third party DDoS Protec-
tion Service (DPS) provider, which is implying in additional costs and a decrease
in service performance. B. Rodrigues et al. B. Rodrigues et al. Software-defined Network (SDN) is an effective solution to enable customiz-
able security policies and services in a dynamic fashion. The centralized network
control and its deployment based on the OpenFlow [11] protocol facilitates the
enforcement of high-level security policies moving away from current approaches
based on SNMP (Simple Network Management Protocol) and CLI (Command
Line Interface). With SDN, flow-rules can be applied to block DDoS attacks, and
the closer these rules are applied, and those malicious packets can be dropped,
the less DDoS traffic occurs. This work uses SDN-based networks as a use case
to perform in a more rapid fashion in ASes the definition and verification of flows
to mitigate DDoS attacks. However, the presented solution is not limited to the
usage of an SDN-based network, being compatible with detection/monitoring
tools able to export attack information to be published in the blockchain. This paper is structured as follows. Section 2 introduces basic concepts and
related work on blockchain and smart contracts. Section 3 presents related collab-
orative DDoS mitigation strategies. Section 4 presents the architecture detailing
its components and basic functioning, as well as describing the implementation
details of the proposed solution. Section 5 provides a discussion on the develop-
ment and results obtained so far. The work is concluded in Sect. 6 highlighting
the significant contributions and discussing future work. 1
Introduction This paper presents the architecture and design of a collaborative mecha-
nism using smart contracts and investigates the possibility of mitigating a DDoS
attack in a fully decentralized manner. Thus, service providers interested in
shared protection, can not only signal the occurrence of attacks but also share
detection and mitigation mechanisms. The objective is to create an automated,
and easy-to-manage DDoS mitigation. Three major building blocks are identified
to build such a mechanism. Blockchains and Smart Contracts. This approach proposes an architecture
and an implementation of an approach to signaling white or blacklisted IP
addresses across multiple domains based on blockchains and smart contracts. The advantage of using smart contracts in a blockchain is: (a) to make use of
an already existing infrastructure to distribute rules without the need to build
specialized registries or other distribution mechanisms/protocols, (b) to apply
rules across multiple domains, which means that even if the AS (Autonomous
System) of the victim is not applying these rules, some traffic can still be filtered,
and (c) the victim or its AS can control which customers get blocked. The only
central element remaining is to show proof of IP ownership. 18 2
Background In addition to SDN,
the OpenFlow protocol [11] leverages network management by providing a pro-
grammable and standardized interface between the data plane and the control
plane. It has been recognized that the decoupling of the data plane and the
control plane makes SDN a promising solution to enable the enforcement of cus-
tomizable security services and policies. Various SDN-based solutions have been
proposed to deal with DDoS attacks [24]. A survey on these issues is provided in
[17]. However, each security/concern category can be sub-divided in fine-grained
aspects e.g., authentication, integrity, network communications. In the following
are presented mainly research efforts addressing DDoS attacks in SDN networks. To analyze the impact of DDoS attacks on network performance, the works in
[18] and [8] have shown how such attacks may impact on several parameters like
the control plane bandwidth (i.e., controller-switch channel), latency, switches
flow tables and the controller performance. Other works as [22] and [4] use the
SDN capabilities to implement schemes that allow to detect and mitigate DDoS
attacks through packet analysis and filtering. These solutions reduce the impact
of attacks, but they may cause an overhead in the flow-tables and the SDN con-
troller. Also, they do not provide any solution to address these particular SDN
performance issues as proposed in [5] (e.g., flow-tables, and controller overload-
ing). Furthermore, they also do not consider DDoS attacks and the collaboration
with AS customers as [16]. To analyze the impact of DDoS attacks on network performance, the works in
[18] and [8] have shown how such attacks may impact on several parameters like
the control plane bandwidth (i.e., controller-switch channel), latency, switches
flow tables and the controller performance. Other works as [22] and [4] use the
SDN capabilities to implement schemes that allow to detect and mitigate DDoS
attacks through packet analysis and filtering. These solutions reduce the impact
of attacks, but they may cause an overhead in the flow-tables and the SDN con-
troller. Also, they do not provide any solution to address these particular SDN
performance issues as proposed in [5] (e.g., flow-tables, and controller overload-
ing). Furthermore, they also do not consider DDoS attacks and the collaboration
with AS customers as [16]. SDN-based solutions allow greater agility to enforce decisions that require a
global network view. Therefore, intra-domain security policies and mechanisms
to prevent and react to DDoS attacks can be made agiler. 2
Background By combining the
intra-domain capabilities provided by SDN and the inter-domain advantages
provided by blockchains and smart contracts, the efficiency to mitigate DDoS
attacks in both inter- and intra-domains can be improved. 2
Background Smart contracts are a piece of software made to facilitate the negotiation or
performance of a contract, being able to be executed, verified or enforced on
its own. A smart contract alone is not ”smart” as it needs an infrastructure
that can implement, verify, and enforce the negotiation or performance of a
contract by particular computer protocols. It has gained attention in the context
of blockchains that provide a fully decentralized infrastructure to run, execute,
and verify such smart contracts [2]. Therefore, smart contracts need to run on a
blockchain to ensure (a) its permanent storage and (b) obstacles to manipulate
the contract?s content. A node participating in the blockchain runs a smart
contract by executing its script, validating the result of the script, and storing
the contract and its result in a block. Although the Bitcoin [12] blockchain was the first fully decentralized dis-
tributed ledger, it is primarily designed for transfer of digital assets, and it is
not Turing-complete (e.g., it does not support loops). Such a Turing-complete
contract language allows defining rules to allow or block IP addresses that can
be interpreted by an SDN controller. While several projects try to address these
issues, the Ethereum [23] blockchain is the most popular that supports a Turing-
complete contract language, empowering more sophisticated smart contracts. In Ethereum, smart contracts run in a sand-boxed Ethereum Virtual Machine
(EVM) and every operation executed in the EVM has to be paid for to prevent
Denial-of-Service (DoS) attacks. SDN characteristics provide better network visibility by decoupling the con-
trol plane from the data plane and by the centralized management to perform Blockchain-Based Architecture for Collaborative DDoS Mitigation 19 tasks such as network diagnosis and troubleshooting [9]. In addition to SDN,
the OpenFlow protocol [11] leverages network management by providing a pro-
grammable and standardized interface between the data plane and the control
plane. It has been recognized that the decoupling of the data plane and the
control plane makes SDN a promising solution to enable the enforcement of cus-
tomizable security services and policies. Various SDN-based solutions have been
proposed to deal with DDoS attacks [24]. A survey on these issues is provided in
[17]. However, each security/concern category can be sub-divided in fine-grained
aspects e.g., authentication, integrity, network communications. In the following
are presented mainly research efforts addressing DDoS attacks in SDN networks. tasks such as network diagnosis and troubleshooting [9]. B. Rodrigues et al. B. Rodrigues et al. Among the collaborative DDoS mitigation techniques, there are two main
approaches using resource management to react against bandwidth attacks [14]. The first takes effect within the victim’s domain and the second within the
domain of the victims ISP, i.e. the AS. Both techniques apply traffic classifica-
tion and define specific actions for those classes. Both customer and AS resource
management schemes need to classify traffic into several types, and then treat
them differently. However, it is rather difficult to give an accurate classification
as DDoS attacks can mimic any legitimate traffic. In this regard, some sophis-
ticated techniques can be implemented to classify traffic, but a unified reaction
strategies implemented both at the AS and the customer can be more efficient
than applying just one. Other works exist for cooperative defense against DDoS attacks. However,
it is still an open issue since DDoS attacks are growing in scale, sophistica-
tion, duration and frequency [10]. The IETF is currently proposing a pro-
tocol [13] called DOTS (DDoS Open Threat Signaling) covering both intra-
organization and inter-organization communications to advertise attacks. The
protocol requires servers and clients DOTS agents, which can be organized in
both centralized and distributed architectures to advertise black or whitelisted
addresses. A DOTS client should register to a DOTS server in advance send-
ing provision and capacity protection information and be advertised of attacks. Then, the DOTS protocol is used among the agents to facilitate and coordinate
the DDoS protection service as a whole. Also, a similar approach to the IETF
proposal is presented in [19]. The authors use a similar architecture but using
an advertising protocol based on FLEX (FLow-based Event eXchange) format,
which is used to simplify the integration and deployment of the solution and
facilitate the communication process between the involved domains. The proposed standard advertises the need for defensive measures in antici-
pation of or response to attack. The main drawback compared to the approach
presented herein is the requirement of additional infrastructure requiring trust
and collaboration between ISPs. A collaborative defense approach using VNF
(Virtual Network Functions) is presented in [15]. The authors propose a coop-
eration between domains that implements VNFs to alleviate DDoS attacks by
redirecting and reshaping excessive traffic to other collaborating domains for fil-
tering. 3
Related Work There are four broad categories of defense against DDoS attacks according to
[14]: (1) attack prevention, (2) attack detection, (3) attack source identification,
and (4) attack reaction. (1) Tries to prevent attacks before they become a problem, i.e. as close to the
sources as possible. The obvious method to achieve this for amplified or
reflected attacks is for the access provider to filter spoofed packets; (1) Tries to prevent attacks before they become a problem, i.e. as close to the
sources as possible. The obvious method to achieve this for amplified or
reflected attacks is for the access provider to filter spoofed packets; (2) Can be a difficult task since certain attacks mask themselves as legitimate
user traffic or use various traffic types. Due to this complexity, it can be
hard to make a confident decision if traffic is part of an attack or special
user behavior, e.g. a flash crowd; (3) Is applied after an attack was detected. This step is important to efficiently
contain or re-route the attack as close to its source as possible; (4) The final step involves taking concrete measures against the attack. The
better the result from (3) the more efficiently this can be done. (4) The final step involves taking concrete measures against the attack. The
better the result from (3) the more efficiently this can be done. 20 4
Proposed System Architecture This section presents the design principles considered in the architectural design. First, Sect. 4.1 exemplifies a deployment scenario. Section 4.2 provides a detailed
description of its main components. Implementation details are presented in
Sect. 4.3. B. Rodrigues et al. In [24], a gossip-based communication mechanism is proposed to exchange
information about attacks between independent detection points to aggregate
information about the overall observed attacks. The system is built as a peer-to-
peer overlay network to disseminate attack information to other listening users
or systems rapidly. A similar approach was presented in [21], formalizing a gossip-based pro-
tocol to exchange information in overlay network using intermediate network
routers. A different approach is presented in [16], which proposes a collaborative
framework that allows the customers to request DDoS mitigation from ASes. However, the solution requires an SDN controller implemented at customer side
interfaced with the AS, which can change the label of the anomalous traffic and
redirect them to security middle-boxes. In the approach presented in this paper Blockchain-Based Architecture for Collaborative DDoS Mitigation 21 customers and ISPs can take action to mitigate an attack by interfacing directly
with a blockchain providing the necessary trust. Instead of making use of an existing infrastructure such as the blockchain
and smart contracts, approaches mentioned above proposes the development of
specific gossip-based protocols. In this sense, the deployment and integration of
such solutions become complex since existing solutions need to be modified to
support these protocols. The IETF proposal focuses on standardizing a protocol
to facilitate its deployment. However, its implementation complexity still exists
in distributed and centralized architectures to support the different types of
communication. Instead, some of the requirements can be inherited from the
natural characteristics of blockchains, smart contracts, and SDN, avoiding the
complexities of development and adoption of new protocols. 4.1
Application Scenario A scenario is presented in Fig. 1 illustrating the system architecture. A web server
hosted at AS C is under a DDoS attack from devices hosted at various domains
(ASes A, B, and C). With a non-collaborative DDoS mitigation approach, the
web server relies on defense mechanisms that are implemented at the AS where
it is allocated, which in many cases may be distant from the origin of the attack
traffic and therefore overloading several domains with attack traffic. Participants of the collaborative defense (ASes and customers) first need to
create a smart contract, that is promptly linked with a registry-based type of
smart contracts. Therefore, when attackers overload web server, the customer or
the AS under attack stores the IP addresses of attackers in the smart contract. In an Ethereum blockchain a new block is created every 14 s, so subscribed ASes
will receive updated lists of addresses to be blocked and confirm the authenticity
of the attack by analyzing the traffic statistics and verifying the authenticity of
the target’s address. Once other ASes retrieve the list of attackers and confirm the attack, differ-
ent mitigation strategies can be triggered according to the security policies and
mechanisms available in the domain. Also, it can block malicious traffic near of
its origin. Near-source, defense is ideal for the health of the Internet because
it can reduce the total cost of forwarding packets which, in the case of DDoS
attacks mostly consist of useless massive attack traffic [13]. In scenarios involving multiple domains, once collaborative defense nodes
receive information about attacks, these can apply mitigation actions in agree-
ment with their security policies. In this sense, an incentive mechanism is nec-
essary to prevent domains from abusing cooperative defense. B. Rodrigues et al. 22 Fig. 1. Application scenario Fig. 1. Application scenario Blockchain-Based Architecture for Collaborative DDoS Mitigation Blockchain-Based Architecture for Collaborative DDoS Mitigation 23 Fig. 2. Proposed system architecture Fig. 2. Proposed system architecture (3) To efficiently aid coordinated attack responses, Blockchain DDoS Mitiga-
tion modules are running on the entities (customers or ASes) reporting IP
addresses and listening to the blockchain;
( ) (3) To efficiently aid coordinated attack responses, Blockchain DDoS Mitiga-
tion modules are running on the entities (customers or ASes) reporting IP
addresses and listening to the blockchain; (3) To efficiently aid coordinated attack responses, Blockchain DDoS Mitiga-
tion modules are running on the entities (customers or ASes) reporting IP
addresses and listening to the blockchain;
( ) (3) To efficiently aid coordinated attack responses, Blockchain DDoS Mitiga-
tion modules are running on the entities (customers or ASes) reporting IP
addresses and listening to the blockchain; (4) Only customers or ASes with proof of ownership of their IP may report
addresses to the smart contract; (4) Only customers or ASes with proof of ownership of their IP may report
addresses to the smart contract; (5) Different domains implement different security policies as well as different
underlying management systems. Once notified of a DDoS attack in which
the customer has its authenticity confirmed, countermeasures are defined
according to the domain security policies and available actions. To mitigate DDoS attacks (1) different techniques can be used upon the
detection by ASes or customers, which typically involves analyzing Internet traf-
fic with sophisticated attack detection algorithms, followed by filtering. In this
regard, a collaborative approach decreases the overhead of such algorithm in the
detection phase using information from other domains. Blockchain DDoS Miti-
gation appliances (2) both on the customer and ASes are simpler as Ethereum
is public and already available technology, which can be used to perform rapid
and widespread DDoS advertisement using smart contracts. Services with chal-
lenge/response authentication can be utilized by an AS to ensure that the IP
address (3) of the customer reporting the attack is the customer under attack,
and to enforce the necessary countermeasures (4) by the security policies imple-
mented in the domain. The smart contract logic illustrated in Fig. 3 is deployed as a complementary
solution to existing DDoS mitigation mechanisms. However, domains implement-
ing the system should consider the principles mentioned above in its design. 4.2
Architectural Design As DDoS attacks continue to increase and vary in their patterns, the need
for coordinated responses also increases to detour the attacks efficiently. How-
ever, it is important to note that only the collaboration between customers and
ASes is an additional approach to existing defense mechanisms. The architecture
depicted in Fig. 2 is composed of three components: – Customers: may report white or blacklisted IP addresses to the Ethereum
blockchain via smart contracts; – Customers: may report white or blacklisted IP addresses to the Ethereum
blockchain via smart contracts; – ASes: may publish white or blacklisted IP addresses and retrieve lists contain-
ing the published IP addresses, and may implement their DDoS mitigation
mechanisms; – Blockchain/Smart Contract: the public Ethereum blockchain (Ethereum
Virtual Machine nodes) running Solidity smart contracts, which comprises
the logic to report IP addresses in the blockchain. The architecture is built considering the following principles: The architecture is built considering the following principles: (1) DDoS detection and mitigation countermeasures are provided as on-demand
services by either the ASes or third-party services; (2) To report/receive attack information, it is necessary for the domain to ded-
icate a node connected to the blockchain. This can be dedicated hardware
exclusively for this purpose or virtualized to minimize resource consump-
tion; Blockchain-Based Architecture for Collaborative DDoS Mitigation First,
any domain (e.g., customers or ASes) participating must create a smart contract
identified with an IP address or range of addresses certified by an authority. Then, the smart contract is registered in a registry-based type of smart contract
so that participation can be easily tracked and thus relevant smart contracts can
be identified. B. Rodrigues et al. 24 Fig. 3. Proposed system flowchart Fig. 3. Proposed system flowchart Traffic arriving at both the customer and AS can be analyzed and filtered
using existing monitoring tools (e.g., NetFlow, sFlow, custom SDN implemen-
tations). The Blockchain appliance can be deployed as an additional security
feature to any system that implements an apparatus to advertise black or
whitelisted IP addresses to the blockchain. The analysis of traffic in a gateway is
facilitated by SDN, and therefore the approach is intended to use a monitoring
framework based on the OpenFlow protocol. 4.3
Implementation Details Listing 1.1 and 1.2 outlines current implemented features of the smart contract
to store source IP addresses that should be blocked or allowed. For simplicity,
only IPv4 addresses are shown here. Either the customer or the AS can cre-
ate the smart contract. In any case, a certificate of IP ownership is required. For the customer, the certificate can be created with an automated challenge-
response system, while the AS requires a certificate matching their entry in the
AS registration. The one that created the smart contract (owner of the account that created
the contract) can add other addresses that are also allowed to add IPs to block. Before such address is added, it is checked if the address matches its parent
subnet. Both AS and the customer can store src IP with an expiration time. The time is measured in blocks, and the access to the stored data is public and
can be viewed by anyone. Before retrieving a list of IP pairs (source/destination), the verifyIP() func-
tion needs to be called to make sure that the target IP address has a proof of
ownership. The issuing of a certificate (certOwnerIPv4) is the only remaining
central entity in the architecture. After that, any AS (does not need to be the
customers AS) can use these IPs to block traffic on its network. The smart contract needs first to register itself in another smart contract
Registry, which stores all relevant smart contracts that should be watched. Thus
an AS listens for these changes, and any addition can be monitored and assessed
against the network properties of the AS and apply a blocking rule if necessary. Blockchain-Based Architecture for Collaborative DDoS Mitigation Blockchain-Based Architecture for Collaborative DDoS Mitigation Blockchain-Based Architecture for Collaborative DDoS Mitigation 25 Blockchain-Based Architecture for Collaborative DDoS Mitigation
25
1
contract
SDNRulesAS {
2
struct
ReportIPv4 {
3
uint32
expiringBlock ;
4
uint32
src_ip;
5
DstIPv4
dst_ip;
6
} ReportIPv4 []
report_src_ipv4 ;
7
8
struct
DstIPv4 {
9
uint32
dst_ipv4;
10
uint8
dst_mask;
11
} DstIPv4
dstIPv4 ;
12
bytes
certOwnerIPv4 ; address
owner;
13
mapping (address => DstIPv4 ) customerIPv4 ;
14
bool
flag; // Indicate
black
or
whitelisted
addresses
15
function
SDNRulesAS(uint32
dst_ipv4 , uint8
dst_ipv4_mask ,
bytes
_certOwnerIPv4 , bool
_flag) {
16
owner = msg.sender;
17
certOwnerIPv4
= _certOwnerIPv4 ;
18
dstIPv4 = DstIPv4 (dst_ipv4 , dst_ipv4_mask );
19
flag = _flag;
20
// TODO: register
in a registry
contract
21
}
22
// suicide
and
deregistering
function
here
23
function
createCustomerIPv4 (address
customer , uint32
dst_ipv4 ,
uint8
dst_ipv4_mask ) {
24
if(msg.sender
== owner
&&
25
isInSameIPv4Subnet (dst_ipv4 , dst_ipv4_mask )) {
26
customerIPv4 [customer] =
27
DstIPv4 (dst_ipv4 , dst_ipv4_mask );
28
}
29
}
30
function
isInSameIPv4Subnet (uint32
dst_ipv4 , uint8
dst_mask)
constant
returns (bool) {
31
// true if
customer
IP is in same
subnet
32
}
33
}
Listing 1.1. Blockchain-Based Architecture for Collaborative DDoS Mitigation 26 17
function
verifyIP(bytes
pubKey) constant
returns (bool) {
18
// check
if
signature
in
certOwnerIPv4
is
correct
19
}
20
21
function
reportedIPv4 ()
constant
returns (uint32 [] src_ipv4 ,
22
uint32 [] dst_ipv4 , uint8 [] mask) {
23
uint32 []
memory
src; uint32 [] memory
dst; uint8 []
memory
msk;
24
for (uint i = 0; i < report_src_ipv4 .length; i++) {
25
if( drop_src_ipv4 [i]. expiringBlock
> block.number) {
26
src[src.length] = report_src_ipv4 [i]. src_ip;
27
dst[dst.length] = report_src_ipv4 [i]. dst_ip.dst_ipv4;
28
msk[msk.length] = report_src_ipv4 [i]. dst_ip.dst_mask;
29
}
30
}
31
return (src , dst , msk);
32
}
Listing 1.2. Smart contract IP reporting functions 7
function
verifyIP(bytes
pubKey) constant
returns (bool) { 17
function
verifyIP(bytes
pubKey) constant
returns (bool) {
18
// check
if
signature
in
certOwnerIPv4
is
correct
19
}
20
21
function
reportedIPv4 ()
constant
returns (uint32 [] src_ipv4 ,
22
uint32 [] dst_ipv4 , uint8 [] mask) {
23
uint32 []
memory
src; uint32 [] memory
dst; uint8 []
memory
msk;
24
for (uint i = 0; i < report_src_ipv4 .length; i++) {
25
if( drop_src_ipv4 [i]. expiringBlock
> block.number) {
26
src[src.length] = report_src_ipv4 [i]. src_ip;
27
dst[dst.length] = report_src_ipv4 [i]. dst_ip.dst_ipv4;
28
msk[msk.length] = report_src_ipv4 [i]. dst_ip.dst_mask;
29
}
30
}
31
return (src , dst , msk);
32
}
Listing 1.2. Smart contract IP reporting functions 18
// check
if
signature
in
certOwnerIPv4
is
correct Listing 1.2. Smart contract IP reporting functions Listing 1.2. Smart contract IP reporting functions Blockchain-Based Architecture for Collaborative DDoS Mitigation Smart contract structures and core functionality
1
function
reportIPv4(uint32 [] src , uint32
expiringBlock ) {
2
if (msg.sender
== owner) {
3
for (uint i = 0; i < src.length; i++) {
4
drop_src_ipv4 .push( ReportIPv4(
5
expiringBlock , src[i], dstIPv4 ));
6
}
7
}
8
DstIPv4
customer = customerIPv4 [msg.sender ];
9
if(customer.dst_ipv4
!= 0) {
10
for (i = 0; i < src.length; i++) {
11
report_src_ipv4 .push( ReportIPv4(
12
expiringBlock , src[i], customer));
13
}
14
}
15
} 1
contract
SDNRulesAS {
2
struct
ReportIPv4 {
3
uint32
expiringBlock ;
4
uint32
src_ip;
5
DstIPv4
dst_ip;
6
} ReportIPv4 []
report_src_ipv4 ;
7
8
struct
DstIPv4 {
9
uint32
dst_ipv4;
10
uint8
dst_mask;
11
} DstIPv4
dstIPv4 ;
12
bytes
certOwnerIPv4 ; address
owner;
13
mapping (address => DstIPv4 ) customerIPv4 ;
14
bool
flag; // Indicate
black
or
whitelisted
addresses
15
function
SDNRulesAS(uint32
dst_ipv4 , uint8
dst_ipv4_mask ,
bytes
_certOwnerIPv4 , bool
_flag) {
16
owner = msg.sender;
17
certOwnerIPv4
= _certOwnerIPv4 ;
18
dstIPv4 = DstIPv4 (dst_ipv4 , dst_ipv4_mask );
19
flag = _flag;
20
// TODO: register
in a registry
contract
21
}
22
// suicide
and
deregistering
function
here
23
function
createCustomerIPv4 (address
customer , uint32
dst_ipv4 ,
uint8
dst_ipv4_mask ) {
24
if(msg.sender
== owner
&&
25
isInSameIPv4Subnet (dst_ipv4 , dst_ipv4_mask )) {
26
customerIPv4 [customer] =
27
DstIPv4 (dst_ipv4 , dst_ipv4_mask );
28
}
29
}
30
function
isInSameIPv4Subnet (uint32
dst_ipv4 , uint8
dst_mask)
constant
returns (bool) {
31
// true if
customer
IP is in same
subnet
32
}
33
}
Listing 1.1. Smart contract structures and core functionality
1
function
reportIPv4(uint32 [] src , uint32
expiringBlock ) {
2
if (msg.sender
== owner) {
3
for (uint i = 0; i < src.length; i++) {
4
drop_src_ipv4 .push( ReportIPv4(
5
expiringBlock , src[i], dstIPv4 ));
6
}
7
}
8
DstIPv4
customer = customerIPv4 [msg.sender ];
9
if(customer.dst_ipv4
!= 0) {
10
for (i = 0; i < src.length; i++) {
11
report_src_ipv4 .push( ReportIPv4(
12
expiringBlock , src[i], customer));
13
}
14
}
15
} 1
contract
SDNRulesAS { 2
struct
ReportIPv4 { 3
uint32
expiringBlock ; 4
uint32
src_ip; nction
SDNRulesAS(uint32
dst_ipv4 , uint8
dst_ipv4_mask , B. Rodrigues et al. 5
Discussion The use of the Ethereum Virtual Machine (EVM) allows the multiple domains
involved in an attack scenario to invoke functions in a smart contract reporting
attacks or maintaining a list of trusted addresses to be operating in case of attack. The support of white or blacklisted IP addresses is a decision that depends on the
policies and security mechanisms available in each domain. Therefore, the smart
contract was developed to support both lists using a flag indicating which type
of address is being reported. The existing and distributed storage infrastructure
reduces the complexity in the development and adoption of the approach as it
supersedes the design and standardization process of a gossip-based protocol,
which needs to be embraced by the various ASes and customers. Also, the EVM
smart contracts support in a decentralized and native way the logic to control
who is reporting an attack and who are the attackers. Through a high-level comparison with the ongoing IETF proposal (the DOTS
protocol) [13], instead of making use of an existing infrastructure such as the
blockchain and smart contracts, the IETF proposes from scratch the develop-
ment of such protocol with several requirements (e.g., extensibility, resilience) to
be deployed in a distributed architecture. In this sense, the protocol development
becomes complex since it must be deployed in distributed and centralized archi-
tectures to support different types of communication (inter and intra domain,
i.e, inside the domain of an AS and between ASes). Instead, it is argued that
some of the requirements can be inherited from the natural characteristics of
blockchains, smart contracts and SDN, avoiding the complexities of develop-
ment and adoption of new protocols. However, this smart contract works well for a small number of attacks, while
for large-scale attacks, the approach is currently costly the contract size, but will
be addressed this issue in a future work to make reference to a larger list of IP
addresses. Therefore, to keep the complexity of the architecture low, only the Blockchain-Based Architecture for Collaborative DDoS Mitigation 27 data (e.g., IP addresses) should be stored in the contract, and it may become nec-
essary to add a reference as shown in Listing 1.3, where the full list of addresses
can be retrieved. 5
Discussion The cost of adding 50 source IPs directly in a freshly deployed
contract is 2.5 mio gas (gas is the internal pricing for running a transaction or
contract in Ethereum) at the current gas price [6] of 20 gwei, which is 0.05 ETH
at the current market price of 9.3 USD is in total 0.46 USD, while 100 source IPs
cannot be mined in one contract and multiple contracts have to be used as it
exceeds the 4 mio gas limit. 1
struct
ReportIPv4 {
2
uint32
expiringBlock ;
3
uint32
src_ip;
4
//e.g. https :// example.com/blockedips.txt
5
string
src_ipv4_ref;
6
DstIPv4
dst_ip ;}
7
ReportIPv4 [] report_src_ipv4 ;
Listing 1 3 Storing references 1
struct
ReportIPv4 { 2
uint32
expiringBlock ; 3
uint32
src_ip; _ p;
4
//e.g. https :// example.com/blockedips.txt 5
string
src_ipv4_ref; 5
string
src_ipv4_ref; 6
DstIPv4
dst_ip ;} 7
ReportIPv4 [] report_src_ipv4 ; Listing 1.3. Storing references B. Rodrigues et al. Another major factor towards the practicability of the approach is the fair-
ness among the cooperative domains. If an AS is targeted more times than
others, means that one would be using resources of others to protect themselves. Therefore, this relevant aspect will be detailed in a future work to propose a
reputation scheme based on the participation of the domains in the cooperative
architecture. 6
Summary and Future Work This paper proposes a collaborative architecture using smart contracts and
blockchain to enable DDoS mitigation across multiple domains. As a distrib-
uted and primarily public storage, the blockchain determines a straightforward
and efficient structure to develop a collaborative approach toward DDoS attacks
mitigation. The proposed architecture can be deployed as an additional security
mechanism to existing DDoS protection schemes. Therefore, it is not intended
to dictate how security mechanisms and policies should be implemented in a
particular domain. Instead, it can be combined with existing solutions to reduce
the DDoS detection and mitigation overhead by involving multiple domains in
the process. Coupled with current solutions, the DDoS detection and mitigation
overhead process comprising multiple domains can be reduced. The architecture enables ASes to deploy their DPS and generate added value
for their customers without transferring control of their network to a third party. The main contributions of this new approach are summarized as (a) the design
and development of an architecture based on blockchains to advertise DDoS
attacks across multiple domains, (b) the adoption and integration of the app-
roach is facilitated since Ethereum and smart contracts are publicly available,
and the ability to enforce rules on the ASes-side by the use of SDN, (c) can be
utilized as an additional security mechanism without modifying existing ones. Future work will investigate ways to compress the list, e.g., with a bloom fil-
ter, and its advantages and disadvantages. Another limitation is that blocking of
destination IPs should be possible only for static IPs. Thus, automated services
issuing these certificates of IP ownership need to check for dynamic IPs first, e.g.,
using services such as SORBS (dul.dnsbl.sorbs.net). Also, the current smart con-
tract supports only one hierarchy. Thus, createCustomerCertIPv4() in Listing 1.1
needs to be extended to allow more hierarchies to map subnets accordingly. 28 References 1. Akamai: How to Protect Against DDoS Attacks - Stop Denial of Service
(2016). https://www.akamai.com/us/en/resources/protect-against-ddos-attacks. jsp. Accessed 10 Jan 2017 2. Bocek, T., Stiller, B.: Smart Contracts - Blockchains in the Wings, pp. 1–16. Springer, Heidelberg. Tiergartenstr. 17, 69121, January 2017 3. CloudFare: Cloudflare advanced DDoS protection (2016). https://www.cloudflare. com/static/media/pdf/cloudflare-whitepaper-ddos.pdf /
/
/
/
4. Dao, N.N., Park, J., Park, M., Cho, S.: A feasible method to combat against
DDOS attack in SDN network. In: 2015 International Conference on Information
Networking (ICOIN), pp. 309–311, January 2015 5. Dridi, L., Zhani, M.F.: SDN-guard: Dos attacks mitigation in SDN networks. In:
2016 5th IEEE International Conference on Cloud Networking (Cloudnet), pp. 212–217, October 2016 6. Fund, E: Ether unit converter. http://ether.fund/tool/converter 7. Jonker, M., Sperotto, A., van Rijswijk-Deij, R., Sadre, R., Pras, A.: Measuring the
adoption of DDoS protection services. In: Proceedings of the 2016 ACM on Internet
Measurement Conference, IMC 2016, Santa Monica, California, USA (2016) 8. Kandoi, R., Antikainen, M.: Denial-of-service attacks in openflow SDN networks. In: 2015 IFIP/IEEE International Symposium on Integrated Network Management
(IM), pp. 1322–1326. IEEE (2015) 9. Kreutz, D., Ramos, F.M., Verissimo, P.E., Rothenberg, C.E., Azodolmolky, S.,
Uhlig, S.: Software-defined networking: a comprehensive survey. Proc. IEEE
103(1), 14–76 (2015) ( )
(
)
10. Mansfield-Devine, S.: The growth and evolution of DDoS. Netw. Secur. 10, 13–20
(2015) (
)
11. McKeown, N., Anderson, T., Balakrishnan, H., Parulkar, G., Peterson, L., Rexford,
J., Shenker, S., Turner, J.: Openflow: enabling innovation in campus networks. ACM SIGCOMM Comput. Communi. Rev. 38(2), 69–74 (2008) 12. Nakamoto, S.: Bitcoin: a peer-to-peer electronic cash system (2008) 13. Nishizuka, K., Xia, L., Xia, J., Zhang, D., Fang, L., Gray, C.: Inter-organization
cooperative DDOS protection mechanism. Draft. https://tools.ietf.org/html/
draft-nishizuka-dots-inter-domain-mechanism-02 14. Peng, T., Leckie, C., Ramamohanarao, K.: Survey of network-based defense mech-
anisms countering the DoS and DDOS problems. ACM Comput. Surv. (CSUR)
39(1), 3 (2007) 15. Rashidi, B., Fung, C.: CoFence: a collaborative DDOS defence using network func-
tion virtualization. In: 12th International Conference on Network and Service Man-
agement (CNSM 16), October 2016 16. Sahay, R., Blanc, G., Zhang, Z., Debar, H.: Towards autonomic DDOS mitigation
using software defined networking. In: SENT 2015: NDSS Workshop on Security
of Emerging Networking Technologies. Internet society (2015) Blockchain-Based Architecture for Collaborative DDoS Mitigation 29 17. Scott-Hayward, S., O’Callaghan, G., Sezer, S.: SDN security: a survey. In: 201
IEEE SDN for Future Networks and Services (SDN4FNS), pp. References 1–7. IEEE (2013) 18. Shin, S., Gu, G.: Attacking software-defined networks: a first feasibility study. In:
Proceedings of the Second ACM SIGCOMM Workshop on Hot Topics in Software
Defined Networking, pp. 165–166. ACM (2013) g, pp
(
)
19. Steinberger, J., Kuhnert, B., Sperotto, A., Baier, H., Pras, A.: Collaborative DDOS
defense using flow-based security event information. In: NOMS 2016–2016 IEEE/I-
FIP Network Operations and Management Symposium, pp. 516–522, April 2016 20. The Associated Press: Hackers Used ‘Internet of Things’ Devices to Cause
Friday’s
Massive
DDoS
Cyberattack. http://www.cbc.ca/news/technology/
hackers-ddos-attacks-1.3817392. Accessed 10 Jan 2017 21. Velauthapillai, T., Harwood, A., Karunasekera, S.: Global detection of flooding-
based DDOS attacks using a cooperative overlay network. In: Network and System
Security (NSS), pp. 357–364. IEEE (2010) (
)
(
)
22. Wei, L., Fung, C.: Flowranger: a request prioritizing algorithm for controller dos
attacks in software defined networks. In: 2015 IEEE International Conference on
Communications (ICC), pp. 5254–5259. IEEE (2015) 23. Wood, G.: Ethereum: a secure decentralised generalised transaction ledger. https://
goo.gl/LG7adX 24. Zhang, G., Parashar, M.: Cooperative defence against DDoS attacks. J. Res. Pract. Inf. Technol. 38(1), 69–84 (2006) 24. Zhang, G., Parashar, M.: Cooperative defence against DDoS attacks. J. Res. Pract. Inf. Technol. 38(1), 69–84 (2006) Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. Achieving Reproducible Network Environments
with INSALATA Nadine Herold, Matthias Wachs, Marko Dorfhuber, Christoph Rudolf,
Stefan Liebald(B), and Georg Carle Department of Informatics, Chair of Network Architectures and Services,
Technical University of Munich (TUM), Munich, Germany
{herold,wachs,liebald,carle}@net.in.tum.de,
{marko.dorfhuber,christoph.rudolf}@tum.de Department of Informatics, Chair of Network Architectures and Services,
Technical University of Munich (TUM), Munich, Germany
{herold,wachs,liebald,carle}@net.in.tum.de,
{marko.dorfhuber,christoph.rudolf}@tum.de Abstract. Analyzing network environments for security flaws and
assessing new service and infrastructure configurations in general are
dangerous and error-prone when done in operational networks. There-
fore, cloning such networks into a dedicated test environment is beneficial
for comprehensive testing and analysis without impacting the operational
network. To automate this reproduction of a network environment in a
physical or virtualized testbed, several key features are required: (a) a
suitable network model to describe network environments, (b) an auto-
mated acquisition process to instantiate this model for the respective
network environment, and (c) an automated setup process to deploy the
instance to the testbed. With this work, we present INSALATA, an automated and exten-
sible framework to reproduce physical or virtualized network environ-
ments in network testbeds. INSALATA employs a modular approach for
data acquisition and deployment, resolves interdependencies in the setup
process, and supports just-in-time reproduction of network environ-
ments. INSALATA is open source and available on Github. To highlight
its applicability, we present a real world case study utilizing INSALATA. Keywords: Infrastructure Information Collection · Automated Testbed
Setup and Configuration · Testbed Management c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 30–44, 2017.
DOI: 10.1007/978-3-319-60774-0 3 1
Introduction The increasing number of attack vectors and the growing complexity of attacks
on computer networks force operators to continuously assess and improve their
networks, services, and configurations. Analyzing, testing, and deploying new
security features and configuration improvements is time-consuming, challeng-
ing, and error-prone. The same holds for general software upgrades or network
infrastructure changes. Performing this on an operational network is often not
suitable, as service continuity has to be ensured and outages cannot be tolerated. Therefore, reproducing a network environment into a self-contained test envi-
ronment is beneficial as the operational network is not influenced. Testing and Achieving Reproducible Network Environments with INSALATA 31 analyzing different options to improve the network and its security can be evalu-
ated and tested sufficiently before deployment. With large and complex network
environments, reproducing such network environments cannot be done manually
as information about the environment and its elements may be unknown, incom-
plete, or not available in a formal description. Hence, an automated process to
reproduce network environments in a physical or virtualized testbed is required. p
p y
q
In this work, we present INSALATA, the IT NetworkS AnaLysis And deploy-
menT Application. INSALATA enables network operators and researchers to
automate reproduction of arbitrary network environments in physical or vir-
tualized testbeds. To represent network environments, we provide a network
model comprising layer-2 network segments, IP networks, connectivity informa-
tion (routing and firewalling), network nodes (routers, hosts), network services
(DNS, DHCP), and host information (network interfaces, memory, disks, oper-
ating system). INSALATA can analyze network environments to obtain a formal
description of the network to track the state continuously or in discrete intervals. Here, INSALATA uses information fusing to provide a comprehensive view on
the network by aggregating information from multiple collector modules. Using
descriptions decouples analysis and deployment and enables re-using, archiving,
and distributing these descriptions. INSALATA can instantiate descriptions on
physical or virtualized testbeds employing a PDDL planner to structure the
setup process and resolve inter-dependencies between setup steps. To minimize
setup steps and reuse existing testbed setups, we support incremental setups by
determining the delta between current and target testbed state. 1
Introduction The key contributions of our work are (a) INSALATA, a fully automated,
modular, and extensible framework to reproduce network environments on test-
beds, (b) the open source implementation of INSALATA available on GitHub,
and (c) a case study showing INSALATA’s applicability to real world scenarios
using exemplary module implementations. The remainder of this paper is structured as follows: First, we describe
our goals and requirements for INSALATA in Sect. 2. Afterwards, we analyze
if related work can fulfill these in Sect. 3. In Sect. 4, we present INSALATA’s
design and introduce its components. Next, we present the main components of
INSALATA in detail, in particular the underlying information model in Sect. 5,
the Collector Component in Sect. 6, and the Deployment Component in Sect. 7. In Sect. 8, we summarize important implementation details. In Sect. 9, we present
a case study to show the applicability of our proposed system. Finally, we give
a conclusion and present future work in Sect. 10. 2
Goals and Requirements Agent-based information collection collects information just-in-time, but
agents need to be deployed and run on the components. Passive scanning has no direct impact on networks, but collected information
is limited and access to all network segments is required. Active scanning creates load in a network and on components, but provides
more detailed information about entities and services in the network. more detailed information about entities and services in the network. Network management protocols need to be available on investigated nodes,
but reduce system’s load and information requests are standardized. Direct access to components, e.g. with SSH, delivers rich information, but
requires appropriate access, to invoke applications, and interpret the output. Agent-based information collection collects information just-in-time, but
agents need to be deployed and run on the components. Network management protocols need to be available on investigated nodes,
but reduce system’s load and information requests are standardized. Direct access to components, e.g. with SSH, delivers rich information, but Network management protocols need to be available on investigated nodes,
but reduce system’s load and information requests are standardized. Direct access to components, e.g. with SSH, delivers rich information, but
requires appropriate access, to invoke applications, and interpret the output. Direct access to components, e.g. with SSH, delivers rich information, but
requires appropriate access, to invoke applications, and interpret the output. Agent-based information collection collects information just-in-time, but Agent-based information collection collects information just-in-time, but
agents need to be deployed and run on the components. (c) Deployment Process: The deployment process has to be incremental, so that
the delta between the current and the target state is computed during setup and
only required configuration steps are executed. Additionally, the deployment
process has to be modular and extensible in order to cope with use case specific
requirements. Therefore, the setup process has to be divided into small, self-
contained steps. To be able to use the deployment process on different testbed
architectures, the process itself needs to be independent from the underlying
architecture as much as possible. 2
Goals and Requirements The overall goal is to reproduce arbitrary network environments into physical or
virtualized testbeds. Therefore, we need (a) a suitable information model reflect-
ing required information, (b) an automated information acquisition process, and
(c) an automated deployment process. N. Herold et al. 32 (a) Information Model: The information model abstracts from the network envi-
ronment. The goal is to reflect network environments up to application layer
of the TCP/IP reference model. The information model has to be extensible
to allow to add use case specific services and additional information elements. Therefore, we require the following information to be present: (a) basic network
nodes, like hosts and routers, (b) networks on layer 2 and 3, including appropri-
ate addressing schemes, (c) connectivity information like routing and firewalling,
(d) basic network services like DNS and DHCP, and (e) host information, includ-
ing network interfaces, disks, memory, CPUs, or operating system. (b) Information Acquisition: The goal is to provide information acquisition that
supports different types of information collection techniques, supports continu-
ous monitoring of the network environment, and is fully automated. We iden-
tified that the following information collection techniques, differing in terms of
intrusiveness and quality of information they provide, have to be supported: Manually specified information is not intrusive, but rarely up-to-date. Including such information is required if other techniques are not applicable. y
p
,
y
p
Including such information is required if other techniques are not applicable. Passive scanning has no direct impact on networks, but collected information
is limited and access to all network segments is required. Active scanning creates load in a network and on components, but provides
more detailed information about entities and services in the network. Network management protocols need to be available on investigated nodes,
but reduce system’s load and information requests are standardized. Including such information is required if other techniques are not applicable. Passive scanning has no direct impact on networks, but collected information
is limited and access to all network segments is required. Active scanning creates load in a network and on components, but provides
more detailed information about entities and services in the network. Network management protocols need to be available on investigated nodes,
but reduce system’s load and information requests are standardized. Direct access to components, e.g. with SSH, delivers rich information, but
requires appropriate access, to invoke applications, and interpret the output. Achieving Reproducible Network Environments with INSALAT Achieving Reproducible Network Environments with INSALATA 33 Data Acquisition Applications are needed to obtain information about the net-
work environment. Here, continuous monitoring is required and information from
different sources needs to be fused. Additionally, tracking changes is a require-
ment. IO-Framework [6,19] does not support continuous monitoring and only
supports intrusive collection methods. The common Network Information Ser-
vice (cNIS) [1] utilizes static information and higher level services (SSH or
SNMP) but does not include less invasive information collection techniques. MonALISA [7,11] and PerfSONAR [29] are not capable of continuously mon-
itoring the network and detect changes. OpenVAS [23] is used to identify vul-
nerabilities within an infrastructure but has limited scanning capabilities. Single
purpose tools like Nmap [20], Traceroute [2], or xprobe2 [35] can be used to col-
lect single aspects of the network environment but do not provide a holistic view. Dedicated network management protocols, like SNMP [8,21] or Netconf [27] can
only be used to retrieve dedicated information from single network components,
but do also not provide a complete view on the network. Testbed Management Frameworks are used to orchestrate and control test-
beds. All presented frameworks do not provide incremental setups but rebuild
the designated network from scratch leading to higher effort within the setup
process and manually configured changes get lost. Additionally, testbed orches-
tration and experiment execution are often tightly coupled. vBET [18] and Laas-
NetExp [24] are both closed source, preventing to extend those frameworks. VNEXT [25] and NEPTUNE [5] do not provide the automated setup of basic
network services, like DNS or DHCP. Emulab [34] or DETER [4] tightly couple
the infrastructure setup and the experiment execution. This requires to rebuild
the network after each experiment. Network Description Languages and Ontologies can be used as information
model. The related work within this field lacks in providing the information
elements needed for a proper mirroring of network environments, especially in
terms of reflecting the connectivity due to routing and the usage of firewalls. IF-
MAP [3,30,31] is not capable of reflecting interfaces or routing information. The
target-centric ontology for intrusion detection [32] does not provide a sufficient
addressing scheme for elements nor reflect routing or firewalls. The Network
Markup Language (NML) [14,15] provides a schema for exchanging network
descriptions on a generic level, but does not provide concepts like network rout-
ing. 3
Related Work To the best of our knowledge, no application to reproduce network environ-
ments exists. Therefore, we examine the two main components of INSALATA,
namely information acquisition and testbed setup for deployment, separately. Next, we investigate network description languages as we need a suitable net-
work model for INSALATA. Finally, we discuss network management protocols
and frameworks to investigate appropriate implementation mechanisms to deploy
the description of the network environment on the testbed. Achieving Reproducible Network Environments with INSALAT The Infrastructure and Network Description Language (INDL) [13,17,28]
extends NML, but is not capable of reflecting routes or firewall rules. Tcl-based
format in Emulab and ns-2 [34] is not capable of modeling networks explicitly
resulting in verbose definitions for large networks. Network Management Protocols and Frameworks can be used to setup and con-
figure the descriptions in testbeds. To do so, the virtualized testbed has to be
setup, e.g. router and hosts as virtual machines, and those components need
to be configured appropriately afterwards, e.g. using adequate routing tables. Dedicated network management protocols, like SNMP [8,21] or Netconf [27] N. Herold et al. 34 can be used to configure components and request dedicated information in a
standardized way. Both protocols do not provide built-in mechanisms to man-
age larger infrastructures as a whole. Ansible [9] is a push-based framework to
configure larger infrastructures using so-called playbooks. Those playbooks need
to be written or adapted for each configuration. Ansible can not be used directly
for our approach, but is suitable as an important building block. Puppet [26]
is a pull-based framework for infrastructure configurations. As the testbed is
reconfigured in irregular intervals, a pull-based mechanism is not suitable. Achieving Reproducible Network Environments with INSALATA 35 Achieving Reproducible Network Environments with INSALATA The Collector acquires information about the network environment and is
used to generate a description for the testbed. It maintains the current state of
the monitored network. Here, the collection process can be done continuously
whereas configuration changes can be tracked and stored with a timestamp in a
database. This approach allows to rebuild a network at each point in time. The Deployment Component utilizes a description that is either obtained by
the Collector or provided by the user. Based on this, the Deployment Component
configures the testbed to reproduce a network environment. To ease writing
descriptions, a Preprocessor is utilized, replacing missing but calculable values
in the description and checking the it for its validity. 4
Approach and System Design INSALATA consists of the Collector and the Deployment Component as its two
main components: The Collector Component is responsible for collecting and fusing informa-
tion of the network environment and the current state of the testbed into a
descriptions (see Sect. 6). The Deployment Component manages configuration changes and the auto-
mated setup process on the testbed (see Sect. 7). Both components utilize the same information model to structure the infor-
mation about the network environment (see Sect. 5) and are managed and orches-
trated by a central controller, the Management Unit. The system architecture
of INSALATA showing these basic components and their interaction is depicted
in Fig. 1. Management Unit
Pre-
processor
upload
User
virtual
physical
Collector
Deployment
Database
load
store
scan
deploy/change
Fig. 1. System overview of INSALATA showing basic components virtual Pre-
processor Fig. 1. System overview of INSALATA showing basic components 5
Description of the Information Model To be able to reflect a network environment in a testbed, a formal description
of this network is required. This description needs to contain information ele-
ments discussed in Sect. 2. Each information element needs to be leviable from
the network environment in an automated manner and has a unique identifier. The proposed information model is shown in Fig. 2. An information element is
represented as box, the identifier of each element is underlined and additional
attributes describing the information element are listed. For FirewalRules and
Routes a combination of attributes is used as identifier. Relations between infor-
mation elements are denoted as arrows in between and additionally denote their
cardinality. The main information element is the Network Component representing a node
in the network environment, e.g. hosts, routers or switches, and are further spec-
ified by the Template attribute. A Network Component is equipped with certain
Disks and Interfaces. Interfaces are needed to interconnect Network Components
in different kinds of networks, e.g. Layer 2 Networks or Layer 3 Networks. Depending on its functionality, a Network Component can maintain Routes
or Firewall Rules. Those express the connectivity between Network Components. The latter can be represented as raw dumps (Firewall Raw) or in a simplified
format (Firewall Rule) to ease transformation between different firewall applica-
tions as proposed in [10]. As a simplification is not free of information loss, the
raw information is stored additionally. To support large-scale test environments consisting of multiple servers, a
Network Element is associated with a Location specifying the testbed server
the Network Component is emulated on. In case of a description reflecting the
network environment, the Location is set to physical. Another important concept of the information model is the Service element. This information element is used to reflect basic network services, like DNS or
DHCP. A Service can be further specified by adding a Product and a Version
to allow a high accuracy. Additionally needed services can be added to the
information model using inheritance, allowing use case specific applications. N. Herold et al. 5
Description of the Information Model 36 1
use
0..*
1..*
in
1
1..*
contains
1..*
1
use
0..*
1
use
0..*
1
has
0..*
0..*
in 1
0..*
1
0..*
connected to 0..1
0..*
0..1
0..*
1
0..*
hop 1
0..*
start 1
0..*
end 1
0..*
destNet
1
0..*
srcNet
1
0..*
interface 1
0..*
hop
1
0..*
interface
1
0..*
1
NetworkComponent
id : String
cpus : Integer
cpuSpeed : Integer
memoryMin : Integer
memoryMax : Integer
powerState : String
template : String
Disk
id : String
size : Integer
Interface
mac: String
rate : Integer
mtu : Integer
Layer2Network
id : String
Location
id : String
Layer3Address
address : String
netmask : String
gateway : String
static : Boolean
Route
genmask : String
metric : Integer
FirewallRule
chain : String
action : String
protocol : String
srcPorts : list
destPorts : list
Layer3Network
address : String
netmask : String
FirewallRaw
firewall : String
data : String
Service
port : Integer
type : String
protocol : String
product : String
version : String
DnsService
domain : String
DhcpService
lease : Integer
connected to
configured on
running on
destination
Fig. 2. Information model of INSALATA Fig. 2. Information model of INSALATA 6
Information Collector Component The Collector Component is capable of managing multiple Environments
describing multiple networks to be monitored. The overview of INSALATA’s
information Collector Component is depicted in Fig. 3. For each Environment, multiple Collector Modules, employing a particular
information collection technique as described in Sect. 2, can be configured to
collect the required information. A Collector Module obtains information about
at least one information element and possible relations between information ele-
ments. Therefore, modules have to obtain the unique identifier of an information
element. The collected information elements are handed over to the Environment
fusing all obtained information into a comprehensive graph describing the net-
work. Here, we assume that a module delivers no false, but potentially incomplete
information. This modular approach has the advantage that different, specialized
information collection techniques can be combined resulting in a more detailed
and more precise view on the network environment. To fuse information, the Collector utilizes the identifier of each object and
the type of the information element. Objects with the same identifier and of
the same type are treated as the same object. Each Collector Module passes
discovered objects to the Collector. Attributes and relations are fused together
in case multiple modules report the same objects. Achieving Reproducible Network Environments with INSALATA 37 Management Unit
Env1
config
Mod1,2
Mod1,1
Mod1,3
Net1
Env2
config
Mod2,1
Mod2,2
Net2
Fig. 3. Information Collector Component of INSALATA Management Unit Fig. 3. Information Collector Component of INSALATA Another challenge is to manage existing information elements in the model
becoming obsolete. Therefore, a deletion scheme needs to be implemented within
an Environment. Each information element in an Environment is equipped with
timers for each Collector module. Each time, an information element is delivered
by a module, the module specific timer is updated. If all timers in the list expire,
the information element and its relations are deleted from the Environment. In
addition, each module can actively set its own timer to zero, if it is capable to
determine the non-existence of an element. To be able to recreate an environment at any (observed) point in time, we
track the network environment’s state over time. Whenever an information ele-
ment or relation is modified, i.e., is added, deleted, or updated, we save this
delta as an event to the database. Such events contain the type of change, the
information element and its properties. 6
Information Collector Component Within an Environment, only the current
state of the network description is maintained. 7
Infrastructure Deployment Component The Deployment Component executes the following steps: (a) determine required
configuration steps using the current testbed state and the given description,
(b) determine a correct execution plan how to achieve the given description, and
(c) deploy the changes on the testbed following the computed execution plan. The overview of the execution flow of INSALATA’s Deployment Component is
depicted in Fig. 4. First, the Deployment Component has to determine what needs to be
changed. Therefore, we need the current state of the testbed in the form a
description as Description D2 that can be determined using the Collector
Component. Additionally, we need the target state in the form a description
as Description D1 provided from the Collector or the user. The Deployment N. Herold et al. Pre-
processor
Description
D1
XML
Collector
Description
D2
Change
detec-
tion
Planning
Problem
parser
Problem
Planner
Domain
Plan
Builder
Builder
module
n
Builder
module
1
... uses
Testbed
builds
scans
Fig. 4. Execution flow of INSALATA’s Deployment Component 38 N. Herold et al. Fig. 4. Execution flow of INSALATA’s Deployment Component Component uses a Change Detection Module to detect added, removed, and
updated information elements in the delta between these states. To determine how to change the testbed, we use automated planning and
scheduling from the domain of artificial intelligence [16]. A planning problem is
described using a dedicated planning language such as the Planning Domain Def-
inition Language (PDDL). PDDL separates the planning problem into a domain
description, describing the problem domain, and a problem description, describ-
ing an instance of the problem [12]. A PDDL domain description describes the
objects’ types, predicates (i.e., properties), and actions. Each action has a defin-
ition of objects it is applicable to, preconditions that have to be fulfilled, and an
effect altering the predicates of objects. With INSALATA, we provide a domain
description for our information model and steps as PDDL actions necessary to
setup a testbed. A detailed overview on the steps we defined and their inter-
dependencies can be found in the domain description provided with the imple-
mentation. The PDDL problem description describes all objects, their type and
their initial state. The problem specifies the goal state of all objects by giving
their desired predicates [22] inside a goal section. 1 https://github.com/tumi8/INSALATA.
2 https://insalata.readthedocs.io/en/latest/index.html.
3 http://www.fast-downward.org/ObtainingAndRunningFastDownward.
4 https://ilab.net.in.tum.de. 8
Implementation INSALATA is written in Python 3 and is available in INSALATA’s GitHub
repository1. A detailed code documentation is available on2. The Management Unit can be controlled using a XML-RPC client commu-
nicating with INSALATA. The presented information model is reflected using
object-oriented Python classes. Besides the Collector Component itself, we provide the following Collector
Modules: (1) a XEN module using to collect information from environments
using XEN virtualization, (2) an XML module for manually provided informa-
tion, (3) a Tcpdump module for passive network scanning using Tcpdump, (4) an
Nmap module for active network scanning using Nmap, (5) an SNMP module
to retrieve information from nodes using SNMP, (6) an SSH module to retrieve
information from nodes via SSH, and (7) a Zabbix module for agent-based infor-
mation collection with the Zabbix network monitoring system. Some modules
have limited functionality, meaning that not all information possible to collect is
implemented, e.g. the SSH module does not collect firewall rules. New collector
modules can be added to INSALATA to cover the desired scanning environment. g
We integrate fast-forward3 as a planner into INSALATA’s Deployment Com-
ponent. The domain file describing the setup process of a testbed we provide, is
given in PDDL. The required problem files are generated for each setup individ-
ually in an automated manner. We provide a framework allowing to add new
Builder Modules in an easy way. Here, we utilize Python annotations to deter-
mine the most suitable Builder Module for a given step within the determined
plan and the configured object. Besides provided Builder Modules to setup the
basic topology on XEN (hosts, routers, layer 2 networks), we utilize Ansible for
additional configurations (routing, firewalling, DNS, and DHCP). 7
Infrastructure Deployment Component While the domain file is static,
the problem file depends on the current and target state and is computed each
time a description is deployed on the testbed. The computed changes, the current testbed state, and the target state are
given to the Problem Parser Module computing a PDDL problem description. This description is given to the Planner Module, an automated planning and
scheduling solver computing an execution plan containing the correct order of
actions that will bring the testbed from the current into the target state. The
execution plan is passed to the Builder to execute the steps on the designated
testbed. Implementations of these steps are provided by architecture-specific
Builder Modules since the implementation of such steps is different for different
testbed architectures and objects. The Builder uses meta-data associated with
the Builder Modules and the objects to identify the correct implementation. Achieving Reproducible Network Environments with INSALATA 39 This allows us to dynamically add new implementations, e.g. for new operating
systems or services, without changing the Deployment Component. 9
Case Study: Chair’s Teaching Infrastructure – iLab To show INSALATA’s applicability in practical scenarios, we assesed INSALATA
in a case study. For this case study, we use a setup adapted from the lab course
iLab4 offered by the TUM’s Chair of Network Architectures and Services. The
iLab is a course to teach student’s practical skills in administering network setups
and configurations for different scenarios using real hardware. A typical setup
students have to work with during an iLab course is shown in Fig. 5. This setup
consists of two Cisco and a Linux router, and four host residing in different
private networks. In our case study, our goal is to reproduce this network envi-
ronment in a network testbed using XEN virtualization. 40 N. Herold et al. Cisco-Router 1
Linux-Router
Cisco-Router 2
Switch
PC1
PC2
PC3
PC4
eth0 10.0.1.254/24
eth0 10.0.2.254/24
eth0 10.0.3.254/24
eth0 10.0.1.1/24
eth0 10.0.1.2/24
eth0 10.0.2.1/24
eth0 10.0.3.1/24
eth2 10.0.6.1/24
eth2 10.0.6.2/24
eth1 10.0.4.1/24
eth1 10.0.5.2/24
eth2 10.0.4.1/24
eth1 10.0.5.1/24
Fig. 5. Typical infrastructure setup within the iLab course Fig. 5. Typical infrastructure setup within the iLab course We provide the IP addresses of all routers to the INSALATA system as a
starting point using manual input in form of XML files. All routers and hosts
are configured to allow INSALATA to access the systems using SSH and SNMP. In the first phase, we obtain the required description of the network environ-
ment using our Collector Component. To expand our infrastructure information We provide the IP addresses of all routers to the INSALATA system as a
starting point using manual input in form of XML files. All routers and hosts
are configured to allow INSALATA to access the systems using SSH and SNMP. In the first phase, we obtain the required description of the network environ-
ment using our Collector Component. To expand our infrastructure information
using our passive Tcpdump Collector Module, we generate traffic on the involved
hosts. Using the SNMP and SSH Collector Modules, missing interfaces, MAC
addresses, and routing information is obtained from routers and hosts. Using
theses Modules, we are able to reflect the network environment shown in Fig. 5
as description. In the second phase, we use our Deployment Component to reproduce the
obtained description into our virtualized testbed using XEN with the xapi tool-
stack. INSALATA computes an execution plan to setup the testbed from scratch
in 0.252 s. 9
Case Study: Chair’s Teaching Infrastructure – iLab The resulting execution plan consists of 92 steps, including setting
up virtual machines and networks, configuring interfaces and deploying routes. Each step is executed in sequence and no parallelization is done. The total time
required to setup the testbed and configure it is 42 min 32 s. Figure 6 visualizes
the setup and configuration process in regard to its execution time. The most time-consuming tasks during the setup process are the creation of
new virtual machines, which happens at the beginning. The reason for this is
that here new virtual machines have to be cloned from the respective template,
including copying hard disk images and required reboot operations. Configura-
tion steps like creation of virtual networks and interfaces are done nearly instan-
taneously. After the setup of our description, we validated the correctness of our
setup using manual inspection, Ping and Traceroute. Achieving Reproducible Network Environments with INSALATA 41 0
200
400
600
800 1,000 1,200 1,400 1,600 1,800 2,000 2,200 2,400
1
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
Time t in s
Step number
Fig. 6. Time distribution of setup steps in the iLab case study Fig. 6. Time distribution of setup steps in the iLab case study References 1. GEANT2 common Network Information Service (cNIS) Schema Specification. http://www.geant2.net //
2. traceroute(8) - Linux man page. linux.die.net/man/8/traceroute 3. Ahlers, V., Heine, F., Hellmann, B., Kleiner, C., Renners, L., Rossow, T.,
Steuerwald, R.: Integrated visualization of network security metadata from het-
erogeneous data sources. In: Mauw, S., Kordy, B., Jajodia, S. (eds.) GraM-
Sec 2015. LNCS, vol. 9390, pp. 18–34. Springer, Cham (2016). doi:10.1007/
978-3-319-29968-6 2 4. Benzel, T.: The science of cyber security experimentation: the DETER project. In: Proceedings of 27th Annual Computer Security Applications Conference, pp. 137–148 (2011) 5. Bifulco, R., Stasi, G.D., Canonico, R.: NEPTUNE for fast and easy deployment of
OMF virtual network testbeds [Poster Abstract] (2010) 6. Birkholz, H., Sieverdingbeck, I., Sohr, K., Bormann, C.: IO: an interconnected asset
ontology in support of risk management processes. In: 7th International Conference
on Availability, Reliability and Security (ARES), pp. 534–541 (2012) 7. Carpen-Amarie, A., Cai, J., Costan, A., Antoniu, G., Boug, L.: Bringing introspec-
tion into the BlobSeer data-management system using the MonALISA distributed
monitoring framework. In: International Conference on Complex, Intelligent and
Software Intensive Systems (CISIS), pp. 508–513 (2010) 8. Case, J.D., Fedor, M., Schoffstall, M.L., Davin, J.R.: Simple network management
protocol (SNMP). RFC 1157, IETF, May 1990 eHaan, M.: Ansible (2012–2016). https:// 9. DeHaan, M.: Ansible (2012–2016). https://www.ansible.com/ 10. Diekmann, C., Hupel, L., Carle, G.: Semantics-preserving simplification of real-
world firewall rule sets. In: 20th International Symposium on Formal Methods, pp. 195–212 (2015) (
)
11. Dobre, C., Voicu, R., Legrand, I.: Monitoring large scale network topologies. In:
IEEE 6th International Conference on Intelligent Data Acquisition and Advanced
Computing Systems (IDAACS), vol. 1, pp. 218–222 (2011) 12. Fox, M., Long, D.: PDDL2.1: an extension to PDDL for expressing temporal plan-
ning domains. J. Artif. Int. Res. 20(1), 61–124 (2003) 13. Ghijsen, M., van der Ham, J., Grosso, P., Dumitru, C., Zhu, H., Zhao, Z.,
de Laat, C.: A semantic-web approach for modeling computing infrastructures. Comput. Electr. Eng. 39(8), 2553–2565 (2013) p
g
( ),
(
)
14. van der Ham, J., Dijkstra, F., Apacz, R., Zurawski, J.: Network Markup Language
Base Schema version 1 (2013) 15. van der Ham, J., Dijkstra, F., Lapacz, R., Brown, A.: The Network Markup Lan-
guage (NML): A Standardized Network Topology Abstraction for Inter-domain
and Cross-layer Network Applications (2013) 16. Hoffmann, J.: Everything you always wanted to know about Planning. In: Bach, J.,
Edelkamp, S. (eds.) KI 2011. LNCS, vol. 10
Conclusions and Future Work In this work we present INSALATA, a system capable of reproducing network
environments in network testbeds. INSALATA enables network operators and
researchers to test and analyze new security features and general configuration
changes in a separated test environment before deployment in operational net-
works. To be able to formalize network environments, INSALATA utilizes an
information model particularly crafted for representing network topologies, enti-
ties, and services in descriptions. To obtain the required information from net-
work environments, we support a modular Collector Component automatically
assessing networks and fusing information from different Collector Modules. The
Deployment Component provides automated planning and scheduling to instan-
tiate descriptions onto a physical or virtualized testbeds. Within our case study,
we show the applicability of our approach reproducing a real world network
environment with several routers and hosts onto a virtualized testbed. To further improve INSALATA, we will continue our work in this field and
on INSALATA and highly appreciate feedback, improvements, and extensions
from the community. To extended INSALATA’s capabilities, additional Collector
and Builder Modules can help to obtain additional properties from the network
environment, such as user information or generic service configurations from
hosts. Existing Collector Modules can be extended to obtain more information
using existing assessment techniques, such as firewall information using SSH. To
reproduce network environments more realistically, Builder Modules to support
Microsoft Windows and additional network services, like mail or ftp are bene-
ficial. One of our main goals is to make the deployment process more efficient
by parallelizing the execution of the setup plan. Since INSALATA only pro-
vides mechanisms to setup and orchestrate a testbed, we aim to integrate our
experiment execution framework GPLMT [33] into the INSALATA system. 42 N. Herold et al. Acknowledgments. This work has been supported by the German Federal Ministry
of Education and Research (BMBF) under support code 16KIS0538 (DecADe). Acknowledgments. This work has been supported by the German Federal Ministry
of Education and Research (BMBF) under support code 16KIS0538 (DecADe). References 7006, pp. 1–13. Springer, Heidelberg
(2011). doi:10.1007/978-3-642-24455-1 1 17. van der Ham, J.J.: A semantic model for complex computer networks: the network
description language. Ph.D. thesis, University of Amsterdam (2010) Achieving Reproducible Network Environments with INSALATA 43 18. Jiang, X., Xu, D.: vBET: A VM-based emulation testbed. In: Proceedings of ACM
SIGCOMM Workshop on Models, Methods and Tools for Reproducible Network
Research, pp. 95–104 (2003) 19. Lorenzin, L., Cam-Winget, N.: Security automation and continuous monitor-
ing (SACM) requirements. Internet-Draft draft-ietf-sacm-requirements-15, Inter-
net Engineering Task Force (2016) (
)
20. Lyon, G.: nmap(1) - Linux man page (2015) 21. McCloghrie, K., Rose, M.: Management information base for network management
of TCP/IP-based internets. RFC 1156, IETF, May 1990 /
22. McDermott, D., Ghallab, M., Howe, A., Knoblock, C., Ram, A., Veloso, M.,
Weld, D., Wilkins, D.: PDDL - the planning domain definition language (1998) 23. OpenVAS: Project Homepage. http://www.openvas.org/ 24. Owezarski, P., Berthou, P., Labit, Y., Gauchard, D.: LaasNetExp: a generic poly-
morphic platform for network emulation and experiments. In: Proceedings of 4th
International Conference on Testbeds and Research Infrastructures for the Devel-
opment of Networks & Communities, no. 24, pp. 1–9 (2008) 25. Pisa, P.S., Couto, R.S., Carvalho, H.E.T., Neto, D.J.S., Fernandes, N.C., Campista,
M.E.M., Costa, L., Duarte, O., Pujolle, G.: VNEXT: Virtual network management
for Xen-based testbeds. In: International Conference on the Network of the Future
(NOF 2011), pp. 41–45 (2011) 26. Puppet Labs: Puppet (2005–2016). https://puppet.com/ 27. Enns, R., Bjorklund, M., Schoenwaelder, J., Bierman, A.: Network configuration
protocol (NETCONF). RFC 6241, IETF, June 2011 28. Taketa, T., Hiranaka, Y.: Network design assistant system based on network
description language. In: 15th International Conference on Advanced Communica-
tion Technology (ICACT), pp. 515–518 (2013) 29. Tierney, B., Metzger, J., Boote, J., Boyd, E., Brown, A., Carlson, R., Zekauskas,
M., Zurawski, J., Swany, M., Grigoriev, M.: perfSONAR: instantiating a global
network measurement framework. In: SOSP Workshop on Real Overlays and Dis-
tributed Systems (ROADS 2009). ACM (2009) (
)
(
)
30. Trusted Network Connect Work Group: TNC IF-MAP Bindings for SOAP, Version
2.2, Revision 10 (2014) 31. Trusted Network Connect Work Group: TNC MAP Content Authorization, Ver-
sion 1.0, Revision 36 (2014) 32. Undercoffer, J., Pinkston, J., Joshi, A., Finin, T.: A target-centric ontology for
intrusion detection. In: Proceeding of 9th Workshop on Ontologies and Distributed
Systems, pp. 47–58 (2004) 33. References Wachs, M., Herold, N., Posselt, S.-A., Dold, F., Carle, G.: GPLMT: a light-
weight experimentation and testbed management framework. In: Karagiannis, T.,
Dimitropoulos, X. (eds.) PAM 2016. LNCS, vol. 9631, pp. 165–176. Springer, Cham
(2016). doi:10.1007/978-3-319-30505-9 13 34. White, B., Lepreau, J., Stoller, L., Ricci, R., Guruprasad, S., Newbold, M., Hibler,
M., Barb, C., Joglekar, A.: An integrated experimental environment for distributed
systems and networks. pp. 255–270 (2002) 35. Yarochkin, F., Arkin, O., Kydyraliev, M.: xprobe2(1) - Linux man page N. Herold et al. 44 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 47–61, 2017.
DOI: 10.1007/978-3-319-60774-0 4 Towards a Software-Defined Security Framework
for Supporting Distributed Cloud Maxime Compasti´e1,2(B), R´emi Badonnel1, Olivier Festor1, Ruan He2,
and Mohamed Kassi-Lahlou2 1 LORIA - Inria, Campus Scientifique, 54600 Villers, France
{maxime.compastie,remi.badonnel,olivier.festor}@loria.fr
2 Orange Labs, 44 Avenue de la R´epublique, 92320 Chatillon, France
{ruan.he,mohamed.kassilahlou}@orange.com Abstract. Cloud computing provides new facilities for building elabo-
rated services hosted through various infrastructures over the Internet. In the meantime, these ones pose new important challenges in terms of
security due to their intrinsic nature. We propose in this paper to detail
a software-defined security framework supporting the protection of these
services, in the context of distributed cloud. These ones require security
mechanisms able to cope with their multi-tenancy and multi-cloud prop-
erties. The foundations of this framework rely on the software-defined
logic to express and propagate security policies to the considered cloud
resources, and on the autonomic paradigm to dynamically configure and
adjust these mechanisms to distributed cloud constraints. In particular,
we describe the main components and protocols of this software-defined
security framework, evaluate this one and discuss implementation con-
siderations, through the analysis of different realistic scenarios. 1
Introduction The cloud computing architectural model permits to build elaborated ser-
vices and applications based on multiple computing resources, such as virtual
machines, network devices, software components, themselves provided as a ser-
vice that can be easily deployed through the Internet. Based on the NIST Insti-
tute [1] definition, this model is mainly characterized by the following features:
on-demand self-service, broad network access, resource pooling, rapid elasticity,
and measured service. It supports an as a service scheme that permits a trans-
parent access to resources and the outsourcing of part of the management to
the cloud provider. This separation enables optimizing the resource allocation
and usage, but may also introduce management complexity due to its distrib-
uted nature. In particular, the cloud infrastructure and its applications may
typically be divided into isolated sets of resources called tenants, correspond-
ing to different ownerships and requirements, defining the multi-tenancy prop-
erty. Another property comes to the facts that the resources may be distributed
among several infrastructures, as each of them may be specialized in a dedicated M. Compasti´e et al. 48 processing. Distributed cloud can be defined by the conjunction of the multi-
tenancy and multi-cloud properties. In this context, security management has
become a major challenge. The dynamics of cloud infrastructures induced by
their on-demand self-service, rapid elasticity and distribution has outrun tradi-
tional security management, while the ubiquity and high availability of cloud
resources make them attractive targets for attackers [2]. Exploiting autonomic and programmability mechanisms opens new perspec-
tives for enabling such a security management. Autonomic computing permits to
address the scalability issues induced by large and distributed cloud infrastruc-
ture resources, by delegating part of the management tasks to the environment
itself. In our context, this concerns more particularly the management tasks
related to self-protection and self-configuration, and aims at maintaining the
security level of a distributed cloud and its services in an adequate manner with
the security threats, based on the activation or deactivation of available coun-
termeasures in a proactive and/or reactive manner. In addition, network pro-
grammability has already shown its advantage for software-defined networking
by separating the network infrastructure into two separate planes, i.e. the data
plane and the control plane, and contributing to its dynamic configuration and
adaptation. Similarly, there is an important need for supporting software-defined
security in the context of distributed cloud. 1
Introduction We have already highlighted the benefits of software-defined security for dis-
tributed cloud environments in [3]. We detail in this paper the different com-
ponents and protocols of our security framework relying on software-defined
and autonomic paradigms, and provides a critical analysis of the proposed solu-
tion considering a set of validation scenarios based on a realistic use case. The
framework permits to specify security policies, and enables their autonomic
enforcement in a multi-tenant and multi-cloud environment. Security mecha-
nisms should be dynamically aligned and adjusted based on changes that may
occur in the distributed cloud. The rest of this paper is organized as follow:
Sect. 2 gives an overview of existing work related to our software-defined secu-
rity solution. The proposed framework, its components and their interactions are
detailed in Sect. 3. We evaluate it and give a critical analysis as well as imple-
mentation considerations in Sect. 4. Finally, we conclude the paper and point
out future research efforts in Sect. 5. 2
Related Work The security of cloud infrastructures has already been largely explored in the
literature. In particular [4] highlights several challenges related to policy-based
security management, such as the specification of a cloud security policy, the
assurance of the security decisions, as well as the certification of security compo-
nents in that context. In the same manner, the TCloud framework [5] proposes
to enforce a security policy with a hardened cloud stack. This one provides
infrastructure-level and platform-level security components, that might be com-
patible with multi-cloud environments, with a hardened build of OpenStack envi-
ronments. However, these solutions do not specifically address self-configuration Towards a Software-Defined Security Framework 49 mechanisms, nor the management issues generated by multi-cloud and multi-
tenancy properties. The Iceman architecture [6] enables secure federated inter-
cloud identity management. The author of [7] proposes a cloud management
framework able to deal with multi-tenancy, but this one is limited to access
control policies and cannot support other security mechanisms. The proposed
architecture is independent from the available security mechanisms and addresses
their self-configuration in a distributed cloud. In the area of programmability, software-defined networking (SDN) permits
to separate the control plane making decisions about where the traffic should be
sent from the data plane forwarding of packets. This paradigm enables a dynamic
and adaptive policy enforcement. It may also serve as a support for chaining secu-
rity functions. For instance, the Flowtags framework described in [8] enables
the integration of middleboxes whose composition is supported by SDN con-
troller. [9] proposes a framework for enforcing a network security policy through
a set of middleboxes. But, this solution only considers middleboxes for instanti-
ating security mechanisms. We have also shown in [10] how to exploit the SDN
paradigm to build a chain of security functions, including intrusion detection
systems and firewalls, to protect smart devices. IETF is also working on SDN-
based security services using interface to network security functions [11]. Such
approaches take advantage of SDN with respect to security policy enforcement. Important efforts have also focused on the verification of security chains. For
instance, VeriCon [12] combines a language for specifying SDN policies with
an approach to check whether a policy verifies invariants expressed in predicate
logic. In the same manner, FlowChecker [13] represents the network as a binary
decision diagram (BDD), whereas properties are expressed in computation tree
logic (CTL). 2
Related Work However, the model based on BDDs requires a certain expertise of
formal methods, which cannot be generally expected from network operators. In
our context, we are focusing on a software-defined security framework to protect
distributed cloud, in line with software-defined networking, but not limited to
network enforcement considerations. The autonomic computing paradigm gives a framework for self-management
activities, and relies on several main areas: self-configuration, self-optimization,
self-protection and self-healing [14]. Although it does not bring a formal dis-
tributed cloud support, it may introduce the negotiation among independent
components. This approach may deal with exhaustive enforcement issues, as
autonomic components can continuously enforce the security policy and adapt
to the changes in their action perimeters. Even if the two previous paradigms
do not directly deal with distributed cloud issues, they provide important build-
ing blocks for supporting security policy enforcement and defining a security
management architecture in that context and in our framework. With respect to security policies, the OASIS consortium introduces two stan-
dardized languages: XACML (eXtensible Access Control Markup Language) for
representing and exchanging security policies [15] and SAML (Security Asser-
tion Markup Language) for specifying security statements [16]. However, they
do not handle any modifications of cloud policies nor its evolution propagation M. Compasti´e et al. 50 to enforcers. This approach remains relevant as the XACML defines modular
components for security enforcement. Besides, an architecture and use-cases fea-
turing XACML and SAML in distributed environment have been detailed in [17]. The latter validates the usability of XACML in distributed systems, underlining
some limitations such as the need for a high granularity of sub-policies and the
difficulty of maintaining an encoded security policy. The languages and formats
introduced by SCAP protocol constitutes also an interesting support, as they
cover many complementary specifications, such as vulnerability descriptions and
scorings, that are exploitable for automating security in distributed cloud [18]. These standards are usable in our solution. In accordance with [3] where we give the basement of our software-defined
security approach, the autonomic paradigm is tied to endorse the continuous
security policy enforcement able to cope with the changes occurring on the
security policy, the tenant configuration and the protected resource state. We
extend our previous work by detailing each components and protocols supporting
our framework, and giving a critical analysis and implementation considerations
based on realistic scenarios. 3
Software-Defined Security Framework We propose a software-defined security (SDSec) framework for protecting distrib-
uted cloud. These one is composed of two main layers, called respectively secu-
rity control plane and security data plane (as depicted on Fig. 1). It relies on a
software-defined scheme to provide a global security policy specification interface
and exploit autonomic mechanisms within distributed cloud infrastructures to
enable cloud resources to be dynamically and exhaustively protected according to
this policy. More precisely, it first consists in a global security policy (GSP) which
formally defines at a business level the security objectives of cloud resources and
is then translated into several tenant-level security policies (TLSP), providing
security statements that must be verified by specified resources at the tenant
level within the distributed cloud. Fig. 1. SDSec framework in a single-infrastructure single-tenant scenario Fig. 1. SDSec framework in a single-infrastructure single-tenant scenario 51 Towards a Software-Defined Security Framework These security statements are then enforced on cloud resources, i.e. virtual-
ized infrastructures and software products. They aim at altering the behavior of
these components and protecting them based on countermeasures available with
distributed cloud. This application can be active if its application requires negoti-
ation with a decisional instance. The enforcement should be performed dynam-
ically, more precisely in an adaptive (it adapts to any change in the enforced
resource state or in the infrastructure), automatic (no operator interventions
are needed for it), and self-configured manner (policy decisions for it are auto-
matically made according to several criteria including the security requirements). The components of the framework part of the security control plane include
the security orchestrator hosting a GSP specified by the system administrator,
exposing through a dedicated interface the TLSPs, and receiving enforcement
feedbacks from the policy decision point (PDP) to adapt them. These interac-
tions are supported by the security discovery protocol enabling the PDP to iden-
tify the security orchestrator and fetch its security policy. The components part
of the security data plane correspond to the policy enforcement points (PEP)
executing the security statements (using the security statement protocol) and
dedicated to the policy enforcement on one type of cloud resources. It may also
solicit the PDP for taking a needed security decision for an active enforcement
(using the security decision requesting protocol). This framework follows a software-defined paradigm to specify security con-
straints, and relies on self-configuration mechanisms to enable a dynamic and
local management. 3
Software-Defined Security Framework Self-configuration enables a lower coupling with respect to
orchestration. Instead of the regular orchestration model addressing requests
and expecting feedbacks, the security orchestrator adopts a passive approach by
exposing security requirements, and letting the PDP to interpret them, accord-
ing to their enforcement contexts. In addition, the framework has been designed
to fit with distributed cloud constraints, in particular the following ones: – multi-tenancy, corresponding to the characteristic for a cloud infrastructure
to be subdivided into different sets of isolated cloud resources called tenants. With that isolation comes the need of regulated access control between each
tenant of the infrastructure, – multi-cloud, corresponding to the capability for cloud infrastructures to
collaborate to enable communications and common treatments on their
resources. With a security-oriented point of view, these treatments come with
a security coordination over potentially heterogeneous infrastructures. In doing so, we detail the role and functioning of its different components,
considering a multi-cloud and multi-tenant context, as depicted on Fig. 2. This
figure makes the assumption that each PDP is dedicated to a tenant, which is
a simple interpretation of software-defined security in this multi-tenant context. We consider the existence of a cloud orchestrator in charge of managing cloud
resources. Even though this component is not meant to be a part of the proposed
security framework, its supposed existence allows taking into account the changes
on cloud resources, which can be done manually by a system administrator or
automatically by one or several potential orchestrators. M. Compasti´e et al. 52 Fig. 2. SDSec framework interacting with a cloud orchestrator, in a multi-cloud multi-
tenant scenario. (1) accounts for the TLSP fetching, (2) for the security statement, (3)
for the enforcement feedback and (4) for the policy decision request. Fig. 2. SDSec framework interacting with a cloud orchestrator, in a multi-cloud multi-
tenant scenario. (1) accounts for the TLSP fetching, (2) for the security statement, (3)
for the enforcement feedback and (4) for the policy decision request. 3.1
Security Orchestrator Amongst the framework components, the security orchestrator is responsible for
the management of the GSP, its interpretation (TLSPs) and distribution. This
policy is meant to be enforced on the distributed cloud, and so, on multiple
collaborating cloud infrastructures with different tenants. The interpretation is
influenced by feedbacks provided by the enforcement. In line with the XACML
terminology [15], the security orchestrator can be seen as a Policy Administra-
tion Point (PAP) allowing the storage of the global policy and generating TLSPs. The changes operated on the global security policy must be propagated to the
whole enforcement perimeter. Contrary to the cloud orchestrator, the security
orchestrator is not meant to manage cloud resources. Consequently, the instan-
tiation, the removal or the reconfiguration of cloud resources is not endorsed by
the security orchestrator. However, this highlights the need for the security orchestrator and the cloud
orchestrator to collaborate. For instance, the security orchestrator requires to
be noticed in case of deployments of new cloud resources, in order to enforce
the security policy on them. In the same manner, the cloud orchestrator must
remove a cloud resource and reconfigure its workflow, when the security orches-
trator requests its removal for security purpose. This collaboration is modeled
on Fig. 2 by the double arrow between the two orchestrators on the leftmost
plane. An overview of the activity diagram of the security orchestrator is given Towards a Software-Defined Security Framework 53 Fig. 3. Overview of the security orchestrator activity diagram Fig. 3. Overview of the security orchestrator activity diagram on Fig. 3. The orchestrator does not push the TLSPs to the PDPs for privacy
purposes, the multi-tenancy property implying the isolation of tenants amongst
each others and with the cloud administrator. These TLSPs must be attached to
meta-datas to enable PDPs to fetch only the policies they are concerned to, by
discriminating each TLSP according to enforcement context criteria. The policy
must be exposed through a dedicated interface accepting incoming connections
from PDPs (with the use of the security discovery protocol). Another interface
assumes the reception of all PDP enforcement feedbacks. The determination of
the exposed TLSPs (as well as the notification sent to the cloud orchestrator) is
correlated to the GSP, the PDP feedbacks and the notifications potentially sent
by cloud orchestrator. 3.2
Policy Decision Points The Policy Decision Point (PDPs) play a central role in this software-defined
security framework, serving as intermediates between the security orchestrator
and the PEPs enforcing policies on resources. More precisely, the PDPs are in
charge of fetching and hosting the TLSPs using the policy security discovery pro-
tocol, and locating their PEPs by invoking the enforcement discovery protocol. Moreover, they support the interactions with PEPs by collecting their feedbacks
and responding to security requests in according to the hosted TLSPs. Accord-
ing to the XACML terminology [15], the PDPs assume different roles: the role of
PDPs providing authorization decisions, but also the role of PAPs with respect
to TLSPs, and the role of PRPs (Policy Retrieval Points). PDPs must take into
account external informations modulating the interpretation of their TLSPs. For
instance, time-regulated access control policy requires an access point to a sys-
tem clock, as this parameter cannot be generalized to all PDPs of the enforced M. Compasti´e et al. 54 perimeter, it is necessary that the PDP proposes an extensible interface able to
communicate with third-party security information providers. In the XACML
terminology [15], these third-party resources are assimilated to Policy Informa-
tion Points (PIPs). Besides, the PDPs maintain several meta-datas describing
their decisional capabilities, which are directly related to their enforcement con-
text. These meta-datas are important for the tenant-level security policy dis-
covery. Consequently, the security statements intended to the PEPs is directly
related to the stored TLSPs, modulated by the preceding feedbacks generated
by the PEPs, and eventually, by the PIP contents. 3.3
Policy Enforcement Points The Policy Enforcement Points (PEPs) are in charge of the enforcement of
TLSPs for a dedicated cloud resource. More precisely, a cloud resource refers
to an instantiated resource on a cloud infrastructure (i.e. a virtual machine,
a service, a set of files, a network function). The considered enforcement con-
sists in (1) the control and modification of security parameters on the resource
according to security statements and (2) the insertion of security event hooks to
handle with state changes and prepare associated security decisional requests. Besides, these objectives correspond the ones defined by the XACML for PEPs. Consequently, the PEPs must expose an interface to the PDP for receiving secu-
rity statements, and be able to contact the PDPs to return feedbacks (after
the execution of a security statement or after an event hook) and to transmit a
security decisional request. The configuration of security parameters is directly
dependent on received security statements. The feedbacks are defined based on
received security statements, states of considered security parameters and event
hook states. Security decisional requests are emitted by PEPs based on event
hook states. 3.4
Interactions Amongst Components The interactions amongst the software-defined security framework components
is supported by different protocols. The Security Policy Discovery Protocol is
a discovery protocol invoked by a PDP to discover the security orchestrator
and fetch a TLSP. The discovery process takes as inputs the PDP meta-data,
and gives back the required TLSPs. Because of the criticality of this protocol,
its specification must integrate technical measures to protect the integrity of
information and remain tamper-proof. In addition, the Enforcement Discovery
Protocol enables a PDP to discover available PEPs in its enforcement perimeter,
and so, to quantify its enforcement capabilities. More precisely, these capabil-
ities are expressed by available PEPs through their enforcement meta-datas,
and brought back to the PDP which determines their potential contributions
to the security enforcement. To prevent security policy information leaks to an
intruder or to prevent an intruder to weaken the security enforcement by pro-
viding false security assessment feedbacks, the protocol must enable the PDP to
verify the authenticity of the discovered PEPs. The Security Statement Protocol Towards a Software-Defined Security Framework 55 enables the PDPs to generate security statements, and send them to PEPs in
their enforcement perimeters. The feedback must be emitted asynchronously, in
case of enforcement statement execution time-out. Hence, to provide a reactive
enforcement, it must be able to emit new feedbacks, when a correctness of a
previously executed security statement changes. Finally, the Security Decision
Requesting Protocol offers to the framework its dynamic enforcement proper-
ties. Indeed, this protocol enables the PEPs to solicit the PDPs for handling a
security decision. This security decision request occurs when a security hook of
a PEP is triggered and verification of the issued security statement cannot be
handled by the PEP itself. The security of these different protocols is out of the
scope of this paper, but is of course a mandatory to guarantee the security of
the whole framework. 4
Framework Evaluation In order to analyze and validate our proposed framework, we have confronted it
to a set of scenarios based on a realistic use case, corresponding to a Cloud Service
Provider (CSP) proposing a Platform-as-a-Service (PaaS) solution to customers,
based on world-wide infrastructures. The multi-tenancy corresponds to the use
of the same infrastructure by several independent customers, while the multi-
cloud property comes from the world-wide location of cloud infrastructures. To
protect its solution, the CSP enforces a security policy on its own infrastructure,
and on its client instantiated cloud resources. In that context, we will consider
the case of a customer, deploying two virtual machines (VM) for hosting two
web applications: one for the European version of his application and one for
the American one. 4.1
Validation Scenarios The scenarios make the following assumptions: the CSP has implemented every
business process in the cloud orchestrator, each customer request is endorsed by
the cloud orchestrator, the customers are unable to remove the PEPs of its cloud
resources, no connection error occurs between PEPs and PDPs, the deployment
of software stacks in the PaaS resources is governed by the cloud orchestrator and
embeds the related PEPs, the cloud resource manager comes with its own PEP
which is managed by the tenant PDP. We have analyzed a set of five scenarios:
the deployment of a new system instance for a customer, the security policy
update by a CSP, a DDoS attack to an instantiated cloud VM, an inter-resource
access request, and the removal of a VM instance. Resource Instantiation Scenario. The customer sets up a dedicated server
associated to his tenant to synchronize and back up the informations of the
instances of his web application. The virtual machines hosting its web applica-
tions are Linux-powered, embeds a SSH server for administrative tasks and a
web server. The chosen technical solution consists in using a SQL server and a M. Compasti´e et al. 56 FTP server in a dedicated VM stored in the European infrastructure, which will
accept connections from the two web application servers. The cloud orchestrator
processes the deployment of these two services with their respective PEPs and
notifies the security orchestrator. As FTP and SQL are newly deployed services
in the tenant, the security orchestrator assumes that the TLSP of the tenant
PDP is not adapted anymore, and modifies the exposed TLSP to this PDP. The
PDP discovers the two new PEPs, fetches the newly available TLSPs from the
security orchestrator, and sends the security statements to the PEPs. Finally,
the PDP transmits a positive enforcement feedback to the security orchestrator. This prevents the security orchestrator to request the cloud orchestrator to take
counter-measures against the tenant. Security Policy Update Scenario. The CSP security administrator enforces
the security of its infrastructure, by restricting the access of critical services only
to the local network and the CSP VPN. The criticality of a service is not defined
in the GSP, but is delegated to the PDP. After the update of the GSP, the PDP
of each tenant detects and collects updated TLSPs. All the PDPs interpret their
TLSPs into security statements restricting the critical service access. 4.1
Validation Scenarios As this module has no precedent records
of hashes, it concludes that the transmitted credentials are newly created and
are allowed to be used. The PDP responses to both security decision requests
are positive, and incoming connections are authorized by respective PEPs. European infrastructure and the production server located in the USA, by using
SQL and FTP transactions: the production server authenticates to the backup
server using a dedicated password. When the production server connects to the
back-up server, the connection attempts trigger the connection hooks of PEPs
related to SQL and FTP servers. Both of them block temporarily the connection
attempts, and make decision requests to the PDP, providing hashes of used
credentials. As the TLSP imposes the verification of the credential lifetime, it
uses its third party module to check it. As this module has no precedent records
of hashes, it concludes that the transmitted credentials are newly created and
are allowed to be used. The PDP responses to both security decision requests
are positive, and incoming connections are authorized by respective PEPs. Resource Removal Scenario. The client wants to update the virtual machine
supporting the American web application by proceeding to a fresh installation. To meet this objective, the client wants to completely remove it and reconfigure
a new virtual machine. He uses the cloud orchestrator to remove this virtual
machine, which is notified to the security orchestrator. The security orchestra-
tor updates its GSP, to take into account the removal of the cloud resource and
checks its consequences on the enforcement: the TLSP is updated. The PDP of
the customer fetches the new TLSP, and stores it. Through the Business Orches-
trator, the security orchestrator starts deallocating resources to the American
VM and the PEP addresses a security decisional request to its PDP for allowing
the removal. According to its TLSP, the PDP grants the request. The PEP lets
the cloud orchestrator to complete the resource removal. This analysis shows that all the presented scenarios can be addressed by our
proposed software-defined security framework. However, some limitations with
respect to the considered use case should be highlighted. First, the use case has
dealt with a GSP set by one security orchestrator. The case of multiple security
administrators, with different enforcement parameters is an addressable issue as
well although we still can abstract it through the single security orchestrator
case. 4.1
Validation Scenarios Second, the use case assumes that one PDP is allocated to one tenant,
corresponding to one customer. This is however only one possible interpreta-
tion of the multi-tenancy notion, but other ones would have made the use case
unnecessarily more complex. 4.1
Validation Scenarios The PDP
associated to the consider customer has deduced that all SSH and SQL servers
were critical. It requests their PEPs to restrict their access and notifies the
security orchestrator of the effective enforcement. If one of the PDPs receives a
PEP negative feedback and has no other counter-measure to apply, it notifies the
security orchestrator which will in turn notify the cloud orchestrator to disable
vulnerable services. Resource Evolution Scenario. The virtual machine in charge of the Euro-
pean version of the web application hosting is targeted by a Distributed Denial
of Service (DDoS) attack. An alert is generated by the PEP to the PDP, indi-
cating the resource consumption is higher than a threshold (initially specified
by the PDP). Consequently, the PDP activates a counter-measure by temporar-
ily increasing the resources allocated to the customer. As this counter-measure
is not efficient, the PDP informs the security orchestrator of its inability to
enforce the GSP. The security orchestrator then relies on the security enforce-
ment stack dedicated to the network infrastructure to perform investigation and
block attacker IP addresses. It requests the tenant PDP to switch the affected
VM into a fail-safe mode. Once the DDoS attack has been countered, the secu-
rity orchestrator reverts back the TLSP exposed to the customer in order to
restore the attacked VM state. Access Request Scenario. The cloud service provider has defined in its GSP
that the used credentials for the connections amongst cloud resources have a
limited lifetime, and have to be regularly changed. The verification of the validity
is committed by the PDP using a third-party module. Meanwhile, the client has
set-up an automatic back-up process between the backup server hosted in the Towards a Software-Defined Security Framework 57 European infrastructure and the production server located in the USA, by using
SQL and FTP transactions: the production server authenticates to the backup
server using a dedicated password. When the production server connects to the
back-up server, the connection attempts trigger the connection hooks of PEPs
related to SQL and FTP servers. Both of them block temporarily the connection
attempts, and make decision requests to the PDP, providing hashes of used
credentials. As the TLSP imposes the verification of the credential lifetime, it
uses its third party module to check it. 4.2
Implementation Considerations After reviewing validation scenarios to evaluate the consistency of our frame-
work, we are discussing in this subsection implementation considerations. Cloud Environment. Before considering a software-defined security stack for
our framework, we focus on the environment and the resources we want to
enforce. We address distributed cloud infrastructure security. The retained tech-
nical solution should be a proven solution in the multi-tenancy area as well as
the multi-cloud one. Moreover, as arisen in the third validation scenario, some M. Compasti´e et al. M. Compasti´e et al. 58 of the countermeasures are likely to rely on infrastructure configuration. This
highlights the need for an extensible cloud stack embedding add-on mechanisms. In both cases, the OpenStack cloud suite is an attractive solution, as it supports
multi-tenancy through the users and region management, and the main compo-
nents of this suite provide plug-in managers. Considering the orchestration, we have to distinguish the need of a security
orchestrator based on a security policy ruling, and a regular cloud one whose
actions are driven by customer solicitation or CSP management tasks. The first
one will be further analyzed in the next subsection. The second has no specific
security expectation except its capability to handle cloud orchestration notifi-
cations, and reciprocally emits notification to it. These two requirements are
related to common orchestrator features as both are linkable to basic messaging
between cloud appliances, each one issuing a request to the other and waiting for
a feedback. Therefore, no more prerequisite other than distributed cloud support
is expected from them. In the cloud resource area, our framework is designed to be resource agnostic
in the sense that the PEPs are the only agents of the architecture depending
on cloud resources. Their interactions are based on resources programmability,
inspection and event handling. Those common features could arise particular
interests the more they are related to dynamic and complex resources. In this
context, virtual machines operating systems and applications are well-suited for
exploring this kind of enforcement, but cannot be generalized as the only type
of resources to be protected. Besides, their nature directly influences the way
PEPs are implemented: an executable cloud resource opens the debate about
whether the PEP should be totally, partially or not at all included in it while a
non-executable one excludes it. Framework Components. Considerations are also raised by the implemen-
tation of the framework itself. 4.2
Implementation Considerations The security orchestration is the component
responsible for the coordination of the PDPs with each others and the cloud
infrastructure (through the cloud orchestrator). As such, it is a highly critical
single point of failure in charge of supervising several tenants and infrastruc-
tures. Such a criticality raises technical issues about redundancy or distribution
among the infrastructure, but also policy concerns such as handling enforce-
ment state transition due to GSP modification: if the modification process is not
properly handled, as cloud tenant-level security policy and cloud-resource state-
ment are not instantly propagated (due to network or processing overhead), we
can conceive that a subset of resources of the cloud infrastructure managed by
the security orchestrator to be trapped into a inconsistent security state. This
eventuality must urge the orchestrator to check the consistency of intermediate
enforcement stated, at the infrastructure level (resource enforcement state can
conflicts) and at the policy-decision level (concurrent low-level security policy
can as well conflicts). Moreover, the privacy concerns is risen with the PDP. Indeed, it can access
all the PEPs it is in charge of, and any data leak may allow an attacker to collect Towards a Software-Defined Security Framework 59 resource data or metadata. Incidentally, the confidentiality of the communication
between PEPs and PDPs is as critical as the isolation between PDPs is. This
statement decides the question of the relation between PDPs and tenants. To
enforce a correct isolation between PDPs, it is necessary that none of them
address several tenants. Otherwise, one tenant could compromise a multi-tenant
PDP, and use-it to fetch data from the other tenant resources. Finally, the variability of the resources this security framework addresses the
enforcement leads to the question of PEP design. Building one PEP for each
type of resource to enforce a TLSP in a cloud is not a sustainable approach as
the workload for a sufficient enforcement coverage would go too far. Thus, we
should consider a more generic approach allowing an automatic adaptation to
cloud resource. A model-driven design and instantiation of PEP is a interesting
response element as the core logic of the PEP could be specified in the model,
before being compiled and adapted on-the-fly to the specificities of the resource
to protect. 5
Conclusions We have proposed in this paper a software-defined security framework for pro-
tecting distributed cloud. It relies on the programmability of software-defined
security, and exploits the autonomic paradigm for addressing the constraints
induced by multi-tenancy and multi-cloud properties. We have detailed the dif-
ferent components of this framework, including a security orchestrator, policy
decision points (PDPs) and policy enforcement points (PEPs) interacting accord-
ing to a dedicated set of protocols. Based on the specification of a security pol-
icy, our framework supports the dynamic configuration of security mechanisms to
adjust to contextual changes, based on available resources and counter-measures. Autonomic methods also enable a lower coupling with respect to orchestration. We have evaluated the proposed solution and discussed implementation consider-
ations, through a set of validation scenarios corresponding to a realistic use case. The proposed solution has raised several challenges with respect to the design of
the considered components, and the specification of security policies in a multi-
cloud and multi-tenant context. The PEPs will apply model-driven scheme to
facilitate the interoperability of heterogeneous enforcements. In the longer term,
the security policy specification of distributed cloud, and the dedicated access
mode will be investigated to complement the security orchestration. 4.2
Implementation Considerations Moreover, such an approach could eventually take advantage of the
cloud resource build environment: if this PEP design and integration process is
able to extract the required information from cloud resources being constructed,
it would lead to an automatic and adaptive design of PEPs tied to cloud resource
dynamics. 1. Mell, P., Grance, T.: The NIST Definition of Cloud Computing (2011)
2. Cloud Security Alliance: Top Threats to Cloud Computing v1. White Paper (2010) References M. Compasti´e et al. 60 3. Compasti´e, M., Badonnel, R., Festor, O. He, R., Kassi-Lahlou, M.: A software-
defined security strategy for supporting autonomic security enforcement in distrib-
uted cloud. In: Proceedings of the 2016 IEEE International Conference on Cloud
Computing Technology and Science (CloudCom), PhD Track, Short Paper, pp. 464–467, December 2016 4. Waller, A., Sandy, I., Power, E., Aivaloglou, E., Skianis, C., Mu˜noz, A., Ma˜na, A.:
Policy based management for security in cloud computing. In: Lee, C., Seigneur,
J.-M., Park, J.J., Wagner, R.R. (eds.) STA 2011. CCIS, vol. 187, pp. 130–137. Springer, Heidelberg (2011). doi:10.1007/978-3-642-22365-5 16 (
)
/
5. Bessani, A., Cutillo, L.A., Ramunno, G., Schirmer, N., Smiraglia, P.: The TClouds
platform: from the concept to the implementation of benchmark scenarios. ACM
SIGOPS Oper. Syst. Rev. 48(2), 13–22 (2014) ( )
(
)
6. Dreo, G., Golling, M., Hommel, W., Tietze, F.: ICEMAN: an architecture for secure
federated inter-cloud identity management. In: Proceedings of the IFIP/IEEE
International Symposium on Integrated Network Management (IM 2013), May
2013 7. Runsewe, O.A.: A policy-based management framework for cloud computing secu-
rity. Master’s thesis, University of Ottawa (2014) 8. Fayazbakhsh, S.K., Sekar, V., Yu, M., Mogul, J.C.: Flowtags: enforcing network-
wide policies in the presence of dynamic middlebox actions. In: Proceedings of the
Second ACM SIGCOMM Workshop on Hot Topics in Software Defined Network-
ing, pp. 19–24. ACM (2013) 9. Koorevaar, T.: Dynamic enforcement of security policies in multi-tenant cloud
networks. Master’s thesis, ´Ecole Polytechnique de Montr´eal (2012) 10. Hurel, G., Badonnel, R., Lahmadi, A., Festor, O.: Behavioral and dynamic security
functions chaining for android devices. In: Proceedings of the 11th IFIP/IEEE/In
Assoc. with ACM SIGCOMM International Conference on Network and Service
Management (CNSM 2015) (2015) 11. Park, J., Jeong, J., Kim, H.: Software-Defined Networking Based Security Services
using Interface to Network Security Functions, October 2015 12. Ball, T., et al.: Vericon: towards verifying controller programs in software-defined
networks. In: Proceedings of 35th ACM SIGPLAN International Conference Pro-
gramming Language Design (PLDI 2014), Edinburgh, UK, pp. 282–293 (2014) 13. Al-Shaer, E., Al-Haj, S.: FlowChecker, configuration analysis and verification of
federated OpenFlow infrastructures. In: Proceedings of the 3rd ACM Workshop
on Assurable and Usable Security Configuration (CCS 2010) (2010) (
) (
)
14. Kephart, J.O., Chess, D.M.: The vision of autonomic computing. Computer 36(1),
41–50 (2003) 15. References Godik, S., Moses, T., Anderson, A., Parducci, B., Adams, C., Flinn, D., Brose, G.,
Lockhart, H., Beznosov, K., Kudo, M., et al.: EXtensible Access Control Markup
Language (XACML) version 1.0 (2003) 16. Maler, E., et al.: Assertions and Protocols for the OASIS Security Assertion
Markup Language (SAML). OASIS, September 2003 17. Golbeck, J.: Trust on the World Wide Web: a survey. Found. Trends Web Sci. 1(2),
131–197 (2006) 18. Waltermire, D., Quinn, S., Scarfone, K., Halbardier, A.: The technical specification
for the security content automation protocol (SCAP): SCAP version 1.2. NIST
Spec. Publ. 800, 126 (2011) Towards a Software-Defined Security Framework 61 61 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. Optimal Service Function Chain Composition
in Network Functions Virtualization Andr´es F. Ocampo1, Juliver Gil-Herrera1, Pedro H. Isolani2, Miguel C. N
Juan F. Botero1(B), Steven Latr´e2, Lisandro Zambenedetti3,
Marinho P. Barcellos3, and Luciano P. Gaspary3 Andr´es F. Ocampo1, Juliver Gil-Herrera1, Pedro H. Isolani2, Miguel C. Neves3,
Juan F. Botero1(B), Steven Latr´e2, Lisandro Zambenedetti3,
Marinho P. Barcellos3, and Luciano P. Gaspary3 1 University of Antioquia, Cl. 67 #53 - 108, Medell´ın, Colombia
{andres.ocampop,juliver.gil,juanf.botero}@udea.edu.co
2 University of Antwerp - imec, Middelheimlaan 1, 2020 Antwerp, Belgium
{pedro.isolani,steven.latre}@uantwerpen.be
3 Federal University of Rio Grande do Sul, Paulo Gama, 110, Porto Alegre, Brazil
{mcneves,granville,marinho,paschoal}@inf.ufrgs.br 1 University of Antioquia, Cl. 67 #53 - 108, Medell´ın, Colombia
{andres.ocampop,juliver.gil,juanf.botero}@udea.edu.co Abstract. Network Functions Virtualization (NFV) is an emerging ini-
tiative where virtualization is used to consolidate Network Functions
(NFs) onto high volume servers (HVS), switches, and storage. In addi-
tion, NFV provides flexibility as Virtual Network Functions (VNFs) can
be moved to different locations in the network. One of the major chal-
lenges of NFV is the allocation of demanded network services in the
network infrastructures, commonly referred to as the Network Functions
Virtualization - Resource Allocation (NFV-RA) problem. NFV-RA is
divided into three stages: (i) Service Function Chain (SFC) composition,
(ii) SFC embedding and (iii) SFC scheduling. Up to now, existing NFV-
RA approaches have mostly tackled the SFC embedding stage taking the
SFC composition as an assumption. Few approaches have faced the com-
position of the SFCs using heuristic approaches that do not guarantee
optimal solutions. In this paper, we solve the first stage of the problem
by characterizing the service requests in terms of NFs and optimally
building the SFC using an Integer Linear Programming (ILP) approach. Keywords: Network
Function
Virtualization ·
Virtual
Network
Functions · Service Function Chain · VNFs chain composition Keywords: Network
Function
Virtualization ·
Virtual
Network
Functions · Service Function Chain · VNFs chain composition c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 62–76, 2017.
DOI: 10.1007/978-3-319-60774-0 5 1
Introduction Network Functions Virtualization is an emerging network management framework
for service deployment, which allows Network Functions to be allocated onto gen-
eral purpose servers [2]. It enables to dynamically compose chains of Virtual Net-
work Functions and embed them anywhere in the network according to a prede-
fined objective. For instance, network functions such as firewalls, load balancers,
and deep packet inspection systems can be placed at the most appropriate location
in the network to support users demand, Quality of Service (QoS) requirements, or
management needs. NFV has grabbed the attention from industry because it has Optimal Service Function Chain Composition in NFV 63 the potential to reduce both CAPEX and OPEX, by the dynamic deployment of
VNFs to commodity hardware, avoiding vertically integrated solutions. In acad-
emia, NFV has been a hot topic because there are interesting technical challenges
to be overcome [1–3], such as the NFV allocation problem [4,5]. To better understand our proposed model in later sections, we introduce the
most important terms used throughout this paper. Network Service (NS): It is an offering provided by an operator that is delivered
using one or more network functions. Network service is a complete, end-to-
end functionality provided by the network operator, such as network protection. A network service may comprise one or more VNFs, for example, a firewall, a
deep packet inspector (DPI), and a data monitor, as in the case of a network
protection system. Virtual Network Function (VNF): It is a function responsible for a specific treat-
ment of data flows. A VNF can act at various network layers of the protocol
stack. As a logical component, a VNF can be realized as a virtual element or be
embedded in a physical network appliance. One or more VNFs can be embedded
in the same physical element. Service Function Chain (SFC): It is an ordered or partially ordered set of VNFs. The implied order may not be a straight line, since the architecture allows SFCs
that send traffic to more than one branch, and also allows cases where there is
flexibility in the order in which VNFs need to be applied. SFCs may be unidi-
rectional or bidirectional, depending on the state requirements of the network
functions. Many common functions such as DPI and firewalls often require bidi-
rectional chaining in order to ensure that the flow state is consistent. 1 http://www.etsi.org/deliver/etsi gs/NFV/001 099/003/01.02.01 60/gs NFV003v010201p.pd 1
Introduction An SFC,
in ETSI’s terminology, is called VNF Forwarding Graph (VNF-FG)1. Efficient network services require the optimal allocation of resources in NFV
(NFV-RA), a challenging problem [5]. A chain of VNFs must be intelligently
composed and allocated to the infrastructure to provide end-to-end QoS guaran-
tees for the applications. However, given the VNFs dependencies, the allocation
is extremely challenging. The above mentioned NFV-RA problem can be sub-divided into three sub-
problems: (i) SFC composition, (ii) SFC embedding and (iii) SFC scheduling. Due to the fact that several chains can fulfill the same NS, the order of VNFs
is often flexible; that is, some VNFs have to be placed in a specific order (e.g.,
the network flow first has to be decrypted before it can be further processed),
while others are flexible in that regard (i.e. they don’t depend from one another). Therefore, the composition of the best possible chain (SFC composition) for each
NS is very important for the operator. However, SFC composition has been so
far overlooked by the scientific community, typically taken as an assumption. Besides, to the best of our knowledge, previous solutions are heuristic in nature
and, therefore, do not guarantee optimal solutions. In this paper, our contributions are twofold: (1) we propose a way to formally
describe network services as a set of VNFs considering the dependences among A.F. Ocampo et al. 64 them and how they can be concatenated in SFCs and (2) we propose an ILP-
based approach to optimally solve this sub-problem by characterizing the service
requests in terms of virtual network functions and solving the SFC composition
problem. The remainder of this paper is organized as follows. In Sect. 2, we describe the
main approaches tackling the NFV-RA problem. In Sect. 3, we define the SFC
composition problem. Section 4 specifies our ILP formulation in detail. Section 5
presents the performance evaluation of our proposed approach. Finally, in Sect. 6,
we conclude the paper with final remarks and perspectives for future work. 2
Related Work The majority of current NFV-RA approaches starts from the assumption that
the chain of VNFs has been already composed, i.e., the important stage of SFC
composition is taken for granted. Few approaches have been proposed to solve
the SFC composition stage so far. Mehraghdam et al. [7] formulate a context-free
language for formalizing chaining requests. They propose a greedy heuristic that
tries to minimize the total data rate of the resulting chain by composing first
the VNF that reduces the data rate of the flows in each step. Recently, Beck and
Botero [1] proposed a scalable recursive heuristic that, at each step, composes
a VNF in the service chain and, at the same time, embeds it in the substrate
network (SN) trying to rapidly find a feasible solution. Most of the existing NFV-RA approaches deal just with the embedding stage
as they consider the VNF-FG as a given input of the problem [5]. For instance,
Bari et al. [4] propose exact Integer Linear Programming (ILP) and heuristic
based approaches trying to minimize the OPEX caused by the SFC embed-
ding. Also, Elias et al. [3] formulate the SFC embedding as a non-linear integer
optimization model where the objective function is to minimize the network
congestion. The aforementioned review shows that, up to now, little research has been
performed in the composition stage of the SFC problem. Current solutions are
heuristic-based and no optimal solution for the problem has been proposed so
far. An optimal solution results in the best possible composition of the service
chain with regard to a predefined objective. In this paper, we propose an optimal
approach to solve the problem based on Integer Linear Programming. 3
The SFC Composition Problem When allocating resources for a given NS, service providers receive a chain of
VNFs and apply an embedding strategy for placing and linking these functions
on the physical substrate. Despite being automatic, this process is rigid for clients
and service providers, respectively. While clients must deal with complex func-
tion dependencies when specifying services, providers are not able to structure
the chains in order to find the best fit for their infrastructure. The result is the Optimal Service Function Chain Composition in NFV 65 allocation of suboptimal chains, which may require many more VNF instances
or network bandwidth than necessary, leading to high costs and expenditures. Although some dependencies and connections among VNFs in a service must
be considered, the order of the VNFs (i.e., the structure of the chain) is often
flexible. For example, normally there is no explicit dependency between a leakage
prevention system and a traffic shaper or between a proxy server and a WAN
optimizer. As a consequence, it is possible for several different chains to fulfill the
same service. We call the problem of finding the most appropriate VNF chain,
given a network service specification and a set of resource constraints, the SFC
composition problem. Figure 1(a) shows our proposal on how a network service specification (i.e., a
Virtual Network Functions Request - VNFR) looks like. Instead of providing the
VNF chain structure as a whole, clients have to inform only the necessary infor-
mation for allowing network service providers to derive the best chain according
to some predefined goal (e.g., to minimize the number of NF instances or the
bandwidth demand). Fig. 1. VNF chain composition (Color figure online) Fig. 1. VNF chain composition (Color figure online) Essentially, a VNFR has five elements: (i) the initial data rate of the network
flow (rinit), (ii) the set of VNFs that compose the service, each one with their
respective processing requirement (drel), (iii) the VNFs where the flow initiates
(ninit) and terminates (nterm), (iv) a number of outgoing links (solid purple
arrows) at each VNF, and (v) mandatory dependencies (dotted blue arrows). Outgoing links can be used to represent scenarios where traffic is split (e.g.,
bifurcations). If a VNF has more than one outgoing link, then it splits the traffic
flow into the same number of sub-flows. For example, consider a load balancer or
a DPI server separating TCP and non-TCP traffic. 3
The SFC Composition Problem Each sub-flow has a relative A.F. Ocampo et al. 66 traffic rate (rrel), which can be higher than 100% if the function replicates or
encodes traffic. Dependencies, in turn, may be of two different types: between a VNF and
an outgoing link, or between two distinct VNFs. Outgoing link dependencies
represent VNFs that should selectively be placed on one of the sub-flows (e.g., a
firewall that succeeds an anomaly-based IDS only for suspicious traffic). Depen-
dencies between VNFs, on the other hand, indicate that the dependent function
must be present in each and every sub-flow in the chain (e.g., the cache servers
succeeding a load balancer). In Fig. 1(b), we represent two possible chains for the service described in
Fig. 1(a). Notice that VNFs 2 and 3 selectively appear in the sub-flows of VNF
1, while VNF 4 is present in both subpaths. Moreover, bandwidth and processing
demands are determined according to the relative traffic of each outgoing link. Although structurally similar, the left chain (i.e., VNF-FG 1) requires less net-
work and computing resources, which at the end results in lower costs for both
clients and providers. As an outcome of the SFC composition problem, VNF-FG
1 would be sent for embedding in the provider infrastructure. Table 1 details the notation used in the model proposed in the following
section. The first part of the table introduces the parameters of a Network Service
Request. The second part explains the sets of nodes of an augmented graph used
to build the ILP model. Finally, the table shows the ILP variables. Before jumping into the next section, it is worth mentioning that the outcome
of the chain composition stage is one complete service chain (VNF-FG) with
regard to one predefined objective and that the amount of required capacities
depends on the amount of data handled by that VNF instance. Optimal Service Function Chain Composition in NFV Optimal Service Function Chain Composition in NFV 67 Table 1. SFC composition inputs and variables
V NF R(V, L)
Is the service request formed by V VNFs and L VNF links
V
Set of VNFs
L
Set of VNF links; this is the set of all links coming out the different VNFs belonging to
the VNFR
Li
out ⊂L
Determines the (outgoing) VNF links of the VNF i ∈V ; a VNF with multiple links
splits the network flow into several sub-flows
Sets
Nterm ⊂V
VNFs where the service terminates
rinit : ZZZ
Initial data rate of the VNFR
rrel(i) : V →ZZZ
Total traffic (in percentage) departing from node i ∈V
ninit ∈V
Initial VNF of the service
drel(i) : V →RRR
Relative processing capacity demands of the VNF i ∈V
rrel(i, b) : L →ZZZ
Relative link traffic rate of link (i, b) ∈L, here i ∈V
identifies the link source VNF
and b is the link number of i
Virtual Network Functions Request (VNFR)
req(i) : V →L∗
Dependencies of the VNF i ∈V ; defined as the incoming edges of a VNF and refer to
outgoing links of other VNFs. This allows the specification of VNFs that get selectively
deployed on specific sub-flows. An assignment of a VNF instance is only valid if traffic
has first been routed through instances of all the required VNFs. Parameters
MI(i)
Minimum number of instances of the VNF i ∈V in the VNF-FG
Gext = (V ext, Lext) This augmented graph is created from the VNFR. 2 For the sake of clarity, directed links are drawn with arrows, so a link with arrows
in its extremes a and b represents a pair of directed links; one from a to b, and the
other from b to a. Optimal Service Function Chain Composition in NFV The final service chain (VNF-FG) will
be a subgraph of Ga
V ext
Set of nodes of the augmented graph, each node in V may have one or more instances
in V ext
Lext
Set of links of the augmented graph
Sets
P
Set of paths from the node in V ext that correspond to the instance of ninit ∈V to the
set of instances of the terminating nodes Nterm in the augmented graph im ∈V ext :
i ∈Nterm
P osp
im
Position of the augmented node im ∈V ext in path p ∈P , if im is not part of the
path, then 0
Augmented Graph
δp
im,jn
Binary parameter that indicates if augmented link (im, jn) ∈Lext is part of path p
Parameters
yi,b
im,jn
Binary variable that says whether the link (i, b) ∈L is mapped in the link (im, jn) ∈
Lext of the augmented network
yim,jn
Binary variable that says whether the link (im, jn) ∈Lext of the augmented network
is chosen as a part of the resulting VNF-FG
xm
i
Binary variable that says whether the instance im ∈V ext of the augmented network
is part of the resulting VNF-FG
LDi,b
im,jn (BW )
Bandwidth required by link (i, b) ∈L assigned to (im, jn) ∈Lext in the augmented
network
LDim,jn (BW )
Total bandwidth required by (im, jn) ∈Lext in the augmented network
T Dim (BW )
Total bandwidth arriving to node (im) in the augmented network
zi,b
im,jn
Auxiliary
variable
to
perform
the
following
product
between
variables zi,b
im,jn
=
T Dim (BW ) · yi,b
im,jn
Variables
yi,b,p
im,jn
Binary variable that says whether the link (i, b) ∈L : i ∈V k is mapped in the link
(im, jn) belonging to the path p of the augmented network
up
Binary variable that says whether the path p ∈P is used in the augmented network Table 1. SFC composition inputs and variables 4.1
Augmented Graph To solve the SFC composition problem, we propose an ILP model that is built
based on an augmented graph created from the VNFR as follows: – For the first VNF (ninit) of the VNFR, one node is placed in the augmented
directed graph Gext = (V ext, Lext); – For the first VNF (ninit) of the VNFR, one node is placed in the augmented
directed graph Gext = (V ext, Lext); – For each of the remaining VNFs, we create as many nodes as the maximum
number of instances that a VNF may have. For example, in Fig. 2, the maxi-
mum number of instances is 2, as the VNF 1 splits the traffic in two sub-flows. To ease the notation, the node im ∈V ext denotes the m-th instance of the
node i ∈V ; – Then, we place directed links between each pair of nodes of the graph except
for: • Instances of the same VNFs: (im, jn) ∈Lext, ∀im ∈V ext, jn ∈V ext ⇐⇒
i ̸= j; • Instances of the same VNFs: (im, jn) ∈Lext, ∀im ∈V ext, jn ∈V ext ⇐⇒
i ̸= j; ̸
• Links directed to ninit: (im, jn) ∈Lext, ∀im ∈V ext, jn ∈V ext ⇐⇒j ̸=
ninit. • Links directed to ninit: (im, jn) ∈Lext, ∀im ∈V ext, jn ∈V ext ⇐⇒j ̸=
ninit. Table 1. SFC composition inputs and variables Binary parameter that indicates if augmented link (im, jn) ∈Lext is part of path p
aram Figure 2, where ninit =VNF1, shows how an augmented graph (see Fig. 2b2)
is created from a VNFR (see Fig. 2a). Figure 2, where ninit =VNF1, shows how an augmented graph (see Fig. 2b2)
is created from a VNFR (see Fig. 2a). Our ILP model is based on the fact that any possible chaining is a subgraph
of the augmented graph, so the ILP variables (cf. Table 1) mainly denote which
nodes and links of the augmented graph are considered to be parts of the chain,
and the demands they will have due to the chosen chaining. Figure 3a shows two
possible chains that can be created out of VNFR in Fig. 2a, and how they are
present in the augmented graph (see Fig. 3b). 2 For the sake of clarity, directed links are drawn with arrows, so a link with arrows
in its extremes a and b represents a pair of directed links; one from a to b, and the
other from b to a. 68
A.F. Ocampo et al. Fig. 2. VNFR and augmented graph
Fig. 3. VNFR and augmented graph 68 A.F. Ocampo et al. Fig. 2. VNFR and augmented graph Fig. 2. VNFR and augmented graph Fig. 2. VNFR and augmented graph Fig. 2. VNFR and augmented graph Fig. 3. VNFR and augmented graph Fig. 3. VNFR and augmented graph 4.2
ILP Formulation Equation 6 ensures that an outgoing link (i, b) ∈L of the VNFR should be
mapped in an augmented link of the augmented graph just for one path:
p∈P
yi,b,p
im,jn ≤1 : (im, jn) ∈Lext, (i, b) ∈L
(6) (6) Establishment of yim,jn: yim,jn =
(i,b)∈L
yi,b
im,jn ≤1 : (im, jn) ∈Lext
(7) (7) yim,jn ≤xm
i : (im, jn) ∈Lext
(8)
yim,jn ≤xn
j : (im, jn) ∈Lext
(9) yim,jn ≤xm
i : (im, jn) ∈Lext
(8) yim,jn ≤xm
i : (im, jn) ∈Lext yim,jn ≤xm
i : (im, jn) ∈Lext
(8) (8) yim,jn ≤xn
j : (im, jn) ∈Lext
(9) (9) Constraints 7, 8, and 9 establish the belonging of a link of the augmented
graph (im, jn) ∈Lext to the resulting service chain (VNF-FG). Node Mapping Constraints: 4.2
ILP Formulation The following is the ILP formulation of the chain composition problem. That is
to say, the ILP that models how to insert a VNFR in the augmented graph with
respect to a predefined objective (e.g. minimize the number of VNF instances of
the resulting VNF-FG). Constraints: Link Mapping Constraints: yi,b
im,jn ≤
p∈P
yi,b,p
im,jn : (im, jn) ∈Lext, (i, b) ∈L
(1)
H · yi,b
im,jn ≥
p∈P
yi,b,p
im,jn : (im, jn) ∈Lext, (i, b) ∈L
(2) (1) H · yi,b
im,jn ≥
p∈P
yi,b,p
im,jn : (im, jn) ∈Lext, (i, b) ∈L
(2) (2) Optimal Service Function Chain Composition in NFV 69 yi,b,p
im,jn ≤δp
im,jn : (im, jn) ∈Lext, (i, b) ∈L, p ∈P
(3) (3) Equations 1, 2, and 3 indicate the relationship between variables yi,b
im,jn and
yi,b,p
im,jn (here H is a big number that enforces the binary constraint). They indicate
that the augmented link (im, jn) ∈Lext can map the outgoing link (i, b) ∈L
when it is mapped using a predefined path p ∈P. Remember that the same
outgoing link (i, b) ∈L may be mapped in several links (im, jn) ∈Lext. For
example, in Fig. 1a, in the VNFR, the outgoing link of VNF 4 is mapped in two
different links between VNF 4 and VNF 5 in the VNF-FG 1 (see Fig. 1b). Equations to ensure that if a path p is mapped to the augmented graph, all
its links have to be assigned.
(im,jn)∈Lext
(i,b)∈L
yi,b,p
im,jn ≤H · up : p ∈P
(4) (4)
(i,b)∈L
yi,b,p
im,jn ≥δp
im,jn · up : (im, jn) ∈Lext, p ∈P
(5) (5) Equation 4 ensures that if a path is not mapped in the augmented graph, then
variable yi,b,p
im,jn should be zero for all links belonging to that path. Equation 5
ensures that if path is mapped in the augmented graph, then the variable yi,b,p
im,jn
should be one for all links belonging to that path. Dependencies fulfillment constraints:
(im,j,n)∈Lext
yi,b
im,jn ≥1 : l ∈V k, l ̸= ninit, (i, b) ∈req(l)
(12) (12) Constraint 12 ensures that for each node l ∈V , all the dependencies are
mapped. Node Mapping Constraints: Node Mapping Constraints: Establishment of xm
i : xm
i =
(im,jn)∈Lext
yi,b
im,jn : i ∈V k, 1 ≤m ≤M i, (i, b) ∈Li
out
(10) (10) A.F. Ocampo et al. 70 Lower bound in the possible number of instances for i: Mi
m=1
xm
i ≥MI(i) : i ∈V k
(11) (11) Constraints 10 and 11 establish the belonging of a node im ∈V ext in the
resulting service chain (VNF-FG). Dependencies fulfillment constraints: Dependencies fulfillment constraints: Position constraints: yi,b,p
im,jn · Posp
im ≤Posp
lr : l ∈V k, l ̸= ninit, (i, b) ∈req(l),
lr ∈Vext, (im, jn) ∈Lext, p ∈P, δp
im,jn ̸= 0
(13) (13)
lr∈Vext
p∈P :lr∈P
(im,jn)∈Lext
yi,b,p
im,jn ≥1 : l ∈V k, l ̸= ninit, (i, b) ∈req(l) (14 Constraint 13 ensures the precedence of the dependencies. If one VNF
instance l ∈V is mapped in the augmented graph, then the set of its dependent
links should be mapped in a prior position in the path going from ninit to l. Constraint 14 ensures that the path being used to map the VNF’s dependencies
also contains an instance of the VNF. Incoming links for each xm
i :
(jn,im)∈Lext
yjn,im ≥xm
i : i ∈V k, 1 ≤m ≤M i, i ̸= ninit
(15)
(jn,im)∈Lext
yjn,im ≤H · xm
i : i ∈V k, 1 ≤m ≤M i
(16) t
yjn,im ≥xm
i : i ∈V k, 1 ≤m ≤M i, i ̸= ninit
(15) (15)
(jn,im)∈Lext
yjn,im ≤H · xm
i : i ∈V k, 1 ≤m ≤M i
(16) (16) Equations 15 and 16 state that if a link of the augmented graph is part of
the resulting service chain (VNF-FG) then the end nodes of this link should be
part of the chain too. The following Equations to establish zi,b
im,jn linearize the following product
i b The following Equations to establish zi,b
im,jn linearize the following product
zi,b
im jn = TDim(BW) · yi,b
im jn. zi,b
im,jn = TDim(BW) · yi,b
im,jn. zi,b
im,jn = TDim(BW) · yi,b
im,jn. Position constraints: zi,b
im,jn ≤yi,b
im,jn · H : (im, jn) ∈Lext, (i, b) ∈L zi,b
im,jn ≤yi,b
im,jn · H : (im, jn) ∈Lext, (i, b) ∈L
(17) (17) Optimal Service Function Chain Composition in NFV
71 zi,b
im,jn ≤TDim(BW) : (im, jn) ∈Lext, (i, b) ∈L
(18) zi,b
im,jn ≥TDim(BW) −(1 −yi,b
im,jn) · H : (im, jn) ∈Lext, (i, b) ∈L
(19) zi,b
im,jn ≥TDim(BW) −(1 −yi,b
im,jn) · H : (im, jn) ∈Lext, (i, b) ∈L
(19) Constraints to set demands: Constraints to set demands: LDi,b
im,jn(BW) = rrel(i, b)zi,b
im,jn : (im, jn) ∈Lext, (i, b) ∈L
(20)
LDim,jn(BW) =
(i,b)∈L
LDi,b
im,jn(BW) : (im, jn) ∈Lext
(21) TDjn(BW) =
(im,jn)∈Lext
LDim,jn(BW) : jn ̸= ninit ∈Vext
(22)
TDn1
init(BW) = rinit · xn1
init
(23) TDjn(BW) =
(im,jn)∈Lext
LDim,jn(BW) : jn ̸= ninit ∈Vext
(22) (22) TDn1
init(BW) = rinit · xn1
init
(23) (23)
(im,jn)∈Lext t
LDim,jn(BW) = TDim(BW) · rrel(i) : im ∈Vext, i /∈Nterm (24) (24) These set of equations simply set the bandwidth demand of each link in
the augmented graph and also the complete load received by each node in the
augmented graph. Here, H is just a big number to force binary variables to take
0 or 1 values. 5
Performance Evaluation In this section, a performance evaluation of the ILP is presented. Our evalu-
ation focuses primarily on minimizing the total bandwidth demanded by the
constructed service chain (VNF-FG). Three scenarios are configured in order to
analyze our ILP model following typical cases for service chaining [8]. 5.1
Simulation Scenario The ILP model for SFC composition presented in the previous section was imple-
mented in the Gurobi solver [6] which provides an exact solution. To the best
of our knowledge, just one work [7] has dealt separately with the composition
phase of NFV-RA. This work heuristically tries to allocate those VNFs with
flexible order following an ascending order according to their ratio of outgoing
to incoming data rate. Here, we compare our ILP-based exact model with this
heuristic proposal. Simulations are performed for three VNFRs based on typical use cases of
networks chains [8]. Figure 4 illustrates the simulation settings for each scenario. The first request VNFR 1 is given for service with NAT64 functions where the A.F. Ocampo et al. 72 Fig. 4. Virtual network function requests Fig. 4. Virtual network function requests Fig. 4. Virtual network function requests traffic is processed by a subchain (composed of VNF 1, VNF 2 and VNF 3),
then the NAT function (VNF 4) for IP capabilities (e.g., mapping from IPv6
to IPv4), and then processed by another subchain (VNF 5). The first subchain
could have VNFs with optional order, so a good planning of such functions in
the final VNF-FG would impact the entire network performance. traffic is processed by a subchain (composed of VNF 1, VNF 2 and VNF 3),
then the NAT function (VNF 4) for IP capabilities (e.g., mapping from IPv6
to IPv4), and then processed by another subchain (VNF 5). The first subchain
could have VNFs with optional order, so a good planning of such functions in
the final VNF-FG would impact the entire network performance. The second scenario (VNFR 2) follows the structure of a service chain used
to split service paths where service providers enable content awareness. VNF
1 splits the traffic into two sub-flows through two outgoing links. On the one
hand VNF 2 is disposed at the sub-flow of VNF 1 to process 60% of its outgoing
traffic; on the other hand, VNF 3 is located at the second sub-flow of VNF 1 to
process 40% of its outgoing traffic. VNF 4 has to be processed by both sub-flows
and, subsequently, the final function of the VNFR is VNF 5. Finally, VFNR 3 is given for scenarios in the Gi Interface for mobile network
environments. We define an scenario with seven VNFs disposed as follows: VNF
1 is the initial function to be performed, this function divides the incoming
traffic into three sub-flows through three links with relative bandwidth demands
of 20%, 40% and 40% respectively. VNF 2 depends on the first sub-flow while
VNFs 3 and 4 depend on the second sub-flow and VNFs 5 and VNF 6 depend on
the last sub-flow. Finally, Each sub-flow must to be processed by the terminal
function VNF 7. Optimal Service Function Chain Composition in NFV 73 5.2
Results The solution of our model is given in terms of a VNF-FG to be embedded on
the physical network. The main objective is to find a VNF-FG with the minimal
bandwidth demand on its links. Therefore, the objective function of our ILP is
to minimize:
(im,jn)∈Lext
LDim,jn(BW)
(25) (25) In order to validate the effectiveness of our solution, we compare it with the
heuristic mentioned before, in terms of the total bandwidth demanded by VNF-
FG, that is, the sum of all bandwidth demands on each link of the VNF-FG. In order to validate the effectiveness of our solution, we compare it with the
heuristic mentioned before, in terms of the total bandwidth demanded by VNF-
FG, that is, the sum of all bandwidth demands on each link of the VNF-FG. Fig. 5. Total bandwidth demanded by VNF-FG Fig. 5. Total bandwidth demanded by VNF-FG Figure 5 shows the comparison between the ILP model and the heuristic in the
aforementioned scenarios. On the one hand, for scenarios where the traffic is split
out into several links (e.g. VNFR 2), our solution yields better results than the
VNF-FG of the heuristic, demanding around 5 Mbps less of the total bandwidth. On the other hand, for scenarios such as VNFR 1 as well as on each bifurcation
of VNFR 3, where all the possible chains are disposed in a monotonic order, i.e.,
following and straight concatenation, both solutions yields same results in terms
of demanded bandwidth. For VNFRs splitting the traffic into several bifurcations (e.g., scenarios
VNFR 2 and VNFR 3), the ILP model would be able to obtain VNF-FGs with
more than one instance of the same VNF for those cases in which such function
must be performed by each sub-flow. This is the case of VNFR 2: for instance,
where VNF 4 has to be performed by each sub-flow, our solution establishes that
it must be implemented in two separately instances (one per sub-flow). Thus,
the load of traffic arriving at VNF 4 is divided requiring less processing device
capabilities into the physical network before the embedding process. As shown
in Fig. 6b, VNF 4 in our ILP solution VNF 4 is created with two instances with
incoming traffic loads of 400 Mbps and 600 Mbps respectively whereas the heuris-
tic solution maps VNF 4 to the same instance for both sub-flows processing an A.F. Ocampo et al. 5.2
Results 74 (a) VNFR 1
(b) VNFR 2
(c) VNFR 3
Fig. 6. Total bandwidth arriving to VNFs ( ) VNFR 1
(b) VNFR 2 (b) VNFR 2 (c) VNFR 3
Fig. 6. Total bandwidth arriving to VNFs Fig. 6. Total bandwidth arriving to VNFs incoming load traffic of 1 Gbps. The fact that VNFs are created in more than
one instance would facilitate the subsequent embedding phase of NFV-RA. Similarly, in VNFR 3, VNF 7 has to be located after all three sub-flows. Our
ILP solution generates two instances of this function; instance 1 receives a load
traffic of 400 Mbps from one sub-flow, while instance 2 receives a traffic load of
600-Mbps from two sub-flows, as shown in Fig. 6c. In VNFR 1 where the VNF-FG in any combination is a monotonic graph
without bifurcations, only one instance of each function is mapped. Figure 6a
illustrates the traffic arriving at each VNF; for both solution the results were
exactly the same. Summarizing, results show that, for the simulated scenarios, our ILP model
always performs better or equal than the evaluated heuristic proposal [7]. Specif-
ically, the proposed ILP model provides better behavior when VNFRs present
traffic bifurcations as it results in less total demanded bandwidth than the heuris-
tic approach. Also, our solution creates several instances per VNF when bifurca-
tion is present in the VNFR which would ease the subsequent embedding phase Optimal Service Function Chain Composition in NFV 75 of NFV-RA as instances with less arriving bandwidth are easier to be embed-
ded in the substrate network. The mean run time of our model considering all
performed scenarios was 1.33 s, in comparison to the heuristic with a mean run
time of 0.028 s. Also, it is important to note that the objective function of our
ILP was restricted here to the minimization of the total link bandwidth in the
resulting VNF-FG (to be comparable with the existing heuristic). However, this
objective may change depending on the operator’s goals to several ones, such
as: minimization of the number of created instances, minimization of the total
processing capacities, etc. 6
Conclusion and Future Work This paper introduces a novel approach to optimally solve the SFC composi-
tion problem based on Integer Linear Programming. Evaluation results indicate
that the proposed approach outperforms existing heuristic-based approaches. Specially, when bifurcation of VNFs is present in the VNFR, the proposed ILP
model reduces the total incoming bandwidth and creates lighter instances of
VNFs in the VNF-FG in order to facilitate the subsequent embedding phase. Scalability issues of the proposed approach are still to be tested. Simulation
scenarios were based in current IETF drafts that show only small VNFRs. A
evaluation on larger scenarios to test the scalability of the approach is left for
future work. Also, the extension of the ILP model to include the embedding phase
of NFV-RA is an exciting branch of future research. In this way, a coordinated
model including SFC composition and embedding may be created to optimally
solve the two phases of NFV-RA. Acknowledgment. This work has been funded by COLCIENCIAS, the University of
Antioquia and by the Flemish fund for scientific research (FWO) and the EMD and
5GUARDS project, co-funded by imec and VLAIO. 7. Mehraghdam, S., Keller, M., Karl, H.: Specifying and placing chains of virtual net-
work functions. In: 2014 IEEE 3rd International Conference on Cloud Networking
(CloudNet), pp. 7–13, October 2014 8. Liu, W., Li, H., Huang, O., Boucadair, M., Leymann, N., Qiao, F., Qiong, Q.,
Pham, C., Huang, C., Zhu, J., He, P.: Service function chaining (sfc) general use
cases. Internet-Draft draft-liu-sfc-use-cases-08, IETF Secretariat, September 2014 A.F. Ocampo et al. References 1. Beck, M.T., Botero, J.F.: Coordinated allocation of service function chains. In: 2015
IEEE Global Communications Conference (GLOBECOM), pp. 1–6, December 2015 2. Beck, M.T., Botero, J.F.: Scalable and coordinated allocation of service function
chains. Comput. Commun. (2016) 3. Elias, J., Martignon, F., Paris, S., Wang, J.: Efficient orchestration mechanisms for
congestion mitigation in NFV: models and algorithms. IEEE Trans. Serv. Comput. PP(99) (2015) 4. Bari, M.F., Chowdhury, S., Ahmed, R., Boutaba, R.: On orchestrating virtual net-
work functions. In: 2015 11th International Conference on Network and Service
Management (CNSM), pp. 50–56, November 2015 5. Gil-Herrera, J., Botero, J.F.: Resource allocation in NFV: a comprehensive survey. IEEE Trans. Netw. Serv. Manag. 13(3), 518–532 (2016) 6. Gurobi Optimization, Inc.: Gurobi optimizer reference manual (2016). http://www. gurobi.com A.F. Ocampo et al. 76 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 79–93, 2017.
DOI: 10.1007/978-3-319-60774-0 6 An Optimized Resilient Advance Bandwidth
Scheduling for Media Delivery Services Maryam Barshan(B), Hendrik Moens, Bruno Volckaert, and Filip De Turck Department of Information Technology, Ghent University – IMEC,
Technologiepark-Zwijnaarde 15, 9052 Gent, Belgium
maryam.barshan@intec.ugent.be Department of Information Technology, Ghent University – IMEC,
Technologiepark-Zwijnaarde 15, 9052 Gent, Belgium
maryam.barshan@intec.ugent.be Abstract. In IP-based media delivery services, we often deal with pre-
dictable network load and traffic, making it beneficial to use advance
reservations even when network failure occurs. In such a network, to offer
reliable reservations, fault-tolerance related features should be incorpo-
rated in the advance reservation system. In this paper, we propose an
optimized protection mechanism in which backup paths are selected in
advance to protect the transfers when any failure happens in the net-
work. Using a shared backup path protection, the proposed approach
minimizes the backup capacity of the requests while guaranteeing 100%
single link failure recovery. We have evaluated the quality and complex-
ity of our proposed solution and the impact of different percentages of
backup demands and timeslot sizes have been investigated in depth. The
presented approach has been compared to our previously-designed algo-
rithm as a baseline. Our simulation results reveal a noticeable improve-
ment in request acceptance rate, up to 9.2%. Moreover, with fine-grained
timeslot sizes and under limited network capacity, the time complexity
of the proposed solution is up to 14% lower. Keywords: Advance bandwidth reservation · Resilient reservation ·
Fixed timeslot size · Media delivery service 1
Introduction Currently, in the media-centric industries, the distribution of media content is
generally performed by either people transporting the content on a physical stor-
age media or over dedicated point-to-point high-speed optical links. However,
these are highly inefficient and costly methods. In order to support decentral-
ized collaboration, reduce capital expenditures and increase network resource
utilization, media related environments tend to switch to the cost-effective IP-
based WAN approaches. Deploying a shared IP-based WAN solution enables
the existing media content owners and their collaborators to work together in
a cost effective way, while new actors can more easily find new collaboration
opportunities, thus fostering the whole industry’s further growth. As media-centric networks usually offer predictable network traffic, this
knowledge of future transmissions can be exploited to use advance reservation M. Barshan et al. 80 (AR) services. This makes it easier to offer guarantees in advance, improves
the number of admitted requests and increases network utilization. In AR tech-
niques, users submit requests for future data transfers, generally encompassing a
start time in the future, a deadline, and total data transfer size or rate. To allo-
cate the necessary resources (more specifically network bandwidth), a schedul-
ing algorithm is needed to ensure that all admitted requests finish before their
specified deadline, while admitting as many requests as possible. Clearly, AR
has advantages for next generation media related networks: it allows network
operators to better plan resource usage, leading to greatly increased resource
utilization and guaranteed Quality of Service (QoS). Reliability of the transport is also of crucial importance in the digital-centric
media transfer process, when different media actors are geographically located
far apart. Therefore, strategies to deal with network dynamics such as failures
should be defined to enable reliable transmission of accepted requests without
any loss in QoS upon occurrence of a failure. For example, in media production
networks, meeting transfers’ deadlines is of crucial importance. Consider a live
show or a news program which is broadcasted everyday at a specific time. Clearly,
even slight delays in transfer of pre-production contents are intolerable in such
a setting. Media-centric networks impose requirements not supported by existing AR
scheduling techniques, such as different types of video or audio transfers, flexible
or unspecified start or end times, strict deadlines, interdependent requests, reli-
ability, etc. Addressing these unexplored aspects was the main focus of our pre-
vious contributions. 2
Related Work The authors in [3] survey the AR algorithms, mostly focusing on Wavelength-
Division Multiplexing (WDM) networks, and provide a taxonomy for classifying
these algorithms. Advance reservation requests can be classified in 4 individual
categories, which are also valid for different types of requests in media related net-
works. However, based on this survey, only two works offer variable-bandwidth
reservation in their scheduling process [4,5]. While both approaches consider a
fixed start time for the requests, all four classes for requests with specified or
unspecified time and duration are supported in our work. Current research on AR
scheduling mostly focuses on rescheduling [6–8], multi-domain reservations [9], and
real-life deployments [10–13]. Nevertheless, reliability and fault tolerance proper-
ties have not been investigated. Resilient AR systems can be deployed either through restoration or protection
failure recovery mechanisms [14]. In protection approaches, backup resources are
reserved in advance before any failure happens in the network, while in case of
restoration backup resources are found upon failure detection. The former results
in more resource consumption but the recovery time is quite fast. In [15], the
authors propose a restoration technique to deal with link failures. In their work,
the active requests and the scheduled requests for the future which are affected
by a failure are restored. In [16,17], optimal ILP-based solutions were proposed
to provide shared and dedicated path protection. Authors in [18,19] also provide
resiliency through shared path protection. Since meeting strict deadlines and
QoS requirements is of great importance in our approach, we have made use of
protection mechanisms. The work detailed in this paper presents a significant optimization to our
previous works on bandwidth reservation approaches [1,2]. In [1], we proposed a
theoretical Integer Linear Programming (ILP) based model and heuristic algo-
rithms for advance bandwidth reservation with no support for failure recovery. In [2], the media production reservation system is enhanced by following a pro-
tection mechanism and provisioning backup reservations for each request. This
resilient solution guarantees 100% recovery against any single link failure. While
this approach strives to minimize the needed bandwidth for the backup paths
and determination of allocations is fast it does not fully utilize the network
capacity. This work aims at optimizing the resilient solution proposed in [2] and
improving the request admittance ratio by reducing wasted network capacity
and thus improving network utilization. 3
Background We briefly explain about the specific properties of media delivery networks and
provide a summery of the resilient advance bandwidth reservation approach. 3.1
Media Delivery Networks An Optimized Resilient Advance Bandwidth Scheduling An Optimized Resilient Advance Bandwidth Scheduling 1
Introduction First optimal and near optimal AR scheduling algorithms,
customized for media production networks, have been proposed [1]. To offer reli-
able reservations, we have further presented the resilient version of our approach
based on a protection mechanism to improve the reliability of the AR systems [2]. The proposed scheme is capable of covering single link failures using pre-reserved
disjoint backup paths. Additionally, the resilient solution improves the scheme’s
availability compared to the non-resilient approach. Continued research has shown us that the resilient bandwidth allocation algo-
rithm, in [2], can be further improved and this is the main contribution of the
work presented in this paper. This algorithm is optimized to reduce network
reservation waste by proposing a more efficient solution to finding an optimal
allocation of bandwidth for each file transfer request. We have made a tradeoff
between the complexity and performance of the resilient bandwidth allocation
algorithm for file-based transfers. This results in better network utilization and
consequently higher request admittance ratio. The remainder of this paper is structured as follows. In Sect. 2, we discuss
the related work. Section 3, provides brief information about the media deliv-
ery services and elaborates on the resilient AR approach. The proposed solution
to improve the performance of resilient AR scheduling approach is described in
Sect. 4. Designed algorithms are explained in Sect. 5. Section 6 provides simula-
tion results and finally, Sect. 7 concludes the paper. 81 3.1
Media Delivery Networks In the digital-centric media related industries, various actors are connected to
a shared wide area network to build a collaboration over IP media contents. M. Barshan et al. 82 This shared network supports the exchange of different media contents, e.g. raw
and encoded video and audio files and streaming transfers. We refer to each
transfer as a request. A request can have a fixed or unspecified start and end
times. Media delivery requests, supported in our work, are of 4 different classes:
independent streaming requests, independent file transfers, dependent streams
and dependent file based transfers. The requests of independent type can be
started at the specified start time but dependent requests have to wait until
the requests upon which they are dependent have finished. Dependency among
different transfers implies that either all or none of the interdependent requests
must be admitted. We refer to a set of interdependent requests as a scenario. We assume that volume for file-based requests, and duration for streaming
requests duration must be specified. The allocated bandwidth for the streams
must be equal to their required bandwidth demand, from the start time to the end
time, because their demand is fixed. However, for file-based requests, the volume of
file is the determinative factor. The file can be transferred whenever possible from
the time the file is ready to be transferred till its deadline. The residual demand
of file-based transfers is modified whenever a part of the file is transferred. 3.2
Resilient AR Scheduling In order to have a quick response to sudden changes such as failures in the net-
work, we use a protection mechanism which finds backup paths for connections
in advance, before the occurrence of any failure to ensure there is enough capac-
ity left when failures occur. The objective is to minimize the resource usage by
the protection paths while full recovery is guaranteed against any single link fail-
ure. In this scheme, first the primary paths for a given request are determined
using an advance reservation algorithm which we presented in [1]. Then disjoint
backup paths are found corresponding to these primary paths [2]. Note that in
the proposed schemes, the user can indicate the amount of required backup for
each request. This way, the higher priority requests can be fully protected while
the ones with lower priority can remain partially protected or unprotected. Using shared backup path protection (SBPP) [19] and multi-path approaches,
to provide full protection against a single link failure, the backups have to provide
the maximum bandwidth allocated on the links of the primary paths. In the
resilient approach, when bandwidth is allocated to a request, we look for disjoint
paths to be reserved as backup paths for that request. In practice, a request
may not ask for full recovery of bandwidth demand upon occurrence of a failure
and it is sufficient that a portion of the request demand is transmitted to the
destination. Therefore, we compare the maximum primary allocation and the
amount of requested backup and select the smaller value as the limit to be
fulfilled by the backups. If this backup limit can not be found, it backtracks to
the initial state and retries the bandwidth allocation with half of the primary
bandwidth demand. This division by 2 is repeated until both primary and backup
demands are satisfied. If the file can not be accommodated by its deadline, the
scenario to which this request belongs, is rejected and all of its reservations have
to be sent back to the network resource pool. Full details can be found in [2]. An Optimized Resilient Advance Bandwidth Scheduling 83 4
Optimized Resilient AR Scheduling In this section, we elaborate on how the reliable scheduling approach has been
improved to achieve a higher request admittance ratio. In the resilient approach, if the requested backup can not be found, it retries
the primary allocation with fraction of the primary bandwidth demand (50% in
our case). Although this is a fast approach, we found that halving the request
demand does not always lead to an optimal solution because we may miss the
opportunity to transfer a higher volume of the file and the network capacity
may not be fully utilized if other concurrent requests can not make use of it. As such, we propose to make better use of leftover capacities by deploying the
binary search mechanism [20] for finding the maximum value which satisfies
both primary allocation and the requested bandwidth demand. This way, per-
step complexity of the reservations increases while a higher amount of allocations
will potentially be achieved. We elaborate more on this with two examples. Example 1: consider Fig. 2 with a file-based request of 300 Gb and timeslot size
of 5 min. The limit component sets 1 Gbps (300 Gb/300 s) for the limitation of
bandwidth reservations for this request. The resilient approach, first checks if it
can fulfill both a 1 Gbps primary allocation and the requested backup demand. The amount of backup allocation depends on the percentage of requested pro-
tection and also the way the primary reservations are allocated. If the request’s
backup demand can not be fulfilled, the limit of the request is divided by 2. Then the same procedure is repeated for the lower limit of 500 Mbps. If both
500 Mbps primary demand and its corresponding backup are available, the pri-
mary and backup allocations are reserved and the request demand is updated
by reducing reserved network capacity and the rest of the file has to wait for the
next timeslots to be processed. Otherwise, this division by 2 is continued until
the request limit is fulfilled or the file is finally rejected. However, this division
by 2 is not efficient if the network is able to provide e.g. 400 Mbps. Based on this
approach the resilient reservation approach can only provide 250 Mbps. This is
shown in Fig. 2b following a multi-path allocation approach, i.e. 200 + 50 Mbps
for primary and 200 Mbps (to cover single link failure) for shared backup paths. 3.3
Resilient AR Scheduling Architecture In this section, we briefly explain the heuristic-based resilient AR architecture,
detailed in [2], for the reliable bandwidth scheduling problem. There are differ-
ent blocks in this approach, presented in Fig. 1. As can be seen, new scenarios
enter the scheduler through an API which can be transformed using the input
transformation block. Then the resilient scheduling algorithm is invoked which
follows a timeslot-based scheme. A timeslot is defined as a period of time in which
reservations remain invariant. The new scenario is admitted and the schedule is
updated provided that this algorithm succeeds in allocating bandwidth to all
the scenario’s requests. Otherwise, the previous scheduling remains unchanged. Fig. 1. Components of the resilient advance bandwidth reservation approach. Fig. 1. Components of the resilient advance bandwidth reservation approach. In the resilient scheduling component, first the requests within the scenario
are prioritized and then the FixedTimeSlot algorithm is called. In the prioriti-
zation algorithm, the requests’ priorities are assigned first based on the dead-
line, and then the request’s demand. Requests with sooner deadline and higher
demand receive higher priorities. The FixedTimeSlot algorithm iterates over the time slots with 5 sub-modules. (1) TimeSlotRequests: determines the eligible requests which can be served at
the current time slot. (2) Limit: defines a limitation for each request. For the
streams this limit is equal to their demand which is fixed. For file-based requests
the residual demand divided by timeslot size is considered as the limit. (3) Sort-
ing: sorts the requests based on their priorities, assigned by the prioritization
algorithm. (4) BWallocationResilient: responsible for resilient bandwidth alloca-
tion to the requests depending on their types. (5) Update and check for feasi-
bility: Once the required demands are allocated to the requests, the schedule is
updated if all the deadlines are met. Otherwise this schedule is infeasible. Full
details can be found in [2]. M. Barshan et al. 84 4
Optimized Resilient AR Scheduling Although the 150 Mbps can be occupied by other requests, this is not optimal
and may results in a waste of network resources. In order to improve the network utilization, we propose to make use of Binary
Search to find an optimal value for the amount of reservations. This way if the
algorithm recognizes that 500 Mbps is not available but 250 Mbps can be offered,
instead of returning 250 Mbps, the algorithm tries to find the maximum available
value between 250 Mbps and 500 Mbps. The proposed solution first takes the
middle value (375 Mbps) and checks the possibility of reservations again. As this
is available, the algorithm again checks for the middle value between 375 Mbps
and 500 Mbps which is 437.5 Mbps. This trend is continued with 406.25 and then
390.625, etc. until the difference between the upper and lower bounds is smaller
than a given margin, which we refer to as ϵ. Assuming 2 Mbps as this margin the An Optimized Resilient Advance Bandwidth Scheduling 85 Fig. 2. An example of primary and backup reservations based on the original and the
optimized resilient AR approaches. (Black: network capacity, Blue: Primary reserva-
tion, Red and dashed: Backup reservation) (Color figure online) Fig. 2. An example of primary and backup reservations based on the original and the
optimized resilient AR approaches. (Black: network capacity, Blue: Primary reserva-
tion, Red and dashed: Backup reservation) (Color figure online) algorithm stops at 399.4 Mbps. The reservation based on the optimized resilient
AR approach is shown in Fig. 2c. This margin can be altered to make a tradeoff
between achieving a precise optimal value and solution complexity. algorithm stops at 399.4 Mbps. The reservation based on the optimized resilient
AR approach is shown in Fig. 2c. This margin can be altered to make a tradeoff
between achieving a precise optimal value and solution complexity. Fig. 3. Comparing the primary allocations of the original and optimized versions of
resilient timeslot-based advance bandwidth reservation approach. Fig. 3. Comparing the primary allocations of the original and optimized versions of
resilient timeslot-based advance bandwidth reservation approach. Example 2: Figure 3 shows an example of a schedule for 3 file-based requests,
R1, R2 and R3. The timeslot size is 1 h and in both figures only primary
reservations are shown. In Fig. 4
Optimized Resilient AR Scheduling 3a and b, the reservations are made using the
original and optimized version of the resilient advance bandwidth reservation
approaches respectively. Figure 3b reveals how the optimized resilient approach
can improve network utilization and increase the probability of admittance for
future requests. As can be seen, by allocating a higher volume of a given file,
this file can be potentially transferred earlier compared to the original approach. This way, higher capacity is available for requests in future and the request
admittance ratio will be potentially increased. Algorithm 1: BWallocationResilient+ Depending on the backup demands and primary allocations, the amount of
backup demand is first calculated. Both algorithms check if the backup can be
fulfilled. In order to cover single failures, the backups have to be disjointed from
the primary paths. As such, the links used in the primary paths are removed from
the network and the bandwidth allocation algorithms are reused on the residual
network to find the backup paths for that request. If the backups can be found,
the primary and backup paths can be successfully allocated for the request. Otherwise, the primary paths have to be removed. Again if the backups for the
streaming request are not fulfilled, the scheduling is not successful. However,
for file transfers if the backup can not be provided, the algorithm tries with
a lower primary bandwidth demand. This is repeated until both primary and
backup demands of the file are satisfied. If this algorithm is being executed in the
timeslot prior to the request deadline, and both primary and backup demands
can not be fulfilled, the entire scenarios to which the file belongs, is rejected. 5
Resilient Timeslot-Based AR Algorithms In this section, we elaborate on the optimized BWallocationResilient+ algo-
rithm, shown in Algorithm 1, which first assigns a cost to each network link
using a cost allocation module. We have previously designed two algorithms for
resilient bandwidth allocation depending on the type of the request, which we
refereed to as BWallocationFBResilient and BWallocationVSResilient for file-
based and streaming requests respectively. As we have optimized the resilient M. Barshan et al. 86 approach for file transfers, we do not elaborate on the BWallocationVSResilient
algorithm. The common part for both algorithms is repeatedly finding the least-
cost paths between source and destination of a given request until the limit
of that request is fulfilled. However, provided that the limit of the request is
not available, a different trend is followed by each approach. For the file-based
request, maximum available capacity is reserved as the rest of the file can be
processed during the next timeslots. For the streams, if there is not enough
capacity to allocate to the request, it can not be served and thus the feasibility
is set to false. The next step in the resilient algorithms is to find the backup
paths. Data: sortedReqList
costAllocation(Links);
for req ∈sortedReqList do
if req is F B then
reservation←BWallocationFBResilient+(req);
else
reservation←BWallocationVSResilient(req);
end
end
return reservation; Algorithm 1: BWallocationResilient+ Algorithm 2: BWallocationFBResilient for file-based requests Algorithm 2: BWallocationFBResilient for file-based requests finds that value X can satisfy both primary and backup demands, instead of
returning this value, which was the case in the BWallocationFBResilient algo-
rithm, a higher value based on the binary search approach is investigated and
this is repeated until a near-optimal value (within an ϵ margin) is calculated and
returned. This process is shown in Algorithm 3. finds that value X can satisfy both primary and backup demands, instead of
returning this value, which was the case in the BWallocationFBResilient algo-
rithm, a higher value based on the binary search approach is investigated and
this is repeated until a near-optimal value (within an ϵ margin) is calculated and
returned. This process is shown in Algorithm 3. Data: an FB request
currentState ←Save the current network state
upperBound ←Limit(req);
optimalLimit ←Limit(req);
lowerBound ←0;
while optimalLimit > 0.1 do
reservation ←BWallocationFB(req, currentLimit, graph);
maxBW ←max Bandwidth(reservation);
graphReduced ←remove the links in reservation from the network graph;
backupLimit ←min(maxBW, requestedBackup(req));
backupReservation ←BWallocationFB(req, backupLimit, graphReduced);
if !backupReservation then
set current network state to currentState;
upperBound ←optimalLimit;
optimalLimit ←lowerBound+upperBound/2;
else
if upperBound −lowerBound >= ϵ then
set current network state to currentState;
lowerBound ←optimalLimit;
optimalLimit ←lowerBound+upperBound/2;
else
return reservation, backupReservation;
end
end
end
Algorithm 3: BWallocationFBResilient+ for file-based reques upperBound ←Limit(req); optimalLimit ←Limit(req); if upperBound −lowerBound >= ϵ then
set current network state to currentState;
lowerBound ←optimalLimit; if upperBound −lowerBound >= ϵ then optimalLimit ←lowerBound+upperBound/2;
else return reservation, backupReservation;
end Algorithm 3: BWallocationFBResilient+ for file-based requests Algorithm 3: BWallocationFBResilient+ for file-based requests 5.1
BWallocationFBResilient+ Algorithm The main idea behind proposing the BWallocationFBResilient+ is to improve
the performance of the BWallocationFBResilient algorithm. In the BWalloca-
tionFBResilient algorithm, if the backup for a given request can not be provided,
the limit of primary allocations is repeatedly halved and the possibility of the
reservation is checked with this lower limit. We argue that this halving cycle can
be improved by deploying a binary search algorithm. That is, given a file-based
request, we seek for maximum available bandwidth which satisfies both primary
and backup demands. Therefore, if the BWallocationFBResilient+ algorithm An Optimized Resilient Advance Bandwidth Scheduling An Optimized Resilient Advance Bandwidth Scheduling 87 Data: an FB request
currentState ←Save the current network state
currentLimit ←Limit(req);
while currentLimit > 0.1 do
reservation ←BWallocationFB(req, currentLimit, graph);
maxBW ←max Bandwidth(reservation);
graphReduced ←remove the links in reservation from the network graph;
backupLimit ←min(maxBW, requestedBackup(req));
backupReservation ←BWallocationFB(req, backupLimit, graphReduced);
if !backupReservation then
set current network state to currentState;
currentLimit ←currentLimit/2;
else
return reservation, backupReservation;
d return reservation, backupReservation;
end
nd 6.1
Evaluation Setup The network topology used for this evaluation contains 8 nodes and 16 bidirec-
tional links. After discussion with our industrial partners in media production
industry, 3 scenario types are defined: a soccer after-game discussion program,
an infotainment show and a news broadcast program, consisting of 5, 18 and 8
interdependent file-based and video streaming requests respectively. A detailed
overview of the randomized variables of requests and network topology can be
observed from [1] and [21] respectively. In the fixed size timeslot-based solution,
timeslot granularities of 5, 15 and 30 min and backup demand of 50% and 100%
are used. Throughout this section, SARA[XX%, YYmin] denotes that backup demand
of XX% and timeslot size of YY minutes is considered in the fixed-size timeslot-
based advance reservation algorithm. SARA+ refers to the optimized resilient
bandwidth reservation approach. In this approach the margin, which we referred
to as ϵ, equals 2 Mbps. Each simulation run covers a 24-h period. All results are
averaged over 50 runs with different randomized inputs, error bars denote the
standard error. All algorithms in this section are implemented in Java 8. 6
Performance Evaluation This section evaluates the quality and execution time of the proposed solu-
tion, compared to our previously designed resilient timeslot-based scheduling M. Barshan et al. 88 algorithms. For this analysis, SARA (Static Advance Reservation Approach)
is evaluated in which all requests are known in advance, before the start of
scheduling. The influence of the available network capacity, network load, backup
demand, timeslot granularity and execution times are assessed. 6.2
Evaluation Results Evaluation of Network Capacity, Backup Demands and Timeslot Sizes:
In Figs. 4 and 5, the network infrastructure has been configured for different
available bandwidths, respectively for 100% and 50% backup demands to inves-
tigate the impact of available network capacity, backup demands and timeslot
sizes on the performance of our algorithms in terms of percentage of admitted
requests. In both evaluations, 7 scenarios (67 requests) are submitted to the
bandwidth reservation system and the network capacities vary from 50 Mbps to
400 Mbps. What we can observe in these evaluations is as follows: first, the finer the
timeslot size, the higher gain achieved by the SARA+ approach. As can be
observed from Fig. 4, the SARA+ approach is able to achieve on average up to
3.6%, 7.3% and 9.2% higher admittance ratio in 30-, 15- and 5-min timeslot sizes,
respectively. Second, with higher backup demand, the performance of SARA+
is more significant. In Fig. 5, with 50% backup demand, SARA+ is able to out-
perform the SARA approach on average up to 8.5% with 5-min timeslots. Com-
paring Figs. 4c and 5c, the SARA+ with 100% back up demand improves the
request admittance ratio on average up to 2.8 times comparing to the 50% backup
demand. An Optimized Resilient Advance Bandwidth Scheduling An Op 89 Fig. 4. Impact of timeslot size with 100% backup demand in the timeslot-based advance
bandwidth reservation approach. Fig. 4. Impact of timeslot size with 100% backup demand in the timeslot-based advance
bandwidth reservation approach. Fig. 5. Impact of timeslot size with 50% backup demand in the timeslot-based advance
bandwidth reservation approach. Fig. 5. Impact of timeslot size with 50% backup demand in the timeslot-based advance
bandwidth reservation approach. Evaluation of Network Load, Timeslot Sizes and Execution Times:
Figures 6 and 7 compare the influence of the network load and timeslot sizes on
the quality and time complexity of our algorithms. Backup demand of 100% and
network capacity of 200 Mbps are used. p
y
p
As can be seen in Fig. 6, by increasing the number of scenarios, the percent-
age of admitted requests decreases and the SARA+ approach performs better
with fine-grained timeslot sizes. We notice that the advance bandwidth reserva-
tion system gains more by deploying the SARA+ approach and with the 5-min
timeslot size, shows up to 7.3% higher request admittance ratio. The time complexity of the approaches are evaluated in Fig. 6.2
Evaluation Results 7 for an increas-
ing range of scenarios. This figure reveals that the granularity of timeslot size
impacts the execution times of both approaches differently. While with 30-min
timeslot size, the execution time of SARA+ is up to 147 milliseconds higher com-
pared to the SARA approach, with 5-min timeslots, this time is up to 4.5 sec-
ond lower. These results indicate that the quality and complexity of the advance
bandwidth reservation system can be improved by deploying the SARA+ app-
roach with fine-grained timeslot sizes. For further investigation of the execution time, we have assessed the impact of
network capacity on the execution time, when the timeslot granularity of 5 min
is used. This has been shown in Fig. 8. The number of scenarios is 7 and 14 90
M. Barshan et al. Fig. 6. Impact of network load in the fixed size timeslot-based advance bandwidth
reservation approach. 90
M. Barshan et al. Fig. 6. Impact of network load in the fixed size timeslot-based advance bandwidth
reservation approach. M. Barshan et al. M. Barshan et al. 90 arshan et al. Fig. 6. Impact of network load in the fixed size timeslot-based advance bandwidth
reservation approach. Fig. 7. Comparing the execution times in the fixed size timeslot-based advance
bandwidth reservation approach. Fig. 7. Comparing the execution times in the fixed size timeslot-based advance
bandwidth reservation approach. Fig. 8. Comparing the execution times in the function of network capacity in the fixed
size timeslot-based advance bandwidth reservation approach. Fig. 8. Comparing the execution times in the function of network capacity in the fixed
size timeslot-based advance bandwidth reservation approach. in Fig. 8a and b respectively. This evaluation shows that when there is enough
bandwidth capacity available, the SARA approach is able to perform faster
while SARA+ can better manage the time under stressed network conditions,
i.e. limited network capacity. 7
Conclusions In this paper, we have optimized the resilient scheduling algorithms, previously
presented for advance bandwidth reservation in media-centric networks. In the An Optimized Resilient Advance Bandwidth Scheduling 91 original version, for a given file transfer, if both primary and backup demands
can not be fulfilled, the algorithm is repeatedly executed with 50% of primary
demand until both demands are fulfilled or the request is rejected. We proposed
to make use of binary search instead of halving the primary demand and showed
that this optimization improves the performance of the timeslot-based advance
reservation system in terms of request admittance ratio. The impact of available
capacity, network load, timeslot sizes and backup demands is evaluated. Based
on the results, we can conclude that the proposed solution specifically performs
well under limited network capacity and with fine-grained timeslot sizes. The
proposed approach outperform the original one both in terms of the execution
time, with 5-min timeslot size, and the percentage of admitted requests, up to
9.2%. Acknowledgment. This work has been performed within the context of ICON
MECaNO, a project co-funded by iMinds, a digital research institute founded by
the Flemish Government. Project partners are SDNsquare, Limecraft, VideoHouse,
Alcatel-Lucent and VRT, with project support from IWT under grant agreement no. 130646. References 1. Barshan, M., Moens, H., Famaey, J., De Turck, F.: Deadline-aware advance reser-
vation scheduling algorithms for media production networks. Comput. Commun. 77, 26–40 (2016) 2. Sahhaf, S., Barshan, M., Tavernier, W., Moens, H., Colle, D., Pickavet, M.:
Resilient algorithms for advance bandwidth reservation in media production net-
works. In: International Conference on the Design of Reliable Communication Net-
works (DRCN), pp. 130–137. IEEE (2016) 3. Charbonneau, N., Vokkarane, V.M.: A survey of advance reservation routing and
wavelength assignment in wavelength-routed WDM networks. IEEE Commun. Surv. Tutor. 14(4), 1037–1064 (2012) ( )
(
)
4. Burchard, L.-O., Heiss, H.-U., De Rose, C.: Performance issues of bandwidth reser-
vations for grid computing. In: Symposium on Computer Architecture and High
Performance Computing, pp. 82–90 (2003) (
)
5. Naikstam, S., Figueira, S.: Elastic reservations for efficient bandwidth utilization
in lambdagrids. Fut. Gener. Comput. Syst. 23(1), 1–22 (2007) 6. Rajah, K., Ranka, S., Xia, Y.: Advance reservations and scheduling for bulk trans-
fers in research networks. IEEE Trans. Parallel Distrib. Syst. 20, 1682–1697 (2009) 6. Rajah, K., Ranka, S., Xia, Y.: Advance reservations and scheduling for bulk trans-
fers in research networks. IEEE Trans. Parallel Distrib. Syst. 20, 1682–1697 (2009)
7. Xie, C., Alazemi, H., Ghani, N.: Rerouting in advance reservation networks. Com- fers in research networks. IEEE Trans. Parallel Distrib. Syst. 20, 1682–1697 (2009)
7. Xie, C., Alazemi, H., Ghani, N.: Rerouting in advance reservation networks. Com-
put. Commun. 35(12), 1411–1421 (2012) (
)
7. Xie, C., Alazemi, H., Ghani, N.: Rerouting in advance reservation networks. Com-
put. Commun. 35(12), 1411–1421 (2012) (
)
(
)
8. Zuo, L., Zhu, M.M., Wu, C.Q.: Fast and efficient bandwidth reservation algorithms
for dynamic network provisioning. J. Netw. Syst. Manag. (2013) 9. Alazemi, H., Xu, F., Xie, C., Ghani, N.: Advance reservation in distributed multi-
domain networks. IEEE Syst. J. (2013) 10. Guok, C., Engineer, E.N., Robertson, D.: Esnet on-demand secure circuits and
advance reservation system (oscars). Internet2 Joint (2006) M. Barshan et al. 92 11. Gibbard, B., Katramatos, D., Yu, D.: Terapaths: end-to-end network path qos con-
figuration using cross-domain reservation negotiation. In: 3rd International Con-
ference on Broadband Communications, Networks and Systems (BROADNETS),
pp. 1–9. IEEE (2006) 12. Gu, J., Katramatos, D., Liu, X., Natarajan, V., Shoshani, A., Sim, A., Yu, D.,
Bradley, S., McKee, S.: Stornet: integrated dynamic storage and network resource
provisioning and management for automated data transfers. J. Phys.: Conf. References Ser. 331, 012002 (2011). IOP Publishing (
)
13. Sharma, S., Katramatos, D., Yu, D., Shi, L.: Design and implementation of an
intelligent end-to-end network qos system. In: Proceedings of the International
Conference on High Performance Computing, Networking, Storage and Analysis,
SC 2012, USA, pp. 68:1–68:11. IEEE Computer Society Press (2012) 14. Watanabe, T., Omizo, T., Akiyama, T., Iida, K.: Resilientflow: deployments of
distributed control channel maintenance modules to recover SDN from unexpected
failures. In: 11th International Conference on the Design of Reliable Communica-
tion Networks (DRCN), pp. 211–218. IEEE (2015) 15. Tanwir, S., Battestilli, L., Perros, H., Karmous-Edwards, G.: Dynamic schedul-
ing of network resources with advance reservations in optical grids. Int. J. Netw. Manag. 18(2), 79–106 (2008) ( )
(
)
16. Li, T., Wang, B., Xin, C., Zhang, X.: On survivable service provisioning in WDM
optical networks under a scheduled traffic model. In: IEEE Global Telecommuni-
cations Conference, GLOBECOM 2005, vol. 4, p. 5-pp. IEEE (2005) 17. Li, T., Wang, B.: On optimal survivability design in WDM optical networks under
a scheduled traffic model. In: 5th International Workshop on Design of Reliable
Communication Networks (DRCN), p. 8-pp. IEEE (2005) 18. Cavdar, C., Tornatore, M., Buzluca, F., Mukherjee, B.: Dynamic scheduling of sur-
vivable connections with delay tolerance in WDM networks. In: IEEE INFOCOM
Workshops 2009, pp. 1–6. IEEE (2009) 19. Cavdar, C., Tornatore, M., Buzluca, F., Mukherjee, B.: Shared-path protection
with delay tolerance (SDT) in optical WDM mesh networks. J. Lightwave Technol. 28(14), 2068–2076 (2010) 20. Lehmer, D.H.: Teaching combinatorial tricks to a computer. In: Proceedings of
Symposium in Applied Mathematics Combinatorial Analysis, vol. 10, pp. 179–193
(1960) 21. Barshan, M., Moens, H., Volckaert, B.: Dynamic adaptive advance bandwidth
reservation in media production networks. In: IEEE NetSoft Conference and Work-
shops (NetSoft), pp. 58–62. IEEE (2016) An Optimized Resilient Advance Bandwidth Scheduling 93 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. © The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 94–107, 2017.
DOI: 10.1007/978-3-319-60774-0_7 The Evaluation of the V2VUNet Concept
to Improve Inter-vehicle Communications Lisa Kristiana1,2(&), Corinna Schmitt1, and Burkhard Stiller1 1 Communication Systems Group CSG, Department of Informatics IfI,
University of Zürich, Binzmühlestrasse 14, 8050 Zurich, Switzerland
{kristiana,schmitt,stiller}@ifi.uzh.ch
2 The Department of Informatics, National Institute of Technology,
Bandung 40124, Indonesia
lisa@itenas.ac.id 1 Communication Systems Group CSG, Department of Informatics IfI,
University of Zürich, Binzmühlestrasse 14, 8050 Zurich, Switzerland
{kristiana,schmitt,stiller}@ifi.uzh.ch
2 The Department of Informatics, National Institute of Technology,
Bandung 40124, Indonesia
lisa@itenas.ac.id Abstract. Due to the high mobility behavior in inter-vehicle communications
(IVC), packet forwarding among vehicles becomes an important issue. For IVC
in a traditional packet forwarding setting, it was observed that the ratio between
packets received and the packets transmitted is often very low, sometimes less
than 50%. This ratio is highly influenced by the environment, especially by road
topologies and obstructions (e.g., buildings or overpasses). Further influences
encompass the number of driving vehicles on streets offering burdens for the
IVC as well as serving as relay candidates. In order to improve IVC this paper
introduces a Vehicular-to-Vehicular Urban Network (V2VUNet) to overcome
inevitable obstructions and frequent network changes by selecting the best relay
candidate. The V2VUNet implemented was evaluated in an IVC with the focus
on three-dimensional road topologies including overpasses with a different
number of driving lanes. The result shows that the developed V2VUNet pro-
vides about 30% better packet transmission performance compared to traditional
packet transmission in IVC. References If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. 1
Introduction Inter-vehicle networks as a part of Vehicular Ad-hoc Network (VANET) are expected
to support communications with multiple participating vehicles [3]. Thus, information
exchanges in a vehicular network communication require stable and reliable connec-
tions. During packet transmissions and receptions the communication path has to be
maintained in any cases. For IVC in a traditional packet forwarding setting, it was
observed that the Packet Delivery Ratio (PDR) as the indicator of network performance
is often less than 50% due to path failures [19]. These path failures are mainly caused
by the road topology complexity of the environment, such as overpass constructions
and buildings at intersection roads [9]. For inter-vehicle communications, a position-based forwarding scheme is generally
used, since it offers an advantage of not relying on packet broadcasting in its routing
mechanisms. Since position information is already made available, the approach pro-
posed in this paper, the Vehicular-to-vehicular Urban Network (V2VUNet), emphasizes
in filtering unnecessary participant nodes [18] and predicting the routing path based on The Evaluation of the V2VUNet Concept 95 the position information and on the calculated direction information [17]. Therefore, the
packet forwarding scheme is expected to become more efficient. In this research work,
the packet forwarding is tested in two types of road environment models. The first type is
the cross road model and the second type is the parallel road model. Both types reflect
the three-dimensions cases with three coordinate axes x, y, and z. The advance beyond
state-of-the-art in this three-dimensional area is determined by the z axis, which in many
of VANET scenarios is rarely included. In addition, the three-dimensional case in
VANET is significantly influenced by objects placed between signal transmitting
devices. These objects can be a building or an overpass. Therefore, it is important to
investigate the packet forwarding in these three-dimensional environments. Besides a
better performance, the packet forwarding scales best in case of non-safety or
non-real-time applications, which can be considered as delay constraint. Thus, delays
are evaluated here as a less significant characteristic. Another relevant aspect today is the use of Multiple Input Multiple Output (MIMO)
technology in Inter-vehicle Communication (IVC). MIMO is a method to increase the
radio link capacity and becomes a promising solution, since it increases the number of
transmitted data by embedding multiple transmitters and receivers [15]. 2
Related Work Successful communication requires an efficient packet forwarding. Packet forwarding
is considered as efficient, when the packet is broadcasted with a smaller probability of
errors. Packet forwarding for non-safety applications refers to numerous size of data
and is assumed to be distributed at a high rate [1–3]. Therefore, dealing with frequent
topology changes in IVC’s behavior, the packet forwarding is based on the method of
forwarding [2, 5–7]. The first idea is to avoid collision in a dense network [8], thus, the
packet forwarding is designed to reduce the number of relay candidates by restricting
the area of transmission [18]. The second idea is to predict the direction of relay
candidates by calculating the relative direction of a relay and by selecting the candidate
that has the same direction with the destination’s direction [17]. It is obvious that the
relay having the same direction with the destination increases the possibility to prolong
the duration of a connection between communicating vehicles. Thus, in a large city
environment with its road topology and traffic complexity [5], both approaches will be
evaluated to perform a reliable data transmission in VANETs, more specifically in IVC. The two concepts of packet forwarding in IVC are studied in a survey that shows
relevant literatures [4–7, 16]. While the first concept follows an angle-based forwarding
approach, the second concept is defined as direction-based forwarding approach. 1
Introduction This method is
useful to be implemented in a non-safety application, such as for infotainment, since it
requires high data rates and large amount of data interchanges. However, as this work
in this paper here focuses on the network layer, MIMO is better to be evaluated in the
data link layer. Fig. 1. Inter vehicle communication in a three-dimensional road topology Fig. 1. Inter vehicle communication in a three-dimensional road topology In previous works as shown in [17, 18], the evaluation of network performance of
each approach is compared to a traditional location-based routing. This paper ensues to
evaluate both forwarding approaches in V2VUNets by implementing various size of
packets, different network densities, and speed of vehicles. The remainder of this paper
is organized as follows: Sect. 2 describes related work of the packet forwarding model
used. Section 3 introduces the key idea of the vertical angle forwarding scheme being
part of a V2VUNet. Additionally, the evaluation of the V2VUNet is discussed in
Sect. 4, followed by the summary and future work in Sect. 5. L. Kristiana et al. 96 2.2
Direction-Based Forwarding In IVC, vehicles are assumed to move on a predefined path such a straight or inter-
section road. Thus, vehicles can have heterogeneous directions depend on the road
types. In order to cover route loss due to the ‘free’ movement of vehicles, the
direction-based forwarding mechanism involves direction as a weight value to deter-
mine the next relay node. The direction has a dynamic value since it depends on the
road topology and time-based factor. The direction value is calculated based on the
sender and receiver position as illustrated in Fig. 3. Fig. 3. A relative direction of a vehicle Fig. 3. A relative direction of a vehicle 3
V2VUNet Concept The concept used in V2VUNet is an enhancement of the selection method in order to
find the best relay node of available candidates [1, 4]. Previous work of V2VUNet
determines the area restriction of transmission [18] and the path prediction [17]. Both
area restriction and path prediction schemes in V2VUNet utilize the angle measurement
on the same road level, i.e., Horizontal Relative Angle (HRA) measurement and dif-
ferent road levels i.e., Vertical Relative Angle (VRA) measurement. The implemen-
tation of V2VUNet in this work is performed to compare each algorithm in various
parameters. The Evaluation of the V2VUNet Concept 97 2.1
Angle-Based Forwarding An angle-based forwarding mechanism utilizes angle measurement to reduce the area
of transmission. The idea of implementing angle is to locate relay candidates within the
transmission range of a sender S. Thus, under the assumption that the location coor-
dinate of a receiver R is known, the imaginary line is drawn in order to scale the angle
as shown in Fig. 2. The angle-based forwarding mechanism selects one of the relay
candidates as the intermediate relay based on the location where it has the smallest
angle calculation respect to an imaginary line [11, 12]. As illustrated in Fig. 2, S selects
node C as the relay node since C has the smallest angle value. The advantage of this
mechanism suits on the dense network because of the efficient route path in terms of
time [12, 13]. The comparison of existing angle forwarding schemes has been done
in [21]. Fig. 2. Angle forwarding scheme Fig. 2. Angle forwarding scheme Fig. 2. Angle forwarding scheme 97 The Evaluation of the V2VUNet Concept 3.1
Area Restriction Scheme In the area restriction scheme, the V2VUNet operates in two steps. The first step is to
define the HRA with value of 30°, which is intuitively based on the width of a road in
two-dimensional area and the closest distance between two vehicles. The second step is
to adjust HRA based on the available relay candidates position. If the number of relay
candidates is more than one, then the V2VUNet algorithm selects the relay based on the
smallest value of HRA. The algorithm for the area restriction scheme is shown as in
Algorithm 1 (cf. Fig. 4). The previously mentioned two steps are also applied in a
three-dimensional area. Similar to HRA, VRA will be first defined as 30°, which is
indicated as the preliminary angle value based on the transmission range. This 30°
value is then increased gradually as part of the searching mechanism. 98 L. Kristiana et al. Fig. 4. Area restriction algorithm Fig. 4. Area restriction algorithm 3.2
Path Prediction Scheme In this scheme, the HRA is used to predict the direction of relay candidates. The
prediction algorithm is designed to overcome the disconnection possibility due to the
transmission coverage in two-dimensional area. In case of VRA, the algorithm is
designed to encounter the disconnection due to obstruction by the overpass. Further-
more, HRA and VRA in this scheme use the relative direction of each vehicle. This
relative direction provides the actual direction in three-dimensional scheme. The path
prediction scheme is shown in Algorithm 2 (cf. Fig. 5). Fig. 5. Path prediction algorithm Fig. 5. Path prediction algorithm For the sake of a precise prediction, the direction that is used in V2VUNet algo-
rithm determines a relative direction since the direction of each vehicle is measured in
vehicle’s current position. This relative direction is changed whenever a vehicle
changes its position as illustrated in Fig. 3. Thus, the traditional direction calculation
cannot be implemented in this prediction as it is done in [17]. Those two proposed
algorithms are expected to provide a network performance as indicated by high PDR
and low end-to-end (E2E) delays. In order to compare all algorithms implemented in
this work, Table 1 shows a short description of each scheme and all related factors. The Evaluation of the V2VUNet Concept 99 Table 1. Comparison of angle, area restricted, and path prediction forwarding algorithms
Factor
Angle-based
forwarding
Area restriction
Path prediction
Coordinate
location
x-, and y-axis
x-, y-, and z-axis
x-, y-, and z-axis
Weight value
HRA
HRA and VRA
HRA and VRA
Relative direction No
No
Yes
Routing based
Greedy
Greedy
Greedy
Road topology
2D intersection,
highway
3D intersection, 3D
parallel
3D intersection, 3D
parallel As the first factor in this comparison table the coordinate location describes the
coordinate axis which is used in measuring the current location of a node. The second
factor is the weight value which determines the angle schemes, HRA for two-dimensional
area and VRA for three-dimensional area. The third factor is the relative direction which
is added in the path prediction algorithm in order to improve the calculation of the
location coordinate. This relative direction factor is suitable when nodes move randomly
and is useful to indicate the current direction of a node. Greedy routing is used in all
algorithms because greedy routing uses the distance factor to do packet forwarding. Basically the greedy approach work best the many routing protocol mechanisms [10]. 3.2
Path Prediction Scheme The last factor that influences all forwarding schemes is the road topology, which
becomes the main idea of forwarding packet improvement. The angle-based forwarding
scheme is used in two-dimensional intersection, where the direction factor becomes an
important value, and in the highway, where the speed of a vehicle is highly considered. However, it is necessary to consider about the complexity of a road topology. Thus, the
area restriction and path prediction schemes raise the three-dimensional road topology
indicated as three-dimensional intersection (i.e., cross road) and parallel roads. As previously mentioned in the introduction section, this work evaluates and
compares the area restriction and path prediction schemes as the improvement of
traditional greedy routing. 4
Performance and Evaluation The simulation in this work aims to validate the theoretical analysis of the proposed
algorithms in IVC. Two simulation scenarios of a road environment with parallel and
cross road topology are selected. In parallel road topology, the difference in vehicles’
direction is more extreme than in a cross road topology. In a parallel topology, there are
less chance that one vehicle can meet another vehicle once they pass each other. In a
cross road topology, there is a segment of the road that is under another segment of the
road, which could potentially contributes to disconnection at particular moment. In order to obtain a realistic city environment, typical parameters for the influencing
factors are chosen as shown in Table 2. The Network Simulator-3 (NS-3.25) [14] is
used to simulate wireless technologies (i.e., IEEE 802.11p), the routing protocol (i.e.,
Greedy Perimeter Source Routing (GPSR) [10]), the mobility, the road topology, and L. Kristiana et al. 100 the network density. The IEEE 802.11p is a well known technology since it is designed
to cope the frequent topology changing in IVC. During 200 s of simulation time, each
vehicle is expected to run under and on the overpass in the first case, and on the
different road level in the second case. S, R, and I are placed randomly both on two
different road levels and SUMO [20] is used to generate the realistic mobility of each
vehicle. Moreover, the number of S and R are generated equally, which means a
10-vehicle network contains of 5 senders and 5 receivers. The simulation area covers
an environment which involves crossing and parallel overpass scenarios (c.f. Fig. 1) in
order to show many cases in three-dimensional area. y
In previous works, these two algorithms have not been evaluated over various
packet sizes, thus, the packet size for the first evaluation is varied from 1–10 kB,
especially for non-safety applications: a half page of unformatted email is 1 kB, one
typical HTML webpage is 30 kB, 1 min of near-CD quality audio as MP3 or a
2048 1536 (4 megapixel) JPEG photo is 1 MB, to evaluate the size of packet that
can be successfully transmitted in two algorithms. However, this simulation focuses
only on transmitting email, with the size of the packet from 1 to 10 kB. The first result (cf. Fig. 4
Performance and Evaluation 6) shows the PDR of all algorithms when different packet
sizes are applied. This performed evaluation simulates 40 vehicles that move with an
average speed of 40 km/h. The V2VUNet area restriction algorithm gives 20% better
PDR compared to the greedy forwarding scheme which in the figure is indicated as No
V2VUNet. V2VUNet
area
restriction
also
shows
10%
better
compared
to
V2VUNet-path prediction scheme. The showed PDR in overall algorithms decreases as
the packet size increases, which indicates that more participating vehicles and simu-
lation time are required to successfully complete the packet forwarding mechanism. V2VUNet indicates that HRA and VRA weight values have significant impacts in
packet transmission. Table 2. Parameter Settings
Parameter
Unit
Transmission range IEEE 802.11p Up to 300 m
Routing protocols
GPSR
Number of vehicles
20–100
Simulation area
500 m 500 m
Upper road height
20 m
Vehicle velocity
30–70 km/h
Packet size
1 kB–10 kB
Simulation time
200 s
Number of driving lanes
4
DSRC data rates
6 Mbps The second result shows the E2E delay of all algorithms (cf. Fig. 7) when different
data sizes are applied. The E2E delay increases as the packet size increases. This is
because more time are required to transmit packet with bigger size. The E2E delay of
200 ms (highest delay) is indicated by path prediction scheme. This is because the path
finding mechanism in the V2VUNet path prediction scheme requires more time The Evaluation of the V2VUNet Concept 101 Fig. 6. Evaluation of a packet delivery ratio Fig. 6. Evaluation of a packet delivery ratio compared to the area restriction mechanism. In overall E2E delay results, the traditional
forwarding scheme provides lowest delay compared to other two algorithms. This is
because the traditional forwarding scheme does not need an additional mechanism to
perform packet forwarding. Fig. 7. Evaluation of end-to-end delays Fig. 7. Evaluation of end-to-end delays The third set of results shows PDR of all algorithms in various number of partic-
ipating vehicles in cross road scenario (cf. Fig. 8). These various number of vehicles
are used to evaluate V2VUNet schemes when dealing with the network density and
network speed. The V2VUNet-path prediction and V2VUNet-area restriction shows
10% and 20% better PDR, respectively, compared to a greedy forwarding without
V2VUNet. As the network density grows, the PDR decreases in all forwarding
schemes. 4
Performance and Evaluation From all simulation trials that have been performed, the highest PDR that can
be achieved is about 40%. This is due to the path failure that occurs when the com-
municating vehicles are under the overpass. This path failure cannot be avoided since 102 L. Kristiana et al. Fig. 8. Evaluation of packet delivery ratios for varying numbers of vehicles and vehicle speeds Fig. 8. Evaluation of packet delivery ratios for varying numbers of vehicles and vehicle speeds Fig. 8. Evaluation of packet delivery ratios for varying numbers of vehicles and vehicle speeds the overpass disturbs the transmission, thus it will never reach 100% of PDR. The
network density with 60–80 vehicles shows the maximum 40% result at speed
50 km/h. This indicates that 60 vehicles with 30 pairs of S and R are the ‘best’
condition where the packet transmission is performed. However, in high speed mobility (i.e. 50 km/h–70 km/h), the PDR reaches higher
results compared to PDR in low speed mobility (i.e., 30 km/h and 40 km/h). The main
reason for this is that in higher speed mobility, the path reconstruction is even more
possible than maintaining the old path. In this case, V2VUNet area restriction scheme The Evaluation of the V2VUNet Concept 103 provides PDR 10% higher compared to V2VUNet path prediction forwarding scheme. This is caused by the overpass construction which blocks the packet transmission, thus,
it becomes difficult to complete path finding process. The fourth set of results shows E2E delays of all algorithms when various number
of participating vehicles are involved in a cross road scenario (cf. Fig. 9). This E2E
delay reaches 350 ms at 70 km/h speed. The high E2E delay can be caused by two
reasons: the first reason is that the intermediate node which moves in the opposite
direction (e.g., vehicle that changes its direction or turns back), has impact to the
searching mechanism. Thus, the mechanism starts to find a new path and the trans-
mission is delayed because of this reason. The second reason is that the connections
between two vehicles are interrupted or discontinued, when one of the vehicles is
located under the overpass. However, the E2E delays decrease for mobility with higher
speeds i.e. 50 km/h–70 km/h. The similar explanation as in the PDR can also be
applied to explain the E2E delays. 4
Performance and Evaluation The required period of time to find a new path is less
than in the mobility with low speeds (i.e., 30 km/h and 40 km/h). In overall results, the Fig. 9. Evaluation of end-to-end delays for varying numbers of vehicles and vehicle speeds Fig. 9. Evaluation of end-to-end delays for varying numbers of vehicles and vehicle speeds 104 L. Kristiana et al. Fig. 10. Evaluation of packet delivery ratios for varying numbers of vehicles and vehicle speeds g. 10. Evaluation of packet delivery ratios for varying numbers of vehicles and vehicle speed Fig. 10. Evaluation of packet delivery ratios for varying numbers of vehicles and vehicle speeds E2E delays are considered as drawbacks in order to obtain better PDR by applying
V2VUNet. Thus, this becomes an open question. The fifth set of results in Fig. 10 shows the evaluation of PDR in the parallel road
scenario. Here, the path prediction scheme reaches 40%. When compared to other
schemes, PDR of the path prediction is found to be the highest because in parallel road
scenarios the direction of vehicles is predictable i.e., either in the same direction or
opposite direction. Thus, in the parallel scenario, path prediction scheme works well in
predicting the relay candidate’s direction. In overall, the PDR decreases accordingly to
number of participating nodes. This is because of the collision due to the network
density. The sixth set of results shows E2E delays (cf. Fig. 11) in the parallel road scenario. The traditional greedy routing shows the lowest delay compared to other scheme
because the scheme does not include additional searching mechanism as previously
mentioned. However, the path prediction scheme shows reasonable E2E delays of
50 ms as the prediction mechanism works well in the parallel road scenario. 105 The Evaluation of the V2VUNet Concept Fig. 11. Evaluation of end-to-end delays for varying numbers of vehicles and vehicle speeds Fig. 11. Evaluation of end-to-end delays for varying numbers of vehicles and vehicle speeds 5
Summary and Future Work The combination of these two
schemes will be considered as a hybrid scheme, thus, both algorithms in V2VUNet are
expected to improve the packet forwarding scheme depending on the use case. 5
Summary and Future Work This work covers and ensues the evaluation of V2VUNet through a three-dimensional
road topology in a large city. Important parameters of V2VUNet have been evaluated:
packet size, speed, and number of vehicles. The V2VUNet takes into account HRA and
VRA as additional weight values, which are applied in area restriction and path pre-
diction algorithms. The network performance as indicated by PDR and E2E delay
values shows to be reliable in non-safety applications. The PDR in an overall perfor-
mance shows that V2VUNet provides 20% better result compared to traditional routing
algorithms. However, the E2E delays in the overall evaluations are slightly higher than
for traditional routing algorithms. Thus, these E2E delays are considered to determine
the trade-off in V2VUNet, even though a non-safety application is assumed to be a
delay tolerant scheme. Additionally, the path prediction scheme is less suitable to be
adopted in the cross scenario, however, it performs better in the parallel scenario. Thus, L. Kristiana et al. 106 it can be concluded that the V2VUNet path prediction works better in the parallel
scenario, since the direction of vehicles is homogeneous. The V2VUNet area restriction
performs better in the cross scenario, since it restricts the number of relay candidates. Further research in improving V2VUNet, the area restriction, and path prediction
t
b
f
d f
iti
b
d
ti
h
h
th di t it can be concluded that the V2VUNet path prediction works better in the parallel
scenario, since the direction of vehicles is homogeneous. The V2VUNet area restriction
performs better in the cross scenario, since it restricts the number of relay candidates. Further research in improving V2VUNet, the area restriction, and path prediction
concept can be performed for any position-based routing scheme, where the distance
and direction indicate the influencing weight value. The combination of these two
schemes will be considered as a hybrid scheme, thus, both algorithms in V2VUNet are
expected to improve the packet forwarding scheme depending on the use case. it can be concluded that the V2VUNet path prediction works better in the parallel
scenario, since the direction of vehicles is homogeneous. The V2VUNet area restriction
performs better in the cross scenario, since it restricts the number of relay candidates. Further research in improving V2VUNet, the area restriction, and path prediction
concept can be performed for any position-based routing scheme, where the distance
and direction indicate the influencing weight value. References 1. Mauve, M., Scheuermann, B.: VANET Convenience and Efficiency Applications, VANET
Vehicular Applications and Inter-Networking Technology, pp. 81–105. Wiley, Hoboken
(2010) (
)
2. Aoki, M., Fujii, H.: Inter-vehicle communication: technical issues on vehicle control
application. IEEE Commun. Mag. 34(10), 90–93 (2002). New York, NY, U.S.A. doi:10. 1109/35.544327 3. Papadimitratos, P., La Fortelle, A.D., Paristech, M., Evenssen, K., Asa, Q.: Vehicular
communication systems: enabling technologies, applications, and future outlook on
intelligent transportation. IEEE Commun. Mag. 47(11), 84–95 (2009). New York, NY, U. S.A. doi:10.1109/MCOM.2009.5307471 4. Jerbi, M., Senouci, S.M., Meraihi, R., Ghamri-Doudane, Y.: An improved vehicular ad-hoc
routing protocol for city environments. In: IEEE International Conference on Communi-
cations, New York, N.Y, U.S.A., pp. 3972–3979, June 2007. doi:10.1109/ICC.2007.654 5. Tavakoli, R., Nabi, M.: TIGeR: a traffic-aware intersection-based geographical routing
protocol for urban VANETs. In: IEEE 77th Vehicular Technology Conference (VTC
Spring), New York, N.Y., U.S.A., pp. 1–5, June 2013. doi:10.1109/VTCSpring.2013. 6692786 6. Lin, Q., Li, C., Wang, X., Zu, L.: A three-dimensional scenario oriented routing protocol in
vehicular ad-hoc networks. In: IEEE Vehicular Technology Conference (VTC), New York,
N.Y., U.S.A., pp. 1–5, June 2013. doi:10.1109/VTCSpring.2013.6691848 7. Wang, H., Tan, G.Z., Yang, J.X.: An improved VANET intelligent forward decision-making
routing algorithm. J. Netw. 7(10), 1546–1553 (2012) 8. Mohimani, G.H., Ashtiani, F., Javanmard, A., Hamdi, M.: Mobility modeling, spatial traffic
distribution, and probability of connectivity for sparse and dense vehicular ad-hoc networks. IEEE Trans. Vehicular Technol. 58(4), 1998–2007 (2009). New York, N.Y., U.S.A. doi:10. 1109/TVT.2008.2004266 9. Walfisch, J., Bertoni, H.: Theoretical model of UHF propagation in urban environments. IEEE Trans. Antennas Propag. 36(12), 1788–1796 (1988). New York, N.Y., U.S.A. doi:10. 1109/8.14401 10. Karp, B., Kung, H.T.: GPSR: greedy perimeter stateless routing for wireless network. In:
International Conference on Mobile Computing and Networking (MobiCom), pp. 243–254. ACM, New York, August 2000. doi:10.1145/345910.345953 11. Spachos, P., Toumpakaris, D., Hatzinakos, D.: Angle-based dynamic routing scheme for
source location privacy in wireless sensor network. In: IEEE Vehicular Technology
Conference (VTC), New York, N.Y., U.S.A., pp. 1–5, May 2014. doi:10.1109/VTC-Spring. 2014.7022833 107 The Evaluation of the V2VUNet Concept 12. Banka, R.K., Xue, G.: Angle routing protocol: location aided routing for mobile ad-hoc
networks using dynamic angle selection. In: IEEE Premier International Conference for
Military Communications (MILCOM), New York, N.Y., U.S.A., vol. 1, pp. 501–506,
October 2001. doi:10.1109/MILCOM.2002.1180493 13. Huang, C., Chiu, Y., Wen, C.: Using hybrid angle/distance information for distributed
topology control in vehicular sensor networks. c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 108–122, 2017.
DOI: 10.1007/978-3-319-60774-0 8 Towards Internet Scale Quality-of-Experience
Measurement with Twitter Dennis Kergl(B), Robert Roedler, and Gabi Dreo Rodosek Department of Computer Science, Universit¨at der Bundeswehr M¨unchen,
85577 Neubiberg, Germany
{dennis.kergl,robert.roedler,gabi.dreo}@unibw.de
http://www.unibw.de Abstract. At present, Quality of Experience (QoE) measurements are
accomplished by interrogating users for the perceived quality of a ser-
vice they just have used. Influenced by many factors and often limited
by domain or geographical region, this technique has several drawbacks
when a general state of QoE for the internet as a whole is prospected. To achieve such a general metric, we leverage user complaints that we
observe in real-time in social media. Such approaches have been success-
fully applied for the monitoring of specific and single services. We aim
to extend existing methods in order to create an overall metric, define
an internet wide QoE baseline, monitor changes and hence, provide a
context for assessing smaller scale findings against a ground truth. The
contribution of this work is to demonstrate the feasibility of using social
media analysis for generating a meaningful value for quantifying the
actual QoE of the internet. References Sensors 14(11), 20188–20216 (2014). Basel,
Switzerland, doi:10.3390/s141120188 14. NS3. https://www.nsnam.org/. Accessed 28 Jan 2017 15. Moser, S., Behrendt, L., Slomka, F.: MIMO-enabling PHY layer enhancement for vehicular
ad-hoc networks. In: IEEE Wireless Communications and Networking Conference
Workshops (WCNCW), New York, N.Y., U.S.A., pp. 142–14 (2015). doi:10.1109/WC-
NCW.2015.7122544 16. Kristiana, L., Schmitt, C., Stiller, B.: Survey of angle-based forwarding methods in VANET
communications. In: IEEE Wireless Days (WD), Toulouse, France, pp. 1–3, March 2016. doi:10.1109/WD.2016.7461505 17. Kristiana, L., Schmitt, C., Stiller, B.: Predictive forwarding scheme in three-dimensional
vehicular communication scenarios. In: International Conference on Selected Topics in
Mobile and Wireless Networking (MoWNeT), Avignon, France, pp. 1–6, May 2017 18. Kristiana, L., Schmitt, C., Stiller, B.: V2VUNet - a filtering out concept for packet
forwarding decision in three-dimensional inter-vehicular communication scenarios. In: IEEE
27th Annual International Symposium on Personal, Indoor, and Mobile Radio Communi-
cations (PIMRC), New York, N.Y., U.S.A., pp. 1–6, September 2016. doi:10.1109/PIM-RC. 2016.7794600 19. Kolici, V., Oda, T., Spaho, E., Barolli, L., Ikeda, M., Uchida, K.: Performance evaluation of
a VANET simulation system using NS-3 and SUMO. In: IEEE 29th International
Conference on Advanced Information Networking and Applications Workshops (WAINA),
New York, N.Y., U.S.A., pp. 348–353, March 2015. doi:10.1109/WAINA.2015.121 20. Simulation of Urban MObility, SUMO. http://www.dlr.de/ts/en/desktopdefault.aspx/tabid-
9883/16931_read-41000/. Accessed 16 Mar 2017 21. Kristiana, L., Schmitt, C., Stiller, B.: Survey of angle-based forwarding methods in VANET
communications. In: IEEE Wireless Days, New York, N.Y., U.S.A., pp. 1–3, March 2016. https://doi.org/10.1109/WD.2016.7461505 Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appro-
priate credit to the original author(s) and the source, provide a link to the Creative Commons
license and indicate if changes were made. The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly
from the copyright holder. Towards Internet Scale Quality-of-Experience Measurement with Twitter There are many different application domains of the QoE concept resulting
in slightly different understandings. Basically, existing work can be classified
in either concentrating on specific network technologies, e.g., mobile networks
like 3G [20], 4G [28], 5G [22], on specific media (e.g., video [32], voice [12]),
on specific services like Internet Protocol television (IPTV) [31], Mobile Social
Networks (MSN) [6] and YouTube [33], or on the type of service deployment
like cloud services [2] or peer-to-peer networks [11]. Also combinations of the
aforementioned categories are actively researched, e.g., in [5]. [ ]
QoE can be thought of as QoS plus a human factor. This simple defini-
tion might be misleading, as modeling of the human factor is an unresolved
issue, that includes different fields of psychology like cognitive psychology, mem-
ory psychology, and psychophysics [30]. That is why QoE measurements often
include conducting real-world experiments with test persons, asking them in var-
ious ways for their opinion on used services. Using this black box test setting, the
inscrutability of the human mind is bypassed and the aimed value is achieved. The disadvantage is the requirement of strictly controlled testing procedures,
high personnel demand and lack of scalability. In order to address these shortcomings, this paper is about the question of
whether it is possible to turn complaints of globally distributed social media
users into valuable signals for inferring perceived levels of web service quality. With a positive outcome, continuous QoE measurement at large scale would
become feasible and advanced questions might be identified. With Twitter, peo-
ple can publish messages (tweets) using various devices and follow other users to
subscribe to their tweets. The public Twitter Application Programming Inter-
faces (APIs) offer programmatic access to public tweets in a well-documented
JavaScript Object Notation (JSON) format. Twitter is the most widespread
service of its kind and, due to its openness and popularity, current subject of
research in several disciplines covering a broad range of examined topics [34]. In this work, we investigate the feasibility of using tweets as an indicator for
QoE drops. Twitter users are not a representative share of all internet users. More than that, we expect the population to be biased in various ways. This is a
restriction to our approach, which let us detect only complaints about problems
without providing an unbiased base line. 1
Introduction Management and operation of communication and networking services rely on
holistic knowledge of interrelationships between technical values and perceived
service quality. Especially perceived quality of internet services is fundamental
for both developing web applications and planning network infrastructure [1]. The shift from technology-oriented to user-centric development, operation and
measurement correlates with the trends of network architectures that evolve
from host-oriented to information-centric models, and infrastructure networks
from static to Software Defined Network (SDN) technologies [22]. On a technical level, there exist plenty of measured parameters that can
describe characteristics of network links, protocols, connected systems and appli-
cations and form the well-defined Quality of Service (QoS) concept [16]. Although
the International Telecommunication Union (ITU) definition of QoS includes the
ability to satisfy stated and implied needs of the user of the service, the QoS
concept does not provide insights to users’ satisfaction on consumed services. To
close this gap and to provide specific and measurable objectives to application
and infrastructure developers, the concept of QoE was introduced. 109 Towards Internet Scale Quality-of-Experience Measurement with Twitter Towards Internet Scale Quality-of-Experience Measurement with Twitter Nevertheless, we expect the outcome
to be actionable in a way that subsequent work can build upon it and support
the presented use cases. The remainder of this paper is organized as follows. Section 2 gives an
overview of QoE metrics for web services, how measurements are performed,
and which shortcomings exist in current methods. Also, use cases are presented
and requirements to the solution are derived. In Sect. 3 we investigate existing
approaches on leveraging social media content for detecting disruptions of web
services. A description of our experimental setup, message processing and signal
extraction methods to receive correlations between web services and user expe-
rience are presented in Sect. 4. In Sect. 5, we give insights to the generated data
and evaluate our findings with respect to the research question that is raised. The results are concluded in Sect. 6 and future work is outlined. D. Kergl et al. 110 2
Problem Statement Assessing accurate QoE metrics is a challenging problem. Schatz et al. give
comprehensive insights in [29]. They define two different kinds of testing tech-
niques for QoE: The first kind is made of subjective tests, typically conducted
in a controlled laboratory setting but also as field tests or using crowdsourcing
methods. All of these methods aim to gather answers from humans to prede-
fined questions and include the downside of being costly, time-consuming and
require careful planning. The second kind of tests is objective measurement that
include measuring physiological aspects of test persons or technical parameters
of the utilized systems and infrastructure. These assessment methods need to be
mapped to a resulting user experience score, requiring a proper model. Whilst
also being affected by the drawbacks of the subjective methods, the big advan-
tage of objective methods is the possibility of automation and therefore, some
degree of scalability. 2.1
Modeling and Formalization of QoE QoE is defined as a metric for the relationship of a person that interacts as
user with an application [21,25]. While QoS focuses on the relationship between
systems, the authors recognize that a change in QoS only affects QoE if a person’s
expectation is affected. Analogous to this approach, also the concept of Quality
of Business (QoBiz) is introduced, the value of which only is affected by changes
in QoE if a company’s revenue is impaired. The key finding of these publications
is that values of different quality aspects can be seen independent, even though
they build upon each other, so that only weak coupling between these metrics
can be assumed. 2.3
Challenges of Objective Methods for Network Related QoE The ITU outlines a framework for estimating end-to-end-performance in IP net-
works in [15] and recommends to focus on technical metrics like bandwidth,
delay and packet loss rate in order to gain insights to perceived web quality. While concluding that perceived quality can be derived with a correlation that
is high enough for most use cases (>0.9), more detailed methods for addressing
factual challenges and considering a higher number of variables have been pub-
lished during the last years. Most of these approaches incorporate in some way
the complexity of human emotion, that are not considered in the framework of
the ITU. Some of the human mind’s complex relationships have been researched
in context of QoE : Egger et al. show in [8] the direct applicability of the Weber-
Fechner Law (see [10] for a brief historical outline) to the relationship of waiting
time and download experience. They proof this finding empirically for simple
waiting tasks and furthermore, they also investigate the applicability of loga-
rithmic relations between bandwidth and mean opinion score for more complex
tasks like web browsing. Instead of a logarithmic relationship, rather an exponen-
tial relationship was discovered, as has also been shown before by Fiedler et al. in [9]. The explanation for this outcome lies in the complex, non-linear models
of network-level page load times, which were investigated in detail by Belshe [4]. Also, a memory effect has to be considered as psychological influence factor as
described in [13]. With [7] Egger et al. provide a condensed summary of many
of the intertwined aspects. From these insights into technical and psychologi-
cal background of perceived web quality, we can derive that a purely technical
approach to measure web QoE is a hard problem. 2.2
Web QoE Metrics and Assessment In contrast to QoS that is well defined and standardized [16], and even adapted
to specific technologies like mobile networks [17], QoE is much harder to quantify. A common factor of most approaches, is the assignment of an average value for
perceived quality on a scale from 1 to 5 (representing bad, poor, fair, good, and
excellent), what is known as the Mean Opinion Score (MOS) [14]. Streijl et al. give a comprehensive summary of methods, applications, limi-
tations and alternatives of the MOS in [30]. They describe the influence of psy-
chological aspects, test design, testing methods and even the choice of scales to
the result of MOS measurement. Stating the costly and time-consuming nature
and limited scope of subjective quality tests, they also review objective models
that exist in various types (e.g., arithmetic models, statistical models, paramet-
ric network planning models). While these models can be considered correct, as
long as the calculated MOS lies within the confidence interval of the subjective
MOS, the authors conclude that slight broadening of distortions results in higher
complexity and disagreement between perceived qualities. Towards Internet Scale Quality-of-Experience Measurement with Twitter 111 2.4
Use Cases To demonstrate the tangibility of the problem statement, we look at the following
exemplary stakeholders that can benefit from internet wide QoE measurements. Network Providers need to optimize investments on new infrastructure in a way
that costs are minimized while turnover is maximized, aiming at ultimately
maximizing profit. QoE is a valid metric for customer satisfaction, which in
turn we imply is positively correlated with turnover. Due to this correla-
tion, optimizing for QoE is more target-oriented than optimizing for technical
QoS parameters. The knowledge of a base level of customer satisfaction and
the ability to detect changes is key either to assess the effect of investments
already carried out and to identify weak points in network infrastructure that
are most in need for further investments. Network Providers need to optimize investments on new infrastructure in a way
that costs are minimized while turnover is maximized, aiming at ultimately
maximizing profit. QoE is a valid metric for customer satisfaction, which in
turn we imply is positively correlated with turnover. Due to this correla-
tion, optimizing for QoE is more target-oriented than optimizing for technical
QoS parameters. The knowledge of a base level of customer satisfaction and
the ability to detect changes is key either to assess the effect of investments
already carried out and to identify weak points in network infrastructure that
are most in need for further investments. Service Providers that offer their business to worldwide customers, often rely on
both own and third-party infrastructure to deliver contents. Ensuring con-
tinuous availability and convenient response times, as two key service level
metrics, is business-critical to them. Their challenge in monitoring customer
experience is manifold: Services are frequently added and changed so that
automatic or synthetic monitoring of technical key performance indicators Service Providers that offer their business to worldwide customers, often rely on
both own and third-party infrastructure to deliver contents. Ensuring con-
tinuous availability and convenient response times, as two key service level
metrics, is business-critical to them. Their challenge in monitoring customer
experience is manifold: Services are frequently added and changed so that
automatic or synthetic monitoring of technical key performance indicators D. Kergl et al. 112 often lags behind and covers only a small fraction of all service functions. Also a service provider would like to be aware of a shift of customers’ expe-
rience during the lifetime of a service. 2.5
Requirements To conclude the former stated shortcomings and limitations of current
approaches, we derive the following requirements on real-time QoE measurement
at internet scale, matching the demands of the presented use cases. 1. Identify an overall baseline for web service QoE. 2. Recognize changes in customer experience with web services, especially drops. 3. Monitor for QoE problems independently of underlying network technology. 4. Monitor new services immediately after deployment and adapt to changes. 2. Recognize changes in customer experience with web services, especially dro g
g
p
,
p
y
p
Monitor for QoE problems independently of underlying network technology. M
d
l
f
d
l
d
d
h 3. Monitor for QoE problems independently of underlying network technology. 4. Monitor new services immediately after deployment and adapt to c Provide continuous insights to changes and affected service. 5. Provide continuous insights to changes and affected service. 6. Provide measurements near real-time. 3
Related Work In order to examine to which extend the identified requirements are met by
existing approaches and also to eventually identify the open points that have to
be considered, we give the necessary overview of the most significant work in the
relevant fields. 2.4
Use Cases Challenging is that the underlying
infrastructure is very heterogeneous in most times, not only because of third-
party services but also because of implementing novel cloud technologies as
demanded by service expansion. In case of a problem, they also want to iden-
tify whether the problem affects only their own service or services of other
providers as well to communicate accordingly to their customers. Security Actors may observe disruptions of network segments or services of cen-
tral importance for the reliability of internet infrastructure as a result of
large-scale attacks. In such scenarios, it is crucial to gain as much informa-
tion as possible as quickly as possible. This is to make up the information
advantage of the attackers and become able to successfully deploy counter
measures in a timely manner. To know whether, which, where and to what
extend web services are affected, can support this process effectively. Security Actors may observe disruptions of network segments or services of cen-
tral importance for the reliability of internet infrastructure as a result of
large-scale attacks. In such scenarios, it is crucial to gain as much informa-
tion as possible as quickly as possible. This is to make up the information
advantage of the attackers and become able to successfully deploy counter
measures in a timely manner. To know whether, which, where and to what
extend web services are affected, can support this process effectively. Towards Internet Scale Quality-of-Experience Measurement with Twitter Mok et al. investigate how network path qualities (i.e., bandwidth, roundtrip
time (RTT) and loss rate) affects QoE of Hypertext Transfer Protocol (HTTP)
video streaming [23]. They measure the MOS in a sophisticated experimental
setup under strictly controlled test conditions. Furthermore, they present first
results for a correlation between video category (i.e., sports, news, comedy,
music video) and the perceived quality, while keeping technical attributes like
stall times and re-buffering frequency fixed. The dependency between MOS and
video category is a good example of the human factor in MOS measurement and
shows the non-linear connection between technical values and perceived qual-
ity. Both publications show the need for access to core network components to
automatically measure a QoE score. While fulfilling some of the requirements,
these approaches cannot adapt to new services and are strongly dependent on
the underlying network topology. The same authors investigate in a recent work the quality of crowd-sourced
approaches to QoE measurement [24]. Though being relatively cost-effective, for
long running settings, costs are still a disadvantage. Advantages over one-time
experiments are, e.g., the ability of conducting an ongoing assessment of certain
services, and due to using humans as sensors, adaptability to changes in the
assessed services. A disadvantage is still the management effort for planning,
supporting and evaluating the questionnaires. Also the quality can be an issue,
as the authors investigate in the paper. 3.1
Measuring QoE There are several approaches to derive MOS for specific applications from mea-
surable network parameters and traffic monitoring, most of which include elab-
orate field trials interviewing test persons. In [5] Casas et al. present YOUQ-
MON, an approach to calculate the MOS for YouTube videos in 3G networks
by passively monitoring network packets within the network core. To evaluate
the model, they conducted a field trial with 16 different videos to compare the
calculated MOSs with the ones perceived by test persons. 113 Towards Internet Scale Quality-of-Experience Measurement with Twitter D. Kergl et al. Qiu et al. evaluated the relationship between tweets and customer care tick-
ets that both address mobile network experience issues [27]. They found that
tweets, relating to the same problem, preceded customer care tickets by approxi-
mately 10 min. Furthermore, tweets reported a wider range of problems while also
addressing a slightly different set of problems. Qiu et al. mapped the problems
reported via Twitter to incidents they knew from the ticket system. In addition
to the already known incidents, they were able to identify short-term problems
that have not been reported via the ticket system. Summing up their findings, we
emphasize that these correspond with our motivation to exploit tweets for mea-
suring QoE in real-time: Timely detection of drops in experience, high sensitivity
for a broad range of problems and open availability of continuous monitoring data. 3.2
Using Twitter to Detect Outages Principally, not only Twitter is suited as a data source for detecting opinions
about web services. Other social media platforms offer also a wealth of user
generated content. The decisive criterion for choosing Twitter is easy accessibility
of data. This is meant in a technical manner, as Twitter offers a well-documented
API with free and open access for many use cases. Apart from that, using Twitter
is motivated in the text-focused format of the data that can be exploited with
well-established techniques. Motoyama et al. were the first to leverage the unique characteristics of Twit-
ter messages for detecting outages of internet services [26]. They identified terms
that qualify tweets to report about service outages by investigating tweets that
occurred in temporal correlation with major service outage reports in the media. To further refine their filter, they developed a heuristic that leveraged customs
of Twitter users, like using hashtags that include the word fail. To clean up the
derived signals, they made use of exponential smoothing and gave insights into
their chosen parameters to achieve optimal results. Their outcome is to be able
to identify outages of online services by observing between 4 and approximately
200 reports about a specific service outage. The suggested solution was later
implemented by Augustine and Cushing [3]. They used the approach to moni-
tor outages and network problems of the NETFLIX content delivery network. They were able to evaluate the accuracy of their system because a list of outages
of the monitored web service was available to them and showed the practical
applicability and value of leveraging tweets for their use case. 114 D. Kergl et al. Towards Internet Scale Quality-of-Experience Measurement with Twitter 115 4.1
Experimental Setup In this section, we briefly describe the source of the analyzed data, the ETL
process (i.e., extract, transform, load), and used methods of feature isolation,
data smoothing and signal extraction. All steps were performed using two mid-
class notebooks and one office workstation, equipped with 4–12 CPU cores at
2.7 GHz–3.16 GHZ and 8 GB–32 GB RAM. For loading and analyzing the data,
these systems formed a small cluster running Elasticsearch on Apache Lucene as
main database supported by a powerful indexing and search environment, and
Kibana for gaining insights into the data. Once the data has been loaded, typical
requests involving a keyword filter took approximately 60–80 s. 4
Internet Scale QoE Measurement We address the identified open points in the following way. In contrast to exist-
ing work, our approach aims to isolate a signal that is suitable for inferring an
internet wide QoE score, rather than concentrating on a specific service. Fur-
thermore, we add distinction between a total loss of availability and response
time of a service, which can be used to derive a graduated score of disruption,
rather than a binary decision. According to existing approaches that use social
media content and therefore humans as sensors, we also use tweets to meet the
remaining requirements. 3.3
Open Points We conclude the review of related work with summarizing how the formulated
requirements are met in Table 1. A global baseline for an internet wide QoE score
is not provided by any of the mentioned publications. Furthermore, to the best
of our knowledge, there is no such approach in existing scientific literature. While
the approaches that use network parameter measurements to obtain an MOS are
able to map the results to a continuous scale between 1 and 5, the approaches
that leverage social media messages to detect outages are only able to make a
binary decision between service available and not available. Also to the best of
our knowledge, there is no approach so far that would investigate other service
disruptions like increased latency. Network measurement based approaches have
an obvious dependency on the underlying technology. In contrast, approaches that
use humans as sensor are free of this dependency. Also, human based test methods
are able to adapt to new services and service changes. In the case of crowd-sourced
test methods, questionnaires and manuals have to be adapted. Provided that ser-
vices and technology conditions are stable, all methods can be used for continu-
ous monitoring. Though crowd-sourcing methods have limited real-time response
times, as the setup and management overhead can be significant. Table 1. Assessment how the requirements are met by existing work. Requirement (see Sect. 2.5)
Casas
[5]
Mok’11
[23]
Mok’16
[24]
Motoyama
[26]
Qiu
[27]
Augustine
[3]
1: Global baseline
2: Detect score changes
3: Independent of technology
4: Adapt new services
5: Continuous monitoring
6: Results in real-time
Requirement
=not met,
=partially met,
=fully met Table 1. Assessment how the requirements are met by existing work. Towards Internet Scale Quality-of-Experience Measurement with Twitter Towards Internet Scale Quality-of-Experience Measurement with Twitter 4.2
Data Source In order to gather a reasonable data set for our analysis, we used Twitter’s public
API. Combining several API methods, we captured and requested tweets that
have been created in the time interval from 13-Feb-2017 12:00:00.000 UTC to
02-Mar-2017 08:59:59.999 UTC. The obtained data set is not complete in the
sense that a complete data set would include every single tweet that has been
published during the considered period. First reason is, for being able to analyze
textual content efficiently, we dropped all tweets with a language attribute that
differs from en (i.e., English). This restriction is not as strict as it may seem. First, English is the most used language on Twitter, and second, internet related
issues are often reported in English, even if it is not the native language of the
reporting user. A possible reason might be that error messages are mostly in
English and are simply cited by reporting users. We have observed this behavior
very often in our data set. In addition, there are several more reasons for the data
set not being complete: Besides public tweets there are direct tweets between
users. Direct tweets are private and not accessible by anyone else but the sending
and the receiving user. Another category of tweets that we missed consists of such
tweets that have already been deleted at the time of our request. Ultimately our
data set lacks of tweets that we simply did not cover with our query parameters. Taking into account that a productive implementation of our findings, if
appropriate, would most certainly also use the public Twitter API, working on
an equivalent data set with common shortfalls appears to be justified. Beyond 116 D. Kergl et al. this, an eventual implementation should be able to work on a much smaller data
set than we used to examine the feasibility of the concept. Figure 1(a) shows a typical Twitter-Day for English content of our data set. We are able to show the amount of total tweets in the firehose (i.e., the stream
of all public tweets), taking advantage of the nature of Twitter’s sample stream
that we described in [19] and also captured for this analysis. Also the proportion
of our data set can be derived from the figure. Furthermore, we have the requirement to our data of being as random as
possible. 4.2
Data Source This is due to a limitation in Twitter’s Streaming API: If requested
with a set of keywords (that seems appropriate for our problem at the first
glance), Twitter will deliver all tweets that contain these keywords, applying a
logical OR between the requested keywords and there is no possibility to use
a NOT operator for this request. However, according to the documentation of
the Streaming API, Twitter caps the delivery rate of this stream to 1% of the
current firehose rate. Hence, if the hit rate of our filter exceeds this limit, we
would not be able to derive an accurate result value due to capped measurement
values. Whether this limitation is a real problem or practical systems could rely
on Twitter’s Streaming API, can be derived from Fig. 1(b): The histogram shows
the distribution of the resulting percentage of the firehose for all 5 min intervals
of our data set when we use the keywords that are presented in the next section. According to this analysis, we would hit the limit of one percent in more than a
half of the time. Fig. 1. Properties of the derived data set Data Extraction. The extraction challenge is typically twofold: First, we have
to identify relevant terms, that appear in tweets of our interest. Second, tweets
containing the identified terms have to be collected from the data set to form
generating components for the desired signal. Inevitably, at first we have to define what characterizes a tweet of interest. As Twitter is to be leveraged as a sensor for drops in web QoE , tweets that
represent a suitable signal combine the following Properties: 117 Towards Internet Scale Quality-of-Experience Measurement with Twitter Related to a specific web service (no need to mention which in particular). 1. Related to a specific web service (no need to mention which in particular). 1. Related to a specific web service (no need to mention which in particula (
2. Describe present reduction of availability or speed of (parts of) a web ser (
)
scribe present reduction of availability or speed of (parts of) a web service. 3. May be formulated as a question, may use humor, sarcasm, or irony. May be formulated as a question, may use humor, sarcasm, or irony. 4. Do not notify about intentional down time. Hereinafter the occurrence of a tweet of interest will be referenced as event. As a starting point for suitable terms to identify events, we analyzed and used
phrases presented in [26]. The first filter approach used the following Conditions
for tweet content (not case sensitive): 1. Must contain: website OR site OR server 2. Must contain: down OR unreachable OR error While Condition 1 identifies tweets about web services, Condition 2 restricts
the result set to terms that most likely describe the problem component of
Property 2. Using only these two conditions, a one-time training set of 400
tweets showed a selectivity of 0.77. After this manual review we optimized the filter choice for selectivity. As a
result, these additional Conditions were added to the filter: 3. Must NOT contain: (going to be OR will be OR was OR is not) down . Must NOT contain: (close* OR shut* OR take* OR took OR torn) down 5. Must NOT contain: (clean OR count* OR dress O
OR settle OR sit OR top OR written) down 5. Must NOT contain: (clean OR count* OR dress OR low OR right OR scroll
OR settle OR sit OR top OR written) down 6. 5
Evaluation For determining a baseline of reports about service outages or service restrictions
like increased response time, we need to apply an appropriate metric. Due to
the nature of tweet distribution that is not unique across a day, we cannot
simply count the number of event occurrences and use this value as a baseline. Since reports of events underlie the same daily rhythm as the firehose and are
correlated with the biological rhythm of the sensors, we have to normalize the
event count to the current activity. This is achieved by using the proportion
between all events and all non-event tweets as metric. Hence, the ratio of event
occurrences in a specific time interval in comparison to the total number of tweets
in the same time interval qualifies as the desired metric. The simple formula for
the wanted score in time interval n is event-ration =
|events|n
|EWMA(tweets)|n −|events|n
. (1) (1) To address temporal spikes in the total tweet number, we applied the smoothing
described in the preceding section to the total number of tweets and Fig. 2(a)
show the distribution of the event ratio in our data set for the outage event
defined above. Figure 2(b) shows the event ratio distribution for slow events, for
that we changed Condition 2 to the term slow only. To address temporal spikes in the total tweet number, we applied the smoothing
described in the preceding section to the total number of tweets and Fig. 2(a)
show the distribution of the event ratio in our data set for the outage event
defined above. Figure 2(b) shows the event ratio distribution for slow events, for
that we changed Condition 2 to the term slow only. Fig. 2. Event ratios of the analyzed data set. The right most spikes, that constitute
a multiple of the average ratio, are related to Amazon’s S3 outage during the capture
time of the data set. Fig. 2. Event ratios of the analyzed data set. The right most spikes, that constitute
a multiple of the average ratio, are related to Amazon’s S3 outage during the capture
time of the data set. Fig. 2. Event ratios of the analyzed data set. The right most spikes, that constitute
a multiple of the average ratio, are related to Amazon’s S3 outage during the capture
time of the data set. Fig. 1. Properties of the derived data set Must not be a retweet of an original tweet. Condition 3 was added to ensure the temporal relation of Property 2. To address
Property 4, Condition 4 was introduced. The conditions were further restricted
by introducing Condition 5, covering the most common semantic ambiguities
of the term down, and by Condition 6 as retweets in the training set in most
instances did not fulfill the listed properties. Finally, this set of conditions was
evaluated against a test set of 840 tweets that is distinct to the training set. For the refined condition set, a selectivity of 0.88 was identified, yielding 12%
false positives. We consider this rate as being sufficient for conducting a proof
of concept, considering the simplicity of our approach. Data Smoothing. The time resolution we applied for this analysis is 5 min. We
could have chosen smaller intervals, but then occasional random disconnections
of the Twitter stream, network anomalies or other random errors, would have a
more significant effect on the results. Hence, we have chosen this general smooth-
ing. Furthermore, we can observe increased activity at every full hour and also at
every half hour. The latter does not weight as much as the first. To address the
artificially generated bursts in the data set, we applied an exponential weighted
moving average (EWMA) with a half-life period of 15 min (α ≈0.206) and used
this value for further calculations. The frequency of the analyzed tweets also
shows a typical variation during a day. 118
D. Kergl et al. D. Kergl et al. 118
D. Kergl et al. 118 5
Evaluation The mean value for the down report ratio can be identified as 2.25 × 10−5,
while the median is 1.66 × 10−5. Using the median as a baseline seems appro-
priate, as there have been major outages taken place during the capturing of
the data set. The mean value for the slow report ratio is 1.58 × 10−6, while the
median is 2.56 × 10−7. The significant difference between median and mean is a
clear indicator for outliers, that can be confirmed by the event ratio histogram Towards Internet Scale Quality-of-Experience Measurement with Twitter 119 that shows occurrences of event ratios that are multiples of the median. Table 2
lists an excerpt from the top 50 event ratios and informs about the causing event,
that we identified by textually analyzing the specific period. Table 2. Excerpt from top 50 highest event ratio intervals. Events manually evaluated. #
Time
Event ratio
Median
multiplier
Causing event
1
28-Feb-17 18:30
0.000456
27.5
Amazon S3 outage
2
28-Feb-17 18:10
0.000424
25.6
Amazon S3 outage
3
28-Feb-17 18:15
0.000374
22.5
Amazon S3 outage
4
28-Feb-17 18:00
0.000371
22.4
Amazon S3 outage
5
28-Feb-17 18:25
0.000352
21.2
Amazon S3 outage
16
27-Feb-17 11:30
0.000275
16.6
Error message on hilton.com
24
02-Mar-17 10:50
0.000246
14.8
Vainglory game server maintenance
37
02-Mar-17 12:25
0.000204
12.3
Booking problem on qatarairways.com
42
27-Feb-17 12:10
0.000188
11.3
Booking problem on klm.com
50
02-Mar-17 12:35
0.000141
8.5
Amazon S3 outage ble 2. Excerpt from top 50 highest event ratio intervals. Events manually evaluated. References 1. Ahmad, A., Floris, A., Atzori, L.: QoE-aware service delivery: a joint-venture app-
roach for content and network providers. In: 2016 Eighth International Conference
on Quality of Multimedia Experience (QoMEX), pp. 1–6. IEEE (2016) (
)
(
)
2. Al-Shammari, S., Al-Yasiri, A.: Defining a metric for measuring QoE of SaaS cloud
computing. In: Proceedings of PGNET, pp. 251–256 (2014) 3. Augustine, E., Cushing, C., Dekhtyar, A., McEntee, K., Paterson, K., Tognetti,
M.: Outage detection via real-time social stream analysis: leveraging the power of
online complaints. In: Proceedings of the 21st International Conference Companion
on World Wide Web, pp. 13–22. ACM (2012) (
)
4. Belshe, M.: More Bandwidth Doesn’t Matter (much) (2010) 5. Casas, P., Seufert, M., Schatz, R.: YOUQMON: a system for on-line monitoring
of YouTube QoE in operational 3G networks. ACM SIGMETRICS Perform. Eval. Rev. 41(2), 44–46 (2013) ( )
(
)
6. Dong, M., Kimata, T., Sugiura, K., Zettsu, K.: Quality-of-Experience (QoE) in
emerging mobile social networks. IEICE Trans. Inf. Syst. E 97D(10), 2606–2612
(2014) 7. Egger, S., Hossfeld, T., Schatz, R., Fiedler, M.: Waiting times in quality of expe-
rience for web based services. In: 2012 Fourth International Workshop on Quality
of Multimedia Experience, pp. 86–96 (2012) 8. Egger, S., Reichl, P., Hossfeld, T., Schatz, R.: ‘Time is bandwidth’? Narrowing
the gap between subjective time perception and quality of experience. In: IEEE
International Conference on Communications, pp. 1325–1330 (2012) 9. Fiedler, M., Hossfeld, T., Tran-Gia, P.: A generic quantitative relationship between
quality of experience and quality of service. IEEE Netw. 24(2), 36–41 (2010) 10. Hecht, S.: The visual discrimination of intensity and the Weber-Fechner law. J. Gen. Physiol. 7(2), 235–267 (1924) 11. Hei, X.H.X., Liu, Y.L.Y., Ross, K.: IPTV over P2P streaming networks: the mesh-
pull approach. IEEE Commun. Mag. 46(February), 86–92 (2008) 12. Ho, T., Hock, D., Tran-gia, P., Tutschku, K., Fiedler, M.: Testing the IQX hypoth-
esis for exponential interdependency between QoS and QoE of voice codecs iLBC
and G.711. In: Proceedings of the 18th ITC Specialist Seminar on Quality of Expe-
rience, pp. 105–114 (2008) 13. Hossfeld, T., Biedermann, S., Schatz, R., Platzer, A., Egger, S., Fiedler, M.: The
memory effect and its implications on Web QoE modeling. In: 2011 23rd Interna-
tional Teletraffic Congress (ITC), pp. 103–110 (2011) 14. International Telecommunication Union: P.800: Methods for subjective determi-
nation of transmission quality. Technical report, International Telecommunication
Union (1996) 15. 6
Conclusion and Future Work We have been able to define a global baseline for down report and slow report
frequencies. Therefore, there are two main contributions of this work to mention:
1. A practical system for monitoring the overall internet web QoE is feasible and
can be implemented using Twitter analysis. This fulfills Requirement 1 that most
likely has not been addressed by any existing work. 2. Not only outages of web
services, but also degradation of web service quality can be detected. This fulfills
Requirement 2 that has not been completely covered by existing publications. The remaining requirements have been matched by using humans as sensors. The presented primary findings about the feasibility of using social media
posts for gaining internet wide insights to QoE aspects in real-time denote an
important step towards more detailed analysis of affected networks, domains
and technologies, constituting a necessary requirement for novel approaches to
improve overall network and internet security, e.g., as suggested in [18]. As follow
up research questions, we are already investigating whether root causes of drops
in QoE can be identified by using additional information contained in tweets, for
instance, analyzing geographical origin of the complaints might lead to insights
about regional problems and using the contained information about which client
software was used to create the tweet might give further hints on whether mobile
or fixed networks or both are affected by drops in perceived web service quality. Furthermore, mapping complaints to specific web services in an automated fash-
ion seems to become feasible, while still being a complex problem. This would
allow to drill down the QoE measurements to individual domains and accord-
ingly to underlying networks and technologies. D. Kergl et al. 120 References International Telecommunication Union: G.1030: Estimating end-to-end perfor-
mance in IP networks for data applications. Technical report, International
Telecommunication Union (2005) 16. International Telecommunication Union: E.800: Definitions of terms related to
quality of service. Technical report, International Telecommunication Union (2008) 16. International Telecommunication Union: E.800: Definitions of terms related to
quality of service. Technical report, International Telecommunication Union (2008)
17. International Telecommunication Union: E.804: QoS aspects for popular services in 17. International Telecommunication Union: E.804: QoS aspects for popular services in
mobile networks. Technical report, International Telecommunication Union (2014) Towards Internet Scale Quality-of-Experience Measurement with Twitter Towards Internet Scale Quality-of-Experience Measurement with Twitter 121 18. Kergl, D., Roedler, R., Rodosek, G.D.: Detection of zero day exploits using real-
time social media streams. In: Pillay, N., Engelbrecht, A.P., Abraham, A., Du
Plessis, M.C., Sn´aˇsel, V., Muda, A.K. (eds.) Advances in Nature and Biologically
Inspired Computing. AISC, vol. 419, pp. 405–416. Springer, Cham (2016). doi:10. 1007/978-3-319-27400-3 36 /
19. Kergl, D., Roedler, R., Seeber, S.: On the endogenesis of Twitter’s Spritzer and
Gardenhose sample streams. In: 2014 IEEE/ACM International Conference on
Advances in Social Networks Analysis and Mining (ASONAM 2014), pp. 357–364. IEEE (2014) (
)
20. Ketyk´o, I., De Moor, K., Joseph, W., Martens, L., De Marez, L.: Performing QoE-
measurements in an actual 3G network. In: 2010 IEEE International Symposium on
Broadband Multimedia Systems and Broadcasting (BMSB), pp. 1–6. IEEE (2010)
21. Kilkki, K.: Quality of experience in communications ecosystem. J. Univ. Comput. Sci 14(5) 615 624 (2008) 20. Ketyk´o, I., De Moor, K., Joseph, W., Martens, L., De Marez, L.: Performing QoE-
measurements in an actual 3G network. In: 2010 IEEE International Symposium on
Broadband Multimedia Systems and Broadcasting (BMSB), pp. 1–6. IEEE (2010) Broadband Multimedia Systems and Broadcasting (BMSB), pp. 1–6. IEEE (2010)
21. Kilkki, K.: Quality of experience in communications ecosystem. J. Univ. Comput. Sci. 14(5), 615–624 (2008) 21. Kilkki, K.: Quality of experience in communications ecosystem. J. Univ. Comput. Sci. 14(5), 615–624 (2008) 22. Liotou, E., Elshaer, H., Schatz, R., Irmer, R., Dohler, M., Passas, N., Merakos, L.:
Shaping QoE in the 5G ecosystem. In: 2015 Seventh International Workshop on
Quality of Multimedia Experience (QoMEX), pp. 1–6 (2015) 23. Mok, R.K.P., Chan, E.W.W., Chang, R.K.C.: Measuring the quality of experi-
ence of HTTP video streaming. In: 2011 IFIP/IEEE International Symposium on
Integrated Network Management (IM), pp. 485–492 (2011) 24. References Mok, R.K., Chang, R.K., Li, W.: Detecting low-quality workers in QoE crowdtest-
ing: a worker behavior based approach. IEEE Trans. Multimedia 19(3), 530–543
(2016) (
)
25. Moorsel, A.V.: Metrics for the internet age: quality of experience and quality of
business. Quantitative evaluation in the internet age: what is different? Perspective
34, 26–31 (2001) (
)
26. Motoyama, M., Meeder, B., Levchenko, K., Voelker, G.M., Savage, S.: Measuring
online service availability using twitter. In: Proceedings of the 3rd Conference on
Online Social Networks (WOSN 2010) (2010) (
) (
)
27. Qiu, T., Feng, J., Ge, Z., Wang, J., Xu, J., Yates, J.: Listen to me if you can: track-
ing user experience of mobile network on social media. In: Proceedings of the 10th
ACM SIGCOMM Conference on Internet Measurement, Melbourne, Australia, pp. 288–293. ACM (2010) (
)
28. Rengaraju, P., Lung, C.H., Yu, F.R., Srinivasan, A.: On QoE monitoring and E2E se-
rvice assurance in 4G wireless networks. IEEE Wirel. Commun. 19(4), 89–96 (2012) (
)
28. Rengaraju, P., Lung, C.H., Yu, F.R., Srinivasan, A.: On QoE monitoring and E2E se-
rvice assurance in 4G wireless networks. IEEE Wirel. Commun. 19(4), 89–96 (2012)
29. Schatz, R., Hoßfeld, T., Janowski, L., Egger, S.: From packets to people: quality
of experience as a new measurement challenge. In: Biersack, E., Callegari, C.,
Matijasevic, M. (eds.) Data Traffic Monitoring and Analysis. LNCS, vol. 7754, pp. 219–263. Springer, Heidelberg (2013). doi:10.1007/978-3-642-36784-7 10 29. Schatz, R., Hoßfeld, T., Janowski, L., Egger, S.: From packets to people: quality
of experience as a new measurement challenge. In: Biersack, E., Callegari, C.,
Matijasevic, M. (eds.) Data Traffic Monitoring and Analysis. LNCS, vol. 7754, pp. 219–263. Springer, Heidelberg (2013). doi:10.1007/978-3-642-36784-7 10 30. Streijl, R.C., Winkler, S., Hands, D.S.: Mean opinion score (MOS) revisited: meth-
ods and applications, limitations and alternatives. Multimedia Syst. 22(2), 213–227
(2016) (
)
31. Takahashi, A., Hands, D., Barriac, V.: Standardization activities in the ITU for a
QoE assessment of IPTV. IEEE Commun. Mag. 46(2), 78–84 (2008) g
( )
(
)
32. Venkataraman, M., Chatterjee, M.: Inferring video QoE in real time. IEEE Netw. 25(1), 4–13 (2011) 33. Wamser, F., Casas, P., Seufert, M., Moldovan, C., Tran-Gia, P., Hossfeld, T.: Mod-
eling the YouTube stack: from packets to quality of experience. Comput. Netw. 109, 211–224 (2016) (
)
34. Zimmer, M., Proferes, N.J.: A topology of Twitter research: disciplines, methods,
and ethics. Aslib J. References Inf. Manag. 66(3), 250–261 (2014) 122
D. Kergl et al. D. Kergl et al. 122 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. Hunting SIP Authentication Attacks Efficiently
Tom´aˇs Jansky1, Tom´aˇs ˇCejka2(B), and V´aclav Bartoˇs2
1 FIT, CTU in Prague, Thakurova 9, 160 00 Prague 6, Czech Republic
janskto1@fit.cvut.cz
2 CESNET, a.l.e., Zikova 4, 160 00 Prague 6, Czech Republic
{cejkat,bartos}@cesnet.cz 1 FIT, CTU in Prague, Thakurova 9, 160 00 Prague 6, Czech Republic
janskto1@fit.cvut.cz
2 CESNET, a.l.e., Zikova 4, 160 00 Prague 6, Czech Republic
{cejkat,bartos}@cesnet.cz Abstract. Extended flow records with application layer (L7) informa-
tion allow for detection of various types of malicious traffic. Voice over IP
(VoIP) is an example of technology that works on L7 and many attacks
against it cannot be reliably detected using just basic flow information. Session Initiation Protocol (SIP), which is commonly used for VoIP sig-
nalling, is a frequent target of many types of attacks. This paper proposes
and evaluates a novel algorithm for near real time detection of username
scanning and password guessing attacks on SIP servers. The detection is
based on analysis of L7 extended flow records. c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 125–130, 2017.
DOI: 10.1007/978-3-319-60774-0 9 2
SIP Attacks This work focuses on two types of network attacks by an unauthenticated exter-
nal attacker against a SIP server – extension scanning (i.e. finding valid user-
names) and password guessing. Both are based on sending large amount of requests (usually REGISTER) to
the server. When a client sends the request requiring authentication, server chal-
lenges it with a response code 401 Unauthorized. Normally, the client sends
valid credentials and server responds with 200 OK. If the username is not valid,
server responds with 404 Not Found or 401 Unauthorized, depending on con-
figuration1. In case of correct username but wrong password, 401 Unauthorized
is returned. Therefore, both types of attacks are characterized by a high number of
REGISTER requests and 401 Unauthorized (or 404 Not Found) responses, using
either different extensions (extension scanning) or a single extension but differ-
ent passwords (passwod guessing). Combination of both is also possible. More
details about these SIP attacks can be found in [4]. 1 The former is considered insecure since it eases the extension scanning as it imme-
diately discloses existence of the extension on the server. 1
Introduction Voice over IP (VoIP) is a technology that replaces classic telephone services and
is used to transfer multimedial data such as voice or video over common packet
switched networks. One of the core protocols used in VoIP services is Session
Initiation Protocol (SIP), which is used for signalling between communicating
parties. There are many types of attacks against SIP infrastructure. The most dan-
gerous attacks often compromise Private Branch Exchange (PBX) devices and
cause a significant financial loss to the owner of PBX. According to [3], a total
worldwide loss due to VoIP hacking and calling to premium rate services goes
to billions of dollars per year. Even though there are standards that describe security considerations and
extensions of the SIP protocol, it is still often observed unencrypted in real
network traffic. This allows for security analysis of SIP traffic at a network
level using a network passive monitoring. The analysis may detect malicious SIP
traffic so that a network operator can inform owners of the target device about
a potential threat or take appropriate actions to mitigate malicious traffic. Network traffic monitoring in large networks is usually done using so called
flow records, i.e. aggregated information about communicating hosts that is com-
puted from observed packets. A typical flow record consists of information from
packet headers up to the transport protocol. This approach is feasible and it
allows for detection of various types of malicious traffic. However, as it was
presented in [2], many types of attack at application protocol (L7) cannot be
reliably detected using just the basic flow records. This paper shows usage of T. Jansky et al. 126 application layer flow records [6], in this case flows extended by L7 information
about SIP traffic, for detection of brute-force password guessing and scanning
for user accounts (called extensions in SIP terminology) on PBX. This work is
a continuation of [2] and an improvement of detection abilities of the previous
detection mechanism. 3
Detection Algorithm In line with the L7 flow monitoring approach, our monitoring probes use a plugin
which is able to extract necessary SIP information from traffic (response code, To
and CSeq). As it is shown in Fig. 1, flow records are sent from probes to a collector
in the IPFIX format and afterwards analyzed by the detection algorithm which
is implemented as a part of the NEMEA [1] system. The detection method is designed to work without any prior knowledge of
VoIP infrastructure or existing extensions. It is based on an analysis of 401
responses from SIP servers. By aggregating these responses by a PBX IP address,
an extension (username) and a client IP address, the detection algorithm can
detect non-standard and potentially malicious traffic. The algorithm shifts between two stages. In the first stage, it receives data
and stores it into data structures. For each SIP server (i.e. IP address sending SIP
responses), the following data is stored – a list of client IPs, a list of usernames,
and a mapping between them that tells which clients tried which usernames and
a number of such attempts. Hunting SIP Authentication Attacks Efficiently 127 Fig. 1. Monitoring infrastructure. Fig. 1. Monitoring infrastructure. After a certain time period, the algorithm gets to the second stage where
it evaluates the stored data. First a type of (potential) attack is determined. If a single client attempts to register one certain extension, it is classified as a
brute-force attack. This attack can be reclassified as a distributed brute-force
attack if more clients attempt to register the particular extension on the same
server. When a client tries to register more than one extension, the behavior is
classified as a scan. When the number of attempts exceeds a threshold, the attack
is reported. If 200 OK response code is detected as part of the communication,
the attack is considered successful. If no communication between the server and
the client is observed for a certain amount of time, the corresponding structures
are released from memory. The algorithm was implemented as a module for open-source NEMEA system
and published at GitHub2. 3 CESNET2 network is monitored at all its 7 peering links at the 10 and 100 Gbps
wire speeds. Average total amount of traffic: 110,000 flows/s, average SIP traffic:
1,500 flows/s. 2 https://github.com/CESNET/Nemea-Detectors/.
3 4
Evaluation Since the algorithm is threshold based, it was necessary to estimate some key
values based on the behavior on a real network. We temporarily captured SIP
traffic from CESNET2 network3. After the analysis of the captured data, we discovered that more than 99.9 %
of all successful register attempts use 20 messages or less. We therefore set 20
attempts as a threshold for deciding whether the communication is malicious or
not. We also examined the frequency of malicious requests in individual attacks
and discovered that only 0.01 % have more than 30 min delay between individ-
ual requests. Therefore an information about a communication is released from p //g
/
/
/
3 CESNET2 network is monitored at all its 7 peering links at the 10 and 100 Gbps
wire speeds. Average total amount of traffic: 110,000 flows/s, average SIP traffic:
1,500 flows/s. T. Jansky et al. 128 the program memory if no new message is observed for 30 min. It also means that
an elapsed attack is reported after this delay since the last observed message. Finally, we counted unique extensions attempted by every client in 30 min
windows. Most observed clients attempted to register as less than 10 unique
extensions on a certain server. This value is surprisingly high, but it is possible
that the client is actually a proxy server or there are multiple SIP clients hidden
behind NAT. We used 10 distinct extensions as a threshold for extension scanning
detection. First, the detection module was tested on a real network with generated
malicious traffic using auditing tool SIPVicious [5]. All generated attacks were
successfully distinguished from other SIP communication and reported. Then, the module was run for one week to capute real attacks in the CES-
NET2 network. Total number of 7,008 events were reported. Table 1 shows some
statistics about reported events. One of the most interesting findings is that
46.3 % of all 200 and 401 SIP responses to REGISTER requests are a malicious
traffic and are directly related with one of reported alerts. Table 1. Statistics after one week of flow detection
Brute-force events
6,488 (92.6 %)
Extension scanning events
520 (7.4 %)
Successful brute-force events
7
Strongest brute-force
6,930,911 attempts
Largest scan
9,360 extensions
SIP flows observed
718,627,758
SIP flows analyzed (401 & 200 responses) 40,909,352 (5.7 %)
Number of malicious flows
18,945,291 (46.3 %) Table 1. 5
Conclusion We designed a method for detection of SIP attacks, namely username scanning
and password guessing, based on an analysis of SIP headers in extended flow
records. The algorithm works without any prior knowledge of VoIP infrastruc-
ture. Its key parameters and thresholds can be adjusted by network admin-
istrators in accordance to the characteristics of their network to reach optimal
detection results. It is efficient and it is able to process data from an NREN-sized
network (several 10 and 100 Gbps links) in real time. Using the algorithm, we were able to detect thousands of scanning and pass-
word guessing against SIP infrastrucutre. The software is also capable of detect-
ing distributed guessing of user’s password, however, this type of attack was
not observed in our network yet. Some of the attacks, which were identified as
successful, were reported to network administrators who subsequently confirmed
the attacks. Analysis of detection results showed only a small amount of false
positive reports with frequency around 0.1% of all reported events. Most of the
false positives are caused by a few clients that communicate in an unusual way
and can be easily filtered using a whitelist. Acknowledgments. This work was supported by Packet analysis based network diag-
nostics (DISTANCE) project No. TH02010186 granted by Technology Agency of the
Czech Republic, project Reg. No. CZ.02.1.01/0.0/0.0/16 013/0001797 co-funded by
the MEYS of the Czech Republic and ERDF and the CTU grant No. SGS17/212/
OHK3/3T/18 funded by the MEYS of the Czech Republic. Hunting SIP Authentication Attacks Efficiently Hunting SIP Authentication Attacks Efficiently Hunting SIP Authentication Attacks Efficiently 4
Evaluation Statistics after one week of flow detection Detection results were stored to a log file during the week. Thorough examina-
tion showed that most attackers perform either brute-force attacks or extension
scanning. However, some of the attackers combine these two attacks to one, usu-
ally trying a small number of password guesses (between 20 to 100) to a large
number of extensions. This behavior indicates that these attackers use some sort
of a set of common and frequently used passwords. To confirm that the detection module is working correctly, we manually ana-
lyzed traffic of some of the reported attacks. Most of them are certainly scanning
or brute-force attempts. In just a few cases were the traffic did not look like any
of the attacks and can be viewed as false positive (we estimate total FP rate to
0.1%), however, it was still an unusual traffic, probably caused by misconfigura-
tion of some devices, which is worth inspecting. To prove practical usefulness of
the detection, we chose one of the attacks marked as successful and contacted
the administrator of the attacked PBX. He confirmed that, indeed, the account
was compromised and informed us that appropriate steps to fortify the PBX will
be taken. 129 MoDeNA: Enhancing User Security for Devices
in Wireless Personal and Local Area Networks Robert M¨uller1(B), Marcel Waldvogel1, and Corinna Schmitt2 1 Distributed Systems Laboratory,
Department of Computer and Information Science, University of Konstanz,
78457 Konstanz, Germany
{robert.mueller,marcel.waldvogel}@uni-konstanz.de
2 Communication Systems Group CSG, Department of Informatics IfI,
University of Zurich UZH, Binzm¨uhlestrasse 14, 8050 Zurich, Switzerland
schmitt@ifi.uzh.ch Abstract. Today most used devices are connected with each other
building the Internet of Things (IoT). A variety of protocols are used
depending on the underlying network infrastructure, application (e.g.,
Smart City, eHealth), and device capability. The judgment of the security
feeling of the data sharing depends on personal settings (e.g., easy to use,
encrypted transmission, anonymization support). MoDeNA – a Mobile
Device Network Assistant – was developed offering an opportunity for
understanding the judgment of security by bringing the user’s concerns
and their technology understanding of used devices and protocols into
relation. MoDeNA provides a transparent overview over the used wire-
less security of the user’s device giving concrete advices for improving
the connection security and usability of mobile device security. c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 131–136, 2017.
DOI: 10.1007/978-3-319-60774-0 10 References 1. Cejka, T., Bartos, V., Svepes, M., Rosa, Z., Kubatova, H.: NEMEA: a framework for
network traffic analysis. In: 12th International Conference on Network and Service
Management (CNSM 2016), Montreal, Canada, October 2016 2. Cejka, T., Bartos, V., Truxa, L., Kubatova, H.: Using application-aware flow mon-
itoring for sip fraud detection. In: Latr´e, S., Charalambides, M., Fran¸cois, J.,
Schmitt, C., Stiller, B. (eds.) AIMS 2015. LNCS, vol. 9122, pp. 87–99. Springer,
Cham (2015). doi:10.1007/978-3-319-20034-7 10 (
)
/
3. Communication Fraud Control Association: Global fraud loss survey (2015). http://
www.cfca.org/pdfsurvey/2015 CFCA Global Fraud Loss Survey Press Release.pdf
4
Dwivedi H : Hacking VoIP: Protocols Attacks and Countermeasures No Starch 4. Dwivedi, H.: Hacking VoIP: Protocols, Attacks, and Countermeasures. No Starch
Press, San Francisco (2009) 4. Dwivedi, H.: Hacking VoIP: Protocols, Attacks, and Countermeasures. No Starch
Press, San Francisco (2009) 5. Gauci, S.: SIPVicious. Tools for auditing sip based VoIP systems (2012). https://
code.google.com/p/sipvicious/
ˇ 5. Gauci, S.: SIPVicious. Tools for auditing sip based VoIP systems (2012). https://
code.google.com/p/sipvicious/
ˇ 6. Velan, P., ˇCeleda, P.: Next generation application-aware flow monitoring. In:
Sperotto, A., Doyen, G., Latr´e, S., Charalambides, M., Stiller, B. (eds.) AIMS
2014. LNCS, vol. 8508, pp. 173–178. Springer, Heidelberg (2014). doi:10.1007/
978-3-662-43862-6 20 6. Velan, P., ˇCeleda, P.: Next generation application-aware flow monitoring. In:
Sperotto, A., Doyen, G., Latr´e, S., Charalambides, M., Stiller, B. (eds.) AIMS
2014. LNCS, vol. 8508, pp. 173–178. Springer, Heidelberg (2014). doi:10.1007/
978-3-662-43862-6 20 T. Jansky et al. 130 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. 1
Motivation The Internet of Things (IoT) not only includes servers, computers, and routers
anymore, but also personal “smart” devices that everyone uses frequently, such as
smartphone, sensors, tags, and tablets. All devices collect many data in different
application areas and are connected to share the data [1,2]. It is envisioned
that the variety of devices will grow in the future as well as the number of
participating devices in the IoT [3]. Usually, a user is just a user of the device
or the application, trusting in the pre-installed security mechanisms. In order to allow a judgment of the used security, MoDeNA—our Mobile
Device Network Assistant—was developed addressing the aforementioned views
of the users abilities and the deployed network infrastructure in a smart city envi-
ronment. MoDeNA is an operating system independent application based on a
classification algorithm taking into account all available security information
from user’s device and used infrastructure to make the security setting transpar-
ent to the user. Further it recommends the user updates of security settings to
improve the mobile device security for the current situation without requiring
in-depth know-how. The overall goal of MoDeNA is to raise the user’s awareness
of security lacks when using WPANs and WLANs to provide countermeasures
to avoid data theft. R. M¨uller et al. 132 2
Related Work While there are calls for novel security challenges for the services of the IoT like
encryption and authentication [4], proposals for securing the IoT with protocols
like Lithe [5], TinyDTLS [6,7] are available. Additionally, analyzes exist that
investigate the technical challenges and limitations of the IP-based IoT [8,9],
though the aspect of involving the user in the security of the connection between
IoT devices is not considered. To our knowledge there is no known approach to
involve the user in the wireless network security, particularly not for IoT devices. Work in the field of discovering network topology without network assistance
is described in [10]. A user study analyzing security and privacy habits as well
as willingness to apply countermeasures is provided by [11]. Another interesting
approach is investigated in [12] by moving privacy-sensitive tasks to remote
security servers which offer higher protection capabilities than smartphones. 3
MoDeNA’s Security Classification Algorithm Based on the presented challenges in Sect. 2 with existing solutions, the following
goals were set for MoDeNA to build a security classification scheme: (1) Central
Overview of connected IoT devices, (2) Automatic Identification of applied
security requirements, (3) User Interaction support when no automatic iden-
tification happens, and (4) Control Wireless Radio Connections to keep
track of own IoT devices. In order to address the first goal the connected IoT devices are classified
according to the security standard required by the data transmitted. Reading
device specific information, such as shared services for communication, applica-
tions used, and identifying device classes can achieve this without user interac-
tion required for an automatic identification. Additional information provided
by the user about the pairing process, if available, is used for a more precise
identification of security requirements. The classification itself is a process that needs to be adopted for the var-
ious available device types and WPAN/WLAN protocols. Therefore, existing
parameters for classification were used building the “static input (e.g., device
identifiers, announced services, Universally Unique Identifier UUID) and if nec-
essary “dynamic input” based on the user’s manual input. The general security
classification algorithm is illustrated in a flow diagram in Fig. 1. The MoDeNA classification algorithm takes the protocol type, device type
and application of the device to be connected with as input values. They are
obtained automatically by the WPAN/WLAN network sensors and connection
information published in the network (e.g. via network service) by the device. If
there is input regarding connection purpose available from the user (“Dynamic
Information”), this information is considered for a dynamic risk level calculation. Otherwise a static risk level calculation without additional user input is applied. Afterwards the newly established connection is displayed together with its secu-
rity classification. If new user input becomes available (i.e. the user confirms a MoDeNA: Enhancing User Security for Devices 133 security improvement measure within the application) a new dynamic risk level
calculation is executed. Otherwise the algorithm terminates. Four levels for application security requirements are distinguished: (1) High
(green) - key exchange mechanism with no design flaws and transmission encryp-
tion, (2) Acceptable (yellow) - key exchange mechanism with design flaws and
transmission encryption, (3) Low (red) - insufficient data security, and (4) Unde-
termined (grey) - by default accepted. 3
MoDeNA’s Security Classification Algorithm This grading can be seen in the Overview
Screen shown in the shown in the upper part of Fig. 2. It presents the user with
an Security classification state per connected smart device. Clicking on a row
opens the device’s Detail View Screen shown in the lower part of Fig. 2. It
advises the user with practical security hints and asks for input of environmental
parameters to improve classification. The application back-end provides adapter
implementations for the supported physical network interfaces and listens asyn-
chronously for connected devices available. First, it identifies whether a device
was previously connected. For new detected devices, the MoDeNA application
collects the protocol - and device specific information and creates a new entry in
the devices database. Previously known devices can be recognized and the secu-
rity classification is based on the available device history. For each device the
database stores a dataset consisting of: device name, type, address, last security
classification, performed security improvements by the user and used application. Based on this information, the MoDeNA algorithm is applied to determine the
security requirements and obtain the security classification. This is then used to
provide the user with recommendations for each specific combination of device
type and security requirement (e.g., Smart watch + WiFi and/or Bluetooth indi- Fig. 1. Classification algorithm
Fig. 2. MoDeNA views (Color figure
online) Fig. 1. Classification algorithm Fig. 1. Classification algorithm Fig. 2. MoDeNA views (Color figure
online) Fig. 1. Classification algorithm Fig. 2. MoDeNA views (Color figure
online) R. M¨uller et al. 134 cating High security, wireless mouse + WiFi or Bluetooth indicating Acceptable
security, hearing aid + Bluetooth indicating Low security). An example for improving the security of a connection is shown for the
WLAN “kerrigan-2.4”. The WLAN is automatically detected by the smartphone
with activated Wi-Fi service as someones private network, which does not require
authentication and the smartphone connects to it automatically. When a user of
MoDeNA application detects it in the Overview Screen, it is listed as a network
interfaced with. Since there is no authentication provided, it is rated not secure
by the MoDeNA application. The three bullets indicator is used to show the max-
imum possible grading available. 3
MoDeNA’s Security Classification Algorithm If the user now clicks on the list entry, he/she is
brought to the detail view, which shows the reason for this security classification
(red indicator) and what measures can be applied to improve connection secu-
rity with “kerrigan-2.4” by adopting them. Settings and measurements made for
known networks can be saved automatically by MoDeNA. Further information
about the security risk of using specific wireless technologies is provided with
links to useful web pages that provide background information and educate the
user. 4
User Study A prototype of the application MoDeNA is realized on the Android OS platform,
since it is the most widely used operating system to date for smartphones. We conducted a two-part user study to analyze usage of IoT devices con-
nected to smartphones via WPAN/WLAN and to rate the use of our applica-
tion. The participants were asked to fill in a questionnaire with 23 questions
while using the application MoDeNA for the second part of the study. For the
evaluation, we used a mock-up of our proposed application without the imple-
mentation of the classification of the real network connections. (1) Wireless Network Smartphone Security. 48% of our participants have
a technical background (work or education). The interest rate in under-
standing wireless smartphone communication is 91% for non-technical and
67% for technical users. 87% of participants would rate data on their smart-
phones as private data. 70% know about security concerns of data stored
on smartphones but they accept the possible risks. 87% of the participants
ask for more protection of their personal data stored on their smartphone. Asking the users if they turn offunused wireless protocols showed that 65%
do turn offradio, but for reasons like battery, radiation and others, only 22%
of them do it also because of security concerns. 83% of participants state
that they would apply security measures, if their smartphone recommended
them to do so. (1) Wireless Network Smartphone Security. 48% of our participants have
a technical background (work or education). The interest rate in under-
standing wireless smartphone communication is 91% for non-technical and
67% for technical users. 87% of participants would rate data on their smart-
phones as private data. 70% know about security concerns of data stored
on smartphones but they accept the possible risks. 87% of the participants
ask for more protection of their personal data stored on their smartphone. Asking the users if they turn offunused wireless protocols showed that 65%
do turn offradio, but for reasons like battery, radiation and others, only 22%
of them do it also because of security concerns. 83% of participants state
that they would apply security measures, if their smartphone recommended
them to do so. (2) “Application Specific Wireless Security”. The users were requested to
play around and evaluate our prototype implementation of the application
MoDeNA. Thus, this received feedback was user-specific and highly influ-
enced by individual knowhow. 5
Conclusions and Future Work 5 We present MoDeNA, a framework for detection and classification of WPAN/
WLAN connection security and a prototype smartphone application for Android
OS to (semi-)automatically rate the security of connected WPAN/WLAN
devices and provide advices to the user. In our user study with 23 participants
we observed that 70% of participants are generally aware of security risks when
transmitting data wirelessly from a smartphone to any other device but never-
theless use the functionality. 78% of our participants have heard or know about
security risks for WPAN/WLAN protocols. MoDeNA is rated by 90% of our
user study participants to be helpful to feel more secure with smart devices in
WPAN/WLAN. 4
User Study 74% of participants state that they gained (2) “Application Specific Wireless Security”. The users were requested to
play around and evaluate our prototype implementation of the application
MoDeNA. Thus, this received feedback was user-specific and highly influ-
enced by individual knowhow. 74% of participants state that they gained MoDeNA: Enhancing User Security for Devices 135 insight in the security of wireless smartphone communication. The same per-
centage of participants also claimed, that they think the application MoD-
eNA would improve the security when used. 87% expect MoDeNA would
improve the WPAN/WLAN security of their smartphones. References 1. Greengard, S.: The Internet of Things (MIT Press Essential Knowledge). The MIT
Press, Cambridge (2015) 2. International Telecommunication Union: The internet of things. ITU Internet
Reports (2005) 3. Vesola, A., Schulte, W., Lheureux, B.: Hype cycle for the internet of things, 2016. Technical report, Gartner Inc., July 2016 4. Roman, R., Najera, P., Lopez, J.: Securing the internet of things. IEEE Comput. J. 44(9), 51–58 (2011) 5. Raza, S., Shafagh, H., Hewage, K., Hummen, R., Voigt, T.: Lithe: lightweight
secure CoAP for the internet of things. IEEE Sens. J. 13(10), 3711–3720 (2013) 6. Schmitt, C., Kothmayr, T., Hu, W.: Two-way authentication for the internet-of-
things. In: Acharjya, D., Kalaiselvi Geetha, M. (eds.) Internet of Things: Novel
Advances and Envisioned Applications, pp. 27–56. Springer, Heidelberg (2017) 7. Kothmayr, T., Schmitt, W., Hu, C., Bruenig, M., Carle, G.: DTLS based security
and two-way authentication for the internet of things. ELSEVIER Ad Hoc Netw. 11(8), 2710–2723 (2013) ( )
(
)
8. Heer, T., Garcia-Morchon, O., Hummen, R., Keoh, S.L., Kumar, S.S., Wehrle, K.:
Security challenges in the IP-based internet of things. Wirel. Pers. Commun. 61(3),
527–542 (2011). http://dx.doi.org/10.1007/s11277-011-0385-5 9. Hummen, R., Shafagh, H., Raza, S., Voigt, T., Wehrle, K.: Delegation-based
authentication and authorization of the IP-based internet of things. In: 11th
Annual IEEE International Conference on Sensing, Communication, and Network-
ing, SECON, pp. 1–9, June/July 2014 10. Black, R., Donnelly, A., Fournet, C.: Ethernet topology discovery without network
assistance. In: 12th IEEE International Conference on Network Protocols, ICNP,
pp. 328–339, October 2004 R. M¨uller et al. 136 11. Chin, E., Felt, A.P., Sekar, V., Wagner, D.: Measuring user confidence in smart-
phone security and privacy. In: 8th Symposium on Usable Privacy and Security,
SOUPS, pp. 1–16. ACM, July 2012 12. Portokalidis, G., Homburg, P., Anagnostakis, K., Bos, H.: Paranoid android: versa-
tile protection for smartphones. In: 26th Annual Computer Security Applications
Conference, ACSAC, pp. 347–356. ACM, December 2010 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. 12. Portokalidis, G., Homburg, P., Anagnostakis, K., Bos, H.: Paranoid android: versa-
tile protection for smartphones. In: 26th Annual Computer Security Applications
Conference, ACSAC, pp. 347–356. ACM, December 2010 11. Chin, E., Felt, A.P., Sekar, V., Wagner, D.: Measuring user confidence in smart-
phone security and privacy. In: 8th Symposium on Usable Privacy and Security,
SOUPS, pp. 1–16. ACM, July 2012 c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 137–142, 2017.
DOI: 10.1007/978-3-319-60774-0 11 References The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. Flow-Based Detection of IPv6-specific Network
Layer Attacks Luuk Hendriks1(B), Petr Velan2, Ricardo de O. Schmidt1,
Pieter-Tjerk de Boer1, and Aiko Pras1 1 Faculty of Electrical Engineering, Mathematics and Computer Scien
University of Twente, Enschede, The Netherlands
{luuk.hendriks,r.schmidt,p.t.deboer,a.pras}@utwente.nl
2 CESNET, a.l.e, Zikova 4, 160 00 Prague 6, Czech Republic
petr.velan@cesnet.cz 1 Faculty of Electrical Engineering, Mathematics and Computer Scien
University of Twente, Enschede, The Netherlands
{luuk.hendriks,r.schmidt,p.t.deboer,a.pras}@utwente.nl
2 CESNET, a.l.e, Zikova 4, 160 00 Prague 6, Czech Republic
petr.velan@cesnet.cz Abstract. With a vastly different header format, IPv6 introduces new
vulnerabilities not possible in IPv4, potentially requiring new detection
algorithms. While many attacks specific to IPv6 have proven to be pos-
sible and are described in the literature, no detection solutions for these
attacks have been proposed. In this study we identify and characterise
IPv6-specific attacks that can be detected using flow monitoring. By con-
structing flow-based signatures, detection can be performed using avail-
able technologies such as NetFlow and IPFIX. To validate our approach,
we implemented these signatures in a prototype, monitoring two produc-
tion networks and injecting attacks into the production traffic. 2.1
Threat Selection Process The comprehensive overview in [3] functions as a starting point in our selection
process. In that paper, Tables II and III list Security Vulnerabilities and Privacy
Vulnerabilities, respectively, indicating the origin of each threat. Step 1: We only
consider threats originating from the design of IPv6, and not any threats based
on implementation or configuration mistakes. We continue by looking at Table V
of that same paper, which is a matrix linking threats to detective, preventative
and/or reactive countermeasures. Step 2: We only consider threats that have
either no forms of countermeasure, or only a reactive countermeasure, as our
goal is detection of attacks. Lastly, we rule out threats that are not actually in
IPv6 itself, but merely in other (supporting) protocols. Step 3: Dismiss threats
based on DNS and ICMP6. 1
Introduction Monitoring network traffic is an essential aspect in today’s Internet. With the
ever-growing collection of possible network-based threats, security officers need
to stay up to date and be aware of what is possibly coming towards their net-
works and services. Intrusion Detection Systems (IDSs) play a critical role in this
scenario, offering the first insight into malicious traffic, e.g. brute-force attacks
on SSH daemons [2], or large numbers of DNS responses caused by a Distrib-
uted Denial of Service (DDoS) attack. Currently, the adoption and deployment
of IPv6 in the Internet is increasing: 16.4% of users of Google’s services have
IPv6 connectivity. North-America and Germany feature an adoption of around
30%, and Belgium is almost at 50%. With the increasing amount of IPv6 traffic
in mind, we want to know whether the flow-based detection approaches from
IPv4 are applicable, and moreover, fully covering the spectrum of IPv6 attacks. In this paper, we ask ourselves 1. which new threats are introduced by these
changes in the network layer; 2. how fundamental these threats are; and 3. how
flow-based monitoring solutions should be adapted in order to enable detection
of these new attacks. L. Hendriks et al. 38
L. Hendriks et al. 138 2
Methodology We focus on a subset of threats: we inquire the literature, and select (Sect. 2.1)
the vulnerabilities that are expected to be a long-term threat not easily miti-
gated. With the selection of threats at hand, we analyze (Sect. 2.2) their packet-
based forms, to construct flow-based signatures. The signatures are implemented
and tested on flows collected on two production networks: (1) the National
Research and Educational Network (NREN) CESNET, with 8 vantage points,
totalling 2.5G of flows (87G packets, 81.2Ti bytes); and (2) the campus net-
work UTNET, with a single vantage point, with 2.2G of flows (158.6G packets,
140.7Ti bytes). 2.2
Threat Analysis Process For each threat, the following steps are carried out: 3
Attack Signatures Our selection process described in Sect. 2.1 yields six attacks (Table 2), cate-
gorized as covert channels (exfiltration of information), DoS attacks (aiming
to overload and impair functioning of systems) and middlebox evasion (getting
around e.g. firewalls). The constructed signatures, along with their formal expla-
nation, are listed in Table 1. Note that we describe signatures from the per-
spective of the collector, not aggregation by the flow cache on the exporter. The Denial of Service (DoS) signatures have two variants: the multi-flow kind
describes an attack where a large number of destination addresses is generated
randomly, as opposed to the kind where a single destination address is used. Nat-
urally, different destination addresses lead to different flow records, and therefore
different signatures. Table 1. Signatures and notation explanation
fi
Field in packet, e.g. Source Address
5t
Shorthand for the 5-tuple flow-key
{f1, . . . fn}
Flow-key based on fields f1 . . . fn
FL
Flow Label (IPv6 header field)
#
Number of flows for flow-key or set
TC
Traffic Class (IPv6 header field)
ppf
Packets per flow
prn
Protocol Number n
pps(S)
Packets per second in flow set S
τ
Threshold, relative to context
(FK|F+)
Set of flows aggregated on FK filtered on one or more filters F
F
Selection filter, e.g. ppf = 1 for flows with a single packet, or pr0 for Protocol 0
Flow Label Covert Channel
#({FL, 5t}|FL > 0, ppf = 1) −#{5t} > τflow diff
Traffic Class Covert Channel
#({TC, 5t}|TC > 0, ppf = 1) −#{5t} > τflow diff
Multi-flow Flow Label DoS
S = ({src ip}|FL > 0, ppf = 1) , pps(S) > τpps
Multi-flow Fragmentation ID DoS
S = ({src ip}|pr44, ppf = 1) , pps(S) > τpps
Multi-flow Hop-by-Hop DoS
S = ({src ip}|pr0, ppf = 1) , pps(S) > τpps
Flow Label DoS
#{F L, 5t} −#{5t} > τflow diff
Fragmentation ID DoS
S = ({5t}|pr44, ppf > τppf ) , pps(S) > τpps
Hop-by-Hop DoS
S = ({5t}|pr0, ppf > τppf ) , pps(S) > τpps
Fragmentation Overlap
{5t|0 < fragMinOffset ≤20} Table 1. Signatures and notation explanation An overview of requirements for flow exporters is presented in Table 2. These
requirements include certain fields to be incorporated in the flow cache key
(distinguishing flows on those fields), and a new IPFIX Information Element to
be implemented. Flow-Based Detection of IPv6-specific Network Layer Attacks Flow-Based Detection of IPv6-specific Network Layer Attacks 139 For each threat, the following steps are carried out: For each threat, the following steps are carried out: 1. At the packet-level, pinpoint the protocol fields and their respective values
that make up the essence of the vulnerability. 1. At the packet-level, pinpoint the protocol fields and their respective values
that make up the essence of the vulnerability. 2. Determine if the essential features found in the previous step are still avail-
able in the aggregated form (flow level). N.B.: availability of these features
depends on which Information Elements are exported by the flow exporter. Furthermore, the flow cache should in some cases use these fields in its cache
key, in order to distinguish and export separate flow records. More details on
this follow in Sect. 3. 3. If an attack is not distinguishable based on information of a single flow,
determine the relationship between malicious flows, as well as the relationship
between the malicious and benign flows. 3. If an attack is not distinguishable based on information of a single flow,
determine the relationship between malicious flows, as well as the relationship
between the malicious and benign flows. 4. Formalize a signature based on the previous two steps, resulting in a per-
attack detection approach. 4. Formalize a signature based on the previous two steps, resulting in a per-
attack detection approach. 3
Attack Signatures Note that not all IANA assigned fields are exported per se. Table 2. Flow record requirements for implementation of signatures Table 2. Flow record requirements for implementation of signatures Threat
Flow key IANA
New IE
Flow Label CC
{FL, 5t}
id31
Traffic Class CC
{TC, 5t} id5
Flow Label DoS
{FL, 5t}
id31
Fragmentation ID DoS
{5t}
id4, id54
Hop-by-Hop Option DoS {5t}
id4
Fragmentation Overlap
{5t}
id4
minFragOffset 140 L. Hendriks et al. 4
Evaluation of the Signatures The proposed signatures are evaluated on real production traffic, in which we
inject generated attacks. As the DoS attacks could harm the routers, a safe
number of packets is used, likely lower than a real attack but still allowing
verification of our signatures. We describe the generated attacks, and discuss
the performance of the signatures with respect to both these attacks and the
production traffic, below. Generated Attacks: Flow Label and Traffic Class Covert Channels: Sending 100, 500, 1000 packets,
within a 5 min time-frame, towards a single host. Flow Label and Traffic Class Covert Channels: Sending 100, 500, 1000 packets,
within a 5 min time-frame, towards a single host. Flow Label, Fragmentation ID, Hop-by-Hop Option DoS: Sending 500 packets
at line rate, towards a single host; Sending 500 packets at line rate, towards
randomly generated hosts in a /64 network. Fragmentation Overlap: Sending flows of 2, 10, 20 packets, with second packet
offsets of 1, 4, 10, 20 towards a single host. Performance: Flow Label Covert Channel: The flow records related to the covert channel are
successfully distinguished, using a threshold of τpkt = 50. No other positives
were found in the dataset, meaning the signature has a low false positive rate
but possibly a non-zero false negative rate. Traffic Class Covert Channel: The Traffic Class can hold different values within
a single flow, and we do observe this in production traffic. Most commonly, these
are a zero and a non-zero value: including the TC-field in the aggregation thus
results in two flows. Using τfl = 10, i.e. marking flows with 10 or more different
Traffic Class values as attacks, the signature distinguishes all the injected attacks
from the production traffic. Similar to the Flow label Covert channel, no other
positives where marked, pointing out a low false positive rate but a possible
non-zero false negative rate. Flow Label Flood: Detection of a Flow Label flood to a single destination address
is similar to detecting a Flow Label covert channel, thus results are equivalent. Distinguishing the covert channel from the DoS attack is challenging. Multi-flow
signature has a false positive rate, albeit because it marks other threats and not
benign traffic. For example, a SYN scan has, on the flow level, vast similarities
when compared to the flow label flood attack: a large number of end hosts is
being connected to from a single source address, with every initiated connection
having a new (thus different) Flow Label. Hop-by-Hop Flood: As the Hop-by-Hop Options are not widely used (most of it
is link local traffic, with only one or two packets per flow), simplistic thresholds
for detection work: τppf = 10 suffices. This means scalable detection without the
need for extra Information Elements or extra processing at the exporter is trivial. A possible form of false positives exists however, as we observed two times on the Flow-Based Detection of IPv6-specific Network Layer Attacks 141 NREN: ping sweeps with Hop-by-Hop Options match this signature. Marking
the spread version of the attack is successful, without any other positives. NREN: ping sweeps with Hop-by-Hop Options match this signature. Marking
the spread version of the attack is successful, without any other positives. Fragmentation Flood: Detection of flooding based on the Fragmentation ID has
several caveats. By definition, a flow with fragmented packets consists of more
than one packet, but an exception of this characteristic are the atomic fragments. 5
Conclusions IPv6 comes with a plethora of threats specific to this new version of the IP
protocol. By systematically characterising threats described in literature, we
found six of these threats to be fundamental, i.e. based on the protocol specifi-
cation and without detection approaches for attacks as of yet. In this study, we
proposed flow-based signatures to perform such detection. By implementing a
prototype, we proved the validity and limitations of these signatures, and defined
the requirements for flow measurement equipment to allow for applying detec-
tion of attacks based on these signatures. These requirements show adaptations
to flow equipment are necessary to enable for detection of these new attacks. By deploying our prototype on two production networks and injecting attacks
into the production traffic, we showed our signatures are able to successfully dis-
tinguish the attacks from benign traffic without any false negatives. We provide
both the detection prototype as well as the code used for generation of attacks
as free and open source software [1]. Acknowledgments. This work has been supported by the project Reg. No. CZ.02.1.01/0.0/0.0/16 013/0001797 co-funded by the Ministry of Education, Youth
and Sports of the Czech Republic and European Regional Development Fund, and the
Ministry of the Interior of the Czech Republic under project no. VI20162019029. Performance: Signatures based on fragmented but single packet flows therefore yield false
positives. As the sending rate and number of sent packets are crucial in the
success of a flooding attack, we can choose thresholds that eliminate these false
positives: τpps = 5000/s, τppf = 200. Our attacks are identified without any other
flows being marked, again pointing out a low false positive rate but a possible
false negative rate. The case where destination addresses were generated and
the flooding attack was hidden in a large number of different 5-tuple flows, is
successfully detected. Fragmentation Overlap: The approach based on fragMinOffset marks all our
injected attacks. The lowest value observed in the production traffic was 64, so
no positives other than our injected attacks were marked. 1. IPv6 L3 Threat Detection. https://github.com/ut-dacs/IPv6-L3-threat-detection/
2. Hofstede, R., Hendriks, L., Sperotto, A., Pras, A.: SSH compromise detection using
NetFlow/IPFIX. ACM SIGCOMM CCR 44(5), 20–26 (2014)
3. Ullrich, J., Krombholz, K., Hobel, H., Dabrowski, A., Weippl, E.R.: IPv6 security:
attacks and countermeasures in a nutshell. In: USENIX WOOT (2014) c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 143–148, 2017.
DOI: 10.1007/978-3-319-60774-0 12 References /
( )
(
)
3. Ullrich, J., Krombholz, K., Hobel, H., Dabrowski, A., Weippl, E.R.: IPv6 security:
attacks and countermeasures in a nutshell. In: USENIX WOOT (2014) 142
L. Hendriks et al. L. Hendriks et al. 142 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. Department of Computer Science,
Oslo and Akershus University College of Applied Sciences, Oslo, Norway
harek.haugerud@hioa.no Department of Computer Science,
Oslo and Akershus University College of Applied Sciences, Oslo, Norway
harek.haugerud@hioa.no Abstract. Hybrid cloud technology is becoming increasingly popular
as it merges private and public clouds to bring the best of two worlds
together. However, due to the heterogeneous cloud installation, facilitat-
ing a hybrid cloud setup is not simple. Despite the availability of some
commercial solutions to build a hybrid cloud, an open source implemen-
tation is still unavailable. In this paper, we try to bridge the gap by
providing an open source implementation by leveraging the power of
Apache Mesos. We build a hybrid cloud on the top of multiple cloud
platforms, private and public. Keywords: Opensource hybrid cloud · Apache Mesos · Data segmen-
tation · Fault tolerance Towards a Hybrid Cloud Platform
Using Apache Mesos Noha Xue, H˚arek Haugerud(B), and Anis Yazidi Department of Computer Science,
Oslo and Akershus University College of Applied Sciences, Oslo, Norway
harek.haugerud@hioa.no 1
Introduction The use of cloud computing is becoming more common, bringing along the advan-
tages of flexibility and abundance of available resources, but also a higher degree
of complexity along with privacy and security concerns. The concepts of multi-
cloud and hybrid cloud are not new and several companies explore and capitalize
these concepts. Most of the available solutions are commercial. Different vendors
including Dell, IBM and HP provide hybrids cloud solutions [1,2]. The MODA-
Clouds project [3] utilizes several tools to provide an environment for utilizing
multiple cloud providers. Several large companies are offering hybrid cloud solu-
tions, often in conjunction with their existing product portfolio. VMWare is
offering a hybrid cloud solution called vRealize suite which provides one inter-
face to manage the entire hybrid cloud platform [4,5]. Other companies like
Cisco [6], IBM [7] and RackSpace [8] are also offering hybrid cloud solutions. Another attempt addresses the challenges of managing heterogeneous virtual
environments to create a hybrid cloud platform [9]. PaaSage is an interesting
initiative for building a hybrid cloud solution using a defined deployment model,
Cloud Application Modeling and Execution Language (CAMEL) [10]. However,
there has been no practical demonstration of using open-source and freely avail-
able clustering technology to attempt to address the multitude of challenges N. Xue et al. 144 when creating a hybrid cloud platform that is available and supports data seg-
mentation. This paper outlines an attempt to prototype such a solution in addi-
tion to facilitate cloud bursting using spot price instances. Borja et al. intro-
duced OpenNebula in [11], which is one of the most popular open source Virtual
Infrastructure Manager (VIM). OpenNebula permits to abstract resources of an
existing local Grid and a cloud infrastructure. At the heart of OpenNebula we
find Haizea [12] which is a VM-based Lease Manager that enhances the resource
scheduling manager with advanced reservation of resources and queueing of best
effort requests. Nevertheless, OpenNebula suffers from the single point failure
problem [13]. In this paper, we present a lightweight solution, that is tolerant to
different failure scenarios. Similarly, it is also possible to create a Hybrid Cloud
With AWS and Eucalyptus. This paper will explore and document the attempts at designing and proto-
typing a possible opensource solution for constructing a computer cluster built
on top of private servers and external cloud providers. Towards a Hybrid Cloud Platform Using Apache Mesos Towards a Hybrid Cloud Platform Using Apache Mesos Towards a Hybrid Cloud Platform Using Apache Mesos
145
Fig. 1. Communication flow between an Apache Mesos slave node and master node
with the registration attempt failing due to how public IP-addresses are handled in
cloud platforms. Fig. 1. Communication flow between an Apache Mesos slave node and master node
with the registration attempt failing due to how public IP-addresses are handled in
cloud platforms. Fig. 2. Prototype 1: Maximizing availability. Distributing the master nodes and thereby
the risks. Fig. 2. Prototype 1: Maximizing availability. Distributing the master nodes and thereby
the risks. 2
Design and Implementation An Apache Mesos cluster including both master nodes and slaves nodes were
successfully installed and configured in Altocloud, with slave nodes correctly
registering themselves to the cluster through the leading master node. However,
when attempting to register a slave node running at Amazon Web Services EC2
peculiar activity was observed. The traffic from the slave node located at EC2
managed to successfully send a registration request to the leading master node,
passing through multiple layers of network abstraction including two layers of
NAT. Although the master node receives the registration requests, no regis-
tration acknowledge is ever sent back. Eventually, the cause was discovered to
be a combination of the use of NAT and the way Mesos nodes communicates
between each other. When a slave node sends a registration request, it includes
information about the resources available and an IP-address. The IP-address
sent along is the one that is defined on the network interface bound by the
Apache Mesos process. Furthermore, in a cloud environment like Altocloud and
Amazon Web Services EC2, the public IP-addresses are loosely coupled with the
virtual machine and functions similarly as NAT does. Consequently, the Mesos
master attempts to send the acknowledgement and other internal traffic meant
for that slave node to the non-routable private IP-address. The communication
flow is illustrated in Fig. 1. By using VPN tunneling, the need for allocating public IP-addresses for each
node disappears for the purpose of maintaining the cluster, as the private IP-
addresses becomes routable within the hybrid cloud platform. With the exception
of the extra infrastructure to maintain a VPN, the prototype is identical to the
proposed proposed design. Figure 2 illustrates the final implementation of the
prototype, showing how the Mesos master nodes are distributed between the
different availability regions. 145 Test Scenarios A Mesos Slave Process Becomes Unavailable. In the event of a Mesos slave node
becoming unavailable for some reason, the Mesos master node allows a default
timeout period of 75 s to pass before procedures for deactivating the slave node N. Xue et al. 146 is begun. Should the slave node start responding within this timeout period,
nothing will happen and both the Mesos master node and the slave node simply
ignores the temporary unavailability. However, if the timeout period is exceeded and the slave nodes is still unavail-
able, the Mesos master node will attempt to deactivate the Mesos slave process
on the slave node before it from the list of available slave nodes. Tasks that were
lost will be rescheduled to other slave nodes with available capacity. Should a slave node simply be temporarily disconnected from the master
node, but exceed the timeout period, the Mesos master will forcibly shut the
Mesos slave node down. To account for such scenarios, the official Apache Mesos
documentation recommends monitoring the Mesos slave process and restart if it
should be terminated for any reason. In this case, this is achieved with a simple
check using Monit. In Listing 1.1 log events of such a case is listed. The Working Mesos Master Instance Cease to Function. ZooKeeper maintains
an active connection to the participants of the quorum and will after a very short
timeout lasting a few seconds, will initiate a new leader electing for choosing a
new leading Mesos master node. As long as the number of functional Mesos
master nodes is equal or higher than the quorum size, a new leader will be
elected and will replace the unresponsive Mesos master node. 1 17:34:23.298998 Shutting down slave ...5050-5669-S3 due to health check timeout
2 17:34:23.300134 Removing slave ...5050-5669-S3 at slave(1)@192.168.187.205:5051
(192.168.187.205)
→
3 17:34:23.301009 Removed slave 20150501-230056-2407081856-5050-5669-S3
4 17:34:23.536837 Notifying framework ...5050-27030-0006 (marathon) at
...473b-b57a-83121a00a01c@128.39.121.140:43217 of lost slave ...5050-5669-S3
(192.168.187.205) after recovering
→
→
5 17:34:29.017205 Slave ...5050-5669-S3 at slave(1)@192.168.187.205:5051 (192.168.187.205)
attempted to re-register after removal; shutting it down
→
6 17:34:57.329751 Registering slave at slave(1)@192.168.187.205:5051 (192.168.187 .205) with
id ...5050-5669-S4
→
Listing 1.1: Excerpt from /var/log/mesos/mesos-master.INFO showing the forced shut down of
the Mesos slave process at 192.168.187.205 and the registration as new slave at end. Truncated for
increased readability. 1 17:34:23.298998 Shutting down slave ...5050-5669-S3 due to health check timeout
2 17:34:23.300134 Removing slave ...5050-5669-S3 at slave(1)@192.168.187.205:5051
(192.168.187.205)
→
3 17:34:23.301009 Removed slave 20150501-230056-2407081856-5050-5669-S3
4 17:34:23.536837 Notifying framework ...5050-27030-0006 (marathon) at
...473b-b57a-83121a00a01c@128.39.121.140:43217 of lost slave ...5050-5669-S3
(192.168.187.205) after recovering
→
→
5 17:34:29.017205 Slave ...5050-5669-S3 at slave(1)@192.168.187.205:5051 (192.168.187.205)
attempted to re-register after removal; shutting it down
→
6 17:34:57.329751 Registering slave at slave(1)@192.168.187.205:5051 (192.168.187 .205) with
id
5050-5669-S4
Test Scenarios id ...5050-5669-S4
→ Listing 1.1: Excerpt from /var/log/mesos/mesos-master.INFO showing the forced shut down of
the Mesos slave process at 192.168.187.205 and the registration as new slave at end. Truncated for
increased readability. This scenario was tested with a simple reboot of the instance where the
leading Mesos master was running. The backup Mesos masters quickly discovers
the loss of connection to the leading Mesos master and promptly, with the use of
ZooKeeper elects a new leading Mesos master node. The rebooted Mesos master
node later joins the cluster as a backup node after coming back online. The setup proposed in this prototype has three Mesos master nodes, with
the quorum size set to two. This means that among the Mesos master nodes, one
can fail without crippling the cluster, as the quorum size dictates the number of
election participants that has to be able to communicate to be able to elect a
new leader. Towards a Hybrid Cloud Platform Using Apache Mesos 147 An Entire Region within the Hybrid Cloud Becomes Unavailable. If an entire
region becomes unavailable, the Mesos nodes located within those regions will by
extension also become unavailable. In this particular case, the loss of one single
site equals the loss of one Mesos master node and four slave nodes. Each node,
depending on the type, is handled as specified in the test scenarios mentioned
above. This was tested by taking down the VPN tunnels at the VPN gateway of
the concerned region. This cuts all communication between the affected region
and the other ones. As expected the Mesos master nodes continued without any
issues, as the current leader was not the affected one. As for the affected Mesos
slave nodes, after the timeout of 75 s, the leading Mesos master node determined
that the slave nodes were unresponsive deactivated them. The Hybrid Cloud Splits and Semi-isolates Part of the Platform. In the event
of split in the hybrid cloud, resulting in a partly isolated availability region, the
quorum mechanics will prevent inconsistencies of the cluster and avoid issues
like the split-brain problem. To test this scenario, two simple iptables DROP rules was added on the
Mesos master node located in Frankfurt with the IP address 192.168.0.5. 3
Conclusion This paper presents a prototype of a hybrid cloud platform using Apache Mesos
to weave together heterogeneous clouds and geographical locations into a uni-
fied platform. The prototype proposed focuses on a specific perspective, namely,
maximizing availability. Test Scenarios iptables -A INPUT -s 10.0.19.5 -j DROP
iptables -A OUTPUT -d 10.0.19.5 -j DROP The leading Mesos master node at the current time was 10.0.19.5, with
nothing occurring immediately as a result of the iptables DROP rules. The
leading master continued with no issues and other two standby Mesos masters
correctly redirected to the leading master node. However, after rebooting the
ZooKeeper process and Mesos master process on the master nodes, the cluster is
unable to elect a new leader. Immediately after the iptables DROP rules were
removed, a new leading Mesos master was elected and operations continued as
normal. 1. Connor, T.R., Southgate, J.: Automated cloud brokerage based upon continuous
real-time benchmarking. In: 2015 IEEE/ACM 8th International Conference on
Utility and Cloud Computing (UCC), pp. 372–375. IEEE (2015) c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 149–152, 2017.
DOI: 10.1007/978-3-319-60774-0 13 References 1. Connor, T.R., Southgate, J.: Automated cloud brokerage based upon continuous
real-time benchmarking. In: 2015 IEEE/ACM 8th International Conference on
Utility and Cloud Computing (UCC), pp. 372–375. IEEE (2015) N. Xue et al. 148 2. Breiter, G., Naik, V.K.: A framework for controlling and managing hybrid cloud
service integration. In: 2013 IEEE International Conference on Cloud Engineering
(IC2E), pp. 217–224. IEEE (2013) 3. MODAClouds. Modaclouds. http://www.modaclouds.eu 4. VMWare, Inc. vrealize suite. http://www.vmware.com/products/vrealize-suite/
features.html 5. VMWare, Inc., Cloud Computing. http://www.vmware.com/cloud-computing/
hybrid-cloud.html 6. Butler, B.: Re-examining ciscos intercloud strategy, January 2015. http://www. networkworld.com/article/2864857/cloud-computing/re-examining-cisco-s-intercl
oud-strategy.html gy
7. IBM. Private and hybrid cloud. http://www.ibm.com/cloud-computing/uk/en/
private-cloud.html 8. Rackspace, Inc., Hybrid cloud computing, hybrid hosting by rackspace. http://
www.rackspace.com/cloud/hybrid 9. Breiter, G., Naik, V.: A framework for controlling and managing hybrid cloud
service integration. In: 2013 IEEE International Conference on Cloud Engineering
(IC2E), pp. 217–224, March 2013 (
)
10. PaaSage. Paasage: Model-based cloud platform upperware. http://www.paasage. eu 11. Borja, S., Ruben, M., Ignacio, M., Ian, F.: An open source solution for virtual
infrastructure management in private and hybrid clouds. IEEE Internet Comput. 1, 14–22 (2009) 12. Kov´acs, ´A., Lencse, G.: Modelling of virtualized servers. In: 2015 38th International
Conference on Telecommunications and Signal Processing (TSP), pp. 241–245. IEEE (2015) 13. Feller, E., Rilling, L., Morin, C.: Snooze: a scalable and autonomic virtual machine
management framework for private clouds. In: 2012 12th IEEE/ACM International
Symposium on Cluster, Cloud and Grid Computing (CCGRID 2012), pp. 482–489,
May 2012 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. Visual Analytics for Network Security and
Critical Infrastructures Karol´ına Bursk´a(B) and Radek Oˇslejˇsek Faculty of Informatics, Masaryk University, Brno, Czech Republic
burska@mail.muni.cz, oslejsek@fi.muni.cz Faculty of Informatics, Masaryk University, Brno, Czech Republic
burska@mail.muni.cz, oslejsek@fi.muni.cz Abstract. A comprehensive analysis of cyber attacks is important for
better understanding of their nature and their origin. Providing a suffi-
cient insight into such a vast amount of diverse (and sometimes seem-
ingly unrelated) data is a task that is suitable neither for humans nor
for fully automated algorithms alone. Not only a combination of the
two approaches but also a continuous reasoning process that is capable
of generating a sufficient knowledge base is indispensable for a better
understanding of the events. Our research is focused on designing new
exploratory methods and interactive visualizations in the context of net-
work security. The knowledge generation loop is important for its ability
to help analysts to refine the nature of the processes that continuously
occur and to offer them a better insight into the network security related
events. In this paper, we formulate the research questions that relate to
the proposed solution. Keywords: Visual analytics · Network security · Knowledge generation 2
Research Questions and Proposed Approaches How to Model Cyber-Security Data and Its Semantics? Cyber security
data has a strong heterogeneous nature. Data sets can be temporal, geospatial,
multivariable, or graph-based, for instance. And also, mixed together. Although
there exist some formalizations that describe how various data types can be
mapped to visual properties [8] in general, a clear taxonomy of data types used
in cyber security domain is missing. However, a formal classification scheme
is necessary if we want to build an adaptive data gathering and construct a
knowledge base – two mandatory parts of any visual analysis loop. In our research, we initially focus on the design of taxonomies for cyber
security data and corresponding analytical processes. We plan to utilize for-
mal OWL ontologies to provide semantically correct vocabulary enabling as to
(semi)automatically construct adaptable data sets and derived knowledge mod-
els. Using existing taxonomies and approaches, e.g. those described in [1,6,13],
we aim to unite the different perspectives and apply them in the visual analysis
loop in the cyber security domain. How to Provide Insight into Cyber Security Processes via Exploratory
Visualizations? Many works confirm that the involvement of the human fac-
tor in the process of data analysis may contribute to revealing new informa-
tion in a significant way [5,12]. One of the basic principles used in this field is
the visual analytics process by Keim et al. [7], which is described as an app-
roach that combines data analysis, visualization, and human factor, as well
as the areas of cognition and perception. This approach follows the Shneider-
man’s visual information-seeking mantra: “Overview first, zoom and filter, then
details-on-demand” [11]. By applying this mantra in the visual analysis domain,
Sacha et al. [10] proposed an approach enabling the visual analytic theories to
go beyond the inclusion of the human factor in the process, to the theory where
human is a part of the loop [3]. Our approach to the cyber security knowledge management and its visual
analysis would combine the Keim’s and Sacha’s approaches. Their models have to
be significantly adapted since the cyber security domain requires a wide range of
network-related manipulation techniques. Our model would consist of two parts. The first part would deal with the automated processes connected to data mon-
itoring and knowledge management, while the second part would involve human
interactions by means of exploratory visualizations. 1
Introduction Although network security is strongly connected with technology (e.g., network
infrastructure, cloud computing), the context is usually much broader and must
be mediated by human interaction. While some of the known attack methods
may be detectable rather easily, many attacks can be identified only with the
participation of a human, by analysis. The analysts’ goals are to identify, track,
and understand these attacks. One of the viable approaches is to combine the
human flexibility, creativity, and background knowledge with the enormous stor-
age and processing capacities of today’s computers to gain insight into complex
problems and to understand causality. Especially, when involving large and com-
plex data sets that require a high degree of interaction, the support of knowledge
generation techniques is likely to prove as very beneficial. In what follows, we formulate research questions that are related to the loop
of exploratory visual analysis in the context of cyber security. Each question
aims to describe a broader motivation and current state and then formulates
approaches enabling us to tackle the goals in proposed PhD thesis. K. Bursk´a and R. Oˇslejˇsek 150 2
Research Questions and Proposed Approaches Unfortunately, there is no
clear separation between the two parts since the whole model for exploratory
visual analysis attempts to connect the benefits of both – humans are creative
and able to find subtle connections between two seemingly unrelated events, but
they miss the ability to deal with large data sets. On the contrary, computers
offer large storage spaces and fast data processing, but they lack the human rea-
soning and the background knowledge of the problem domain. Therefore, finding
a balanced solution based on the feasible technical background makes this goal
challenging. Visual Analytics for Network Security 151 How to Utilize Exploratory Visualizations for Efficient Protection of
Critical Information Infrastructures? Protection of critical information
infrastructures is ensured by security experts. Their skills and the ability to react
to incidents quickly and correctly are affected by two factors: a training and an
online situation awareness. In general, decision making is viewed as consisting
of an analyst’s state of knowledge in a dynamically changing environment [4]. [ ]
To facilitate a cyber protection training and to evaluate benefits of visu-
alization techniques for situation awareness, we attempt to use KYPO Cyber
Range [9], where various attacks and threats can be easily simulated. KYPO
enables us to focus on linking the knowledge base with suitable visualizations
and to evaluate their benefits. New approaches can be tested and evaluated by
means of cyber defense exercises focused on improving skills of participants [2]. Acknowledgements. This research was supported by the Security Research Pro-
gramme of the Czech Republic 2015–2020 (BV III/1 VS) granted by the Ministry of
the Interior of the Czech Republic under No. VI20162019014 Simulation, detection,
and mitigation of cyber threats endangering critical infrastructure. References 1. Chi, E.H.: A taxonomy of visualization techniques using the data state reference
model. In: IEEE Symposium on Information Visualization 2000 (2000)
ˇ
ˇ 2. ˇCeleda, P., ˇCegan, J., Vykopal, J., Tovarˇn´ak, D.: KYPO - a platform for cyber
defence exercises. In: M&S Support to Operational Tasks Including War Gaming,
Logistics, Cyber Defence. NATO Science and Technology Organization (2015) 3. Endert, A., et al.: The human is the loop: new directions for visual analytics. J. Intell. Inf. Syst. 43(3), 411–435 (2014) 4. Endsley, M.R.: Toward a theory of situation awareness in dynamic systems. Hum. Factors: J. Hum. Factors Ergon. Soc. 37(1), 32–64 (1995) g
( ),
(
)
5. Fischer, F.: Visual analytics for situational awareness in cyber security (2016) 6. Gao, J., et al.: Ontology-based model of network and computer attacks for security
assessment. J. Shanghai Jiaotong Univ. (Sci.) 18(5), 554–562 (2013) (
)
( )
(
)
7. Keim, D.A., Mansmann, F., Stoffel, A., Ziegler, H.: Visual Analytics. Springer,
Heidelberg (2009) 8. Kott, A., Wang, C., Erbacher, R.F.: Cyber Defense and Situational Awareness. Springer, New York (2014) 9. Kouˇril, D., et al.: Cloud-based testbed for simulation of cyber attacks. In: IEEE
Network Operations and Management Symposium (NOMS), pp. 1–6, May 2014 10. Sacha, D., et al.: Knowledge generation model for visual analytics. IEEE Trans. Vis. Comput. Graph. (Proc. Vis. Anal. Sci. Technol.) 20(12), 1604–1613 (2014) p
p
(
)
(
),
(
)
11. Shneiderman, B.: The eyes have it: a task by data type taxonomy for information
visualizations. In: Proceedings 1996 IEEE Symposium on Visual Languages (1996)
12. Sun, G., Wu, Y., et al.: A survey of visual analytics techniques and applica- (
)
(
)
(
)
11. Shneiderman, B.: The eyes have it: a task by data type taxonomy for information
visualizations. In: Proceedings 1996 IEEE Symposium on Visual Languages (1996) 12. Sun, G., Wu, Y., et al.: A survey of visual analytics techniques and applica-
tions: state-of-the-art research and future challenges. J. Comput. Sci. Tech. 28(5),
852–867 (2013) 13. Zareen, S., et al.: UCO: a unified cybersecurity ontology. In: Proceedings of the
AAAI Workshop on Artificial Intelligence for Cyber Security (2016) 52
K. Bursk´a and R. c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 153–156, 2017.
DOI: 10.1007/978-3-319-60774-0 14 References Oˇslejˇsek 152 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. Preserving Relations in Parallel Flow Data
Processing Tom´aˇs ˇCejka1(B) and Martin ˇZ´adn´ık2 Tom´aˇs ˇCejka1(B) and Martin ˇZ´adn´ık2 1 FIT, CTU in Prague, Prague, Czech Republic
cejkato2@fit.cvut.cz
2 CESNET, a.l.e., Prague, Czech Republic
zadnik@cesnet.cz Abstract. Network monitoring produces high volume of data that must
be analyzed ideally in near real-time to support network security opera-
tions. It is possible to process the data using Big Data frameworks, how-
ever, such approach requires adaptation or complete redesign of process-
ing tools to get the same results. This paper elaborates on a parallel
processing based on splitting a stream of flow records. The goal is to cre-
ate subsets of traffic that contain enough information for parallel anom-
aly detection. The paper describes a methodology based on so called
witnesses that helps to scale up without any need to modify existing
algorithms. 1
Introduction Common architecture of monitoring large networks contains multiple observation
points measured by monitoring probes and a central collector with captured data. This approach creates a global view of the network traffic. In addition, it allows
for analysis and detection of global events that are less visible from a local view. This approach works well on small networks, however, since the network
traffic grows, processing all data on one place reaches limits of resources such
as memory capacity. In addition, various network events produce data that
reach maximal performance of a single machine. Altogether, network monitoring
becomes a Big Data processing and some scalable approach must be considered. As it is described later in Sect. 2, Big Data principles are being studied for
last years. However, a general methodology of splitting network data into subsets
is missing. The aim of this work is to describe a principle how to split data with
respect to internal relations and used processing algorithms in order to analyze
balanced data subsets while the needed information still remains together. The goal is to allow processing huge amounts of flow data using analysis
tools that are not designed for a distributed environment. A correct selection of
data subsets can improve current mechanisms of data distribution or sampling
without loosing information needed by detection algorithms. T. ˇCejka and M. ˇZ´adn´ık 154 2
Related Work MapReduce was used for network data processing e.g. in [6,7], however, the
authors used a distributed database or a distributed filesystem to store data
files. Paper [4] analyzes IP, TCP and HTTP traffic stored in offline files using
Hadoop and MapReduce. Paper [1] analyzes campus network using several types
of MapReduce jobs (e.g. measuring volume of traffic per subnet). Authors of [5] try to use Apache Spark framework with Netmap to extract
traffic features for a packet-based detection of different types of DDoS attacks
in real-time. The detection uses machine learning methods. The authors rely on
a distributed storage and an abstraction of objects called Resilient Distributed
Dataset, however, no efficient data splitting is discussed. The paper notes that
usage of sampled data produces many false-positives. Semantic relations in data and possible negative effects of splitting data were
mentioned in [3]. The authors present experiments with Hashdoop, an improved
Hadoop, that splits data using CRC hashes of src and dst IPs. The authors chose
a simple packet counting and ASTUTE algorithm for parallel processing. The
splitting based just on IP addresses is a single case in our methodology. 3
Proposed Approach The main requirement is to identify which parts of data must stay together to
preserve data relations and which parts can be split into subsets. A detection algorithm can be described as a function with data about
network traffic as its input and alerts (detected events) as its output. Generally,
the input data is a mixture of benign and malicious traffic. The goal of a detec-
tion algorithm is to identify the malicious traffic and to generate an alert that
describes the detected event. The algorithm is successful if it observes at least
a minimal subset of malicious traffic which triggers the alert. Lets call the
instance of a minimal subset a witness. If a witness gets divided, the malicious
traffic is not detectable with the same detection algorithm anymore because
there is not enough information for decision. As a result, data can be divided
for parallel processing in any way that does not break witnesses. In practice, there are many different detection algorithms processing the same
data to detect various types of malicious traffic. As a result, multiple different
witnesses must be preserved at the same time which complicates data splitting. Data can contain many witnesses that identify the same malicious traffic,
while any of them is sufficient for a successful detection. In order to design a
data splitter a particular type of witness should be characterized. This kind of
characterization describes what data an algorithm analyzes, how the malicious
traffic looks like and what is the configuration of an algorithm. 5
Conclusion This paper addressed a network traffic analysis in a distributed environment. There are many papers focusing on existing Big Data frameworks but, to our best
knowledge, a general methodology of splitting a stream of flow data is missing. This research aims to describe data relations that must be preserved for the
parallel analysis. The data relations, types of malicious traffic and used detection
algorithms with their parameters define so called witnesses. Since this research
is rather a work-in-progress, we have some preliminary results. However, the
experiments with real data show that respecting witnesses allow for distributed
processing without significant impact on detection results. As a future work, we are going to explore the principle of witnesses in more
detail. Moreover, based on witnesses, an algorithm of real-time reconfiguration
of the splitter to scale up the distributed system would be useful. Acknowledgments. This work was supported by the Technology Agency of the Czech
Republic under No. TA04010062 Technology for processing and analysis of network
data in big data concept and grant No. SGS17/212/OHK3/3T/18 funded by Ministry
of Education, Youth and Sports of the Czech Republic. 4
Evaluation To evaluate the witness-based splitting, we analyze data distribution among
computation nodes and overall detection results. We need to compare results Preserving Relations in Parallel Flow Data Processing 155 of a single instance and results of a distributed environment. As the distrib-
uted processing generates some alerts multiple times a deduplication based on
timestamps, type of events and other information contained in the alerts is
necessary. For the evaluation, we use a NEMEA framework [2] which can be easily run
in a single instance as well as in a distributed configuration. There are several
detection modules in NEMEA and some of them were presented in our previous
work. However, the presented principle can be used with any system that allows
to modify an algorithm of data splitting. The first experiments with splitting flow data with respect to potential wit-
ness showed that it is possible to distribute flow data almost uniformly and there
is no significant difference between detection results of a single instance and the
distributed environment. The measured difference was about 1% which is caused
by inaccurate timing of stream-wise real-time analysis during our experiments. 2. Cejka, T., et al.: NEMEA: a framework for network traffic analysis. In: Proceedings
of CNSM (2016). doi:10.1109/CNSM.2016.7818417 1. Bumgardner, V.K., el al.: Scalable hybrid stream and Hadoop network analysis
system. In: Proceedings of the 5th ACM/SPEC ICPE (2014). doi:10.1145/2568088.
2568103 References 1. Bumgardner, V.K., el al.: Scalable hybrid stream and Hadoop network analysis
system. In: Proceedings of the 5th ACM/SPEC ICPE (2014). doi:10.1145/2568088. 2568103 1. Bumgardner, V.K., el al.: Scalable hybrid stream and Hadoop network analysis
system. In: Proceedings of the 5th ACM/SPEC ICPE (2014). doi:10.1145/2568088. 2568103 1. Bumgardner, V.K., el al.: Scalable hybrid stream and Hadoop network analysis
system. In: Proceedings of the 5th ACM/SPEC ICPE (2014). doi:10.1145/2568088. 2568103 2. Cejka, T., et al.: NEMEA: a framework for network traffic analysis. In: Proceedings
of CNSM (2016). doi:10.1109/CNSM.2016.7818417 2. Cejka, T., et al.: NEMEA: a framework for network traffic analysis. In: Proceedings
of CNSM (2016). doi:10.1109/CNSM.2016.7818417 3. Fontugne, R., et al.: Hashdoop: A MapReduce framework for network anom-
aly detection. In: Proceedings of INFOCOM (2014). doi:10.1109/INFCOMW.2014. 6849281 3. Fontugne, R., et al.: Hashdoop: A MapReduce framework for network anom-
aly detection. In: Proceedings of INFOCOM (2014). doi:10.1109/INFCOMW.2014. 6849281 T. ˇCejka and M. ˇZ´adn´ık 156 4. Ibrahim, L.T., et al.: A study on improvement of internet traffic measurement and
analysis using Hadoop system. In: Proceedings of ICEEI (2015). doi:10.1109/ICEEI. 2015.7352545 5. Karimi, A.M., et al.: Distributed network traffic feature extraction for a real-time
IDS. In: Proceedings of EIT (2016). doi:10.1109/EIT.2016.7535295 6. Lee, Y., et al.: Toward scalable internet traffic measurement and analysis with
Hadoop. ACM SIGCOMM Comput. Commun. Rev. 43(1), 5–13 (2013). doi:10. 1145/2427036.2427038 /
7. Zhang, J., Zhang, Y., Liu, P., He, J.: A spark-based DDoS attack detection
model in cloud services. In: Bao, F., Chen, L., Deng, R.H., Wang, G. (eds.)
ISPEC 2016. LNCS, vol. 10060, pp. 48–64. Springer, Cham (2016). doi:10.1007/
978-3-319-49151-6 4 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 159–164, 2017.
DOI: 10.1007/978-3-319-60774-0 15 SmartDEMAP: A Smart Contract Deployment
and Management Platform Markus Knecht(B) and Burkhard Stiller Communication Systems Group CSG, Department of Informatics IfI,
University of Z¨urich, Binzm¨uhlestrasse 14, 8050 Z¨urich, Switzerland
markus.knecht2@uzh.ch Communication Systems Group CSG, Department of Informatics IfI,
University of Z¨urich, Binzm¨uhlestrasse 14, 8050 Z¨urich, Switzerland
markus.knecht2@uzh.ch Abstract. Smart contracts on a blockchain behave exactly as specified
by their code. To be sure that a smart contract behaves as expected, the
end-user has to either analyze its code or trust a potentially anonymous
developer or auditor to do so. This approach proposes a smart contract
deployment and management platform that can execute development
tools and code quality tools in a trusted way and uses this to reduce the
trust required into the smart contract developer or auditor. Additionally,
such a platform can provide new capabilities for developers aiding them
in the creation of smart contracts. 2
Hypotheses An investigation into the current state of smart contract development has shown
that there currently is a high risk for end-users when interacting with smart
contracts, as shown by the “The DAO” incident [10], where an attacker exploited
a bug to steel 3.6 million in ether. The following hypotheses are premises for
developing and analysing SmartDEMAP. In the project it should be researched
how well these premises can mitigate the respective risks. Hypothesis 1: A platform on the blockchain, which provides access to trusted
execution of development and code quality tools, enables the development of
smart contracts which can manage, verify, and analyze the code of other smart
contracts in order to increase their robustness as well as reducing the trust
required in developers and auditors. Sub-hypothesis 1.1: The ability to associate attributes with a smart contract
based on a trusted analysis of its code, where the results can be queried and
analyzed by other smart contract or external sources, enables the detection of
misbehaving smart contracts. Sub-hypothesis 1.2: A smart contract that controls the compilation and
deployment of other smart contracts by using development and code quality
tools, allows a developer to add new features or bug fixes to a smart contract
after it has been deployed, without the need for end-users to trust the developer. Sub-hypothesis 1.3: A custom smart contract programming language with the
ability of accessing code analysis at run-time can prevent certain exploits. 1
Introduction Smart contracts are programs which run in a trusted execution environment pro-
vided by a blockchain [2]. The code of smart contracts can dictate how valuable
assets, associated with a smart contract, are handled. A flaw in the code can lead
to the loss or theft of the handled assets [10]. Developing bug-free software is
challenging even for skilled professionals [7]. Programming languages and tools
like formal verification or automated tests can support that process. Before a
smart contract is trusted with assets, such as cryptocurrency coins or a owner-
ship certificate, it must be ensured that the code implements the expected and
specified behavior. A end-user can ensure this by analyzing the code, by trust-
ing the developer to have implemented the specified behavior, or by trusting an
auditor to verify that the code implements the specified behavior. Analyzing the
code is not an option for most end-users, because of the complexity of the task
as well as the required time. This paper proposes SmartDEMAP a smart contract deployment and man-
agement platform which reduces the trust required into smart contract devel-
opers and auditors by imposing restrictions on the smart contracts that can be
deployed on it. The restrictions are enforced by executing formal verifiers [1,4],
compilers, automated bug-finders [8], or other development and code quality
tools on smart contracts. Such restrictions could consist of a formal proof of some
specified properties, enforcing a programming language, or requiring a negative
result from an automatic bug finder. SmartDEMAP can reduce the trust needed
into third parties, without requiring expertise in software auditing. 160 M. Knecht and B. Stiller SmartDEMAP allows to run development and code quality tools in a trusted
way to do deploy-time and run-time checks to increase smart contract quality
and robustness. To accelerate the integration into the development process, we
propose to develop a custom smart contract programming language that is aware
of the existence of SmartDEMAP. Such a language can generate code that facil-
itates the provided functionality and gives a developer easy access to it. Existing
languages can integrate SmartDEMAP by providing libraries to interact with it. 3
Related Work There are two categories of work related to SmartDEMAP. On one hand there are
smart contract specific development and code quality tools including program-
ming languages. On the other hand there is research on how resource intensive
computations can be executed in a trusted way despite the resource limitations
of smart contract enabled blockchains [2,11]. For the tools it is important that they work in a reliable way and can not be
fooled by a fine tuned input. If a compiler guarantees a certain semantic which
do not hold in the generated byte code, then a trusted execution of the compiler
will not help either. An earlier Solidity version had such a problem [9]. Formal SmartDEMAP: A Smart Contract Deployment and Management Platform 161 verification [1,4] and automatic bug-finding [8] are other relevant research topics
for SmartDEMAP. Research into these topics is relatively new and the devel-
oped tools are not in wide use and geared more towards trained professionals. SmartDEMAP could change that by allowing users not trained in these tools to
still benefit from their results. Theoretical results already exist concerning the execution of complex com-
putations in a way, such that the results can be trusted [3,5]. Their currently
is a project developing a concrete implementation [6] based on the theoretical
foundations from [3,5], and will allow smart contracts to trigger a trusted com-
putation and access the result. One part of the current research promises new tools that can be used to
improve the development process and reduce the exploitability of smart con-
tracts. Another part promises ways to run complex computations in a trusted
way, which can be utilized during the execution of smart contracts. There is
no research investigating if and how these two approaches could be combined. SmartDEMAP will close that research gap. 4
Smart Contract Exploits In recent years, different exploits have been found which are usable against
some of the existing smart contracts [8]. The most prominent example is the
“The DAO” theft [10]. Contrary to centralized software development, smart con-
tracts operate in an open environment where arbitrary adversaries can exist [8]
and thus attacks can originate from inside the same virtual machine. Addition-
ally, it is substantially harder to correct a bug because smart contract code is
unchangeable after it is deployed on a blockchain [2,11]. Most problems occur
when unknown code is executed, because it may have been deployed by an adver-
sary. Such vulnerabilities can lead to loss or theft of valuable assets and are often
hard or even impossible to fix. Most users of such smart contracts do not have
the expertise and time to ensure that it is safe to trust the smart contract with
their assets. 6
Improved Smart Contract Programming Language SmartDEMAP benefits from a custom programming language, which is aware
of it and uses its features during compilation or at run-time. Such a custom pro-
gramming language can incorporate SmartDEMAP to give additional guarantees
by generating the respective run-time checks based on the platforms capabilities. Further it provides a simple way for developers to access the platforms services. A new programming language provides the opportunity to analyze existing lan-
guages as well as common exploits of smart contracts programmed in these
languages. A smart contract programming language covering this aspect could
prevent some exploits and common pitfalls by design. One currently preferred approach is a language based on a process calculi
as suggested in [4]. This does prevent by design some of the common exploits,
such as the reentrancy exploit that brought “The DAO” to its knees [10]. This
exploit is prevented because no state is shared between different processes and
unlike a function a process cannot be called again if it is still running, and thus,
no unexpected state change can occur. This work evaluates if such a language
efficiently can be compiled to existing smart contract virtual machines and which
exploits could be prevented on the language level. 5
Platform-Based Smart Contract Management The auditors and developers of smart contracts are often anonymous and their
trustworthiness is unknown. Some smart contracts include code which allows a
privileged entity to exchange parts of the code. This is done to make it possible
to replace code containing a flaw with a fixed version. On the other hand this
could be used to inject code that violates a specified behavior. SmartDEMAP determines a new mechanism, which allows only code to be
deployed that does not violate an associated behavior specification. The behavior
can be specified as a formal specification and on deployment needs a proof that
it conforms to the specification. Other approaches like defining a test suite and
only allow code to be deployed that passes the test suite will be investigated
in addition to the formal verification approach. This has only a benefit if the M. Knecht and B. Stiller 162 formal verification tool or test suite can be run in a trusted way. This reduces
the trust required in developers and auditors and replaces it with trust into the
tools and their input (formal specification or test suite). To achieve this, the approach to be designed will follow a blockchain-based
path, with a platform for management, analysis, and deployment of smart con-
tracts (SmartDEMAP). The platform will manage a set of tools and use them
to enforce that smart contracts deployed on it fulfill as set of specifiable crite-
ria. Such a tool set contains formal verifiers, compilers, automatic test suites,
and automatic bug-finders. These tools are often complex and SmartDEMAP
will provide a way to ensure that these tools fulfill their purpose. This indicates
that each SmartDEMAP instance needs a entity fulfilling this role. This may
be another instance or a known third party (e.g. Microsoft, Amazon, Google) as
well as a consortium of people founded exclusively for that purpose. This system
reduces the trust needed in the code quality tools and replaces it with trust in
the tool verification entity. SmartDEMAP: A Smart Contract Deployment and Management Platform 163 martDEMAP: A Smart Contract Deployment and Management Platform The project is approached by developing a model of SmartDEMAP and the
custom programming language that describes their respective capabilities and
guarantees. Beside the model the platform as well as a compiler for the language
are implemented as a proof of the practical feasibility. The biggest risk involved in the project is that the currently developed tools
like formal verifiers as well as the trusted execution infrastructure will not be
available in time or do not satisfy the needs of SmartDEMAP. The developed
models are used to prove that certain exploits can be prevented fully or at least
to which degree if the described platform and the custom programming language
is used. The expected proofs are: 1. A proof that it is possible to decide if a unknown smart contract can be called
without the risk of becoming vulnerable to certain exploits. 1. A proof that it is possible to decide if a unknown smart contract can be called
without the risk of becoming vulnerable to certain exploits. 2. A proof that a developer can only deploy code that result in a behavior that
conforms to a formal specification. 2. A proof that a developer can only deploy code that result in a behavior that
conforms to a formal specification. 3. A proof that smart contracts programmed in the custom language are not
vulnerable against certain exploits. 3. A proof that smart contracts programmed in the custom language are not
vulnerable against certain exploits. The evaluation of which exploits are preventable this way is another expected
result from this project. Beside the theoretical results an implementation of
SmartDEMAP on the EVM [11] is expected. 7
Methodology This project is of high importance if smart contract should become safe to use
by non-experts. On one hand SmartDEMAP can give them a higher degree
of certainty, that it is safe to interact with a smart contract without the risk
of unexpected behavior. On the other hand SmartDEMAP and the custom pro-
gramming language help the developer to deliver smart contracts that are harder
to exploit. SmartDEMAP: A Smart Contract Deployment and Management Platform References 1. Bhargavan, K., Delignat-Lavaud, A., Fournet, C., Gollamudi, A., Gonthier, G.,
Kobeissi, N., Kulatova, N., Rastogi, A., Sibut-Pinote, T., Swamy, N., Zanella-
B´eguelin, S.: Formal verification of smart contracts: short paper. In: 2016 ACM
Workshop on Programming Languages and Analysis for Security, PLAS 2016,
Vienna, Austria, pp. 91–96 (2016) 1. Bhargavan, K., Delignat-Lavaud, A., Fournet, C., Gollamudi, A., Gonthier, G.,
Kobeissi, N., Kulatova, N., Rastogi, A., Sibut-Pinote, T., Swamy, N., Zanella-
B´eguelin, S.: Formal verification of smart contracts: short paper. In: 2016 ACM
Workshop on Programming Languages and Analysis for Security, PLAS 2016,
Vienna, Austria, pp. 91–96 (2016) 2. Buterin, V.: A next-generation smart contract and decentralized application plat-
form. Technical report (2014). Accessed 15 Nov 2016 2. Buterin, V.: A next-generation smart contract and decentralized application plat-
form. Technical report (2014). Accessed 15 Nov 2016 3. Canetti, R., Riva, B., Rothblum, G.N.: Practical delegation of computation using
multiple servers. In: 18th ACM Conference on Computer and Communications
Security, CCS 2011, Chicago, Illinois, USA, pp. 445–454 (2011) 3. Canetti, R., Riva, B., Rothblum, G.N.: Practical delegation of computation using
multiple servers. In: 18th ACM Conference on Computer and Communications
Security, CCS 2011, Chicago, Illinois, USA, pp. 445–454 (2011) 4. Edstroem, R., Pettersson, J.: Safer smart contracts through type-driven devel-
opment. Master’s thesis, Chalmers University of Technology and University of
Gothenburg (2016) 4. Edstroem, R., Pettersson, J.: Safer smart contracts through type-driven devel-
opment. Master’s thesis, Chalmers University of Technology and University of
Gothenburg (2016) 5. Jain, S., Saxena, P., Stephan, F., Teutsch, J.: How to verify computation with a
rational network, June 2016 5. Jain, S., Saxena, P., Stephan, F., Teutsch, J.: How to verify computation with a
rational network, June 2016 6. Teutsch, J., Reitwiessner, C.: A scalable verification solution for blockchains (2017). http://people.cs.uchicago.edu/%7Eteutsch/papers/truebit.pdf 7. Anand, K., Rai, K., Madan, L.: Software crisis. Int. J. Innov. Res. Technol. 1
(2014) 8. Luu, L., Chu, D.-H., Olickel, H., Saxena, P., Hobor, A.: Making Smart Contracts
Smarter. Cryptology ePrint Archive, Report 2016/633 (2016) M. Knecht and B. Stiller 164 9. Reitwiessner, C.: Security alert solidity variables can be overwritten in storage
(2016). https://blog.ethereum.org/2016/11/09/analysis-storage-corruption-bug. Accessed 03 Dec 2016 10. Vessenes, P.: Deconstructing the DAO Attack: A Brief Code Tour (2016). http://vessenes.com/deconstructing-thedao-attack-a-brief-code-tour. Accessed 03
Dec 2016 11. Wood, G.: Ethereum: A Secure Decentralised Generalised Transaction Ledger
(2015). http://gavwood.com/paper.pdf. 9. Reitwiessner, C.: Security alert solidity variables can be overwritten in storage
(2016).
https://blog.ethereum.org/2016/11/09/analysis-storage-corruption-bug.
Accessed 03 Dec 2016 10. Vessenes, P.: Deconstructing the DAO Attack: A Brief Code Tour (2016).
http://vessenes.com/deconstructing-thedao-attack-a-brief-code-tour. Accessed 03
Dec 2016 11. Wood, G.: Ethereum: A Secure Decentralised Generalised Transaction Ledger
(2015). http://gavwood.com/paper.pdf. Accessed 03 Dec 2016 References Accessed 03 Dec 2016 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 165–170, 2017.
DOI: 10.1007/978-3-319-60774-0 16 Keywords: Cloud modeling languages · Service orchestration · Juju Keywords: Cloud modeling languages · Service orchestration · Juju Optimizing the Integration of Agent-Based
Cloud Orchestrators and Higher-Level
Workloads Merlijn Sebrechts(B), Gregory Van Seghbroeck, and Filip De Turck IDLab, Department of Information Technology, Ghent University - imec,
Technologiepark-Zwijnaarde 15, 9052 Ghent, Belgium
merlijn.sebrechts@ugent.be Abstract. The flexibility of cloud computing has put significant strain
on operations teams. Manually installing and configuring applications in
the cloud simply isn’t an option anymore. Configuration management
automation solves the issue of getting a single application into a certain
state automatically and reliably. However, the issue of automatic depen-
dency management between multiple applications is still an “open, hard
problem” according to researchers at Google. Agent-based modeling and
orchestration tools like Juju solve the issue of getting from zero to a
working set of correctly clustered and connected frameworks. The short-
comings of these state-of-the-art tools are that they don’t provide effi-
cient ways to model and orchestrate workloads running on top of these
frameworks. This paper presents a number of ways to deploy and orches-
trate workloads with Juju, compares their performance and overhead,
and suggests how this overhead can be minimized. Keywords: Cloud modeling languages · Service orchestration · Juju 1
Introduction There is a big need to make IT operations easier. Take the field of data science for
example. There is an ever-growing set of tools and platforms that support data
scientists. The prevalence of open-source software in that field has shifted the
barrier of entry from licensing costs to operations costs. The tools are available
and free to use, but actually running them in production requires a team of
system administrators that have expert knowledge on both the tools themselves
and IT operations in general. Even industry-standard companies such as Google
state that the issue of automatic dependency management between multiple
services is still an “open, hard problem” [1]. The devops world has spawned a number of useful tools that help operations
teams. Configuration management systems help automate the task of installing,
configuring and managing applications. Automating these tasks reduces errors
and saves a lot of time when scaling an application. This process, called M. Sebrechts et al. 166 infrastructure as code, allows businesses to quickly react to changes in usage
of their application. These languages are less suited to lower the time to market
because each new application requires new management code. Moreover, these
tools don’t really abstract away the complexity of operations. This means that
operators using these tools now have to be experts in three fields: Configuration
management, IT operations and the applications they’re maintaining. Cloud modeling languages aim to reduce complexity and time to market
by providing an abstraction layer on top of IT operations. Instead of changing
the applications themselves, the operator changes a model that represents the
application. The orchestrator then translates actions on the model into actions
on the application. This is a great step forward to manage the complexity of IT
operations. The current generation of cloud modeling languages such as OASIS
TOSCA [2] also improve flexibility and re-usability of operations code by dividing
the operations code of an entire cloud application into a number of reusable
isolated pieces connected to each other using dependencies. Monolithic cloud orchestrators have a tendency to become very complex [3]. This results hard-to-maintain and hard-to-scale bottlenecks. Agent-based orches-
trators such as Juju [6] are the solution to this problem. All the dependency
resolution and operations logic is put into a series of agents that communicate
with each other over predefined interfaces. 1 https://jujucharms.com/u/tengu-team/limeds-bigdata/.
2 http://limeds.be/. 2
Modeling High-Level Workloads in Juju The authors’ previous work proposed the workflow component as a way to
model and manage high-level workloads with Cloud Modeling Languages [4]. Each workflow component is a Charm that contains both the workload itself
and a workflow agent that manages the workload. This approach provides a lot
of flexibility without adding any additional logic to the Juju orchestrator itself. The tricky part of this approach is that each workload requires at least one
agent, and this agent needs to run somewhere. Juju provides two ways to run
additional agents: co-locate the workflow agent and the framework agent with-
out any isolation and isolate the workflow agent from the framework agent by
running it inside an LXD container. Both methods aren’t ideal. It clearly shows that Juju is not built with such
use in mind. The issue with co-location is that Juju doesn’t allow two co-located
agents to run in parallel. This is to avoid conflicts when two agents try to man-
age the same machine at the same time. This significantly slows down the agents
because each agents needs to wait for the other agents to finish executing. Iso-
lating the agents using LXD containers solves this issue but introduces a new
one: the overhead of the LXD container. In many cases the overhead of the LXD
container is larger than the resources used by the actual workload. 1
Introduction The only responsibility of the orches-
trator is to install the agents and set up communication channels between them
[5]. The actual dependency resolution happens in the agents. This has the added
benefit that the implementation of the agent is hidden. This makes it possible
for two agents that manage services using two different configuration manage-
ment tools to communicate with each other, exchange information, and feed that
information into the config management tools. The combination of agent-based cloud orchestrators and cloud modeling lan-
guages makes IT operations a lot easier but there is still a lot of work to be done. All the aforementioned tools have a strong focus on the operations of an appli-
cation as a combination of services. What is left out are the actual workloads
running on top of these services. It’s great that orchestrators allow an operator
to setup a MySQL database, but what about the tables in the database? It’s
easy to model and orchestrate an Apache Hadoop cluster, but what about the
jobs running on top of that Hadoop cluster? This isn’t only about creating the
table and submitting the job. The MySQL table will be used by some software
or algorithm and the Hadoop job will get data from somewhere and put the
extracted information somewhere else. Configuring all these workloads by hand
isn’t a viable option due to the same reason that running the operations of an
entire application isn’t a viable option: it’s error-prone, it slows innovation down
to a crawl, and requires a very competent team with highly specialized skills. Since agent-based cloud orchestrators solve these challenges for the operation
of services and applications, they form a great start to explore solutions for the
operation of high-level workloads. Optimizing the Integration of Agent-based Cloud Orchestrators 167 4
Evaluation The deploy-time overhead is measured as the time it takes from the model to
scale. The tests start with a running LimeDS Big Data cluster with two units. This cluster is then scaled to n units, and the time until the scaling action is
complete is measured and compared. The results in Fig. 1 show that the deploy-time overhead is initially greater
for the isolated setup than for the co-located setup. This is due to the overhead
of spinning up an LXD container for the new agent. However, when more units
are requested, the isolated setup scales faster than the co-located setup due to
the sequential nature the co-located setup. Only one co-located agent is allowed
to execute actions at any given moment. For the runtime overhead of the agents, the memory and disk usage of
the agent are recorded as shown in Fig. 2. Here the disadvantage of the isolated
setup is clearly visible, it has a much bigger runtime overhead. The +200 MB of
RAM usage per agent is especially worrisome since the LimeDS container itself
uses about 300 MB of RAM. Fig. 1. The deploy-time overhead of
the agents; LXD vs co-located. Fig. 2. The runtime overhead of the
agents; LXD vs co-located. Fig. 2. The runtime overhead of the
agents; LXD vs co-located. Fig. 1. The deploy-time overhead of
the agents; LXD vs co-located. Fig. 1. The deploy-time overhead of
the agents; LXD vs co-located. Fig. 2. The runtime overhead of the
agents; LXD vs co-located. 3
LimeDS Big Data Model This paper evaluates both methods for running additional agents in order to
get a better grasp on what the actual overhead is and how it compares to the
resources used by the workload. The evaluation is done using the LimeDS Big
Data model1. This model and its components is further explained in this section. LimeDS is a modular platform to create and run data-driven services2. The
LimeDS Big Data model is perfect for validating the flexibility of modeling work-
loads for a number of reasons. First of all, LimeDS is both a workload and
a platform. The LimeDS Docker container is a workload running on top of
the Docker host, but it is also a host to services and modules running on top
of LimeDS. It is important to support such flexibility. Having LimeDS and the
Docker runtime be two different Charms also has the advantage that you can
swap out the single Docker host and plug in for example a Kubernetes clus-
ter. Secondly, workloads running on LimeDS need to connect to other
services, for example external datastores or load balancers. These connections
require a workload agent communicating with other services to exchange the cor-
rect information and to resolve possible dependencies such as the workload hav-
ing to wait for MongoDB to start. Lastly, LimeDS needs to run in a scaled-
out setup to handle Big Data workloads. The agents make this incredibly easy. An operator specifies how many instances of LimeDS are needed. The Orches-
trator installs an agent for each LimeDS instance, and the agents communicate M. Sebrechts et al. 168 with the Docker host agent to deploy LimeDS correctly. Since each LimeDS agent
implements the http interface, the agents don’t need any additional clustering
logic. Each agent connects to the agent managing the HAProxy load balancer,
and that agent configures the proxy correctly to loadbalance requests over the
LimeDS cluster. 5
Conclusion and the Road Forward Neither of the solutions has satisfactory performance. The co-located setup com-
promises heavily on deploy-time overhead and both setups compromise on run-
time overhead. There are a few advantages to these solutions. Having the abil-
ity to write arbitrary logic in the agent enables complex dependency resolution
without adding complexity to the orchestrator itself. Containers successfully stop Optimizing the Integration of Agent-based Cloud Orchestrators 169 the workload agents from accessing or changing the machine where the frame-
work is running. This forces agents to communicate using the relationships. This
enables other frameworks to implement the same relationship, making the solu-
tion pluggable. The ability to model higher-level workloads as a combination of
components related to each other gives operators a clear view of what is actu-
ally running, and allows the workloads themselves to be pluggable. The challenge
will be to find a solution that addresses the performance issues described here
without compromising on the stated advantages. Future research will explore
the road forward in a few directions. Agentless Agents: The advantage of the agents is that they allow run-
ning arbitrary dependency handling code, thus keeping the orchestrator simple. A possible solution might be to have a way for giving snippets of dependency
handling code to other agents instead of spinning up new agents. Slim Agents: Instead of reducing the amount of agents, another path forward
is to investigate if the overhead of the agent itself can be reduced. This approach
requires thorough investigation into where the overhead comes from. There is
also potential to use more lightweight process containers such as Docker instead
of the full-blown operating system containers that LXD provides. Parallel co-located Agents: Enabling co-located containers to run in parallel
is a possible solution to the deploy-time overhead of co-located agents. This
would need each agent to specify what kind of operations the agent will execute. The orchestrator can then use that information to determine whether or not two
agents are allowed to run at the same time. Acknowledgment. Part of this work has been funded by the iFest project, cofunded
by imec and VLAIO. 5. Sebrechts, M., Vanhove, T., Van Seghbroeck, G., Wauters, T., Volckaert, B., De
Turck, F.: Distributed service orchestration: eventually consistent cloud operation
and integration. In: Proceedings of the 2016 IEEE International Conference on
Mobile Services (MS 2016). IEEE (2016) (
)
(
)
6. Tsakalozos, K., Johns, C., Monroe, K., VanderGiessen, P., Mcleod, A., Rosales, A.:
Open big data infrastructures to everyone. In: 2016 IEEE International Conference
on Big Data (Big Data), pp. 2127–2129, December 2016 References 1. Burns, B., Grant, B., Oppenheimer, D., Brewer, E., Wilkes, J.: Borg, Omega, and
Kubernetes. Queue 14(1), 70–93. http://dl.acm.org/citation.cfm?doid=2898442. 2898444 2. OASIS:
TOSCA
Simple
Profile
in
YAML
Version
1.0,
August
2016. https://docs.oasis-open.org/tosca/TOSCA-Simple-Profile-YAML/v1.0/
TOSCA-Simple-Profile-YAML-v1.0.html 2. OASIS:
TOSCA
Simple
Profile
in
YAML
Version
1.0,
August
2016. https://docs.oasis-open.org/tosca/TOSCA-Simple-Profile-YAML/v1.0/
TOSCA-Simple-Profile-YAML-v1.0.html 3. Schwarzkopf, M., Konwinski, A., Abd-El-Malek, M., Wilkes, J.: Omega: flexible,
scalable schedulers for large compute clusters. In: Proceedings of the 8th ACM
European Conference on Computer Systems, EuroSys 2013, NY, USA, pp. 351–364
(2013). http://doi.acm.org/10.1145/2465351.2465386 3. Schwarzkopf, M., Konwinski, A., Abd-El-Malek, M., Wilkes, J.: Omega: flexible,
scalable schedulers for large compute clusters. In: Proceedings of the 8th ACM
European Conference on Computer Systems, EuroSys 2013, NY, USA, pp. 351–364
(2013). http://doi.acm.org/10.1145/2465351.2465386 (
)
//
/
/
4. Sebrechts, M., Borny, S., Vanhove, T., Seghbroeck, G.V., Wauters, T., Volckaert, B.,
Turck, F.D.: Model-driven deployment and management of workflows on analytics
frameworks. In: 2016 IEEE International Conference on Big Data (Big Data), pp. 2819–2826, December 2016 4. Sebrechts, M., Borny, S., Vanhove, T., Seghbroeck, G.V., Wauters, T., Volckaert, B.,
Turck, F.D.: Model-driven deployment and management of workflows on analytics
frameworks. In: 2016 IEEE International Conference on Big Data (Big Data), pp. 2819–2826, December 2016 M. Sebrechts et al. 170 5. Sebrechts, M., Vanhove, T., Van Seghbroeck, G., Wauters, T., Volckaert, B., De
Turck, F.: Distributed service orchestration: eventually consistent cloud operation
and integration. In: Proceedings of the 2016 IEEE International Conference on
Mobile Services (MS 2016). IEEE (2016) 6. Tsakalozos, K., Johns, C., Monroe, K., VanderGiessen, P., Mcleod, A., Rosales, A.:
Open big data infrastructures to everyone. In: 2016 IEEE International Conference
on Big Data (Big Data), pp. 2127–2129, December 2016 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 173–178, 2017.
DOI: 10.1007/978-3-319-60774-0 17 Situational Awareness: Detecting Critical
Dependencies and Devices in a Network Martin Laˇstoviˇcka(B) and Pavel ˇCeleda Institute of Computer Science and Faculty of Informatics,
Masaryk University, Brno, Czech Republic
lastovicka@ics.muni.cz, celeda@ics.muni.cz Institute of Computer Science and Faculty of Informatics,
Masaryk University, Brno, Czech Republic
lastovicka@ics.muni.cz, celeda@ics.muni.cz Abstract. Large-scale networks consisting of thousands of connected
devices are like a living organism, constantly changing and evolving. It
is very difficult for a human administrator to orient in such environment
and to react to emerging security threats. With such motivation, this
PhD proposal aims to find new methods for automatic identification of
devices, the services they provide, their dependencies and importance. The main focus of the proposal is to find novel approaches to building
cyber situational awareness in an unknown network for the purpose of
computer security incident response. Our research is at the initial phase
and will contribute to a PhD thesis in four years. Keywords: Situational awareness · Cybersecurity · Device importance
evaluation · Threat impact estimation · Graph theory · Network moni-
toring 1
Introduction The impacts of cyber threats became more serious with organisations increasing
dependency on computer infrastructure. To defend against such threats, system
administrators must build situational awareness which allows them to under-
stand and orient in the complex networks [6]. The aim of this PhD thesis is to
find new ways to automatically build situational awareness to help administra-
tors understand possible impacts of a cyber threat. Situational awareness means the knowledge and understanding of the current
situation. It is possible for a system administrator to know what is going on in a
small network, but with the growing number of connected devices, this becomes
more and more difficult. A basic solution is to manually create a list of all devices
in the network. But it is impossible to maintain such list throughout time and
keep it updated with the dynamic changes of the network. Moreover, the trend
of nowadays networks, containing mobile devices or IoT (Internet of Things),
and cloud environments goes directly against the idea of device list and makes it
useless in practice An automated approach is needed to deal with the constantly
changing environment [6]. 174 M. Laˇstoviˇcka and P. ˇCeleda The current approach for device and service identification focus on very spe-
cific networks, e.g., industrial control systems, or selected subset of services [1,2]
which reduces their value in modern networks described above. Manual evalua-
tion by security expert is still prevalent in the field of dependency detection and
importance estimation. These risk assessment methods are not automated [5] or
need active cooperation of the devices [8]. In our work, we intend to find new methods of building situational awareness
based on data from network monitoring that will not depend on a specific type of
network. We will define a computer network model containing information about
devices and services, their dependencies and importance for the organisation. The importance of a device can then be expressed as how the device outage or
compromise would impact other devices and goals of the host organisation. The
nature of continuous information gathering from the network also overcomes
the ever-changing nature of large networks and allows us to evaluate the data
throughout time. 2
Research Questions This research aims to discover new ways of threat impact estimation with respect
to current situation, devices and services. To achieve this goal we attempt to
answer following research questions: 3
Proposed Approach Our first step towards the building of situational awareness will be the defini-
tion of a network model. The natural representation of a computer network is
a graph, where each node stands for a device in the network. Edges between
nodes represent device communication, while another type of edge can represent
dependencies or the presence of a cyber threat. This model allows us to separate
the mostly static nature of what device is from its dynamic behaviour on the
network. 1. How can device and its services be identified in a complex network
using passive network monitoring?
Many devices are not willing (end-user devices) or not able to (IoT) provide
information about themselves in large networks. But every device commu-
nication over network could be analysed [3] and used to identify the type
of the device, its operating system and provided services. However, current
trends in modern networks, e.g., encrypted communication, port obfuscation,
high transfer rate, make such identification hard. We plan to investigate those
issues and propose methods to handle them. 1. How can device and its services be identified in a complex network
using passive network monitoring? Many devices are not willing (end-user devices) or not able to (IoT) provide
information about themselves in large networks. But every device commu-
nication over network could be analysed [3] and used to identify the type
of the device, its operating system and provided services. However, current
trends in modern networks, e.g., encrypted communication, port obfuscation,
high transfer rate, make such identification hard. We plan to investigate those
issues and propose methods to handle them. 2. How can device dependencies be detected in a network? To understand the situation in a network, it is not enough to know only what
a device is and what services it provides. It is important to know which devices
it depends on and how many devices depend on it. To answer this research
question, we will study relationships between devices in internal network and
propose new methods for their detection from network monitoring data. 3. How can device importance be estimated from the perspective of
reaction to cyber threats? The importance of a particular device for organisation mission differs accord-
ing to the provided services and the number of clients depending on the device. We plan to take these factors into account to build a model for importance
estimation and we will find new ways of automatic importance evaluation
based on traffic monitoring. 3. How can device importance be estimated from the perspective of
reaction to cyber threats? The importance of a particular device for organisation mission differs accord-
ing to the provided services and the number of clients depending on the device. We plan to take these factors into account to build a model for importance
estimation and we will find new ways of automatic importance evaluation
based on traffic monitoring. Detecting Critical Dependencies and Devices 175 3.2
Dependency Detection and Importance Estimation of a Devi The problem of asset criticality evaluation is known as vital for proper decision
making during cyber-attacks but is difficult to achieve [5]. Research in this area
is mainly focused on finding ways how a group of security experts can determine
criticality by following prepared guidelines just like in risk assessment. But this
approach is very time-consuming and cannot be repeated very often which leads
to the data being outdated. On the contrary, automatic evaluation is able to run continuously and can
provide results when needed. We are aware that some important services or
dependencies can be discovered only during exceptional operations or back-up
servers become active only after failure of the main one. Automatic detections
can still provide good staring point for risk assessment and save resources. More-
over, automatic system can identify operations that the administrators do not
know about as presented in [9]. We propose three components to combine in
order to estimate the device importance: 1. Traffic Statistics – Analysis of ongoing traffic in the network can point out
the most used services in the terms of connected clients and data transfer
volume. We will link these volumetric statistics to identified services to give
them the dynamic context for importance evaluation, e.g., heavily loaded
web server will be set as more important than another one scarcely visited. Our research will focus on real-time statistics computations so that it will be
possible to dynamically adjust the evaluation as the network usage changes
in time. 1. Traffic Statistics – Analysis of ongoing traffic in the network can point out
the most used services in the terms of connected clients and data transfer
volume. We will link these volumetric statistics to identified services to give
them the dynamic context for importance evaluation, e.g., heavily loaded
web server will be set as more important than another one scarcely visited. Our research will focus on real-time statistics computations so that it will be
possible to dynamically adjust the evaluation as the network usage changes
in time. 2. Dependency detection – Based on the identification of device type and
traffic statistics, the basic dependency between client and server will be mod-
elled. Using graph centrality algorithms we can then estimate the servers
importance and the impact of its outage as the number of affected clients
weighted by their own criticality. More complex dependencies can be dis-
covered by clique detection. 3.1
Identification of Devices and Services The knowledge of what a device is and what services it provides is a funda-
mental part of understanding the network. The goal of this part is to research
methods of processing network traffic data to identify the type of the device
(server, workstation, mobile, IoT), its operating system (Windows, Linux) and
its services (web, mail, database). Easiest way to determine a device type and services is to simply ask it. To do
it in an organised way, many Service Discovery Protocols have been implemented
[7] and deployed. They build a directory of all devices and their services, as
an example, we can name well-known protocols such as BitTorrent or UPnP. However, this approach require active cooperation of the devices and hence we
will not focus on them. Another way is to use active scanning. Our plan is to focus
on passive methods only, yet we can use outputs of network scanning projects,
e.g., Shodan, Censys, as a verification or an enhancement of our methods. To achieve passive classification described above a sophisticated method must
be used. Simple methods using protocol and port numbers currently falls short
in classifying services with a dynamic port assignment or port obfuscation, e.g.,
hiding behind TCP port 80 [10]. To overcome these issues, more characteristics
need to be taken into consideration. The current trend of traffic encryption makes the analysis of its content
hard, but on the other hand, it opens new ways of host identification. A client
needs to send a lot of data to establish encrypted communication. For example,
supported ciphersuites can be used to identify communicating clients during TLS
(Transport Layer Security) handshake [4]. Similarly, we plan to investigate other
properties of encrypted communication to identify the client device. The most promising service identification method nowadays is the use of
machine learning algorithms to classify the network traffic. Current methods
perform well in a controlled environment where every application is known in
advance, but cannot efficiently handle unknown traffic. Zhang et al. [11] pre-
sented an iterative method to improve identification accuracy, yet this field is
still not fully explored. The two challenges we plan to address are the accuracy of
identification in real network and performance of such algorithms when process-
ing large amounts of data continuously coming from the monitored network. M. Laˇstoviˇcka and P. ˇCeleda 176 4
Conclusion In this research, we focus on building situational awareness from passive network
observation without the necessity of active device probing. From those data, we
intend to determine what a device is, what services it provides, what are its
dependencies and how important it is for the network. Our methods will evalu-
ate the situation continuously in order to follow changes in network and will be
designed to be autonomic to minimise the need for human administrator assis-
tance. Achieving our goals will help system administrators to better understand
the situation in their network and to perceive the possible impacts of cyber
threats. Acknowledgement. This
research
was
supported
by
the
Security
Research
Programme of the Czech Republic 2015–2020 (BV III/1 VS) granted by the Ministry
of the Interior of the Czech Republic under No. VI20172020070 Research of Tools for
Cyber Situation Awareness and Decision Support of CSIRT Teams in the Protection
of Critical Infrastructure. Martin Laˇstoviˇcka is Brno Ph.D. Talent Scholarship Holder – Funded by the Brno
City Municipality. 3.2
Dependency Detection and Importance Estimation of a Devi Dependencies forming a clique between servers
can indicate strong relationship and exploitation of one will affect the whole
group. The first steps towards automatic dependency detection using graph
algorithms were made in [9], but they rely on active probing (i.e., Nagios
system) to discover effects of service failure and backup detections, whereas
we plan to achieve the same with passive network monitoring. 3. Attacks Statistics – Network attack is a manifestation of a cyber threat. The understanding of attack targets and discovery of most attacked devices
should lead to raising the protection level of those devices. Our assumption is
that parts of critical infrastructure will be targeted by attackers more often
than user stations. Moreover, the type of the attack should differ and these
differences could help to identify the most important devices. However, such
assumption needs to be carefully verified before using in the criticality calcula-
tions. For example, attackers could target the most vulnerable device instead
of critical infrastructure. In that case, such observation should be used as an
advisory for the administrator rather than for criticality estimation. Detecting Critical Dependencies and Devices 177 References 1. Callado, A., Kamienski, C., Szab´o, G., Gero, B.P., Kelner, J., Fernandes, S., Sadok,
D.: A survey on internet traffic identification. IEEE Commun. Surv. Tutorials
11(3), 37–52 (2009) 2. Franke, U., Brynielsson, J.: Cyber situational awareness - a systematic review of
the literature. Comput. Secur. 46, 18–31 (2014) 3. Hofstede, R., ˇCeleda, P., Trammell, B., Drago, I., Sadre, R., Sperotto, A., Pras, A.:
Flow monitoring explained: from packet capture to data analysis with NetFlow and
IPFIX. IEEE Commun. Surv. Tutorials 16(4), 2037–2064 (2014, Fourthquarter) 4. Hus´ak, M., ˇCerm´ak, M., Jirs´ık, T., ˇCeleda, P.: HTTPS traffic analysis and client
identification using passive SSL/TLS fingerprinting. EURASIP J. Inf. Secur. 2016(6), 1–14 (2016) 5. Kim, A., Kang, M.H.: Determining asset criticality for cyber defense. Technical
report, Naval Research Lab, Washington DC (2011) 6. Kott, A., Wang, C., Erbacher, R.F.: Cyber Defense and Situational Awareness. Springer, Heidelberg (2014). ISBN: 978-3-319-11390-6 7. Meshkova, E., Riihij¨arvi, J., Petrova, M., M¨ah¨onen, P.: A survey on resource dis-
covery mechanisms, peer-to-peer and service discovery frameworks. Comput. Netw. 52(11), 2097–2128 (2008) 8. Weintraub, E., Cohen, Y.: Continuous monitoring system based on systems’s envi-
ronment. In: Proceedings of the Conference on Digital Forensics, Security and Law,
p. 151. Association of Digital Forensics, Security and Law (2015) 9. Zand, A., Houmansadr, A., Vigna, G., Kemmerer, R., Kruegel, C.: Know your
achilles’ heel: automatic detection of network critical services. In: Proceedings of
the 31st Annual Computer Security Applications Conference, pp. 41–50. ACM
(2015) M. Laˇstoviˇcka and P. ˇCeleda 178 10. Zander, S., Nguyen, T., Armitage, G.: Automated traffic classification and appli-
cation identification using machine learning. In: The IEEE Conference on Local
Computer Networks 30th Anniversary (LCN 2005), pp. 250–257. IEEE (2005) 11. Zhang, J., Chen, C., Xiang, Y., Zhou, W.: Robust network traffic identification with
unknown applications. In: Proceedings of the 8th ACM SIGSAC Symposium on
Information, Computer and Communications Security, pp. 405–414. ACM (2013) Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. 11. Zhang, J., Chen, C., Xiang, Y., Zhou, W.: Robust network traffic identification with
unknown applications. In: Proceedings of the 8th ACM SIGSAC Symposium on
Information, Computer and Communications Security, pp. 405–414. ACM (2013) 10. Zander, S., Nguyen, T., Armitage, G.: Automated traffic classification and appli-
cation identification using machine learning. In: The IEEE Conference on Local
Computer Networks 30th Anniversary (LCN 2005), pp. 250–257. IEEE (2005) References The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 179–184, 2017.
DOI: 10.1007/978-3-319-60774-0 18 Lucas Bondan1,2(B), Tim Wauters2, Bruno Volckaert2, Filip De Turck2,
and Lisandro Zambenedetti Granville1 1 Institute of Informatics (INF), Federal University of Rio Grando do Sul,
Porto Alegre, Brazil {lbondan,granville}@inf.ufrgs.br {
}
2 Department of Information Technology (INTEC), Ghent University,
Ghent, Belgium 2 Department of Information Technology (INTEC), Ghent University,
Ghent, Belgium {tim.wauters,bruno.volckaert,filip.deturck}@intec.ugent.be Abstract. Industry and academia have increased the deployment of
Network Functions Virtualization (NFV) on their environments, either
for reducing expenditures or taking advantage of NFV flexibility for ser-
vice provisioning. In NFV, Service Function Chainings (SFC) composed
of Virtualized Network Functions (VNF) are defined to deliver services
to different customers. Despite the advancements in SFC composition
for service provisioning, there is still a lack of proposals for ensuring the
integrity of NFV service delivery, i.e., detecting anomalies in SFC oper-
ation. Such anomalies could indicate a series of different threats, such
as DDoS attacks, information leakage, and unauthorized access. In this
PhD, we propose a framework composed of an SFC Integrity Module
(SIM) for the standard NFV architecture, providing the integration of
anomaly detection mechanisms to NFV orchestrators. We present recent
results of this PhD regarding the implementation of an entropy-based
anomaly detection mechanism using the SIM framework. The results
presented in this paper are based on the execution of the proposed mech-
anism using a realistic SFC data set. Keywords: Service function chaining · Network functions virtualiza-
tion · Anomaly detection A Framework for SFC Integrity in NFV
Environments Lucas Bondan1,2(B), Tim Wauters2, Bruno Volckaert2, Filip De Turck2,
and Lisandro Zambenedetti Granville1 2
SFC Integrity Framework The NFV MANO architecture does not consider security-related tasks to protect
functions and services. In this PhD research, we seek to guarantee the integrity of
SFC operation for service delivery. Our proposal is designed to operate in NFV
networks ruled by NFVOs according to the standard NFV MANO architecture. 1.1
Motivation In virtualized environments, vulnerabilities and exploits can lead to different SFC
threats, since virtualization elements of NFV environments are susceptible to
exploits. Examples of exploitable elements are container engines [4], hypervisors
[5], and virtual machines [6]. Therefore, solutions have been proposed to detect
anomalies in different NFV elements, such as VNFs [7], NFV services [8], and
SLA violations [9]. However, there is still a lack of proposals dealing with security
and integrity issues in the context of SFC [10]. In this PhD, we consider both
the lack of solutions for SFC integrity and the potential vulnerabilities of NFV
environments as research opportunities to be properly explored. To do so, we
first investigated and proposed a framework that allows the implementation of
anomaly detection techniques based on the NFV MANO information model. 1
Introduction Network Functions Virtualization (NFV) was proposed to deal with the virtu-
alization of network functions usually performed by dedicated hardware devices
(e.g., firewalls, session border controllers, load balancers) [1]. In NFV, Virtual
Network Functions (VNF) are connected to each other, composing Service Func-
tion Chainings (SFC) for service delivery. Any anomaly in SFC operation, such
as missing elements, misconfiguration, and redirection, could lead to the inter-
ruption of the service delivery and, in some cases, could indicate attacks to the L. Bondan et al. 180 network. For this reason, in this PhD, we propose an additional SFC Integrity
Module (SIM) to the NFV architecture [2]. SIM is a framework that allows the
implementation of different anomaly detection mechanisms and the integration
of such mechanisms into any NFV network under the control of NFV Orchestra-
tors (NFVO). In this PhD, our focus resides in: (i) the applicability of existent
and new anomaly detection mechanisms for SFC integrity in NFV environments,
(ii) how to integrate such mechanisms to the NFV Management and Orchestra-
tion (MANO) architecture [3], and (iii) the evaluation of anomaly detection
solutions in realistic NFV scenarios using the proposed SIM framework. 2.2
Methodology SIM was designed with specific elements for processing, analyzing, and filtering,
enabling the design and implementation of different anomaly detection mech-
anisms. In this paper, we advance our first investigation using entropy-based
anomaly detection [2] in two ways: (i) evaluating our solution using realistic
NFV data sets [11] and (ii) improving the entropy-based anomaly detection
mechanism to work with the current data set. These improvements enabled us
to analyze each customer individually, increasing the accuracy of the anomaly
detection mechanism. The data set was generated based on realistic information
regarding the number of network functions composing SFCs on lager scale enter-
prise networks (with around 100 VNFs) [11]: 2 to 7 VNFs per SFC, mostly 2 to 5
[12]. So the number of VNFs for a given customer follows a truncated power-low
distribution with exponent 2, minimum 2 and maximum 7. Following enterprise
reports, anomalies were injected in the data set with a likelihood of 60% [13]. We considered three anomaly types: (i) unregistered SFCs, (ii) missing SFCs,
and (iii) unauthorized changes in the SFC, such as additional or missing VNFs. 2.1
Proposed Approach The NFVO sends cataloged and monitored information to an Orchestrator
Abstraction Driver (OAD), depicted in Fig. 1 along with all SIM internal compo-
nents. The information is then processed and analyzed according to the anomaly
detection mechanisms implemented in the Detector component. If no anomalies
are detected, the results are stored in the Library for further access. Otherwise,
the results are filtered using the Filter module to specify the sources of such
anomalies. Once identified, SIM stores it in the Library and forwards a report
message to NFVO with the filtered results and suggestions from the Advisor
module for overcoming such anomalies, e.g., turn offunregistered VNFs. A Framework for SFC Integrity in NFV Environments 181 SIM
Operator’s
Data Center
ConfiguraƟon
InformaƟon Flow
Network
Operator
Specifier
Yes
Detector
Analyzer
Processor
Anomaly? Advisor
Historic
Library
Alerts
Cataloged Values
Filter
! No
Cataloged Values
Values
Catalogs
NFVO
Interval
OAD
Fig. 1. Detailed SIM architecture [2] – The SIM communicates directly with NFVOs,
using standard northbound APIs for requesting information regarding NFV elements
operation and also to forward the results of the anomaly detection analysis. Fig. 1. Detailed SIM architecture [2] – The SIM communicates directly with NFVOs,
using standard northbound APIs for requesting information regarding NFV elements
operation and also to forward the results of the anomaly detection analysis. 2.3
Results Obtained Figure 2 shows the entropy results of the anomaly detection mechanism consid-
ering 4 customers with different sets of SFCs. The detector creates a merged
list with cataloged and monitored information. As the number of elements with
low probability increases in the list, i.e., highly uncertain elements, the merged
entropy changes, indicating a disorder in the monitored elements. The merged
entropy varies according to the number and type of anomalies detected (repre-
sented by markers). In our experiments, anomalies of type (i) and (ii) decreased
the entropy value, since they involve adding or subtracting information, while
anomalies of type (iii) (changes in existing values) increased the entropy value. It may lead to situations where anomalies of type (i) and (ii) cancel the entropy 182 L. Bondan et al. Fig. 2. Entropy results per customer. When anomalies occur (represented by markers),
the entropy values varies, according to the amount of anomalies and their type. Fig. 2. Entropy results per customer. When anomalies occur (represented by markers),
the entropy values varies, according to the amount of anomalies and their type. variations caused by anomalies of type (iii) and vice-versa. Despite rare to occur,
this problem should be properly addressed to avoid false negatives. With the
two-level approach of SIM (detection and filtering) it is possible to avoid false
negatives with fine-grained filters comparing monitored and cataloged informa-
tion. After each analysis the entropy values go back to normal (cataloged). variations caused by anomalies of type (iii) and vice-versa. Despite rare to occur,
this problem should be properly addressed to avoid false negatives. With the
two-level approach of SIM (detection and filtering) it is possible to avoid false
negatives with fine-grained filters comparing monitored and cataloged informa-
tion. After each analysis the entropy values go back to normal (cataloged). 3
Conclusions and Future Work This PhD aims to propose efficient solutions for maintaining the integrity of ser-
vice delivery in NFV environments. As first step, we proposed a SIM framework
that allows the implementation of different anomaly detection mechanisms to
analyze the network operation. The SIM modular architecture has the ability to
operate with different NFVOs, requiring only to adapt one specific block. For
future research, we foresee the following topics as good directions to follow. Detection on Different Information Levels. SIM was designed to operate
at different levels of information. In this way, we foresee the possibility to analyze
information regarding real-time resource consumption by virtual machines (e.g.,
CPU, RAM, disk) and network information (e.g., SFC traffic flows, bandwidth). Evaluation of Different Detection Mechanisms and Network Scenarios. Different anomaly detection mechanisms could be more suitable for a given net-
work scenario, according to its characteristics. Analyzing the operation of different
mechanisms in different environments will lead to important insights. A Framework for SFC Integrity in NFV Environments 183 Deployment on Production Networks. Our results are based on realistic
data sets generated according to real-world observations. However, production
networks may present unpredicted behaviors, such as communication problems
between NVFOs and other network elements. In this way, analyzing SIM oper-
ation in production networks is another important step of this PhD. Acknowledgements. This research was performed partially within the FWO project
“Service-oriented management of a virtualised future internet”. References 1. Chiosi, M., et al.: Network Functions Virtualisation (NFV). White Paper 1, ETSI
NFV ISG (2012). https://portal.etsi.org/NFV/NFV White Paper.pdf 2. Bondan, L., Wauters, T., Volckaert, B., Turck, F.D., Granville, L.Z.: Anomaly
detection framework for SFC integrity in NFV environments. In: IEEE Conference
on Network Softwarization (NetSoft), (July 2017, to appear) (
) (
)
3. Quittek, J., et al.: Network Functions Virtualisation (NFV) - Management and
Orchestration. White paper, ETSI NFV ISG (2014) 4. Combe, T., Martin, A., Pietro, R.D.: To docker or not to docker: a security per-
spective. IEEE Cloud Comput. 3(5), 54–62 (2016) 5. Thongthua, A., Ngamsuriyaroj, S.: Assessment of hypervisor vulnerabilities. In:
International Conference on Cloud Computing Research and Innovations (ICC-
CRI), pp. 71–77, May 2016 6. Wang, Z., Yang, R., Fu, X., Du, X., Luo, B.: A shared memory based cross-VM side
channel attacks in IaaS cloud. In: IEEE Conference on Computer Communications
Workshops (INFOCOM WKSHPS), pp. 181–186, April 2016 7. Giotis, K., Androulidakis, G., Maglaris, B.S.: A scalable anomaly detection and
mitigation architecture for legacy networks via an openflow middlebox. Secur. Commun. Netw. 9, 1958–1970 (2015) ,
(
)
8. Xilouris, G.K., Kourtis, M.A., Gardikis, G., Koutras, I.: Statistical-based anomaly
detection for NFV services. In: IEEE Conference on Network Function Virtualiza-
tion and Software Defined Networks (NFV-SDN) (2016, to appear) (
) (
)
9. Sauvanaud, C., Lazri, K., Kaˆaniche, M., Kanoun, K.: Anomaly detection and root
cause localization in virtual network functions. In: IEEE International Symposium
on Software Reliability Engineering (ISSRE), pp. 196–206, October 2016 (
)
10. Briscoe, B., et al.: Network Functions Virtualisation (NFV) - NFV Security: Prob-
lem Statement. White paper, ETSI NFV ISG (2014) 11. Rankothge, W., Le, F., Russo, A., Lobo, J.: Data modelling for the evaluation of
virtualized network functions resource allocation algorithms. Computing Research
Repository (CoRR) abs/1702.00369 (2017). http://arxiv.org/abs/1702.00369 12. Sherry, J., Hasan, S., Scott, C., Krishnamurthy, A., Ratnasamy, S., Sekar, V.:
Making middleboxes someone else’s problem: network processing as a cloud service. In: ACM SIGCOMM Conference on Applications, Technologies, Architectures, and
Protocols for Computer Communication, pp. 13–24 (2012) 13. Anstee, D., Bowen, P., Chui, C., Sockrider, G.: Worldwide infrastructure secu-
rity report. Technical report, Arbor Networks (2017). https://www.arbornetworks. com/insight-into-the-global-threat-landscape L. Bondan et al. c
⃝The Author(s) 2017
D. Tuncer et al. (Eds.): AIMS 2017, LNCS 10356, pp. 185–190, 2017.
DOI: 10.1007/978-3-319-60774-0 19 References 184 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. Multi-domain DDoS Mitigation
Based on Blockchains Bruno Rodrigues(B), Thomas Bocek, and Burkhard Stiller Communication Systems Group (CSG), Department of Informatics (IfI),
University of Z¨urich (UZH), Z¨urich, Switzerland
{rodrigues,bocek,stiller}@ifi.uzh.ch Communication Systems Group (CSG), Department of Informatics (IfI),
University of Z¨urich (UZH), Z¨urich, Switzerland
{rodrigues,bocek,stiller}@ifi.uzh.ch Abstract. The exponential increase of the traffic volume makes Distrib-
uted Denial-of-Service (DDoS) attacks a top security threat to service
providers. Existing DDoS defense mechanisms lack resources and flexi-
bility to cope with attacks by themselves, and by utilizing other’s compa-
nies resources, the burden of the mitigation can be shared. Technologies
as blockchain and smart contracts allow distributing attack information
across multiple domains, while SDN (Software-Defined Networking) and
NFV (Network Function Virtualization) enables to scale defense capabil-
ities on demand for a single network domain. This proposal presents the
design of a novel architecture combining these elements and introducing
novel opportunities for flexible and efficient DDoS mitigation solutions
across multiple domains. Keywords: Distributed
Denial-of-Service (DDoS)
·
Security
·
Blockchain · Software-defined Networks (SDN) · Network management 2
Problem Description As DDoS attacks become progressively sophisticated and coordinated, the
defense from such attacks likewise needs distribution and coordination. To pre-
vent or reduce damages caused by these DDoS attacks, different detection and
mitigation methods are available. Typical implementation is based on dedicated ASIC-based appliances to ana-
lyze flow records exported from edge routers, and further filtering or load bal-
ancing traffic. Cloud-based solutions such as Cloudfare [3] and Akamai [1] can
take away the burden of detection and mitigation, serving as a proxy able to
load balance, reroute, or drop the traffic in case of DDoS attacks. Many centralized defense systems lack of hardware resources or software
capabilities to detect and mitigate attacks themselves. Traditional or cloud-based
defenses can become a communication bottleneck due to the need to download
and process all the traffic measurements at a single location. Thus, if an attack
is highly sophisticated and there is no countermeasure available, legitimate users
may be impaired until the attack stops. An alternative is sharing hardware and defense capabilities with other sys-
tems, an approach called cooperative DDoS mitigation. However, existing coop-
erative approaches involve the proposal of a particular distributed architecture
and protocols that usually require the modification of existing hardware and
software in its support. 1
Introduction and Motivation A Distributed Denial-of-Service (DDoS) attack is a large-scale, coordinated
attempt to make a target system’s resources unavailable. Although being a
known category of attack, it remains as one of the major causes of concern for
service providers. The increasing number of unsecured connected devices (sta-
tionary and portable) and their growing processing capacity, allow attackers to
take control of a vast amount of unsecured devices that ranges from connected
cameras to smart fridges to generate malicious attacks. Major causes of concern for service providers is that not only the volume
of traffic of DDoS attacks is growing, but also their complexity. Botnets taking
advantage of unsecured IoT (Internet of Things) devices are the primary cause
of these large-scale attacks. The Mirai botnet [4], for example, exploits default
and weak security credentials to spread itself for other devices. In an attack launched on Krebs Security [1] website in September 2016, Mirai
peaked 623 Gbps in volume of traffic. Akamai, the service hosting the website,
had to shut down the site because defending it during three days became too
costly. It was reported that so many devices were used that the attacker did not
have to use any sophisticated strategy. B. Rodrigues et al. 186 4
Research Questions Many research challenges are found in the current scenario to improve current
DDoS defense mechanisms not only in a single domain, but in a cross-domain
perspective. Expected contributions of this work are categorized herein into three
major stages of the DDoS protection: (1) analysis and detection, (2) collabora-
tion across multiple domains, and (3) scalability of the proposed solution. There-
fore, contributions of this work are expected to answer the following research
questions: RQ1: How to efficiently identify traffic types avoiding that, in presence of
attacks, legitimate users may be hampered by the traffic of attackers? This
proposal involves the identification of techniques based on machine learning
to promote the signaling of attacks. RQ1: How to efficiently identify traffic types avoiding that, in presence of
attacks, legitimate users may be hampered by the traffic of attackers? This
proposal involves the identification of techniques based on machine learning
to promote the signaling of attacks. RQ2: How to simplify existing cooperative DDoS architectures and protocols so
minimal hardware and software modifications are necessary to advertise DDoS
information across multiple domains? In addition, it is necessary to investigate
an incentive scheme to balance the relationship between cooperative entities,
preventing a domain from abusing the cooperative scheme. RQ2: How to simplify existing cooperative DDoS architectures and protocols so
minimal hardware and software modifications are necessary to advertise DDoS
information across multiple domains? In addition, it is necessary to investigate
an incentive scheme to balance the relationship between cooperative entities,
preventing a domain from abusing the cooperative scheme. RQ3: How does the solution scale to report a number of addresses given the
scale of devices involved in large-scale attacks? RQ3: How does the solution scale to report a number of addresses given the
scale of devices involved in large-scale attacks? Multi-domain DDoS Mitigation Based on Blockchains 187 Multi-domain DDoS Mitigation Based on Blockchains 3
State-of-the-Art Although there are several related works, concepts and technologies guiding
the development of the proposal, for brevity in this section we highlight only
three main related works. Internet Engineering Task Force (IETF) is proposing
a protocol [5] named DOTS (DDoS Open Threat Signaling) covering both intra-
organization and inter organization communications. DOTS requires servers and
clients organized in both centralized and distributed architectures to advertise
black or whitelisted addresses. However, DOTS presents a complex architectural
design, which hinders your deployment without the complete standardization
of the DOTS protocol. A different approach is [7], proposing a collaborative
framework that allows the customers to request DDoS mitigation from ASes. However, the solution requires an SDN controller at customer side interfaced
with the service provider, which can change the label of the anomalous traffic
and redirect them to security middle-boxes. A similar approach is seen in [6]. The authors propose a cooperation between domains that implements VNFs to
alleviate DDoS attacks by redirecting and reshaping excessive traffic to other col-
laborating domains for filtering. However, the proposal still requires the support
of a gossip-based protocol by the network infrastructure to exchange information
about attacks. 5
Approach and Next Steps A novel approach is presented herein to mitigate DDoS attacks across mul-
tiple domains. Recent advances on networking technology, such as Software-
defined Networking (SDN) and Network Function Virtualization (NFV) are
gaining attention towards the establishment of software-defined infrastructures. Blockchain and Smart Contracts may be used to advertise information across
multiple domains, reducing the complexity of distributed protocols and architec-
tures for gossiping DDoS attacks information. Figure 1 illustrates the proposed
architecture of the system. – Software-Defined Networks (SDN): enable the development of customiz-
able security policies and services managed in a dynamic, software-based fash-
ion. Among the available SDN controllers, Ryu is an open-source controller
providing an well defined API for interacting and managing applications. – Software-Defined Networks (SDN): enable the development of customiz-
able security policies and services managed in a dynamic, software-based fash-
ion. Among the available SDN controllers, Ryu is an open-source controller
providing an well defined API for interacting and managing applications. – Network Function Virtualization (NFV): enforce the security policies
of the centralized control through virtualized functions provisioned in generic
hardware. The VNF-BC is the virtual appliance deployed both on the net-
work domain and customers, that may interface with network management
systems, and optionally import flow-records of widely used network monitor-
ing tools, as sFlow or NetFlow. – Blockchain: Ethereum-based blockchain, which is public, decentralized and
provide a trusted consensus in which data of DDoS attacks can be advertised
and accessed between the cooperative domains. In an Ethereum blockchain, – Blockchain: Ethereum-based blockchain, which is public, decentralized and
provide a trusted consensus in which data of DDoS attacks can be advertised
and accessed between the cooperative domains. In an Ethereum blockchain, B. Rodrigues et al. 188 Fig. 1. Proposed architecture Fig. 1. Proposed architecture VNF-BC appliances listening to the blockchain may see new addresses
reported within a new 14 s, which is the time a block is mined. – Smart Contracts: a Solidity-based contract implementing the logic of the
collaborative approach, advertising of white or blacklisted IP addresses of
certified customers, as well as information on the reporting entity and attack
characteristics. – Smart Contracts: a Solidity-based contract implementing the logic of the
collaborative approach, advertising of white or blacklisted IP addresses of
certified customers, as well as information on the reporting entity and attack
characteristics. 6
Summary An architecture for multi-domain DDoS Mitigation based on Blockchains, SDN
and VNFs was presented. Although designed based on SDN, a VNF appliance
to read/write in the blockchain could be integrated with different networking-
systems that exports flow records with sFlow or NetFlow. Expected contributions
are not limited to the collaborative perspective of the DDoS defense, but also
on the detection and mitigation of these attacks in a single domain based on key
technologies such as SDN and NFV. 5
Approach and Next Steps SDNs optimize the management of flows in response to attacks by enabling
the deployment of sophisticated traffic analysis based on global network aware-
ness given by a centralized controller. Aligned, SDN and NFV offer flexible
and programmable network infrastructures toward generic network hardware
deployed on open software, in which functions of the centralized control can be
performed through virtualized network functions (VNF) and capabilities from
NFV. Security policies and thresholds may be defined based on historical records
directly obtained from southbound protocols such as OpenFlow or SNMP, or
exported from monitoring tools as sFlow or NetFlow. In response to attacks,
the SDN controller may dynamically provision virtual functions for firewalling,
packet inspection (e.g., Snort), or black-holing malicious traffic. Blockchain and Smart Contracts can be used to advertise DDoS attacks infor-
mation across multiple domains [2]. This simplifies existing cooperative DDoS
mechanisms by using an existing distributed infrastructure to broadcast black or
whitelisted addresses without the need to build specialized registries or other dis-
tribution mechanisms/protocols. The Ethereum blockchain supports a Turing-
complete contract language [2], such as Solidity. Therefore, a node participating
in the Ethereum blockchain runs a Solidity smart contract by executing a script, Multi-domain DDoS Mitigation Based on Blockchains 189 which is used to store references to the advertised addresses. The contracts is
further processed checking if the entity reporting addresses is certified and its
result is stored in a block. However, entities issuing addresses need to have its identity certified. Sim-
ilar to IETF-DOTS, certificates can be used to ensure authenticity of entities. Therefore, a network domain may issue to its customers a an authentication
service to its customers through a registered VNF appliance able to report black
or whitelisted addresses to the blockchain. For example, an LDevID certificates
signed by the device owner may encode an owner assigned unique identifier and a
PKI matching a private key held within the VNF appliance. Inter-domain trust
can be established through any of the multi-PKI trust models in use today [8]. Then, information on the registered appliances will be hashed and referenced in
the smart contract. References 1. Akamai: How to Protect Against DDoS Attacks - Stop Denial of Service (2016). https://goo.gl/pfcWph. Accessed 10 Jan 2017 //
/
2. Bocek, T., Stiller, B.: Smart Contracts - Blockchains in the Wings, pp. 1–16. Springer, Heidelberg (2017). Tiergartenstr. 17, 69121 (
)
. CloudFare: Cloudflare advanced DDoS protection (2016). Accessed 10 Jan 2017 (
)
3. CloudFare: Cloudflare advanced DDoS protection (2016). Accessed 10 Jan 2017
4. Gamblin: Source code of the mirai botnet available on github, January 2016. https://
goo.gl/CB5vx4. Accessed 14 Mar 2017 (
)
3. CloudFare: Cloudflare advanced DDoS protection (2016). Accessed 10 Jan 2017
4
Gamblin: Source code of the mirai botnet available on github January 2016 https:// 4. Gamblin: Source code of the mirai botnet available on github, January 2016. https://
goo.gl/CB5vx4. Accessed 14 Mar 2017 5. Nishizuka, K., Xia, L., Xia, J., Zhang, D., Fang, L., Gray, C.: Inter-organization
cooperative DDoS protection mechanism. Draft, December 2016. https://goo.gl/
szsalO 6. Rashidi, B., Fung, C.: Cofence: a collaborative DDoS defence using network function
virtualization. In: 12th International Conference on Network and Service Manage-
ment (CNSM 2016), October 2016 7. Sahay, R., Blanc, G., Zhang, Z., Debar, H.: Towards autonomic DDoS mitigation
using software defined networking. In: SENT 2015: NDSS Workshop on Security of
Emerging Networking Technologies. Internet Society (2015) 8. Shimaoka, M., Hastings, N., Nielsen, R.: Memorandum for multi-domain public key
infrastructure interoperability (2008) B. Rodrigues et al. 190 Open Access This chapter is licensed under the terms of the Creative Commons
Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/),
which permits use, sharing, adaptation, distribution and reproduction in any medium
or format, as long as you give appropriate credit to the original author(s) and the
source, provide a link to the Creative Commons license and indicate if changes were
made. The images or other third party material in this chapter are included in the chapter’s
Creative Commons license, unless indicated otherwise in a credit line to the material. If
material is not included in the chapter’s Creative Commons license and your intended
use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. Author Index Badonnel, Rémi
47
Barcellos, Marinho P. 62
Barshan, Maryam
79
Bartoš, Václav
125
Bocek, Thomas
16, 185
Bondan, Lucas
179
Botero, Juan F. 62
Burská, Karolína
149
Carle, Georg
30
Čejka, Tomáš
3, 125, 153
Čeleda, Pavel
173
Compastié, Maxime
47
de Boer, Pieter-Tjerk
137
de O. Schmidt, Ricardo
137
De Turck, Filip
79, 165, 179
Dorfhuber, Marko
30
Festor, Olivier
47
Gaspary, Luciano P. 62
Gil-Herrera, Juliver
62
Granville, Lisandro Zambenedetti
179
Haugerud, Hårek
143
Hausheer, David
16
He, Ruan
47
Hendriks, Luuk
137
Herold, Nadine
30
Isolani, Pedro H. 62
Jansky, Tomáš
125
Kassi-Lahlou, Mohamed
47
Kergl, Dennis
108
Knecht, Markus
159
Kristiana, Lisa
94
Lareida, Andri
16
Laštovička, Martin
173
Latré, Steven
62
Liebald, Stefan
30
Moens, Hendrik
79
Müller, Robert
131
Neves, Miguel C. 62
Ocampo, Andrés F. 62
Ošlejšek, Radek
149
Pras, Aiko
137
Rafati, Sina
16
Rodosek, Gabi Dreo
108
Rodrigues, Bruno
16, 185
Roedler, Robert
108
Rudolf, Christoph
30
Schmitt, Corinna
94, 131
Sebrechts, Merlijn
165
Stiller, Burkhard
16, 94, 159, 185
Švepeš, Marek
3
Van Seghbroeck, Gregory
165
Velan, Petr
137
Volckaert, Bruno
79, 179
Wachs, Matthias
30
Waldvogel, Marcel
131
Wauters, Tim
179
Xue, Noha
143
Yazidi, Anis
143
Žádník, Martin
153
Zambenedetti, Lisandro
62 Lareida, Andri
16
Laštovička, Martin
173
Latré, Steven
62
Liebald, Stefan
30
Moens, Hendrik
79
Müller, Robert
131
Neves, Miguel C. 62
Ocampo, Andrés F. 62
Ošlejšek, Radek
149
Pras, Aiko
137
Rafati, Sina
16
Rodosek, Gabi Dreo
108
Rodrigues, Bruno
16, 185
Roedler, Robert
108
Rudolf, Christoph
30
Schmitt, Corinna
94, 131
Sebrechts, Merlijn
165
Stiller, Burkhard
16, 94, 159, 185
Švepeš, Marek
3
Van Seghbroeck, Gregory
165
Velan, Petr
137
Volckaert, Bruno
79, 179
Wachs, Matthias
30
Waldvogel, Marcel
131
Wauters, Tim
179
Xue, Noha
143
Yazidi, Anis
143
Žádník, Martin
153
Zambenedetti, Lisandro
62 Badonnel, Rémi
47
Barcellos, Marinho P. 62
Barshan, Maryam
79
Bartoš, Václav
125
Bocek, Thomas
16, 185
Bondan, Lucas
179
Botero, Juan F. 62
Burská, Karolína
149 Badonnel, Rémi
47
Barcellos, Marinho P. 62
Barshan, Maryam
79
Bartoš, Václav
125
Bocek, Thomas
16, 185
Bondan, Lucas
179
Botero, Juan F. 62
Burská, Karolína
149 Badonnel, Rémi
47
Barcellos, Marinho P. 62
Barshan, Maryam
79
Bartoš, Václav
125
Bocek, Thomas
16, 185
Bondan, Lucas
179
Botero, Juan F. Author Index 62
Burská, Karolína
149
Carle, Georg
30
Čejka, Tomáš
3, 125, 153
Čeleda, Pavel
173
Compastié, Maxime
47
de Boer, Pieter-Tjerk
137
de O. Schmidt, Ricardo
137
De Turck, Filip
79, 165, 179
Dorfhuber, Marko
30
Festor, Olivier
47
Gaspary, Luciano P. 62
Gil-Herrera, Juliver
62
Granville, Lisandro Zambenedetti
179
Haugerud, Hårek
143
Hausheer, David
16
He, Ruan
47
Hendriks, Luuk
137
Herold, Nadine
30
Isolani, Pedro H. 62
Jansky, Tomáš
125
Kassi-Lahlou, Mohamed
47
Kergl, Dennis
108
Knecht, Markus
159
Kristiana, Lisa
94 Žádník, Martin
153
Zambenedetti, Lisandro
62 192
Author Index © The Editor(s) (if applicable) and The Author(s) 2017. This book is an open access publication. Open Access This book is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appro-
priate credit to the original author(s) and the source, provide a link to the Creative Commons
license and indicate if changes were made. The images or other third party material in this book are included in the book’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the book’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly
from the copyright holder.
|
https://openalex.org/W1965840739
|
https://sefarad.revistas.csic.es/index.php/sefarad/article/download/520/618/1127
|
English
| null |
Some Pitfalls of Translation Greek
|
Sefarad
| 2,004
|
cc-by
| 8,748
|
Sefarad 64 (2004) págs. 341-362
©CSIC
ISSN 037-0894 Ponencia leída en el xii Congreso de la International Organization for Septua-
gint and Cognate Studies (lOSCS), Leiden, 30 de julio de 2004.
natalio@filol.csic.es
^ «La critica testuale non scopre 'il vero' se non in quanto caccia il 'falso'»: cf.
G. CONTINI, Breviario di Ecdotica (Torino 1990) p. 147. Or, in words of the famous
text critic A. E. HouSMAN: «Textual criticism is the science of discovering error in
texts and the art of removing it»: cf. A. E. HouSMAN, «The Application of Thought
to Textual Criticism», Proceedings of the Classical Association 18 (1922) pp. 67-84:
p. 68. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) ^ «La critica testuale non scopre 'il vero' se non in quanto caccia il 'falso'»: cf.
G. CONTINI, Breviario di Ecdotica (Torino 1990) p. 147. Or, in words of the famous
text critic A. E. HouSMAN: «Textual criticism is the science of discovering error in
texts and the art of removing it»: cf. A. E. HouSMAN, «The Application of Thought
to Textual Criticism», Proceedings of the Classical Association 18 (1922) pp. 67-84:
p. 68. Ponencia leída en el xii Congreso de la International Organization for Septua-
gint and Cognate Studies (lOSCS), Leiden, 30 de julio de 2004. Some Pitfalls of Translation Greek Natalio FERNÁNDEZ MARCOS
Instituto de Filología - CSIC, Madrid One of the primary functions of textual criticism is to detect the
genesis of errors. The «true» reading can only be discovered when
the «false» is unmasked ^ Similarly, if the Septuagint is to be used
critically, it is essential to start by unmasking the corruptions,
misreadings or mistranslations that lie within. The completion of a Greek-Hebrew Index of the Antiochene Text
in the Historical Books is an excellent opportunity to go through the
whole translation process and detect the most common mistakes, the
main difficulties met by the translators and the mechanisms emplo-
yed to overcome them. It is like looking at the reverse side of the
weave, giving an insight into the high degree of literal and formal
correspondence between the Greek translation and the Hebrew pa-
rent text in most of the historical books. At the same time it offers http://sefarad.revistas.csic.es 342 NATALIO FERNÁNDEZ MARCOS 5^/64:2 (2004) 5^/64:2 (2004) the opportunity to appreciate the limits of the formal equivalence
and, in some cases, to get a glimpse into the presumed Vorlage of
the translators. For the Greek, our edition of the Antiochene text has been used
as the basis of the analysis, and for the Hebrew the text of the Bi-
blia Hebraica Stuttgartensia (BHS) ^. We are aware that the Maso-
retic Text (MT) is not identical to the Hebrew Vorlage of the trans-
lators. We use it in the comparison for practical reasons, since it is
the only complete Hebrew text available for those books. Moreover,
it should be emphasized that the MT agrees to a large degree with
the Vorlage of the translators, as can be inferred from a look at the
Index in which the formal, extant equivalent, is of first concern. One may wonder why priority is not given to Qumran texts of Sa-
muel. The fact is, that apart from the fragmentary character (ca. 8%
of the text of Samuel) and occasional agreements with the Antio-
chene text, there is little evidence to define the textual affiliation of
the Qumran fragments •^. Much has been written recently on the use of formal or presumed
equivalents in an index or concordance. E. Tov and T. Muraoka
have diversely criticized the Hatch & Redpath's procedure, because
these scholars adhered very closely to the formal, almost mechani-
cal equivalence in their Concordance ^. ^ E. TOV, The Text-Critical Use of the Septuagint in Biblical Research (2"^^ ed.
Jerusalem 1997) pp. 91-100; T. MURAOKA, Hebrew/Aramaic Index to the Septuagint
Keyed to the Hatch-Redpath Concordance (Grand Rapids, MI 1998) p. 8. ^ «However, insufficient evidence was found to affirm any link between L and
4QSam", except for L's dependence upon LXX, which was in turn dependent upon
4QSam"»: cf. E. D. HERBERT, «4QSam" and its Relationship to the LXX: An
Exploration in Stemmatological Analysis», in IX Congress of the International
Organization for Septuagint and Cognate Studies. Cambridge, 1995, SCS 45, ed.
B. A. TAYLOR (Atlanta, GA 1997) pp. 37-55: p. 49. - N. FERNANDEZ MARCOS and J. R. BUSTO SAIZ, with the collaboration of M.^ V
SPOTTORNO DÍAZ-CARO and S. P. Co WE, El texto antioqueno de la Biblia Griega I-
III, TECC 50, 53, 60 (Madrid 1989-1996), and K. ELLIGER and W. RUDOLPH (eds.),
Biblia Hebraica Stuttgartensia (5th corrected ed., Stuttgart 1997). (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.es "^ Just as Abraham Tromm did in his Concordance published in Amsterdam 1718
(Abrahaini Trommii Concordantiae Graecae Versionis Vulgo Dictae LXX Interpre-
turn..., I-II [Amstelodami et Trajecti ad Rhenum MDCCXVIII]).
We hope that a mine of useful information has been added through the new
Greek words and new Hebrew equivalents preceded by an asterisk in the Index.
Likewise we suggest the presumed reading in a number of obvious equivalences
signaled with an obelus by Hatch and Redpath: auxiiobÔrjç, 'dry' in 1 Sam 23:14.15
does not translate niiD, but is a doublet of the unknown geographic name f^n;
Poppâç is a stereotype equivalent for \\ù^\ hence, it can be presumed that in 2 Chr
14: 9 the translator read m^ù)i instead of nriDii; in 1 Sam 13:7 it can be presumed
that the translator read oniVT as participle of nnv (oi ôiapaívovieç) instead of the
substantive 'the Hebrews' (on:;iv)).
^ R. SMEND, Griechisch-Syrisch-Hebrãischer
Index zur Weisheit des Jesus Sirach
(Berlin 1907) p. X. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) ^ R. SMEND, Griechisch-Syrisch-Hebrãischer
Index zur Weisheit des Jesus Sirach
(Berlin 1907) p. X. http://sefarad.revistas.csic.es Some Pitfalls of Translation Greek For the Index of the http://sefarad.revistas.csic.es 343 SOME PITFALLS OF TRANSLATION GREEK 5^/64:2 (2004) Antiochene text we have adopted a middle road. We adhere to the
formal equivalence when it looks plausible after a scrutiny of the
sentence in both languages. We believe that by using this procedure
we are ensuring that the reader can make the best use of the Index
without going down the path of the subjectivity of the presumed
equivalences. It is common knowledge that the abuse of presumed
equivalents, while they may be tempting, can lead to suggestions
that can be analysed in different ways by scholars, thus producing
different reconstructions. Nevertheless, in some cases and as an aid
to the reader, the presumed equivalent preceded by the abbreviation
leg (= legit) is suggested between brackets ^. In our search for the correct correspondence, the meticulous study
of both the Hebrew and the Greek texts leads us to some further
considerations. In some cases it is extremely difficult to decide
whether the extant Masoretic text reflects a new equivalent for the
extant Greek or whether, in fact, the Greek is being translated from
a different Vorlage. Indeed, our knowledge of the Hebrew and Ara-
maic as well as of the Alexandrian Greek is limited, and I subscribe
to the sound statement of R. Smend that «Eine Konkordanz muss in
der Gleichsetzung, soweit eine solche überhaupt durchführbar ist, so
mechanisch wie moglich verfahren und das Urteil der Zukunf iiber-
lassen» ^. This appreciation is also valid for an Index. Both extre- http://sefarad.revistas.csic.es 344 NATALIO FERNÁNDEZ MARCOS Sef 64:2 (2004) mes should be avoided: the inclusion of Hebrew words among the
new equivalents whose meaning is well outside the semantic field
of the Greek word ^, and the systematic exclusion of a new Hebrew
equivalent because it is not attested in other parts of the Septuagint. The good number of new equivalents marked with an asterisk in our
Index attests to the richness and variety of the translation manifes-
ted through several new plausible correspondences. These equiva-
lents are lacking in the Hatch & Redpath Concordance, be it because
the Antiochene Greek terms are only attested in the deuterocanoni-
cal or apocryphal books (some of them without Hebrew Vorlage), or
because they appear in the three Jewish translators whose Hebrew
equivalents are not recorded in this Concordance. '^ The translation may be idiomatic or metaphorical, or may correspond to a
different Vorlage, or may conceal a complex text critical problem. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.es Some Pitfalls of Translation Greek Moreover, Hatch
& Redpath follow the Greek text of the codices Vaticanas, Alexan-
drinus and Sinaiticus plus the Sixtine edition (1587), but they
ignore the Antiochene text which is very different from these ma-
nuscripts in the historical books. A careful use of the Index allows the user to draw certain conclu-
sions in relation to the different problems of textual criticism. The
stereotype correspondence between two terms in Hebrew and Greek
may lead to the restoring of a different reading from that of the MT
for the passage in question. Thus, in 2 Chr 33:7 aicov has been
introduced in the Greek translation for the MT Oìb^v. Taking into
account that 99% of the ocurrences of oblv have been regularly
translated by aicov, it can be deduced, in all confidence, that the
Greek translator of this passage also read obiv, as was the case in
the other ancient versions, and, consequently, it can be restored as a
genuine reading instead of the dubious and uncertain Dlb^v of MT. On the contrary, in 1 Chr 17:16 we come across a different text
critical panorama. MT reads «and what is my house, that you have
brought me thus far (obn-iv)?». The entire Greek tradition inter- http://sefarad.revistas.csic.es 345 SOME PITFALLS OF TRANSLATION GREEK Sef6A:2 (2004) prêts the last part of the sentence as scoç aicòvoç. But this reading
results from a phonetic confusion between the guturais v and n and,
consequently, cannot be invoked as a sound witness to change the
reading of the MT that makes sense. The frequent occurrence of the
expression ëcoç aicòvoç in the Greek Bible has contributed to
consolidate this reading in the Greek transmission. This is just a sample of the kind of textual criticism that can be
made with the aid of the Index, and which is valid for a high pro-
portion of common, abstract and concrete names where an almost
stereotype equivalence is recorded. Notwithstanding, the critical
judgement is more difficult to exercise in other names (for which
the translator liked the variatio or the metaphorical or stylistic
equivalence), and more especially in the verbs, where the array of
equivalents is highly diversified: for instance, sïôcoXov corresponds
to no less than ten Hebrew words, and Xajiipávsiv translates
eighteen different Hebrew forms ^. ^ As Muraoka observes, «It is obviously ill-advised to attempt to establish
mechanical patterns of correspondence between Greek tenses and those of Hebrew»:
T. MURAOKA, «Translation Techniques and Beyond», in Helsinki Perspectives on
the Translation Technique of the Septuagint, eds. R. SOLLAMO and S. SlPlLA
(Helsinki - Gottingen 2001) pp. 13-22: p. 20. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.es (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) ^ The English translation of the Hebrew, when no otherwise said, is taken from
B. M. METZGER and R. E. MURPHY (eds.), The New Oxford Annotated Bible with the
Apocrypha (New York 1989). '° Interestingly, the correct alternative reading Kai CKauaav aùiòv Kaûaiv
jicyáÀiiv has been preserved exclusively in the Alcalá polyglot Bible. Some Pitfalls of Translation Greek I shall now move on and try to identify the incorrect readings of
the Greek tradition, especially of the Antiochene text, in order to
highlight the genuine reading. I will deal with the translation pro-
cesss and the text transmission in a reverse order, going back from
a) the paléographie errors of transmission (inner-Greek corruptions),
through b) the different vocalization performed by the translator and
c) the interchange or confusion of similar letters in the Hebrew Vor-
lage, to d) some variant readings supported by a different Vorlage. For the last case, the Qumran fragments of Samuel open a window
towards actual readings of the Greek confirmed by an extant
Hebrew Vorlage different from MT. http://sefarad.revistas.csic.es 346 5^/64:2 (2004) 5^/64:2 (2004) NATALIO FERNÁNDEZ MARCOS http://sefarad.revistas.csic.es A. INNER-GREEK CORRUPTIONS A. INNER-GREEK CORRUPTIONS A. INNER-GREEK CORRUPTIONS Using the parent text as a control, some Greek corruptions can be
detected that have contaminated a part of or the entire manuscript
tradition. A few examples taken from the Antiochene text, shared
occasionally by the whole Greek tradition, will suffice to illustrate
this phenomenon: — Aï^,-y6ç is the regular equivalent for the Hebrew iv 'goat'. However, in 2 Chr 31:6 we come across a formal equivalent of aiyœv
for the Hebrew DWlp in the sequence «the tithe of cattle and sheep,
and the tithe of the dedicated things that had been consecrated to
the Lord their God» ^. In all probability the whole Greek tradition
has been corrupted from àyícov to aiyœv. However, following the
manuscript tradition we have restored aiyœv as did A. Rahlfs in his
manual edition. The reason why this new reading, so alien to the origi-
nal meaning of the Hebrew, succeeded in the text reception, is that
it makes sense also in the Greek chain of words joined to the cattle
and sheeps: Kai aòxoì iívsyKav STTiôéKaxa jiióaxcov Kaì TipoPáxcov
Kaì éíciSsKaxa aiyoov, Kaì fiyíaaav xoò Kupíco Bsœ auxœv. — The Antiochene reading of 2 Chr 16:14 must be characterised
as an inner-Greek corruption: Kai SKXauaav aòxcò KÀ^aCaiv
|Lisyá/lT|v for the Hebrew nbn> nonvy ib-iDn\yn ('and they made a
very great fire in his honor'). KaÍ£iv and Kaûaiç correspond better
to the meaning of the Hebrew root ^n'\y, while K?taisiv translates
regularly the root riDi. However, the paléographie confusion
8K?taüaav / SKauaav and KXaûaiv / KaCaiv, easy to detect in the
cursive Greek writing, provoked the new reading in the Antiochene
branch of manuscripts '°. The fact that the new reading makes sense
in the context of the verse lead to its consolidation within the Greek
tradition. In fact it is a doublet or alternative translation that Antio- http://sefarad.revistas.csic.es 347 SOME PITFALLS OF TRANSLATION GREEK Sef6A:2 (2004) chene added to the reading of the current Septuagint resulting in the
following sentence: Kai 87ioir|aav auioò ¿Kcpopàv \xzyòXr{V Kaì
8KXai)aav auxoD K^aCaiv jueyá^riv. In a text conceived for public
reading it is essential that it has meaning. On several occasions the
doublets of Antiochene fulfil this function by completing the sense
or clarifying the context by means of an alternative reading attached
to the reading of the majority. ^" This term is the right Greek translation in the parallel passage of 2 Kings 22:14.
Did the author of the Antiochene text take loO íiJ,aTio(pú?iaKoç from this parallel
passage? There is no trace of Hexaplaric reading to 2 Chr 34:22; Field's reference
points to alia exemplaria, in fact the reading of the Complutensian Polyglot which
follows the Lucianic manuscript 108: cf. F. FIELD, Origenis Hexaplorum quae super-
sunt I (Oxford 1875 = Hieldesheim 1964). ^^ As Montfaucon realized, instead of (puXáaaouaav Tàç èvToXáç the original
reading should be (pU/Váaaovioç làç aioXáç, and these words should refer not to
Huldah but to Shallum, her husband: of. F. SCHLEUSNER, Novus Thesaurus
philologico-criticus sive lexicon in LXX et reliquos interpretes graecos ac scriptores
apocryphos Veteris Testamenti (Lipsiae 1820) sub voce èvio^^fj. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.es A. INNER-GREEK CORRUPTIONS The same phenomenon of a doublet
or alternative reading based on a paléographie, inner-Greek, corrup-
tion can be detected in the Antiochene text of 2 Chr 21:19: Kai OÙK
¿Tioíriasv aÙTOÒ ò À-aòç aûxoC èKcpopàv Kai Kka\)üiv Kaià Tf|v
K^^auaiv TODv Tiaxspcov auioû. Again, this alternative reading has
succeeded in the text transmission because it fits the context of
mourning for Asa's death. — In 2 Chr 34:22 the MT speaks of «the prophet Huldah, the wife
of Shallum... son of Hasrah, keeper of the wardrobe (on>in nni\y>>. In the target language Shallum is no longer the keeper of the
wardrobe but 'a prophet Huldah... cpuX^aaaouaav xàç évxoXáç'
('that observes the commands'!). Already Montfaucon ^^ detected
the corruption of the original reading axoXáç, restored by A. Rahlfs, against the ívxokáq of the manuscript tradition. The Antio-
chene text, following his tendency to incorporate as doublets alter-
native readings, retains the corrupted reading of the current
Septuagint and introduces a correct translation of the Hebrew,
restoring Shallum (Sellem) as keeper of the wardrobe (ifiaxio-
(puXa^): Kai STiopsiiBri Xe^KÍaç... Tipòç "OX,ôav xf|v 7ipo(pfìxiv
yuvaÎKa ILzkXrwi DÍOC ©SKÓDS UÍOÜ ' Aaèp xoû íjiaxiocpúXaKoç http://sefarad.revistas.csic.es 348 NATALIO FERNÁNDEZ MARCOS Sef 64:2 (2004) Tf|v (pvXáaaovaav xàç èvxoXáç. The reading of IDW as feminine
participle by the translator (joining to the participle the article of
the following word) generated an embarrassing interpretation in the
target language and probably contributed to the succcess of the
corrupt reading èvxoXáq instead of axoXáç. — When the queen of Sheba visits king Solomon, she contem-
plated and admired all his wisdom and among other things in his
palace she was amazed by the clothing of his servants, and by his
cupbearers (vpv)Di, 1 Kings 10:5). The current Septuagint translates
this part of the sentence xòv í|iaxia)nòv aùxoû Kai xoî)ç oivoxóouç
aùxoC, in exact correspondence with the meaning of the root npv) in
hiphil, 'give to drink'. However, in Antiochene we come across xòv
í|Liaxia|Liòv aùxoC Kai xoi)ç eóvoúxouç auxoû ('his clothing and his
eunuchs'). It is the reading of Antiochene without variants, with a
meaning far different from the original. It is not plausible to imagi-
ne a paléographie confusion at the level of the Hebrew between
n¡7^D, the regular equivalent for oivoxóoç and ono, the stereotype
equivalent for euvoûxoç. '^ Ms /• of Brooke-McLean has still another corruption, iivioxouç, 'chariot-
drivers'. There are no Hexaplaric remains to this passage. Two late minuscles (243
and 244) solve the problem by creating the doublet: Kai TOÎ)Ç oivoxóouç Kai TOÙÇ
EÚvoúxouç: cf. F. FIELD, Oìigenis Hexaplorum, ad ¡oc. A. INNER-GREEK CORRUPTIONS However, at the level of the Greek trans-
mission the phonetic corruption by similar pronounciation of both
terms due to the phenomenon of itacism provides a reasonable
explanation. The reading also makes sense, and is consolidated in
the text transmission of the Antiochene family of manuscripts ^'\ (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.es ^'^ The Old Latin can be consulted in the apparatus of our edition of the
Antiochene text quoted in note 2. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.es B. A DIFFERENT READING OF THE CONSONANTAL TEXT The numerous passages characterised as aliter in the Index provi-
de some information on misleading translations caused by a dif-
ferent reading of the Hebrew text and, occasionally, by homophonic
translation. I emphasize that it is a typical phenomenon that occurs http://sefarad.revistas.csic.es 349 SOME PITFALLS OF TRANSLATION GREEK Sef6A:l (2004) in the translation process at the first level of encounter of the two
languages. Translation is a kind of reading and concretely the Sep-
tuagint is the first interpretation of an unvocalised Hebrew text. It
is a kind of performance of the consonantal text, like a score, to use
a musical metaphor. No doubt, in several cases it is clear that the
translators were following a different reading tradition or an exege-
tical device, but in many other cases the end product can be analy-
sed simply as a misreading. — In 1 Sam 2:31 the Hebrew word :^'ny, 'arm', is read twice as
V3J, 'seed'. The Hebrew sentence «See... I will cut off your arm and
the arm of your ancestor's family» becomes in Greek Kai i8oì)... è^oXoGpeúaco xò arcépina aou Kai xò aTiépiua xou OIKOU XOC
Tiaxpóç aou. — In 1 Sam 15:9, the different vocalization plus the confusion of
similar consonants leads to a new diverse sense in the target langua-
ge quite different from the parent text. Saul and the people spared
Agag, and «the best of the sheep... and the lambs (o^lDn), and all
that was valuable». The plural of ID, 'lamb' or 'ram' is read and
interpreted by the whole Greek tradition as xoov ajuTusXcovcov =
D'>)D'i|, the plural of 0*15. — In 1 Sam 16:20 the Hebrew onb mnn ^v)-) npn («And Jesse
took a donkey with bread») is translated in Antiochene: Kai zXa^zv
'Isaaai ovov, Kai cTiéOiiKev aúxoò y^M-op âpxcov. In all proba-
bility this sentence arose from the double translation of a single
word niDn with different vocalization as donkey ('liDq) and as a
measure (nD n). Antiochene utilizes this recourse to double inter-
pretation in order to solve the brachilogy of the Hebrew; the ma-
jority text of the Septuagint understood it as yó|uop. But only Antio-
chene makes a difficult sentence in Hebrew explicit. Interestingly,
the Old Latin retains the Antiochene reading: Et accepit lesse asi-
niini et imposuit super gonior panis ''^. ^^ MT: «A little cloud no bigger than a person's hand is rising out of the sea» (nbv
D>)3). The Old Greek àvótyouaa i35cüp probably read o>a nb^JD. Antiochene conflates
both readings and obtains a meaningful sentence. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) ^^ People from Gebal called Byblos by the Greeks. ^^ MT: «A little cloud no bigger than a person's hand is rising out of the sea» (nbv
D>)3). The Old Greek àvótyouaa i35cüp probably read o>a nb^JD. Antiochene conflates
both readings and obtains a meaningful sentence.
^^ People from Gebal called Byblos by the Greeks. B. A DIFFERENT READING OF THE CONSONANTAL TEXT http://sefarad.revistas.csic.es 350 5^/64:2(2004) NATALIO FERNÁNDEZ MARCOS — In 2 Sam 14:17 the Hebrew word nnìD)p, 'resting-place' is
translated by the entire Septuagint tradition by Guata, reading
nn;)D, 'offering', and changing the sense of the Hebrew sentence
«the word of my lord the king will set me at rest» into the new
r8VT|9f|TCÛ Ôf| ó XÓyOÇ TOÛ KUpíOU jUOÜ TOC Paai>^8CÛÇ
8ÍÇ
9uaíav. — The same source of confusion can be detected in 2 Chr 10:10:
iD^bVD bpn nriN, where Antiochene reads the preposition b:^ as b'v
'yoke' and translates accordingly: Kai ah vCv Kot3(piaov ànò TOO
CuyoC f||i(5v in contrast with the current Septuagint Kai oh ácpeç
àcp' fjjLlODV. An alternative reading of the consonantal text may produce a
double interpretation that Antiochene incorporates willingly into the
textual chain, as in 1 Kings 18:44, where the MT o>n, 'out of the sea',
has been translated twice by Antiochene: uSœp aTcò GaXáaarjç ^^. — The different vocalization of the MT may result in the inter-
pretation of some proper names as common names or verbs, with a
sense far from that of the original. In 1 Kings 6:3 (= MT 5:32a) it is
stated that in the building of the temple «Solomon's builders and
Hiram's builders and the Giblites did the stonecutting» (ibt7D>i
o>!7i:\ni Dn>n >m riDbv) >n), translated by Antiochene: Kai fjvey-
Kav oí üíoi Zo/lo|LioovToc Kai oí uíoi X8ipá|Li, Kai èvépallov
aÓToúç («and Solomon's sons and Hiram's sons brought the stones
and fashioned their borders»). The majority text of the LXX reads
èneXsKriuav ('did the stonecutting') instead of fjveyKav, and puts
the simple verb ëpallav instead of èvépaXov. But, what is more
important, the translator read with different vocalization 'sons'
(>Í5), not 'builders' (^A'l), and interpreted the proper name in the
plural 'Giblites' '^ as a hiphil of bn>, 'circumscribe', probably read- http://sefarad.revistas.csic.es 351 5^/64:2 (2004) SOME PITFALLS OF TRANSLATION GREEK ing Oib^^n, 'they fashioned their borders', a different reading
accepted as emendation to the MT by some modern dictionaries ^^. The homophonic translation may also explain some unusual equi-
valences in the Index. In these examples there is no reason for pos-
tulating a paléographie confusion at the level of the Hebrew lan-
guage. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) ^^ Cf. D. J. A. CLINES (éd.), The Dictionary of Classical Hebrew, vol. II (Sheffield
1995). Translation of proper names is very frequent in Antiochene. But, oc-
casionally, they are interpreted as common names: cf. èv TOÍ%Ü) TEKTOVIKÌÌC for 'at
Qir Hareset' in 2 Kings 3:25. '^ Of an unknown origin it means 'guard-chamber' for the outrunners: cf. L.
KOEHLER and W. BAUMGARTNERT, The Hebrew and Aramaic Lexicon of the Old
Testament [translated by M. E. J. RICHARDSON] (Leiden - Boston - Koln 1999). '^ H. B. ROSÉN, L'hébreu et ses rapports avec le monde classique. Essai
d'évaluation culturelle (Paris 1979) pp. 25-46 and F. VlNEL, La Bible d'Alexandrie.
18 L'Ecclesiaste (Paris 2002) pp. 55-57. ^^ Cf. D. J. A. CLINES (éd.), The Dictionary of Classical Hebrew, vol. II (Sheffield
1995). Translation of proper names is very frequent in Antiochene. But, oc-
casionally, they are interpreted as common names: cf. èv TOÍ%Ü) TEKTOVIKÌÌC for 'at
Qir Hareset' in 2 Kings 3:25.
'^ H. B. ROSÉN, L'hébreu et ses rapports avec le monde classique. Essai
d'évaluation culturelle (Paris 1979) pp. 25-46 and F. VlNEL, La Bible d'Alexandrie.
18 L'Ecclesiaste (Paris 2002) pp. 55-57.
'^ Of an unknown origin it means 'guard-chamber' for the outrunners: cf. L.
KOEHLER and W. BAUMGARTNERT, The Hebrew and Aramaic Lexicon of the Old
Testament [translated by M. E. J. RICHARDSON] (Leiden - Boston - Koln 1999). http://sefarad.revistas.csic.es "° In the parallel passage of 1 Kings 14:42 (= MT 14:28), the majority text of the
Septuagint transliterates 0££, and Antiochene OsKoüe. B. A DIFFERENT READING OF THE CONSONANTAL TEXT However the similar phonetics of the Hebrew and Greek
word may have influenced the selection of terms in the translation
process in passages such as 2 Chr 33:6 sv yi] B6V8VVÓ|LI for the He-
brew 0)n"il *>>! («in the valley of the son of Hinnom»), or 2 Chr 30:10
èv TOD opsi 'E9pái|Li Kaì Mavaaax] for the Hebrew DnDN""^(nNi
nv)3ni («in the country of Ephraim and Manasseh»). A phonetic
connection exists between n and yf], ^IN and õpoç in Hebrew and
Greek that might reasonably explain these uncommon translations. There may also be an underlying, diffuse conscience among Helle-
nistic Jews that Hebrew and Greek had something in common ^^ In 2 Chr 12:11 the guard of Rehoboam, whenever he went into the
house of the Lord, would come along bearing the shields of bronze,
«and would then bring them back to the guardroom» (NÎT^N Oìiv)m
o>i¿in). The verse has been diversely interpreted by the Greek
tradition. It is clear that the word Nn, 'guardroom' was not transparent
for the translators ^^. They resolve the difficulty with a puzzling
translation; the majority text of the Septuagint reads Kaì oí
87i:iaTpé(povT8ç sic àTiávTiiaiv xœv TiapaxpcxóvTODv. But the
homophonic translation appears clear enough in the double sentence
of Antiochene that includes 8Íç Tf|v xá^iv xcov Trapaxpexovxœv, a http://sefarad.revistas.csic.es 352 NATALIO FERNÁNDEZ MARCOS 5e/64:2 (2004) guess translation induced by the phonetic similarity of xáÇiç with
HT\ ^°. Hatch and Redpath insert an obelus instead of the Hebrew
equivalent of the septuagintal àTiávirjaiç, although the formal equi-
valence is beyond doubt. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.es C- INTERCHANGE OF SIMILAR LETTERS Another source of misleading translations lies in the confusion of
similar letters or groups of letters. The Index provides a mine of
information on unusual equivalences going back eventually to a
misreading of some consonants in the early square script. It is an
accident of reading or copying; in the first case it arises in the
course of the translator's deciphering of the Vorlage; in the second,
it reflects a Vorlage already at variance with the MT. It is not to be
excluded that a genuine textual difference underlies some of these
variants, but in general it can be stated that the paléographie
confusion at the level of the Hebrew script is the most plausible
explanation. The most frequent interchange of similar letters occurs
between n/i. — In 1 Sam 23: 15-16.18-19, the city where David remains
hidden in the wilderness of Ziph, Horesh (nv)nn), is translated
systematically in Antiochene by Kaivi^, obviously read as nv>in. — In 1 Sam 19:13.16 the uncertain Hebrew expression onvn
n^lD, translated commonly as 'net of goat's hair' (Vulgate pellem
pilosam caprarum), is interpreted in the whole Greek tradition as
i^Tiap aiyoDv, by reading the first term as 115, iiver'. — In 1 Sam 24:3 Saul went to look for David and his men •>DD b^
0'>bv^n m:^, ususally translated as «in the direction of the Rocks of
the Wild Goats». The majority text of the Septuagint reads STIÌ
TCpóacoTCOV 'Eôôaié|Li, that is, a transliteration (cum variantibus). http://sefarad.revistas.csic.es 353 SOME PITFALLS OF TRANSLATION GREEK 5^/64:2 (2004) But Antiochene interprets Kaxà TcpóacoTiov xfjç 9f|paç TCDV kXàcpcD
This interpretation is confirmed by the reading of the Old Latin
transmitted by Lucifer of Cagliari ante faciem venationis cervorum. No doubt, the translator read i>^, 'hunting' instead of the niii
'rocks' from MT. In this example the two most frequent interchan-
ges of similar letters concur: n/i and '>/^. But Antiochene interprets Kaxà TcpóacoTiov xfjç 9f|paç TCDV kXàcpcD
This interpretation is confirmed by the reading of the Old Latin
transmitted by Lucifer of Cagliari ante faciem venationis cervorum. y
g
No doubt, the translator read i>^, 'hunting' instead of the niii
'rocks' from MT. In this example the two most frequent interchan-
ges of similar letters concur: n/i and '>/^. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) Josephus, Ant. VI, 125: laxaiai ôè Kai aòiòç aùv TÔ) Tiaiôî Kai' àXXo juépoç. http://sefarad.revistas.csic.es C- INTERCHANGE OF SIMILAR LETTERS — In 2 Sam 22:21b it is stated «according to the cleanness of my
hands he [the Lord] recompensed me» ('>'p n^v)^ >i> *ii3), translate
literally by the current Septuagint as Kaxd Trjv Ka6apiÓTr|Ta xoòv
Xsipœv |Liou ávxaTiéôcoKév |uoi. However, Antiochene gives a dif-
ferent interpretation of the sentence: ôó^av X£ipá>v ¡noi) àvxaTio-
Sc5a8i jLioi, reading the first word as liiD. A similar graphic con-
fusion underlies the Antiochene term ôo£,ao"]Lióç in 2 Sam 22:25:
ôo^aajLióç |j.oi) àTiévavxi xœv o(pQaXiiôv aùxoC for the Hebrew
VD>V l>3b '>^2D. — In 1 Sam 14:40 Saul says to all Israel: «You shall be on one
side, and I and my son Jonathan will be on the other side» (*iivb
mH nivb ...inN). The whole Greek tradition transmits in both cases
£Íç Soü?t£Íav, reading 12V instead of niv. Interestingly, the Antio-
chene text adds, as a doublet, a new sentence with the correct sense
according to the MT: Kai SÍTIS I^aohX Tipòç xòv Xaòv
'Y)LI£ÎÇ
eGSuQe eiq êv |Liépoç, Kai éyò Kai 'IcovaGàv SGoixsQa 8Îç êv
JLispoç. The alternative reading, in agreement with MT, is not sup-
ported by any Hexaplaric witness, and we are probably dealing with
an early correction, already known to Josephus ^^ The double
reading SouX-eíav/juiépoç, based on the interchange of i/n generated
a new sentence. As is well known, a trend of the Antiochene text
consists of joining double readings with small redactional retouches
to clarify the meaning so that all the information of the preserved
variants can be explicit for public reading. The misreading of other graphically similar letters like >/i, i/\,
n/D; D/n, Ù/2, T/i, ^/y, letters with similar phonetics like the sibi-
lants Ì, V, \J, V); or the gutturals H, n, n, v, is also reflected in the http://sefarad.revistas.csic.es 354 NATALIO FERNÁNDEZ MARCOS Sef 64:2 (2004) Sef 64:2 (2004) Index. These phenomena have been recently dealt with by T. Mu-
raoka, E. Tov and A. Gelston ^^. The examples abound, especially in
the transmission of the proper names and other transliterated words. -- T. MURAOKA, «A New Index to Hatch and Redpath», ETL 74 (1998) pp. 257-
276; E. Tov, «Interchanges of consonants between the Masoretic Text and the
Vorlage of the Septuagint», in Sha'arei Talmon, eds. M. FiSHBANE and E. Tov
(Winona Lake 1992) pp. 255-267, and A. GELSTON, «Some Hebrew Misreadings in
the Septuagint of Amos», VT 52 (2002) pp. 493-500. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) -^ T. MURAOKA, Hebrew/Aramaic Index, p. 54. http://sefarad.revistas.csic.es ~^ Only the manuscripts AN followed by a few cursives transmit õpGpioç, 'of the
morning' or matutimis: cf. A. E. BROOKE, N. MCLEAN and H. St. J. THACKERAY,
The Old Testament in Greek. Part I, I and II Samuel (Cambridge 1927). (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) C- INTERCHANGE OF SIMILAR LETTERS Herewith a handful of illustrations: Reference
MT
Presumed Reading
Antiochene
1 Kings 12:16
r)Hi(qal)
nvn
póaKSiv
2 Kings 10:11
v'yiy
v':?H>
xoùç àyxiCTTeúovTaç
2 Kings 16:18
^vm
ivm
QeixeXiov
2 Kings 21:9
nvn {hiph.)
3vn
pSeJ^uaaeiv
1 Chi-4:10
nvn
nv*T
yvœaiç
1 Chr 12:33
niv {qal)
liV
poriGew In 1 Kings 21:38 (MT 20:38), the prophet waited for the king of
Israel along the road, «disguising himself with a bandage over his
eyes» (v)^^"!:?^ noNi \yDnnn). The entire Greek tradition reads this
part of the sentence Kai Kaxeôéaaxo év T8?ta|Liouvi TOÎ)Ç ócpOall-
|Lioí)ç auToC. Hatch and Redpath give as the Hebrew equivalent of
KaxaSsiv the hithpael of V^DD with a question mark. Muraoka ^^ put
this root between double brackets signifying that the equivalent
given by Hatch and Redpath is implausible. He pointed with an
arrow to the qal of DDV as the true equivalent that should replace
that of Hatch and Redpath. However, I think it is more plausible
that the translator read in this passage the qal of v)nn, regularly
translated in the Septuagint by Ô8Îv, Kaxaôsîv. The confusion of
D/n in the Hebrew script is frequent and also between the sibilants
v/\),
while ÜDV in qal is regularly translated by aipeiv, éTiiysiLií- In 1 Kings 21:38 (MT 20:38), the prophet waited for the king of http://sefarad.revistas.csic.es 355 Sef 64:2 (2004) SOME PITFALLS OF TRANSLATION GREEK Csiv and, in my opinion, its confusion with v^Dn is less probable. In
any case, it is just an example of how the presumed equivalents can
be seen differently by diverse scholars. The different reading based on paléographie confusion may affect
not only isolated consonants but also a group of letters, the phe-
nomenon of metathesis included: — In 1 Sam 8:16 the Hebrew reads «He will take your male and
female slaves and the best of your young men (oDmni'TiNi, iuve-
nes óptimos in the Vulgate) and donkeys», while the Greek tradition
interprets: Kai TOÎ)Ç SOUXODÇ \)\x,(hv Kai xàç ôouX^aç \)\x(bv Kai x
PoüKÓXia \)\x(bv xà àyaBà Kai xoùç õvouç ÙJLIOÒV. In view of the
regular equivalence between npi and POUKOA^IOV, it can reasonably
be presumed that the translator read ODnpi"TiNi. http://sefarad.revistas.csic.es ~^ KŒÎ 8KTia£V Ò 0£ÒÇ TÒV CCvGpCOTlOV èv SÍKÓVI ÔlttCpÓpO), ÒpÔlOV ó BeÒÇ CKTiaEV
aÙTÓv, cf. J. W. WEVERS, Septuaginta. I Genesis (GoUingen 1974) p. 6.
-^ Cf. A. SALVESEN, Symmachus in the Pentateuch (Manchester 1991) pp. 2-6.
^^ Cf. F. SCHLEUSNER, NOVUS Thesaurus, sub voce õpOioç and P. K. MCCARTER,
/ Samuel, AB 8 (Garden City, NY 1984) p. 419. McCarter prefers the reading of the
Old Greek as more genuine.
-^ W. A. VAN BEUKEN, «I Samuel 28: The Prophet as a 'Hammer of Witches'»,
7507 6(1978) pp. 3-17: p. 9. C- INTERCHANGE OF SIMILAR LETTERS — In 1 Sam 8:16 the Hebrew reads «He will take your male and
female slaves and the best of your young men (oDmni'TiNi, iuve-
nes óptimos in the Vulgate) and donkeys», while the Greek tradition
interprets: Kai TOÎ)Ç SOUXODÇ \)\x,(hv Kai xàç ôouX^aç \)\x(bv Kai x
PoüKÓXia \)\x(bv xà àyaBà Kai xoùç õvouç ÙJLIOÒV. In view of the
regular equivalence between npi and POUKOA^IOV, it can reasonably
be presumed that the translator read ODnpi"TiNi. — In 1 Chr 22:9 the king Jehu searched for Ahaziah, «who was
captured while hiding in Samaria» (^nDV)! NinxiD Nim iniDb>i). But the Greek tradition interprets unanimously: Kai KaxéÀ.aPov
auxòv íaxpeuójLisvov èv Sajuapeía. Hatch and Redpath insert an
obelus of uncertainty by iaxp8üó|Li8Vov as equivalent of Nin. However, given the regular equivalences of Kpúpeiv, KpÚTixsiv for
the hithpael of Nin, and, likewise, the regular equivalence of
iaxpeusiv for the hithpael of Non, it can be presumed that the
translator read NDnriD. Moreover, the confusion of final ì and final ^ leads to a quite
different interpretation in the Greek of 1 Sam 28:14, when the
woman medium evokes Samuel's spirit to Saul. To Saul's question
concerning Samuel's appearance, the woman answers according to
the MT: «An old man is coming up; he is wrapped in a robe» (v)>N
b>v)3 nuv Nini nbv ^pT). However, Antiochene translates with the
rest of the Greek witnesses "'^: âvôpa õpBiov àvaPaívovxa ÒLTÍÒ xf| http://sefarad.revistas.csic.es 356 NATALIO FERNÁNDEZ MARCOS 5^/64:2 (2004) yfiç, àvaPsPÀ,r||iévov ôiTiXoíôa. "OpBioç is a hápax of the Septua-
gint in this passage. The Hexaplaric witnesses represented by oí
À.oi7roí read, according to the MT, npeG^mr\v. But õpBioç is used
by Symmachus for the translation of Gen 1:27, the man's crea-
tion ^^. In this passage of Genesis Symmachus inserts an explana-
tory note relying probably on an exegetical tradition that empha-
sizes the most peculiar feature of the human being in contrast with
animals, his upright stance, a tradition that can be traced back to
Justin Martyr and other rabbinic sources ^^. In contrast, in 1 Samuel
28:14 it seems that the origin of the Greek reading is not exegetical
but paléographie. The verb ^p\ is translated by ávop9o6v in the two
occurrences of the Bible (Psalm 144:14 and 145:8). Moreover, it is
well attested with the meaning of 'stand upright, erect' in
postbiblical Hebrew as well as in Aramaic and Akkadian ^^. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) ^^ Less plausible, in my opinion, is the conjecture in BHS of ^\>V), as metathesis
of X)SùV in 2 Chr 20:9, based on the Antiochene or Lucianic text of the Septuagint
It occurs in a sequence of calamities announced, «if disaster comes upon us, sword,
judgement, or pestilence or famine», and was translated by Antiochene: éàv é7i£À9r|
ècp'fmâç KttKá, pofxcpaia, àKpíç, Oávatoç, Xijióc. The majority text of the
Septuagint reads Kpíaiç according to the MT, instead of àKpíç. The regular equi-
valent for àKpíç, 'locust', is r^TM, while the regular translation for <)S?V, 'flood',
KaiaKÀuafj-óç. Given the stereotype correspondence of these two words, I rather
consider àKpíç a secondary variant resulting from an inner-Greek corruption from
Kpíaiç. Again, this variant reading succeeded and consolidated in the text transmis-
sion because it was inserted in a sequence of disasters that made sense. C- INTERCHANGE OF SIMILAR LETTERS In any
case, an exegetical tradition may have influenced this version since,
according to the Midrash, when the spirits of dead people are evoked
from the netherworld, only the kings appear upright, face first; the
other persons rise feet first. This is, no doubt, why the woman recog-
nised Samuel ^^. Metathesis can be detected in some unusual translations, but it is
especially visible in the transliteration of proper names. In 2 Chr 28:3: Kai Sifiyaye xa xéKva auxou sv Trupi, for the MT
\!)Nn i^n-nN *ivi'>i reflects a different reading from the verb v)Ki ...niv">i,
in hiphil, a stereotype expression for «make pass through fire». In 2 Sam
22:13 it is said that «coals of fire flamed forth» (\i)N-'>bn> nvi). The
cuiTcnt LXX translates literally: è^8Kaú9r|0"av âv6paK8ç Tiupóç. However, Antiochene interprets the whole sentence as ôif|X-9ov xáX-aCa http://sefarad.revistas.csic.es 357 SOME PITFALLS OF TRANSLATION GREEK 5^/64:2 (2004) Kai áv9paK8ç Tiupóç. As a matter of fact, a reading m v underlies the
Antiochene translation. It is also probable that %àXa(,a, which com-
monly translates the Hebrew m i in the Septuagint, originated as a
double reading of this very word. Some items of metathesis in the transliteration of proper names
are the following: 'ApooCá for nuiv (1 Chr 2:18,19); 'Aôapí for
nnn (2 Sam 23:25); 'Acpapei for nvo ( 2 Sam 23:25); 0opya|iá
for nDn>in (1 Chr 1:6), and 'Pácpeç for ^^n (2 Kings 19:12) ^^ I believe that most of the commented phenomena can be explai-
ned as misreadings during the process of translation due to the in-
correct desciphering of the Hebrew Vorlage. Consequently, they are
of secondary character arising from an accident of the transmission,
be it in the copying of the Hebrew text itself or produced by a
misreading of the translator. It cannot be excluded, however, that
some of these variants conceal a genuine reading. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.es ^^ S. BROCK, The Recensions of the Septuagint Version of 1 Samuel (Torino 1996)
pp. 167-169.
^' Cf. E. D. HERBERT, «4QSam" and Its Relationship to the LXX: An Exploration
in Stemmatological Analysis», p. 46.
^- Cf. E. ULRICH, The Qumran Text of Samuel and Josephus, HSM 19 (Missoula,
MO 1978) p. 95 and A. FiNCKE, The Samuel Scroll From Qumran. 4QSam" restored
and compared to the Septuagint and 4QSam% STDJ 43 (Leiden - Boston - Koln
2001) p. 12. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) ^^ S. BROCK, The Recensions of the Septuagint Version of 1 Samuel (Torino 1996)
pp. 167-169. ^- Cf. E. ULRICH, The Qumran Text of Samuel and Josephus, HSM 19 (Missoula,
MO 1978) p. 95 and A. FiNCKE, The Samuel Scroll From Qumran. 4QSam" restored
and compared to the Septuagint and 4QSam% STDJ 43 (Leiden - Boston - Koln
2001) p. 12. http://sefarad.revistas.csic.es ^' Cf. E. D. HERBERT, «4QSam" and Its Relationship to the LXX: An Exploration
in Stemmatological Analysis», p. 46. D. TRACES OF A DIFFERENT VORLAGE It is common knowledge that the Antiochene text is rooted in the
Hebrew not only as part of the Septuagint tradition, but also due to
the fact that it incorporates a set of Hexaplaric corrections accord-
ing to the MT. Sometimes it is even closer to the MT that the rest
of the Septuagint tradition. Moreover, S. Brock realized that not all
the approximations to the Hebrew in Antiochene were of Hexaplaric http://sefarad.revistas.csic.es http://sefarad.revistas.csic.es 358 NATALIO FERNÁNDEZ MARCOS 5^/64:2(2004) 5^/64:2(2004) provenance ^^. Thanks to the discovery of the Qumran documents for
Samuel this statement has been confirmed. There are a few Antio-
chene deviations from the MT that are supported by 4QSam^ The relationship between the textual witnesses of the book of
Samuel is very complex and, therefore, it is dangerous to make any
kind of generalization. On the other hand, only with the full publi-
cation of the fragments and a thorough comparative study of all the
witnesses can the net of relationship be ascertained. Provisionally, it
can be stated that 4QSam'' was not the Vorlage of the Antiochene
text; the lack of secondary agreements or conjunctive errors between
both texts do not allow such a close relationship to be established ^\
For our purpose it will suffice to point out some agreements of
Antiochene with 4QSam'^ leaving a full comparison of both witnes-
ses for a further study. — 1 Sam 5:9: «And it ocurred that after they had brought it [the
ark of God]» (iriN iipn nnN ^n>i), in the majority text of the Sep-
tuagint the translation is Kai éysvfiSii ixsxà xò fiexe^^Geîv aÚTf|v. However, in Antiochene we come across the following interpre-
tation: Kai ¿yevexo ¿v xœ iiexeXQslv xf|v KIPCOXÒV Tipòç xoi)ç
Y899aíouç. This version makes explicit the noun of the ark,
translated literally by the pronoun in the Septuagint, but, what is
more important, it mentions Gath (xoi)ç yeGôaíouç) as read in
4QSam'' (nn:^ 110 nnN ^n-^i), but absent in MT ^l — 2 Sam 12:16: When the Lord struck the child that Uriah's wife
bore to David, the king fasted «and went in and lay all night on the
ground» (ni¿nN 1DV>1 p^ HII). The Vaticanus and his group of http://sefarad.revistas.csic.es 359 SOME PITFALLS OF TRANSLATION GREEK Sef 64:2 (2004) manuscripts translate Kai 8ÌafìX08v Kai T]X)XÍGQT] EKÌ Tf|ç yf|ç. ^^ A. FlNCKE, The Samuel Scroll, p. 202. The Antiochene reading is supported by
the Old Latin: donnivit in cilicio.
^^ E. ULRICH, The Qumran Text, pp. 100-101.
^' E. ULRICH, The Qumran Text, p. 104 and A. FiNCKE, The Samuel Scroll, p. 261.
^^ E. ULRICH, The Qumran Text, p. 113 and A. FiNCKE, The Samuel Scroll, p. 263. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.es D. TRACES OF A DIFFERENT VORLAGE However, Antiochene renders: Kai siasXQcòv
SKáQevbsv èv
aáKKCO 871Ì TTÍv yfiv, a version that corresponds exactly to the
reading of 4QSam'' to this passage, n^^H pM)2 2Dm Ninn ^l The
Alexandrinus and the group of MN plus some minuscles read
r\\)XiaQT] Kai 8KOi|Lní6T| according to the MT. Only Antiochene
uses Ka08uÔ8iv the frequent equivalent for IDV) in the Septuagint ^'^. — 2 Sam 22:43: «I beat them fine like the dust of the earth»
(opnv)Hi ^^nN'nDVD), has been translated in the current Septuagint:
Kai S/^éava auxoùç cbç xvovv jf\q. But Antiochene translates the
sentence differently: ôiaaKopTiiGÒ aùxoijç òç x^^Cv 87iì TcpóacoTiov
àvéjiou. The Antiochene version is closer to the Qumran reading
(niH >3D ÙV noVD opnvym) than to the MT ^\ Interestingly, the
Vorlage of Antiochene was not identical to that of the Qumran
fragments; it probably read n n 'wind' instead of nn"H 'wanderer',
the reading of the parallel passage in Psalm 18 (17):43. — In 2 Sam 23:1, the majority text of the LXX is close to the MT,
while the Antiochene version follows the reading of 4QSam^: Oracle
of David, son of Jesse, «and oracle of the man who was exalted on the
anointed of the God of Jacob» (npv^ ^nl^H n>v)n bv opn nn>n ONDI), is
rendered in the current Septuagint as Kai Tiiaxòc àvfjp ôv àvéa-
TT|a8v Kúpioç 87ii xpiGTOv 98oû 'IaKOOp. However, the Antio-
chene family of manuscripts translates Tiiaxòç àvfjp ôv àv8aTr|a8V
ó 08ÒÇ xptcjTÒv, 08ÒÇ 'laKcop, a literal rendering of 4QSam'': 2p]V'> ^nnbN] nv^D bN o>pn I2>n [OND ^^ It is clear that Antiochen
read a Hebrew text similar to the fragment of Qumran, that is, bn
(68ÓÇ), instead of bv (= èrcí) of the rest of the Septuagint tradition. http://sefarad.revistas.csic.es 360 NATALIO FERNÁNDEZ MARCOS 5^/64:2 (2004) In two other cases, the reading underlying the whole Greek
tradition is witnessed in Qumran, not in the MT: 1 Sam 2:8-9 the
use of soXoyeïv in the Septuagint is transparent of the Qumran
reading *jnn'>i ^^, not of a different or corrupted MT. And in 1 Sam
2:20 the current text of the Septuagint with àTioxíveiv as well as
the Antiochene variant with àvxaTioSiôóvai are supported by the
Qumran reading obv^^ ^^ instead of the ov^^ of the MT. ' ATCOTÍVSIV
and àvxaTioSiSóvai are regular equivalents for the piel of obv) in
the Septuagint, while these two verbs are never used for ow. ^^ E. ULRICH, The Qumran Text, p. 119 and A. FiNCKE, The Samuel Scroll, p. 9.
^^ E. ULRICH, The Qumran Text, p. 72 and A. FiNCKE, The Samuel Scroll, p. 10. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) D. TRACES OF A DIFFERENT VORLAGE These agreements between the Greek text, especially the Antio-
chene, and an extant, non-Masoretic, Hebrew, lead us to the conclu-
sion that, in all probability, several other deviations of Antiochene
are also rooted in the Hebrew. In this context I would like to point
out a series of doublets in the Antiochene text whose origin can
only be explained at the level of the Hebrew, a Hebrew text dif-
ferent from the MT. Such cases also confirm, from another pers-
pective, that the Antiochene text is rooted in the Hebrew. A typical
example will serve as an illustration: — In 2 Kings 2:23 while Elisha was going up on the way to
Bethel, «some small boys came out of the city and jeered at him,
saying (i!? nDNn in-ipt^pnn), «Go away, bald-head! Go away, bald-
head!». The current Septuagint renders literally: Kai Tiaiôápia
|iiKpà é^f|/^9ov èK xfiç TióÀ-scoç Kai KaxéTiaiCov aùxoû Kai SÍTCOV
aùxœ. Notwithstanding, Antiochene emphasizes that the boys not
only mocked him but also threw stones at him: è^fjXBs Tiaiôápia
jLiiKpà SK xf|ç TióXscoç Kaì éXíOaCov aùxòv Kai KaxéTcaiCov
aùxou Kai eXsyov auxoò 'Avápaivs, (paXaKpé, àvápaive,
(paXaKpé. The use of é|U7iaíC£iv, KaxaTiaíCsiv for the hihtpael of
obp is consolidated in the Septuagint. The use of XiBáCsiv,
XiQo^oXelv for all the forms of bpo is also well attested among the
Greek-Hebrew equivalences. Consequently, it can be deduced that http://sefarad.revistas.csic.es 361 SOME PITFALLS OF TRANSLATION GREEK 5^/64:2 (2004) this curious doublet ultimately relies on a different Vorlage with the
reading bpp, or on the extant MT read with metathesis of conso-
nants by the translator. Interestingly, the Old Latin retains only this
second interpretation of the Antiochene text: pueri pusilli exierunt
de civitate et lapidabant ilium dicentes: Ascende calve, Ascende
calve. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) E. CONCLUSIONS Through the lens of translation, particularly of the Antiochene
text in the historical books, I have tried to point out some of the
pitfalls that may have occurred in the process of translation and
transmission. An awareness of these mistranslations is the only way
of correctly evaluating the Greek variants for the restoration of the
genuine text. Some mistakes have been produced, such as inn^r-
Greek corruptions, through the frequent copying of the manuscripts. Several mistranslations arose as a result of a different vocalization
on the side of the translators. Other variant readings were produced
by the confusion of similar consonants or groups of letters; these
variants or alternative readings can be explained only at the level of
the Hebrew. And finally, in a few cases, an extant, non-Masoretic,
Vorlage has been detected in the Hebrew fragments of 4QSam^
These agreements open a window toward a textual stage when
different Hebrew texts were in circulation. The Vorlage of the
Septuagint (Old Greek) was one of them. MT is the only complete
Hebrew text available, but we must be aware that the Greek
tradition, when it deviates from the MT, may conceal another text,
with a striking resemblance, but not identical to the MT. Some
scholars maintain that the Vorlage of the Septuagint in the books of
Kings is older and probably more genuine than the MT. There are numerous passages in the Index preceded by the
mention of the mark aliter. These draw our attention to the specific
texts which should allow a continuous exercise of textual criticism http://sefarad.revistas.csic.es 362 NATALIO FERNÁNDEZ MARCOS 5^/64:2 (2004) with all the evidence at our disposal, and this, in the knowledge that
not every scholar will come to the same conclusions in a great many
of these text-critical problems. RESUMEN En crítica textual es muy importante descubrir la génesis de los errores; a veces
la lectura verdadera sólo se descubre desenmascarando la falsa. De igual manera,
para usar críticamente la Septuaginta es imprescindible descubrir primero las
corrupciones y los errores de traducción. La confección de un índice griego-hebreo
del texto antioqueno en los libros históricos es una ocasión excelente para analizar
el proceso de traducción y detectar los errores más comunes cometidos por los
traductores. En el artículo se estudian algunos ejemplos con relación a los
siguientes fenómenos: corrupciones internas al griego y traducciones equivocadas
motivadas por la confusión gráfica de letras (paleografía) o sonidos (fonética)
semejantes y por una vocalización diferente del texto consonantico. En varios
casos este análisis permite vislumbrar un texto base hebreo distinto del masorético. PALABRAS CLAVE: Crítica textual, técnicas de traducción, hebreo y griego. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.es SUMMARY In textual criticism it is important to detect the genesis of mistakes; sometimes the
tme reading is only reached through the unmasking of the wrong one. Likewise, in order
to use critically the Septuagint it is indispensable to find out first its conuptions and
mistranslations. The making of a Greek-Hebrew Index of the Antiochene Text in the
Historical Books is an excellent occasion to observe the translation process and find out
the most common eiTors made by the translators. A few examples will be commented
concerning the following issues: inner-Greek comaptions and misleading translations
caused by the graphic confusion of similar letters (paleography) or sounds (phonetics),
and by a different reading or vocalization of the consonantal text. In several cases this
analysis may open a window towards a non-Masoretic Hebrew Vorlage. KEYWORDS: Textual criticism, translation technique, Hebrew and Greek. (c) Consejo Superior de Investigaciones Científicas
Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.es http://sefarad.revistas.csic.es
|
https://openalex.org/W4309048490
|
https://jssidoi.org/ird/article/download/113
|
English
| null |
Implementation of effective solutions to the crisis tasks and its regional management
|
Insights into regional development
| 2,022
|
cc-by
| 9,457
|
Received 18 August 2022; accepted 20 October 2022; published 30 December 2022 Abstract. Because of unexpected negative social and natural phenomena, there is a growing need to prepare crisis structures of society,
also due to technical and technological progress, which directly affect these facts. The facts defined in the article encourage the social
requirement for practical, creative, and interactive training of individuals and teams in the public interest and public administration. The
need for adequate preparation should reflect the logical and systematic analysis of dependencies and contexts in all procedures and
processes of identifying negative phenomena to encourage and create thought stimuli of crisis staff, which were able to minimize losses
and maximize opportunities. The potential of individual types of simulation technologies and their simulation tools is a possible means to
achieve the already mentioned requirements for practical training of crisis staff. Keywords: crisis management; crisis manager; education; simulation technology; financial calculation Keywords: crisis management; crisis manager; education; simulation technology; financial calculation Reference to this paper should be made as follows: Grega, M., Nečas, P. 2022. Implementation of effective solutions to the crisis tasks and
its regional management. Insights into Regional Development, 4(2), 21-35. http://doi.org/10.9770/IRD.2022.4.4(2) JEL Classifications: I18, J28, K10, K22, K37 JEL Classifications: I18, J28, K10, K22, K37 JEL Classifications: I18, J28, K10, K22, K37 Additional disciplines political sciences; sociology; information and communication; informatics * This work was supported by the Slovak Research and Development Agency under the contract No. APVV-20-0334. E-mails: 1 matus.grega@aos.sk; 2* pavel.necas@umb.sk (Corresponding author) E-mails: 1 matus.grega@aos.sk; 2* pavel.necas@umb.sk (Corresponding author) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/IRD/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/IRD/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/IRD/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/IRD/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2)
Publisher
http://jssidoi.org/esc/home
IMPLEMENTATION OF EFFECTIVE SOLUTIONS TO THE CRISIS TASKS AND ITS REGIONAL
MANAGEMENT * 2022 Volume 4 Number 4 (December
http://doi.org/10.9770/IRD.2022.4.4(2
Publisher
http://jssidoi.org/esc/home Publisher http://jssidoi.org/esc/home IMPLEMENTATION OF EFFECTIVE SOLUTIONS TO THE CRISIS TASKS AND ITS REGIONAL
MANAGEMENT * IMPLEMENTATION OF EFFECTIVE SOLUTIONS TO THE CRISIS TASKS AND ITS REGIONAL
MANAGEMENT * Matúš Grega ¹, Pavel Nečas ²* 1, Armed Forces Academy, Demanová 393, 031 01 Liptovský Mikuláš, Slovakia
2 Matej Bell University, Kurzmányho 1, 974 01 Banská Bystrica, Slovakia 1. Introduction Society and the individual, which is an integral part of it, have constantly encountered security conditions that
affected him existentially throughout his existence. From a historical point of view, it is observable that each
society was purposefully preparing itself, in a certain way, for already existing and potential threats. Society was
able to face, to some extent, avoid, eliminate and even predict the specifics of security conditions. Hazards arising
from natural patterns, such as windstorms with torrential rains and subsequent floods, devastating fires,
earthquakes, etc., but also direct anthropogenic threats, such as war clashes, industrial cataclysms and catastrophic
and apocalyptic manifestations thereof, were accompanying features when the individual and society learned to 21 INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) manage and control the patterns arising from them, they were forced to face them directly. They had to take
measures to ensure historically conditioned security. (Blažek, 2014) manage and control the patterns arising from them, they were forced to face them directly. They had to take
measures to ensure historically conditioned security. (Blažek, 2014) Following the threats, the field of public administration has structurally and in terms of content created and
gradually became an essential tool of management function. This management function was used and shared by
the elements of social system management, which ensure the operation of the society in all its areas of existence
and represent public administration and the public interest. The facts mentioned above directly affected and, in a
way, created the conditions for actions and works associated with the management function of responsible
individuals or groups to begin to fulfil the tasks that result from the specific functions of crisis management as a
functioning unit. (Hašanová, Dudor, 2014) Ensuring basic internal security processes and areas, such as defence, protection, external and internal security,
and security of the social apparatus (state or individual) represents important attributes that activate, stimulate and
thus directly affect the components of public administration. From a historical point of view, analysis of
knowledge and experience and scientific research, the fundamental pillars of public administration - the
authorities, in terms of opinion, and legislation, jointly or separately, participated in creating and ensuring the
security of states, their citizens and other elements that make up the security environment. 1. Introduction The development of a
public administration system is an ongoing process in which ties are permanently managed, responsibilities
updated, new relationships defined, competencies and implemented administrative national and international
standards, the application of which is subject to a specific component of public administration, are strengthened. Despite the constant improvement of ties, procedures, and processes in matters of communication, decision-
making, control, planning, executive, execution, etc., historical development in the area has failed to prevent the
emergence of newly created crises, phenomena and situations that public administration authority has to face and
for which they must be well prepared. In these intentions, it is necessary to realize that crisis phenomena and situations have been and will continue to
accompany individuals and communities. They are a particular representative of the interaction of the individual
with society, the mutual exchange of societies and individuals, and humanity's interaction with its external
environment. This is the impact of the adverse effects on the individual and community. Their specifics and
common denominators are: −
complicated prediction, −
destabilization of social and natural balance, −
destabilization of system subjects and their subsystems, −
a direct threat to individuals, society, and their mutual ties, −
Destruction of cultural, social, and historical values. Crisis phenomena directly but indirectly affect the economy, politics, and infrastructure of the community
negatively; they also affect the lives of individuals, biosocial groups, their health, and material goods and, to a
significant extent, negatively destabilize the natural and socio-social aspects of an individual's life. Experience
shows that the invoice value of lost or degraded goods, health and capacity constraints of logistics infrastructure
represent only a visible part of the costs and losses of companies. (Korecki, Adámková, 2020) Security system y y
The security system is representative of a complex of elements and means to ensure the safety and protection of
lives, human health and property. It represents an integrator of elementary beams of security, institutional and
system elements for its security. It is created based on risk identification, analysis, evaluation and taking the
following measures to ensure security. Using the method of multitude quantitative risk analysis using mutual
correlation, the dependency of the management structure of the selected critical infrastructure area is then
determined, which is expressed by the mutual relation of identified risks. (Korecki, Adámková, 2018) The created security system must represent a set of applications to have the qualities, which will enable it to
ensure different levels of survival and functioning of the social whole in case of expected or unexpected threats. To ensure the required basic tasks, this system is designed to be open, flexible and evolving, with multiple
interconnected networks of relationships in individual levels and levels of system and subsystems subjects, which
allows it to respond adequately, to react to potential as well as natural threats (Ivančík, 2019). The primary
qualitative attributes of a social security system include, for example, the ability to: analyze the security
environment; identify and classify national security risks and threats and trends; identify procedures and measures
to prevent or eliminate security risks and threats; ensure their continued readiness and ability to act
(Ivančík, 2021); identify crisis solutions in line with current resources and capacities; ensure effective
management of forces and aids in prevention, readiness to manage crises, adequate reactions and responses to the
activated source of threat during an emergency, to mitigate and eliminate the consequences of situations and to
create conditions for the renewal and development of society. Many of the tasks and requirements placed on the
security system and its creation were often carried out in contrast to the situation, that the readiness of the system
to act, to be adequately prepared for real as well as potential threats, could only be verified during an actual
emergency or crisis. Experimenting with the security of society, states, and humanity is risky and possible only to
a limited extent, so attention in this area has also focused on using crisis scenarios. Public administration In general, an opinion has been settled that public administration consists of two primary components, namely
state administration and self-governance, which realize public interest activities. Its main features are activity,
initiative, determination, and focus on shaping the future while complying with the law and learning general tasks. Public administration can also be understood as a complex of methods, measures, ways and procedures that put
the mechanism of economic policy in place, ensuring its objectives, regulation and development. This clearly
shows the interrelationship between economic policy and public administration, as economic policy cannot exist
without public administration and public administration loses its importance without economic policy. In this
context, a more precise definition by the authors Strecková and Malý has a deeper meaning and substance: "The
public administration is represented by the performance of self-administration and state administration, because
the sector is financed from public finances and public finances are formed from sources of the population, there is
no entity available for public sector management other than the public administration”. (Strecková, Malý, 1998,
p.25). 2. Crisis management as a complex system The concept of crisis management in a regional public arena has resonated for a long time. It has been the subject
of frequent scientific controversies and interpretations in the theory of law, politics, economics, and the
community. One may consider the following entities as being relevant elements of the system. 22 INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) Crisis scenario In the theory and practice of crisis management in public administration, we are increasingly facing the issue of
crisis scenarios, which are the result of knowledge and risk assessment, as well as the development of the security
environment. Crisis scenarios in public administration are based on historical experience and the degree of
knowledge of their nature. The development of scientific knowledge in fields such as security risk theory,
sociology, public administration theory, etc., enabled a deeper penetration into the essence of security, processes
and mechanisms of ensuring it (security) as much as possible. At the same time, it enabled the transfer of
scientific knowledge into crisis scenarios. Throughout historical development, humans and their various social communities have encountered multiple
forms of danger, threat, risk and the resulting crises. The historical stage of development of human society
determined to what extent a particular community consciously prepared for natural as well as potential forms of
threat and was able to prevent, confront or mitigate their harmful effects. Man, and human society, throughout
their lives, have always encountered specific states of the security environment that have influenced it
existentially as well as security, but also which he created by himself. In the historical development, emerged,
structurally and in terms of content grew the field of public administration, which became a type of management
activity by which the subjects of social system management carry out the course of the society in all areas of its
existence and which constitutes public administration and management in the "public interest." The attention of society was focused on the primary sources of danger, threats, risks, and the resulting crises,
which it most frequently encountered in its historical development and whose destructive factors and extent was
the greatest. Initially, supernatural forces were considered sources of threats depending on the development of
science and human knowledge. They were gradually replaced by an approach that accepts the knowledge of
natural and social pressures and their development patterns. Based on the gradual scientific penetration into their
essence, three large groups of threat sources were created in the system of opinions: a congenial group of natural
sources of threat, a congenial group of civilization sources of threat, and a disparate group of their combinations. We also encounter a simplified approach that presents the dangers, hazards, risks, and crises resulting from them
in two large disparate groups, namely non-military and military threats. Security system The gained experience and knowledge from personal crises of human society - natural disasters (floods, volcanic
eruptions, earthquakes, extreme climatic fluctuations, etc.), technical and technological accidents (accidents of
nuclear power plants, chemical plants, etc.), war conflicts (world and local war conflicts, the possibility to use
weapons of mass destruction, etc.) were scientifically systematized. Attention was paid to the use of crisis
scenarios in the crisis management system of public administration at all levels and areas. The assumption that crisis scenarios are an effective and inexpensive tool for "managing" the adverse effects of
crises, protection of lives, health and property of citizens and society, environment, defence and security interests
has been confirmed. The growth of the close connection between the gained experience and scientific knowledge
increased the importance of crisis scenarios for society. They became an essential part of theoretical and practical 23 INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) activity in the security field, an important starting point for security policy and the creation of national security
systems. In the current understanding, crisis scenarios result from recognizing the need to prepare society and
individuals for threats and crises, the starting point for creating threats and problems identification tools, and
mitigation or elimination of their adverse effects. activity in the security field, an important starting point for security policy and the creation of national security
systems. In the current understanding, crisis scenarios result from recognizing the need to prepare society and
individuals for threats and crises, the starting point for creating threats and problems identification tools, and
mitigation or elimination of their adverse effects. Crisis scenario 24 INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) These groups of threats also correspond to crises: These groups of threats also correspond to crises: These groups of threats also correspond to crises:
−
Non-military crisis - a crisis caused by the impact of natural disasters, technical and technological
accidents, as well as the action of social forces. This type of crisis is in the legal norms of our state
expressed in terms of an emergency state of emergency. −
Non-military crisis - a crisis caused by the impact of natural disasters, technical and technological
accidents, as well as the action of social forces. This type of crisis is in the legal norms of our state
expressed in terms of an emergency state of emergency. −
Military crisis - a crisis that arose because of the threat of use or use of military force, which is expressed in
our legal system in terms of the state of war, war. −
Military crisis - a crisis that arose because of the threat of use or use of military force, which is expressed in
our legal system in terms of the state of war, war. Crises have different durations and intensities of harmful and destructive phenomena and affect different sizes of
territories and numbers of people. They are the carriers of processes with a high degree of uncertainty for society
and man. They hurt various areas of society; they can accumulate several random phenomena and trigger other
destructive cycles. The core of the crisis consists of unexpected negative phenomena and their impact on the
human, social system, social infrastructure, and the environment. Crisis management in public administration Crisis management in public administration represents a specifically hierarchical and fully functional system of
public authorities and organizations, their ties, where their competencies and responsibilities are set. The output of
this system is the implementation of approaches, opinions, experiences, recommendations, measures, and
decisions. The general terminology about crisis management refers to solving various undesirable conditions. The
first use of the term crisis management is highly debatable, but consensus prevails in the area where it was applied
for the first time, and that is politics. In the 1970s, when the term crisis management became more widely used,
the term began to be used in theories of business economics. However, its definitions differed, and the
interpretation took different forms. Howard Chase and his colleague Barry Jones were the first, who try to define
the term. They described crisis management as "an unsettled matter which is ready for decision" (Chase, 1984). Today several definitions depending on the field in which it is applied (Bilczak, 2021; Rak et al., 2022). However,
the essence remains unchanged. Crisis management represents a specific form of management that has a high
degree of priority, and its main tasks are (Šenovský, Adamec, 2004): y
g
y
(
,
)
Today several definitions depending on the field in which it is applied (Bilczak, 2021; Rak et al., 2022). However,
the essence remains unchanged. Crisis management represents a specific form of management that has a high
degree of priority, and its main tasks are (Šenovský, Adamec, 2004): −
crisis state prevention, −
crisis state overcoming. Prevention means undergoing critical processes using appropriate forecasting, analytical and management
methods that allow the system to be protected and do not allow these essential processes to expand. Overcoming
means a precisely oriented, thoughtful, and targeted reaction that recovers an incompetent, destroyed, or damaged
system. Crisis scenario These are understood as: −
Non-military threats are real destructive potencies, the activation of which can lead to significant
deterioration and weakening of the economic and security function of the state due to the action of natural,
economic, internal, and international forces. Non-military threats must also be understood and analyzed in
terms of mutual determination, accumulation, and the possibility of the emergence of an individual threats
chain reaction, which may result in larger-scale threats. Non-military forces and means are used in the first
place to eliminate them. −
Non-military threats are real destructive potencies, the activation of which can lead to significant
deterioration and weakening of the economic and security function of the state due to the action of natural,
economic, internal, and international forces. Non-military threats must also be understood and analyzed in
terms of mutual determination, accumulation, and the possibility of the emergence of an individual threats
chain reaction, which may result in larger-scale threats. Non-military forces and means are used in the first
place to eliminate them. −
Military threats represent such groups of activities of a state or coalition of states or organizations
(especially terrorist ones) that endanger the security of coalitions of states or the state, comprising mainly
the immediate use of military forces and means, as well as the mediated indirect use of military forces and
means. The military threats associated with armed violence have become an area of exquisite growth in the
historical development of humanity, even though it reflects a sad place in our history. −
Military threats represent such groups of activities of a state or coalition of states or organizations
(especially terrorist ones) that endanger the security of coalitions of states or the state, comprising mainly
the immediate use of military forces and means, as well as the mediated indirect use of military forces and
means. The military threats associated with armed violence have become an area of exquisite growth in the
historical development of humanity, even though it reflects a sad place in our history. Public administration versus crisis management In terms of universal definitions, it is possible to derive the meaning of crisis management for public
administration. Crisis management in the context of public administration is understood as a set of tasks and
measures that the administration performs independently or in cooperation with other organizations before, during
and after crises to ensure the protection and security of the population. In a broader sense, these are: −
preservation and maintenance of public administration functionality, −
preservation and maintenance of public administration functionality, −
maintenance of population health, −
ensuring the availability of vital services to the population, −
securing private and public property, −
ensuring public order, −
preservation of cooperative activities of state rescue and security forces, −
ensuring informedness and −
provision of humanitarian aid. 25 INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) Crisis management in public administration represents a specifically hierarchical and fully functional system of
public authorities and organizations, their ties, where their competencies and responsibilities are set. The output of
this system is the implementation of approaches, opinions, experiences, recommendations, measures, and
decisions. Thus, crisis management represents a complex task whose primary goal is correction, prevention,
contraction, reduction, and reconstruction. The problematic tasks and their goals should be based on a holistic
point of view, i.e., that crisis state prevention and preparedness are both part of the process and partial tactical
steps taken until the crisis occurs. This complex of tasks and their goals represents a closed circle because, after
the elimination of a crisis state, there is a period of preparation for the future crisis state. (Figure 1) Fig. 1. Complex tasks and their goals by period
Source: Jaques (2007) Fig. 1. Complex tasks and their goals by period
Source: Jaques (2007) Source: Jaques (2007) A specific group of managers, namely crisis managers, was also identified by introducing and defining the term
crisis management. Tasks connected to crisis management performed by a crisis manager are not usually a
decision of one person but rather are served by several crisis managers (experts in specific fields) who are
members of the crisis staff. The crisis staff implements complex tasks and measures before, during and after the
crisis. Public administration versus crisis management The competent
authorities need an erudite and practical approach based on early warning, notification, activation, analysis, and
adoption of appropriate measures and decisions. Counter(re)action - active involvement in the emergence and gradation of the crisis state. The competent
authorities need an erudite and practical approach based on early warning, notification, activation, analysis, and
adoption of appropriate measures and decisions. p
pp
p
Reduction - minimizing the outbreaks of crisis states and eliminating their adverse effects. The executive needs a
coordinated approach and an active process by the managing authorities. Reduction - minimizing the outbreaks of crisis states and eliminating their adverse effects. The executive needs a
coordinated approach and an active process by the managing authorities. Reconstruction means eliminating the consequences of a crisis state and returning to the state before the crisis
state (stabilized condition). The need to stop the effects that caused a crisis state and take such decisions and
measures that ensure that the recurrence of the crisis state is minimized. Reconstruction means eliminating the consequences of a crisis state and returning to the state before the crisis
state (stabilized condition). The need to stop the effects that caused a crisis state and take such decisions and
measures that ensure that the recurrence of the crisis state is minimized. The institutional dimension of crisis management of public administration authorities is presented as a
competently integrated system of elements of crisis management. It is divided into management and executive
components, which have a certain economic autonomy. Public administration versus crisis management The security system, as a representative of a complex of elements and means in solving the tasks of crisis
management in public administration, must have elementary attributes without which such a system would not
work: −
Broad-spectrum and centrality define the participatory approaches of individual components. The
components are in coherence with each other, where the structure, hierarchy and responsibilities are
precisely defined, and these components are effectively influenced and guided. −
Legitimacy and legality, which are defined by the law order of the Republic, by respecting the results of
regional and parliamentary elections, as well as by applying international standards and obligations. −
Legitimacy and legality, which are defined by the law order of the Republic, by respecting the results of
regional and parliamentary elections, as well as by applying international standards and obligations. 26 INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) −
Control and transparency exist where effective feedback mechanisms allow for evaluating the taken
decisions and the planned decisions, especially in the competencies and responsibilities of crisis managers. −
Control and transparency exist where effective feedback mechanisms allow for evaluating the taken
decisions and the planned decisions, especially in the competencies and responsibilities of crisis managers. −
Versatility is representative of the continuous functioning of the system in both crisis and "non-crisis"
states. −
Versatility is representative of the continuous functioning of the system in both crisis and "non-crisis"
states. Functions of crisis management of public administration authorities functionally (functional dimension) carry out
a complex of tasks. These functions represent another elementary framework that cannot be neglected and whose
primary goal is: which means minimizing the resources or causes that precede the crisis state. Prevention - which means preparing for action in a crisis state. The need to apply preventive measures, renewal
and revision of emergency, alarm, crisis plans, joint operational procedures, etc. Prevention - which means preparing for action in a crisis state. The need to apply preventive measures, renewal
and revision of emergency, alarm, crisis plans, joint operational procedures, etc. p p
g
pp y p
and revision of emergency, alarm, crisis plans, joint operational procedures, etc. Counter(re)action - active involvement in the emergence and gradation of the crisis state. Crisis staff and crisis manager The crisis staff is a working body of the statutory crisis management body serving their need to coordinate the
intervening components and other elements in the cooperation of the components of the integrated rescue system
in joint intervention and in resolving the crisis state. An inseparable assumption for ensuring the correct and
efficient functioning of the crisis staff is its erudite and well-prepared staff. The crisis manager, as a person, a member of the crisis staff, should be a representative of an effective, efficient
and vigorous response, a representative of reception, analysis and correct evaluation of information and reports, as
well as a representative of rationality, tolerance and complexity, when resolving a specific crisis state. The
members of the crisis staff should form a group of erudite people who are efficient and united, where each
member-element knows what his duties and competencies are. He is also aware that in resolving a crisis state, he
must make quick and especially correct decisions under the stress factor of not only time. These reasons also
represent a constant need for effective and efficient training for crisis managers. Training crisis managers is one of
the essential elements to acquiring the necessary expertise and technical and technological readiness for solving
potential crisis states and their management. (Grega, 2014) The area crisis managers training should be prepared to prevent the occurrence of possible crisis states, ensure
preparation for dealing with potential emergencies, ensure a well-founded solution to troubles within their
competence at individual levels, solve tasks from superior authorities, also implementing measures to eliminate
consequences of crisis states and the distribution of instructions to the executive elements involved in this process,
is at a superficial level. 27 INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) Crisis manager competencies The abovementioned circumstances and facts point to the need for well-prepared staff. Nevertheless, what makes
a crisis manager the right element in this system, what competencies should he have, and what negative factors
should he eliminate when dealing with crises? Decision-making is the most important function of managers at all levels of management, and all managerial
procedures include decision-making. Decision-making, acceptance of a decision represents the principle of choice
in the presence of several options by which a common goal can be achieved. Crisis staff and crisis manager Bad decisions by the crisis staff are
often the result of an inability to consider all the contexts and doubts about the assumptions. The decision results
in a consensus. However, it is acceptable only if it has not been adopted very quickly, given that unanimity does
not prevail in the crisis staff because it suppresses the motivation to evaluate alternatives
realistically The public administration crisis manager must consider the worst-case scenario when dealing with a crisis state. He must be adequately prepared for this state, and whereas he works in the public interest, he is the first to
prevent the spread of panic, and he must be the first to inform the population and the media. The crisis manager
takes preventive measures in the public interest to prevent the gradation of negative phenomena and must be able
to evaluate what is a priority and paramount in a given situation. 3. Simulations The method and methodological procedure of 28 INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) its implementation depend mainly on the type of simulation used and the parameters of the created model at
which the simulation is performed. Simulation has many advantages that encourage its constant development and
expansion into new spheres and areas. Influenced by the development of the information-communication base, its
potential is constantly enhancing. At present, simulation technologies have fully established themselves in the
educational and training environment in the form of: 1. instrumented simulation (also referred to as live simulation), 1. instrumented simulation (also referred to as live simulation), 2. virtual simulation, 3. constructive simulation and 4. distributed simulation, resp. blended simulation Of course, there are many divisions and views on simulations. They can be classified according to approaches,
according to the capacity of primary and secondary training elements (staff, tactical, operational, etc.), according
to calculation methods (deterministic, stochastic, static, or dynamic), or also according to the end-user
determination (military, police, medical, etc.). The simulation generally consists of five main parts (Cayirci,
Marincic, 2009), resp. should consider the following aspects: −
simulation of performances that are performed in the simulation, −
models that provide real representations, −
input data that provide model definition, −
a visualization that shows the results of the simulation, −
communication that serves as a data interface. 3. Simulations People want to be informed and prepared for the possible existence of incoming threats that could affect families,
municipalities and cities, districts, or entire countries in any way. How to protect lives and properties, societies
and their values? These denominators characterized the perception of security and readiness to respond correctly. The level of protection of specific fields, such as demography, the economy, the social and environmental field,
and the political, reflects every society. In an unexpected, unplanned emergency, society must take appropriate
measures to minimize negative phenomena (Cayirci, Marincic, 2009). The forces and means that must be used to
protect population, property and social values are elements of the national security system. These elements must
include all managerial, executive, operational and other components that must have different skills, such as the
ability to communicate, make effective decisions, cooperate, cope with stress load, etc. Maintenance of these
skills requires continuous and regular training using a variety of training methods based on simulation
technologies. (Žentek, Nečas, 2020) Simulations from the beginning, when implemented into the education and training environment, reflected the
need to increase the effectiveness of staff training (from individual to coordinated staff) for events and situations
that are very difficult to implement - to prepare in a natural environment. The benefits resulting from the
implementation of individual types of simulations and their tools point to the growing interest and social need for
their usage in the process of staff training in the field of management coordination and cooperation: −
improving the decision-making and planning process of staff (both civilian and military) in the preparation,
planning and management of operations, −
improving the decision-making and planning process of staff (both civilian and military) in the preparation,
planning and management of operations, −
obtaining alternative and backup solutions, −
risk reduction, −
saving staff capacity and time, −
minimizing adverse effects on the ecological environment, and −
saving costs and materials. −
saving costs and materials. Any simulation of activities in terms of military and non-military operations is carried out on an appropriate,
specific model based on statistical, mathematical, analytical methods, or a combination thereof, and on the
widespread use of modern computer and information technologies. Constructive simulation The main advantages of productive
simulation methods include that (Grega, Bučka, 2013): −
it allows for simulating sources of threat and destructive processes of destruction during a crisis
economically and efficiently, −
it allows the creation of an environment that is very similar to the real conditions of the emergency and the
stages of the crisis, −
it will enable the simulation of a situation taking place in different geographical regions, in different
climatic conditions, with various forces and means, −
constructive simulation models form important components also in other types
management activities; they represent a universal unifying basis for all groups of s −
reduce the damages caused by training from an ecological point of view, and the environment is not
bothered by such training, −
reduce the damages caused by training from an ecological point of view, and
bothered by such training, −
its universality results from the applied mathematical basis, −
quantification of phenomena and processes enables their more accurate analysis in the decision-making
activity of management authorities, −
it allows to model and simulation, at any level of similarity or with any degree of generalization, objects,
phenomena and processes at the selected level (tactical, operational, strategic), −
it allows an objective expression of the influence of the terrain and other environmental factors on the
processes of the crisis, as well as on the activity of crisis management, −
it gives the possibility of quality registration of the course of training and their use for a more objective
evaluation of their results and formulation of conclusions and recommendations for theory and practice, gives the possibility of quality registration of the course of training and their use for a
valuation of their results and formulation of conclusions and recommendations for theory a −
the usability of constructive simulation in problem-solving crisis management is ev −
it mediates the possibility of interconnecting constructive simulation systems wi
other categories, −
sets a measurable factor to determine the results of the operation of individuals and −
it documents the course of solving the situation in 2D and 3D display, −
offers partial and complex statistical data, −
it records the planned and actual activity of the units, −
it allows searching for optimal solutions using forces and resources, −
it will enable crisis managers to take risks due to their own decisions. Constructive simulation Constructive simulation is the most used, in terms of time, the longest applied and the most used simulation, with
a universal utilization and a wide range of uses. It is often referred to as the universal method. “The core of the
constructive simulation is the usage of logical-mathematical models which are expressed usually by equations or
systems of mathematical equations, inequalities (algebraic, differential, integral, etc.) and by algorithms or
systems of closed or opened algorithms, while parameters (constant, variables) and interrelationships of these
systems have deterministic or stochastic interpretation of material-energetic or thought-information objects and
processes of armed encounter or conflict situation.“ (Rybár, 2000). Today, computer technologies and software solutions are fully applied, where the relationships and dependencies
of mathematical-logical interpretation are implemented. It is usually used in a distributed form, where several
computing systems are connected via a computer network, or it can also be applied separately on a specialized
hardware element. Constructive simulation is thoroughly used in training staff responsible for an operation or
situation's planning and decision-making phase. In this way, the staff is trained in individual staff positions and
management functions in the process of command, control, and verification of planning and decision-making
activities. A constructive simulation can also be specified as an artificial entity (model) representative of a man with actual
behaviour, a natural technique, or specific units or wholes. Such a model is defined according to real data
regarding its tactical-technical data, behaviour, and representation. In a constructive simulation, a real object
(being, vehicle, technical device, system, living creature, etc.) is replaced by a model - an entity. It is, therefore, a
simulation where synthetic beings, vehicles, systems, or technical devices move in a virtual environment and,
depending on the simulated activities, also appropriately perform the assigned tasks. (Hubáček, Hausner, Vráb,
2013) Artificial entities behave according to defined algorithms, programmed in separate property classes (entity
behaviour level). Constructive simulation These models - entities based on interactions and defined semi-automatics can be divided into:
−
deterministic - those that contain only precisely determining relationships (algorithms), 29 INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) −
stochastic - those that have at least one relationship with the generated variable; other relationships are
deterministic, −
stochastic - those that have at least one relationship with the generated variable; other relationships are
deterministic, −
an entity with intelligence imitated by the properties of application programs, i.e., Constructive simulation is based on mathematical methods and has many advantages that stimulate its constant
development and expansion into new application areas in education and training. Like any other simulation
method, constructive simulation has certain drawbacks and limitations. 4. Training effectiveness and evaluation The use of constructive simulation for staff and managers training will expand only if its usage is practical
(especially more financially advantageous) compared to other training methods, by which the intervening
components of crisis management are prepared to fulfil their tasks. The exact methodology for calculating the cost
of conducting training by live simulation (training in natural conditions) and for the relevant calculation of the
expenses of constructive simulation does not yet exist. The evaluation mentioned in (1) is focused on the content
of the training and is not applicable to evaluate the financial demands of the training. However, a simplified
model can be developed that can be used for comparative analysis in certain circumstances. The basis of the
training effectiveness calculation model is the premise that in the analytical relationship for cost calculation are
accepted identical variables characterizing a given phenomenon, process, or behaviour from both constructive and
live simulations. (Žentek, Bučka, 2019) The relationship for calculating the costs associated with the Ncv training can be express Ncv = Na + Nb + Nt + Np + Nn,
Wherein:
Na are the costs of using the machinery in training in [€],
Nb are the costs of renting the premises for training in [€],
Nt are the costs of using the technical means for intervention (action) in [€],
Np are the costs of supporting the training in [€],
Nn are (for now) unspecified costs in [€]. Ncv = Na + Nb + Nt + Np + Nn, Ncv = Na + Nb + Nt + Np + Nn,
Wherein:
Na are the costs of using the machinery in training in [€],
Nb are the costs of renting the premises for training in [€],
Nt are the costs of using the technical means for intervention (action) in [€],
Np are the costs of supporting the training in [€],
Nn are (for now) unspecified costs in [€]. cv
a
b
Wherein:
Na are the costs of using the machinery in training in [€],
Nb are the costs of renting the premises for training in [€],
Nt are the costs of using the technical means for intervention (actio
Np are the costs of supporting the training in [€],
Nn are (for now) unspecified costs in [€]. Constructive simulation As mentioned above, constructive simulation is usually implemented in a distributed form. Its basis is the
implementation of the so-called Computer Assisted Exercise (CAX). Such form of training can be characterized
as a sophisticated method of training commanders, managers, officials, and crisis staff members of the state, by
which simulated processes and phenomena are potentially emerging or developing in carrying out the specified
activity of the decision-making process, planning process or command and control process. Computers simulate
processes and phenomena in the real or specified time and environment to give training entities the impression of
performing real operations and activities as if they were performed in a natural setting and actual conditions. Constructive simulations and their computer-assisted exercises are excellent training and education in crisis
management, mainly due to their versatility, broad application support and ease of implementation. 30 INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) Simulation tools and their technologies are currently a global leader in training commanders, staff and
participating units in their subordination and cooperation to solve crisis management tasks. There is no doubt that
a well-prepared crisis staff and its members are the basis for the successful resolution of crises, primarily in
matters of domestic crisis management and in cases of crisis management at the international level. 4. Training effectiveness and evaluation Nt are the costs of using the technical means for intervention (action) in [€], The variable Na represents the costs of using all types of machinery in training (passenger, transport, goods and
special vehicles, aircraft, reconnaissance aircraft, etc.). Its value can be expressed by equation (2). Wherein:
aj is the number of j machinery used in training,
lj is the number of kilometres the j machinery will cover (covered) during the training in [km],
ck are the costs of j machinery per 1 km of driving in [€/km]. Live simulation (training in real terrain) can take place on premises, the use of which is conditioned by payment
for rent. The calculation of its value is based on the price per m2 of the j area (cj) multiplied by the total used j area
(Pj), or the total invoiced value is substituted into equation 2. 31 INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) The Nb calculation can be performed according to equation (3). The Nb calculation can be performed according to equation (3). Wherein:
Pj is the area of the j land used for training in [m2],
cj is the price per 1 m2 of the j land in [€/m2]. Pj is the area of the j land used for training in [m2], cj is the price per 1 m2 of the j land in [€/m2]. The costs of using technical means Nt represents the sum of funds, that must be spent for their use in training. The
category of technical means refers to, for example, power plants for special power supplies and for powering
other means (communication technology, computing technology, etc.). Equation (4) can be used to calculate these
costs. Wherein:
ti is the time of use of the i technical means for training in hours,
ci is the price of using the i technical means (operating costs) per 1 hour [€/hour]. The costs for supporting the Np training in Equation 4 represent the estimated or spent funds needed to support the
training. 4. Training effectiveness and evaluation This group can include, for example, the fees for training figurants, simulation system operators, etc.,
i.e., for people who are not trainers but whose work is necessary for the course of the training. These costs can be
calculated using equation (5). Wherein:
ti is the total time of deployment of i person to support the training in [hours],
ci is the I person's hourly fee in [€/hour]. Wherein: The Nn costs represent unforeseen, unpredictable, additional costs that may arise during training. This category of
costs can include costs for dealing with damage caused to materials, machinery, property of the population, etc.,
in direct and indirect connection with the training. A coefficient k can increase the total cost of the training or its components Nx, wherein k is the coefficient of
tolerance of the population and its value is in the range [k ≥ 1]. The coefficient k is dimensionless, and the value
"1" means that the training is not perceived negatively by the population that does not participate in the training. The population participating in the training is included in the variable Np (for example, in the category of
figurants). If the population does not tolerate the training (because of restrictions, etc.), the coefficient k is greater
than 1. The limit of the coefficient cannot be determined precisely yet, and its value will be estimated. 32 INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) The overall effectiveness of the training is then evaluated as the ratio of the financial costs of live simulation
training and training performed by constructive simulation. The verification of the described model was not
exactly verified in the training conditions. Only an empirical comparison of training by constructive simulation
with the equivalent of exercise performed in natural conditions (e.g., command and staff training in the military
training area) was performed. The calculation showed that the training performed with the support of constructive
simulation reaches the cost of about 10% of the cost of the training performed in real terrain. It can be assumed
that in training on the components of crisis management using constructive simulation, a comparable value of
training efficiency will be achieved. 4. Training effectiveness and evaluation The costs associated with the catering for trainees and support elements, with
their accommodation, the wage costs of trainees were not included in the calculations of the estimates, nor in the
above model, as these costs are in many cases comparable regardless of whether the exercise is performed by
constructive or live simulation. Accurate cost and training efficiency calculations are not yet possible because the
data needed for the analysis are not available in 100% of cases. Training using constructive simulation has been used by the armies of many countries for many years and brings
results in the exercise of commanders and staff. Even in the case of this technology, it is becoming more and more
widespread. In connection with this, its price is gradually falling, and the technology of constructive simulation is
slowly beginning to be used outside the army. It is logical that the training of police and security forces, as well as
rescuers, is another area into which this technology penetrates. Of course, the acquisition costs of simulation tools
are not minor and will initially seem significantly higher than the cost of training in real conditions. Such activity
cannot be replaced even by the perfect simulator, and the training of powerful elements must take place in a real
environment with real equipment. However, only a limited number of people and machinery can participate in
such training. Thus, lower management and performance elements are usually participating. A considerable
complex training, with the participation of hundreds or thousands of intervening persons and dozens of special
machineries, cannot be prepared in real conditions. Therefore, even the highest levels of management do not have
adequate resources with an excellent response to their training. For these levels, the training using constructive
simulation tools is tailor-made. Individual training can be modified; it is possible to prepare various scenarios, move in space and not limit regular
operation. For essential functions, live training remains the most important. Training using constructive
simulation provides new opportunities in an almost real environment and expands training opportunities to an
unprecedented extent. It allows preparing new members who can make mistakes during training without the risk
of life-threatening and causing great material or other losses. The participants of such complex training can thus
prepare to deal with real crises. 4. Training effectiveness and evaluation The simulation can be repeated countless times with different staff, compare
various solutions and learn from each other, gain new information, learn, look for weak points in the plans and
improve before deploying. As in the case of the training of military rescuers, in future, the usage of constructive
simulation can be expected for firefighters and civilian rescuers, including members of the crisis staff. Well-
prepared control components can better coordinate rescue operations and prevent loss of life and material damage. As shown by the experience of floods in recent years, preparedness for crises is one of the primary conditions for
successful interventions in a large area. However, it is better to gather this experience in a simulator than in a
natural disaster at the cost of loss of life, destroyed houses, damaged infrastructure and other damage, especially if
the actual training using the simulation is much cheaper than the same training in terrain. (Žentek, Grega, 2017) Conclusions The system of regional public administration is responsible for the human individual, their potential and the action
of a particular societal group. It is the one which is the bearer of progress and modernization. But despite
optimization procedures and methods, modern methods of work, the individual or the whole is unable to act 33 INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/
2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) effectively in a matter of crises management unless it is qualified, well professionally erudite, technically and
technologically prepared to deal with potential concerns (Andrassy, Grega, Nečas, 2018). The functionality of
public administration in connection with organizational and managerial problems is integrated into the principal
theoretical and practical demands, which are constantly discussed in the past and present. Interventions in this
system, in its organizational structures, tend to be fundamental and substantial because their functional impact has
a broad scope in the life of society in all its areas. (Šimák, 2006). Modelling and simulation, aimed at optimizing activities using experimentation with a created computer model
built into a real synthetic environment, provide an effective and efficient form of preparation and development of
the characteristics and behaviour of individuals and groups. They appear to be a progressive means of training
commanders and staff to find solutions to crises and acquiring habits and skills in individual and particular
activities. Simulation technologies are therefore becoming a significant subject increasing the preparedness of
individuals and groups in the frame of regional public administration, as a key to achieving training priorities
because they can realistically simulate activities and prepare personnel to deal with complicated non-military and
military operations of national as well as domestic crisis management. INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/ INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/ INSIGHTS INTO REGIONAL DEVELOPMENT
ISSN 2669-0195 (online) http://jssidoi.org/jesi/ 2022 Volume 4 Number 4 (December)
http://doi.org/10.9770/IRD.2022.4.4(2) Korecki, Z., Adámková, B. 2018. Implement policies and processes to increase the level of security of the multi-tier logistics chain. In: In:
2018 XIII International Scientific Conference-New Trends in Aviation Development (NTAD). IEEE, 2018. p. 1-6. ISBN 978-153867918-
0. https://doi.org/10.1109/NTAD.2018.8551684 Korecki, Z., Adámková, B. 2018. Implement policies and processes to increase the level of security of the multi-tier logistics chain. In: In:
2018 XIII International Scientific Conference-New Trends in Aviation Development (NTAD). IEEE, 2018. p. 1-6. ISBN 978-153867918-
0. https://doi.org/10.1109/NTAD.2018.8551684 Rak, R., Kopencova, D., Sulc, F., Vlach, F., Hudecova, V. 2022. Crisis Development and its management. Entrepreneurship and
Sustainability Issues, 9(3), 414-428. http://doi.org/10.9770/jesi.2022.9.3(25) Strecková, Y., Malý, I. et col. 1998. Veřejná ekonomie pro školu i praxi. (Public economics for school and practice) Praha: Published by
Computer Press. Šimák, L. 2006. Manažment rizík (Risk management). Žilina: Published by University of Žilina. 2004. Základy krízového manažmentu (Basics of crisis management). Ostrava: Published by SPBI. Šenovský, M., Adamec, V. 2004. Základy krízového manažmentu (Basics of crisis management). Ostrava: P Žentek, M., Bučka, P. 2019. Safety model scenarios in the synthetic air traffic management environment on bases of LETVIS real
information system. In: New Trends in Aviation Development 2019. Institute of Electrical and Electronics Engineers. ISBN 978-1-7281-
4078-0, 218-222. https://doi.org/10.1109/NTAD.2019.8875569 Žentek, M., Grega, M. 2017. Simulation tools in the armed forces interfaced by the operational design. Revista Academiei Fortelor
Terestre. ISSN 1582-6384. 22(2 (86)), 134-140. https://doi.org/10.1515/raft-2017-0019 Žentek, M., Nečas, P. 2020. European air C2 system national requirements. Incas Bulletin. ISSN 2066-8201 12(1), 249-260. https://doi.org/10.13111/2066-8201.2020.12.1.24 Funding: This work was supported by the Slovak Research and Development Agency under the contract No. APVV-20-
0334. References Andrassy, V., Grega, M., Nečas, P. 2018. Crisis management and simulations: scientific monograph. Ostrowiec Świętokrzyski: Published
by Wyższa Szkoła Biznesu i Przedsiębiorczości w Ostrowcu Świętokrzyskim. ISBN 978-83-64557-33-0. Andrassy, V., Grega, M., Nečas, P. 2018. Crisis management and simulations: scientific monograph. Ostrowiec Świętokrzyski: Published
by Wyższa Szkoła Biznesu i Przedsiębiorczości w Ostrowcu Świętokrzyskim. ISBN 978-83-64557-33-0. Bilczak, M. 2021. Crisis diagnosis in cross-border region: lessons from the pandemic. Entrepreneurship and Sustainability Issues, 9(2),
223-237. http://doi.org/10.9770/jesi.2021.9.2(15) Bilczak, M. 2021. Crisis diagnosis in cross-border region: lessons from the pandemic. Entrepreneurship and Sustainability Issues, 9(2),
223-237. http://doi.org/10.9770/jesi.2021.9.2(15) Blažek, V. 2014. Východiská tvorby krízových scenárov pre potreby vzdelávania na APZ v Bratislave (The starting points for the creation
of crisis scenarios for educational needs at APZ in Bratislava). Bratislava: Published by Academy of Police Foces. Cayirci, E., Marincic, D. 2009. Computer assisted exercises and training: A reference guide. Hoboken, (N.J.): Published by John Wiley. https://doi.org/10.1002/9780470498620 Grega, M. 2014. Vzdelávanie krízových manažérov s podporou simulačných technológií (Education of crisis managers with the support of
simulation technologies). In: Bezpečnost, spolehlivost a rizika. Liberec: Technical University. Hašanová, J., Dudor, L. 2014. Základy správneho práva (Basics of administrative law). Plzeň: Published by Aleš Čeňek. Hubaček, M., Hausner, D., Vráb, V. 2013. The Use of Simulation Technologies in the Preparation for New Types of Operations. Vojenské
Rozhledy, 22. https://doi.org/10.3849/2336-2995.22.2013.01.149-159 Chase, W. H. 1984. Issue Management - Origins of the Future. Stamford, CT: Published by Issue Action Publications. Ivančík, R. 2019. Quo Vadis európska obrana a bezpečnos (Quo Vadis European defense and security). Politické Vedy, 22(3), 47-67. ISSN
1335-2741. https://doi.org/10.24040/politickevedy.2019.22.3.47-67 Ivančík, R. 2021. Informačná vojna - jeden z multidisciplinárnych fenoménov súčasnej ľudskej spoločnosti (Information war - one of the
multidisciplinary
phenomena
of
contemporary
human
society). Politické
Vedy,
24(1),
135-152. ISSN
1335-2741. https://doi.org/10.24040/politickevedy.2021.24.1.135-152 Jaques, T. (2007). Issue management and crisis management: An integrated, non-linear, relational construct. Public Relations Review, 33,
147-157. https://doi.org/10.1016/j.pubrev.2007.02.001 Korecki, Z., Adámková, B. 2020. Multi-criteria evaluation of critical infrastructure resilience and economic impacts in selected airports in
the
Czech
Republic. Zeszyty
Naukowe. Transport/Politechnika
Śląska. ISSN
0209-3324. 108,
95-+. https://doi.org/10.20858/sjsutst.2020.108.9 34 34 Copyright © 2022 by author(s) and VsI Entrepreneurship and Sustainability Center
This work is licensed under the Creative Commons Attribution International License (CC BY).
http://creativecommons.org/licenses/by/4.0/ Author Contributions: The authors contributed qually. Matúš GREGA, Dipl.Eng, PhD. is the researcher in Simulation Centre in Armed Forces Academy in Liptovský Mikuláš,
where he deals with the issue of simulation tools and their application to the environment of education and training of crisis
management personnel. He is the author and co-author of many scientific and professional monographs, articles, scripts and
studies dealing with modeling and simulation, security and crisis management. He is the author of many research reports of
national and international science and research projects. He was 6 years active in the Science and Technology Organization
of NATO in the Modeling and Simulation Group. g
p
ORCID ID: https://orcid.org/0000-0003-1065-2355 Pavel NEČAS, Dr.h.c. Prof. Dipl. Eng. PhD., MBA is professor at the Department of Security Studies and the Vice-Dean at
the Faculty of Political Science and International Relations of Matej Bel University in Banska Bystrica. His research is
focused on political-military strategic issues, particularly in defence and security policy, international relations, and
international security studies. HE is an experienced European researcher and the author of several monographs, books, papers
and articles related to the political sciences, security and defence published worldwide. O C
5 Copyright © 2022 by author(s) and VsI Entrepreneurship and Sustainability Center
This work is licensed under the Creative Commons Attribution International License (CC BY). http://creativecommons.org/licenses/by/4.0/ 35
|
https://openalex.org/W4309360828
|
https://www.nature.com/articles/s41380-022-01857-4.pdf
|
English
| null |
Rare tandem repeat expansions associate with genes involved in synaptic and neuronal signaling functions in schizophrenia
|
Molecular psychiatry
| 2,022
|
cc-by
| 8,681
|
Received: 22 February 2022 Revised: 14 October 2022 Accepted: 24 October 2022
1Department of Genetics, University of North Carolina, Chapel Hill, NC 27599, USA. 2Genetics and Genome Biology, The Hospital for Sick Children, Toronto, ON M5G 1X8, Canada.
3The Centre for Applied Genomics, The Hospital for Sick Children, Toronto, ON M5G 1X8, Canada. 4Department of Psychiatry, University of Florida College of Medicine, Gainesville,
FL 32610, USA. 5Department of Biostatistics, University of North Carolina, Chapel Hill, NC 27599, USA. 6Department of Medical Epidemiology and Biostatistics, Karolinska Institutet,
Stockholm 17177, Sweden. 7Department of Immunology, Genetics and Pathology, Science for Life Laboratory, Uppsala University, Uppsala 75185, Sweden. 8Clinical Genetics
Research Program and Campbell Family Mental Health Research Institute, Centre for Addiction and Mental Health, Toronto, ON M6J 1H4, Canada. 9The Dalglish Family 22q Clinic
for Adults with 22q11.2 Deletion Syndrome, Toronto General Hospital, and Toronto General Hospital Research Institute, University Health Network, Toronto, ON M5G 2C4,
Canada. 10Department of Psychiatry, University of Toronto, Toronto, ON M5S 1A8, Canada. 11Department of Psychiatry, University of North Carolina, Chapel Hill, NC 27599, USA.
12Department of Molecular Genetics, University of Toronto, Toronto, ON M5S 1A8, Canada. 13These authors contributed equally: Jia Wen, Brett Trost, Worrawat Engchuan. 14These
authors jointly supervised this work: Ryan K. C. Yuen, Jin P. Szatkiewicz. ✉email: ryan.yuen@sickkids.ca; jin_szatkiewicz@med.unc.edu IMMEDIATE COMMUNICATION
OPEN
Rare tandem repeat expansions associate with genes involved
in synaptic and neuronal signaling functions in schizophrenia
Jia Wen1,13, Brett Trost
2,3,13, Worrawat Engchuan2,3,13, Matthew Halvorsen1, Linda M. Pallotto2, Aleksandra Mitina2,
NaEshia Ancalade1, Martilias Farrell
1, Ian Backstrom
2, Keyi Guo2, Giovanna Pellecchia
2,3, Bhooma Thiruvahindrapuram
2,3,
Paola Giusti-Rodriguez4, Jonathan David Rosen1, Yun Li1,5, Hyejung Won
1, Patrik K. E. Magnusson
6, Ulf Gyllensten7,
Anne S. Bassett
8,9,10, Christina M. Hultman6, Patrick F. Sullivan1,6,11, Ryan K. C. Yuen
2,12,14✉and Jin P. Szatkiewicz
1,11,14✉ Molecular Psychiatry (2023) 28:475–482; https://doi.org/10.1038/s41380-022-01857-4 Molecular Psychiatry (2023) 28:475–482; https://doi.org/10.1038/s41380-022-01857-4 Received: 22 February 2022 Revised: 14 October 2022 Accepted: 24 October 2022
Published online: 16 November 2022 Molecular Psychiatry Molecular Psychiatry Molecular Psychiatry www.nature.com/mp IMMEDIATE COMMUNICATION
OPEN
Rare tandem repeat expansions associate with genes involved
in synaptic and neuronal signaling functions in schizophrenia
Jia Wen1,13, Brett Trost
2,3,13, Worrawat Engchuan2,3,13, Matthew Halvorsen1, Linda M. Pallotto2, Aleksandra Mitina2,
NaEshia Ancalade1, Martilias Farrell
1, Ian Backstrom
2, Keyi Guo2, Giovanna Pellecchia
2,3, Bhooma Thiruvahindrapuram
2,3,
Paola Giusti-Rodriguez4, Jonathan David Rosen1, Yun Li1,5, Hyejung Won
1, Patrik K. E. Magnusson
6, Ulf Gyllensten7,
Anne S. Bassett
8,9,10, Christina M. Hultman6, Patrick F. Sullivan1,6,11, Ryan K. C. Yuen
2,12,14✉and Jin P. Szatkiewicz
1,11,14✉ Jia Wen1,13, Brett Trost
2,3,13, Worrawat Engchuan2,3,13, Matthew Halvorsen1, Linda M. Pallotto2, Aleksandra Mitina2,
NaEshia Ancalade1, Martilias Farrell
1, Ian Backstrom
2, Keyi Guo2, Giovanna Pellecchia
2,3, Bhooma Thiruvahindrapuram
2,3,
Paola Giusti-Rodriguez4, Jonathan David Rosen1, Yun Li1,5, Hyejung Won
1, Patrik K. E. Magnusson
6, Ulf Gyllensten7,
Anne S. Bassett
8,9,10, Christina M. Hultman6, Patrick F. Sullivan1,6,11, Ryan K. C. Yuen
2,12,14✉and Jin P. Szatkiewicz
1,11,14✉ Jia Wen1,13, Brett Trost
2,3,13, Worrawat Engchuan2,3,13, Matthew Halvorsen1, Linda M. Pallotto2, Aleksandra Mitina2,
NaEshia Ancalade1, Martilias Farrell
1, Ian Backstrom
2, Keyi Guo2, Giovanna Pellecchia
2,3, Bhooma Thiruvahindrapuram
2,3,
Paola Giusti-Rodriguez4, Jonathan David Rosen1, Yun Li1,5, Hyejung Won
1, Patrik K. E. Magnusson
6, Ulf Gyllensten7,
Anne S. Bassett
8,9,10, Christina M. Hultman6, Patrick F. Sullivan1,6,11, Ryan K. C. Yuen
2,12,14✉and Jin P. Szatkiewicz
1,11,14✉ Jia Wen1,13, Brett Trost
2,3,13, Worrawat Engchuan2,3,13, Matthew Halvorsen1, Linda M. Pallotto2, Aleksandra Mitina2,
NaEshia Ancalade1, Martilias Farrell
1, Ian Backstrom
2, Keyi Guo2, Giovanna Pellecchia
2,3, Bhooma Thiruvahindrapuram
2,3,
Paola Giusti-Rodriguez4, Jonathan David Rosen1, Yun Li1,5, Hyejung Won
1, Patrik K. E. Magnusson
6, Ulf Gyllensten7,
Anne S. Bassett
8,9,10, Christina M. Hultman6, Patrick F. Sullivan1,6,11, Ryan K. C. Yuen
2,12,14✉and Jin P. Szatkiewicz
1,11,14✉ © The Author(s) 2022 © The Author(s) 2022 Tandem repeat expansions (TREs) are associated with over 60 monogenic disorders and have recently been implicated in complex
disorders such as cancer and autism spectrum disorder. The role of TREs in schizophrenia is now emerging. In this study, we have
performed a genome-wide investigation of TREs in schizophrenia. Using genome sequence data from 1154 Swedish schizophrenia
cases and 934 ancestry-matched population controls, we have detected genome-wide rare (<0.1% population frequency) TREs that
have motifs with a length of 2–20 base pairs. We find that the proportion of individuals carrying rare TREs is significantly higher in
the schizophrenia group. There is a significantly higher burden of rare TREs in schizophrenia cases than in controls in genic regions,
particularly in postsynaptic genes, in genes overlapping brain expression quantitative trait loci, and in brain-expressed genes that
are differentially expressed between schizophrenia cases and controls. We demonstrate that TRE-associated genes are more
constrained and primarily impact synaptic and neuronal signaling functions. These results have been replicated in an independent
Canadian sample that consisted of 252 schizophrenia cases of European ancestry and 222 ancestry-matched controls. Our results
support the involvement of rare TREs in schizophrenia etiology. INTRODUCTION a major source of genetic variation in the human genome [9, 10]. The repetitive sequence in a tandem repeat can cause DNA
slippage during DNA replication or repair, leading to increases in
repeat
size
across
generations. Pathogenic
tandem
repeat
expansions (TREs) are currently known to cause over 60 disorders,
most of which affect the central nervous system [7, 11–13]. TREs
can alter both coding and non-coding regions of genes and exert
harmful effects via a variety of pathophysiological mechanisms
(reviewed in [7, 12]). TREs in the coding sequence can result in
abnormally long stretches of polyglutamine or polyalanine,
leading to protein misfolding and aggregation (e.g., expanded
CAG repeats in HTT in Huntington’s disease). TREs in the
noncoding sequence can occur in 5’ or 3’ untranslated regions
(UTRs), introns, promoters, or enhancers. The impact of noncoding
TREs depends on the type, length, and locations of the repeats,
which may include epigenetic gene silencing, RNA toxicity
mediated by protein titration, repeat-associated non-AUG transla-
tion,
modulation
of
enhancer
activity,
and
disruptions
of
boundaries demarcating 3D chromatin domains [7, 12–15]. Schizophrenia is a chronic and debilitating mental disorder. Twin,
family, and adoption studies consistently support a genetic basis
for schizophrenia, with estimates of heritability in the range of 60-
65% from pedigree data, and around 81% from twin data [1–3]. Tremendous progress has been made toward the identification of
genetic variants that confer schizophrenia risk, including 270 loci
harboring common variants, 8 large rare copy number variants,
and 10 genes implicated from exome sequencing studies of rare
coding variants [4–6]. Nonetheless, the genetic variants identified
thus far for schizophrenia confer less risk than its heritability
estimates [4–6]. Tandem repeats are a plausible source for some of
this missing heritability [4, 7, 8]; however, until recently tandem
repeats were difficult to interrogate due to the technical difficulty
of resolving complex variants in repetitive regions from short-read
sequencing [4] and in genotyping/imputing tandem repeats using
biallelic SNP arrays. y
Tandem repeats occur in DNA where sequences of one or more
nucleotides are repeated directly adjacent to each other. They are J. Wen et al. 476 Due to their repetitiveness and abundance in the human
genome, tandem repeats are challenging to study at a genome-
wide scale [7]. Recent studies by Trost et al. and Mitra et al. Statistical analyses y
A detailed description of quality control, genome annotations and
statistical analyses is provided in Supplementary Methods. Genome-wide detection of tandem repeats, TREs, and rare
TREs We used ExpansionHunter Denovo (EHdn) [28], an efficient catalog-free
method for genome-wide tandem repeat detection from short-read WGS
data. We applied the density-based spatial clustering of applications with
noise (DBSCAN) algorithm to identify TREs whose lengths were outliers
compared with other members of the cohort [16, 29]. We defined rare TREs
as TREs that were found in less than 0.1% of the 1000GP population
controls. All genomic coordinates are given in NCBI Build 38/UCSC hg38. A
detailed description of the detection methods is provided in Supplemen-
tary Methods. p
In this study, we carry out a genome-wide investigation in the
role of rare TREs in schizophrenia etiology. We first apply a
genome-wide TRE detection pipeline [16] to identify TREs from
WGS data in a Swedish sample of schizophrenia cases and
controls. We then assess the possible functional effects of these
variants by comparing burdens of rare TREs between cases and
controls in genic and intergenic regions, in different parts of
genes, in gene-sets previously identified to increase risk for
schizophrenia or neurodevelopmental disorders, as well as in
conserved sequences and epigenomic annotations empirically
derived from the human brain. Finally, we replicate significant
associations in an independent Canadian cohort [18]. Our results
suggest that rare TREs collectively contribute to the genetic risk of
schizophrenia. MATERIALS AND METHODS
Subject recruitment and ethics approval We have complied with all relevant ethical regulations. The study protocol
and all procedures on data from human research subjects were approved
by the appropriate ethical committees in Sweden and the United States
(Karolinska Institutet [Regionala Etikprövningsnämnden, Stockholm], Uni-
versity of Uppsala [Regionala Etikprövningsnämnden, Uppsala], and
University of North Carolina Institutional Review Boards). All participants
gave written informed consent. INTRODUCTION developed computational methods to evaluate genome-wide
tandem repeats in autism spectrum disorder (ASD) using data
generated from short-read whole genome sequencing (WGS), and
together they provide compelling evidence that TREs contribute
to ASD susceptibility [16, 17]. For schizophrenia, genome-wide
evaluations of tandem repeats are now emerging [18]. Published
WGS studies examined TREs in a number of specific loci known to
be associated with monogenic neurological diseases and found
several of these variants in schizophrenia patients [19, 20]. In an
attempt to identify causative variants underlying CACNA1C, one of
the loci being replicated in multiple common variant associations
for schizophrenia and bipolar disorder, Song et al. found that an
intronic tandem repeat was associated with a higher risk of
developing psychiatric disorders and decreased enhancer activity
[21]. Most recently, Mojarad et al applied methods developed in
ASD studies and evaluated genome-wide tandem repeats in
252 schizophrenia cases and 222 controls [18]. Despite their
relatively small sample size, the results from Mojarad et al suggest
that rare TREs are an important class of variants contributing to
the etiology of schizophrenia. These examples, as well as the
known genetic overlap between schizophrenia and ASD [22, 23],
highlight the need for further systematic investigations of TREs in
schizophrenia. Whole genome sequencing data g
q
g
Individuals from the schizophrenia case-control cohort were previously
sequenced by our group at the National Genomics Infrastructure platform
in Sweden [19]. DNA was extracted from blood. DNA libraries were
prepared from ~1 μg DNA using Illumina TruSeq PCR-free DNA sample
preparation kits targeting an insert size of 350 bp in accordance with the
manufacturer’s instructions. Libraries were sequenced on the Illumina
HiSeq X platform using 2 ×150 base pair (bp) cycles (2 × 150 bp paired-end
reads) to a target depth of 30x (minimum 21x, median 37x). The 1000GP
sample collection was sequenced by The New York Genome Center [27]. DNA was extracted from lymphoblastoid cell lines. DNA libraries were
prepared from 1 μg DNA using the Illumina TruSeq DNA PCR-free (450 bp)
Library Preparation Kit in accordance with the manufacturer’s instructions. Libraries were sequenced on an Illumina NovaSeq 6000 sequencer using
2 ×150 bp cycles to a target depth of 30x (minimum 27x, mean 34x). Gel electrophoresis p
Selected tandem repeats predicted by EHdn were validated with gel
electrophoresis size separation of PCR products of the regions of interest. The following coordinates (hg38) were amplified using the reagents,
primers. and conditions specified. Takara PrimeSTAR GXL DNA Polymerase
(GXL) and Qiagen HotStarTaq DNA Polymerase (HotStar) kits were used. General thermocycling conditions are as follows: GXL - 1 min 98 °C, 37x
(10 s 98 °C, 15 s variable, 1 min 30 s 68 °C), 10 min 68 °C. HotStar – 15 min
95 °C, 37x (30 s 95 °C, 30 s variable, 1 min 15 s 72 °C), 10 min 72 °C. Target
specific information: PDIA5 chr3:123151603-123152369, F-GCCTTCATAGC
AGACATAAGCC, R - TCTGCCAGAGGTTGAGTCAC, HotStar with Q solution,
Anneal 63 °C; GABRA1 chr5:161663263- 161663867, F - GCAAGAAAGGGGA
GTTACCG, R - CCTAACACCTCATGCTGTACC, GXL, Anneal 60 °C. The sample
NA12878 available from Coriell was used as the reference control. A 100 bp
ladder was used for size reference (FroggaBio 100 bp). Replication We obtained replication association results from an independent dataset
from Canada that included 252 unrelated adult cases with schizophrenia of
European ancestry and 222 ancestry-matched individuals with no major
neuropsychiatric disorders (after removal of 8 samples being outliers of
genome-wide tandem repeat count) [18]. This dataset is well-suited for
replication because the sequencing technology and TRE detection
methods used for the replication samples were identical to those used
for our Swedish case-control samples. Replication was attempted for all
significant association results in the Swedish case-control comparisons. For
each attempted region, we performed burden testing of rare TREs in cases
versus controls in the replication samples and obtained association
summary statistics. We then used METAL [30] to perform a fixed-effect
meta-analysis using the inverse-variance-based method to merge the
findings between the original and the replication studies. Details of the
replication samples, TRE detection, quality control and statistical analyses
are documented in Supplementary Methods. g
The primary WGS dataset used in this study consists of 1159 Swedish
schizophrenia
cases
and
936
ancestry-matched
population
control
individuals [19]. Full descriptions of the cohort are available elsewhere
[19] and are briefly summarized here. The schizophrenia cases were
selected from the Swedish Schizophrenia Study [24] to have typical
Swedish ancestry, unequivocal schizophrenia case status (>8 inpatient or
outpatient psychiatric treatment contacts for schizophrenia or schizoaffec-
tive disorder, ≥30 inpatient days for schizophrenia, ≥5 redeemed
prescriptions for antipsychotics, and few or no treatment contacts for
bipolar disorder), and without any known pathogenic CNVs (e.g., 22q11.2
deletion). Carriers of known pathogenic CNVs were previously identified
from SNP array genotyping [25] and were not included in this study
because here we aim to evaluate the contribution of novel loci to
schizophrenia risk. Controls were group matched to cases by ancestry and
were selected from the SweGen project (unrelated individuals originating
from the Swedish Twin Registry [26]). We also included WGS data derived
from 2504 unrelated samples from the phase three panel of the 1000
Genomes Project (1000GP [27]). The 1000GP cohort included individuals
from 26 populations, representing five continental regions of the world. Contribution of rare tandem repeat expansions in
schizophrenia p
We defined a tandem repeat-containing region as a genomic
location containing tandem repeats with one or more different
motifs overlapping by at least 1 bp. Across all 4592 samples, we
identified 21,153 unique tandem repeat motifs in 16,723 distinct
regions (Supplementary Table 1). The technical characteristics of
our tandem repeat data, including the distributions of motif, motif
size, and tandem-repeat-containing regions, were consistent with
those reported previously [16, 18] (Supplementary Figs. 2–4). Tandem-repeat-containing regions were enriched in GC-rich
regions (odds ratio [OR] = 1.04, P < 2.2e-16) but depleted in
conserved DNA sequences defined by phyloP [31] (OR = 0.015,
P < 2.2e-16) and phastCons [32] (OR = 0.208, P < 2.2e-16, Supple-
mentary Table 2, Supplementary Fig. 5). We compared our data to
the known simple sequence repeat regions in the human
reference genome and to tandem repeat loci reported in ASD
[16]. Of the 16,723 tandem-repeat-containing regions reported
here, 3447 (20.6%) of them have not been previously reported
(Supplementary Fig. 6). p
To assess the possible functional effects of rare TREs, we used
burden testing to evaluate whether rare TREs are enriched in
different genomic annotations in schizophrenia cases versus
controls. Only autosomal TREs were retained for burden analysis. Our power calculation suggested that we had ≥80% power to
detect
association
signals
with
burden
testing
when
the
aggregated minor allele frequency was 0.01 (i.e. aggregated
minor allele count of 20), the genotypic relative risk was ≥4.9, and
assuming a type I error level of 1×10−5 (Supplementary Fig. 8). Burden testing was performed using logistic regression models
that allowed us to correct for confounding factors that may cause
spurious association signals as described in Supplementary
Methods. To identify potential confounding variables, we carried
out a principal component (PC) analysis of the normalized
anchored in-repeat-read counts which suggested the inclusion
of PC2, PC3, and PC8 as covariates in the logistic regression
models (Supplementary Fig. 9 and Supplementary Table 4). After
correcting for these PCs along with sex, we did not find evidence
of inflation based on the estimated effect measured by the burden
of rare TREs in intergenic regions (see below Genome-wide
burden), which is consistent with a prior report [16]. Furthermore,
for burden testing in target regions (i.e. RESULTS We used WGS data from 1159 schizophrenia cases and 936
ancestry-matched population controls from Sweden that were
previously generated by our group [19] (Fig. 1). To estimate
population frequency of tandem repeats, we included WGS data Molecular Psychiatry (2023) 28:475 – 482 J. Wen et al. Fig. 1
Study overview. This flowchart summarizes the study design and analytic workflow. J. Wen et al. 477 Fig. 1
Study overview. This flowchart summarizes the study design and analytic workflow. t summarizes the study design and analytic workflow. for 2504 genomes sequenced by the 1000 GP [27]. All WGS data
were sequenced on Illumina platforms using 150 bp paired-end
reads to a similar mean coverage per sample. for 2504 genomes sequenced by the 1000 GP [27]. All WGS data
were sequenced on Illumina platforms using 150 bp paired-end
reads to a similar mean coverage per sample. algorithm, to tandem repeat calls across the 4592 post-QC
samples. Outliers for repeat length of each tandem repeat motif
were deemed to be TREs. A total of 2890 TREs were identified, and
1559 (53.9%) of those were novel in comparison to the TREs
reported in Trost et al. [16]. Identification of novel tandem repeats expanded in
schizophrenia We deemed TREs in the schizophrenia case-control cohort to be
rare when found in less than 0.1% of the 1000 GP samples. This
resulted
in
603
rare
TREs
for
subsequent
burden
testing
(Supplementary Table 3). We examined the distribution of the
count of rare TREs per sample using a stratified histogram
(Supplementary Fig. 7). We did not observe any samples that were
outliers based on rare TRE count (Supplementary Fig. 7). Using EHdn [28] we performed genome-wide detection of tandem
repeats that have motifs between 2 and 20 bp and total length
greater than the read length (150 bp) across all the samples. Seven
samples (5 schizophrenia cases and 2 controls) were removed
because they were outliers in terms of genome-wide tandem
repeat count (Supplementary Fig. 1), resulting in 4592 genomes
for subsequent analyses (1154 schizophrenia cases, 934 controls,
and 2504 samples from 1000GP). Molecular Psychiatry (2023) 28:475 – 482 Contribution of rare tandem repeat expansions in
schizophrenia We observe a higher burden of rare
TREs in sub-TAD boundaries – level 1 (OR = 4.977, P = 0.012), sub-
TAD boundaries – level 2 (OR = 4.476, P = 0.022), and enhancer-
promoter anchors (OR = 1.506, P = 0.026) in schizophrenia cases,
but they were not statistically significant after multiple testing
correction (BH-corrected P = 0.105, Supplementary Table 8). Fig. 2
Genome-wide burden of rare TREs in schizophrenia. The
error bars denote 95% confidence interval. regression models to correct for any confounding factors that may
have not been accounted for by the PC2, PC3, and PC8, and to
increase the specificity of the tests in target regions. Genome-wide burden. We first compared the total number of rare
TREs in schizophrenia cases versus controls in three ways:
genome-wide, in genic regions only, and in intergenic regions
only. When a rare TRE overlapped with any gene/transcript by at
least 1 bp, it was defined as genic; otherwise, it was intergenic. We
observed a significantly increased burden in schizophrenia for rare
TREs (OR = 1.403, P = 0.004) and genic rare TREs (OR = 1.549,
P = 0.003; Fig. 2). No statistically significant difference between
cases and controls was detected for rare TREs in intergenic regions
(OR = 1.182, P = 0.232; Fig. 2). To understand how repeat motifs
may influence case-control burden comparison, we stratified genic
TREs into CpG-containing (any rare genic TRE that contains the
sequence of CG, GC, CGC, CGG, CCG, GCC, GGC or GCG) or non-
CpG-containing, and performed burden testing for each group
separately. There were 249 CpG-containing rare genic TREs and
694 non-CpG-containing rare genic TREs across 1,154 schizophre-
nia and 934 control samples. The burdens of both CpG-containing
and non-CpG-containing genic rare TREs were significantly
increased in schizophrenia cases compared to controls. (Supple-
mentary Table 5). Burden in gene sets previously implicated in schizophrenia and
neurodevelopmental
disorders. In
fragile
X
syndrome,
both
abnormally expanded CGG repeats and point mutations in FMR1
have been
reported [39]. Motivated by this example,
we
hypothesized that schizophrenia-associated TREs may be enriched
in genes known to increase risk for schizophrenia, previously
identified via common variant association [5], copy number
variation [4], exome sequencing [6], or gene expression studies
[40, 41]. Contribution of rare tandem repeat expansions in
schizophrenia global genic regions, gene
parts,
gene
sets,
epigenomic
annotations),
we
additionally
included global intergenic burden as a covariate in the logistic Pathogenic TREs are typically significantly longer than what is
observed in the general population. For example, patients with
fragile X syndrome typically carry >200 CGG repeats in the 5’ UTR
of FMR1, while unaffected individuals generally have 6 to 53
repeats. Following Trost et al. [16], we defined a TRE as a tandem
repeat that is much larger than most other members of the study
cohort. We applied DBSCAN [29], a non-parametric clustering Molecular Psychiatry (2023) 28:475 – 482 J. Wen et al. 478 Fig. 2
Genome-wide burden of rare TREs in schizophrenia. The
error bars denote 95% confidence interval. 8 Burden in different parts of genes. Previous studies found that, in
ASD, rare exonic TREs and rare TREs affecting splicing were
enriched, while de novo tandem repeat mutations were enriched
in brain regulatory regions [16, 17]. Motivated by these examples,
we first compared the total number of rare TREs in schizophrenia
cases versus controls in different parts of protein coding genes
(Fig. 3 and Supplementary Table 7). Interestingly, we found a
higher burden of rare intronic (OR = 1.436, P = 0.030) and rare
splicing TREs (OR = 2.174, P = 0.024), although the excess was not
statistically significant after multiple testing correction (BH-
corrected P = 0.105 for both). Burden in brain epigenomic annotations. We then compared the
total number of rare TREs in schizophrenia cases versus controls
within functional annotations experimentally derived from human
brain tissue known to affect gene expression (Supplementary
Table 8). These annotations include open chromatin regions from
ATAC-seq [33], chromatin binding factor CTCF from ENCODE [34],
boundaries of topologically associating domains (TADs) [35]
including level 1 sub-TAD boundaries and level 2 sub-TAD
boundaries, differential neuronal cell specific histone modifica-
tions (H3K27ac and H3K4me3) peaks [36], neuronal frequently
interacting regions (FIRE, superFIRE) [37], and neuronal chromatin
interactions [37]. The bin size was 40 kb for sub-TADs and FIREs,
10 kb for enhancer-promoter annotations, and for the remaining
annotations ranged from 0.126 kb to 880 kb. We leveraged
neuronal annotations when available as previous work has
indicated neurons as the central cell type harboring genetic risk
for schizophrenia [5, 36–38]. Contribution of rare tandem repeat expansions in
schizophrenia Given the known genetic overlap between schizophrenia
and ASD [22, 23], we also included risk genes previously
implicated in neurodevelopmental disorders via exome sequen-
cing [42–44], copy number variation [45–47] or tandem repeats
studies [6]. Burden testing compared the total number of rare TREs
in cases versus controls within each of the 21 gene sets
considered (Methods, Supplementary Table 9). We found an
excess of rare TREs in schizophrenia cases in brain-expressed
genes that are differentially expressed (DEGs) between schizo-
phrenia and controls as determined by the Common Mind
Consortium (i.e. CMC brain DEGs; OR = 6.63, P = 0.005, BH-
corrected P = 0.063), the genes with expression quantitative trait
loci (eQTLs) in human brain as identified by PsychENCODE
Integrative
Analysis
(OR = 1.73,
P = 2.14e-3,
BH-corrected
P = 0.063), as well as in the SynGO ontology category postsynapse
process (OR = 27.94, P = 0.004, BH-corrected P = 0.063). These are
further supported by the fact that many of the CMC brain DEGs
with rare TREs and the genes with brain eQTLs overlapping with
CMC brain DEGs are highly connected with other known
schizophrenia genes and they are involved in similar functions
in synaptic or neuronal signaling (Fig. 3). While examining specific
genes within the three significant gene sets, we found that the
rare TREs mostly affected introns (Supplementary Table 10). Given our finding that tandem repeats were depleted in
conserved DNA sequences, we next compared the total conserved
base pairs affected by rare TREs in cases versus controls and
observed a modestly elevated burden in schizophrenia (OR =
1.005, P = 0.021; Supplementary Table 6). We further performed
burden analysis for coding conserved base pairs and non-coding
conserved base pairs separately and found that the elevated
burden is significantly contributed by conserved base pairs in non-
coding regions (Supplementary Table 6). We estimated the proportion of samples carrying rare TREs
using the residuals of rare TRE counts after controlling for
confounding factors. Using this approach, the estimated sample
proportions were 11.35% in schizophrenia cases and 4.39% in
controls, i.e., a 6.96% excess in schizophrenia (Wilcoxon ranked
sum test P = 8.83e-9). To further explore the potential mechanisms by which TREs may
regulate the underlying genes, we compared the level of Molecular Psychiatry (2023) 28:475 – 482 J. Wen et al. 479 Fig. 3
Network of genes with rare TREs that are differentially expressed in schizophrenia. Fig. 3
Network of genes with rare TREs that are differentially expressed in schizophrenia. Interactions between genes were extracted from
GeneMANIA. Top 50 scoring connected genes were pulled from pathways or physical interactions from the network data with automatic
weighting. Node color: gray for known schizophrenia genes (defined in Methods), red for eQTL genes with rare TREs that are differentially
expressed in schizophrenia (CMC genes with eQTLs and rare TREs); and blue for additional connected genes (top 50 genes) pulled from
GeneMANIA. Node size is proportional to the strength of predictions for a given gene function. Fig. 3
Network of genes with rare TREs that are differentially expressed in schizophrenia. Interactions between genes were extracted from
GeneMANIA. Top 50 scoring connected genes were pulled from pathways or physical interactions from the network data with automatic
weighting. Node color: gray for known schizophrenia genes (defined in Methods), red for eQTL genes with rare TREs that are differentially
expressed in schizophrenia (CMC genes with eQTLs and rare TREs); and blue for additional connected genes (top 50 genes) pulled from
GeneMANIA. Node size is proportional to the strength of predictions for a given gene function. mutational
constraints
(determined
by
the
observed
over
expected number of loss of function variants in gnomAD [48])
within CMC DEGs, within brain eQTL genes and in all genes with
and without rare TREs found in schizophrenia cases. We found
that rare TREs impacting CMC brain DEGs were significantly more
constrained than other CMC brain DEGs (Fig. 4a) and rare TREs
impacting genes with eQTLs are significantly more constrained
than other CMC brain DEGs (Fig. 4b). We also found that the genes
with rare TREs found in schizophrenia cases were significantly
more constrained than genes without rare TREs found in
schizophrenia cases (Fig. 4c). Replication We sought replications for significant associations detected in our
Swedish data in an independent dataset from Canada that
consisted of 252 schizophrenia cases of European ancestry and
222 ancestry-matched controls [18]. We excluded post-synapse
genes from final testing because the number of rare TREs
observed in this gene-set in the replication samples was too
small (<5). For each of the remaining significant associations, we
conducted an association analysis in the replication samples and
then a meta-analysis of the Swedish and independent Canadian
samples for the total sample size of 1,436 schizophrenia cases and
1,156 controls. As shown in Supplementary Table 12, there is a
high concordance between discovery and replication samples and
all significant associations detected in our discovery samples have
been replicated. Our study is underpowered to detect individual loci of rare TREs
associated
with
schizophrenia
at
genome-wide
significance
(Supplementary Fig. 8). We compared the individual rare TRE loci
from our data to the 39 top-ranking rare TRE loci identified by
Mojarad et al. [18]. Nine of the 39 loci were found in the
schizophrenia cases from this study (Supplementary Table 11). To
explore whether pleiotropic TRE loci may exist across ASD and
schizophrenia, we compared the individual rare TRE loci from our
data to the top 57 rare TRE loci identified by Trost et al that were
in the top two gene sets enriched in ASD and had a higher
frequency in individuals with ASD than siblings without ASD [16]. 14 of the 57 ASD loci were found in the schizophrenia cases from
this study (Supplementary Table 11). DISCUSSION
W
h
i We have carried out one of the first genome-wide investigations
of rare TREs in schizophrenia. We found a significant excess of rare
TREs in schizophrenia in genic regions, in genes with brain eQTLs,
in brain-expressed genes that are differentially expressed between
schizophrenia cases and controls, and in genes involved in
postsynaptic processes. These results have been replicated in
independent samples [18]. Our results suggest that rare TREs
collectively play a role in schizophrenia etiology. Contribution of rare tandem repeat expansions in
schizophrenia Interactions between genes were extracted from
GeneMANIA. Top 50 scoring connected genes were pulled from pathways or physical interactions from the network data with automatic
weighting. Node color: gray for known schizophrenia genes (defined in Methods), red for eQTL genes with rare TREs that are differentially
expressed in schizophrenia (CMC genes with eQTLs and rare TREs); and blue for additional connected genes (top 50 genes) pulled from
GeneMANIA. Node size is proportional to the strength of predictions for a given gene function. 4 Confirmation of EHdn detected tandem repeats Confirmation of EHdn detected tandem repeats
The detected tandem repeats from EHdn were validated using
multiple complementary methods by Trost et al, which showed
that the validation rate of detected tandem repeats was 77%
compared to long-read sequencing [16]. In this study, we used an
identical pipeline as Trost et al. [16], so we expected the same rate
of validation in our calls. Furthermore, we selected two TRE regions
for validation based on (1) high odds ratio from burden testing, (2)
located close to genes of interest (i.e. loss of function intolerant
genes and/or known risk genes for schizophrenia), and (3) absence
of complex repetitive sequences nearby. We attempted to confirm
TREs in the chosen regions by Integrative Genomics Viewer (IGV)
visualization of WGS reads and gel electrophoresis size separation
of PCR products of the regions of interest. All TREs detected were
confirmed (Fig. 5 and Supplementary Fig. 10). We estimated ~7% excess of rare TREs in schizophrenia
compared to controls, which may approximate their collective
contribution to schizophrenia risk. This is modest compared to the
estimated contribution of common variation to schizophrenia risk
[5], but is comparable in magnitude to the previously estimated
contribution of tandem repeat expansions to disease risk in
schizophrenia and ASD [16, 17]. Mojarad et al estimated that rare
exonic TREs contribute to ~4% of the risk in schizophrenia based
on excess of variants detected [18]. However, we note that we
employed a more stringent rare variant frequency filter (0.1%)
than that used in Mojarad et al (0.5%), and, therefore, the higher
excess observed in our study is as expected [18]. Trost et al and
Mitra et al estimated that rare TREs contribute to ~4% of ASD
cases based on excess of variants detected [16, 17]. It is Molecular Psychiatry (2023) 28:475 – 482 J. Wen et al. 480 Fig. 4
Constraint scores for genes with and without rare TREs. Constraint scores (extracted from gnomAD’s observed/expected (o/e) upper
bound LOEUF values) of genes with rare TREs are compared against those of the other genes without rare TREs in schizophrenia. A Only genes
differentially expressed between schizophrenia and controls as determined by the Common Mind Consortium are compared, B only genes
with eQTL are compared and (C) comparison is done for all protein coding genes. Box plots show Q1-1.5×IQR, Q1, median, Q3 and
Q3 + 1.5×IQR. Confirmation of EHdn detected tandem repeats A, B, Images of the gel electrophoresis showing bands that
correspond to the expanded alleles in schizophrenia (SCZ) cases
and the unexpanded allele in the reference sample NA12878. A
100 bp ladder is shown for size reference. g
p
Although we observed a mildly elevated burden of rare TREs in
schizophrenia in non-coding conserved bases, we did not identify
significant enrichment of rare TREs within various categories of
active regulatory elements in human brain. We note that there
may be a disproportionately low representation of regions
containing tandem repeats in currently available functional
genomic annotations. For example, the use of exclusive regions
of “blacklists” have been employed by the ENCODE project to
remove signal artifact regions in next-generation sequencing
experiments [34, 49]. These blacklisted regions may be due to
repetitive elements or other anomalies where genome assembly
has been difficult resulting in problematic read alignment and
erroneous signals. Such blacklist filtering is a widely-used quality
control step for functional genomics assays. noteworthy however that the risk estimation from rare TREs in
ASD was based on comparison of sample proportions between
affected probands and unaffected siblings within families, while
our risk estimation in schizophrenia was based on comparison of
sample proportions between unrelated individuals. The risk
estimation in ASD may be more conservative as some unaffected
siblings within ASD families may have inherited shorter, yet
expanded, tandem repeats [16]. Apart from the comparable
collective contributions from rare TREs to ASD and schizophrenia,
we did not identify with statistical significance TRE gene-pathways
shared between ASD and schizophrenia, though we were likely
underpowered to detect small enrichment given the rarity of the
variants. noteworthy however that the risk estimation from rare TREs in
ASD was based on comparison of sample proportions between
affected probands and unaffected siblings within families, while
our risk estimation in schizophrenia was based on comparison of
sample proportions between unrelated individuals. The risk
estimation in ASD may be more conservative as some unaffected
siblings within ASD families may have inherited shorter, yet
expanded, tandem repeats [16]. Apart from the comparable
collective contributions from rare TREs to ASD and schizophrenia,
we did not identify with statistical significance TRE gene-pathways
shared between ASD and schizophrenia, though we were likely
underpowered to detect small enrichment given the rarity of the
variants. We acknowledge several constraints in variant detection owing
to limitations in existing algorithms. Confirmation of EHdn detected tandem repeats P-values reported were calculated from one-sided Wilcoxon rank-sum test assuming lower gnomAD o/e upperbound in genes
with rare TREs. In all three categories assessed, the genes with rare TREs found in schizophrenia were more constrained (i.e., had on average
lower LOEUF values) than the genes without rare TREs found in schizophrenia. 0 Fig. 4
Constraint scores for genes with and without rare TREs. Constraint scores (extracted from gnomAD’s observed/expected (o/e) upper
bound LOEUF values) of genes with rare TREs are compared against those of the other genes without rare TREs in schizophrenia. A Only genes
differentially expressed between schizophrenia and controls as determined by the Common Mind Consortium are compared, B only genes
with eQTL are compared and (C) comparison is done for all protein coding genes. Box plots show Q1-1.5×IQR, Q1, median, Q3 and
Q3 + 1.5×IQR. P-values reported were calculated from one-sided Wilcoxon rank-sum test assuming lower gnomAD o/e upperbound in genes
with rare TREs. In all three categories assessed, the genes with rare TREs found in schizophrenia were more constrained (i.e., had on average
lower LOEUF values) than the genes without rare TREs found in schizophrenia. genes were found to involve synaptic functions and signaling
which is consistent with a prior report of tandem repeats in
schizophrenia [18]. Our finding that genes with TREs tend to be
more constrained is consistent with the literature [18] and
suggests that TREs may affect the underlying genes in the same
manner as loss of function variants. Our finding that rare TREs are
enriched in schizophrenia in genes involved in postsynapse
process, a gene set indicated by the largest schizophrenia GWAS
[5], suggests that studies of tandem repeats may pinpoint shared
underlying biology that is dysregulated across the spectrum of
variant type and allele frequency. As our present study is
correlative, a future direction would be to test the role of
schizophrenia-associated TREs in patient-derived stem cells and
model organisms in order to understand their precise functional
effects on gene expression. Fig. 5
Confirmation
of
EHdn
detected
tandem
repeats. A, B, Images of the gel electrophoresis showing bands that
correspond to the expanded alleles in schizophrenia (SCZ) cases
and the unexpanded allele in the reference sample NA12878. A
100 bp ladder is shown for size reference. Fig. 5
Confirmation
of
EHdn
detected
tandem
repeats. Molecular Psychiatry (2023) 28:475 – 482 Confirmation of EHdn detected tandem repeats First, EHdn detects tandem
repeats only with motifs of 2-20 bp and total length larger than
>150 bp. We were unable to evaluate tandem repeats with motif
size or total length beyond the detection ranges. Second, as
demonstrated by Trost et al. [16], EHdn likely underestimated the
number of repeat units for larger tandem repeats when the total
length of a tandem repeat is greater than the sequencing insert Our finding that rare TREs are enriched in schizophrenia in brain
expressed genes with eQTLs or with different expression levels
between schizophrenia cases and controls is not surprising. Tandem repeats can directly affect coding sequences or play an
important role in the regulation of gene expression via a variety of
mechanisms [7, 12]. The functional profiles of the implicated Molecular Psychiatry (2023) 28:475 – 482 J. Wen et al. 481 size (approximately 350 bp for the Swedish schizophrenia cohort). Third, very small expansions, such as expansions with one or two
additional repeat units, are missed. A few such small TREs, such as
those found in spinocerebellar ataxia types 1, 2, and 6, are known
to be disease causing [7, 50]. Fourth, our method does not resolve
zygosity or orientation of the tandem repeats (only an aggregated
length was estimated). 20. Mojarad BA, Yin Y, Manshaei R, Backstrom I, Costain G, Heung T, et al. Genome
sequencing broadens the range of contributing variants with clinical implications
in schizophrenia. Transl Psychiatry. 2021;11:84. y
y
21. Song JHT, Lowe CB, Kingsley DM. Characterization of a Human-Specific Tandem
Repeat Associated with Bipolar Disorder and Schizophrenia. Am J Hum Genet. 2018;103:421–30. 22. Brainstorm C, Anttila V, Bulik-Sullivan B, Finucane HK, Walters RK, Bras J, et al. Analysis of shared heritability in common disorders of the brain. Science. 2018;360:eaap8757. With the current sample size, we lack the power to detect
individual tandem repeat loci that are associated with schizo-
phrenia risk at genome-wide significance. Prior studies in ASD that
had sample size larger than ours but in similar magnitude (~2000
cases) also failed to implicate individual loci of tandem repeats
with ASD risk [16, 17]. Much larger cohorts (on the order of
Ncase/control > 10,000), combined with future improvements in
variant detection methods, will likely pinpoint specific tandem
repeat loci [19]. Substantial collaborative efforts will be critical in
the pursuit of larger sample sizes. REFERENCES Rare coding
variants
in ten genes confer
substantial risk for schizophrenia. Nature. 2022;604:509–16. 34. Consortium EP. An integrated encyclopedia of DNA elements in the human
genome. Nature. 2012;489:57–74. 7. Depienne C, Mandel JL. 30 years of repeat expansion disorders: What have we
learned
and
what
are
the
remaining
challenges? Am
J
Hum
Genet. 2021;108:764–85. 7. Depienne C, Mandel JL. 30 years of repeat expansion disorders: What have we
learned
and
what
are
the
remaining
challenges? Am
J
Hum
Genet. 2021;108:764–85. 35. Giusti-Rodriguez PMD, Sullivan PF. Using three-dimensional regulatory chromatin
interactions from adult and fetal cortex to interpret genetic results for
psychiatric disorders and cognitive traits. bioRxiv https://doi.org/10.1101/406330. 8. Hannan AJ. Repeat DNA expands our understanding of autism spectrum dis-
order. Nature. 2021;589:200–2. 36. Girdhar K, Hoffman GE, Jiang Y, Brown L, Kundakovic M, Hauberg ME, et al. Cell-
specific histone modification maps in the human frontal lobe link schizophrenia
risk to the neuronal epigenome. Nat Neurosci. 2018;21:1126–36. 9. Gymrek M. A genomic view of short tandem repeats. Curr Opin Genet Dev. 2017;44:9–16. 37. Hu B, Won H, Mah W, Park RB, Kassim B, Spiess K, et al. Neuronal and glial 3D
chromatin architecture informs the cellular etiology of brain disorders. Nat
Commun. 2021;12:3968. 10. Willems T, Gymrek M, Highnam G, Genomes Project C, Mittelman D, Erlich Y. The
landscape of human STR variation. Genome Res. 2014;24:1894–904. 11. Hannan AJ. Tandem repeats mediating genetic plasticity in health and disease. Nat Rev Genet. 2018;19:286–98. 38. Skene NG, Bryois J, Bakken TE, Breen G, Crowley JJ, Gaspar HA, et al. Genetic
identification
of
brain
cell
types
underlying
schizophrenia. Nat
Genet. 2018;50:825–33. 12. Usdin K. The biological effects of simple tandem repeats: lessons from the repeat
expansion diseases. Genome Res. 2008;18:1011–9. 39. Handt M, Epplen A, Hoffjan S, Mese K, Epplen JT, Dekomien G. Point mutation
frequency in the FMR1 gene as revealed by fragile X syndrome screening. Mol
Cell Probes. 2014;28:279–83. 13. Gall-Duncan T, Sato N, Yuen RKC, Pearson CE. Advancing genomic technologies
and clinical awareness accelerates discovery of disease-associated tandem repeat
sequences. Genome Res. 2022;32:1–27. 40. Fromer M, Roussos P, Sieberts SK, Johnson JS, Kavanagh DH, Perumal TM, et al. Gene expression elucidates functional impact of polygenic risk for schizophrenia. Nat Neurosci. 2016;19:1442–53. 14. Grunewald TG, Bernard V, Gilardi-Hebenstreit P, Raynal V, Surdez D, Aynaud MM,
et al. Chimeric EWSR1-FLI1 regulates the Ewing sarcoma susceptibility gene EGR2
via a GGAA microsatellite. Confirmation of EHdn detected tandem repeats Our data are meant to be
included through such collaborative efforts in the future in meta-
analyzing whole genome sequence datasets in schizophrenia and
other psychiatric disorders. 23. Mah W, Won H. The three-dimensional landscape of the genome in human brain
tissue unveils regulatory mechanisms leading to schizophrenia risk. Schizophr
Res. 2020;217:17–25. 24. Ripke S, O’Dushlaine C, Chambert K, Moran JL, Kahler AK, Akterin S, et al. Genome-wide association analysis identifies 13 new risk loci for schizophrenia. Nat Genet. 2013;45:1150–9. 25. Szatkiewicz JP, O’Dushlaine C, Chen G, Chambert K, Moran JL, Neale BM, et al. Copy
number
variation
in
schizophrenia
in
Sweden. Mol
Psychiatry. 2014;19:762–73. 26. Lichtenstein P, De Faire U, Floderus B, Svartengren M, Svedberg P, Pedersen NL. The Swedish Twin Registry: a unique resource for clinical, epidemiological and
genetic studies. J Intern Med. 2002;252:184–205. 27. Byrska-Bishop M, Evani US, Zhao X, Basile AO, Abel HJ, Regier AA, et al. High-
coverage whole-genome sequencing of the expanded 1000 Genomes Project
cohort including 602 trios. Cell. 2022;185:3426–40. REFERENCES 28. Dolzhenko E, Bennett MF, Richmond PA, Trost B, Chen S, van Vugt J, et al. ExpansionHunter Denovo: a computational method for locating known and
novel
repeat
expansions
in
short-read
sequencing
data. Genome
Biol. 2020;21:102. 1. Sullivan PF, Kendler KS, Neale MC. Schizophrenia as a complex trait: evidence
from a meta-analysis of twin studies. Arch Gen Psychiatry. 2003;60:1187–92. 1. Sullivan PF, Kendler KS, Neale MC. Schizophrenia as a complex trait: evidence
from a meta-analysis of twin studies. Arch Gen Psychiatry. 2003;60:1187–92. 2. Lichtenstein P, Yip BH, Bjork C, Pawitan Y, Cannon TD, Sullivan PF, et al. Common
genetic determinants of schizophrenia and bipolar disorder in Swedish families: a
population-based study. Lancet. 2009;373:234–9. 2. Lichtenstein P, Yip BH, Bjork C, Pawitan Y, Cannon TD, Sullivan PF, et al. Common
genetic determinants of schizophrenia and bipolar disorder in Swedish families: a
population-based study. Lancet. 2009;373:234–9. 29. Ester Martin, K H-P, Sander Jiirg, Xu Xiaowei. A density-based algorithm for dis-
covering
clusters
in
large
spatial
databases
with
noise. KDD-96
Proc. 1996;96:226–31. 3. Wray NR, Gottesman II. Using summary data from the danish national registers to
estimate heritabilities for schizophrenia, bipolar disorder, and major depressive
disorder. Front Genet. 2012;3:118. 30. Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genome-
wide association scans. Bioinformatics. 2010;26:2190–1. 31. Zoonomia C. A comparative genomics multitool for scientific discovery and
conservation. Nature. 2020;587:240–5. 4. Marshall CR, Howrigan DP, Merico D, Thiruvahindrapuram B, Wu W, Greer DS,
et al. Contribution of copy number variants to schizophrenia from a genome-
wide study of 41,321 subjects. Nat Genet. 2017;49:27–35. 32. Siepel A, Bejerano G, Pedersen JS, Hinrichs AS, Hou M, Rosenbloom K, et al. Evolutionarily conserved elements in vertebrate, insect, worm, and yeast gen-
omes. Genome Res. 2005;15:1034–50. 5. Trubetskoy V, Pardinas AF, Qi T, Panagiotaropoulou G, Awasthi S, Bigdeli TB, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nature. 2022:604. 33. Bryois J, Garrett ME, Song L, SafiA, Giusti-Rodriguez P, Johnson GD, et al. Eva-
luation of chromatin accessibility in prefrontal cortex of individuals with schi-
zophrenia. Nat Commun. 2018;9:3121. 6. Singh T, Poterba T, Curtis D, Akil H, Al Eissa M, Barchas JD, et al. Rare coding
variants
in ten genes confer
substantial risk for schizophrenia. Nature. 2022;604:509–16. 6. Singh T, Poterba T, Curtis D, Akil H, Al Eissa M, Barchas JD, et al. ACKNOWLEDGEMENTS This project is funded by NIMH R01 MH106611 relating to National Institute of Mental
Health (to J.P.S.) and SciLifeLab National Project under the project identifier 2015-R2
(to PFS.). PFS gratefully acknowledges support from the Swedish Research Council
(Vetenskapsrådet, award D0886501). JW acknowledges support by a fellowship from
the
NHLBI
BioData
Catalyst
program
(1OT3HL142479-01,
1OT3HL142478-01,
1OT3HL142481-01, 1OT3HL142480-01, 1OT3HL147154). PGR was supported by NIMH
K01 MH109772. YL is supported by R01MH12523 (to PFS). HW acknowledges
DP2MH122403 and U01MH122509. The Sweden Schizophrenia Study was supported
by NIMH R01 MH077139 (to PFS). Prior whole genome sequencing of the Swedish
schizophrenia cases and controls was supported by National Genomics Infrastructure
(NGI) Sweden, Science for Life Laboratory, the Swedish Research Council and the
Knut and Alice Wallenberg Foundation. The SweGen Project was funded by Science
for Life Laboratory (SciLifeLab) as a National Project, supported by the Knut and Alice
Wallenberg Foundation (2014.0272), and The National Research Council (PI: UG). We
acknowledge The Swedish Twin Registry for access to data. The Swedish Twin
Registry is managed by Karolinska Institutet and receives funding through the
Swedish Research Council under the grant number 2017-00641. The computations/
data handling/[SIMILAR] were/was enabled by resources provided by the Swedish
National Infrastructure for Computing (SNIC) at [SNIC CENTER] partially funded by the
Swedish Research Council through grant agreement no. 2018-05973. RKCY is
supported by The Hospital for Sick Children’s Research Institute, SickKids Catalyst
Scholar in Genetics, Brain Canada, The Azrieli Foundation, the University of Toronto
McLaughlin Center and the Nancy E.T. Fahrner Award. This work was also supported
by the Canadian Institutes of Health Research (CIHR) (PJT-175329 to RKCY; PJT-
178161 to RKCY and ASB; MOP-89066 and MOP-111238 to ASB). BT was funded by
the Canadian Institutes for Health Research Banting Postdoctoral Fellowship and the
Brain Canada Canadian Open Neuroscience Platform Research Scholar Award. A.S.B. holds the Dalglish Family Chair in 22q11.2 Deletion Syndrome at the University
Health Network and University of Toronto. Supplementary information The online version contains supplementary material
available at https://doi.org/10.1038/s41380-022-01857-4. Correspondence and requests for materials should be addressed to Ryan K. C. Yuen
or Jin P. Szatkiewicz. Reprints and permission information is available at http://www.nature.com/
reprints AUTHOR CONTRIBUTIONS 45. Bragin E, Chatzimichali EA, Wright CF, Hurles ME, Firth HV, Bevan AP, et al. DECIPHER:
database for the interpretation of phenotype-linked plausibly pathogenic sequence
and copy-number variation. Nucleic Acids Res. 2014;42:D993–D1000. (Database issue) The study was designed by JPS and RKCY. Funding was obtained by JPS, PFS and
RKCY. Statistical analysis was performed by JW, B Trost, WE, MH, JPS, and RKCY. Experimental validation was performed by NA., MF, LMP, AM, KG, and IB. PGR
provided annotations for CTCF and boundaries of topologically associating domains
from human adult brain. HW provided annotations for Hi-C enhancer-promoter
anchors from human brain. PFS, GP, JDR, YL, and B Thiruvahindrapuram provided
bioinformatics assistance. PKEM, UG, and the SweGen Project provided SweGen data. CMH and PFS provided the Swedish schizophrenia samples. ASB provided the
Canadian schizophrenia samples used for replication. JW and JPS wrote the
manuscript. B Trost, WE, MF, and RKCY contributed to writing. All authors reviewed
and approved the final version of the manuscript. 46. Coe BP, Witherspoon K, Rosenfeld JA, van Bon BW, Vulto-van Silfhout AT, Bosco P,
et al. Refining analyses of copy number variation identifies specific genes asso-
ciated with developmental delay. Nat Genet. 2014;46:1063–71. 47. Stefansson H, Meyer-Lindenberg A, Steinberg S, Magnusdottir B, Morgen K,
Arnarsdottir S, et al. CNVs conferring risk of autism or schizophrenia affect cog-
nition in controls. Nature. 2014;505:361–6. 48. Karczewski KJ, Francioli LC, Tiao G, Cummings BB, Alföldi J, Wang Q, et al. The
mutational constraint spectrum quantified from variation in 141,456 humans. Nature. 2020;581:434–43. 49. Amemiya HM, Kundaje A, Boyle AP. The ENCODE blacklist: identification of pro-
blematic regions of the genome. Sci Rep. 2019;9:9354. REFERENCES Nat Genet. 2015;47:1073–8. 41. Wang D, Liu S, Warrell J, Won H, Shi X, Navarro FCP, et al. Comprehensive
functional genomic resource and integrative model for the human brain. Science
2018;362:eaat8464. 15. Sun JH, Zhou L, Emerson DJ, Phyo SA, Titus KR, Gong W, et al. Disease-associated
short tandem repeats co-localize with chromatin domain boundaries. Cell. 2018;175:224–38 e215. 42. Satterstrom FK, Kosmicki JA, Wang J, Breen MS, De Rubeis S, An JY, et al. Large-
scale exome sequencing study implicates both developmental and functional
changes in the neurobiology of autism. Cell. 2020;180:568–84.e523. 16. Trost B, Engchuan W, Nguyen CM, Thiruvahindrapuram B, Dolzhenko E, Back-
strom I, et al. Genome-wide detection of tandem DNA repeats that are expanded
in autism. Nature. 2020;586:80–86. 43. Kochinke K, Zweier C, Nijhof B, Fenckova M, Cizek P, Honti F, et al. Systematic
phenomics analysis deconvolutes genes mutated in intellectual disability into
biologically coherent modules. Am J Hum Genet. 2016;98:149–64. 17. Mitra I, Huang B, Mousavi N, Ma N, Lamkin M, Yanicky R, et al. Patterns of de novo
tandem repeat mutations and their role in autism. Nature. 2021;589:246–50. 18. Mojarad BA, Engchuan W, Trost B, Backstrom I, Yin Y, Thiruvahindrapuram B, et al. Genome-wide tandem repeat expansions contribute to schizophrenia risk. Mol
Psychiatry. 2022:1–7. 44. Kaplanis J, Samocha KE, Wiel L, Zhang Z, Arvai KJ, Eberhardt RY, et al. Integrating
healthcare
and
research
genetic
data
empowers
the
discovery
of
49
novel developmental disorders. bioRxiv preprint https://doi.org/10.1101/797787. 19. Halvorsen M, Huh R, Oskolkov N, Wen J, Netotea S, Giusti-Rodriguez P, et al. Increased burden of ultra-rare structural variants localizing to boundaries of
topologically associated domains in schizophrenia. Nat Commun. 2020;11:1842. Molecular Psychiatry (2023) 28:475 – 482 J. Wen et al. 482 Reprints and permission information is available at http://www.nature.com/
reprints Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims
in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative
Commons licence, and indicate if changes were made. The images or other third party
material in this article are included in the article’s Creative Commons licence, unless
indicated otherwise in a credit line to the material. If material is not included in the
article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly
from
the
copyright
holder. To
view
a
copy
of
this
licence,
visit
http://
creativecommons.org/licenses/by/4.0/. © The Author(s) 2022 © The Author(s) 2022 COMPETING INTERESTS 50. McMurray CT. Mechanisms of trinucleotide repeat instability during human
development. Nat Rev Genet. 2010;11:786–99. PF Sullivan reports the following potentially competing financial interests: Neumora
(advisory board, shareholder). The other authors declare no competing interests. ADDITIONAL INFORMATION Molecular Psychiatry (2023) 28:475 – 482
|
https://openalex.org/W2082820264
|
https://epress.lib.uts.edu.au/journals/index.php/portal/article/download/2568/2884
|
English
| null |
Moorings: Indian Ocean Creolizations
|
Portal
| 2,012
|
cc-by
| 21,822
|
f
p
y
,
,
,
y
Indian Ocean Traffic Special Issue, guest edited by Lola Sharon Davidson and Stephen Muecke. p
, g
y
ISSN: 1449-2490; http://epress.lib.uts.edu.au/ojs/index.php/portal ;
p
p
j
p p p
PORTAL is published under the auspices of UTSePress, Sydney, Australia. PORTAL Journal of Multidisciplinary International Studies, vol. 9, no. 1, January 2012. Françoise Vergès & Carpanin Marimoutou
Translated by Stephen Muecke & Françoise Vergès Françoise Vergès & Carpanin Marimoutou
Translated by Stephen Muecke & Françoise Vergès ‘Amarres’ [moorings] in Réunion Creole means many different things: link, attachment,
bewitchment, spellbinding, to be in love, to be captivated, to be in a relationship, to care for [amar
lë ker], to enliven the senses [i amar la boush]. And a few other things as well … Indian Ocean Traffic Special Issue, guest edited by Lola Sharon Davidson and Stephen Muecke.
ISSN: 1449-2490; http://epress.lib.uts.edu.au/ojs/index.php/portal Ocean region, a community that is both imagined and concrete, ancient and still being
created. Ocean region, a community that is both imagined and concrete, ancient and still being
created. When Europe used to think of itself as the centre of the world, and organised the world
around this centre, we were somewhere over there at the end of the world. Then, we
were still moored to France, but it was an imposed mooring that strangled us on
occasions. Today, now that Europe has become one of the provinces of the world, we
are rethinking our moorings. Our project is now one of decentering the gaze and
redrawing the cartography of the world from the Indian Ocean viewpoint, here where
France, Africa, Europe, Asia and the Muslim worlds cross paths. We want to inscribe
our island into a network of meetings and exchanges, at the crossroads of African,
European, Asian, and islander worlds. Sure, this is a somewhat peripheral world, but it
is one we can think with, work over, and transform into an asset or an advantage. We
are not in the centre of the world, we never will be. We will always be a little on the
sidelines, in the margins, but so what? We suggest a mode of reinscription in diversity,
and think of globalisation as a series of meetings and exchanges in a multipolar world. Moorings—so that we can anchor ourselves in the ocean, and then—we slip the
moorings, to enter into relationships. When Europe used to think of itself as the centre of the world, and organised the world
around this centre, we were somewhere over there at the end of the world. Then, we
were still moored to France, but it was an imposed mooring that strangled us on
occasions. Today, now that Europe has become one of the provinces of the world, we
are rethinking our moorings. Our project is now one of decentering the gaze and
redrawing the cartography of the world from the Indian Ocean viewpoint, here where
France, Africa, Europe, Asia and the Muslim worlds cross paths. We want to inscribe
our island into a network of meetings and exchanges, at the crossroads of African, European, Asian, and islander worlds. Sure, this is a somewhat peripheral world, but it
is one we can think with, work over, and transform into an asset or an advantage. We
are not in the centre of the world, we never will be. Vergès & Marimoutou Vergès & Marimoutou Moorings Ocean region, a community that is both imagined and concrete, ancient and still being
created. PORTAL, vol. 9, no. 1, January 2012. PORTAL, vol. 9, no. 1, January 2012. The island Oh no
I am not writing goodbye again
no
I am not rewriting the ledger of returns
to the homeland
but
playing knucklebones
for childhood renewed
Patrice Treuthardt, Pointe et Complainte des Galets Oh no
I am not writing goodbye again
no
I am not rewriting the ledger of returns
to the homeland
but
playing knucklebones
for childhood renewed
Patrice Treuthardt, Pointe et Complainte des Galets Oh no
I am not writing goodbye again
no
I am not rewriting the ledger of returns
to the homeland
but
playing knucklebones
for childhood renewed Patrice Treuthardt, Pointe et Complainte des Galets We are native to an island that is often left off maps of the world, and often confused
with other French overseas territories. We want to bring to bear a problematic starting
with this forgetfulness and confusion. Since it is true that it is the lot of so many peoples
and groups to be forgotten, or not to count, we have to ask: ‘Forgotten by whom, and
why? Counting for whom, and why?’ So starting from this forgetfulness, this ‘non-
existence,’ and asking oneself the fundamentally political question of ‘who counts and
for whom,’ is to go straight to the heart of what brings sociality into being, that is, being
accepted by the community of citizens. But this community is not a purely national one,
it relates also to what it means to live together, on the soil of Réunion and in the Indian Vergès & Marimoutou Ocean region, a community that is both imagined and concrete, ancient and still being
created. We will always be a little on the
sidelines, in the margins, but so what? We suggest a mode of reinscription in diversity,
and think of globalisation as a series of meetings and exchanges in a multipolar world. Moorings—so that we can anchor ourselves in the ocean, and then—we slip the
moorings, to enter into relationships. What is our motivation for writing this text? One might say that everything has already
been said on questions of Réunion, métissage, the intercultural, and other cultural
crossovers, and that we could only go over the same ground, but less effectively. There
would be nothing to add, or at least very little. Do we have the capacity to renew all this,
or should we wait for the ‘next generations’ who would naturally be identified with the
‘new’? The need to write this text emerged as a response to several things: the
increasingly significant presence of artists and culture in the Réunion landscape, and the
questions arising from that presence; the new interest taken by Paris in these What is our motivation for writing this text? One might say that everything has already
been said on questions of Réunion, métissage, the intercultural, and other cultural
crossovers, and that we could only go over the same ground, but less effectively. There
would be nothing to add, or at least very little. Do we have the capacity to renew all this,
or should we wait for the ‘next generations’ who would naturally be identified with the
‘new’? The need to write this text emerged as a response to several things: the
increasingly significant presence of artists and culture in the Réunion landscape, and the
questions arising from that presence; the new interest taken by Paris in these
Réunionese artistic developments; the lack of thinking about them; the superficiality of
public debate; the aggressively masculine ethic of the Réunionese discourses on art,
culture, politics, and the social; new questions and new practices arising from the
profound changes of the last thirty years; and the need to take part in the postcolonial
debate. The language of the Other For all that, we will not idealise
mixing, nor indulge in wide-eyed celebration of creolité, but we will pay attention to the
conflicts and tensions, and to the impulses that are always possible out of compensatory
ethnic identification. The language of the Other The language of the Other
This text was written by two people from Réunion, a woman and a man who grew up
here, who feel native to the island, and who have taken part in, and continue to take part
in, the cultural and political debate. To be native for us means not just to be born on the
island but also to care for it as regards its place in the Indian Ocean; the reassessment of
its local practices and modes of expression; and of the reclaiming of its territory. For us,
being here means (without having any choice in the matter) being bilingual and
pluricultural. Europe has privileged monolingualism and monoculturalism for a long
time. The European didn’t have need for others’ languages and cultures since in his eyes
his own culture or language was considered universal. In the colonies plural languages
and cultures were an inevitable fact of life, but made out by imperialism to be signs of
backwardness. Yet, they are now the necessary condition for intercultural practice. The
language of the other has become ours—we are not proud or ashamed of it—and we
have not lost our native tongue. It is also without fear or favour that we borrow the
techniques and conceptual tools useful to us, and that we express our love for the arts
and literatures of the West. We want to take on board the saying of the great African-
American poet Audre Lorde: Using the master’s tools to dismantle the master’s house. Despite that, we are not into binary oppositions: Europe is one of our worlds. Africa,
Asia, the Moslem and island worlds are all part of our matrix. We are simply beginning
a critical inventory. Binary oppositions can be dispensed with, all cultures interpenetrate
and nourish one another, and none is isolated and pure. For all that, we will not idealise
mixing, nor indulge in wide-eyed celebration of creolité, but we will pay attention to the
conflicts and tensions, and to the impulses that are always possible out of compensatory
ethnic identification. Asia, the Moslem and island worlds are all part of our matrix. We are simply beginning
a critical inventory. Binary oppositions can be dispensed with, all cultures interpenetrate
and nourish one another, and none is isolated and pure. Vergès & Marimoutou Moorings psychoanalysis, feminism, contemporary art and architecture; and on the other hand, the
theoreticians of postcolonialism, among them Aimé Césaire, Frantz Fanon, C. L. R. James, Stuart Hall, Edward Saïd, Paul Gilroy, Arjun Appadurai, Arundhati Roy, Sara
Suleri, Salman Rushdie, Achille Mbembe. It is on the basis of all this that we hope to
develop a critical movement on Réunion. Ocean region, a community that is both imagined and concrete, ancient and still being
created. On the one hand, we will base our work in part on the structuralism and
poststructuralism of Michel Foucault, Claude Lévi-Strauss, Jacques Derrida, Tzvetan
Todorov, James Clifford, Régine Robin, Chantal Mouffe, and the insights of Réunionese artistic developments; the lack of thinking about them; the superficiality of
public debate; the aggressively masculine ethic of the Réunionese discourses on art,
culture, politics, and the social; new questions and new practices arising from the
profound changes of the last thirty years; and the need to take part in the postcolonial
debate. On the one hand, we will base our work in part on the structuralism and
poststructuralism of Michel Foucault, Claude Lévi-Strauss, Jacques Derrida, Tzvetan
Todorov, James Clifford, Régine Robin, Chantal Mouffe, and the insights of PORTAL, vol. 9, no. 1, January 2012. 2 2 Vergès & Marimoutou PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou knowingly and willingly left others to one side. As we wrote we were confronted by the
multiple meanings of ‘we’: the ‘we’ that refers to the authors, the ‘we’ that stands in
opposition or confrontation to a ‘they’ (be it on or beyond the island) and the ‘we’ that
brings together the island’s inhabitants. We are conscious of the exclusion effects
brought about by the ‘we,’ but we know that no group is brought into being without
some strategy of exclusion. There can be no process of identification without the
establishment of a frontier between a ‘we’ and a ‘they.’ But this in no way signifies that
frontiers cannot be crossed, that the other does not constitute the self, and that such
identifications are not constantly subject to negotiable transformations. This also means
that one has to be wary of the transformation of the ‘they’ into a fantastical threat to our
existence, bringing about the emergence of closed identities. The common ‘we’ of this
essay is in the process of becoming. There are two reefs to avoid sailing into, on the port
the nationalist and/or communalist we, and on the starboard the risk of being diluted in
the abstract and ahistorical universal of the so-called ‘global village.’ If nationalism—as
invention of European romanticism—was quite rightly used as a force in decolonisation
struggles, if the nation-state remains the dominant model with all the problems that it
brings with it (repression of languages and cultures), we know, then, from the examples
of the rise and fall of nationalisms in the second half of the 20th century, that this model
is not the only one we might use for our own emancipation and reappropriation. Réunion, as a part of the French Republic, therefore in Europe, and a region anchored in
its Indian Ocean environment, has to invent its own postcolonial model. The absence of
an enlightened bourgeoisie in Réunion—apart from some rare exceptions, few have
anything to do with any movement to reinvest in the territory and its cultures—is
associated with both communitarian1 tendencies and a fascination for the ‘global
village,’ both of which are ahistorical attitudes. The rapid growth of communications
and the accelerating access to consumer goods, along with the disappearance of
scandalously visible poverty have inevitably produced the illusion that all this has come
to pass without friction. 1 Communaliste, a term used in Mauritius, as noted by the authors, is used in the original. PORTAL, vol. 9, no. 1, January 2012. Frontiers of the ‘we’ The ideas we are presenting are open to debate; they are partisan without seeking either
to be exhaustive or neutral. This is a text that highlights certain points. We have The ideas we are presenting are open to debate; they are partisan without seeking either
to be exhaustive or neutral. This is a text that highlights certain points. We have 3 PORTAL, vol. 9, no. 1, January 2012. 3 Vergès & Marimoutou Vergès & Marimoutou Moorings Vergès & Marimoutou It is rare to find people who have
passed on to their children the family genealogy, the struggles and battles, which have
opened the door to the current phenomenon of compensatory identities. These social
classes are quite like the postcolonial middle classes that have, consciously or not, taken
part in what Sarat Maharaj calls ‘multicultural management’ (2001), which can accept a
little cultural difference but not too much, and especially if it is well-framed by a strict
separation between the social and the cultural, the cultural and the political. So we are
looking to develop a ‘we’ that would avoid cultures of recrimination, the
mythologisation of history, self-referenced identity or the fundamentally static notion of
identity and choose responsibility, the present, the heterogeneous and creolisation. This
is a ‘we’ that remembers the past but is not enclosed there; it is situated in a genealogy
of struggles for justice equity and democratisation rapist) are put forward by way of explanation for Réunion’s ‘lack’ of progress, while the
inheritances of colonialism with its brutality and violence, are moved to the background. For us, only an analysis that works the engagements between the political and the social,
the economic and the cultural, the private and the public, can allow us to understand the
complexity of life on Réunion. The Réunion bourgeoisie, for the most part, has
i
d
ithd
hil th l
iddl
l
th
hild
f f
k shopkeepers, labourers, clerks, quickly wanted to forget, in the rush to the ‘metropolis,’
where they came from, as they grasp the secondary signs of Frenchness—cars, holidays
in Mauritius, contempt for the poor—while at the same time being unaware of cultural
and intellectual movements in the region or in Europe. It is rare to find people who have
passed on to their children the family genealogy, the struggles and battles, which have
opened the door to the current phenomenon of compensatory identities. These social
classes are quite like the postcolonial middle classes that have, consciously or not, taken
part in what Sarat Maharaj calls ‘multicultural management’ (2001), which can accept a
little cultural difference but not too much, and especially if it is well-framed by a strict
separation between the social and the cultural, the cultural and the political. Vergès & Marimoutou Moorings rapist) are put forward by way of explanation for Réunion’s ‘lack’ of progress, while the
inheritances of colonialism with its brutality and violence, are moved to the background. For us, only an analysis that works the engagements between the political and the social,
the economic and the cultural, the private and the public, can allow us to understand the
complexity of life on Réunion. The Réunion bourgeoisie, for the most part, has
remained withdrawn, while the lower middle class, the children of farm workers,
shopkeepers, labourers, clerks, quickly wanted to forget, in the rush to the ‘metropolis,’
where they came from, as they grasp the secondary signs of Frenchness—cars, holidays
in Mauritius, contempt for the poor—while at the same time being unaware of cultural
and intellectual movements in the region or in Europe. It is rare to find people who have
passed on to their children the family genealogy, the struggles and battles, which have
opened the door to the current phenomenon of compensatory identities. These social
classes are quite like the postcolonial middle classes that have, consciously or not, taken
part in what Sarat Maharaj calls ‘multicultural management’ (2001), which can accept a
little cultural difference but not too much, and especially if it is well-framed by a strict
separation between the social and the cultural, the cultural and the political. So we are
looking to develop a ‘we’ that would avoid cultures of recrimination, the
mythologisation of history self referenced identity or the fundamentally static notion of rapist) are put forward by way of explanation for Réunion’s ‘lack’ of progress, while the
inheritances of colonialism with its brutality and violence, are moved to the background. For us, only an analysis that works the engagements between the political and the social,
the economic and the cultural, the private and the public, can allow us to understand the
complexity of life on Réunion. The Réunion bourgeoisie, for the most part, has
remained withdrawn, while the lower middle class, the children of farm workers,
shopkeepers, labourers, clerks, quickly wanted to forget, in the rush to the ‘metropolis,’
where they came from, as they grasp the secondary signs of Frenchness—cars, holidays
in Mauritius, contempt for the poor—while at the same time being unaware of cultural
and intellectual movements in the region or in Europe. Vergès & Marimoutou The social and cultural struggles that brought about a better
way of life have been quickly forgotten. We have slipped under the clouds of an
amnesia that has obscured and personalised social difficulties and conflicts (the media
have by and large reflected and reproduced this amnesia). In this way certain typical
characterisations of the Réunionese (the hopeless man, the possessive mother, Dad the 4 4 Vergès & Marimoutou Vergès & Marimoutou So we are
looking to develop a ‘we’ that would avoid cultures of recrimination, the mythologisation of history, self-referenced identity or the fundamentally static notion of
identity and choose responsibility, the present, the heterogeneous and creolisation. This
is a ‘we’ that remembers the past but is not enclosed there; it is situated in a genealogy
of struggles for justice, equity and democratisation. I invoke you, land of sapodillas
In the reverberations of the riverbanks
[…]
I invoke you, land of Babel
hidden like a shameful illness
bazaar of erased alphabets
beatitude of cargo God
dried entrails
monuments of corrugated iron
I invoke you, scorched land
crossroads of nowhere bringing in contraband portulans I invoke you, land of sapodillas
In the reverberations of the riverbanks
[…]
I invoke you, land of Babel
hidden like a shameful illness
bazaar of erased alphabets
beatitude of cargo God
dried entrails
monuments of corrugated iron
I invoke you, scorched land
crossroads of nowhere bringing in contraband portulans I invoke you, land of sapodillas
In the reverberations of the riverbanks
[…]
I invoke you, land of Babel
hidden like a shameful illness
bazaar of erased alphabets
beatitude of cargo God
dried entrails
monuments of corrugated iron
I invoke you, scorched land
crossroads of nowhere bringing in contraband portulans Riel Debars, L’Oriflamme léthargique Riel Debars, L’Oriflamme léthargique The island archipelago
We propose to begin from what made us: the land where we grew up—the volcanic peak,
the uninhabited land, isolated in the Indian Ocean, known to the Arabs, avoided by the
Portuguese, colonised by the French—by retracing the trading routes criss-crossing the PORTAL, vol. 9, no. 1, January 2012. 5 5 Vergès & Marimoutou Moorings worlds that made it. The first inhabitants were French colonisers with their Malagasy
and Indian slaves. It was an island of slaves and masters, then an island of masters and
indentured labourers. It is an island where History has thrown together Malagasy,
Africans, Comorans, Indians, Chinese, Vietnamese, Malays, Europeans and French,
atheists, Catholics and Moslems, Buddhists and Hindus, animists and polytheists. But
this would be no simple business of juxtaposition. The island allows for people to be at
the same time Christian and Hindu, Christian and animist, or Hindu and animist. It is an
island of the Creole world, on the route between Africa and Asia, a ‘French’ island, an
island-archipelago. Vergès & Marimoutou It is an island of the Indian Ocean world, an island of Indian-
oceanic creolisations. PORTAL, vol. 9, no. 1, January 2012. PORTAL, vol. 9, no. 1, January 2012. Riel Debars, L’Oriflamme léthargique Riel Debars, L’Oriflamme léthargique Exiles, reappropriations Exile and deportation framed our birth. We have had to give things up, give up the right
to return, give up the story of the European travellers of the 17th and 18th centuries. Their myth was biblical—paradise or hell. We wanted to renounce the hegemonic story
of pain and suffering because we didn’t want to inspire pity. We live on the island. There would be no reparation … We were born in violence, the violence of slavery,
indentured labour and colonialism. Nothing can repair this primordial violence. Nothing. Only by building the present will our land be ours. So it is from this angle that we have
to consider questions of repair. That in itself is not enough to constitute the ground or
the foundation for our identity. Because there is no doubt about what is the issue: a
reappropriation of the land and the creating of a ‘living together.’ Let’s revise these
facts and turn once again to the scenes of subjection to slavery or indentured labour. The
island was a long way away and isolated; the French colonised it and it quickly became
the ‘sister-island’ of the Île de France (Mauritius). The French were not particularly
interested in the island: they had no great plans for it; it would simply be a stopover on
the way to India. A few settlers were sent out from France, but they did not think they
were actors in a great colonial scheme. They were forced to leave because of poverty in
rural France. They were poor landless peasants, pirates, impoverished aristocrats, and a
few tradesmen. They were not allowed to develop on their own account, they had to
follow decisions taken for them in Paris—monocultures of coffee or sugar cane—and
were forgotten on this island; the Empire forgot them. It was rather in San Domingo, the
pearl of the Antilles, that imperial glory flourished. Over there, courtly balls were put on
that rivalled in luxury those of the metropolis. This was to be Versailles in the PORTAL, vol. 9, no. 1, January 2012. 6 6 PORTAL, vol. 9, no. 1, January 2012. PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou Vergès & Marimoutou Moorings Caribbean. In Réunion the masters tried one way or another to imitate the life of those
masters: furniture from India, large estates, silks, black domestics, black nannies,
reserved pews in church … all the outward signs of the life of white masters, in the
same mould, from one plantation to another across the world. 2 The ‘big white [families]’ are big land-holders from the colonial and slave era, now mainly involved in
the import-export business, marketing, banking, automobile franchises, etc. Land of the banished and the deported f
p
This land of masters, slaves, maroons, freed slaves, indentured labourers, is first of all a
land occupied by men dominating other men. There were few women, a third of the
population for centuries. Slavery, like indenture, was predominantly a masculine affair,
because growing coffee and sugar cane demanded physical strength. You had to know
how to wield the machete all day long, repeating the same movement: grabbing the cane
with the left hand, cutting it with one blow, trimming the leaves and throwing it down. Cane dust irritated the skin, the leaves were sharp, ants bit the feet and legs, the wall of
cane hindered movement, and the cane-cutting season was during the summer … you
had to go on. Then the cane had to be baled and taken to the mill … The scenario of
oppression was of one man submitted to another, possessed by the other, like his
household goods, one of his heritage items. This was a masculine world where nature
yielded to an economy of plunder. Once the war against the maroons was won, once the
uprisings were crushed, once the abolition of slavery had been achieved, colonial power
was in a position to put its own stamp on the territory, on the periphery of the great
French colonial empire. Lives crouched in the grass at the water’s edge
Devour the horizon with wide open eyes
Claire Karm, Rue d’Après Lives crouched in the grass at the water’s edge
Devour the horizon with wide open eyes Claire Karm, Rue d’Après at Sainte-Suzanne at the beginning of the 19th century are outstanding proof of this. I was born over on Zanzibar
Who am I now? I will go look from high on the cliffs
I will climb to Dimitile and be reborn Danyèl Waro, Bwéo Danyèl Waro, Bwéo PORTAL, vol. 9, no. 1, January 2012. Moorings Moorings Vergès & Marimoutou at Sainte-Suzanne at the beginning of the 19th century are outstanding proof of this. PORTAL, vol. 9, no. 1, January 2012. Scenes of subjection: A cartography of power The scene of subjection marked the territory. The apparatus of colonial slavery and
post-slavery had to be inscribed again and again. It had to be repeated, to be made
visible for public display. The territory was marked by this display; the territory of the
masters’ houses or those of the lesser whites were distinguished from the territory of the
subjected. The latter lived in a camp. As the years went by, he squeezed from the master
the cultivation of a small plot of land, for his vegetables, his animals or his garden, but
all that remained precarious. The island was divided into territories: plantations on the
coast, with the territories of the master, the freed man and the slave, then in the centre,
in the mountains, the territory of the runaway, the maroon. Let’s not forget the territory
that remains uninhabited, not yet entirely crafted by humans: that of the volcano, the
forests, the ravines, a natural territory that has not yet come under human hands. The
territory of the plantations marked out the social and cultural cartography of the island. There was a cartography of power that imitated the territory of colonial power: towns
organised around the Church and State institutions (first monarchical, then republican). A capital with the Governor’s square and the institutions of power—Government House,
the cathedral, churches, customs houses, police, municipalities, schools—with the
botanic gardens, the main street, the shops … the other towns followed the same pattern. The cartography of counter-power, that of the maroons, shaped an island interior where
the toponymy retains the traces of the warrior chiefs. The war against the maroons
destroyed the vestiges of their villages, so the spatial organisation of their power
remains in the realm of speculation. The raids they carried out on plantations were
evidence of their capacity to develop strategies of resistance. But there was another
cartography of resistance: conspiracies among slaves to organise revolts, with
designated targets. The time was at night, the site the plantation. What they wanted was
freedom, to take back their lives even at the risk of death. The uprisings at Saint-Leu or 7 Vergès & Marimoutou Signs of the cross at the great division of the world
between fidelities and humilities
a house
only
so the evenings might pray
passage with passengers towards unknown lands
Alain Lorraine, Sur le black Signs of the cross at the great division of the world
between fidelities and humilities
a house
only
so the evenings might pray
passage with passengers towards unknown lands
Alain Lorraine Sur le black Signs of the cross at the great division of the world
between fidelities and humilities
a house
only
so the evenings might pray
passage with passengers towards unknown lands Signs of the cross at the great division of the world
between fidelities and humilities
a house
only
so the evenings might pray
passage with passengers towards unknown lands y
so the evenings might pray
passage with passengers towards unknown lands so the evenings might pray
passage with passengers towards unknown lands Alain Lorraine, Sur le black Alain Lorraine, Sur le black 3 [boutique chinois] Local grocery shop which also doubled as a bar. For a long time this was an essential
social and economic space, before the appearance and development of supermarkets. PORTAL, vol. 9, no. 1, January 2012. PORTAL, vol. 9, no. 1, January 2012. Peripheries Réunion remained on the periphery of the Empire, despite the efforts of certain grand
Blancs.2 For a long time the island did not appear on all the maps. No one was
interested, no one cared. It is still somewhat neglected, of secondary interest compared
to Mauritius, Madagascar, India, the Antilles. It is on the last rung of the imperial ladder. It remains in the margin, still confused with (or at best placed in comparison or 8 Vergès & Marimoutou Moorings Vergès & Marimoutou competition with) the Antilles, already a minor player. Throughout its history there were
those looking to create a destiny as the ‘colonising colony.’ It would work for imperial
triumph in the region, they said, it would proudly fly the flag of the colonial army, and
subjugate other peoples for the greater glory of the colonial empire. Men from Réunion
took part in imperial conquests and won fame as colonial mercenaries. Others realised
they were colonised when they encountered other colonised peoples and became anti-
colonial militants: George Garros, Raymond Vergès, Lucien Barquissau, Paul Dussac. After the abolition of slavery the territory was reorganised for further colonisation. The
freed slaves and poor whites [petits Blancs] were moved off the more fertile and easier
to cultivate lands to the high parts and isolated corners of the island. The island was
divided between les Hauts and les Bas. Sent to the margins of public colonial space
were the labourers, the small plantation owners, the laundresses. Our modernity was
shaped by this territorialisation of wealth and power. Shantytowns transformed the
urban areas, marking the territory of the poor and the excluded around the centre of the
city. In the course of the last fifty years the invention of new territories—the beach, the
road, the shopping centre, the far South, the East and the West—compounds the
phenomenon of territorialisation. Spaces disappear or are marginalised: the large
plantation, the factory, the railway, the boutik sinwa3 … The social, imaginary, cultural
and economic space has been deeply changed by all this. Vergès & Marimoutou would be integrated without too much controversy. What is at stake with this resistance? While France, despite everything, was still able to finally confront Vichy or the
Algerian war, in Réunion it still seems too dangerous to confront the past because that
would threaten either the ties that bind (‘dangers of communitarianism’) or the ties to
France (asking the question about ‘being linked to those who subjected you’) and would
bring about a rereading of notions such as citizenship, equality, and national identity
starting from the position of who is in or out. Along with this resistance is a tendency to mythologise. There is an absence of evidence,
and there is difficulty in accessing the archives (neglected for a long time, totally
abandoned, some even no doubt destroyed). The fact that the majority of the eye-
witness accounts are those of the masters, of the magistrates or other administrators of
the colonial and slave-owning society, has promoted a compensatory history that rejects
complexity, grey areas and complicity in order to highlight only suffering and heroic
deeds. The absence of concrete traces paradoxically brings about an inflation of
memory, as if only this inflation could make sense of the suffering. As Paul Ricoeur has
pointed out, the imperatives of memory run the risk of sitting in opposition to the
imperatives of history and lead to what Régine Robin defined as ‘saturated memory’
that no longer knows how to sort out myth from history. From our point of view, it is
the writing of history that is crucial. It is accompanied by memory, but it is not reduced
to it, because memory is a social construction following its own logic. In any society we
can observe memories superimposing themselves on each other or opposing each other. Our proposal is to build a ‘common story’ that would make place for memories, pointing out at every turn that the story can be subject to critical revision. We want to do
an inventory of the places of memory, of which too many have already been lost, defend
them against destruction, preserve them, make them known. We want to inscribe the
historical threads and genealogies so that transmission can occur. We want to confront
gaps in memory and continue to develop scientific rigour in the research. PORTAL, vol. 9, no. 1, January 2012. Writing histories Our past, even though it is the object of commemorations, studies and reappropriations,
still remains a polemical field. Colonialism remains a minor research topic, and if
slavery is indeed studied, it still has not become part of the common story even though
we might assume that the voices of victims, the place of genocide, of crimes against
humanity in politics and the law, within the development of international criminal law, 3 [boutique chinois] Local grocery shop which also doubled as a bar. For a long time this was an essential
social and economic space, before the appearance and development of supermarkets. PORTAL, vol. 9, no. 1, January 2012. 9 9 Vergès & Marimoutou Moorings Vergès & Marimoutou To speak
of welfare dependence or of parasitism, as certain kinds of well-known discourses like
to, returns to a denial of the need for reparation; it stigmatises people already in a
precarious state, and excludes a significant number of Réunionese from citizenship. But
we should also analyse the lethal consequences of aid that is not accompanied by
responsibility, and that organises a life predicated on the wait for the next dole cheque
or piece-meal job. The growing vulnerability of a sector of the population, hidden by
welfare and a kind of getting by, can nevertheless be read in the daily violence that
articulates the anxiety of a present going nowhere and a future that appears absent. 4 Pejorative and discriminatory term used to designate rough young men, particularly those from the
suburbs.
5 Pejorative and discriminatory term used to designate young women who are supposed to like riding 4 Pejorative and discriminatory term used to designate rough young men, particularly those from the
suburbs.
5 Pejorative and discriminatory term used to designate young women who are supposed to like riding
around in flashy sports cars. Vergès & Marimoutou Moorings Vergès & Marimoutou Figures of exclusion
Work must be done to locate the figures of exclusion in order to deconstruct them, to
make an inventory of the stereotypes and insults: l’argent braguette [social security
cheques], l’assisté [someone on the dole], le léspèrkui [parasite], parents in retirement,
les cagnards,4 la tantine larou.5 The descendants of slaves and indentured labourers do
not have the same fate as those descended from the masters. This is a fact, and we need
to study the effects of that heritage today, showing the various discriminations,
stigmatisations, and racial stereotypes. We have to think about the social politics of
reparation, already begun thanks to political and social movements (unions, women’s
movements and political parties) and put into practice through social welfare. To speak
of welfare dependence or of parasitism, as certain kinds of well-known discourses like
to, returns to a denial of the need for reparation; it stigmatises people already in a
precarious state, and excludes a significant number of Réunionese from citizenship. But
we should also analyse the lethal consequences of aid that is not accompanied by
responsibility, and that organises a life predicated on the wait for the next dole cheque
or piece-meal job. The growing vulnerability of a sector of the population, hidden by
welfare and a kind of getting by, can nevertheless be read in the daily violence that
articulates the anxiety of a present going nowhere and a future that appears absent. Work must be done to locate the figures of exclusion in order to deconstruct them, to
make an inventory of the stereotypes and insults: l’argent braguette [social security
cheques], l’assisté [someone on the dole], le léspèrkui [parasite], parents in retirement,
les cagnards,4 la tantine larou.5 The descendants of slaves and indentured labourers do
not have the same fate as those descended from the masters. This is a fact, and we need
to study the effects of that heritage today, showing the various discriminations, stigmatisations, and racial stereotypes. We have to think about the social politics of
reparation, already begun thanks to political and social movements (unions, women’s
movements and political parties) and put into practice through social welfare. suburbs.
5 Pejorative and discriminatory term used to designate young women who are supposed to like riding
around in flashy sports cars. Vergès & Marimoutou We want to
reconstitute this particular history of violence, of plunder and of dehumanisation in the
general history of violence, plunder and dehumanisation. We want to live with absence,
so that this history stops being the history of lost souls, but rather gives meaning to the
lives of the women and men who have lived in this land, and to the present. 10 PORTAL, vol. 9, no. 1, January 2012. 10 Vergès & Marimoutou PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou Vergès & Marimoutou Moorings Contemporary debates around the world suggest a whole series of responses on the
question of reparation: tribunals, Truth and Reconciliation Commissions, positive
discrimination, financial compensation, restitution (of objects, bodies, documents,
lands), the construction of memorials, and teaching. We have noticed that young people
in Réunion have no idea about the recent past, nor of the history of decolonisation that
is so important for their understanding of the world and for their place in it and their
ability to take part in its debates. For us, a first step in the process of reparation would be for the education system
genuinely to take that history into account—there is no reason why it shouldn’t be
included in the national curriculum. The production and distribution of visual
documentation could be supported and encouraged by the government, there could be a
department of African and Indian Ocean Studies at the University of Réunion, all sorts
of archives could be made more easily available, a critical sociology of Réunionese
society could be developed and, as we have said, a survey of sites of memory should be
done in order to protect them. We have a responsibility. The responsibility of writing history and of living together. The conditions under which slaves and indentured workers lived cannot explain the
current situation. To do so would be to disrespect them. And yet, can one ignore the
morbid social effects that such systems have produced? The cleaving of social links, the
rupture of transmission between generations, the reduction of the human being to a
thing, the inevitable violence of human relations, all that has had an impact, as has been
proven elsewhere, on the fabric of society, on the relations between women and men,
and parents and children. It is for this reason that history has to be explored and not
repressed, why we must not be afraid of the excessiveness of certain discourses: these
expressions must be understood as part of a desire for integration and recognition rather
than a desire for exclusion. We consider it indispensable to maintain a tension between
the sometimes turbulent emergence of these voices for so long kept out of the
Réunionese master narrative, and the need to integrate them into the new narrative we
will construct together. This is what we call being moored in history and not to history. PORTAL, vol. 9, no. 1, January 2012.
1 The politics of reparation The politics of reparation
Past events do not save us from critical interpretation. No one can occupy the moral
high ground because of the suffering of their ancestors. When a crime has affected the
whole society and can compromise its future, reparation needs to be collective. The
Truth and Reconciliation Commission in South Africa showed how a society could
collectively heal a crime by putting executioners and victims face to face while
highlighting the question of responsibility towards the past and the future. We don’t
want to make that the sole model, but we can see there a collective will to confront past
evil, and move beyond revenge in order to live together. Frantz Fanon did not want to
be a prisoner of the past: ‘I am not the slave of the Slavery that dehumanised my
fathers,’ he wrote (Fanon 1986: 230). For him, the important thing was to be a man in
the present, who would not have to carry the burden of the victimisation of his ancestors. 4 Pejorative and discriminatory term used to designate rough young men, particularly those from the
suburbs. 5 Pejorative and discriminatory term used to designate young women who are supposed to like riding
around in flashy sports cars. 4 Pejorative and discriminatory term used to designate rough young men, particularly those from the
suburbs. 5 11 11 PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou Vergès & Marimoutou of monstrous structures, let the landscapes be spoilt, buy thousands of cars every month. Spoiling the island continues the predatory practices of colonial imperialism, since it is
founded on the right to exploit while minimising the responsibility to pass things on, the
consciousness that what one does today can have irreparable effects on resources and
landscapes. Since the landscape as a social construction is part of our imaginary,
destroying it threatens our imaginary and culture. Because the landscape is linked to
and part of our social, psychological real-life experience. So when we speak of
reparation, it is about all this: it is as much about the social spaces as it is about the
historic, economic, linguistic, cultural and imaginary scapes. This politics of reparation
is one of the moorings we are suggesting. It is explicitly located in relation to the
countries that surround us, the continents we have come from: Africa, Asia, Europe and
the islands. of monstrous structures, let the landscapes be spoilt, buy thousands of cars every month. Spoiling the island continues the predatory practices of colonial imperialism, since it is
founded on the right to exploit while minimising the responsibility to pass things on, the
consciousness that what one does today can have irreparable effects on resources and
landscapes. Since the landscape as a social construction is part of our imaginary,
destroying it threatens our imaginary and culture. Because the landscape is linked to
and part of our social, psychological real-life experience. So when we speak of
reparation, it is about all this: it is as much about the social spaces as it is about the
historic, economic, linguistic, cultural and imaginary scapes. This politics of reparation
is one of the moorings we are suggesting. It is explicitly located in relation to the
countries that surround us, the continents we have come from: Africa, Asia, Europe and
the islands. Indianoceanics We plough the waves of the Indian Ocean, searching for
the most marvellous clouds, the most enchanting breezes
the most iridescent flasks,
songs, the subtlest colours, blue is
our idol, and we know how to break
waves on the sand and the reefs
Jean Albany, Fare Fare We plough the waves of the Indian Ocean, searching for
the most marvellous clouds, the most enchanting breezes
the most iridescent flasks,
songs, the subtlest colours, blue is
our idol, and we know how to break
waves on the sand and the reefs
Jean Albany, Fare Fare We plough the waves of the Indian Ocean, searching for
the most marvellous clouds, the most enchanting breezes
the most iridescent flasks,
songs, the subtlest colours, blue is
our idol, and we know how to break
waves on the sand and the reefs PORTAL, vol. 9, no. 1, January 2012.
1 PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou This is why we are
answerable to the present. It is us who throw rubbish into the ravines, who destroy the
lagoons, ruin the soil with pesticides, destroy flora and fauna, agree to the construction PORTAL, vol. 9, no. 1, January 2012. 12 Vergès & Marimoutou Moorings Vergès & Marimoutou Vergès & Marimoutou Moorings In recent years, transnational and transcontinental exchanges have undergone a renewal. They are uncovering new routes and itineraries. A new cartography is being drawn with
the emergence of new global cities like Johannesburg, Dubai and Singapore. The study
of such spaces presupposes the study of the ‘production’ of the space, which is a social
and cultural production. The Indian Ocean is a space without any precise
supranationality or territorialisation. It is a cultural space overarched by several
chronotopes, where temporalities and territorialities are constructed and deconstructed. An ocean linking continents and islands. A space that is Afro-Asiatic, Moslem,
Christian, Animist, Buddhist, Hindu, and creolised. An ocean of trade winds, monsoons,
cyclones and winds. In recent years, transnational and transcontinental exchanges have undergone a renewal. They are uncovering new routes and itineraries. A new cartography is being drawn with
the emergence of new global cities like Johannesburg, Dubai and Singapore. The study
of such spaces presupposes the study of the ‘production’ of the space, which is a social
and cultural production. The Indian Ocean is a space without any precise PORTAL, vol. 9, no. 1, January 2012. Seascapes Our understanding of land includes the ocean. The notion of seascape, untranslatable
into French, is useful here: the ocean is an immense, imaginary landscape, a space of
slave trading, of forced migrations, of deportation and of ties. It is a place of crime, of
separation, but also a place of primary transformation, of the first creolisation that unites
diversities. Exchanges, encounters, commerce, new languages and cultures; all took place in the
Indian Ocean long before the arrival of the Europeans. There were cosmopolitan cities,
genuine global towns where Jews, Armenians, Arabs, Indians, Chinese, Malagasy
rubbed shoulders… prefiguring (as evocative singular figures rather than models),
contemporary global cities. If the arrival of the Europeans profoundly changed the
Indian Ocean world, it did not destroy it completely. The decolonisation period, followed by the construction of nation-states, consolidated the nationalisation of space. 13 PORTAL, vol. 9, no. 1, January 2012. 13 PORTAL, vol. 9, no. 1, January 2012.
1 PORTAL, vol. 9, no. 1, January 2012. Time/space—world The Indian Ocean contains several historic time zones. Successive globalisations have
produced regionalisations that go back to Antiquity, to about the 4th to the 6th centuries. At that time this part of the Indian Ocean progressively entered a ‘time-world’
characterised by a variable multipolarity. What was often at stake was control of the
routes of communication and exchange. Its vastness, via numerous seas lapping
numerous bordering lands, qualify it more than any other ocean for the name of cross-
roads of civilisation, with the existence also of various fringe civilisations flourishing in
many archipelagos and islands. As a contact zone, the Indian Ocean still contains the
most significant maritime sea-lanes linking the Middle East, Africa, Asia, Europe and
America. A lot of the crude oil and related products from the Persian Gulf and
Indonesian oil wells goes through this ocean. It is not a homogenous space. Its diversity and heterogeneity is highly visible. It looks
like a transnational, transcontinental world in formation, with its inequalities, tensions,
potential wars, its cosmopolitanism, its multipolarity, its dynamism and its creativity. It
is piecemeal and fragmented, but also traversed by common itineraries; this ocean is
marked by the different temporalities found there: Malayo-Indonesian, globalisation, the
Muslim economic world, European thalassocracy, pre-European global empires, trade
and slavery, and European empires. As a commercial vector among cultures and peoples
from the earliest times, it is today undergoing a new ‘globalisation.’ The geopolitical,
cultural and economic stakes are doubled in this situation. Tensions are exacerbated by
the strong American military presence, civil wars and ethnic cleansing, environmental PORTAL, vol. 9, no. 1, January 2012. 14 14 Vergès & Marimoutou Vergès & Marimoutou Moorings degradation, demographic growth, pandemic diseases, the struggle to control natural
resources (water, forests, oil, gemstones, minerals), and entrenched identities and
religious positions. It is important to observe how changes register. Those that make
new exchange routes visible while submerging others follow a broad social logic. By
looking at modes of affirmation, legitimation and strategic identification, we can
analyse their interaction. The renewal of diasporic identities sometimes encouraged by
their nation-states of origin, new circuits of exchange and traffic, including the mafia,
should be analysed. Réunion is not sheltered from reconfigurations of power or from
ways of contesting it. Indianoceanness: Anchorage and moorings PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou Moorings creolisation demands or requires room to manoeuvre where tensions and conflicts are
resolved without being dissolved. Something has to be given up to find space for the
other, for the stranger, to share the land, the island. We explore processes of
I di
i
li
ti
i l
d t th
d
f th
ti
t
d it
t d i creolisation demands or requires room to manoeuvre where tensions and conflicts are
resolved without being dissolved. Something has to be given up to find space for the
other, for the stranger, to share the land, the island. We explore processes of
Indianoceanic creolisation on an island at the edge of the continents and situated in an
oceanic space where civilisations have experienced multiple territorialisations. Indianoceanic creolisation on an island at the edge of the continents and situated in an
oceanic space where civilisations have experienced multiple territorialisations. Bricolage and Borrowings Postcoloniality does not just refer simply to an historical period, but to a way of
rereading the world. The world is multivalent, and modernity is not the prerogative of
the West alone, with the rest of the world trying to catch up. It asserts that other
modernities exist, that mixing is inevitable, that the intercultural is a feature of any
civilisation, and that there is both conflict and exchange in the relationship between
coloniser and colonised. Postcolonial theory is sensitive to regimes of representation
and identification (masculine/feminine), the constructions of subjectivities of self and
other (orientalism, the black body, insularity) and to strategies of creolisation and
hybridisation. Without assuming uninterrupted traditions, it speaks of borrowings, Postcoloniality does not just refer simply to an historical period, but to a way of
rereading the world. The world is multivalent, and modernity is not the prerogative of
the West alone, with the rest of the world trying to catch up. It asserts that other
modernities exist, that mixing is inevitable, that the intercultural is a feature of any
civilisation, and that there is both conflict and exchange in the relationship between
coloniser and colonised. Postcolonial theory is sensitive to regimes of representation
and identification (masculine/feminine), the constructions of subjectivities of self and
other (orientalism, the black body, insularity) and to strategies of creolisation and
hybridisation. Without assuming uninterrupted traditions, it speaks of borrowings,
bricolage and reformulations. It is in itself a theory based on the idea of borrowing,
drawing ideas both from local expressions and practices and Western thought. It is wary
of totalising discourses. It recognises that several different temporal regimes can coexist
in a space layered with different territories. It does not oppose ‘tradition’ and
‘modernity,’ rather it asserts that a situated subject constantly negotiates the interaction
between traditions and modernities. The gaze is necessarily decentered within and
among transnational, transcontinental and diasporic movements, as these are translated
in the imagination and the praxis of artists. Post-colonial theory does not forget that the
world is beset by conflicts, wars and inequalities. age and reformulations. It is in itself a theory based on the idea of borrowing, drawing ideas both from local expressions and practices and Western thought. It is wary
of totalising discourses. It recognises that several different temporal regimes can coexist
in a space layered with different territories. PORTAL, vol. 9, no. 1, January 2012.
1 Indianoceanness: Anchorage and moorings We want to suggest an Indianoceanness that comprises both anchorages and moorings. We highlight the metaphor of anchorage because it helps us think about exile and
displacement, movement and flux, without forgetting about the territory we have left. We want to work with an identity that is anchored yet travelling (concretely or through
the imagination), marking or recognising routes and itineraries where exchanges and
meetings happen. The reappropriation of territory liberates the imagination, allows us to
take leave without fear or sorrow and to set sail. The island remembers its continents. We see a to-and-fro movement, a hither and thither, between continents and the island,
between the island and the world of islands. The presence of the horizon means that one
cannot forget what is over there beyond it. The horizon, that which is not yet known,
that which arrives, the unpredictable, the unexpected, that is to say history. This
geographical line is the metaphor for our political horizon, which is always subject to
modification, to new contradictions, new conflicts, new challenges. This horizon that
tricks the eye by appearing curved is a good metaphor for our position: the horizon
recedes, the curve approaches. Indianoceanness is not just cultural, or rather it
recognises the cultural as an element of geopolitics and economics. Our island, on the Asia-Africa axis, has been a crossing point of different economies
and world-cultures. It is a space shaped by the successive territorialisation practices that
cross each other, destroy each other, get mixed up and reordered. Indian Ocean
creolisations are always being reworked, they are never finished. Its dynamics are
controlled by negotiation, as things necessarily get lost or relinquished. There is no
creolisation without loss, just as it cannot happen without inequality because PORTAL, vol. 9, no. 1, January 2012. 15 15 Vergès & Marimoutou Boris Gamaleya, Vali pour une reine morte Boris Gamaleya, Vali pour une reine morte Postcoloniality I salute you my queen
[…]
from the holds a black wind opening up the Asian dawn
a nova on the portulan of my oceanic island I salute you my queen
[…]
from the holds a black wind opening up the Asian dawn
a nova on the portulan of my oceanic island 6 Union of Réunion Women. Founded in 1958 and closely linked to the Réunion Communist Party, this
organisation was the spearhead of a great number of social movements. g
p
g
7 The Association of Réunion Writers. With other organisations, like the Réunion Cultural Movement, th
Mouvman pou nout droi kozé, les Editions des Chemins de la liberté, this group took up the cause of
Réunion literature and the Creole language.
8 g
g
8 The Union for Réunion Broadcasting. An association of writers with a long history in the créolie
movement, which was initiated by Gilbert Aubry and Jean-François Samlong, beginning with the
neologism forged by Jean Albany.
9 Bricolage and Borrowings It does not oppose ‘tradition’ and
‘modernity,’ rather it asserts that a situated subject constantly negotiates the interaction
between traditions and modernities. The gaze is necessarily decentered within and
among transnational, transcontinental and diasporic movements, as these are translated
in the imagination and the praxis of artists. Post-colonial theory does not forget that the
world is beset by conflicts, wars and inequalities. We date ‘post-coloniality’ in Réunion from the beginning of the 1960s. The colonial
order was breaking up. Voices were raised to assert the solidarity of the Réunionese
people with other Indian Ocean peoples, with social movements in France, to affirm the PORTAL, vol. 9, no. 1, January 2012. 16 Vergès & Marimoutou Moorings 7 The Association of Réunion Writers. With other organisations, like the Réunion Cultural Movement, the
Mouvman pou nout droi kozé, les Editions des Chemins de la liberté, this group took up the cause of
Réunion literature and the Creole language.
8 The Union for Réunion Broadcasting. An association of writers with a long history in the créolie
movement, which was initiated by Gilbert Aubry and Jean-François Samlong, beginning with the
neologism forged by Jean Albany.
9 The General Union of Réunion Workers in France. This union had a long history of federating student
and progressive worker unions of people from Réunion in France, especially between the 1960s and the
1990s. g
g
y
y
9 The General Union of Réunion Workers in France. This union had a long history of federating student
and progressive worker unions of people from Réunion in France, especially between the 1960s and the
1990s. 10 The maloya is song and dance coming from slave and indentured culture, practised either at mystical
ceremonies celebrating the ancestors (servis kabaré, servis makwalé), or in more profane versions on
musical evenings (kabar). While the maloya was marginalised for a long time, it was revived and
celebrated by the Réunion communist party militants and officials at the end of the 1960s.
11 A Game-show host [trans.] Vergès & Marimoutou une colonie française or La Réunion, département français belied the toned-down
image of ‘the little France of the Indian Ocean.’ None of this was easy. The repression
was vicious: censorship, discrimination. Conservative newspapers, radio and TV never
reported what was going on, attacks never ceased. A little reminder: the bicentennial
celebrations of colonisation in 1965 excluded the Madagascan inhabitants; they refused
to broadcast the maloya,10 celebration of a ‘white’ Réunion, and frankly disdained the
Creole language. une colonie française or La Réunion, département français belied the toned-down
image of ‘the little France of the Indian Ocean.’ None of this was easy. The repression
was vicious: censorship, discrimination. Conservative newspapers, radio and TV never
reported what was going on, attacks never ceased. A little reminder: the bicentennial
celebrations of colonisation in 1965 excluded the Madagascan inhabitants; they refused
to broadcast the maloya,10 celebration of a ‘white’ Réunion, and frankly disdained the
Creole language. 10 The maloya is song and dance coming from slave and indentured culture, practised either at mystical PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou existence of a Creole language, and a culture moored to Africa, Asia, Europe and the
other islands of the Indian Ocean. Intellectual history merged with political demands
that countered an hysterical campaign for ‘Frenchness.’ The hegemonic discourse of
Frenchness took the line that Réunion had no history, culture or language; that the
inhabitants could not count as a people; only France could give it meaning, identity and
existence. We could not exist without a France that was, as conservatives politicians
proclaimed, ‘our sun.’ The women and men of Réunion countered this cultural attack using the vocabulary of
decolonisation, dignity and identity. They affirmed that our language exists, that we
have a culture, a history that could not be subsumed to the French (‘hexagonal’) one,
and even if we were connected to France, we were not reduced to it. Indeed, events
might have taken place that had a greater effect here than there. Obviously the abolition
of slavery and the mass importation of indentured workers had no impact in France:
they are still not fully inscribed in French historiography. Social and cultural
movements emerged and began to spawn innovative and creative action. From the
1970s through the 1980s, journals (Les Cahiers de l’île de la Réunion, Bardzour), poets
(Boris Gamaleya, Alain Lorraine, Gilbert Aubry, Agnès Guéneau, Alain Armand,
Patrice Treuthardt), novelists (Alain Cheynet, Axel Gauvin, Daniel Honoré, Jean-
François Samlong), musicians (Firmin Viry, Lo Rwa Kaf, Granmoun Lélé, Danyèl
Waro, Ziskakan), cultural and political activists (Laurent Vergès, Firmin Lacpatia,
Reynolds Michel), researchers working on language, history and anthropology, created
the broad outlines to think the foundations and the modes of expression of Réunionese
cultural, historical, and social identities. There were movements (the UFR,6 the
Southern Cultural Front, the Réunion Cultural Movement, the ADER,7 the UDIR,8 the
UGTRF9) and places (Farfar at La Rivière-des-Galets). All brought together eyewitness
testimonies, offered meeting places and drafted demands. Books such as Réunion 69, 17 PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou Moorings The maloya is song and dance coming from slave and indentured culture, practised either at mystical
ceremonies celebrating the ancestors (servis kabaré, servis makwalé), or in more profane versions on
musical evenings (kabar). While the maloya was marginalised for a long time, it was revived and
celebrated by the Réunion communist party militants and officials at the end of the 1960s.
11 A Game-show host [trans ] musical evenings (kabar). While the maloya was marginalised for a long time, it was revived and
celebrated by the Réunion communist party militants and officials at the end of the 1960s.
11 A Game-show host [trans.] 12 Métissage. Batar in Réunion Creole does not have the same connotations as bâtard [bastard] in French PORTAL, vol. 9, no. 1, January 2012.
1 Vergès & Marimoutou Moorings between ‘us’ and ‘them’ was even picked up by the anti-colonialist movement, opening
the way for the emergence of an ‘us’ from Réunion, which played two roles: one to try
to get away from the two-bloc logic (campaign for the ‘Indian Ocean Zone of Peace’),
and the other to try get recognition for the fact that on Réunion there were networks and
practices more complex than the simple Creole-French opposition. The cultural movement of the last decades mobilised the society. Vernacular music
became a space of resistance. We rediscovered our voices in literature, art, theatre,
music and poetry. It was a period of militancy and engagement. The cultural debate was
turned around and forever changed. It was a remarkable time of creativity and
dynamism. It was possible to celebrate exile and métissage, and the Creole language
was strengthened. The stakes in present debates cannot be understood unless one goes
back to the economics of the dissent, discontent and disturbances of those years. Now
we have the task of a postcolonial/sociological rereading of the politics and culture of
those years to reinsert the complexity of the real. For instance, we think that the
opposition of French and Creole languages ignores the presence of other languages like
Gujarati, Chinese, Malagasy, Swahili, Tamil, Urdu or their mixtures on the island; in
Réunion there exist languages. I have no need to look, myself
the pot of my identity
overflows
with my mixing
I have no need to look, myself
the container of my identity
overflows
Danyèl Waro, Métissage I have no need to look, myself
the pot of my identity
overflows
with my mixing
I have no need to look, myself
the container of my identity
overflows Danyèl Waro, Métissage A mythic France The assertion of abstract and ahistorical Frenchness sought to alienate (in the Fanonian
sense: to despise oneself) Réunion society, relegating our dreams, our lives, to non-
existence. The ‘France’ that came our way was mediocre, stupid and sub-cultural. We
saw nothing of the New Wave cinema, nor the poetic, literary or dramatic avant-garde,
nor the renewal of thought around structuralism and post-structuralism, nor the critical
Marxist debates, nor contemporary art, nor the debates in the human and social sciences,
nor psychoanalysis, nor the sexual revolution, nor the women’s movement, nor the gay
movement … these things were blocked. Not to mention what was going on in Africa
and Asia. It was forbidden to talk of torture in Algeria, to do so led to prosecution. Civil
servants who led trade unions, who joined anti-hegemonic cultural and political
movements were punished and children deported by the thousands. We only got a
dumbed-down mass culture: Guy Lux,11 Connaissance du monde tours, B-grade movies,
photo-novels … it was a France of mythic, provincial, timid and inward-looking
dimensions. Even if its domination was not completely hegemonic, it was given
nonetheless considerable powers of distribution and legitimation by the state. The world
was read according to the binary logic of the Cold War, of the battle between the ‘free
world’ versus the ‘communist world.’ This was not without its effect in Réunion, which
also had its binary structures hard-wired into its society (master/slave,
master/indentured, colonialist/anti-colonialist). This concatenation of binaries brought
b
i h
di
h
k d
i l
d
l
l
l
i
i h master/indentured, colonialist/anti-colonialist). This concatenation of binaries brought
about a Manichean discourse that masked social and cultural complexity. Manichean
discourses became the dominant explanation of life on Réunion. A strong opposition PORTAL, vol. 9, no. 1, January 2012. 18 Vergès & Marimoutou Rereadings and reinterpretations The 1960s and 1970s movements have been a source of references for our thinking ever
since. Batarsité,12 Créolie, banyan people, coral people … ideas have emerged to define
our presence. The official celebration (since 1981) of the 20th December—the date of
the liberation of the slaves—has opened the way for a rereading of history. We have
spoken once again of slavery, of our forebears who were enslaved or indentured. We
have dared to denounce crimes and lies. No, slavery was not ‘softer’ on Réunion than in
the Caribbean. Yes, we are all inheritors of this history. No, the slaves were not just 12 Métissage. Batar in Réunion Creole does not have the same connotations as bâtard [bastard] in French. 19 19 PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou Moorings Africans. Yes, there was resistance—maroons, abortion, suicides, revolts, poisoning of
masters, sabotage. No, it was not all resistance, slavery and indenture also succeeded in
breaking peoples’ will and in their dehumanisation. Yes, indenture work was brutal,
perhaps as much as slavery. No, Réunion was not a Garden of Eden. Yes, there was
racism. Historians—Michèle Marimoutou, Prosper Eve, Claude Wanquet, Sudel
Fuma—and economists—Hai Quang Ho—are reconstructing this period. The airplane, the biggest Boeing ever, lands
at the huge aerodrome at Grillot. When it arrives, this airplane
makes a noise like hell, and when I speak softly,
some say that I am noisy. Christian Jalma, Le Pouvoir éphémère des lapsus The airplane, the biggest Boeing ever, lands
at the huge aerodrome at Grillot. When it arrives, this airplane
makes a noise like hell, and when I speak softly,
some say that I am noisy. PORTAL, vol. 9, no. 1, January 2012.
2 13 A term of uncertain etymology, designating people born outside of Réunion (most often in continental
France), residing in and/or working on the island. PORTAL, vol. 9, no. 1, January 2012.
2 Vergès & Marimoutou Vergès & Marimoutou Moorings The vocabulary of complaint and accusation is far too limiting, while assimilationist
language has worked hard to deprive us of a sense of self. And if the anti-colonialist
language opened us to the world and led us to tie together forms of solidarity once again,
it also induced a habit of simplifying the facts. Face to face with the violence and racism
of a rigid society, with the denial of justice (electoral fraud, discrimination), we learned
to be on the defensive. Today we will have to forge new alliances, leave behind
simplistic approaches, and recognise complexity. Christian Jalma, Le Pouvoir éphémère des lapsus Christian Jalma, Le Pouvoir éphémère des lapsus Yet, we were still too much a people ‘spoken’ through others. Our space has been filled
with representations in which we didn’t recognise ourselves. The French are telling us:
‘We will explain to you who you are.’ A whole series of contradictory statements are
blocking us in, spatially. We are at the same time always too much of this (violent,
mute) and not enough of that (responsible, hard-working). We are ‘The France of the
Indian Ocean’ yet ungrateful. This double-bind situation of contradictory injunctions (in
the psychoanalytic sense, as an ‘effort to make the other mad’) is a feature of
colonisation. As long as one is the stereotypical ‘Creole’ it is acceptable: charming,
singing, dancing, ‘nice,’ but not ‘angry’ and acting. This brings about a paradoxical
situation where the subject is drawn into conforming to the stereotype the better to be
indicted for following it. We affirm that it is preferable to assert our autonomy, to reject
the slick touristy image of the ‘intense’ and ‘mixing pot’ island, in order to bring out the
rough edges, the conflicts along with the wonders; life as a whole, basically. Still,
negative stereotypes are long-lived and even in 1985 one was able to read this
interpretation: ‘The simplified gestures of miming, as symbols of expression appearing
in the two dance types, maloya and sega, and intellectual introversion as refuge of the
self, underpin the relations that Réunion people have with the outside world’ (‘Report
de la Commission traitant des violence intra-familiale’ 1985). Our voices are regularly interrupted and our silences over-interpreted; we scarcely begin
to say something and the gag is put in place. We have learnt to speak in an oblique or
indirect fashion. We will have to invent a conceptual vocabulary to speak our world. PORTAL, vol. 9, no. 1, January 2012. 20 20 Vergès & Marimoutou We are not pure deserts Agnès Guéneau, La Réunion: une île, un silence PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou Vergès & Marimoutou Vergès & Marimoutou We have to know how to perceive the ‘small’ or the ‘minor,’ the scarcely visible
manifestations of adaptation to change. The urban-rural dichotomy will henceforth have
less meaning, since the urban fabric and its representations dominate, carving out
territories that are interlaced but also confront each other. An urban sociology is called
for. How did the inhabitants acquire public housing? What are the new urban cultures,
or the modern semiotics of belonging to one group or another? What is the language of
young people? What are the signs of belonging? What are the inhabitants’ uses of
spatial strategies? What have the thousands of Comorans, and their children, now
Réunionese, brought to Réunion creolisation? What have the zoreys brought? What are
the new social networks? Who is doing what, and why? How is knowledge and social
status being transmitted? How are masculinity and femininity structured? We need an
anthropology of the political, of ‘how we live together.’ Who is speaking for whom? Who is being heard? How does one become ‘Réunionese’? Certainly the work of young
researchers is beginning to build new knowledge, but we sense a lack of transfer, of
fruitful exchanges, of a useful and vigorous public debate. Obstacles exist. We witness a campaign whose goal is to put a stop to new
developments and to impose a particular way of thinking. Here and there histories are
rewritten so that past tensions and conflicts are erased or relegated to ‘prehistory.’
Medias (newspapers, radio, TV), the university, the schools are tools of pacification,
which is threatened by any appearance of cultural diversity (understood as
communitarian threats). The aim of this rewriting and revisionism is to make people
forget their responsibilities towards each other. However, archives (both material and
immaterial) inevitably reveal, and are revealing, multiple forms of complicity in brutal
repressive strategies of silencing. Preaching about novelty is also a way to mask the fact
that novelty is a new disguise for past repressive measures. My country is a crazy ship
Where are they taking us? Axel Gauvin, Chants pour ouvrir la langue et le Coeur My country is a crazy ship
Where are they taking us? Axel Gauvin, Chants pour ouvrir la langue et le Coeur Axel Gauvin, Chants pour ouvrir la langue et le Coeur Agnès Guéneau, La Réunion: une île, un silence Today we acknowledge the need for a new movement, the need to think up new
approaches and ways of seeing. We have to take into account the profound
transformations that have taken place over the last twenty years. They demand new
concepts and a new methodology. Tens of thousands of unemployed, thousands of
young people out of work, almost half the population on social benefits, the hegemony
of oil and cars, bad city planning, spoilt environment, a daily life of domestic violence,
rising delinquency, disappearing agriculture, 753,600 inhabitants of whom 40 percent
are under 25, the region of France with the highest unemployment (31 percent), nearly
40 percent of the population below the poverty line, the continual rise of people with the
right to access unemployment benefits (67,915 enrolled); rising numbers of women
working outside the home, domestic violence, rape, incest, drugs, large numbers of
suicides (especially young men, and especially in a region—the south of the island—
famous for having ‘kept its traditions,’ for its ‘authenticity’), alcoholism, murderous
road accidents. These are some indication that Réunion society has problems. To this
we can add the rapid changes taking place in the region and globally. Yet these facts
and data do not tell us everything. What we need is a sociology of these transformations,
of the effects on society of the number of active women, of the constant rise in the
proportion of zoreys,13 of the presence of mobile and other new forms of
communication, of the car, of television, of the number of students, of the emergence of
rap and hip-hop in Réunion. communication, of the car, of television, of the number of students, of the emergence of
rap and hip-hop in Réunion. 21 Moorings Vergès & Marimoutou suffer from becoming uprooted. They are disillusioned, and from this disillusion, they
extract a knowledge of what should be ‘good’ for Réunion. These are yaka fokon (must
do) experts, setting up shop for their grievances in the media. Often sexist, they speak
freely of the offended or humiliated ‘Réunionese man’; there are no women in their
universe. No fathers in their world, not to mention mothers. They speak of a long-lost
past so the present can be pasted over and the future made to look rosy. Trading on
others’ struggles in which they did not take part, they take few risks. They believe that
‘speaking is doing’; speaking, speaking, holding forth, these ‘kings of the word’ are free
with words, but refuse to recognise that they are confounded by complex situations. They have solutions; they ‘know.’ They hide their power behind a low-risk populism,
going about accusing others of being ‘intellectuals’ cut off from ‘the people.’
‘Intellectual’ is even an insult. Over the last few years, a set of rhetorical strategies have been presented as ‘truths.’ It is
important to deconstruct them in order to make visible the conditions under which these
opinions are transformed into incontrovertible truths, into commonplaces. We have to
understand how and why some of these expressions are picked up by the media, Over the last few years, a set of rhetorical strategies have been presented as ‘truths.’ It is
important to deconstruct them in order to make visible the conditions under which these
opinions are transformed into incontrovertible truths, into commonplaces. We have to
understand how and why some of these expressions are picked up by the media,
circulate, and become dominant, why a whole series of opinions are transformed into
points of reference, the commonplaces of common sense, so that one no longer even
asks where they have come from, and who created them. The function of these opinions
is to construct a ‘truth effect.’ One of these classical strategies consists in stereotyping
adversaries and denigrating them. In the public space that should be open to the
agonistic encounter of diverse positions, insult and defamation move in. Demonising the
adversary seeks to take away his or her autonomy in action and thought. Adversaries of
hegemonic thought are presented as being controlled by forces greater than themselves
(conspiracies abound), always suspected of having secret motivations. PORTAL, vol. 9, no. 1, January 2012.
2 PORTAL, vol. 9, no. 1, January 2012. Conservatisms, fears Bitterness and resentment characterise too often the attitudes and discourses of
intellectuals from Réunion. In their opinion, communitarianism is a danger, the
politicians are not up to it, people are alienated by supermarkets and consumerism and PORTAL, vol. 9, no. 1, January 2012. 22 Vergès & Marimoutou Moorings 14 Malbars are people in Réunion who recognise themselves as having, or going back to, a South Asian
origin (Bengal or southern India), or who follow the rituals linked to the forms of Hinduism in Réunion. PORTAL, vol. 9, no. 1, January 2012.
2 Vergès & Marimoutou This anti-
intellectual populism is the inheritance of a world where might created right. One often
hears, for the most part from the middle classes, ‘This is too intellectual, people don’t
want to hear that,’ which tells us a lot about the kind of opinion they hold about those
who are not allowed to think. This current ‘caste,’ the petit bourgeoisie, often civil
servants, gained social and economic status thanks to the union struggles of the 1950s
and 1960s (from which it is still reaping the benefits) and thanks to the French state
handing over public responsibilities to them. This caste (rather than a ‘class’ since it
does not even fight for its rights, it begs the State to protect them) gives itself all too circulate, and become dominant, why a whole series of opinions are transformed into
points of reference, the commonplaces of common sense, so that one no longer even
asks where they have come from, and who created them. The function of these opinions
is to construct a ‘truth effect.’ One of these classical strategies consists in stereotyping
adversaries and denigrating them. In the public space that should be open to the agonistic encounter of diverse positions, insult and defamation move in. Demonising the
adversary seeks to take away his or her autonomy in action and thought. Adversaries of
hegemonic thought are presented as being controlled by forces greater than themselves
(conspiracies abound), always suspected of having secret motivations. This anti-
intellectual populism is the inheritance of a world where might created right. One often
hears, for the most part from the middle classes, ‘This is too intellectual, people don’t
want to hear that,’ which tells us a lot about the kind of opinion they hold about those
who are not allowed to think. This current ‘caste,’ the petit bourgeoisie, often civil
servants, gained social and economic status thanks to the union struggles of the 1950s
and 1960s (from which it is still reaping the benefits) and thanks to the French state
handing over public responsibilities to them. This caste (rather than a ‘class’ since it
does not even fight for its rights, it begs the State to protect them) gives itself all too PORTAL, vol. 9, no. 1, January 2012. 23 Vergès & Marimoutou Vergès & Marimoutou Moorings easily the legitimate right to be the spokesperson of the people, a people that is being
constructed as generic, rather than diverse and complex. At the centre of this rhetorical
machinery, ‘the truth’ is the obsessively marked term that is able to mask the conditions
of its own production. For instance, a ‘truth’ was that Blacks were lazy, Malbars
thieves,14 Réunion women lascivious and men violent. The traces of racism and the
denial of difference still impact upon the society. Another common rhetorical strategy: ‘Africa’ refers to a homogenised, stereotyped and
racialised space. Africa is a space of catastrophe, victimisation, misery, purged of any
content; the artistic, intellectual, economic and philosophical production of a rich and
diverse continent, along with the productive conflicts and the philosophical and legal
contributions, are wiped out. In May 2004, an open letter was published in the Réunion
press that claimed that ‘The African civilisation has lost everything.’ Now, for one,
there are African civilisations, and two, extremely alive cultures. Think, for instance, of
the great writers: Hampâté Bâ, Amadou Kourouma, Wole Soyinka, Nadine Gordimer,
Naguib Mahfouz; of the great filmmakers: Med Hondo, Youssef Chahine, Cheick
Oumar Sissoko; of the philosophers and social science researchers: Jean Godefroy
Bidima, Achille Mbembe, Ato Quayson, Rehanna Vally; of the artists: Ousmane Sow,
Sotigui Kouyaté, Fela, Youssou N’Dour; of Nelson Mandela. Such a statement flies in
the face of all the decolonisation struggles, as well as the contributions to international
law, to theories and cultural emergences. At the beginning of 2004, South Africa celebrated ten years of democracy. Now without
diminishing in any way the problems this country has been facing, like the rest of the
nations of Africa, and with respect for the peoples of Africa and for what they have
achieved, in admiration for their artistic and cultural expression, we reject totally the
idea that ‘an African civilisation would have lost everything.’ Africans have deemed
this opinion racist, as they have denounced the ‘Afro-pessimism’ that banished them to
the void so that Western superiority could settle in more comfortably. We are talking
about African worlds whose creolised traces have contributed and are continuing to
contribute to the richness of Réunion society. 24 Vergès & Marimoutou PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou Vergès & Marimoutou Moorings The paternalist position towards the African continent (rich, complex, multiple) goes
hand in hand with the opinions of those who refuse to see current mutations. Too often,
in the peoples of the region, in the civilisations that have given birth to the groups that
constitute the peoples of Réunion island, Réunion paternalists see only minor
civilisations, only minor peoples among the people. When they recognise their
connections to the continent, they rarely acknowledge its contribution to Réunion
culture. The history of the island is recent, its culture is young, they say, and its survival
will depend entirely on its being moored to France. They have a nostalgia for tan lontan
[a long time ago], they are fearful of a new, more complex, more brutal world, of young
people who ‘have no respect,’ and they hide under the wing of ‘Mother France.’ They
seek its parental protection, worried about having to take risks or assume
responsibilities. They do not even notice that France has deeply changed, that she
herself has been subject to profound transformations over the last fifty years, that she is
facing up to new facts, and to unavoidable social and political changes. Because of these
changes, it is more important than ever that we make our voices heard. We are a society
born in the first globalisation (slavery and colonisation), and from the beginning the
society was multi-religious and multi-ethnic. (We will not digress on the demographic
use that is made in Réunion of the term ‘ethnic’; today anthropologists have agreed on
the use of the term ‘ethno-cultural constitutions’). We can explain how we have
managed, how we have made do to live together in this little land. We do not embrace
any kind of idealism. We know that ethnic stereotypes, racist insults, and racialised
fears exist, but we also know that Réunion would not be Réunion without the
confrontation of differences. The idea of confrontation implies a cultural and political
debate; a cultural debate that is not frightened of the political. Let us underline this point
for all those who would like to put aside politics so as to impose a demagogic populism
(‘the elites are all corrupt’), where it is ‘every man for himself,’ where anonymous
denunciations are encouraged, rumour is spread over the radio, in the papers, while
mediocrity, sexism and racism is the order of the day. Breakaways from the institutionalisation of culture Breakaways from the institutionalisation of culture
In the 1980s, the French State had put quite a bit of money into decentralised cultural
structures. These policies had positive and unintended effects. As numerous artists and
art critics have noted, massive institutionalisation and museumification have the
tendency to empty the political content, in the broadest sense, out of culture, and lead
artists for the most part to focus entirely on the market. This institutionalisation has
muddied the waters on the question of moorings. If it is quite legitimate for an artist to
want to sell her art, to be promoted and therefore enter the circuits of trade, or follow
trends, the question of the relations of culture and politics, art and society, should be
debated. The development of a critical space would be a good start. During a debate
with artists on the 3rd May 2003, the problem of a ‘Réunion art’ was raised. Among the
statements produced were: ‘Wouldn’t the term ‘Réunionese’ restrict the artist’s
expressive scope? Enclosing him or her in a space, whereas they may want to aspire to
the universal? Wouldn’t ‘Réunionese art’ revert to a ‘regionalism’ in which people
would necessarily look for ‘authenticity’?’ Numerous philosophers, historians and anthropologist have reminded us that the notion
of the universal was invented in Europe (as the heritage of secularised Christianity after
the 18th century), which then led to the imposition of the equation Europe = Universal. Europe’s singularity was hidden behind its universalist mask. If, after the second half of
the 20th century, this analogy has been radically thrown into question, its continued use
needs clarification. There is no doubt that the notion of the ‘universal’ is often used to
go beyond things that appear too localised, to contain a space where the differentiations
are fixed, leading to a ‘balkanisation,’ an ‘ethnicisation’ of identities. In the domain of
art, it will be useful to be more precise about the Réunionese contribution to modernity. In Réunion, the republican secular education system imposed the European notion of
universality and modernity. One might ask oneself how the term came about in an
historical context that denied the universal to a large part of the world’s population. What is the Réunionese contribution to the artistic and theoretical work towards new
definitions of universality and modernity? What kinds of artistic works would answer to
this context? Vergès & Marimoutou Gilbert Aubry Rivages d’alizé 25 PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou PORTAL, vol. 9, no. 1, January 2012.
2 PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou Moorings Vergès & Marimoutou Breakaways from the institutionalisation of culture Breakaways from the institutionalisation of culture p
g
16 Michel Debré, a Parisian minister, arrived in Réunion in April 1963, and succeeded in being elected
Député for Saint-Denis on May 6 despite local opposition to the Ordonnance Debré law he had
introduced in 1960, that allowed Civil Servants in the overseas departments and territories of France to be
put into forced exile on mainland France if suspected of disturbing public order. Supported by those who
rejected autonomy, he immediately became the leader of the local right wing. 15 ‘Lend me two words,’ from Axel Gauvin, Chants pour ouvrir la langue et le coeur.
16 Anprèt amoin dé Mo Anprèt amoin dé Mo Axel Gauvin Romans po détak la lang démay lo kèr15 PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou Vergès & Marimoutou Moorings Robér, William Zitte, Gabrielle Manglou; and in the area of new expressive forms like
rap and hip hop, slam, cartoons, video. That said, there is no school of criticism, and it
remains forbidden to suggest a debate on aesthetics. All expressions, all creations
should be celebrated on their own terms. Jealousy, envy and resentment have
contributed to the creation of little feudalities. Yet, new forms of artistic expression also
bring into play what there is to be thought about, and to be translated into concepts and
actions. Young architects, social workers and psychologists have accumulated
knowledge on new sites of negotiations. Scattered knowledges are the foundations of a
new methodology for coming to terms with Réunionese society. ,
,
p
g
16 Michel Debré, a Parisian minister, arrived in Réunion in April 1963, and succeeded in being elected Breakaways from the institutionalisation of culture Over the last few years, artists have explored new arrangements. Photographers—Laurent Zitte, René-Paul Savignan, Frédéric Pothin, Yoyo Gonthier;
plastic artists—Esther Hoareau Colette Pounia Thierry Fontaine Alain Padeau André Photographers—Laurent Zitte, René-Paul Savignan, Frédéric Pothin, Yoyo Gonthier;
plastic artists—Esther Hoareau, Colette Pounia, Thierry Fontaine, Alain Padeau, André PORTAL, vol. 9, no. 1, January 2012. 26 nd me two words,’ from Axel Gauvin, Chants pour ouvrir la langue et le coeur. Vergès & Marimoutou Moorings literature is haunted by the idea of trying to making the space habitable, taking account
of history, speaking this place and this history, as it dreams of the missed meeting
between the ‘white’ and the maroon. The real was there, and no one knew how to talk
about it. Literature goes back, in its encounter between history and place, to the
inventory of its phantoms and fantasies. The always-fictionalised history of the origins of the encounter-as-living-together, of
the very origins of the island, is always set up as a disaster about to happen, as an
unaccomplished possibility, as original sin, of the primal scene as failure. The dream of
paradise in the offing, prior to the occupation of the space, transforming it into a non-
place, is a recurrent effect in Réunionese literature like a pre-origin myth that would
come instead of, and in place of, the impossible (or the unthinkable) origin myth. This
way of presenting the island as a paradise allows it to be situated outside of the history
being made there, and to be thought beyond conflicts, in the classical framework of
colonial paternalism. The island is thought of through the figure of the marvellous, but
above all it is constructed as a foundation myth that puts ‘the Creole’ (white settler) at
the origins of the island and the world, lending that figure a legitimacy that history
would not acknowledge. The fantasy of a pre-human paradise is immediately outweighed by the fantasy of a
scorched earth, of a hellish space. If the island is both hell and heaven, even before man
set foot there, the founding narrative of place will have to take account of this: there is
really no purity to be rediscovered. If Réunionese literature allows itself to be the
space/place of a fictional memory that takes the place of an unthinkable mythic or
historical memory, it is emerging also, at the same time, from another memory, which is
that of the texts and of the languages that are considered forgotten, and which are only
awaiting for a space in which to be heard again. 17 My road is not smoothed/It is not difficult Mikaèl Kourto, Up There. The literature of absence The literature of Réunion has always said one thing: there ‘is no history because there is
no place.’ Or more precisely: ‘there is no appropriation of history because the place is
uninhabitable’; there is ‘no tongue nor language which can really convey this history or
this place, because neither tongue nor language are inhabited by the place or this history,
nor really do they live here in fact.’ This is where the challenge lies: to take charge of
the place and the history, and the languages of the place and the history. Where is the
fiction in trying to come to terms with the history and the place, unless it is through
traces, through ghosts? How can one live in a land of migrants? How can one live when
one is a migrant? Debré, with his willing and systematic denial of history and space, sought to erase
Réunion’s memories.16 The film Sucre Amer [dir. Yann Le Masson, 1963] is worth
looking at again and again: the dead are voting, and as for the living, they get nowhere
near the ballot box. Let’s have a look at the literary status of the slave: a non-person, a
spectral figure; that of the maroon, condemned to wander, like a shadow; that of the
indentured labourer, that of the petit Blanc who successively confronts the maroon, then
the uninhabited that remains uninhabitable, because in history they never met. The PORTAL, vol. 9, no. 1, January 2012. 27 Vergès & Marimoutou PORTAL, vol. 9, no. 1, January 2012. Monstrous and human The revolution in thought and culture that began around the world in the 1960s, with
social, feminist and gay movements that questioned heterosexual and dominant
masculinity, arrived in Réunion twenty years later disconnected from these social
movements. Conformity to the colonial norms and taboos applied by religious
orthodoxy or their insular versions (prohibition on masculine homosexuality, fear of
feminine sexuality) impacted on the forms of emancipation. Even today it is extremely
difficult for a Réunionese homosexual to be out in a society that caricatures, fears or
excludes him or her. The reality of Réunionese homosexuality may be seen in the
personal columns or on the radio, but there is still no social space to express it. Heterosexuality, on the other hand, stills run on masculinist desire and its
representations, and the relations between men and women are structured by predatory
violence. How else can we understand conjugal rape, the rape of infants, of young and
old women? How else can we understand incest? When a father rapes his baby, or a
young man rapes a women forty years older than him, the violence is amplified by
intergenerational confusion. The revolution in thought and culture that began around the world in the 1960s, with
social, feminist and gay movements that questioned heterosexual and dominant
masculinity, arrived in Réunion twenty years later disconnected from these social
movements. Conformity to the colonial norms and taboos applied by religious
orthodoxy or their insular versions (prohibition on masculine homosexuality, fear of
feminine sexuality) impacted on the forms of emancipation. Even today it is extremely
difficult for a Réunionese homosexual to be out in a society that caricatures, fears or
excludes him or her. The reality of Réunionese homosexuality may be seen in the
personal columns or on the radio, but there is still no social space to express it. Heterosexuality, on the other hand, stills run on masculinist desire and its
representations, and the relations between men and women are structured by predatory
violence. How else can we understand conjugal rape, the rape of infants, of young and
old women? How else can we understand incest? When a father rapes his baby, or a
young man rapes a women forty years older than him, the violence is amplified by
intergenerational confusion. Beyond the judicial and moralising discourses on crime, why is a woman’s (or
sometimes a young boy’s) body nullified? PORTAL, vol. 9, no. 1, January 2012.
2 Vergès & Marimoutou Vergès & Marimoutou The ethics of solidarity Mon shemin lé pa galizé
Lï lé pa malizé
Mikaèl Kourto, Lao17 Mon shemin lé pa galizé
Lï lé pa malizé Mon shemin lé pa galizé
Lï lé pa malizé Mikaèl Kourto, Lao17 28 PORTAL, vol. 9, no. 1, January 2012. Moorings Vergès & Marimoutou PORTAL, vol. 9, no. 1, January 2012. Masculine, feminine To this analysis of man/woman relations in Réunion, we have to add that of the
encounter of two forms of masculinity: the one which develops in a colonial world
where physical strength gave the Réunionese man his social status (working hard,
working the land), and the other that arrives with the economic and social To this analysis of man/woman relations in Réunion, we have to add that of the
encounter of two forms of masculinity: the one which develops in a colonial world
where physical strength gave the Réunionese man his social status (working hard,
working the land), and the other that arrives with the economic and social
transformation brought about when office work and the mastery of French began to take
over from the world of rural labour and its values. With unemployment, large numbers
of men have lost their social position. How can we expect a man to occupy a parental
place while at the same time the economy deprives him of all the symbols related to this
function? Working men’s experiences are not found in school or in the media; the
physical strength that made his world is no longer valued. He reacts to this loss with a
compensatory violence. transformation brought about when office work and the mastery of French began to take
over from the world of rural labour and its values. With unemployment, large numbers
of men have lost their social position. How can we expect a man to occupy a parental
place while at the same time the economy deprives him of all the symbols related to this
function? Working men’s experiences are not found in school or in the media; the
physical strength that made his world is no longer valued. He reacts to this loss with a
compensatory violence. Mon vi I vol o van l’an par l’an
Somin domin
In ot kalité
Domoun
Barbara Robert, Fannfoutan18 Mon vi I vol o van l’an par l’an
Somin domin
In ot kalité
Domoun Barbara Robert, Fannfoutan18 18 ‘My life flies away with the wind year after year/the roads of the tomorrow/another kind/of people.’
Barbara Robert, Joking. PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou Vergès & Marimoutou Moorings But it is not only war situations that have rape at the borders of masculinity, it is also
where speech no longer takes place, where sexuality (contradictory, under the spell of
the unconscious, of desires) disappears in favour of predation or gang-bangs. It is the
symptom of a crisis in masculinity where ‘being a man’ means dominating a woman,
branding her body, annihilating her. Brutality becomes the very expression of
masculine identity. Rape on Réunion can be understood both as a sign of an inheritance
and as a symptom of a global crisis in masculinity. In contemporary globalisation the
body has become a commodity once again, with massive increases in prostitution, sale
of organs, the trafficking of women and children, and according to this logic, the weaker
body (child, women, senior) is seen as the first to be the object to be traded, mutilated or
violated. Monstrous and human We have to both judge and condemn the
crime and make the victim aware of her rights, as well as carry out an analysis in which
the Réunionese men are not turned into monsters. Making the crime monstrous puts it
beyond the frame of the human condition by way of saying, ‘This is not a human thing.’
But these particular monsters are human beings, and their acts do cause us to ask about
humanity and its capacity for cruelty and the horrific. Yet we do know about horror,
brutality and cruelty; our society was made of it. We can examine how the crime
manifests itself, in its major expression—rape—by approaching it in a comparative
context. Contextualising rape (which is not to excuse it!) stops it being a ‘Réunionese
crime.’ The observation of post-totalitarian situations (like South Africa after apartheid)
shows how much the inheritance of a politics of violence, segregation and negation of
the other has its effects on relations between men and women. It is as if man has to take
revenge on woman for his past humiliation and make her body the territory of
vengeance for his own body that was ‘feminised’ through humiliation, rape and torture. The observation of war and conflict situations today shows rape to be a massive
phenomenon, whether in Yugoslavia, Rwanda, Afghanistan, Chechnya, Tibet or India. PORTAL, vol. 9, no. 1, January 2012. 29 Vergès & Marimoutou Vergès & Marimoutou Well before the massive effects of free-market globalisation on economies At the heart of our unbridled ravines
the surf has carved itself an islet
where mother springs sing
where the murmurs of a country’s roots entwine
Idriss Issop Banian, Je suis d’ici et d’ailleurs At the heart of our unbridled ravines
the surf has carved itself an islet
where mother springs sing
where the murmurs of a country’s roots entwine
Idriss Issop Banian, Je suis d’ici et d’ailleurs At the heart of our unbridled ravines
the surf has carved itself an islet
where mother springs sing
where the murmurs of a country’s roots entwine Idriss Issop Banian, Je suis d’ici et d’ailleurs Idriss Issop Banian, Je suis d’ici et d’ailleurs Vergès & Marimoutou break down insularity,’ yet there have never been so many conference delegates, visits
of artists and researchers. What’s missing? Is there really a lack? We suggest this
unease has several causes: physical, political and economic limitations. Physical
limitations: only on a small part of the land, can constructions be made. There is a lot of
pressure on fundamental resources. We have to admit that resources are not unlimited,
that nature cannot be extended, and that technology is not the solution to all these
limitations. To local limitations must be added the extra burden brought about by
exterior constraints: climate change, attacks on biodiversity, exhaustion of resources,
and the pressure on resources. Do we have the illusion of being protected from these
pressures? And how are we going to work together to find solutions to these local
limitations? Political and social limitations: the local political elite faces the same
challenges as everywhere else. Populism and demagoguery are easy ways out. When
problems demanding careful consideration present themselves, the political tendency is
to wait on France. Demography poses its own specific problems: from 1930 to 2003, the
Réunionese population quadrupled. Every year 10,000 young people enter a workforce
with annual places for only 3,000. In 2030 the number of retirees will by multiplied by
three. Economic limitations: as a product of the first globalisation (slavery and
colonisation), the island’s economy was limited. Slave owners lived on credit and went
easily into debt to buy their bonded workforce. Few had the willingness to take risks. Well before the massive effects of free-market globalisation on economies
(delocalisation, privatisation of public services, impoverishment of the middle classes,
the rule of finance), the island’s economy went through a period where its industrial and
craft structures were destroyed. The economy depends in large measure on government
contracts, and couldn’t protect itself from liberalisation. The limitations of free-market
globalisation are forcing us to be more inventive. Limits We hear people ask, ‘Where is civil society?’ yet there have never been so many
associations and activities. They say, ‘You have to bring people in from the outside to 18 ‘My life flies away with the wind year after year/the roads of the tomorrow/another kind/of people.’
Barbara Robert, Joking. PORTAL, vol. 9, no. 1, January 2012. 30 30 Vergès & Marimoutou Moorings PORTAL, vol. 9, no. 1, January 2012.
3 Vergès & Marimoutou reading over their writings, we inscribe ourselves in the present. We have no regrets, we
have no vindictive or contemptuous feelings against political ‘elites,’ ‘bureaucrats’ or
‘welfare recipients,’ our starting point is who we are, not what we would have liked to
have been. Thus, with language. It is wearying to constantly have to prove that Creole
should be taken seriously at schools, to point out that it is a language, that it can be
written and read, and that, like all languages, it borrows from others. It is wearying to
see how those who negate Creole are given space in public debates: teachers,
psychologists, social workers, while at the same time the National Education Ministry
has recognised the teaching of Creole. And yet, the same old reactionaries, in the name
of the fantasy of purity, showing their fear of diversity, are doing everything they can to
take us back to a former world. Yet, on this point there is no negotiation. reading over their writings, we inscribe ourselves in the present. We have no regrets, we
have no vindictive or contemptuous feelings against political ‘elites,’ ‘bureaucrats’ or
‘welfare recipients,’ our starting point is who we are, not what we would have liked to
have been. Thus, with language. It is wearying to constantly have to prove that Creole
should be taken seriously at schools, to point out that it is a language, that it can be
written and read, and that, like all languages, it borrows from others. It is wearying to
see how those who negate Creole are given space in public debates: teachers, psychologists, social workers, while at the same time the National Education Ministry
has recognised the teaching of Creole. And yet, the same old reactionaries, in the name
of the fantasy of purity, showing their fear of diversity, are doing everything they can to
take us back to a former world. Yet, on this point there is no negotiation. We also have to take into account new facts that throw into doubt the binary structure of
anticolonialist struggles: the emergence of new identities and cultural and confessional
loyalties, massive exclusion through unemployment and a world of precarious lives, a
more complex fragmentation of society, the relation of young people to new
technologies and cartographies (New York, London, Mumbai, Maputo, Antananarivo,
Chennai, Johannesburg). PORTAL, vol. 9, no. 1, January 2012. Inheritances While the battles of the 1960s for language, history and culture are still to be fought, we
also have to respond to new challenges. We acknowledge our debt to the women and
men who had the courage to assert Réunionese singularity. Standing on their shoulders, PORTAL, vol. 9, no. 1, January 2012. 31 Vergès & Marimoutou Moorings PORTAL, vol. 9, no. 1, January 2012.
3 19 From 1964 to 1982, social services obeying Debré’s orders, took away more than 1,600 Réunionese
children from their families and transported them to the department of Creuse in France’s Massif Central,
ostensibly because of overpopulation of Réunion, under-employment and a large demographic of young
people. These children were often abused. Some families never saw them again. All the attempts to bring
the affair to court were rejected. the affair to court were rejected.
20 Teasing and bullying, often nasty, stigmatising physical characteristics or social behaviour. Vergès & Marimoutou seeking visas for our country (not to mention the different currency exchange values). All Réunionese tourists should be conscious of their positions and their prejudices. As
for the sex tourists who head off from Réunion to Madagascar or other countries in the
region, the idea of a colonising colony is alive and well in their heads. Vergès & Marimoutou People in Réunion have learned to manipulate the tools and
methods of globalisation and its communication technologies (internet, mobile
phones—in 2007, 900,000 cell phones for 810,000 inhabitants). This flexibility in
adapting new tools goes along with the difficulty of adapting to social and economic
mutations and sets up a number of questions that force us to listen patiently. We have
been enmeshed in a catch-up economy (catching up to the ‘metropolitan model’—
statistics are always showing us the comparison between household disposable incomes
in Réunion versus data of ‘the metropole’). In order to justify itself (with the question:
‘are we supposed to leave people as they are?’), it dictates all our actions. We must put
in place a new economy, an economy of time to listen, of time to negotiate. New cartographies are asking us to develop an ethics of responsibility. Our colonial past
should make us more sensitive in our behaviour towards the countries of the region. Artists and researchers should ask themselves questions about the relations of
knowledge and power. There is a need to build a real space for collaboration and
exchange. Any Réunionese, being a European citizen, can move around the region at
will. The reverse is not true: queues and humiliation are the lot for our neighbours 32 PORTAL, vol. 9, no. 1, January 2012. 32 Vergès & Marimoutou Moorings Living together The ethic of responsibility that the metaphor of moorings suggests led us to the analysis
of the laws that regulate the right to speak. We easily find scapegoats, most recently
zoreys and people of the Comoros. We have convinced ourselves that we are owed
something, and we find comfort in the idea that others have nothing to teach us. To
share is to speak and to listen to what others have to say as well as to the silences. To
live somewhere is to build a common space where the very question of the political is
posed: who wants the right to speak? Solidarity is too often cliquey. We need to accept
critique, understand that without a critical space the world shrinks and paternalism is
waiting in the wings. Debate is replaced by verbal abuse or what we call grocer
(jealousy, resentment). The structure of silence is deadly when silence becomes law. We
are struck by the way in which the story of the Creuse children has the contours of a
secret.19 First movement: a silent conspiracy. Second movement: the scandal breaks in
the media. Third movement: justification and legitimation (talk of poverty of the
families from whom the children were taken, of the neglect of the children, of the
possibility of a better life for the children, minimising and banalising, ‘there are some
unfortunate exceptions, but the majority was better off in the end’). Fourth movement:
the finger is pointed at the whistleblower. This reminds us of the plot of the film Festen
[Vinterberg, 1998], it is not the person who commits the crime who upsets the apple cart,
and who is rejected in the end, but the person who names him. He uncovers the family
secret, brings the skeleton out of the cupboard and spoils everyone’s fun. Too often we are our own worst enemies. Too often, we don’t allow ourselves dignity
and pride. The practice of moucatage20 is testimony to a perverted relation to otherness. Moucatage can be funny, yet it can also be destructive, signifying the social prohibition PORTAL, vol. 9, no. 1, January 2012. 33 Vergès & Marimoutou Moorings of crossing the bounds of the normal, and the will to bring the other back to conformity. We read practices like ladilafé,21 moucatage, clannish loyalty, as symptoms of envy and
jealousy. How can we bring back practices of solidarity, that is to say of common
consciousness? Living together Réunionese society is not lacking in compassion, it is a lack of
solidarity that too often lets us down. We have yet to build a notion of territory as a
common good and a space of shared practice. The outlines of the limits of private
property (‘never go through the barreau [gate] without being invited!’) indicates the
importance of security fences in Réunionese daily life. Quarrels among neighbours
(with sudden explosions of violence) underline our susceptibility when it comes to these
barriers. The separation between public and private, which is very important (someone
might leave a pile of rubbish outside their gate, no problem, but one’s own kour [front
yard] will be spotless, an aesthetic jewel) no doubt represents the difficulty in
acknowledging the common good. But perhaps one has first to feel secure before
opening one’s door. This process goes back to what we were suggesting: a
reappropriation of the way the territory of ‘Réunion’ is moored. 21 Rumours, often uncomplimentary. PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou Moorings civilisations a given for the future. Homogenisation of the world is an offence against
interculturality, multiplicity and diversity. We are speaking of areas of civilisation. We reject the particularisation of religions, civilisations, philosophies. We defend a
philosophy of borrowing, forgery, imitation, and a dynamic of patching up, making do,
of fixing up, of mending. A world quick to imitate, but which creolises the imitated thing
to make something else of it, which invents the quotidian. This is a dynamic of alterity
where we see no alienation or submission, rather a creativity of a world subject to
continual conflicting inputs. A society has always recourse to imitation. All social
groups and individuals are constituted by a network of borrowings, debts and
recreations. Creole We want to focus here on two versions of creolisation, bearing in mind, once again, that
this text has no pretensions to completeness. We are interested here in language and
maloya, but there are also rituals, housing, cuisine. Creole was born of the need for
communication among people coming from different places, myths, imaginaries and
languages, of the need for spoken exchanges under work conditions of a settler and
plantation society. Words from masters to slaves, slaves to slaves, masters to indentured
labourers, labourers to masters, labourers to labourers, freed slaves to freed slaves. We want to focus here on two versions of creolisation, bearing in mind, once again, that
this text has no pretensions to completeness. We are interested here in language and
maloya, but there are also rituals, housing, cuisine. Creole was born of the need for
communication among people coming from different places, myths, imaginaries and
languages, of the need for spoken exchanges under work conditions of a settler and
plantation society. Words from masters to slaves, slaves to slaves, masters to indentured
labourers, labourers to masters, labourers to labourers, freed slaves to freed slaves. Discourses and knowledges of the world, delivered to language in the form of meanings
to be made, were produced from perceptions and experiences of place and of
relationships of production in the place. Creole language necessarily carries, in the
heterogeneity that presided over its development, the mark of languages, dreams,
imaginaries, which were there at the very start, delivered unconsciously, subterranean,
cryptic. But they burst forth again, in one way or another, in the everyday exchange of
words, in poetic speech, in the lyrics of ségas and maloyas, in proverbs, word-play,
riddles. Discourses and knowledges of the world, delivered to language in the form of meanings
to be made, were produced from perceptions and experiences of place and of
relationships of production in the place. Creole language necessarily carries, in the
heterogeneity that presided over its development, the mark of languages, dreams,
imaginaries, which were there at the very start, delivered unconsciously, subterranean,
cryptic. But they burst forth again, in one way or another, in the everyday exchange of
words, in poetic speech, in the lyrics of ségas and maloyas, in proverbs, word-play,
riddles. PORTAL, vol. 9, no. 1, January 2012. Indian Ocean creolisations O Ramaloya (raya m’aloya) dia Ralaloya (raha m’aloya)
Alevena ao anatin’ny dihy (dia) re ny lasa fandrao managérat Éric Manana, Maloya Creolisation is not an accumulation or a sum of differences. It has the dynamism of an
unfinished process that is subject to mutations and loss. It borrows mimetically and
creatively. It has no problem with putting down roots, because a root is not necessarily
stultifying, if it is a mooring that allows us to move on more easily. We do not idealise
movement. Mooring is a relation that accepts the link, that has no fear of submitting to
meaning, to desire, and is happy to let things go. The Other can no longer be evaluated according to the rulings of the One. All
civilisations have been in contact with others, none has the monopoly on the universal
or on modernity. Each is traversed by conflicts among different traditions and
modernities, each has movable frontiers, complex configurations. Territories are
overlapping, histories are interlaced. Cultural diversity is a fact, and dialogue among 34 34 PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou 22 A street theatre character who usually appears on the first of January. Jako is an acrobat in
monkey/jaguar costume, most likely the product of the mixing of Indian and Mozambiquian practices. y j g
,
y
p
g
q
p
23 Song and dance theatre probably from the south of India and creolised by Indian indentured labourers
It has been practised for a long time on religious occasions, marriages and other festive events. The
repertoire is borrowed from major Hindu myths. y j g
,
y
p
g
q
p
23 Song and dance theatre probably from the south of India and creolised by Indian indentured labourers.
It has been practised for a long time on religious occasions, marriages and other festive events. The PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou Moorings perception of the supernatural. It bursts forth in a maloya by Firmin Viry where the
heroine of an Indian epic, Sìtà in the Ràmàyana, transformed into a female plantation
worker, meets an ancient French romance. It bursts forth in street theatre that mixes
sacred spaces with profane spaces, as in the jako,22 which brings into its dance style and
repertoire of movements practices that are reinterpretations from Dravidian India and
Mozambique. It bursts forth in the narlgon23—Tamil or Malabar theatre—where what
was ritual in the original context becomes theatrical spectacle in the site of a terukkutu
(traditional Tamil theatre) gathered in the unconscious. It burst forth no doubt without
the knowledge of the performers themselves, who have left the origins aside, but it is
there, always present and immediately to hand. Maloya is a stage for the space of processes and practices of Indian Oceanic creolisation,
a common ground for a Réunionese ethos. The lyrics of the maloya take on meaning
and value in a festive or ceremonial context, where there is internal interaction (singer,
chorus) and external interaction (players/participating audience). The lyrics of the
maloya, often improvised from a base of uncertain origin, are infinitely variable,
according to the conditions under which it is uttered, to the public’s role, how the singer
is feeling, and how the chorus is made up; in short, maloya is a performance. It is both a
social practice and a discursive practice, with its own internal logic. It can be read as a
text, it has deconstructions/reconstructions of established collective speech, it has
lexical shifts, and is the unique text of a unique artist. Maloya is a community-in-living represented linguistically, socially and discursively. What was marginal becomes central. Maloya mixes up and multiplies speaking
positions and identities. The closing ceremony is challenged by its own formal devices. What disappears is the voice of the community and its connivance, which is the only
thing that can assure the control and closure of meaning. Yet the meaning is open to all
voices and pathways. Where the same would want to return to the same, the other is
always present. So what does the text tell us, stripped of authorised language? It shows
us precisely the impossible community; there is no protected historical space, no fixed
time. A story of loss emerges that opens the way to melancholy, to an impossible task of PORTAL, vol. Crossings, meetings Heterogeneity is a fact, transformed by the encounter of the imaginaries that produce the
imaginary of the place; it bursts forth in crossings and appropriations. One legend,
Granmèr Kal, is built from an amalgamation of myths from India, Madagascar, and
Africa in the popular memory of Réunionese oral traditions. This memory is linked to
the apprehension that slaves hold for the master and his powers, and to a specific PORTAL, vol. 9, no. 1, January 2012. 35 Vergès & Marimoutou PORTAL, vol. 9, no. 1, January 2012.
3 aguar costume, most likely the product of the mixing of Indian and Mozambiquian practices.
nd dance theatre probably from the south of India and creolised by Indian indentured labourers. onkey/jaguar costume, most likely the product of the mixing of Indian and Mozambiquian practices.
Song and dance theatre probably from the south of India and creolised by Indian indentured labourers Vergès & Marimoutou Vergès & Marimoutou Moorings mourning. What is narrated, in face of the desired or dreamed of utopia of spatio-
temporal closure and of respect for the communitarian norm, is rather, a contrario, a
declension in major and minor keys, all sorts of violence, the dark side of desire,
loneliness of the self, the permanent friction at the heart of everyday life. No one could
deny the robust modernity of this poetic structure in which the poem is an object
constantly under modification by its conditions of interpretation. The song, thus
understood, carries within it hundred of texts, thousands of possibilities, a practice of
reappropriation founded on the knowledge of loss. mourning. What is narrated, in face of the desired or dreamed of utopia of spatio-
temporal closure and of respect for the communitarian norm, is rather, a contrario, a
declension in major and minor keys, all sorts of violence, the dark side of desire,
loneliness of the self, the permanent friction at the heart of everyday life. No one could
deny the robust modernity of this poetic structure in which the poem is an object
constantly under modification by its conditions of interpretation. The song, thus
understood, carries within it hundred of texts, thousands of possibilities, a practice of
reappropriation founded on the knowledge of loss. 24 My father is a weaver bird/my mother a Bengali bird/I am cast from a weaver bird/I drink water/stalks
of the fataque plant. PORTAL, vol. 9, no. 1, January 2012.
3 Vergès & Marimoutou 9, no. 1, January 2012. 36 Vergès & Marimoutou PORTAL, vol. 9, no. 1, January 2012. Vergès & Marimoutou for new solidarities. We live in the Indian Ocean, one of the most dynamic places in the
world today, a place of conflicts and encounters where new configurations are emerging. New cartographies remain to be drawn, reinterpreting the former cartographies from the
present and inventing future ones. We suggest paying particular attention to vernacular practices and forms of expression. Vernacular does not mean tan lontan, rather negotiation between different modernities
and different traditions. It is for this reason that representations have to be deconstructed,
the urban youth cultures of young boys and girls interrogated, along with the middle
class cultures, and the new creolised forms of rites and customs. To achieve this we
propose putting practices in relation to each other, in tension, practices that in a binary
relation would cancel each other out. We wish to privilege a comparative approach of
coming and going, using complexity, without trying to construct a totalising theory. For
example, comparing hip hop cultures in Maputo and Saint-Denis, diasporic
manifestations in Mauritius and Réunion, the informal economy in Antananarivo and in We suggest paying particular attention to vernacular practices and forms of expression. Vernacular does not mean tan lontan, rather negotiation between different modernities
and different traditions. It is for this reason that representations have to be deconstructed,
the urban youth cultures of young boys and girls interrogated, along with the middle
class cultures, and the new creolised forms of rites and customs. To achieve this we
propose putting practices in relation to each other, in tension, practices that in a binary
relation would cancel each other out. We wish to privilege a comparative approach of
coming and going, using complexity, without trying to construct a totalising theory. For
example, comparing hip hop cultures in Maputo and Saint-Denis, diasporic
manifestations in Mauritius and Réunion, the informal economy in Antananarivo and in
Le Port, the trappings of wealth in the middle classes of the Indian Ocean. This
methodology teaches us to be ready for unexpected developments that might contradict
what we are expecting. It warns us of the danger of transforming the past into a burden
for the present and future generations, making them feel guilty (they have the right to
want to forget in order to reappropriate the territory in their own way). methodology teaches us to be ready for unexpected developments that might contradict
what we are expecting. PORTAL, vol. 9, no. 1, January 2012. Moorings Mon papa moutardié
mon monmon bingali
a moin même batard moutardié
mi boire de l’eau
dane coeur fatak
Firmin Viry, Moutardié24 Mon papa moutardié
mon monmon bingali
a moin même batard moutardié
mi boire de l’eau
dane coeur fatak
Firmin Viry, Moutardié24 Firmin Viry, Moutardié24 Proposing the paradigm of Indianoceanic creolisations, we suggest a problematic of loss
and reappropriation. The picture we have painted may seem depressing, but we think
there can be no programme of reconstruction without the work of turning the critical
gaze on oneself. Moving towards a pedagogy of life together, we have not been afraid to
drag the skeletons from the closet, to put forward dissonant voices, emphasise
contradictions and limitations and zoom in on conformities and taboos. Dreaming of
harmony is out of the question. Democracy needs a space of negotiation where opposed
and divergent interests can find a resolution around a common decision where everyone
is prepared to give up a part of their egoistic goals and come to own a part of the others’
goals. We can find in creolisation a methodology for living together. We want to keep our eyes and ears open to see how creolised people are inventive and
capable of appropriating the unexpected. In Réunion we have moved, in less than fifty
years, from a colonial society to a post-modern post-industrial one. Few peoples in the
world have had to face up to such mutations and such uncertainties without experiencing violence. We are proposing a contemporary inscription, as we live on this
island, in the margins of the world and linked to all the continents. Old links offer space 37 37 Vergès & Marimoutou Moorings December 2003 – June 2004, Réunion – Paris. December 2003 – June 2004, Réunion – Pari Vergès & Marimoutou It warns us of the danger of transforming the past into a burden
for the present and future generations, making them feel guilty (they have the right to
want to forget in order to reappropriate the territory in their own way). The world is increasingly confronted with multilingualism, multi-religious practices,
multicultural practices. Colonial managing of differences made them minor and
marginal and inscribed them in a relationship of inequality. Today, no culture or area of
civilisation would accept being placed on a hierarchical scale. The idea of monoculture
no longer makes sense, if indeed it ever did. Creolised people have a long experience of
the intercultural, or of the negotiation between marked contrasts and a constant doubt
about resolving them into one fixed set of practices. It is a fragile space, always on the
edge. One can easily fall on the side of ethnicisation or assimilation. Creolisation is not
the only model of cultural contact, and is not looking to set itself up as such. We do not
know what will emerge from current globalisation. Creolisation is one of the products of
different globalisations; as such it offers a contribution to the debate. For us it represents
the moorings that, going out from the island, attach us to other islands and continents. December 2003 – June 2004, Réunion – Paris. 38 Moorings Vergès & Marimoutou PORTAL, vol. 9, no. 1, January 2012. Reference List Albany, J. 1978, Fare Fare: ou, ‘Le retour aux Isles.’ J. Albany, Paris. Aubry, G. 1971, Rivages d’alizé. Saint-Denis de La Réunion, Saint-Denis. ‘Report de la Commission traitant des violence intra-familiale’ 1985, N.p, Saint-Denis. Fanon, F. 1986, Black Skin, White Masks. Pluto Press, London. Gamaleya, B. 1986, Vali pour une reine morte. 2nd ed. Bibliothèque départementale, Saint-Denis. Gauvin, A. 1983, Romans po détak la lang démay lo kèr. Presses de développement, Saint-Leu. Guéneau, A. 1979, La Réunion: une île, un silence. Arts Graphiques Modernes, Saint-Denis. Jalma, C. 1997, Le Pouvoir éphémère des lapsus. Grand Océan, Saint-Denis. Karm, C. 1992, Rue d’Après. Association des écrivains réunionnais, Saint-Denis. Le Masson, Y. (dir.) 1963, Sucré Amer, Documentary Film, Yann Le Masson Production. Lorraine, A. 1990, Sur le black. 1, Dieu des ravines. Page libre, Saint-Denis. Maharaj, Sarat 2001, ‘Perfidious Fidelity: The Untranslatability of the Other,’ in S. Hall and S. Maharaj
Modernity and Difference, (eds) S. Campbell & G. Tawadros. Institute of International Visual Arts
London, 28–35. Robert, B. 2000, Fannfoutan. Grand Océan, Saint-Denis. Treuthardt, P. 1988, Pointe et Complainte des Galets. UDIR, Le Port, Village Titan. Vinterberg, T. (dir.) 1998, Festen, Feature Film, Nimbus Film Productions. Viry, F. 1998, ‘Moutardié,’ Ti Mardé, CD, Indigo Records. Albany, J. 1978, Fare Fare: ou, ‘Le retour aux Isles.’ J. Albany, Paris. Aubry, G. 1971, Rivages d’alizé. Saint-Denis de La Réunion, Saint-Denis. ‘Report de la Commission traitant des violence intra-familiale’ 1985, N.p, Saint-Denis. Fanon, F. 1986, Black Skin, White Masks. Pluto Press, London. Gamaleya, B. 1986, Vali pour une reine morte. 2nd ed. Bibliothèque départementale, Saint-Denis. Gauvin, A. 1983, Romans po détak la lang démay lo kèr. Presses de développement, Saint-Leu. Guéneau, A. 1979, La Réunion: une île, un silence. Arts Graphiques Modernes, Saint-Denis. Jalma, C. 1997, Le Pouvoir éphémère des lapsus. Grand Océan, Saint-Denis. Karm, C. 1992, Rue d’Après. Association des écrivains réunionnais, Saint-Denis. Le Masson, Y. (dir.) 1963, Sucré Amer, Documentary Film, Yann Le Masson Production. L
i
A 1990 S
l bl
k 1 Di
d
i
P
lib
S i t D
i Albany, J. 1978, Fare Fare: ou, ‘Le retour aux Isles.’ J. Albany, Paris. y
y
Aubry, G. 1971, Rivages d’alizé. Saint-Denis de La Réunion, Saint-Denis. y
y
Aubry, G. 1971, Rivages d’alizé. Saint-Denis de La Réunion, Saint-Denis. y
g
‘Report de la Commission traitant des violence intra-familiale’ 1985, N.p, Saint-Denis. Lorraine, A. 1990, Sur le black. 1, Dieu des ravines. Page libre, Saint-Denis. Vinterberg, T. (dir.) 1998, Festen, Feature Film, Nimbus Film Productions. Reference List ‘Report de la Commission traitant des violence intra-familiale’ 1985, N.p, Saint-Denis. Fanon, F. 1986, Black Skin, White Masks. Pluto Press, London. d Gamaleya, B. 1986, Vali pour une reine morte. 2nd ed. Bibliothèque départementale, Saint-Denis. auvin, A. 1983, Romans po détak la lang démay lo kèr. Presses de développement, Saint-Leu. Jalma, C. 1997, Le Pouvoir éphémère des lapsus. Grand Océan, Saint-Denis. Karm, C. 1992, Rue d’Après. Association des écrivains réunionnais, Saint-Denis. e Masson, Y. (dir.) 1963, Sucré Amer, Documentary Film, Yann Le Masson Production. Lorraine, A. 1990, Sur le black. 1, Dieu des ravines. Page libre, Saint-Denis. g
Maharaj, Sarat 2001, ‘Perfidious Fidelity: The Untranslatability of the Other,’ in S. Hall and S. Maharaj
Modernity and Difference, (eds) S. Campbell & G. Tawadros. Institute of International Visual Arts,
London, 28–35. Maharaj, Sarat 2001, ‘Perfidious Fidelity: The Untranslatability of the Other,’ in S. Hall and S. Maharaj
Modernity and Difference, (eds) S. Campbell & G. Tawadros. Institute of International Visual Arts,
London, 28–35. Robert, B. 2000, Fannfoutan. Grand Océan, Saint-Denis. ,
,
f
,
Treuthardt, P. 1988, Pointe et Complainte des Galets. UDIR, Le Port, Village Titan. Vinterberg, T. (dir.) 1998, Festen, Feature Film, Nimbus Film Productions. 39 39
|
https://openalex.org/W2299911325
|
https://link.springer.com/content/pdf/10.1007%2Fs11892-016-0731-9.pdf
|
English
| null |
Targeting Homeostatic T Cell Proliferation to Control Beta-Cell Autoimmunity
|
Current diabetes report
| 2,016
|
cc-by
| 7,064
|
Curr Diab Rep (2016) 16: 40
DOI 10.1007/s11892-016-0731-9 Curr Diab Rep (2016) 16: 40
DOI 10.1007/s11892-016-0731-9 OLOGYAND TRANSPLANTATION (L PIEMONTI AND V SORDI, SECTION EDITORS) Targeting Homeostatic T Cell Proliferation to Control Beta-Cell
Autoimmunity Debora Vignali1 & Paolo Monti1 Published online: 16 March 2016
# The Author(s) 2016. This article is published with open access at Springerlink.com Published online: 16 March 2016
# The Author(s) 2016. This article is published with open access at Springerlink.com Abstract Immunomodulation of the autoreactive T cell re-
sponse is considered a major strategy to control beta-cell au-
toimmunity, both in the natural history of type 1 diabetes and
in islet transplantation, which can be affected by autoimmuni-
ty recurrence. So far, these strategies have had modest results,
prompting efforts to define novel cellular and molecular tar-
gets to control autoreactive T cell expansion and activation. Novel findings highlighted the important role of the homeo-
static cytokine interleukin-7 in inducing proliferation and dif-
ferentiation of autoreactive T cell clones that causes beta-cell
autoimmunity. In this review, we discuss recent evidences and
novel findings on the role of IL-7 mediated homeostatic T cell
proliferation in the process of beta-cell destruction and evi-
dences of how targeting IL-7 and its receptor could be an
innovative and effective strategy to control beta-cell
autoimmunity. 1
Transplant Immunology Unit, Diabetes Research Institute (DRI),
IRCCS San Raffaele Scientific Institute, Via Olgettina 60,
20132 Milan, Italy Introduction Type 1 diabetes is caused by the selective destruction of
pancreatic insulin-producing beta-cells by an immune-
mediated reaction, predominantly mediated by autoreactive
T cells [1]. CD4+ and CD8+ T cells able to recognize MHC
class II and class I restricted epitopes of the beta-cell-
associated antigens glutamic acid decarboxilase 65
(GAD65), proinsulin, islet tyrosine phosphatase (IA-2),
and zinc transporter 8 [2] have been found in patients with
type 1 diabetes using proliferation assays, ELISPOT, and
fluorescent class I and class II MHC multimers [3]. These
studies have also highlighted several key observations
concerning the nature of the T cell response toward beta-
cells. First, while naive T cells specific for beta-cell anti-
gens are commonly found in subjects with no signs of beta-
cell autoimmunity [4•], in patients with type 1 diabetes
autoreactive T cells show signs of previous antigen encoun-
ter, such as telomere shortening [5•], activation in the ab-
sence of co-stimulatory signals [6], and the expression of
the memory marker CD45RO [5•]. Second, an autoreactive
memory T cell response is difficult to control with standard
immunosuppression and is highly refractory to immuno-
modulatory molecules. This is testified by the limited effi-
cacy of recent clinical trials aiming to prevent or delay
immune-mediated beta-cell loss using GAD65 vaccination
[7], CTLA4-Ig [8], and humanized anti-CD3 antibody [9]. More aggressive approaches based on profound T cell de-
pletion, although effective early after treatment [10], were
later affected by frequent relapse of the autoimmune re-
sponse [11, 12]. Third, generation and expansion of
autoreactive T cell clones can occur under the influence
of homeostatic proliferation mediated by interleukin-7
(IL-7) [13]. The canonical antigen-specific proliferation
pathway relies on the autocrine production of IL-2 and Keywords Autoreactive Tcells . Homeostatic proliferation . Interleukin-7 . Type 1 diabetes . Islet transplantation
This article is part of the Topical Collection on Immunology and
Transplantation
* Paolo Monti
paolo.monti@hsr.it
Debora Vignali
debora.vignali@hsr.it
1
Transplant Immunology Unit, Diabetes Research Institute (DRI),
IRCCS San Raffaele Scientific Institute, Via Olgettina 60,
20132 Milan, Italy Keywords Autoreactive Tcells . Homeostatic proliferation . Interleukin-7 . Type 1 diabetes . The IL-7/IL-7 Receptor Signaling Pathway the upregulation of the IL-2 receptor alpha (also known as
CD25). Therefore an important class of immunomodulato-
ry molecules was developed to target this pathway includ-
ing calcineurin inhibitors and anti-CD25 antibodies. Recent evidences however clearly showed that T cells can
proliferate and acquire a memory phenotype upon activa-
tion of the IL-7/IL-7R axis [14••]. Moreover, recent find-
ings suggest that IL-7 is involved in the generation of T
cells with a stem cell-like memory phenotype (memory
stem T cells, Tscm) [15••] and in the reprogramming of T
cell bio-energetic metabolism for T cell proliferation [16]. While the interest on the homeostatic T cell proliferation
pathway is increasing in autoimmunity, there is a substan-
tial lack of molecules targeting this pathway in humans and
trials to assess whether this could be an effective approach
to control T cell-mediated beta-cell autoimmunity. In this
article, we discuss how the IL-7/IL-7R pathway can con-
tribute to the development of type 1 diabetes and how pre-
clinical models have demonstrated the efficacy of a selec-
tive targeting of this pathway. Finally, we discuss how, and
in which clinical setting, the targeting of the IL-7/IL-7R
pathway can be a therapeutic option for the prevention
and treatment of beta-cell autoimmunity. The high affinity IL-7 receptor results from the IL-7-mediated
crosslinking of the extracellular domains of the IL-7 receptor
alpha-chain (IL-7Rα; also known as CD127) and the common
cytokine receptor-γ chain (γc, also known as CD132). Tyrosine kinases Janus kinases (JAK1) and JAK3 which are
linked to the intracellular domain of the γc mutually phos-
phorylate and their increased kinase activity, thus inducing
phosphorylation of the IL-7Rα intracellular domain and
allowing the binding of the signal inducer transducer and ac-
tivator of transcription factor 5 (STAT-5) which dimerizes and
translocates to the nucleus activating gene transcription [19]. Once in the nucleus, STAT5 regulates the expression of many
targets genes involved in T cell proliferation and survival,
including the upregulation of prosurvival Bcl-2 family mem-
bers, such as the main targets BCL-2, MCL-1, and BC-XL
[20] and downregulation of proapoptotic members of the
BCL-2 family, such as BAX or BIM [21]. Other STAT5-
regulated genes are Runt-related transcription factor 3
(Runx3), essential for T cell development and differentiation,
and the suppressor of cytokines signaling (SOCS), necessary
for negative feedback regulation of the cytokine signaling and
T cell compartment homeostasis. The IL-7/IL-7 Receptor Signaling Pathway In addition to survival, the IL-7R signaling induces T cell
proliferation through the repression of the cell cycle inhibitor
p27kip1 [22] and preventing degradation of Cdc25A [23]. T
cell proliferation is an energy-demanding process and neces-
sitates the incorporation of nutrient to increase the biomass
needed to produce daughter cells [24]. In order to meet the
bio-energetic demands for proliferation, T cells activate aero-
bic glycolysis. Despite a lower efficiency in ATP production
as compared to oxidative phosphorylation, glycolysis is an
anabolic process and provides also fundamental nutrients for
the synthesis of biological macromolecules [25]. The IL-7
signaling activates the PI3K/Akt pathway, which is funda-
mental for T cell proliferation. To sustain the bio-energetic
and metabolic needs, IL-7R signaling is also involved in the
activation of the glycolytic machinery, including the Akt-
mediated upregulation of the glucose transporter 1 (GLUT1)
[26] on the cell surface and the upregulation of the glycolytic
enzymes hexokinase II (HXKII) [16]. Excision of IL-7R on
mature Tcells in vivo results in size and growth reduction of T
cell compartment due to homeostatic mechanism impairment
and metabolic cellular machinery liability [27]. IL-7 Production and Regulation IL-7 is secreted by stromal cells located in the bone mar-
row, in peripheral lymphoid organs, and in the gastro-
intestinal tract [17]. These cells have yet to be fully char-
acterized; however, it appears that IL-7 transcripts are pro-
duced at a constitutive level and are not influenced by
extrinsic stimuli such as the concentration of IL-7 in a
negative feedback or the size of the T cell compartment. Peripheral IL-7 concentration relies on consumption by IL-
7 receptor positive T cells. IL-7 production at a fixed rate
and consumption keeps the serum concentration of IL-7
below 5 pg/ml in physiological conditions. [18] This rep-
resents a limiting factor for T cell expansion. In an immune
system with a full T cell compartment, the concentration of
IL-7 is sufficient for the survival of a finite number of T
cells. IL-7 also maintains a slow rate of T cell proliferation
to counterbalance physiological T cell death. When the
peripheral T cell number is reduced (lymphopenia), the
serum concentration of IL-7 rises to supraphysiological
levels (Fig. 1). Remaining T cells in a lymphopenic envi-
ronment and in an IL-7 rich milieu enter the cell cycle, and
homeostatic T cell proliferation induced by IL-7 occurs. The process is self-limiting as restoring the number of
peripheral T cells corresponds to an increased consumption
of IL-7. Homeostatic T cell proliferation stops when the
physiological T cell number and IL-7 concentration are
fully restored. IL-7 is secreted by stromal cells located in the bone mar-
row, in peripheral lymphoid organs, and in the gastro-
intestinal tract [17]. These cells have yet to be fully char-
acterized; however, it appears that IL-7 transcripts are pro-
duced at a constitutive level and are not influenced by
extrinsic stimuli such as the concentration of IL-7 in a
negative feedback or the size of the T cell compartment. Introduction Islet transplantation This article is part of the Topical Collection on Immunology and
Transplantation
* Paolo Monti
paolo.monti@hsr.it
Debora Vignali
debora.vignali@hsr.it
1
Transplant Immunology Unit, Diabetes Research Institute (DRI),
IRCCS San Raffaele Scientific Institute, Via Olgettina 60,
20132 Milan, Italy This article is part of the Topical Collection on Immunology and
Transplantation * Paolo Monti
paolo.monti@hsr.it
Debora Vignali
debora.vignali@hsr.it 40
Page 2 of 8 Curr Diab Rep (2016) 16: 40 IL-7 in Beta-Cell Autoimmunity In order to protect an individual from a wide range of different
pathogens, the T cell compartment needs to guarantee a broad
diversity of antigen specificity and a sufficient cell number to
mount an effective immune response [28]. The T cell Page 3 of 8
40 Curr Diab Rep (2016) 16: 40 Fig. 1 IL-7-mediated
homeostatic T cell proliferation. In normal conditions, stromal
cells constitutively produce IL-7
which is consumed by T cells and
kept at low levels in the periphery. After T cell depletion, IL-7 con-
centration becomes elevated and
remaining T cell start to prolifer-
ate in the IL-7 rich milieu. During
expansion, autoreactive T cell
clones can undergo expansion
and are included in a higher
number in the reconstituted T cell
pool Fig. 1 IL-7-mediated
homeostatic T cell proliferation. In normal conditions, stromal
cells constitutively produce IL-7
which is consumed by T cells and
kept at low levels in the periphery. After T cell depletion, IL-7 con-
centration becomes elevated and
remaining T cell start to prolifer-
ate in the IL-7 rich milieu. During
expansion, autoreactive T cell
clones can undergo expansion
and are included in a higher
number in the reconstituted T cell
pool the incidence of diabetes was reported only when lymphope-
nia was induced with cyclophosphamide or by exogenous
administration of IL-7. The diabetogenic effect of cyclophos-
phamide or IL-7 injection can be neutralized by blocking the
IL-7R with a monoclonal antibody. repertoire can also include T cell clones reactive to self-anti-
gens, which escaped the negative selection process in the thy-
mus and circulate as naive T cells. These potentially patho-
genic clones are prevented from activation and proliferation
mainly by their low affinity for the cognate MHC-peptide
antigen complex which can deliver only weak signals as com-
pared to high affinity T cell receptors [29•]. The TCR signal
strength determines the T cell capacity to produce IL-2 and to
upregulate the IL-2Rα (also known as CD25) for an effective
expansion and differentiation in a T cell compartment with a
self-limiting size [30]. Lymphopenia, which is associated with increased circulat-
ing levels of IL-7 represents a typical condition in which
autoreactive T cells can expand [28]. A relatively high risk
of autoimmunity development has been observed in patients
undergoing bone marrow transplantation after conditioning
regimen and profound T cell depletion. IL-7 in Beta-Cell Autoimmunity Immunosuppression
in patients receiving islet allotransplantation was associated
with IL-7-mediated expansion of autoreactive T cell clones
specific for GAD65 [14••]. Patients at risk for or with diag-
nosed type 1 diabetes have normal circulating levels of IL-7. However, single nucleotide polymorphisms (SNPs) of the IL-
7Rα were associated with an increased risk of developing type
1 diabetes and multiple sclerosis [33]. As for many other cy-
tokine receptors, a soluble form of the IL-7Rα has been iden-
tified and characterized for its capacity to bind to and inhibit
the interaction of IL-7 with the surface receptor in T cells. In
conditions of high blood glucose levels, the soluble IL-7Rα
undergoes a non-enzymatic glycation process, losing its buff-
ering capacity of IL-7 biological activity [34]. High levels of
circulating IL-7 can be found in neonates and persist for the
first 6 months of life [35]. Both GAD65 and proinsulin-
specific T cells with a naive phenotype are already present at
birth, and IL-7 was shown to promote in vitro proliferation
and activation of GAD65 and proinsulin-specific T cells ob-
tained from cord blood samples. This competitive advantage of T cells with a high affinity
TCR is lost in an IL-7 rich milieu. While the IL-2Rα has an
inducible expression regulated by Tcell activation through the
TCR, the IL-7Rα is constitutively expressed by all Tcells and
can induce Tcell proliferation when the circulating concentra-
tions of IL-7 reach supra-physiological levels [28]. Moreover,
the low affinity TCR-MHC self-peptide interactions, which
are required for T cell persistence, can give a selective advan-
tage to autoreactive T cell clones [29•]. Perturbations of the
IL-7/IL-7R axis can be associated with the activation of
autoreactive T cells. In animal models, the non-obese diabetic (NOD) mouse
was shown to be affected by a chronic state of lymphopenia,
which is permissive for autoreactive homeostatic Tcell expan-
sion under the influence of IL-21 [31]. Increasing the number
of circulating Tcells by injection of complete Freund adjuvant
(CFA) at 3 weeks of age efficiently reduced the incidence of
diabetes. Direct evidences of the role of IL-7 in the autoim-
mune beta-cell destruction process come from the rat insulin
promoter (RIP)-lymphocytic choriomeningitis virus glycopro-
tein (GP)-transgenic mouse model [32•]. In this model, adop-
tive transfer of Smarta T cells restricted for the I-Ab-restricted
LCMV GP61-80 (p13) epitope together with polyclonal CD8
Tcells did not modify the incidence of diabetes. IL-7 in Beta-Cell Autoimmunity An increase in An IL-7 rich milieu can affect immune tolerance to self-
antigens by acting directly on the CD4+CD25+ regulatory T
cells (Treg) [36•]. The Treg network is essential to limit pro-
liferation of autoreactive T cell clones. Despite a low expres-
sion of the IL-7Rα, which makes Treg unresponsive to 40
Page 4 of 8 Curr Diab Rep (2016) 16: 40 physiological concentrations of circulating IL-7, STAT5 phos-
phorylation can be detected at pathological circulating con-
centrations of IL-7. The Treg response to IL-7 appeared to
correlate with the phenotype. While CD45RO+ memory
Treg remain anergic, CD45RO naive Treg strongly proliferate
in response to IL-7, thus acquiring a CD45RO+ memory phe-
notype. Importantly, once exposed to IL-7, both memory and
naive Treg have a reduced capacity to inhibit proliferation of
conventional Tcells in response to autoantigens. The suppres-
sive capacity is fully restored upon IL-7 withdrawal. This
effect can be of importance in the islet transplantation setting
in which prolonged exposure of Treg to high circulating con-
centrations of IL-7 can impair the Treg capacity to control the
expansion of autoreactive and alloreactive T cells. type 1 diabetes and is then maintained for decades after the
disease onset, as testified in patients with type 1 diabetes re-
ceiving islet or pancreas transplants, in which despite immu-
nosuppression, can be associated with reactivation of autoim-
munity [43, 44]. Given the role of the IL-7/IL-7R axis in Tcell
autoimmunity and the long lasting memory to beta-cell anti-
gens observed in patients with type 1 diabetes, we previously
hypothesized that a T cell precursor with stem cell-like prop-
erties could be generated by autoantigen stimulation in the
presence of IL-7 [45]. Even though the existence of memory
stem Tcells reactive to beta-cell antigens has to be proven, this
population would be an ideal target cell for an innovative
approach to beta-cell autoimmunity. Most strategies were de-
signed to target activation/proliferation of pathogenic effector
Tcells. However, the identification and targeting of Tcells that
preserve long-term memory to beta-cell antigen would be able
to eradicate the pathogenic T cell progeny. This approach is
similar in principle to that hypothesized to treat cancer by
targeting cancer stem cells [46]. In this theory, a permanent
eradication of a tumor mass can be achieved by selective
targeting of rare and slowly proliferating cancer stem cell pre-
cursors. IL-7 and the Generation of Memory Stem T Cells The memory T cell compartment is composed by several sub-
sets in different stages of differentiation [30]. Conventionally,
memory T cell populations can be divided according to the
differential expression of CD45RA, CCR7, and CD62L into
central memory (Tcm), effector-memory (Tem), and terminal-
ly differentiated effector-memory RA (Temra) [37]. All mem-
ory subsets are generated from a common naive T cell (Tn)
precursor according to a progressive differentiation model
[38]. In this model, Tn progress along a differentiation path-
way in the order of Tn, Tcm, Tem, and Temra. The final step of
differentiation generates terminally differentiated effector T
cells that are short lived and undergo to massive apoptosis
(contraction) when the immune response exhausts. It was un-
clear whether long-term T cell memory is maintained by the
conventional memory T cell subsets or by a long-lived T cell
precursor with self-renewal potential. In humans, convention-
al memory T cell subsets have indeed a rapid turnover and a
half-life of few weeks [39, 40]. The existence of a memory
precursor with stem cell-like properties was first hypothesized
and subsequently described in mice [41••], in humans [42••],
and in non-human primates [42••]. Tscm show a CD45RA+
CCR7+ naive phenotype associated with the expression of the
memory markers IL-2Rbeta and CD95 [42••]. Generation of
Tscm from naive Tcells involves the homeostatic cytokine IL-
7. In the classical antigen-specific activation pathway, T cell
receptor engagement and IL-2 provides strong signals for T
cell differentiation toward short-lived effector cells. In con-
trast, the homeostatic cytokine IL-7 sustains T cell prolifera-
tion without the robust differentiating activity of IL-2. In vitro
priming of naive T cells in the presence of IL-7 results in the
generation of T cells with phenotypic, functional, and gene
expression attributes found in naturally arising Tscm cells
[15••]. hi i
l
b
ll
i
i
b
ll IL-7 in Beta-Cell Autoimmunity In contrast, standard radio-chemotherapy targeting
the vast majority of rapidly proliferating cancer cells but not
cancer stem cells results in a high rate of tumor relapse. Targeting IL-7 Mediated Homeostatic T Cell
Proliferation in Preclinical Models The polymorphism determines an increase of
circulating levels of sCD127 and has been associated with an
increased susceptibility to type 1 diabetes. sCD127 can bind to
and inhibit IL-7 and is the only known endogenous regulator
of the IL-7 biological activity [34]. sCD127 display a low
affinity (Kd=10–8 M) for IL-7 which is 3 logs lower than
the affinity of IL-7 for the membrane IL-7Rα-γc-chain [53]. Nevertheless, sCD127 concentration in serum (70–80 ng/mL)
is 4 logs higher than physiological IL-7 concentration IL-7
(5 pg/ml) and is therefore expected to significantly influence
the IL-7 bioactivity. The generation of a modified sCD127
with higher affinity for IL-7 could have therapeutic
applications. An interesting biological compound which acts as a potent supernatants from in vitro cultures. Tat enters the cytoplasm of
CD8+ T cells and interacts with the cytoplasmic tail of the IL-
7Rα causing clustering and downregulation from the cell sur-
face. The internalized IL-7Rα is then directed to the protea-
some for degradation. effector T cells that appeared to be necessary for maintenance
of tolerance, which could be reversed by PD-1 blockade. Interestingly, in both models, IL-7Rα antibody treatment
was associated with an increased frequency of regulatory T
cells. The same monoclonal antibody raised against the IL-
7Rα was also tested on BALB/c mice receiving C57BL/6
islets under the kidney capsule, as model of islet transplanta-
tion in streptozotocin-induced diabetes [49•]. When started
3 weeks before islet infusion, the anti-IL-7Rα treatment in-
duced indefinite pancreatic islet allograft survival. This is of
interest in the field of islet transplantation where a single treat-
ment with an anti-IL-7Rα can potentially control both the
autoimmune and the alloimmune response to islet allografts. Pharmacological modulation of IL-7Rα signaling can be
achieved by targeting downstream signaling such as JAK1
and JAK3. Tofacitinib is a potent inhibitor of JAK1 and
JAK3 and currently approved for the treatment of rheumatoid
arthritis (RA), has demonstrated effectiveness in the treatment
of psoriasis in phase III clinical trials, and is currently tested as
immunomodulator for organ transplantation [55]. Tofacitinib
was shown to impair Th1 and Th17 effector function; howev-
er, its action is broad and includes all γc-chain cytokines. Other JAK inhibitors under evaluation in clinical trials for
autoimmune diseases include the JAK1 and JAK2 inhibitor
Ruxolitinib, the JAK1 inhibitor GLPG-0634, and the JAK3
antagonist VX-509. Targeting IL-7 Mediated Homeostatic T Cell
Proliferation in Preclinical Models Even though none of these compounds is
selective for the IL-7/IL-7Rα signaling pathway, they can be
used in combination with selective inhibitors of the IL-7Rα to
potentiate the effectiveness. Biological and pharmacological regulators of the IL-7/IL-
7R axis with a potential translation into the clinic have been
identified. As for other cytokine receptors, a soluble form of
the IL-7Rα (sCD127) has been identified in humans [50]. sCD127 is generated both by alternative splicing or shedding
of the membrane bound IL-7Rα. The amount of circulating
sCD127 is regulated by single nucleotide polymorphism
(SNP) of the IL-7Rα gene [51]. A SNP found in exon 6
(rs6897932) determines the extent of exon splicing. Transcripts that skip exon 6 (C allele of rs6897932) encode
sCD127 [52]. The polymorphism determines an increase of
circulating levels of sCD127 and has been associated with an
increased susceptibility to type 1 diabetes. sCD127 can bind to
and inhibit IL-7 and is the only known endogenous regulator
of the IL-7 biological activity [34]. sCD127 display a low
affinity (Kd=10–8 M) for IL-7 which is 3 logs lower than
the affinity of IL-7 for the membrane IL-7Rα-γc-chain [53]. Nevertheless, sCD127 concentration in serum (70–80 ng/mL)
is 4 logs higher than physiological IL-7 concentration IL-7
(5 pg/ml) and is therefore expected to significantly influence
the IL-7 bioactivity. The generation of a modified sCD127
with higher affinity for IL-7 could have therapeutic
applications. Targeting IL-7 Mediated Homeostatic T Cell
Proliferation in Preclinical Models Molecules with the ability to directly or indirectly target the
IL-7/IL-7R axis have been discovered and first tested in pre-
clinical models (Table 1). The IL-7/IL-7R axis has been im-
plicated in the development of autoimmune diabetes in the
NOD mouse model. Two reports showed how IL-7Rα block-
ade was effective to prevent and revert diabetes. In the first
report [47••], the selective inhibition of the IL-7Rα using a
monoclonal antibody was able to prevent diabetes after only
2–3 injections starting at week 9 of age. Disease remission
was complete and durable, after a 4-week cycle of anti-IL-
7Rα antibody starting at the time of diabetes onset. Anti-IL-
7R antibody was reported to act by inhibiting diabetogenic
effector-memory T cells (Tem). Diabetogenic Tem cells were
not depleted by IL-7Rα treatment but were instead
reprogrammed to suppress IFN-γ secretion and to upregulate
the inhibitory receptor programmed death 1 (PD-1). Tem cells
from anti-IL-7Rα-treated mice were also adoptively trans-
ferred in recipient mice without causing diabetes and demon-
strating that IL-7-signaling deprivation was related to a state
of cell intrinsic tolerance. In the second report [48••], treat-
ment with the IL-7Rα antibody resulted in the reduction of
IFN-γ-producing CD4+ (TH1) and IFN-γ-producing CD8+
(Tc1) T cells. Targeting IL-7 Mediated Homeostatic T Cell
Proliferation in Preclinical Models As shown also in the first report, IL-7Rα anti-
body treatment induced upregulation of PD-1 expression in This is relevant to beta-cell autoimmunity because T cell
memory to beta-cell antigens is established before the onset of Curr Diab Rep (2016) 16: 40 Page 5 of 8
4 40 Table 1 Molecules targeting the
L-7/IL-7R axis
Study phase
Molecule
IL-7/IL-7R specificity
Setting
Refs
Clinical
Anti-IL-7Rα mAb
Yes
T1D, MS
N/A
Mycophenolate mofetil
No
Islet tx in T1D
[14••]
JAK inhibitors
No
RA
[55]
Preclinical
Anti-IL-7Rα mAb
Yes
Diabetes in NOD
Islet tx
[47••, 48••, 49•]
Bz-423
No
GVHD
[59•]
2-DG+ metformin
No
Lupus
[60•]
Experimental
sCD127
Yes
Human
[34]
Tat protein
Yes
Human
[54]
Curr Diab Rep (2016) 16: 40
Page 5 of 8
40 Table 1 Molecules targeting the
IL-7/IL-7R axis Study phase
Molecule
IL-7/IL-7R specificity
Setting
Refs
Clinical
Anti-IL-7Rα mAb
Yes
T1D, MS
N/A
Mycophenolate mofetil
No
Islet tx in T1D
[14••]
JAK inhibitors
No
RA
[55]
Preclinical
Anti-IL-7Rα mAb
Yes
Diabetes in NOD
Islet tx
[47••, 48••, 49•]
Bz-423
No
GVHD
[59•]
2-DG+ metformin
No
Lupus
[60•]
Experimental
sCD127
Yes
Human
[34]
Tat protein
Yes
Human
[54] effector T cells that appeared to be necessary for maintenance
of tolerance, which could be reversed by PD-1 blockade. Interestingly, in both models, IL-7Rα antibody treatment
was associated with an increased frequency of regulatory T
cells. The same monoclonal antibody raised against the IL-
7Rα was also tested on BALB/c mice receiving C57BL/6
islets under the kidney capsule, as model of islet transplanta-
tion in streptozotocin-induced diabetes [49•]. When started
3 weeks before islet infusion, the anti-IL-7Rα treatment in-
duced indefinite pancreatic islet allograft survival. This is of
interest in the field of islet transplantation where a single treat-
ment with an anti-IL-7Rα can potentially control both the
autoimmune and the alloimmune response to islet allografts. Biological and pharmacological regulators of the IL-7/IL-
7R axis with a potential translation into the clinic have been
identified. As for other cytokine receptors, a soluble form of
the IL-7Rα (sCD127) has been identified in humans [50]. sCD127 is generated both by alternative splicing or shedding
of the membrane bound IL-7Rα. The amount of circulating
sCD127 is regulated by single nucleotide polymorphism
(SNP) of the IL-7Rα gene [51]. A SNP found in exon 6
(rs6897932) determines the extent of exon splicing. Transcripts that skip exon 6 (C allele of rs6897932) encode
sCD127 [52]. Compliance with Ethical Standards Conflict of Interest
Debora Vignali and Paolo Monti declare that they
have no conflict of interest. Human and Animal Rights and Informed Consent
This article does
not contain any studies with human or animal subjects performed by any
of the authors. Targeting IL-7-Mediated Homeostatic T Cell
Proliferation in Type 1 Diabetes and Islet
Transplantation A humanized IgG1 monoclonal antibody that binds to and
inhibit the IL-7Rα has been developed by GlaxoSmithKline
and tested as single ascending doses study phase I clinical trial
on healthy volunteers to determine the safety, tolerability,
pharmacokinetics, pharmacodynamics, and immunogenicity
(NCT02293161, ClinicalTrials.gov identifier). The primary
outcome measures were expected in September 2015 and
not yet released. The anti-IL-7Rα monoclonal antibody PF-06342674
(RN168) developed from Pfizer has been tested in phase 1b
in patients with multiple sclerosis (NCT02045732
ClinicalTrials.gov identifier) with the primary outcome to
evaluate adverse side effects and pharmacokinetic– An interesting biological compound, which acts as a potent
downregulator of IL-7Rα signaling is the 15 kDa Tat protein
produced by the HIV virus [54]. Tat is secreted by HIV infect-
ed cells detectable in the serum of HIV-infected patients or in 40
Page 6 of 8 Curr Diab Rep (2016) 16: 40 glycolysis inhibitor 2-deoxy-D-glucose and the fatty acid ox-
idation inhibitor metformin was able to normalize bio-
energetic metabolism of T cells and reverse SLE [60•]. As
previously reported, there is a strong link between IL-7 sig-
naling and bio-energetic metabolism, especially the IL-7 de-
pendent activation of the glycolysis machinery. The Tcell bio-
energetic metabolism is a novel and quickly developing fron-
tier for developing innovative approaches to control IL-7-
mediated homeostatic T cell proliferation and beta-cell
autoimmunity. pharmacodynamic properties. The trial is reported to have
been terminated in April 2015 due to a corporate decision
not related to safety or tolerability issues. A second trial was
designed to test the same antibody in patients with type 1
diabetes and is estimated to conclude in March 2016. While waiting for the results of these trials in terms of
safety and tolerability and for the design of phase II and phase
III trials to determine whether blockade of the IL-7Rα is ef-
fective in the treatment of type 1 diabetes, it is important to
highlight the interest and the effort to explore the IL-7/IL-7R
axis as a potential target pathway to control beta-cell
autoimmunity. Future Perspectives Recent findings suggest that the immune response can be ma-
nipulated by targeting the bio-energetic metabolism of Tcells. Both naive and memory T cell subsets are quiescent cells and
rely on oxidative phosphorylation to sustain housekeeping
functions [57]. However, proliferating cells like activated
and effector T cells upregulate oxidative phosphorylation to
meet the increased energetic demand and activates aerobic
glycolysis, which is better suited for the biosynthesis of cell
components, proliferation, and effector functions [58]. Bz-423
is a novel small molecule that inhibits the mitochondrial
F1F0-ATPase resulting in the increase of superoxide and in-
ducing selective apoptosis of alloreactive Tcells while sparing
proliferating hematopoietic stem cells using preferentially gly-
colysis [59•]. This strategy was successful in preventing graft
vs host disease while preserving immune system repopulation
in a mouse model. In a mouse model of systemic lupus ery-
thematosus (SLE), a combination treatment with the Open Access This article is distributed under the terms of the Creative
Commons Attribution 4.0 International License (http://
creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided you give appro-
priate credit to the original author(s) and the source, provide a link to the
Creative Commons license, and indicate if changes were made. 1.
Roep BO. The role of T-cells in the pathogenesis of Type 1 diabetes:
from cause to cure. Diabetologia. 2003;46:305–21. Conclusions As we previously reported, islet transplantation in patients
with type 1 diabetes is associated with IL-7-mediated homeo-
static proliferation of memory autoreactive Tcells. This model
also provided important evidences of the effect of classical
immunosuppressive drugs on the IL-7/IL-7R axis. The
mTOR inhibitor rapamycin and the calcineurin inhibitor
FK506 were ineffective on IL-7-mediated homeostatic T cell
proliferation at therapeutic serum level concentrations [14••]. Interestingly, the anti-CD25 antibody daclizumab by blocking
the interaction of CD25 with the γc chain, promotes the asso-
ciation of the IL-7Rα/γc complex resulting in an increased
sensitivity of T cells to IL-7 [56]. On the other hand, myco-
phenolate mofetil (MMF) was shown to reduce by 90 % the
rate of homeostatic T cell proliferation at therapeutic serum
level concentrations [14••]. MMF is a purine synthesis inhib-
itor acting very downstream in the proliferation process. Even
though the mechanisms of action suggest that inhibition of T
cell proliferation is not specific it represents the best therapeu-
tic option to control homeostatic proliferation at least in the
transplantation setting. IL-7-mediated homeostatic proliferation T cell clones have
gained increasing interest in the pathogenesis of type 1 diabe-
tes and in autoimmunity recurrence post pancreas or islet
transplantation. This prompted efforts into a better under-
standing of the molecules and pathways involved in the IL-
7/IL-7R axis signaling that can be selectively targeted to con-
trol homeostatic proliferation. Some of these are in first phase
clinical trials, and we need to consider other ways in which
homeostatic proliferation can be controlled and incorporate
such strategies in the therapy of beta-cell autoimmunity. Acknowledgments
Paolo Monti is supported by a Career Development
Award from the Juvenile Diabetes Research Foundation (JDRF, 5-CDA-
2015-85-A-N). References 2011;378:319–27. 27. Rathmell JC, Farkash EA, Gao W, et al. IL-7 enhances the survival
and maintains the size of naive Tcells. J Immunol. 2001;167:6869–
76. 8. Orban T, Bundy B, Becker DJ, DiMeglio L, et al. Co-stimulation
modulation with abatacept in patients with recent-onset type 1 dia-
betes: a randomised, double-blind, placebo-controlled trial. Lancet. 2011;378:412–9. 28. Khoruts A, Fraser JM. A causal link between lymphopenia and
autoimmunity. Immunol Lett. 2005;98:23–31. 29.• Goldrath AW, Bevan MJ. Low-affinity ligands for the TCR drive
proliferation of mature CD8+ T cells in lymphopenic hosts. Immunity. 1999;11:183–90. This study showed the selective ad-
vantage a T cells with a low affinity T cell Receptor during
homeostatic T cell proliferation. 9. Sherry N, Hagopian W, Ludvigsson J, et al. Teplizumab for treat-
ment of type 1 diabetes (Protégé study): 1-year results from a
randomised, placebo-controlled trial. Lancet. 2011;378:487–97. 10. Voltarelli JC, Couri CEB, Stracieri ABPL, et al. Autologous
nonmyeloablative hematopoietic stem cell transplantation in
newly diagnosed type 1 diabetes mellitus. JAMA. 2007;297:
1568–76. 30. Sallusto F, Geginat J, Lanzavecchia A. Central memory and effector
memory T cell subsets: function, generation, and maintenance. Annu Rev Immunol. 2004;22:745–63. 31. King C, Ilic A, Koelsch K, et al. Homeostatic expansion of T cells
during immune insufficiency generates autoimmunity. Cell. 2004;117:265–77. 11. Couri CEB, de Oliveira MC, Simões BP. Risks, benefits, and ther-
apeutic potential of hematopoietic stem cell transplantation for au-
toimmune diabetes. Curr Diabet Rep. 2012;12:604–11. 32.• Calzascia T, Pellegrini M, Lin A, et al. CD4 T cells, lymphopenia,
and IL-7 in a multistep pathway to autoimmunity. PNAS. 2008;105:
2999–3004. This study directly correlated IL-7 with the devel-
opment of autoimmune diabetes in a mouse model. 12. Couri CEB, Oliveira MCB, Stracieri ABPL, et al. C-peptide levels
and insulin independence following autologous nonmyeloablative
hematopoietic stem cell transplantation in newly diagnosed type 1
diabetes mellitus. JAMA. 2009;301(15):1573–9. 33. Lundmark F, Duvefelt K, Iacobaeus E, et al. Variation in interleukin
7 receptor α chain (IL7R) influences risk of multiple sclerosis. Nat
Genet. 2007;39:1108–13. 13. Monti P, Bonifacio E. Interleukin-7 and type 1 diabetes. Curr Diabet
Rep. 2014;14:518–26. 14.•• Monti P, Scirpoli M, Maffi P, Ghidoli N, De TF, Bertuzzi F, et al. Islet transplantation in patients with autoimmune diabetes induces
homeostatic cytokines that expand autoreactive memory T cells. J
Clin Invest. 2008;118:1806–14. References Papers of particular interest, published recently, have been
highlighted as: •
Of importance
•• Of major importance Curr Diab Rep (2016) 16: 40 Page 7 of 8 40 7, 9, 15, and 21, and their signaling pathways. Immunol Rev. 2004;202:67–83. 2. Di Lorenzo TP, Peakman M, Roep BO. Translational mini-review
series on type 1 diabetes: systematic analysis of T cell epitopes in
autoimmune diabetes. Clin Exp Immunol. 2007;148:1–16. 20. Maraskovsky E, O’Reilly LA, Teepe M, et al. Bcl-2 can rescue T
lymphocyte development in interleukin-7 receptor-deficient mice
but not in mutant rag-1−/−mice. Cell. 1997;89:1011–9. 3. Mannering SI, Wong FS, Durinovic-Belló I, et al. Current ap-
proaches to measuring human islet-antigen specific T cell function
in type 1 diabetes. Clin Exp Immunol. 2010;162:197–209. 21. Pellegrini M, Bouillet P, Robati M, et al. Loss of bim increases T
cell production and function in interleukin 7 receptor-deficient
mice. J Exp Med. 2004;200:1189–95. 4.• Danke N, Yang J, Greenbaum C, Kwok WW. Comparative study of
GAD65-specific CD4+ T cells in healthy and type 1 diabetic sub-
jects. J Autoimmun. 2005;25:303–11. This paper revealed the
presence of GAD65 reactive T cells in subjects with no other
sign of autoimmunity. 22. Li WQ, Jiang Q, Aleem E, et al. IL-7 promotes T cell proliferation
through destabilization of p27Kip1. J Exp Med. 2006;203:573–82. 23. Khaled AR, Bulavin DV, Kittipatarin C, et al. Cytokine-driven
cell cycling is mediated through Cdc25A. J Cell Biol. 2005;169:755–63. 5.• Monti P, Scirpoli M, Rigamonti A, et al. Evidence for in vivo
primed and expanded autoreactive T cells as a specific feature of
patients with type 1 diabetes. J Immunol. 2007;179:5785–92. This
paper showed that autoreactive T cells in patients with type 1
diabetes developed a memory phenotype. 24. Fox CJ, Hammerman PS, Thompson CB. Fuel feeds function: en-
ergy metabolism and the T-cell response. Nat Rev Immunol. 2005;5:844–52. 25. Pearce EL. Metabolism in Tcell activation and differentiation. Curr
Opin Immunol. 2010;22:314–20. 6. Viglietta V, Kent SC, Orban T, Hafler DA. GAD65-reactive T cells
are activated in patients with autoimmune type 1a diabetes. J Clin
Invest. 2002;109:895–903. 26. Wofford JA, Wieman HL, Jacobs SR, et al. IL-7 promotes Glut1
trafficking and glucose uptake via STAT5-mediated activation of
Akt to support T-cell survival. Blood. 2008;111:2101–11. 7. Wherrett DK, Bundy B, Becker DJ, et al. Antigen-based therapy
with glutamic acid decarboxylase (GAD) vaccine in patients with
recent-onset type 1 diabetes: a randomised double-blind trial. Lancet. References Vrisekoop N, den Braber I, de Boer AB, et al. Sparse production but
preferential incorporation of recently produced naive T cells in the
human peripheral pool. Proc Natl Acad Sci U S A. 2008;105:6115–
20. 51. McKay FC, Swain LI, Schibeci SD, et al. Haplotypes of the inter-
leukin 7 receptor alpha gene are correlated with altered expression
in whole blood cells in multiple sclerosis. Genes Immunol. 2008;9:
1–6. 41.•• Gattinoni L, Zhong X-S, Palmer D, et al. Wnt signaling arrests
effector T cell differentiation and generates CD8+ memory stem
cells. Nat Med. 2009;15:808–13. This study reported for the first
time the generation of memory stem T cells by interfering with
the WNT signaling pathway. 52. Teutsch SM, Booth DR, Bennetts BH, et al. Identification of 11
novel and common single nucleotide polymorphisms in the
interleukin-7 receptor-alpha gene and their associations with multi-
ple sclerosis. Eur J Hum Genet. 2003;11:509–15. 42.•• Lugli E, Dominguez MH, Gattinoni L, et al. Brief report Superior
T memory stem cell persistence supports long-lived T cell mem-
ory. J Clin Invest. 2013;123:594–9. This study reported for the
first time the existence and phenotype of memory stem T cells
in humans. 53. Noguchi M, Nakamura Y, Russell SM, et al. Interleukin-2 receptor
gamma chain: a functional component of the interleukin-7 receptor. Science. 1993;262:1877–80. 43. Laughlin E, Burke G, Pugliese A. Autoimmune diabetes after pan-
creas transplantation. Clin Immunol. 2008;128:23–30. 54. Faller EM, Sugden SM, McVey MJ, et al. Soluble HIV Tat protein
removes the IL-7 receptor alpha-chain from the surface of resting
CD8 T cells and targets it for degradation. J Immunol. 2010;185:
2854–66. 44. Roep BO, Stobbe I, Duinkerken G, et al. Auto- and alloimmune
reactivity to human islet allografts transplanted into type 1 diabetic
patients. Diabetes. 1999;48:484–90. 55. Tanaka Y. Recent progress and perspective in JAK inhibitors for
rheumatoid arthritis: from bench to bedside. J Biochem. 2015;158:
173–9. 45. Monti P, Heninger A-K, Bonifacio E. Differentiation, expansion,
and homeostasis of autoreactive T cells in type 1 diabetes mellitus. Curr Diabet Rep. 2009;9:113–8. 56. Monti P, Brigatti C, Heninger A-K, et al. Disengaging the IL-2
receptor with daclizumab enhances IL-7-mediated proliferation of
CD4(+) and CD8(+) T cells. Am J Transplant. 2009;9:2727–35. 46. Beck B, Blanpain C. Unravelling cancer stem cell potential. Nat
Rev Cancer. 2013;13:727–38. 57. Jones RG, Thompson CB. Revving the engine: signal transduction
fuels T cell activation. Immunity. 2007;27:173–8. References This study reported that lympho-
penia is associated with increased circulating IL-7 which drives
expansion of autoreactive T cell clones in patients with type 1
diabetes undergoing to islet allo-transplantation. 34. Monti P, Brigatti C, Krasmann M, et al. Concentration and activity
of the soluble form of the interleukin-7 receptor alpha in type I
diabetes identifies an interplay between hyperglycemia and im-
mune function. Diabetes. 2013;62:2500–8. 35. Heninger A-K, Monti P, Wilhelm C, et al. Activation of islet
autoreactive naive T cells in infants is influenced by homeostatic
mechanisms and antigen-presenting capacity. Diabetes. 2013;62:
2059–66. 15.•• Cieri N, Camisa B, Cocchiarella F, Forcato M, Oliveira G, Provasi
E, et al. IL-7 and IL-15 instruct the generation of human memory
stem T cells from naive precursors. Blood. 2013;121:573–84. This
study demonstrated that homeostatic cytokines IL-7 and IL-15
are involved in the generation of memory stem T cells. 36.• Heninger A-K, Theil A, Wilhelm C, et al. IL-7 abrogates suppres-
sive activity of human CD4+CD25+FOXP3+ regulatory Tcells and
allows expansion of alloreactive and autoreactive T cells. J
Immunol. 2012;189(12):5649–58. This study reported that high
IL-7 concentrations can inhibit the suppressive capacity of
CD4 + CD25 + FOXP3+ regulatory T cells. 16. Chehtane M, Khaled AR. Interleukin-7 mediates glucose utilization
in lymphocytes through transcriptional regulation of the hexokinase
II gene. Am J Physiol Cell Physiol. 2010;298:1560–71. 37. Sallusto F, Lenig D, Förster R, et al. Two subsets of memory T
lymphocytes with distinct homing potentials and effector functions. Nature. 1999;401:708–12. 17. Mazzucchelli RI, Warming S, Lawrence SM, Ishii M, Abshari M,
Washington AV, et al. Visualization and identification of IL-7 pro-
ducing cells in reporter mice. PLoS ONE. 2009;4:e7637. 38. Restifo NP, Gattinoni L. Lineage relationship of effector and mem-
ory T cells. Curr Opin Immunol. 2013;25:556–63. 18. Mazzucchelli R, Durum SK. Interleukin-7 receptor expression: in-
telligent design. Nat Rev Immunol. 2007;7(2):144–54. 39. Macallan DC, Wallace D, Zhang Y, et al. Rapid turnover of effector-
memory CD4(+) T cells in healthy humans. J Exp Med. 2004;200:
255–60. 19. Kovanen PE, Leonard WJ. Cytokines and immunodeficiency dis-
eases: critical roles of the γc-dependent cytokines interleukins 2, 4, 40
Page 8 of 8 Curr Diab Rep (2016) 16: 40 50. Goodwin RG, Friend D, Ziegler SF, et al. Cloning of the human and
murine interleukin-7 receptors: demonstration of a soluble form and
homology to a new receptor superfamily. Cell. 1990;60:941–51. 40. References 47.•• Penaranda C, Kuswanto W, Hofmann J, et al. IL-7 receptor block-
ade reverses autoimmune diabetes by promoting inhibition of
effector/memory T cells. PNAS. 2012;109:12668–73. This study
and the following reference show reversal of autoimmune dia-
betes in the NOD mouse model using a monoclonal antibody
directed to the IL-7Rα. 58. Wang R, Green DR. Metabolic checkpoints in activated Tcells. Nat
Immunol. 2012;13:907–15. 59.• Gatza E, Wahl DR, Opipari AW, et al. Manipulating the bioenerget-
ics of alloreactive Tcells causes their selective apoptosis and arrests
graft versus host disease. Sci Transl Med. 2011;3:1–16. This paper
showed that GVDH can be controlled by targeting the mito-
chondrial f0f1 ATP-ase. 48.•• Lee L, Logronio K, Huan G, et al. Anti–IL-7 receptor-α reverses
established type 1 diabetes in nonobese diabetic mice by modulat-
ing effector T-cell function. PNAS. 2012;109:12674–9. Companion paper of Penaranda et al. ref 45. 60.• Yin Y, Choi SS-C, Xu Z, Perry DJ, Seay H, Croker BP, et al. Normalization of CD4+ T cell metabolism reverses lupus. Sci
Transl Med. 2015;7:274ra18. This paper showed that a combina-
tion of the glycolysis inhibitor 2-DG and the oxidative phos-
phorylation inhibitor metformin can stop lupus in a mouse
model. 49.• Le MH, Boeffard F, Longis J, et al. IL-7 receptor blockade follow-
ing T cell depletion promotes long-term allograft survival. J Clin
Invest. 2014;124:1723–33. This study showed that a monoclonal
antibody directed to the IL-7Rα induce islet graft acceptance in
a non-autoimmune model of diabetes.
|
https://openalex.org/W4362432888
|
https://aacr.figshare.com/articles/journal_contribution/Figure_S3_from_Functional_i_Ex_Vivo_i_Assay_Reveals_Homologous_Recombination_Deficiency_in_Breast_Cancer_Beyond_BRCA_Gene_Defects/22467971/1/files/39919415.pdf
|
English
| null |
Figure S3 from Functional <i>Ex Vivo</i> Assay Reveals Homologous Recombination Deficiency in Breast Cancer Beyond BRCA Gene Defects
| null | 2,023
|
cc-by
| 244
|
WGS WGS WGS Sureselect regions WGS Sureselect regions 0
25
50
75
100
0
20000
40000
60000
80000
PD5937
(n=76097)
PD9063
(n=50045)
PD9568
(n=30395)
PD13167
(n=2405)
PD9591
(n=1296)
Signatures
Signature 1
Signature 2
Signature 3
Signature 5
Signature 6
Signature 8
Signature 9
Signature 12
Signature 13
Signature 14
Signature 16
Signature 20
Signature 21
Filtered (<3%)
Absolute contribution
Absolute contribution
Relative contribution
Relative contribution
PD5937
(n=466)
PD9063
(n=47)
PD9568
(n=27)
PD13167
(n=1683)
PD9591
(n=1068)
0
25
50
75
100
0
500
1000
1500
Signatures
Signature 1
Signature 2
Signature 3
Signature 4
Signature 6
Signature 7
Signature 9
Signature 10
Signature 13
Signature 17
Signature 20
Signature 21
Signature 24
Filtered (<3%)
WGS Sureselect regions Relative contribution Relative contribution Absolute contribution
PD5937
(n=466)
PD9063
(n=47)
PD9568
(n=27)
PD13167
(n=1683)
PD9591
(n=1068)
0
500
1000
1500
Signatures
Signature 1
Signature 2
Signature 3
Signature 4
Signature 6
Signature 7
Signature 9
Signature 10
Signature 13
Signature 17
Signature 20
Signature 21
Signature 24
Filtered (<3%) 0
20000
40000
60000
80000
PD5937
(n=76097)
PD9063
(n=50045)
PD9568
(n=30395)
PD13167
(n=2405)
PD9591
(n=1296)
Signatures
Signature 1
Signature 2
Signature 3
Signature 5
Signature 6
Signature 8
Signature 9
Signature 12
Signature 13
Signature 14
Signature 16
Signature 20
Signature 21
Filtered (<3%)
Absolute contribution 0 4
Absolute contribution Absolute contribution
1 PD13167
(n=2405) Signature 24
Filtered (<3%) Signature 24
Filtered (<3%) Signature 5
Signature 6
Signature 8 Signature 21
Filtered (<3%) Signature 21
Filtered (<3%) Signatures Signatures
|
W4313645609.txt
|
https://ehoonline.biomedcentral.com/counter/pdf/10.1186/s40164-022-00370-2
|
en
|
ALKBH5 prevents hepatocellular carcinoma progression by post-transcriptional inhibition of PAQR4 in an m6A dependent manner
|
Experimental hematology & oncology
| 2,023
|
cc-by
| 8,687
|
Wang et al.
Experimental Hematology & Oncology
(2023) 12:1
https://doi.org/10.1186/s40164-022-00370-2
Experimental Hematology & Oncology
Open Access
RESEARCH
ALKBH5 prevents hepatocellular carcinoma
progression by post‑transcriptional inhibition
of PAQR4 in an m6A dependent manner
Weijian Wang1,2†, Qibo Huang1,2†, Zhibin Liao1,2,3†, Hongwei Zhang1,2,3, Yachong Liu1,2, Furong Liu1,2,
Xiaoping Chen1,2,3, Bixiang Zhang1,2,3, Yan Chen1,2,3* and Peng Zhu1,2,3*
Abstract
Background N6-methyladenosine (m6A) is a prevalent modification of mRNA and is known to play important roles
in tumorigenesis in many types of cancer. The function of N6-methyladenosine (m6A) RNA methylation depends on
a variety of methyltransferases and demethylases. AlkB homolog 5 (ALKBH5) is a demethylase, and its biological function has not been completely explored in HCC.
Results ALKBH5 is downregulated and has antitumor effects in HCC cells. In addition, Progestin and AdipoQ Receptor 4 (PAQR4) was identified as a downstream target of ALKBH5 based on transcriptome sequencing and validation
studies. We found that ALKBH5 decreases PAQR4 mRNA and protein expression in an N6-methyladenosine (m6A)dependent manner. The study also showed that ALKBH5 changes PAQR4 expression via the m6A reader IGF2BP1. In
both in vivo and in vitro experiments, PAQR4 showed a strong association with the development of HCC. Finally, we
found that PAQR4 interacts with AKT and enhances PI3K/AKT pathway activation.
Conclusions ALKBH5 inhibits HCC growth by downregulating PAQR4 expression in an m6A-dependent manner,
therefore suppressing PI3K/AKT pathway activation.
Keywords ALKBH5, Methylation, PAQR4, AKT, HCC
†
Weijian Wang, Qibo Huang and Zhibin Liao are contributed equally to the
study
*Correspondence:
Yan Chen
c_yan66@yahoo.com
Peng Zhu
zhupeng@tjh.tjmu.edu.cn
1
Hepatic Surgery Center, Tongji Hospital, Tongji Medical College,
Huazhong University of Science and Technology, Wuhan 430030, China
2
Hubei Key Laboratory of Hepato-Pancreato-Biliary Diseases,
Wuhan 430030, China
3
Department of General Surgery, Tongji Hospital, Tongji Medical College,
Huazhong University of Science and Technology, Wuhan 430030, China
Background
Hepatocellular carcinoma (HCC) is still one of the most
prevalent malignancies and is the fifth leading cause of
cancer death worldwide [1, 2]. Despite achievements in
the prevention of HCC, the incidence rate of HCC worldwide is still rising, while the incidence rate of HCC has
been declining in some areas in recent years [3–5]. To
date, some patients have benefited from current advances
in standardized treatment. However, the long-term survival rate of patients with HCC is low [6], which is attributed to a weak response to chemotherapeutic agents and
to metastasis and recurrence of hepatic cancers after
resection [7, 8]. The abovementioned situations have
inevitably led to treatment failure in HCC. Therefore, we
need to explore the specific mechanisms involved in the
recurrence and development of HCC.
© The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
As a unique RNA modification, m6A modification was
first identified in the 1970s. However, it was after the discovery of FTO as the first RNA demethylase in 2011 [9]
that the methylation of RNA was regarded as a reversible
process. Since then, research on m6A RNA modification
has arisen to fill the gap in this field. Dynamic regulation of RNA methylation mainly depends on ‘‘writers’’
and ‘‘erasers’’, which function as dedicated RNA methyltransferases and demethylases, respectively [10]. In the
past few years, with the development of high-throughput sequencing technology and new anti-m6A antibodies, the role of m6A modification in hepatic cancers has
been increasingly discussed [11].
In HCC, the m6A methyltransferase methyltransferase-like 3 (METTL3) contributes to HCC progression
by methylation of SOCS2 and Snail [12, 13]. Similarly, the
methyltransferase methyltransferase-like 14 (METTL14)
suppresses metastasis in HCC by modulating the methylation levels of microRNA 126 and USP48 [7, 14]. However, the potential roles of the demethylases fat-mass and
obesity-associated protein (FTO) and AlkB homolog 5
(ALKBH5) in HCC have remained unexamined. There
have been few studies of FTO in liver cancer, while the
results of studies of ALKBH5 in HCC are inconclusive
[15].
ALKBH5 was first reported as a mammalian demethylase in 2013 [16]. Soon after that, many studies were
conducted on ALKBH5. ALKBH5 maintains the tumorigenicity of glioblastoma stem-like cells (GSCs) through
the regulation of FOXM1 in a methylation-dependent
manner [17]. In intrahepatic cholangiocarcinoma (ICC),
ALKBH5 regulates the tumor immune microenvironment and immunotherapeutic efficacy via modulating
the methylation of PD-L1 mRNA [18–20]. In ovarian
cancer, ALKBH5 promotes cisplatin resistance in cancer
cells by modifying its m6A target gene JAK2 to activate
the JAK2/STAT3 signaling pathway[21]. These findings
demonstrate that ALKBH5 is involved in carcinogenesis,
tumor formation and the immune microenvironment in
many types of cancer. In HCC, the malignancy of hepatocellular carcinoma cells is suppressed by ALKBH5 via
m6A-guided suppression of the oncogenic driver LYPD1
[22]. However, another article reported that ALKBH5
promotes hepatocellular carcinogenesis in HBV-related
HCC and catalyzes m6A demethylation of HBx mRNA to
sustain its expression [23]. Thus, the role of ALKBH5 in
HCC remains to be clarified.
Here, we found that ALKBH5 expression was
decreased in HCC tissues. We discovered that ALKBH5
suppressed HCC cell proliferation and invasion in vitro.
By transcriptome sequencing and subsequent validation through filtering, we found that PAQR4 mRNA was
methylated and modified by ALKBH5, which influenced
Page 2 of 16
the stability of PAQR4 mRNA in an IGF2BP1-dependent
manner. Additionally, we suggested that PAQR4 can promote the proliferation, migration, and invasion of HCC
cells in vivo and in vitro. Furthermore, we found that
PAQR4 promoted the malignant behavior of HCC cells
through the PI3K/AKT signaling pathway. Overall, we
found that PAQR4, a target gene of ALKBH5 for demethylation, promotes HCC cell proliferation, migration, and
invasion via the PI3K/AKT signaling pathway.
Results
The ALKBH5 protein level is decreased in HCC
and suppresses HCC proliferation and invasion
To investigate the function of ALKBH5 in HCC, we
first measured the protein level of ALKBH5 in 67 pairs
of patient tumor and adjacent tissues. The results
showed that the ALKBH5 protein level was significantly
decreased in HCC (Fig. 1A), consistent with the results
of previous studies [22]. Herein, we generated ALKBH5
overexpression and knockdown cell lines to investigate
the biological functions of ALKBH5 in HCC. CCK8
and EdU incorporation assays were performed to assess
the proliferation of the corresponding HCC cell lines.
The results showed that overexpression of ALKBH5
inhibited the proliferation of HLF and 97H cells, while
knockdown of ALKBH5 in LM3 cells showed the opposite effect (Fig. 1B–E, Additional file 1A). Similarly, the
results of the migration, invasion, and wound healing
assays indicated that overexpression of ALKBH5 reduced
the migration and invasion of HLF and 97H cells, while
knockdown of ALKBH5 in LM3 cells showed the opposite effects (Fig. 1F–I, Additional file 1B-C).
PAQR4 was downregulated by ALKBH5‑mediated m6A
modification
To investigate the potential mechanisms by which
ALKBH5 inhibits HCC cell proliferation and invasion, mRNA transcriptome sequencing was performed
in HLF vector cells and HLF ALKBH5-overexpressing
cells. RNA-seq analysis was used to identify possible
ALKBH5 target genes (Fig. 2A). We focused on the
top 6 downregulated genes: ISG15, HERC6, SYT14,
PAQR4, TAP1, and FYB1 (Fig. 2B). Then, we performed RT–qPCR analysis to verify the downregulation of these genes in ALKBH5-overexpressing HLF
cells (Fig. 2C). To confirm which genes are regulated by
ALKBH5-mediated N6-methyladenosine (m6A) modification, MeRIP assays were performed. The results
indicated that the m6A level of PAQR4 was the highest
in HLF cells (Fig. 2D). We found that the protein level
of PAQR4 was increased after ALKBH5 knockdown
but decreased after ALKBH5 overexpression (Fig. 2E).
According to a previous study, ALKBH5 H204A
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
Page 3 of 16
Fig. 1 ALKBH5 is decreased in HCC and suppress HCC proliferation and invasion in vitro. A Protein level of ALKBH5 was detected by Western bolt
assay in paired HCC samples (n = 60). CCK8 and EdU assays confirmed that overexpression of ALKBH5 inhibit proliferation of 97H and HLF (B, D and
Additional file 1A), whereas knockdown of ALKBH5 promoted cell proliferation in LM3 (C, E and Additional file 1A). Transwell and wound healing
assays showed that overexpression of ALKBH5 suppress the capability of migration and invasion in 97H and HLF (F, H and Additional file 1B, C). In
contract, knockdown of ALKBH5 promoted migration and invasion in LM3 (G, I and Additional file 1B, C). GAPDH was used as internal controls in
Western blotting analysis. *P < 0.05, **P < 0.01, ***P < 0.001
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
Page 4 of 16
Fig. 2 ALKBH5 suppressed m6A modification of PAQR4 mRNA. A The heat map exhibited the changed genes of transcriptome sequencing after
ALKBH5 overexpression. B, C RT–qPCR analysis showed the six altered target genes with ALKBH5 knockdown or ALKBH5 overexpression. D The
result of m6A abundances in mRNA transcripts of ALKBH5 target genes was shown. E Western blot analysis showed the protein level of PAQR4
with overexpression of ALKBH5 and knockdown of ALKBH5 in HCC cells. F–H RT–qPCR analysis and Western blot verified both mRNA and protein
expression of PAQR4 was decreased when co-transfected with ALKBH5-WT but not with ALKBH5-H204A. GAPDH was used as internal controls in
Western blotting and RT–qPCR analysis. *P < 0.05, **P < 0.01, ***P < 0.001
exhibited complete loss of demethylase activity relative to that of ALKBH5-WT [16]. After transfection of
ALKBH5-WT or ALKBH5-H204A in HEK293T cells,
we found that ALKBH5-WT but not ALKBH5-H204A
decreased the mRNA level of PAQR4 (Fig. 2F), which
indicated that demethylation of ALKBH5 reduced the
expression of PAQR4. The results of co-transfection of
PAQR4 with either ALKBH5-WT or ALKBH5-H204A
in HEK293T cells provided further evidence supporting
our conclusion (Fig. 2G–H).
We constructed the psiCHECK-PAQR4 luciferase
reporter plasmid to detect the effect of m6A modification on PAQR4 expression. To construct the mutant
forms of PAQR4, all five predicted methylation regions
in the PAQR4 CDS and its 3′UTR were replaced
as shown in the Fig. 3A. Then, the PAQR4-WT or
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
PAQR4-MUT plasmid was co-transfected with ALKBH5
or si-ALKBH5. The results indicated that the translation
efficiency of PAQR4 was negatively correlated with the
expression of ALKBH5 (Fig. 3B). MeRIP-qPCR analysis
was then performed to reveal the correlation between the
protein level of ALKBH5 and the m6A level of PAQR4.
The m6A level of PAQR4 decreased when ALKBH5 was
overexpressed. Conversely, when ALKBH5 was knocked
down, the m6A level of PAQR4 increased (Fig. 3C).
Furthermore, we discovered that overexpression of
ALKBH5 reduced the PAQR4 mRNA degradation rate,
whereas knockdown of ALKBH5 accelerated this process
(Fig. 3D). In our research, it was shown that ALKBH5
expression was related to the decreased m6A methylation level of PAQR4, which led to decreased PAQR4
mRNA stability. Previous research has shown that m6A
“readers” in the YTHDF and IGF2BP families might play
key roles in mRNA methylation and in mRNA stability,
translation, and/or localization [24]. Here, we discovered
that when IGF2BP1 was knocked down, the mRNA level
of PAQR4 was reduced (Fig. 3E). This finding was consistent with the role of IGF2BP1 proposed in previous
reports [25]. We then used a RIP assay and observed that
PAQR4 mRNA was enriched in IGF2BP1-overexpressing cells, implying that PAQR4 mRNA may be degraded
through interaction with IGF2BP1 (Fig. 3F). To validated
that ALKBH5 and IGF2BP1 are really relevant with
PAQR4, correlation analyses were performed to show
that ALKBH5 was negative correlated with PAQR4 and
IGF2BP1 was positive correlated with that of PAQR4 in
the HCC samples (Fig. 3G–I). In summary, PAQR4 was
downregulated by ALKBH5-mediated m6A modification
in an IGF2BP1-dependent manner.
PAQR4 promoted HCC cell proliferation in vivo and in vitro
To investigate the biological function of PAQR4 in HCC,
PAQR4 was overexpressed in Hep3B and HLF cells and
knocked down in HLF and 97H cells. The CCK8 and
EdU incorporation assays showed that overexpression
of PAQR4 promoted the proliferation of Hep3B and
HLF cells and that knockdown of PAQR4 suppressed the
proliferation of HLF and 97H cells (Fig. 4A–E). In vivo
Page 5 of 16
experiments were also completed to demonstrate these
findings (Fig. 4F). The tumor weight and tumor volume
in the PAQR4 overexpression group were substantially
higher than those in the control group in the subcutaneous xenograft model (Fig. 4G and H). Correspondingly, PAQR4 overexpression was associated with higher
expression levels of the proliferation indices Ki-67 and
PCNA, according to immunohistochemistry (IHC)
(Fig. 4I).
PAQR4 promoted the migration and invasion of HCC cells
in vitro and promoted HCC metastasis in vivo
The wound healing, migration, and invasion assays
showed that PAQR4 promoted the migration and invasion of Hep3B and HLF cells and that knockdown of
PAQR4 suppressed the migration and invasion of HLF
and 97H cells (Fig. 5A–D and Additional file 2A, B). Since
epithelial-mesenchymal transition (EMT) is of great
importance in the migration and invasion of HCC cells
[26], we hypothesized that PAQR4 may promote migration and invasion by stimulating EMT signaling. Our
data obtained from immunofluorescence staining and
Western blot analysis indicated that the epithelial marker
E-cadherin was upregulated in PAQR4-overexpressing
Hep3B and HLF cells, while the mesenchymal marker
vimentin was downregulated in PAQR4-overexpressing
Hep3B and HLF cells (Fig. 5E and F). Consistent with
these findings, the staining results in tumor tissues also
showed the same trend (Fig. 5I). Collectively, these findings suggested that overexpression of PAQR4 activates
EMT in HCC. Moreover, overexpression of PAQR4 significantly increased the formation of intrahepatic metastatic foci compared to that in the control group in the
intrahepatic HCC implantation model (Fig. 5G and H).
Taken together, these results suggested that overexpression of PAQR4 promotes the metastasis of HCC cells.
To further confirm that the protective effect of
ALKBH5 could be blocked by the overexpression of
PAQR4, corresponding functional rescue assays were
performed. As EdU and CCK8 assays showed, overexpression of ALKBH5 decreased the proliferation capacity in 2 HCC cell lines, while co-overexpressed of PAQR4
(See figure on next page.)
Fig. 3 ALKBH5 downregulated PAQR4-mediated with IGF2BP1 in a m6A dependent manner. A Schematic representation exhibited totally 5
possible sites of m6A motifs in PAQR4 mRNA, and all of these positions were muted as described then constructed into psiCHECK-Vector plasmids
to investigate the m6A roles on PAQR4 expression. PAQR4-WT was constructed into psiCHECK-Vector plasmids as control. B The psiCHECK
PAQR4-MUT and psiCHECK PAQR4-WT plasmids were transfected into wild-type or ALKBH5-overexpression HLF cells and ALKBH5-knockdown LM3
cells for 24 h. Fluorescence intensity represented changes in PAQR4 transcriptional activity. C MeRIP-qPCR analysis indicated ALKBH5 overexpression
abolish m6a modification on PAQR4 mRNA in HLF cells while ALKBH5 knockdown enriched m6a modification on PAQR4 mRNA. D Actinomycin
D (ActD) assay showed overexpression of PAQR4 accelerate the degradation of mRNA whereas this process slowed down after ALKBH5 was
knockdown. E PAQR4 expression was measured via RT–qPCR in two IGF2BP1 knockdown HCC cells. F The result of RIP-qPCR indicated that IGF2BP1
could bind to PAQR4 mRNA in HLF and LM3 cells. GAPDH was used as internal controls in qPCR analysis. G Representative immunohistochemical
staining images of ALHBH5, IGF2BP1 and PAQR4 in human HCC tissues; scale bar: 200 μm. H Correlation analysis between ALHBH5 and PAQR4 in
human HCC tissues. I Correlation analysis between IGF2BP1 and PAQR4 in human HCC tissues. *P < 0.05, **P < 0.01, ***P < 0.001
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
Page 6 of 16
Fig. 3 (See legend on previous page.)
reverted this phenomenon (Additional file 3A and B).
Wound healing and transwell assays showed that overexpression of ALKBH5 decreased the invasion and migration ability, while co-overexpressed of PAQR4 reverted
this phenomenon (Additional file 3C and D). To sum up,
the protective effect caused by upregulation of ALKBH5
in HCC cell lines may achieved by reducing the effect of
PAQR4.
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
Page 7 of 16
Fig. 4 PAQR4 promoted HCC cells proliferation in vivo and in vitro. CCK8 and EdU assays confirmed that overexpression of PAQR4 promote
proliferation in Hep3B and HLF cells (A, C and E), whereas knockdown of ALKBH5 repressed cell proliferation in HLF and 97H cells (B, D and E). F
Representative images of tumor xenograft models using PAQR4-overexpressing HLF cells (n = 5), HLF-Vector cells were used as control. (G and H)
Tumor weight and volume of PAQR4-overexpressing and control group were shown. I The Immunohistochemistry (IHC) assays showed expression
of ki-67 and PCNA in the tumor xenograft model. Scale bar, 20 μm (40 ×). *P < 0.05, **P < 0.01, ***P < 0.001
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
Page 8 of 16
Fig. 5 PAQR4 promoted the migration and invasion capability of HCC cells in vitro and promoted HCC metastasis in vivo. Transwell and wound
healing assays showed that overexpression of PAQR4 promote the capability of migration and invasion in Hep3B and HLF cells (A, C and Additional
file 2A and B). Knockdown of PAQR4 suppress the capability of migration and invasion in HLF and 97H cells (B, D and Additional file 2A and B).
Immunofluorescence and Western blot analysis EMT markers like E-ca, N-ca, Vimentin in Hep3B and HLF cells (E, F). Representative images of
intrahepatic metastasis model using PAQR4-overexpressing HLF cells (n = 5), HLF-Vector cells were used as control (G). The numbers of metastasis
nodules (H). The Immunohistochemistry (IHC) assays showed expression of EMT markers E-ca and N-ca on tumor sections. Scale bar, 20 μm (40 ×). i
*P < 0.05, **P < 0.01, ***P < 0.001
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
PAQR4 promoted the development of HCC by activating
AKT
To further explore how PAQR4 promotes the development of HCC, silver staining and subsequent mass spectrometry were performed to identify potential binding
partners of PAQR4 (Fig. 6A). Interestingly, we found
that the characteristic peptides of AKT were identified by mass spectrometry. In line with this result, we
conducted exogenous co-IP assays in HEK293T cells
(Fig. 6B) and performed an immunofluorescence assay,
thus showing the colocalization of PAQR4 and AKT
in the cytoplasm (Fig. 6C). The results of both of these
assays suggested an interaction between PAQR4 and
AKT. In previous studies, it has been shown that PAQR4
promotes tumorigenesis and metastasis via the PI3K/
AKT pathway in hepatocellular carcinoma and prostate
cancer [27, 28]. We detected the activation of AKT in
wild-type and PAQR4-overexpressing HCC cells stimulated with or without Insulin-like growth factor-1 (IGF1), which is regarded as an activator of the AKT pathway
[29, 30]. The Western blot analysis results revealed that
PAQR4 overexpression increased the phosphorylation of
AKT regardless of IGF-1 stimulation (Fig. 6D). We also
knocked down AKT in PAQR4-overexpressing HCC cells
to evaluate cell proliferation, migration, and invasion.
Knockdown AKT significantly prevented the increases in
proliferation, migration, and invasion induced by PAQR4
overexpression in HCC cells (Fig. 6E–I, Additional
file 4A–C). Taken together, these data suggested that
PAQR4 enhanced proliferation, migration, and invasion
by activating AKT in HCC cells (see Additional file 5).
High expression of PAQR4 is associated with poor
prognosis in HCC
To explore the potential implications of PAQR4 in the
prognosis of patients with HCC, we elected to perform
studies of PAQR4 data from multiple databases. The data
from The Cancer Genome Atlas (TCGA) database indicated that the PAQR4 level was significantly increased in
tumor tissues (Fig. 7A). In the same database, the results
of Kaplan–Meier analysis revealed that high expression
of PAQR4 was significantly correlated with shorter OS
and DFS times (Fig. 7B, C). In addition, in six different
GSE datasets, PAQR4 was found to be highly expressed in
HCC patients (Fig. 7D). Interestingly, our study showed
that a higher expression level of PAQR4 was detected in
HCC tumor tissues than in the adjacent normal tissues
in the diethylnitrosamine (DEN)-induced mouse model
(Fig. 7E). IHC staining of a total of 108 pairs of clinical
HCC tissues demonstrated that the expression of PAQR4
was significantly increased in tumor tissues (Fig. 7F).
Patients with higher PAQR4 expression levels exhibited
worse prognoses, and the same was true in our cohort
Page 9 of 16
(Fig. 7G, H). These clinical data supported our finding
that high expression of PAQR4 was associated with poor
prognosis in HCC. Together, our results demonstrated
that ALKBH5- mediated demethylation of PAQR4
mRNA resulted in downregulation of PAQR4. This study
also implies that PAQR4 plays an important role in HCC
development and is an important biomarker for poor
prognosis in HCC (Fig. 7I).
Discussion
For decades, an increasing number of studies have shown
that m6A mRNA modification affects the biology of
normal and tumor cells [31–34]. STM2457, a new bioavailable inhibitor of METTL3, was developed in 2021
as a new drug to treat myeloid leukemia [35]. As the first
methylase-targeted anticancer drug, it was proven to be
effective and safe in vivo, constituting an example of successfully using m6A as a therapeutic target [36]. In the
present study, we found that the expression of ALKBH5
was decreased in HCC. Overexpression of ALKBH5
suppressed the proliferation, migration, and invasion of
HCC cells in vitro, while knockdown of ALKBH5 exhibited the opposite effects. As shown in our previous publication [22], in vivo experiments to demonstrate the
antitumor effect of ALKBH5 were not repeated in this
study. The data from transcriptome analyses inspired us
to search for possible downstream targets of ALKBH5.
After verification of mRNA and protein levels, we established PAQR4 as our research target of an ALKBH5 substrate molecule. Subsequently, we observed that the effect
of ALKBH5 on PAQR4 was dependent on the activity of
the enzyme and was abolished by the ALKBH5 H204A
mutation, which suggests that the function of ALKBH5
was mediated through m6A modification. The results
of MeRIP analysis and the luciferase reporter assay also
confirmed this conclusion. After identifying the eraser
of m6A in PAQR4 mRNA and clarifying its effects on
PAQR4, we next looked for the m6A reader, which can
“recognize” PAQR4 mRNA methylation. YTH domaincontaining proteins and insulin-like growth factor proteins function as m6A readers to regulate mRNA stability
and translation [12, 24, 37–40]. Surprisingly, we found
that IGF2BP1 bound to PAQR4 mRNA and prevented
its degradation. However, other potential m6A readers
of PAQR4 have not been identified and require further
investigation.
Progestin and AdipoQ Receptor 4 (PAQR4) belongs
to the progestin and adipoQ receptor family. This family includes 11 proteins (PAQR1 to PAQR11) that encode
functional receptors, as their names suggest [41]. In
recent studies, PAQR4 was demonstrated to act as an
oncogene in multiple cancers. PAQR4 was discovered
to disrupt the interaction between Nrf2 and Keap1 in
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
Page 10 of 16
Fig. 6 PAQR4 promoted the development of HCC by activating AKT. A Silver staining indicated PAQR4 may interact with AKT. B
Co-immunoprecipitation (Co-IP) assay identified the interaction between PAQR4 and AKT. C The colocalization of PAQR4 and AKT was observed
by immunofluorescence in HLF and HepG2 cells. D Western blot analysis showed activation of AKT pathway with or without IGF-1 stimulation
when PAQR4 was overexpressed. (E, F and Additional file 4A) CCK8 and EdU assays showed knockdown of AKT reduce proliferation caused by
overexpression of PAQR4. (G, H and Additional file 4B, C) Transwell and wound healing assays showed that knockdown of AKT reduce increased
migration and invasion capability caused by overexpression of PAQR4. GAPDH was used as internal controls in Western blotting analysis. *P < 0.05,
**P < 0.01, ***P < 0.001
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
Page 11 of 16
Fig. 7 High expression of PAQR4 is associated with poor prognosis in HCC. A PAQR4 expression in the normal and HCC data were obtained from
TCGA and the (B, C) Kaplan–Meier survival curves disease-free survival of patients with HCC from TCGA. D PAQR4 expression in human HCC data
were obtained from 6 GEO datasets. E PAQR4 expression in DEN-induced mouse liver cancer. F Representative images of Immunohistochemistry
(IHC) from human HCC samples and the expression of PAQR4 was compared between Adjacent tissue and tumor tissue according to the IHC
scoring. G, H Kaplan–Meier survival curves and disease-free survival of patients with HCC from Tongji cohort. i Schematic representation of a model
for the role of PAQR4 in HCC. *P < 0.05, **P < 0.01, ***P < 0.001
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
non-small-cell lung cancer (NSCLC), preventing Nrf2
protein degradation and thereby decreasing the sensitivity of malignant cells to chemotherapy [42]. PAQR4
inhibits SKP2-mediated ubiquitination of CDK4 by competitive binding to the binding site, contributing to cancer cell proliferation and carcinogenesis [43]. In breast
cancer, PAQR4 was discovered to be located in the Golgi
apparatus, where it reduces cytotoxicity and generates
S1P to aid cell development [44]. In HCC, PAQR4 promotes the development of HCC through the PI3K/AKT
pathway [27], as also shown in our study. In this research,
we demonstrated that PAQR4 promotes the proliferation,
invasion, and metastasis of HCC cells in vivo and in vitro.
We found that the expression of PAQR4 was significantly
upregulated in our patient cohort and in TCGA data.
In the diethylnitrosamine (Den)-induced mouse model
of liver cancer, we also observed high expression levels
of PAQR4 and found that it was correlated with shorter
overall survival and disease-free survival rates. These
data indicate that PAQR4 might be used as a prognostic
factor in HCC patients.
Overall, AKT signaling has many biological functions
in normal and tumor cells [29, 45]. Particularly in HCC,
AKT is regarded as a “driver” of cancer development and
malignant behaviors [46, 47]. In this study, we found that
overexpression of PAQR4 facilitated but knockdown of
PAQR4 counteracted the activation of AKT. In addition,
we knocked down AKT in PAQR4-overexpressing cancer cells in vitro and found that the proliferation, invasion, and migration of these cells were restored. Based
on our existing data, we concluded that PAQR4 may
exert its tumor-promoting effects by activating the AKT
pathway in HCC. This conclusion may be further supported by using the PI3K/AKT signaling-specific inhibitor LY294002 in mice in future experiments.
Conclusions
In our studies, we found that ALKBH5 inhibits HCC
growth by downregulating PAQR4 expression in an
IGF2BP1-dependent manner. Furthermore, we found
that PAQR4 enhanced proliferation, migration, and invasion by activating AKT in HCC cells.
Methods
Clinical samples or patients and specimens
All HCC specimens and paired adjacent normal tissues
were collected from patients who underwent curative
surgical resection between 2012 and 2017 at the Hepatic
Surgery Center, Tongji Hospital of Huazhong University
of Science and Technology (HUST; Wuhan, China). A
set of 108 pairs of HCC and adjacent normal tissues were
used for immunohistochemistry (IHC) and prognostic
analysis. Another 80 pairs of HCC and adjacent normal
Page 12 of 16
tissues were used for Western blotting (WB) to detect the
expression of PAQR4. Informed consent for the use of
human material was obtained from each patient, and the
study was approved by the Ethics Committee of Tongji
Hospital of HUST.
Animal models
The animal experimental protocols and ethical considerations were approved by the Institutional Animal Care
and Treatment Committee of Huazhong University of
Science and Technology. Four-week-old male BALB/c
nude mice were purchased from HFK BioTechnology
and maintained under specific pathogen-free conditions. For the subcutaneous xenograft model, 1 × 106
tumor cells were resuspended in 100 μL of DMEM and
then injected into the axilla of each mouse (n = 5 mice
per group). The mice were sacrificed after 2 weeks, the
tumor weight was recorded, and the tumor volume was
calculated according to the following equation: volume = 1/2*(length × width2). After the experiment,
the specimens were fixed with 4% formaldehyde. For
the intrahepatic tumor implantation model, 1 × 106
tumor cells were resuspended in 100 μL of DMEM and
then inoculated under the capsule of the right lobe of
the liver (n = 5 mice per group). The mice were sacrificed
after 4 weeks, and the tumors were removed and subjected to HE staining.
Cell lines and cell culture
The HCC cell lines LM3, HepG2, and HEK293T were
purchased from the China Center for Type Culture Collection (Wuhan, China), and the HCC cell lines HLF,
Hep3B, and 97H were obtained from the Hepatic Surgery
Center, Tongji Hospital of Huazhong University of Science and Technology (HUST; Wuhan, China). All of the
above cells were cultured in Dulbecco’s modified Eagle’s
medium (DMEM, Gibco, Grand Island, NY) supplemented with 10% fetal bovine serum (FBS; Gibco, Grand
Island, NY) and incubated in 5% CO2 at 37 °C.
In vitro cell proliferation assays
An EdU Cell Proliferation Kit (Beyotime) was used to
evaluate cell proliferation in vitro. Cells were plated into
24-well plates at a density of 4.0 × 105 cells per plate.
After adherent growth, EdU (10 μM) was added to the
cells and incubated at 37 °C for 1 h. Then, PBS was used
to remove the DMEM and preserve the EdU probe. Cells
were fixed with 4% formaldehyde for 30 min. After being
washed twice with PBS, the cells were permeabilized in
0.3% Triton X-100 for 10 min. Click Additive Solution
was then added to the cultured cells and incubated at
room temperature in the dark for 30 min. Nuclei were
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
stained with Hoechst for 10 min. Images were acquired
by microscopy.
Colony formation assay
HCC cells were seeded into 6-well plates (200 cells/well)
and further cultured at 37 °C for 2 weeks. The cells were
then fixed with 4% formaldehyde for 30 min and stained
with 0.1% crystal violet for 15 min. Then, the colonies
were counted.
CCK8 assay
The in vitro HCC cell viability assay was conducted using
a Cell Counting Kit-8 (CCK8). Cells were seeded into
96-well plates (100 cells/well) at 37 °C in 5% CO2. Then,
100 µl of CCK8 solution diluted tenfold with DMEM was
added to each plate and incubated at 37 °C for 1 h. Cell
proliferation activity was then evaluated with a microplate reader.
Migration and invasion assays
According to the manufacturer’s protocol, Transwell
migration and invasion assays were performed using
a 24-well Transwell plate containing membranes with
8-μm pores (Costar, Cambridge, MA). For the migration
assay, 3 × 104 cells resuspended in 100 µl of serum-free
DMEM were seeded in the upper chambers, while 700 µl
of DMEM containing 10% fetal bovine serum (FBS) was
placed in the lower chambers. For the invasion assay, the
membranes in the upper chambers were first coated with
50 μL of a 1:4 mixture of Matrigel (BD Biosciences, USA):
DMEM. Then, 3 × 104 cells resuspended in 100 µl of
serum-free DMEM were seeded in the upper chambers,
while 700 µl of DMEM containing 10% FBS was placed in
the lower chambers. After culture for 24 h in 5% CO2 at
37 °C, the cells that passed through the lower surface
of the membranes were fixed with 4% formaldehyde
for 30 min at room temperature and then stained with
0.1% crystal violet at room temperature for 15 min. The
cells were imaged by bright-field microscopy and then
counted with ImageJ software (NIH open source image
processing software). Each experiment was repeated at
least 3 times.
Wound healing assay
A wound healing assay was used to evaluate the cell
migration ability in vitro. HCC cells were seeded in
6-well plates at 90% to 100% confluence. Sterilized 10
μL pipette tips were used to scratch across the surface
of the plate to create an artificial wound. The cells were
then washed with PBS 3 times to remove detached cells
and cell debris. This step was followed by imaging using
a bright-field microscope 0 h and 24 h after the scratch
was made. Each experiment was repeated at least 3 times.
Page 13 of 16
Western blotting (WB)
Cells were lysed with RIPA buffer containing protease
inhibitor cocktail and phosphatase inhibitor cocktail
(MedChemExpress, USA) at 4 °C for 30 min. Protein
concentrations in the cell lysates were quantified by using
a BCA Protein Assay Kit (Thermo Fisher Scientific).
Briefly, equal amounts of protein were then separated
by SDS–PAGE (Boster Biological Technology, Wuhan,
China) and transferred to polyvinylidene fluoride (PVDF)
membranes (Millipore, USA). The membranes were
blocked with 5% bovine serum albumin (BSA) at 37 °C
for 60 min and then incubated with primary antibodies
at 4 °C overnight. Next, the membranes were incubated
with goat anti-rabbit or goat anti-mouse immunoglobulin as secondary antibodies at 37 °C for 1 h. Finally, the
membranes were visualized with an ECL detection system (Bio-Rad Laboratories). Image Lab™ 4.0 software
was used to analyze the Western blot band densities, and
GAPDH was used as the control for normalization to the
protein content.
Antibodies and reagents
Primary antibodies against PAQR4 (13401-1-AP for WB
and IHC), GAPDH (60004-1-Ig for WB), and ALKBH5
(16837–1-AP for WB) were purchased from Proteintech
(Wuhan, Hubei, China). Primary antibodies against Flag
(F1804 for WB and IP) and HA (H6908 for WB and IP)
were purchased from Sigma (St. Louis, MO, USA). Primary antibodies against AKT (#4691 for WB and IF),
phospho-Akt (Ser473) (#4060 for WB), Vimentin (#5741
for WB, IF and IHC), and Snail (#3879 for WB) were
purchased from CST (Danvers, MA, USA). Primary
antibodies against E-cadherin (#610182 for WB, IF
and IHC) and N-cadherin (#610920 for WB) were purchased from BD Biosciences (NJ, USA). Primary antibody against IGF2BP1 (#A1517 for IHC) was purchased
from ABclonal (Wuhan, Hubei, China). Primary antibody
against ALKBH5 (#ab195377 for IHC) was purchased
from Abcam Technology.
Coimmunoprecipitation (co‑IP)
Cells were lysed in IP lysis buffer containing protease
inhibitor cocktail and phosphatase inhibitor cocktail
(MedChemExpress, USA) at 4 °C for 30 min. Afterward, the cell lysates were immunoprecipitated with the
indicated antibodies at 4 °C overnight. On the second
day, the lysates were incubated with Magnetic Agarose
Beads (Biolinkedin, Shanghai, China) for 1 h prior to 3
washes with Binding/Washing buffer (0.5 M NaCl and
20 mM Na2HPO4, pH 7.0). The washed beads were eluted
with 2 × loading buffer and then subjected to Western
blot analysis. The protein interacting with PAQR4 was
screened by immunoprecipitation, silver staining and
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
mass spectrometry. Cell extracts from 293 T cells transfected with Flag-tagged PAQR4 or Flag-tagged vector
were immunoprecipitated with anti-Flag antibody. The
corresponding IP experiment steps were described above
and samples were carried out silver staining to ensure
samples quality. Then eluted proteins were identified by
mass spectrometry.
Immunofluorescence (IF) analysis
A total of 3 × 105 cells were seeded into glass dishes, and
8 h later, the cells were fixed with 4% formaldehyde for
15 min at room temperature. Then, the cells were washed
3 times and permeabilized with 0.1% Triton X-100 for
30 min. After blocking with 5% bovine serum albumin
(BSA) at 37 °C for 60 min, the cells were incubated with a
primary antibody overnight at 4 °C. The next day, the cells
were washed three times and incubated with a fluorescent secondary antibody for 2 h at room temperature in
the dark. Finally, nuclei were stained with 4′,6-diamidino2-phenylindole (DAPI; Sigma–Aldrich) for 8 min. Images
were acquired with a laser scanning confocal microscope.
Immunohistochemistry (IHC)
Tumor tissues were fixed with 4% formaldehyde for 1 day
at room temperature, embedded in paraffin and sliced
into sections. The paraffin sections were first warmed
to 65 °C for 30 min and then placed sequentially into
xylene, xylene, 100% alcohol, 100% alcohol, 95% alcohol,
90% alcohol, 80% alcohol, and 70% alcohol for dewaxing. Then, the slices were heated to 95 °C for 30 min
for antigen repair. After the slices were cooled to room
temperature and then washed 3 times with PBS, they
were immersed in 3% H
2O2 for 10 min to quench endogenous peroxidase activity. The slices were blocked with
5% bovine serum albumin at 37 °C for 60 min and then
incubated with primary antibodies at 4 °C overnight. On
the second day, after being washed 3 times with PBS,
the slices were incubated with a horseradish peroxidase
(HRP)-conjugated secondary antibody for 1 h. Finally,
DAB (Dako; Agilent Technologies, Inc.) was used to visualize staining. Staining was observed under a microscope
after hematoxylin counterstaining, dehydration, and sealing of the slices. Immunohistochemical staining scoring
of human HCC samples was performed by two pathologists in double blind. In each example, five visual fields
were examined under a microscope, and scores were
assigned based on the number of positive cells and the
staining intensity. If fewer than 5% of the total cells are
stained, the number of positive cells receives a score of 0,
26% to 50%, 51% to 75%, and 76% to 100%, respectively.
According to the staining intensity standard, 0 is colorless, 1 is pale yellow, 2 is brown, and 3 is brown. The final
staining score is the product of two scores.
Page 14 of 16
RNA extraction and RT–qPCR
RNA was extracted by using a FastPure Cell/Tissue Total
RNA Isolation Kit (Vazyme) following the manufacturer’s
instructions. cDNA libraries were prepared by using HiScript II Q Select RT SuperMix for qPCR (Vazyme). Then,
the number of transcripts was quantified with the CFX96
Touch™ Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA) by using ChamQ SYBR qPCR Master Mix
(Vazyme) following the manufacturer’s instructions. The
amounts of target gene transcripts were normalized to the
internal gene GAPDH. Each experiment was repeated at
least 3 times. The All primers used were as follows: GAPDHF, 5′-GACAAGCTTCCCGTTCTCAG-3′; GAPDH-R,
5′-GAGTCAACGGATTTGGTCGT-3′; PAQR4-F, 5′-FTAC
CTGCACAACGAACTGGG-3′; PAQR4-R, 5′ -AAGAGG
TGATAGAGCACGGAG-3′; ISG-15-F, 5′ -CGCAGATCA
CCCAGAAGATCG-3′; ISG-15-R, 5′ -TTCGTCGCATTT
GTCCACCA-3′; HERC6-F, 5′ -CCCTCAGTGGGCGTA
ATGTC-3′; HERC6-R, 5′ -AGAGCGATTGTCTCCAAA
TGTG-3′; SYT14-F, 5′ -AAATACAGTCCTCTATCG
GCAGA-3′; SYT14-R, 5′ -TTGGGCACTTGTTATATG
AGCAT-3′; TAP1-F, 5′ -CTGGGGAAGTCACCCTACC-3′;
TAP1-R, 5′ -CAGAGGCTCCCGAGTTTGTG-3′; FYB1-F,
5′ -GCAGGCCAAAGATTCGGAAC-3′; FYB1-R, 5′ -GGA
GGCCAGGGAAATGTAGG-3′; ALKBH5-F, 5′ -CGGCGA
AGGCTACACTTACG-3′; ALKBH5-R, 5′ -CCACCAGCT
TTTGGATCACCA-3′; IGF2BP1-F, 5′ -GCGGCCAGTTCT
TGGTCAA-3′; IGF2BP1-R, 5′ -TTGGGCACCGAATGT
TCAATC-3′.
Plasmid construction and transfection
The human PAQR4 or ALKBH5 sequence was tagged
with Flag or HA at the N-terminus and inserted into the
pcDNA3.1 vector (Invitrogen, USA). Short hairpin RNAs
(shRNAs) targeting PAQR4 and ALKBH5 were inserted
into the pcDNA3.1 vector to construct knockdown plasmids. The ALKBH5 H204A sequence was constructed by
site-directed mutagenesis. Human PAQR4 and ALKBH5
were also inserted into the pLenti vector (Invitrogen,
USA) to construct overexpression plasmids. All the
above plasmids were confirmed by DNA sequencing.
A lentiviral system was used to construct cell lines with
stable knockdown and overexpression. Small interfering
RNA (siRNA) oligos were designed by RiboBio (Guangzhou, China) for gene silencing. Lipo3000 (Invitrogen)
was used in all of the above transfections according to
the manufacturer’s recommendations. The sequences of
the siRNAs are summarized as follows: PAQR4 si-1, GCA
GGCTCCGTGCTCTATCAC; PAQR4 si-2, CGTCTT
GCTCTGAGAGTTCAA; ALKBH5 si-1, GCTGCAAGT
TCCAGTTCAA; ALKBH5 si-2, CCTCAGGAAGAC
Wang et al. Experimental Hematology & Oncology
(2023) 12:1
AAGATTAGA; IGF2BP1 si-1, GGCTCAGTATGGTAC
AGTA; IGF2BP1 si-2, TGAAGATCCTGGCCCATAA .
RNA immunoprecipitation (RIP) assays
Cells were washed twice with precooled PBS and then
digested with RIP buffer. After ultrasonic lysis and centrifugation at 4 °C for 15 min, 20 µL of each sample was
removed as input. Different antibodies were added separately to the remaining samples, and IgG was also added
as the negative control. After overnight incubation at
4 °C, magnetic beads were added to each sample. After
magnetic beads were added, the samples were incubated
at 4 °C for 4 h. After 4 washes with buffer, TRIzol was
added to the tube to extract RNA. RNA was reverse transcribed into cDNA for subsequent experiments.
Statistical analysis
Statistical analyses of all data were carried out using
GraphPad Prism 8.3.0 (GraphPad, La Jolla, CA, USA)
software. All values are expressed as the means ± SEMs.
Comparisons between two groups were performed
by two-tailed Student’s t test or one-way or two-way
ANOVA. The Kaplan–Meier method with the log-rank
test was used to assess survival in different subgroups.
Differences between groups were considered significant
when the p value was < 0.05.
Abbreviations
HCC Hepatocellular carcinoma
ALKBH5 Alkb homolog 5
m6A N6-methyladenosine
PAQR4 Progestin and AdipoQ Receptor 4
qRT-PCR Quantitative reverse transcription PCR
RIP RNA immunoprecipitation
MeRIP Methylated RNA immunoprecipitation
EMT Epithelial-to-mesenchymal transition
Den Diethylnitrosamine
Supplementary Information
The online version contains supplementary material available at https://doi.
org/10.1186/s40164-022-00370-2.
Additional file 1. ALKBH5 inhibited the proliferation of HLF and 97H cells,
while knockdown of ALKBH5 in LM3 cells showed the opposite effect (A).
Overexpression of ALKBH5 reduced the migration and invasion of HLF and
97H cells, while knockdown of ALKBH5 in LM3 cells showed the opposite
effects (B–C).
Additional file 2. PAQR4 promoted the migration and invasion of HCC
cells in vitro (A–B).
Additional file 3. Overexpression of ALKBH5 decreased the proliferation
capacity in 2 HCC cell lines, while co-overexpressed of PAQR4 reverted this
phenomenon (A and B). Overexpression of ALKBH5 decreased the invasion and migration ability, while co-overexpressed of PAQR4 reverted this
phenomenon (C and D).
Additional file 4. Knockdown AKT significantly prevented the increases
in proliferation, migration, and invasion induced by PAQR4 overexpression
in HCC cells (A–C).
Page 15 of 16
Additional file 5: Table S1. Correlation between PAQR4 and clinicopathological characteristics in HCC (n = 108).
Acknowledgements
The authors would like to thank Miss Lanping Ding (Institute of OrganTransplantation, Tongji Hospital) for animal care.
Author contributions
Designed and conducted the experiments: WJW, QBH, ZBL and PZ. Performing
the experiments: WJW, QBH, and ZBL. Technical or material support: HWZ, ZBL,
YCL, FRL, BXZ, YC, and PZ. Data analysis: WJW, QBH and FRL. Wrote and edited
the manuscript of the manuscript: WJW, QBH, ZBL and YC. Obtained the funding and supervised the study: BXZ, ZBL and PZ. All authors read and approved
the final manuscript.
Funding
This study was supported by The National Natural Science Foundation of
China (No. 81874189 to Bixiang Zhang, No. 82103597 to Zhibin Liao); The State
Key Project on Infectious Diseases of China (No. 2018ZX10723204-003).
Availability of data and materials
All data generated or analyzed during this study are included either in this article or in the additional files and methods, tables, figures, and figure legends
files.
Declarations
Ethics approval and consent to participate
The present study was approved the Institutional Review Board of Tongji
Medical College of Huazhong University of Science and Technology.
Consent for publication
All authors have agreed to publish this manuscript.
Competing interests
No potential competing interest were disclosed.
Received: 9 July 2022 Accepted: 30 December 2022
References
1. Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics. CA Cancer J Clin.
2021;71(1):7–33.
2. Zhang J, He X, Wan Y, Zhang H, Tang T, Zhang M, Yu S, Zhao W, Chen L.
CD44 promotes hepatocellular carcinoma progression via upregulation
of YAP. Exp Hematol Oncol. 2021;10(1):54.
3. Moon AM, Singal AG, Tapper EB. Contemporary epidemiology
of chronic liver disease and cirrhosis. Clin Gastroenterol Hepatol.
2020;18(12):2650–66.
4. Yang S, Zhang H, Yang H, Zhang J, Wang J, Luo T, Jiang Y, Hua H. SEPHS1
promotes SMAD2/3/4 expression and hepatocellular carcinoma cells
invasion. Exp Hematol Oncol. 2021;10(1):17.
5. Zhang SZ, Zhu XD, Feng LH, Li XL, Liu XF, Sun HC, Tang ZY. PCSK9 promotes tumor growth by inhibiting tumor cell apoptosis in hepatocellular
carcinoma. Exp Hematol Oncol. 2021;10(1):25.
6. Zheng Y, Huang C, Lu L, Yu K, Zhao J, Chen M, Liu L, Sun Q, Lin Z, Zheng
J, et al. STOML2 potentiates metastasis of hepatocellular carcinoma
by promoting PINK1-mediated mitophagy and regulates sensitivity to
lenvatinib. J Hematol Oncol. 2021;14(1):16.
7. Ma JZ, Yang F, Zhou CC, Liu F, Yuan JH, Wang F, Wang TT, Xu QG, Zhou WP,
Sun SH. METTL14 suppresses the metastatic potential of hepatocellular
carcinoma by modulating N 6 -methyladenosine-dependent primary
MicroRNA processing. Hepatology. 2017;65(2):529–43.
Wang et al. Experimental Hematology & Oncology
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
(2023) 12:1
Aldrighetti L, Pulitano C, Catena M, Arru M, Guzzetti E, Halliday J, Ferla G.
Liver resection with portal vein thrombectomy for hepatocellular carcinoma with vascular invasion. Ann Surg Oncol. 2009;16(5):1254.
Jia G, Fu Y, Zhao X, Dai Q, Zheng G, Yang Y, Yi C, Lindahl T, Pan T, Yang Y-G,
et al. N6-Methyladenosine in nuclear RNA is a major substrate of the
obesity-associated FTO. Nat Chem Biol. 2011;7(12):885–7.
Huang H, Weng H, Chen J. m6A modification in coding and noncoding RNAs: roles and therapeutic implications in cancer. Cancer Cell.
2020;37(3):270–88.
Shen S, Zhang R, Jiang Y, Li Y, Lin L, Liu Z, Zhao Y, Shen H, Hu Z, Wei Y, et al.
Comprehensive analyses of m6A regulators and interactive coding and
non-coding RNAs across 32 cancer types. Mol Cancer. 2021;20(1):1.
Chen M, Wei L, Law CT, Tsang FH, Shen J, Cheng CL, Tsang LH, Ho DW,
Chiu DK, Lee JM, et al. RNA N6-methyladenosine methyltransferase-like
3 promotes liver cancer progression through YTHDF2-dependent posttranscriptional silencing of SOCS2. Hepatology. 2018;67(6):2254–70.
Lin X, Chai G, Wu Y, Li J, Chen F, Liu J, Luo G, Tauler J, Du J, Lin S, et al. RNA
m6A methylation regulates the epithelial mesenchymal transition of
cancer cells and translation of Snail. Nat Commun. 2019;10(1):1.
Du L, Li Y, Kang M, Feng M, Ren Y, Dai H, Wang Y, Wang Y, Tang B. USP48
is upregulated by Mettl14 to attenuate hepatocellular carcinoma via
regulating SIRT6 stabilization. Cancer Res. 2021;81(14):3822–34.
Huang Q, Mo J, Liao Z, Chen X, Zhang B. The RNA m6A writer WTAP in
diseases: structure, roles, and mechanisms. Cell Death Dis. 2022;13(10):1.
Zheng G, Dahl JA, Niu Y, Fedorcsak P, Huang CM, Li CJ, Vagbo CB, Shi Y,
Wang WL, Song SH, et al. ALKBH5 is a mammalian RNA demethylase that
impacts RNA metabolism and mouse fertility. Mol Cell. 2013;49(1):18–29.
Zhang S, Zhao BS, Zhou A, Lin K, Zheng S, Lu Z, Chen Y, Sulman EP, Xie
K, Bögler O, et al. m 6 A demethylase ALKBH5 maintains tumorigenicity
of glioblastoma stem-like Cells by sustaining FOXM1 expression and cell
proliferation program. Cancer Cell. 2017;31(4):591-606.e596.
Qiu X, Yang S, Wang S, Wu J, Zheng B, Wang K, Shen S, Jeong S, Li Z, Zhu
Y, et al. M6A demethylase ALKBH5 regulates PD-L1 expression and tumor
immunoenvironment in intrahepatic cholangiocarcinoma. Can Res.
2021;81(18):4778–93.
Liu F, Qin L, Liao Z, Song J, Yuan C, Liu Y, Wang Y, Xu H, Zhang Q, Pei Y, et al.
Microenvironment characterization and multi-omics signatures related
to prognosis and immunotherapy response of hepatocellular carcinoma.
Exp Hematol Oncol. 2020;9:10.
Lu JC, Zhang PF, Huang XY, Guo XJ, Gao C, Zeng HY, Zheng YM, Wang
SW, Cai JB, Sun QM, et al. Amplification of spatially isolated adenosine
pathway by tumor-macrophage interaction induces anti-PD1 resistance
in hepatocellular carcinoma. J Hematol Oncol. 2021;14(1):200.
Nie S, Zhang L, Liu J, Wan Y, Jiang Y, Yang J, Sun R, Ma X, Sun G, Meng H,
et al. ALKBH5-HOXA10 loop-mediated JAK2 m6A demethylation and
cisplatin resistance in epithelial ovarian cancer. J Exp Clin Cancer Res.
2021;40(1):284.
Chen Y, Zhao Y, Chen J, Peng C, Zhang Y, Tong R, Cheng Q, Yang B, Feng
X, Lu Y, et al. ALKBH5 suppresses malignancy of hepatocellular carcinoma
via m6A-guided epigenetic inhibition of LYPD1. Mol Cancer. 2020;19(1):9.
Qu S, Jin L, Huang H, Lin J, Gao W, Zeng Z. A positive-feedback loop
between HBx and ALKBH5 promotes hepatocellular carcinogenesis. BMC
Cancer. 2021;21(1):686.
Zaccara S, Ries RJ, Jaffrey SR. Reading, writing and erasing mRNA methylation. Nat Rev Mol Cell Biol. 2019;20(10):608–24.
Huang X, Zhang H, Guo X, Zhu Z, Cai H, Kong X. Insulin-like growth
factor 2 mRNA-binding protein 1 (IGF2BP1) in cancer. J Hematol Oncol.
2018;11(1):88.
Giannelli G, Koudelkova P, Dituri F, Mikulits W. Role of epithelial to
mesenchymal transition in hepatocellular carcinoma. J Hepatol.
2016;65(4):798–808.
Zhao G, Shi X, Sun Z, Zhao P, Lu Z. PAQR4 promotes the development of
hepatocellular carcinoma by activating PI3K/AKT pathway. Acta Biochim
Biophys Sin. 2021;53(12):1602–13.
Ye J, Gao M, Guo X, Zhang H, Jiang F. Breviscapine suppresses the growth
and metastasis of prostate cancer through regulating PAQR4-mediated
PI3K/Akt pathway. Biomed Pharmacother. 2020;127: 110223.
Yang W-L, Wang J, Chan C-H, Lee S-W, Campos AD, Lamothe B, Hur L,
Grabiner BC, Lin X, Darnay BG, et al. The E3 ligase TRAF6 regulates Akt
ubiquitination and activation. Science. 2009;325(5944):1134–8.
Page 16 of 16
30. Wang G, Long J, Gao Y, Zhang W, Han F, Xu C, Sun L, Yang SC, Lan J, Hou
Z, et al. SETDB1-mediated methylation of Akt promotes its K63-linked
ubiquitination and activation leading to tumorigenesis. Nat Cell Biol.
2019;21(2):214–25.
31. Cui Q, Shi H, Ye P, Li L, Qu Q, Sun G, Sun G, Lu Z, Huang Y, Yang CG, et al.
m(6)A RNA methylation regulates the self-renewal and tumorigenesis of
glioblastoma stem cells. Cell Rep. 2017;18(11):2622–34.
32. Li F, Yi Y, Miao Y, Long W, Long T, Chen S, Cheng W, Zou C, Zheng Y, Wu X,
et al. N6-methyladenosine modulates nonsense-mediated mRNA decay
in human glioblastoma. Can Res. 2019;79(22):5785–98.
33. Li Z, Weng H, Su R, Weng X, Zuo Z, Li C, Huang H, Nachtergaele S, Dong
L, Hu C, et al. FTO plays an oncogenic role in acute myeloid leukemia as a
N(6)-methyladenosine RNA demethylase. Cancer Cell. 2017;31(1):127–41.
34. Zhang Z, Liu F, Chen W, Liao Z, Zhang W, Zhang B, Liang H, Chu L, Zhang
Z. The importance of N6-methyladenosine modification in tumor immunity and immunotherapy. Exp Hematol Oncol. 2022;11(1):30.
35. Yankova E, Blackaby W, Albertella M, Rak J, De Braekeleer E, Tsagkogeorga
G, Pilka ES, Aspris D, Leggate D, Hendrick AG, et al. Small-molecule
inhibition of METTL3 as a strategy against myeloid leukaemia. Nature.
2021;593(7860):597–601.
36. Huang J, Shao Y, Gu W. Function and clinical significance of N6-methyladenosine in digestive system tumours. Exp Hematol Oncol.
2021;10(1):40.
37. Zaccara S, Jaffrey SR. A unified model for the function of YTHDF proteins
in regulating m(6)A-modified mRNA. Cell. 2020;181(7):1582–95.
38. Wang W, Shao F, Yang X, Wang J, Zhu R, Yang Y, Zhao G, Guo D, Sun Y,
Wang J, et al. METTL3 promotes tumour development by decreasing APC
expression mediated by APC mRNA N6-methyladenosine-dependent
YTHDF binding. Nat Commun. 2021;12(1):1.
39. Huang H, Weng H, Sun W, Qin X, Shi H, Wu H, Zhao BS, Mesquita A,
Liu C, Yuan CL, et al. Recognition of RNA N6-methyladenosine by
IGF2BP proteins enhances mRNA stability and translation. Nat Cell Biol.
2018;20(3):285–95.
40. Chen X, Zhou X, Wang X. m(6)A binding protein YTHDF2 in cancer. Exp
Hematol Oncol. 2022;11(1):21.
41. Tang YT, Hu T, Arterburn M, Boyle B, Bright JM, Emtage PC, Funk WD.
PAQR proteins: a novel membrane receptor family defined by an ancient
7-transmembrane pass motif. J Mol Evol. 2005;61(3):372–80.
42. Xu P, Jiang L, Yang Y, Wu M, Liu B, Shi Y, Shen Q, Jiang X, He Y, Cheng
D, et al. PAQR4 promotes chemoresistance in non-small cell lung
cancer through inhibiting Nrf2 protein degradation. Theranostics.
2020;10(8):3767–78.
43. Wang L, Zhang R, You X, Zhang H, Wei S, Cheng T, Cao Q, Wang Z, Chen
Y. The steady-state level of CDK4 protein is regulated by antagonistic
actions between PAQR4 and SKP2 and involved in tumorigenesis. J Mol
Cell Biol. 2017;9(5):409–21.
44. Pedersen L, Panahandeh P, Siraji MI, Knappskog S, Lønning PE, Gordillo
R, Scherer PE, Molven A, Teigen K, Halberg N. Golgi-localized PAQR4
mediates antiapoptotic ceramidase activity in breast cancer. Can Res.
2020;80(11):2163–74.
45. Luo X, Cao M, Gao F, He X. YTHDF1 promotes hepatocellular carcinoma
progression via activating PI3K/AKT/mTOR signaling pathway and inducing epithelial-mesenchymal transition. Exp Hematol Oncol. 2021;10(1):35.
46. Rebouissou S, Nault J-C. Advances in molecular classification and precision oncology in hepatocellular carcinoma. J Hepatol. 2020;72(2):215–29.
47. Song M, Bode AM, Dong Z, Lee M-H. AKT as a therapeutic target for
cancer. Can Res. 2019;79(6):1019–31.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
|
|
https://openalex.org/W4246435792
|
https://www.researchsquare.com/article/rs-29239/v2.pdf?c=1599174662000
|
English
| null |
Melatonin promotes adventitious root formation in apple by promoting the function of MdWOX11
|
Research Square (Research Square)
| 2,020
|
cc-by
| 6,929
|
Research article Keywords: activation of adventitious roots, adventitious roots, apple rootstocks, MdWOX11, melatonin,
transgenic plantlets Posted Date: September 3rd, 2020 DOI: https://doi.org/10.21203/rs.3.rs-29239/v2 License:
This work is licensed under a Creative Commons Attribution 4.0 International
License. R
d F ll Li License:
This work is licensed under a Creative Commons Attribution 4.0 International
License. Read Full License Version of Record: A version of this preprint was published on November 26th, 2020. See the published
version at https://doi.org/10.1186/s12870-020-02747-z. Page 1/19 Abstract Background: Melatonin (MT) is important for plant growth and development; however, it is not known
whether MT is involved in apple adventitious root (AR) development. In this study, we treated Malus
prunifolia (MP) at four different stages of AR development, and analysed the level of the endogenous
hormones MT, auxin (IAA), zeatin-riboside (ZR), gibberellins (GA1+3) and abscisic acid (ABA) in all four
treatment groups and in the untreated control group. The expression of genes related to MT, IAA
biosynthesis, transport and signal transduction, the cell cycle, and root development was quantified by
RT-qPCR, and the function of MdWOX11 was analysed in transgenic apple plants. Results: The promotion of AR development by MT was dependent on the stage of AR induction between
0–2 d in apple rootstocks. MT-treatment increased the level of IAA and crosstalk existed between MT and
IAA during AR formation. The expression of MdWOX11 was induced by MT treatment and positively
regulated AR formation in apple. Furthermore, transgenic lines that overexpressed MdWOX11 lines
produced more ARs than ‘GL3’. Phenotypic analysis indicated that MdWOX11 overexpression lines were
more sensitive to exogenous MT treatment than ‘GL3’, suggesting that MdWOX11 regulates AR formation
in response to MT in apple rootstock. Conclusions: MT promotes AR formation mainly during the AR induction stage by inducing IAA levels and
upregulating MdWOX11. Background Apple (Malus domestica) is a major commercial fruit tree that is cultivated globally and apple fruits have
a high nutritional and economic value. Malus prunifolia (MP) is widely known as the easiest-rooting
apple rootstock and it offers advantages such as good graft compatibility, cold resistance, salt and alkali
tolerance, and disease and insect resistance. The induction of adventitious roots (ARs) from stem basal
tissues is a major step in the vegetative propagation of apple rootstocks. ARs are post embryonic roots
that emerge from non-root organs and AR primordia arise from interfascicular cambium cells adjoining
phloem cells [1, 2]. The processes required for AR formation have been identified in different plants,
including rice [3], Arabidopsis [4, 5], and poplar [6]; however, methods for improving AR formation in apple
have not been studied. Melatonin (MT; N-acetyl-5-methoxytryptamine) is a well-known animal hormone and was initially
discovered in plants by two groups of workers in 1995 [7, 8]. Previous studies have shown that MT
functions as a regulatory signal in plants [9], and is important for the growth of roots, shoots, explants,
and in stress responses [10-15]. The relationship between MT and AR formation has mainly been studied
in herbaceous plants; for example, the exogenous application of MT promoted adventitious rooting in
tomato and rice [16, 17], but the mechanism how MT regulates AR formation remains to be elucidated in
woody plants such as apple. AR formation can be categorised as a four-stage process [18-21], and the
stage at which MT is important for AR development remains unknown. In this study, we observed that MT Page 2/19 Page 2/19 promoted AR formation at early stages of AR induction and initiation. Previous studies have
demonstrated the relationship between MT and other plant hormones such as auxin (IAA), cytokinin (CK),
gibberellins (GA), abscisic acid (ABA) [22]: Treatment of plants with MT led to an increase in CK levels
during non-biological stress [23], MT contributed to a high content of active GAs such as GA3 and GA4
[24], and exogenous MT application led to a decrease in the content of ABA [23]. However, the relationship
between MT and these hormones during adventitious rooting remains to be determined. Potentially, MT
acts as a growth-promoting compound, by increasing the level of IAA, IAA synthesis and polar IAA
transport [25-27]. Background Most studies have analysed the ability of MT to stimulate root and shoot growth, in a
similar way to IAA [28]. However, the effect of MT on root growth and differentiation is thought to be
independent of IAA [29]. In this study, we established that MT–IAA crosstalk plays an important role in AR
induction, but the role of plant hormone homeostasis and associated signaling networks during AR
formation in apple rootstock is incompletely understood. The genes involved in MT biosynthesis, such as TDC, SNAT, HIOMT, and ASMT, are induced by MT [30-
34], MzSNAT5 regulates MT synthesis in the mitochondria of apple [33], and overexpression of ASMT
increased MT production in Arabidopsis thaliana [34], in this study, the expression of MT biosynthesis
genes and auxin-related genes, such as AUXIN RESPONSE FACTORS (ARFs) and PINFORMED (PIN) genes
were analysed in apple. In addition, WUSCHEL-RELATED HOMEOBOX GENE 11 (WOX11) functions in
crown root emergence and development [35] and AR development in Arabidopsis [36], but in woody
plants, the regulation of WOX11 during AR development is poorly understood. Furthermore, it is unknown
whether AR formation in transgenic apple plants that overexpress MdWOX11 is regulated by exogenous
MT. Currently, the mechanisms via which AR formation is regulated by MT is not well characterised in apple
rootstock. In this study, we showed that exogenous MT induced AR formation at early stages of AR
induction and initiation by increasing IAA synthesis, transport, and the expression of signalling-related
genes. Apple plantlets treated with MT in tissue culture showed an increase in the expression of genes
related to root development, and therefore, an increase in the number of ARs. Furthermore, we
demonstrated that overexpression of MdWOX11 promoted the emergence and development of ARs, and
treatment with exogenous MT induced AR development in transgenic apple that overexpressed
MdWOX11. The results of this study provide insights into the mechanism how MT regulates AR
formation in apple rootstock. Results The aim of this study was to identify the precise time at which MT promotes AR formation in tissue-
culture plantlets of Malus prunifolia apple rootstocks. The study consisted of five different treatment
groups: MT, MT0–2, MT2–5, MT5–20 and one control group (Fig. 1A). No morphological changes were
evident in any of the groups up to 5 d; however, ARs emerged from basal stem parts at 10 d (Fig. 1B). At
20 d, the greatest number of ARs was observed in the MT0–2 group, and groups MT2–5 and MT5–20
produced more ARs than the control and MT group at 20 d, but no significant difference was observed in Page 3/19 the number of ARs among these groups (Fig. 1B). To observe the anatomy of stems at different stages of
AR formation, sections were made from paraffin-embedded samples and were viewed using light
microscopy. On day 0, cross-sections of the samples revealed the existence of competent cells, but
mitotic cambial cell division was observed at 5 d and cell divisions were visible in the compactly arranged
cells, and AR appeared in sections from stem bases cultured in medium for 10 days (Fig. 1C). No AR
formation was observed in MP treated for 20 d with the auxin inhibitors N-1-naphthylphthalamic acid
(NPA) or triiodobenoic acid (TIBA), but ARs were observed following MT treatment. All phenotypes are
summarised in Figure S1. We also measured the root rate, number of AR, crossings, root length, root surface area and root volume
in all five treatment groups, and the data were consistent with the phenotypes of AR formation. All
measured parameters were higher in the MT0–2 group than in other groups, and the minimum number of
ARs and values for other root parameters were observed in the control group (Fig. 2). The results showed
that MT mainly promoted AR formation at 0–2 d, during the AR induction stage. On the basis of their diameter, ARs were classified into three groups: 0–2.0 mm, 2.0–5.0 mm and >5.0
mm. According to AR number, length and surface area, the 0–2.0 mm category contained the greatest
percentage of the total; however, for root volume, the 2.0–5.0 mm category was the largest for all groups
(Fig. 3). The MT0–2 group contained the greatest number of ARs in the 0–2.0 mm class, which was twice
as many as in the control group (Fig. 3). Results We conclude that most ARs were fine roots (0–2.0 mm). The levels of the hormones MT, IAA, ZR, GA1+3 and ABA were analysed in MP tissue culture plantlets after
treatment with MT. The MT content was higher in the MT0–2 group than in other treatment groups during
the early AR developmental stage, and reached a peak at 5 d in the MT0–2 group. In the MT and MT0–2
treatments, the levels of IAA, ZR and GA1+3 were higher than those in the control group during AR
induction at 1 d and 2 d, but the levels were lower in the MT0–2 group than other groups at 10 d. The
level of ABA responded opposite to that of IAA, ZR and GA1+3 (Fig. 4) in the treatments. The expression of MdTDC1, MdHIOMT2, MdASMT1 and MdASMT2 genes, which are involved in MT and
IAA signal transduction, was higher in plants in the MT0–2 groups than in other groups, except at 0 and
20 d, and expression of MdSNAT and MdHIOMT1 was higher in MT0–2 than in other groups at 1 d, 5 d
and 20 d (Fig. 5). To determine whether an interactive effect between MT and IAA existed, we measured the expression of
genes related to IAA biosynthesis and signal transduction. The expression of MdYUCCA2, MdYUCCA10,
MdARF7 and MdARF19 was higher in plants in the MT0–2 treatment than that in other groups at 1 d, 2 d
and 3 d. The expression level of the IAA transport genes MdAUX1, MdPIN1, and MdPIN3 was also higher
in the MT0-2 than in other groups at 2 d; however, expression of the IAA signal transduction gene MdIAA5
was lower following MT0–2 treatment than in other treatments during the AR induction stage (Fig. 6). To investigate whether MT affects cell division, the expression of the cell-cycle genes MdCYCD1;1 and
MdCYCD3;1 was analysed, and these genes were more highly expressed in MT0–2 at 3 d and 10 d. Page 4/19 Therefore, we conclude that the application of MT promoted AR formation in apple rootstock, and RT-
qPCR analysis showed that the expression of genes related to root development was higher at most time
points in response to MT treatment. We observed that among all root-related genes, the expression of
MdWOX11 following MT treatment was 5.6 times higher than that in control plants at 2 d (Fig. 7). Results This
suggests that MdWOX11 probably plays an important role in AR induction in response to MT treatment. The expression of MdWOX11 was induced by IBA treatment (Fig. 7). We generated the overexpression
(OE) transgenic lines MdWOX11-OE15#, 16# and 20# in ‘GL3’, and confirmed the level of overexpression
MdWOX11 transgenic lines (Figure S2). To confirm whether MdWOX11 transgenic lines exhibited an
enhanced response to MT signalling, wild-type and transgenic apple plantlets growing in tissue culture
were either treated with 3.45 μM IBA as a control, or with MT for 0–2 d (MT0-2). More ARs were observed
in the MT0-2 group than in the control group and the overexpressing MdWOX11 transgenic lines and
‘GL3’, and the MdWOX11 overexpressing lines produced more ARs than ‘GL3’ (Fig. 8A). Overexpression of
MdWOX11 also caused an increase in the rate of ARs (Fig. 8B). Furthermore, MdWOX11 overexpressing
plants were more sensitive to exogenous MT treatment than wild type (Fig. 8A-C), which indicates that
MdWOX11 induced AR formation in response to MT treatment. Discussion AR formation is the key to vegetative propagation and previous studies have divided AR formation from
tissue culture plantlets into four stages: induction, initiation, primordium formation, and emergence [18-
21]. In this study, 0–2 d represents the stage of AR induction, 2–5 d represents AR initiation, and 5–20 d
covers the stages of AR primordium formation and AR emergence in MP. MT affects many biological
processes, including plant growth, flowering and stress responses, and although MT promotes AR
development in tomato [16], no studies have been conducted in woody plants. Therefore, to determine
when and how MT can promote AR formation, we treated MP plants with MT at different times. The
phenotypes of the five treatment groups showed that there was no difference between plants in the
MT2–5 and MT5–20 groups, but the MT0–2 group produced more ARs than other groups (Figs. 1, 2 and
3), which demonstrates that MT promotes AR formation mainly during the AR induction stage at 0–2 d. The effect of interaction between MT and IAA during root development is unclear: some studies have
suggested that a low concentration of exogenous MT can induce an increase in the endogenous IAA
content in plants, and it is believed that the promoting effect of MT on growth might be caused by this
increase in IAA content [25]. However, other research has shown that the regulation of root growth and
differentiation by MT was independent of IAA [29]. To analyse whether MT treatment can elevate the IAA
content and enhance IAA signalling, we measured the IAA content in plants in the five MT treatments. The
IAA content mainly increased during AR induction after MT application, but decreased during the AR
initiation and AR emergence stages (Fig. 4). This might reflect that IAA plays an important role during the
early stage of root development [37-39], MP plantlets were treated with auxin inhibitors and MT during AR
formation (Figure S1), NPA and TIBA treatment inhibits AR formation, and we conclude that exogenous
MT treatment promoted AR formation in the presence of IAA. However, RT-qPCR data suggest that MT is Page 5/19 Page 5/19 involved in IAA signalling pathways, and MT treatment can induce the expression of genes involved in
IAA biosynthesis, transport and signal transduction (Fig. 6). Therefore, MT potentially promotes AR
induction by increasing IAA levels and IAA signalling. Discussion Previous research has shown that MT-treated plants
have increased CK levels during non-biological stress [23], and MT also led to an increase in the content
of active GAs such as GA3 and GA4 [24]. The observed increase in the levels of GA and ZR following MT
treatment also suggests that there is a link between MT and ZR or GA1+3. WOX11 regulates AR formation in response to IAA in Arabidopsis [36] and some studies have
demonstrated that WOX11 is induced by IAA and positively induces the expression of LATERAL ORGAN
BOUNDARIES DOMAIN16 (LBD16) and LBD29 at an early stage of AR development [40]. However, little is
known concerning the function of MdWOX11 in woody plants such as apple, including the morphological
changes that occur in transgenic apple plants in response to MT during AR formation. In this study, we
generated transgenic apple plants that expressed 35S::WOX11-OE, and their phenotype demonstrated
that MdWOX11 is a positive regulator of AR activation and that MT treatment of these MdWOX11
overexpressing plants increased AR development (Fig. 8). Therefore, the induction of MdWOX11 and its
related genes might represent a possible mechanism by which MT promotes AR formation. This is the
first study to use MdWOX11 transgenic lines to investigate the role of MdWOX11 in response to MT
during AR induction. Collectively, the data indicate that MdWOX11 promotes AR formation in response to
MT, and provide insights into the molecular mechanisms that underlies the induction of ARs by MT. Conclusions Melatonin promotes adventitious root formation mainly at the stage of AR induction by increasing auxin
levels and activating the function of MdWOX11. The results represent potential to improve AR formation
to accelerate the asexual reproduction of apple rootstock that is difficult to root. Explant growth conditions and MT treatments Tissue culture plantlets of Malus prunifolia apple rootstock were grown in tissue culture at the Northwest
Agriculture and Forestry University, Yangling (108°04′E, 34°16′N), China, and were used as plantlets for AR
formation. The plantlets of MP were imported from Aomori in Japan and were propagated by asexual
reproduction. The tissue culture plantlets of MP were split into five groups, and plants in all groups were
treated simultaneously for 20 d. Control plants were cultured on inducing-root medium containing half-
strength MS supplemented with 3.45 μM indole-3-butyric acid (IBA) to promote root formation of roots. The second group was cultivated on medium containing half-strength MS supplemented with 1.29 μM
MT and 3.45 μM IBA and was designated as the MT treatment group. The third group was transferred to
root-inducing medium after cultivation on MT medium for 2 d and was called MT 0–2. The fourth group
(MT 2–5) was transferred to MT medium after culture on root-inducing medium for 2 d, and was then
transferred to root-inducing medium after culture on MT medium for 3 d. The fifth group was transferred Page 6/19 Page 6/19 to MT medium after cultivation on root-inducing medium for 5 d, and was called MT5–20. The
composition of the different media used for this study is listed in Supplemental Table S1. Samples were
harvested from all five groups at 0 d, 1 d, 2 d, 3 d, 5 d, 10 d and 20 d (even though some samples were
collected prior to MT-treatment). In total, 3150 cuttings, consisting of 630 cuttings from each of the five
groups were harvested, which in turn, consisted of 90 cuttings sampled at each sampling point. Samples
were collected from the basal portion of the stems, including the AR formation zone (approximately 0.5
cm). The plants in the NPA treatment group were continuously cultivated in 10 μM NPA and 1.29 μM MT
for 20 d, and those in the TIBA treatment were continuously cultivated in 10 μM TIBA and 1.29 μM MT for
20 d, control is the same as above. Overexpression of MdWOX11 transgenic apple (35S::MdWOX11-OE)
and ‘GL3’ were divided into two groups: one group was cultured on 3.45 μM IBA and served as the
controls, and the other group was transferred to root-inducing medium for 20 d after culturing in MT
medium for 2 d. Anatomical observations and morphological measurements Anatomical observations were carried out according to previously described protocols [1, 41, 42]. The
morphological parameters measured included: AR rate, AR length and mean AR number for each cutting
[43]. In addition, crossings, root length, root surface area and root volume were analysed by an Epson
Expression 10000XL scanner (LA l600 scanner, Canada). In total, 90 cuttings were analysed, with 90
cuttings from each group harvested at each sampling point. The collected samples were immediately
frozen in liquid nitrogen and stored at −80 °C for further analysis. Measurement of hormone levels Samples for hormone extraction were harvested at different time points from all five treatment groups. Hormones were purified and extracted according to previously described procedures [44]. Three biological
replicates for each group at each sampling point were analysed. The enzyme-linked immunosorbent
assay (ELISA) technique was used to detect and analyse the level of hormones [44]. Extraction of RNA and synthesis of cDNA Total RNA was isolated using the CTAB-based extraction method [45] and the total RNA integrity was
analysed by electrophoresis of the samples on 2% agarose gels. cDNA was synthesised using the Prime
Script RT Reagent Kit with gDNA Eraser (TaKaRa Bio, Shiga, Japan). Availability of data and materials All data generated or analysed during this study are included in this published article and its
supplementary information files. Statistical analysis Significance differences among each treatment and time-point were determined using (ANOVA), with
significant differences between means determined at the 5% probability level using SPSS11.5 software
(SPSS, Chicago, IL, USA). The figures were generated using SigmaPlot12.0 (Systat Software, Inc.). Author contributions J.M., D.Z. and M.H. designed the research study. J.M., C.N., K.L. and S.C. performed the research. J.M.,
C.N. M.M. and analysed the data. J.M. and D.Z. wrote the paper. All authors approved the manuscript. A k
l d
t J.M., D.Z. and M.H. designed the research study. J.M., C.N., K.L. and S.C. performed the research. J.M.,
C.N. M.M. and analysed the data. J.M. and D.Z. wrote the paper. All authors approved the manuscript. RT-qPCR analyses The expression of genes related to MT, IAA synthesis, transport and signal transduction, the cell cycle and
root development was quantified by RT-qPCR. The gene names and abbreviations, the MDP annotations
in apple, as well as homologous proteins and species on which the identification of the proteins in apple
was based, are listed in Supplemental Table S2. Primer design was based on previous research [43] and
all gene-specific primers for the analysed genes are listed in Supplemental Table S3. Page 7/19 Page 7/19 RT-qPCR assays were performed according to published methods [46]. The apple EF-α gene was used to
normalise expression. Three biological replicates and three technical replicates were performed for each
sample. The relative expression of the analysed genes was calculated by the 2−ΔΔCt method [47]. RT-qPCR assays were performed according to published methods [46]. The apple EF-α gene was used to
normalise expression. Three biological replicates and three technical replicates were performed for each
sample. The relative expression of the analysed genes was calculated by the 2−ΔΔCt method [47]. RT-qPCR assays were performed according to published methods [46]. The apple EF-α gene was used to
normalise expression. Three biological replicates and three technical replicates were performed for each
sample. The relative expression of the analysed genes was calculated by the 2−ΔΔCt method [47]. RT-qPCR assays were performed according to published methods [46]. The apple EF-α gene was used to
normalise expression. Three biological replicates and three technical replicates were performed for each
sample. The relative expression of the analysed genes was calculated by the 2−ΔΔCt method [47]. Abbreviations MT, Melatonin; AR, adventitious root; MP, Malus prunifolia; IAA, auxin; CK, cytokinin; GA, gibberellins; ABA,
abscisic acid; WOX11, WUSCHEL-RELATED HOMEOBOX GENE 11; IBA, indole-3-butyric acid; ELISA,
enzyme-linked immunosorbent assay; LBD16, LATERAL ORGAN BOUNDARIES DOMAIN16; Acknowledgements Not applicable. Funding This work was financially supported by the National Key Research and Development Project
(2018YFD1000101, 2019YFD1001803), the Key Research and Development Project in the Shaanxi
province of China (2017ZDXM-NY-019, 2019TSLNY02-04), Tang Scholarship by the Cyrus Tang
Foundation and Northwest Agriculture and Forestry University, and the China Apple Research System
(CARS-27). The funder is the corresponding author of the article, who designed the research study. Consent for publication Not applicable. References 1. Naija S, Elloumi N, Jbir N, Ammar S, Kevers C. Anatomical and biochemical changes during
adventitious rooting of apple rootstocks MM 106 cultured in vitro. C R Biol. 2008;331(7):518-25. 2. Jásik J, Klerk GJD. Anatomical and ultrastructural examination of adventitious root formation in
stem slices of apple. Biol Plant. 1997;39(1):79-90. 3. Jiang W, Zhou S, Zhang Q, Song H, Zhou DX, Zhao Y. Transcriptional regulatory network of WOX11 is
involved in the control of crown root development, cytokinin signals, and redox in rice. J Exp Bot. 2017;68(11):2787-98. 3. Jiang W, Zhou S, Zhang Q, Song H, Zhou DX, Zhao Y. Transcriptional regulatory network of WOX11 is
involved in the control of crown root development, cytokinin signals, and redox in rice. J Exp Bot. 2017;68(11):2787-98. 4. Gutierrez L, Mongelard G, Floková K, Păcurar DI, Novák O, Staswick P, Kowalczyk M, Păcurar M,
Demailly H, Geiss G. Auxin controls Arabidopsis adventitious root initiation by regulating jasmonic
acid homeostasis. Plant cell. 2012; 24(6):2515-27. 4. Gutierrez L, Mongelard G, Floková K, Păcurar DI, Novák O, Staswick P, Kowalczyk M, Păcurar M,
Demailly H, Geiss G. Auxin controls Arabidopsis adventitious root initiation by regulating jasmonic
acid homeostasis. Plant cell. 2012; 24(6):2515-27. 5. Hu X, Xu L. Transcription factors WOX11/12 Directly Activate WOX5/7 to Promote Root Primordia
Initiation and Organogenesis. Plant Physiol. 2016; 172:2363-73. 5. Hu X, Xu L. Transcription factors WOX11/12 Directly Activate WOX5/7 to Promote Root Primordia
Initiation and Organogenesis. Plant Physiol. 2016; 172:2363-73. 6. Rigal A, Yordanov YS, Perrone I, Karlberg A, Tisserant E, Bellini C, Busov VB, Martin F, Kohler A,
Bhalerao R. The AINTEGUMENTA LIKE1 Homeotic Transcription Factor PtAIL1 Controls the
Formation of Adventitious Root Primordia in Poplar. Plant Physiol. 2012; 160(4):1996. 6. Rigal A, Yordanov YS, Perrone I, Karlberg A, Tisserant E, Bellini C, Busov VB, Martin F, Kohler A,
Bhalerao R. The AINTEGUMENTA LIKE1 Homeotic Transcription Factor PtAIL1 Controls the
Formation of Adventitious Root Primordia in Poplar. Plant Physiol. 2012; 160(4):1996. 7. Dubbels R, Reiter RJ, Klenke E, Goebel A, Schnakenberg E, Ehlers C, Schiwara HW, Schloot W. Melatonin in edible plants identified by radioimmunoassay and by high performance liquid
chromatography-mass spectrometry. J Pineal Res. 1995; 18(1):28-31. 7. Dubbels R, Reiter RJ, Klenke E, Goebel A, Schnakenberg E, Ehlers C, Schiwara HW, Schloot W. Melatonin in edible plants identified by radioimmunoassay and by high performance liquid
chromatography-mass spectrometry. J Pineal Res. 1995; 18(1):28-31. 8. Competing interests All authors declare that they have no competing interests. All authors declare that they have no competing interests. Ethics approval and consent to participate Not applicable. Page 8/19 References Hattori A, Migitaka H, Iigo M, Itoh M, Yamamoto K, Ohtanikaneko R, Hara M, Suzuki T, Reiter RJ. Identification of melatonin in plants and its effects on plasma melatonin levels and binding to
melatonin receptors in vertebrates. Biochem Mol Biol Int. 1995; 35(3):627-34. 8. Hattori A, Migitaka H, Iigo M, Itoh M, Yamamoto K, Ohtanikaneko R, Hara M, Suzuki T, Reiter RJ. Identification of melatonin in plants and its effects on plasma melatonin levels and binding to
melatonin receptors in vertebrates. Biochem Mol Biol Int. 1995; 35(3):627-34. 9. Park WJ. Melatonin as an Endogenous Plant Regulatory Signal: Debates and Perspectives. J Plant
Biol. 2011; 54(3):143-49. 9. Park WJ. Melatonin as an Endogenous Plant Regulatory Signal: Debates and Perspectives. J Plant
Biol. 2011; 54(3):143-49. 10. Sarropoulou V, Dimassi-Theriou K, Therios I, Koukourikou-Petridou M. Melatonin enhances root
regeneration, photosynthetic pigments, biomass, total carbohydrates and proline content in the
cherry rootstock PHL-C (Prunus avium x Prunus cerasus). Plant Physiol Biochem. 2012; 61:162-68. 11. Li C, Liang B, Chang C, Wei Z, Zhou S, Ma F. Exogenous melatonin improved potassium content in
Malus under different stress conditions. J Pineal Res. 2016; 61(2):218-29. Page 9/19 Page 9/19 12. Arnao MB, Hernández-Ruiz J. Melatonin promotes adventitious- and lateral root regeneration in
etiolated hypocotyls of Lupinus albus L. J Pineal Res. 2007; 42(2):147. 13. Yeong Byeon, Kyoungwhan Back. An increase in melatonin in transgenic rice causes pleiotropic
phenotypes, including enhanced seedling growth, delayed flowering, and low grain yield. J Pineal
Res. 2014; 56:408-14. 14. Małgorzata M, Posmyk, Krystyna M, Janas. Melatonin in plants. Acta Physiol Plant. 2009; 31:1-11. 15. Sangkyu Park, Back K. Melatonin promotes seminal root elongation and root growth in transgenic
rice after germination. J Pineal Res. 2012; 53:385-89. 16. Wen D, Gong B, Sun S, Liu S, Wang X, Wei M, Yang F, Li Y, Shi Q. Promoting Roles of Melatonin in
Adventitious Root Development of Solanum lycopersicum L. by Regulating Auxin and Nitric Oxide
Signaling. Front Plant Sci. 2016; 7:718. 17. Liang C, Li A, Yu H, Li W, Liang C, Guo S, Zhang R, Chu C. Melatonin Regulates Root Architecture by
Modulating Auxin Response in Rice. Front Plant Sci. 2017; 8:134. 18. Atkinson JA, Rasmussen A, Traini R, Voss U, Sturrock C, Mooney SJ, Wells DM, Bennett MJ. Branching out in roots: uncovering form, function, and regulation. Plant Physiol. 2014; 166(2):538-
50. 19. Klerk GJD, Arnholdt-Schmitt B, Lieberei R, Neumann KH. References Regeneration of roots, shoots and embryos:
physiological, biochemical and molecular aspects. Biol Plant. 1997; 39(1):53-66. 20. Legue V, Rigal A, Bhalerao RP. Adventitious root formation in tree species: involvement of
transcription factors. Physiol Plant. 2014; 151(2):192-98. 21. Klerk G-J: Rooting of microcuttings. Theory and practice. In Vitro Cell Dev Biol Anim. 2002; 38(5):415-
22. 22. Arnao MB, Hernandez-Ruiz J. Melatonin and its relationship to plant hormones. Ann bot. 2018;
121(2):195-07. 23. Jing Z, Yi S, Zhang X, Du H, Xu B, Huang B. Melatonin suppression of heat-induced leaf senescence
involves changes in abscisic acid and cytokinin biosynthesis and signaling pathways in perennial
ryegrass ( Lolium perenne L.). Environ Exp Bot. 2017; 138:36-45. 24. Hai-Jun Z, Na Z, Rong-Chao Y, Li W, Qian-Qian S, Dian-Bo L, Yun-Yun C, Sarah W, Bing Z, Shuxin R. Melatonin promotes seed germination under high salinity by regulating antioxidant systems, ABA
and GA? interaction in cucumber (Cucumis sativus L.). J Pineal Res. 2015; 57(3):269-79. 25. Chen Q, Qi WB, Russel J, Wei W, Wang BM. Exogenously applied melatonin stimulates root growth
and raises endogenous indoleacetic acid in roots of etiolated seedlings of Brassica juncea. J Plant
Physiol. 2009; 166(3):324-28. 26. Wang Q, An B, Wei Y, Reiter RJ, Shi H, Luo H, He C. Melatonin Regulates Root Meristem by Repressing
Auxin Synthesis and Polar Auxin Transport in Arabidopsis. Front Plant Sci. 2016; 7:1882. 27. Arnao MB, Hernández-Ruiz J. Growth activity, rooting capacity, and tropism: three auxinic precepts
fulfilled by melatonin. Acta Physiol Plant. 2017; 39(6):1-7. Page 10/19 28. Marino B. Arnao, Hernández-Ruiz J. The Physiological Function of Melatonin in Plants. Plant Signal
Behav. 2006; 1(3):89-95. 29. Ramo´ n Pelagio-Flores, Edith Mun˜ oz-Parra, Randy Ortiz-Castro, pez-Bucio. JL. Melatonin regulates
Arabidopsis root system architecture likely acting independently of auxin signaling. J Pineal Res. 2012; 53:279-88. 30. Qiuxiang Ma, Ting Zhang, Peng Zhang, Wang Z-Y. Melatonin attenuates postharvest physiological
deterioration of cassava storage roots. J Pineal Res. 2016; 60:424-34. 31. Sangkyu P, Da-Eun L, Hyunki J, Yeong B, Young-Soon K, Kyoungwhan B. Melatonin-rich transgenic
rice plants exhibit resistance to herbicide-induced oxidative stress. J Pineal Res. 2013; 54(3):258-63. 32. Yeong B, Hyoung Yool L, Ok Jin H, Hye-Jung L, Kyungjin L, Kyoungwhan B. Coordinated regulation of
melatonin synthesis and degradation genes in rice leaves in response to cadmium treatment. J
Pineal Res. 2015; 58(4):470-78. 33. References Wang L, Feng C, Zheng X, Guo Y, Zhou F, Shan D, Liu X, Kong J. Plant mitochondria synthesize
melatonin and enhance the tolerance of plants to drought stress. J Pineal Res. 2017; 63(3). 34. Bixiao Zuo, Xiaodong Zheng, Pingli He, Lin Wang, Qiong Lei, Chao Feng1, Jingzhe Zhou, Qingtian Li,
Zhenhai Han, Jin Kong. Overexpression of MzASMT improves melatonin production and enhances
drought tolerance in transgenic Arabidopsis thaliana plants. J Pineal Res. 2014; 57:408-17. 35. Zhao Y, Hu Y, Dai M, Huang L, Zhou DX. The WUSCHEL-related homeobox gene WOX11 is required to
activate shoot-borne crown root development in rice. Plant Cell. 2009; 21(3):736-48. 36. Liu J, Sheng L, Xu Y, Li J, Yang Z, Huang H, Xu L. WOX11 and 12 are involved in the first-step cell fate
transition during de novo root organogenesis in Arabidopsis. Plant Cell. 2014; 26(3):1081-93. 37. Laskowski M. Lateral root initiation is a probabilistic event whose frequency is set by fluctuating
levels of auxin response. J Exp Bot. 2013; 64(9):2609-17. 37. Laskowski M. Lateral root initiation is a probabilistic event whose frequency is set by fluctuating
levels of auxin response. J Exp Bot. 2013; 64(9):2609-17. 38. Mao J-P, Zhang D, Zhang X, Li K, Liu Z, Meng Y, Lei C, Han M-Y. Effect of exogenous indole-3-butanoic
acid (IBA) application on the morphology, hormone status, and gene expression of developing lateral
roots in Malus hupehensis. Sci Hortic. 2018; 232:112-20. 38. Mao J-P, Zhang D, Zhang X, Li K, Liu Z, Meng Y, Lei C, Han M-Y. Effect of exogenous indole-3-butanoic
acid (IBA) application on the morphology, hormone status, and gene expression of developing lateral
roots in Malus hupehensis. Sci Hortic. 2018; 232:112-20. 39. Himanen K. Auxin-Mediated Cell Cycle Activation during Early Lateral Root Initiation. Plant Cell. 2002;
14(10):2339-51. 39. Himanen K. Auxin-Mediated Cell Cycle Activation during Early Lateral Root Initiation. Plant Cell. 2002;
14(10):2339-51. 40. Liu B, Wang L, Zhang J, Li J, Zheng H, Chen J, Lu M. WUSCHEL-related Homeobox genes in Populus
tomentosa: diversified expression patterns and a functional similarity in adventitious root formation. BMC Genom. 2014; 15:296. 40. Liu B, Wang L, Zhang J, Li J, Zheng H, Chen J, Lu M. WUSCHEL-related Homeobox genes in Populus
tomentosa: diversified expression patterns and a functional similarity in adventitious root formation. BMC Genom. 2014; 15:296. 41. Qiyun XU, Chai F, Xincheng AN, Han S. 44. Dobrev PI, Kamínek M. Fast and efficient separation of cytokinins from auxin and abscisic acid and
their purification using mixed-mode solid-phase extraction. J Chromatography A. 2002; 950(1-2):21-
29. 47. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR
and the 2(-Delta Delta C(T)) Method. Methods. 2001; 25(4):402-08. References Production Method for Paraffin Section of Invasive Species of
Bemisia tabaci. Plant Dis Pests. 2012; (2):46-48. 41. Qiyun XU, Chai F, Xincheng AN, Han S. Production Method for Paraffin Section of Invasive Species of
Bemisia tabaci. Plant Dis Pests. 2012; (2):46-48. 42. Yang JP. Improvement of traditional paraffin section preparation methods. Journal of Biology. 2006. 42. Yang JP. Improvement of traditional paraffin section preparation methods. Journal of Biology. 2006. 43. Mao J, Zhang D, Meng Y, Li K, Wang H, Han M. Inhibition of adventitious root development in apple
rootstocks by cytokinin is based on its suppression of adventitious root primordia formation. Physiol
Plant. 2018; 166:663-76. 43. Mao J, Zhang D, Meng Y, Li K, Wang H, Han M. Inhibition of adventitious root development in apple
rootstocks by cytokinin is based on its suppression of adventitious root primordia formation. Physiol
Plant. 2018; 166:663-76. Page 11/19 Page 11/19 44. Dobrev PI, Kamínek M. Fast and efficient separation of cytokinins from auxin and abscisic acid and
their purification using mixed-mode solid-phase extraction. J Chromatography A. 2002; 950(1-2):21-
29. 45. Gambino G, Perrone I, Gribaudo I. A Rapid and effective method for RNA extraction from different
tissues of grapevine and other woody plants. Phytochem Anal. 2008; 19(6):520-25. 45. Gambino G, Perrone I, Gribaudo I. A Rapid and effective method for RNA extraction from different
tissues of grapevine and other woody plants. Phytochem Anal. 2008; 19(6):520-25. 46. Li G, Ma J, Tan M, Mao J, An N, Sha G, Zhang D, Zhao C, Han M. Transcriptome analysis reveals the
effects of sugar metabolism and auxin and cytokinin signaling pathways on root growth and
development of grafted apple. BMC Genom. 2016; 17:150. 46. Li G, Ma J, Tan M, Mao J, An N, Sha G, Zhang D, Zhao C, Han M. Transcriptome analysis reveals the
effects of sugar metabolism and auxin and cytokinin signaling pathways on root growth and
development of grafted apple. BMC Genom. 2016; 17:150. 47. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR
and the 2(-Delta Delta C(T)) Method. Methods. 2001; 25(4):402-08. 45. Gambino G, Perrone I, Gribaudo I. A Rapid and effective method for RNA extraction from different
tissues of grapevine and other woody plants. Phytochem Anal. 2008; 19(6):520-25. Figures Page 12/19 Figure 1
Anatomical and morphological observations of AR development in paraffin-embedded sections from the
experimental samples generated in the study at three sampling times: 0 d, 3 d, and 10 d. Five treatment
groups were analysed: (A) The control group in which apple tissue culture plantlets were continuously
cultured on root-inducing medium containing 3.45 μM IBA; the MT group was continuously cultivated on
3.45 μM IBA and 1.29 μM MT, and based on different times of MT-treatment, the MT treatment group was
also divided into three groups of MT0–2, MT2–5 and MT5–20; (B) Observations of morphological AR
formation in the five treatment groups at 0 d, 5 d, 10 d and 20 d; (C) Anatomical observations of AR
formation in the five treatment groups at 0 d, 5 d and 10 d. Figure 1 Anatomical and morphological observations of AR development in paraffin-embedded sections from the
experimental samples generated in the study at three sampling times: 0 d, 3 d, and 10 d. Five treatment
groups were analysed: (A) The control group in which apple tissue culture plantlets were continuously
cultured on root-inducing medium containing 3.45 μM IBA; the MT group was continuously cultivated on
3.45 μM IBA and 1.29 μM MT, and based on different times of MT-treatment, the MT treatment group was
also divided into three groups of MT0–2, MT2–5 and MT5–20; (B) Observations of morphological AR
formation in the five treatment groups at 0 d, 5 d, 10 d and 20 d; (C) Anatomical observations of AR
formation in the five treatment groups at 0 d, 5 d and 10 d. Anatomical and morphological observations of AR development in paraffin-embedded sections from the
experimental samples generated in the study at three sampling times: 0 d, 3 d, and 10 d. Five treatment
groups were analysed: (A) The control group in which apple tissue culture plantlets were continuously
cultured on root-inducing medium containing 3.45 μM IBA; the MT group was continuously cultivated on
3.45 μM IBA and 1.29 μM MT, and based on different times of MT-treatment, the MT treatment group was
also divided into three groups of MT0–2, MT2–5 and MT5–20; (B) Observations of morphological AR
formation in the five treatment groups at 0 d, 5 d, 10 d and 20 d; (C) Anatomical observations of AR
formation in the five treatment groups at 0 d, 5 d and 10 d. Page 13/19 Page 13/19 gure 2 Figure 2
Morphological parameters of AR formation in tissue culture plantlets of Malus prunifolia. Tissue culture
plantlets were divided into five groups: Control, MT, MT0–2, MT2–5, MT5–20. The number of ARs, and
their length, surface area and volume were measured in the five treatment groups. Values represent the
mean ± SE for three biological replicates; letters indicate significant differences between means (P <
0.05). Figure 2 Figure 2 Morphological parameters of AR formation in tissue culture plantlets of Malus prunifolia. Tissue culture
plantlets were divided into five groups: Control, MT, MT0–2, MT2–5, MT5–20. The number of ARs, and
their length, surface area and volume were measured in the five treatment groups. Values represent the
mean ± SE for three biological replicates; letters indicate significant differences between means (P <
0.05). Page 14/19 Page 14/19 Figure 3
Morphological parameters of AR formation in tissue culture plantlets of Malus prunifolia apple rootstoc
for different categories of root diameter, according to the number, length, surface area, and volume of
root. The ARs were classified into three groups based on root diameter: 0–2.0 mm, 2.0–5.0 mm and >5.0
mm. Figure 3 Morphological parameters of AR formation in tissue culture plantlets of Malus prunifolia apple rootstock
for different categories of root diameter, according to the number, length, surface area, and volume of
root. The ARs were classified into three groups based on root diameter: 0–2.0 mm, 2.0–5.0 mm and >5.0
mm. Page 15/19 Figure 4 Figure 4 The effect of exogenous application of MT for different lengths of time on the contents of MT, ZR, IAA,
GA1+3, and ABA at different stages of AR development in five treatment groups of Malus prunifolia apple
rootstocks. Page 16/19
Figure 5 Figure 5 Page 16/19 Page 16/19 The effect of exogenous MT treatment on the relative expression of genes involved in MT signal
transduction at different stages of AR development in five treatment groups of Malus prunifolia. transduction at different stages of AR development in five treatment groups of Malus prunifolia. Figure 6
The effect of exogenous MT treatment on the relative expression of genes involved in auxin biosynthes
transport, and signal transduction at different stages of AR development in five treatment groups of
Malus prunifolia. Figure 6 The effect of exogenous MT treatment on the relative expression of genes involved in auxin biosynthesis,
transport, and signal transduction at different stages of AR development in five treatment groups of
Malus prunifolia. Page 17/19 Page 17/19 Figure 7
The effect of exogenous MT treatment on the relative expression of genes involved in the cell recycle and
AR development at different stages of AR formation in five treatment groups of Malus prunifolia. Figure 8 (A) Morphological observations, (B) AR rate and (C) AR number during AR formation in transgenic tissue-
culture plantlets overexpressing MdWOX11 (35S::WOX11-OE) and wild type ‘GL3’ after treatment with
3.45 μM IBA as the control. Another group was treated with MT for 0–2 d (MT0–2 group), the results
represent measurements after culturing the controls and MT0–2 group for 20 d. (A) Morphological observations, (B) AR rate and (C) AR number during AR formation in transgenic tissue-
culture plantlets overexpressing MdWOX11 (35S::WOX11-OE) and wild type ‘GL3’ after treatment with
3.45 μM IBA as the control. Another group was treated with MT for 0–2 d (MT0–2 group), the results
represent measurements after culturing the controls and MT0–2 group for 20 d. (A) Morphological observations, (B) AR rate and (C) AR number during AR formation in transgenic tissue-
culture plantlets overexpressing MdWOX11 (35S::WOX11-OE) and wild type ‘GL3’ after treatment with
3.45 μM IBA as the control. Another group was treated with MT for 0–2 d (MT0–2 group), the results
represent measurements after culturing the controls and MT0–2 group for 20 d. Figure 7 Figure 7 The effect of exogenous MT treatment on the relative expression of genes involved in the cell recycle and
AR development at different stages of AR formation in five treatment groups of Malus prunifolia. Page 18/19 igure 8
A) Morphological observations, (B) AR rate and (C) AR number during AR formation in transgenic tissue-
ulture plantlets overexpressing MdWOX11 (35S::WOX11-OE) and wild type ‘GL3’ after treatment with Figure 8
(A) Morphological observations, (B) AR rate and (C) AR number during AR formation in transgenic tissue-
culture plantlets overexpressing MdWOX11 (35S::WOX11-OE) and wild type ‘GL3’ after treatment with
3 45 μM IBA as the control Another group was treated with MT for 0 2 d (MT0 2 group) the results Supplementary Files This is a list of supplementary files associated with this preprint. Click to download. Supplementalmaterial.docx Page 19/19 Page 19/19
|
https://openalex.org/W2007776162
|
http://www.mdpi.com/1420-3049/18/7/7886/pdf/
|
English
| null |
Triterpenoids of Marine Origin as Anti-Cancer Agents
|
Molecules/Molecules online/Molecules annual
| 2,013
|
cc-by
| 7,895
|
Yong-Xin Li 1, S. W. A. Himaya 2 and Se-Kwon Kim 1,2,* Yong-Xin Li 1, S. W. A. Himaya 2 and Se-Kwon Kim 1,2,* 1 Marine Bioprocess Research Center, Pukyong National University, Busan 608-737, Korea;
E-Mail: lyxycg@hotmail.com
2 2 Department of Chemistry, Pukyong National University, Busan 608-737, Korea;
E-Mail: himayaswa@yahoo.com 2 Department of Chemistry, Pukyong National University, Busan 608-737, Korea;
E-Mail: himayaswa@yahoo.com * Author to whom correspondence should be addressed; E-Mail: sknkim@pknu.ac.kr;
Tel.: +82-51-629-7097; Fax: +82-51-629-7099. * Author to whom correspondence should be addressed; E-Mail: sknkim@pknu.ac.kr;
Tel.: +82-51-629-7097; Fax: +82-51-629-7099. Received: 16 February 2013; in revised form: 10 June 2013 / Accepted: 27 June /
Published: 4 July 2013 Received: 16 February 2013; in revised form: 10 June 2013 / Accepted: 27 June /
Published: 4 July 2013 Received: 16 February 2013; in revised form: 10 June 2013 / Accepted: 27 June /
Published: 4 July 2013 Abstract: Triterpenoids are the most abundant secondary metabolites present in marine
organisms, such as marine sponges, sea cucumbers, marine algae and marine-derived
fungi. A large number of triterpenoids are known to exhibit cytotoxicity against a variety
of tumor cells, as well as anticancer efficacy in preclinical animal models. In this review
efforts have been taken to review the structural features and the potential use of
triterpenoids of marine origin to be used in the pharmaceutical industry as potential
anti-cancer drug leads. Keywords: triterpenoids; anti-cancer; cytotoxic; pharmacological agents; marine origin Molecules 2013, 18, 7886-7909; doi:10.3390/molecules18077886
molecules
ISSN 1420-3049
www.mdpi.com/journal/molecules
Review
Triterpenoids of Marine Origin as Anti-Cancer Agents
Yong-Xin Li 1, S. W. A. Himaya 2 and Se-Kwon Kim 1,2,*
1 Marine Bioprocess Research Center, Pukyong National University, Busan 608-737, Korea;
E-Mail: lyxycg@hotmail.com
2 Department of Chemistry, Pukyong National University, Busan 608-737, Korea;
E-Mail: himayaswa@yahoo.com
* Author to whom correspondence should be addressed; E-Mail: sknkim@pknu.ac.kr;
Tel.: +82-51-629-7097; Fax: +82-51-629-7099.
Received: 16 February 2013; in revised form: 10 June 2013 / Accepted: 27 June /
Published: 4 July 2013
OPEN ACCESS Molecules 2013, 18, 7886-7909; doi:10.3390/molecules18077886 Review 1. Introduction Recently, the search for novel bioactive compounds as anti-cancer agents from marine resources
has gained much attention. Triterpenoids are terpenoid derivatives of natural products containing about
thirty carbon atoms, and their structures are considered to be derived from the acyclic precursor
squalene [1,2]. Triterpenoids are the most abundant secondary metabolite present in marine sources,
such as marine sponges [3,4], sea cucumbers [5], marine algae [6], and marine-derived fungi [7]. During a last few years, a great number of biologically active triterpenoids was found to have
cytotoxicity against a variety of tumor cells [8,9]. More than 20,000 triterpenoids has been isolated and
identified from Nature, which belong to different chemical groups such as squalene, lanostane, Molecules 2013, 18 7887 dammarane, lupane, oleanane, ursane, hopane, etc. [10,11]. This review summarizes the anti-cancer
triterpenoids isolated from marine sponges, sea cucumbers, marine algae, and marine fungi that
includes isomalabaricane-type triterpenoids (stellettins, stelliferins, and geoditins), polyether triterpenes
(sodwanones, raspacionins, sipholenols, sipholenones, and siphonellinols), triterpenoid glycosides
(saponins), and tetracyclic triterpenoids and their potential anti-cancer activity. Therefore, this review
brings insights to marine triterpenoids as potential candidates to be developed as pharmaceuticals
against tumor progression. An ideal anticancer agent is expected to inhibit, delay or reverse the progression of cancer through
its cytotoxicity or apoptosis-inducing properties [12]. The discovery and development of anticancer
drugs, especially cytotoxic agents, differ significantly from the drug development process for any other
indications. Identification of cytotoxic compounds led the development of anticancer therapeutics for
several decades. Cytotoxic drugs are primarily used as anticancer drugs because they are toxic to
cancer cells. These drugs have been associated with human cancers at high (therapeutic) levels of
exposure and are carcinogens and teratogens in many animal species. Cytotoxic drugs have an effect of
preventing the rapid growth and division (mitosis) of cancer cells [13]. During a last few years, great
numbers of biologically active triterpenoids are found to have cytotoxicity against a variety of tumor
cells. Triterpenoids are highly multifunctional and the antitumor activity of these compounds is
measured by their ability to block nuclear factor-kappaB activation, induce apoptosis, inhibit signal
transducer, and activate transcription and angiogenesis [14]. Advances in cancer treatment, however,
continued to be challenged by the identification of unique biochemical aspects of malignancies that
could be exploited to selectively target tumor cells. However, selective elimination of tumor cells
using cytotoxic agents is universally applicable approach of cancer treatment. 1. Introduction This review will
highlight the enormous potential of triterpenoids identified from marine resources as cytotoxic agents
against tumor progression. 2. Triterpenoids from Marine Sponges Isomalabaricane-type triterpenoids are a rare group of triterpenoids with unique skeletons, often
found in marine sponges. The cytotoxic isomalabaricane-type triterpenoids stellettins A-K (1–13,
Figure 1) have been reported from the marine sponge species of the genus Jaspis [15], Stelleta [16–18],
and Rhabdastrella [19]. Stellettin A (1) and B (2), were isolated from the sponge Stelletta tmuis
collected from Hainan Island, China in 1994. Stellettin A was significantly toxic to P388 leukemia
cells, exhibiting an ED50 value of 0.001 μg/mL [20]. Furthermore, Liu et al. have demonstrated that
stellettin A and stellettin B induce cytotoxicity in HL-60 cells treated for 24 h at 3 μM concentration [21]. The cytotoxic isomalabaricane triterpenoids stellettins A-G (1–7) have been examined at the National
Cancer Institute (Australia) against 60 cell lines. Stelletin C (3) and D (4) were the most potent
derivatives with a mean panel GI50 of 0.09 µM. The stelletin E (5) and F (6) pair was approximately
10-times less potent (mean GI50 of 0.98 µM) [16,22]. The isomalabaricane triterpenes, stellettins A-D (1–4), stellettin H (8), stellettin I (9) along with
rhabdastrellic acid-A (14), have been isolated from the marine sponge Rhabdastrella globostellata,
collected from the Philippines. These compounds have shown selective cytotoxicity towards
p21WAF1/Cip1-deficient human colon tumor (HCT-116) cells [23]. 7888 Molecules 2013, 18 ,
Figure 1. Isomalabaricane-type triterpenoids stellettins from marine sponge. R2
O
O
1 R1, R2=O
13(14)=Z
2 R1, R2=O
13(14)=E
3 R1, R2=OCOCH3,
13(14)=Z
4 R1, R2=OCOCH3,
13(14)=E
O
O
R1
O
COOH
5
13(14)=Z,
24(25)=E
6
13(14)=E,
24(25)=Z
7
13(14)=Z,
24(25)=Z
14 13(14)=E,
24(25)=E
13 14
13 14
24
25
△
△
△
△
△
△
△
△
△
△
△
△
O
9
O
O
COOH
O
8
O
O
COOH
HO
O
HO
HO
O
HOOC
10
11
13
O
HO
COOH
12
O
HO
COOH
he cytotoxic isomalabaricane triterpenoids stelletin J (10) and K (11) from Rhabdastrella
ostellata have shown activity in an assay measuring stabilization of the binding of DNA with igure 1. Isomalabaricane-type triterpenoids stellettins from marine sponge. 2. Triterpenoids from Marine Sponges O
O
COOH
5
13(14)=Z,
24(25)=E
6
13(14)=E,
24(25)=Z
7
13(14)=Z,
24(25)=Z
14 13(14)=E,
24(25)=E
13 14
24
25
△
△
△
△
△
△
△
△ 5
13(14)=Z,
24(25)=E
6
13(14)=E,
24(25)=Z
7
13(14)=Z,
24(25)=Z
14 13(14)=E,
24(25)=E
△
△
△
△
△
△
△
△ 1 R1, R2=O
13(14)=Z
2 R1, R2=O
13(14)=E
3 R1, R2=OCOCH3,
13(14)=Z
4 R1, R2=OCOCH3,
13(14)=E
△
△
△
△ 2 R1, R2 O
E
3 R1, R2=OCOCH3,
13(14)=Z
4 R1, R2=OCOCH3,
13(14)=E
7
13(14)=Z,
24(25)=Z
14 13(14)=E,
24(25)=E
△
△
△
△
△
△
O
9
O
O
COOH
O
8
O
O
COOH
HO
O
HO
HO
O
HOOC
10
11 8
COOH
HO
O
HO
10 9
HO
O
HOOC
11 8 11 10 O
10
12
O
HO
COOH 13
O
HO
COOH 13 12 The cytotoxic isomalabaricane triterpenoids stelletin J (10) and K (11) from Rhabdastrella
globostellata have shown activity in an assay measuring stabilization of the binding of DNA with
DNA polymerase β. However, stelletin J (10) and K (11) displayed varying levels of activity toward
the A2780 ovarian cancer cell line, revealing structure-based effects on both the level of cytotoxicity
and DNA-polymerase β binding [24]. Stelletin L (12) and M (13) were isolated from the marine sponge Stelletta tenuis collected in the
South China Sea and both compounds exhibited significant cytotoxic activity against stomach cancer
cells (AGS) in vitro [18]. Stelliferins A–F (15–20, Figure 2), antineoplastic isomalabaricane triterpenes were isolated from
the Okinawan marine sponge Jaspis stellifera [25]. The isomalabaricane triterpenes, stelliferin G (21),
29-hydroxystelliferin A (22), 29-hydroxystelliferin E (23) together with the known triterpenes
3-epi-29-hydroxystelliferin E (24), 13(E)-29-hydroxystelliferin E (25), 29-hydroxystelliferin B (26), 7889 Molecules 2013, 18 Molecules 2013, 18 13(E)-stelliferin G (27), and 13(E)-3-epi-29-hydroxystelliferin E (28), were isolated from the organic
extract of the sponge Jaspis sp. collected in the South Pacific ocean. All compounds were tested
against melanoma (MALME-3M) and leukemia (MOLT-4) cells. The mixtures of 29-hydroxystelliferin
B (26) and 13E-stelliferin G (27) have shown highest growth-inhibitory [(IC50) 0.11, 0.23, μg/mL,
respectively)] activities against MALME-3M [26]. Figure 2. Triterpenoid stelliferins from marine sponges. Figure 2. Triterpenoid stelliferins from marine sponges. Figure 2. Triterpenoid stelli Moreover, the isomalabaricane triterpenoids stelliferin riboside (29) and 3-epi-29-acetoxystelliferin E
(30) were isolated from an extract of the sponge Rhabdastrella globostellata which was active in an
assay measuring stabilization of the binding of DNA with DNA polymerase β. Both compounds have
shown to induce 29% and 23% binding, respectively [24]. 7890 Molecules 2013, 18 Four isomalabaricane triterpenes, geoditin A (31), geoditin B (32), isogeoditin A (33), and
isogeoditin B (34) were isolated from marine sponge Rhabdastrella aff. distincta. All compounds were
tested against a small panel of human tumor cell lines [19]. Geoditin A (31) and geoditin B (32) have
also been isolated from marine sponge Geodia japonica. Geoditin A was the most cytotoxic to HL60
cells [IC 50 Z3 mg/mL (<6.6 mM)], and geoditin B exhibited relatively weak cytotoxicity [27]. Five cytotoxic triterpene glycosides, erylosides F1-F4 (35–38), and erylosides F (39) (Figure 3)
were isolated from the sponge Erylus formosus collected from the Mexican Gulf (Puerto Morelos,
Mexico). Four compounds induced the early apoptosis of Ehrlich carcinoma cells, where erylosides F3
have shown the highest activity at a concentration of 100 μg/mL [28]. Figure 3. Triterpenoid geoditins from marine sponges. Figure 3. Triterpenoid geoditins from marine sponges. Figure 3. Triterpenoid geoditins from marine sponges. β
The special group of triterpenoids named sodwanones: sodwanones A-I (40–48) and sodwanones
K-W (49–61), have been isolated from the Indo-Pacific sponge Axinella wltneri [29]. Sodwanones G
(46), H (47), and I (48) have been found to have cytotoxic activity. The compounds have shown
cytotoxicity activity against cell cultures of P-388 murine leukemia, A-549 human lung carcinoma,
HT-29 human colon carcinoma, and MEL-28 human melanoma. Sodwanones G (46), H (47), I (48)
showed high specificity towards human lung carcinoma cell line A-549, where the specificity of
sodwanone G was prominent (46) [30]. The cytotoxic triterpenes sodwanones K (49), L (50), and M
(51) were found to be cytotoxic to P-388 murine leukemia cells [31]. The biological activity of
sodwanone S (57) was evaluated against 13 human tumor cell lines [32]. Sodwanone V (60) inhibited
both hypoxia-induced and iron chelator (1,10-phenanthroline)-induced HIF-1 activation in T47D
breast tumor cells (IC50 15 μM), and sodwanone V (60) was the only sodwanone that inhibited HIF-1
activation in PC-3 prostate tumor cells (IC50 15 μM). β β oup of triterpenoids named sodwanones: sodwanones A-I (40–48) and sodwanones The special group of triterpenoids named sodwanones: sodwanones A-I (40–48) and sodwanones
K-W (49–61), have been isolated from the Indo-Pacific sponge Axinella wltneri [29]. Sodwanones G
(46), H (47), and I (48) have been found to have cytotoxic activity. The compounds have shown
cytotoxicity activity against cell cultures of P-388 murine leukemia, A-549 human lung carcinoma,
HT-29 human colon carcinoma, and MEL-28 human melanoma. Sodwanones G (46), H (47), I (48)
showed high specificity towards human lung carcinoma cell line A-549, where the specificity of
sodwanone G was prominent (46) [30]. The cytotoxic triterpenes sodwanones K (49), L (50), and M
(51) were found to be cytotoxic to P-388 murine leukemia cells [31]. The biological activity of
sodwanone S (57) was evaluated against 13 human tumor cell lines [32]. Sodwanone V (60) inhibited
both hypoxia-induced and iron chelator (1,10-phenanthroline)-induced HIF-1 activation in T47D
breast tumor cells (IC50 15 μM), and sodwanone V (60) was the only sodwanone that inhibited HIF-1
activation in PC-3 prostate tumor cells (IC50 15 μM). Molecules 2013, 18 7891 Sodwanone A (40) and sodwanone T (58) inhibited hypoxia-induced HIF-1 activation in T47D cells
(IC50 values 20–25 μM), and sodwanone V (60) showed cytotoxicity to MDA-MB-231 breast tumor
cells (IC50 23 μM). Sodwanone derived compounds, 3-epi-sodwanone K (62), 3-epi-sodwanone K
3-acetate (63), 10,11-dihydrosodwanone B (64) have been isolated from Axinella sp., and 62 and 64
also inhibited hypoxia-induced HIF-1 activation in T47D cells (IC50 values 20–25 μM) and 63 was
cytotoxic to T47D cells (IC50 22 μM) [33] (Figure 4). Figure 4. Triterpenoid sodwanones from marine sponges. Figure 4. Triterpenoid sodwanones from marine sponges. Molecules 2013, 18 7892 Raspacionin triterpinoids 65–83 (Figure 5), namely raspacionin (65), raspacionins A (66),
raspacionins B (67), 21-deacetylraspacionin (68), 10-acetoxy-21-deacetyl-28-hydroraspacionin (69),
10-acetoxy-21-deacetyl-4-oxo-28-hydroraspacionin
(70),
10-acetoxy-15,21-dideacetyl-4-oxo-28-
hydroraspacionin (71), 10-acetoxy-15-deacetyl-4-oxo-28-hydroraspacionin (72), 10-acetoxy-4-acetyl-
15-deacetyl-28-hydroraspacionin (73), 10-acetoxy-15-deacetyl-4-21-dioxo-28-hydroraspacionin (74),
10-hydroxy-4,21-dioxo-28-hydroraspacionin (75), 21-oxoraspacionin (76), 15-deacetyl-21-dioxo-
raspacionin (77), 4,21-dioxo-raspacionin (78), 10-acetoxy-4,21-dioxo-28-hydroraspacionin (79),
10-acetoxy-4-acetyl-21-oxo-28-hydroraspacionin (80), 10-acetoxy-4-acetyl-28-hydroraspacionin (81),
10-acetoxy-28-hydroraspacionin (82) and 10-acetoxy-21-deacetyl-4-acetyl-28-hydroraspacionin (83),
have been isolated from red sponge, Raspaciona aculeuta Johnston (family Raspailiidae), and from
the Mediterranean sponge Raspaciona aculeata. All the compounds have showed cytotoxicity against
MCF-7 tumor cell line with IC50 values between 4 and 8 μM [34–36]. Figure 5. Triterpenoid raspacionins from marine sponges. Figure 5. Triterpenoid raspacionins from marine sponges. Molecules 2013, 18 7893 The Red Sea sponge Siphonochalina siphonella is a rich source of sipholane triterpenoids including
sipholenols (A, C-L) (84, 85–94), sipholenones (A, E) (95, 96), and siphonellinols (C, D, E) (97, 98,
99). Sipholenol A (84) and sipholenone A (sipholenol B) are the major sipholane triterpenoids [37]. Sipholenol A was found to have increased the sensitivity of resistant KB-C2 cells [38]. Sipholenol A
(84), sipholenol I (91), sipholenol L (94), sipholenone A (95), sipholenone E (96), siphonellinol C
(97), and siphonellinol D (98) have found to show potent reversal of multidrug resistance in cancer
cells that over expressed P-glycoprotein. These compounds enhanced the cytotoxicity of several
P-glycoprotein substrate anticancer drugs, and significantly reversed the multidrug resistance phenotype
in P-glycoprotein-overexpressing multidrug resistant cancer cells KB-C2 and KB-V1 in a dose-
dependent manner [39,40] (Figure 6). Figure 6. Triterpenoid sipholenols, sipholenones and siphonellinols from marine spon g
p
p
,
p
p
p
g 7894 Molecules 2013, 18 Figure 6. Cont. O
96
OH
OH
O
O
97
OH
OH
OH
OOH
O
98
OH
OH
OH
O
99
OH
OH
OH
HOO
3. Triterpenoids from Sea Cucumbers Figure 6. Cont. Figure 6. Cont. 97 99 3. Triterpenoids from Sea Cucumbers 3. Triterpenoids from Sea Cucumbers Triterpenoid glycosides (saponins) are the major and most abundant type of compounds isolated
from sea cucumbers. Saponins are generally perceived as highly active natural product and the sea
cucumber saponins have been well characterized for their anti-cancer activities. The cytotoxicity of
five triterpene glycosides, fuscocineroside A (100), B (101), and C (102), pervicoside C (103) and
holothurin A (104) isolated from Holothuria fuscocinerea Jaeger on human leukemia HL-60 and
human hepatoma BEL-7402 cells was analyzed and all compounds have shown a potent cytotoxicity
towards both cell lines. However, fuscocineroside C was found to be the most potent (IC50 = 0.88,
IC50 = 0.58 µg/mL) in HL-60 and BEL-7402 cell lines respectively [41] (Figure 7). Figure 7. Triterpenoid glycosides fuscocinerosides, pervicoside C and holothurin A from sea cucumbers. The triterpene glycosides from the sea cucumber Holothuria scabra, namely holothurin A3 (105)
and A4 (106) found to be strongly cytotoxic to cancer cell lines; human epidermoid carcinoma (KB)
and human hepatocellular carcinoma (Hep-G2), with IC50 values of 0.87 and 0.32 µg/mL (for
compound 7) and of 1.12 and 0.57 µg/mL (for compound 8), respectively [42] (Figure 8). The triterpene glycosides from the sea cucumber Holothuria scabra, namely holothurin A3 (105)
and A4 (106) found to be strongly cytotoxic to cancer cell lines; human epidermoid carcinoma (KB)
and human hepatocellular carcinoma (Hep-G2), with IC50 values of 0.87 and 0.32 µg/mL (for
compound 7) and of 1.12 and 0.57 µg/mL (for compound 8), respectively [42] (Figure 8). The triterpene glycosides from the sea cucumber Holothuria scabra, namely holothurin A3 (105)
and A4 (106) found to be strongly cytotoxic to cancer cell lines; human epidermoid carcinoma (KB)
and human hepatocellular carcinoma (Hep-G2), with IC50 values of 0.87 and 0.32 µg/mL (for
compound 7) and of 1.12 and 0.57 µg/mL (for compound 8), respectively [42] (Figure 8). 7895 Molecules 2013, 18 Figure 8. Triterpenoid glycosides holothurin A3 and A4 from sea cucumbers. Arguside A (107) also exhibited significant cytotoxicity against different human tumor cell lines
while showing the highest activity towards human colorectal carcinoma (HCT-116) cells (IC50 = 0.14 µM)
with more potency than the employed positive control, 10-hydroxycamptothecin (IC50 = 0.84 µM) [43]. Argusides B (108) and C (109) have also shown potent cytotoxicity against human tumor cell lines,
adenocarcinomic human alveolar basal epithelial cells (A549), HCT-116, HepG2, and human breast
adenocarcinoma (MCF-7) cell lines. 3. Triterpenoids from Sea Cucumbers The cytotoxicity of the compounds on A549 (108-IC50 =
0.48 µg/mL, 109-IC50 = 0.43 µg/mL) and HCT-116 (108-IC50 = 0.46 µg/mL, 109-IC50 = 0.38 µg/mL)
cells were more potent than the positive control V-16 (Figure 9). However, there was no significant
difference between the cytotoxicity of two compounds [44]. Besides, argusides D (110) and E (111)
have also been tested for their anticancer activities in above human cancer cell lines and revealed a
significant activity with IC50 values in the range of 3.36–7.77 µg/mL [45] (Figure 10). This finding
shows that compounds 108 and 109 are potent cytotoxic agents compared to compounds 110 and 111. It has been reported that the length and type of sugar moieties of glycosides play an important role in
terms of cytotoxic activity against tumor cells and this observation clearly indicates that. Arguside A (107) also exhibited significant cytotoxicity against different human tumor cell lines
while showing the highest activity towards human colorectal carcinoma (HCT-116) cells (IC50 = 0.14 µM)
with more potency than the employed positive control, 10-hydroxycamptothecin (IC50 = 0.84 µM) [43]. Argusides B (108) and C (109) have also shown potent cytotoxicity against human tumor cell lines,
adenocarcinomic human alveolar basal epithelial cells (A549), HCT-116, HepG2, and human breast
adenocarcinoma (MCF-7) cell lines. The cytotoxicity of the compounds on A549 (108-IC50 =
0.48 µg/mL, 109-IC50 = 0.43 µg/mL) and HCT-116 (108-IC50 = 0.46 µg/mL, 109-IC50 = 0.38 µg/mL)
cells were more potent than the positive control V-16 (Figure 9). However, there was no significant
difference between the cytotoxicity of two compounds [44]. Besides, argusides D (110) and E (111)
have also been tested for their anticancer activities in above human cancer cell lines and revealed a
significant activity with IC50 values in the range of 3.36–7.77 µg/mL [45] (Figure 10). This finding
shows that compounds 108 and 109 are potent cytotoxic agents compared to compounds 110 and 111. It has been reported that the length and type of sugar moieties of glycosides play an important role in
terms of cytotoxic activity against tumor cells and this observation clearly indicates that. Figure 9. Triterpenoid glycosides arguside A, B and C from sea cucumbers. 7896 Molecules 2013, 18 7896 Figure 9. Cont. Figure 9. Cont. Figure 9. Cont. Figure 9. Cont. Figure 10. Triterpenoid glycosides arguside D and E from sea cucumbers. Figure 10. Triterpenoid glycosides arguside D and E from sea cucumbers. 3. Triterpenoids from Sea Cucumbers Moreover, the in vitro cytotoxicity of impatienside A (112) and bivittoside D (113) were evaluated
extensively by employing seven human cancer cell lines and the results showed that both glycosides
exhibited in vitro cytotoxicities similar to or better than that of the potent anticancer drug etoposide
(V-16) in four human tumor cells, A549 (112-IC50 = 0.35 µg/mL, 113-IC50 = 0.52 µg/mL),
HCT-116 (112-IC50 = 0.45 µg/mL, 113-IC50 = 0.37 µg/mL), DU-145 (112-IC50 = 1.14 µg/mL,
113-IC50 = 0.937 µg/mL), KB (112-IC50 = 1.6 µg/mL, 113-IC50 = 1.42 µg/mL). The structural
differences between glycosides 112 and 113 limited to their holostane skeleton, and no significant
difference in the cytotoxicity of the two glycosides was found. However, pervicoside C (103), an analogue
of 113 having the same aglycone but a different sugar chain, isolated from Holothuria fuscocinerea
Jaeger, exhibited weak activities against HCT-116 and A549 cancer cells, with IC50 values of 18.7 and
28.6 µg/mL, respectively [46]. According to these results it is again confirmed that the length and type of
sugar moieties of such glycosides play an important role in terms of cytotoxic activity against tumor cells. 17-Dehydroxyholothurinoside A (114) and griseaside A (115) are identified as promising anticancer
agents due to their significantly higher cytotoxicity against four human tumor cell lines, A549
(114-IC50 = 0.886 µM, 115-IC50 = 1.07 µM) , HL-60 (114-IC50 = 0.245 µM, 115-IC50 = 0.427 µM),
BEL-7402 (114-IC50 = 0.97 µM, 115-IC50 = 1.114 µM), and human acute lymphoblastic leukemia cell
line (Molt-4) (114-IC50 = 0.34 µM, 115-IC50 = 0.521 µM) compared to the positive control HCP Molecules 2013, 18 Molecules 2013, 18 7897 (A549 IC50 = 2.35 µM, BEL-7402 IC50 = 2.6 µM, HL-60 IC50 = 1.9 µM, Molt-4 IC50 = 2.2 µM) [47
(Figure 11). 2.35 µM, BEL-7402 IC50 = 2.6 µM, HL-60 IC50 = 1.9 µM, Molt-4 IC50 = 2.2 µM) [47] Figure 11. Triterpenoid glycosides impatienside A, bivittoside D, 17-dehydroxyholothurinoside
A and griseaside A from sea cucumbers. Figure 11. Triterpenoid glycosides impatienside A, bivittoside D, 17-dehydroxyholothurinoside
A and griseaside A from sea cucumbers. A and griseaside A from sea cucumbers. 3. Triterpenoids from Sea Cucumbers Hillaside C (116) has also been tested for its anticancer potential against eight human tumor cell
lines (A-549, MCF-7, human lung carcinoma cells-IA9, human clear cell carcinoma cells—CAKI-1,
human prostate cancer cells—PC-3, KB, KB-VIN, and human colorectal sdenocarcinoma cells—HCT-
8) and has exhibited cytotoxicity with IC50 values in the range of 0.15–3.20 µg/mL [48] (Figure 12). Compared to the positive control HCP the compound 116 has shown more potent cytotoxicity towards
CAKI-1 (IC50 = 0.15 µg/mL) and KB-VIN (IC50 = 2.81 µg/mL) cell lines. Three new triterpene
glycosides, intercedensides A (117), B (118), and C (119) from Mensamaria intercedens Lampert,
were widely studied for their anticancer activity employing 10 human tumor cell lines (A549, MCF-7,
IA9, CAKI-1, human glioblastoma cells—U-87-MG, PC-3, KB, KB-VIN, human skin melanoma
cells—SK-MEL-2, HCT-8). Interestingly all compounds showed a significant cytotoxicity against all
tumor cell lines within the IC50 value range of 0.7–4 µg/mL, and the compounds 117 and 119 showed
similar potencies, while compound 118 was generally more potent in all cell lines. Furthermore,
compound 117 also exhibited significant in vivo antineoplastic activity against mouse Lewis lung
cancer and mouse S180 sarcoma, with 48.39% and 57.48% tumor reduction levels [49] (Figure 13). 7898 Molecules 2013, 18 Figure 12. Triterpenoid glycoside hillaside C from sea cucumber. Figure 12. Triterpenoid glycoside hillaside C from sea cucumber. Figure 12. Triterpenoid glycoside hillaside C from sea cucumber. Figure 13. Triterpenoid glycosides intercedensides A, B, and C from sea cucumbers. Figure 13. Triterpenoid glycosides intercedensides A, B, and C from sea cucumbers. A new sulfated triterpene glycoside from Pentacta quadrangularis, philinopside E (120) showed a
significant cytotoxicity (IC50 = 0.75–3.50 µg/mL) against ten tumor cell lines (mouse lymphocytic
leukemia cells-P388, HL60, A549, lung adenocarcinoma cells-SPC-A4, gastric carcinoma cells -
MKN28, gastric carcinoma cells-SGC7901, BEL7402, human ovarian carcinoma - HO8901, human
fetal lung fibroblasts-W138, human epithelial carcinoma cells-A431) [41]. Furthermore, sulfated
triterpene glycoside intercedenside B (121) from Pseudocolochirus violaceus exhibited significant
cytotoxicity against cancer cell lines MKN-45 (human gastric adenocarcinoma) and HCT-116 with
IC50 values in the range of 0.052–0.442 μM and both compounds showed significantly higher activity
against HCT-116 compared to the positive control HCP [50] (Figure 14). Molecules 2013, 18 7899 Figure 14. Triterpenoid glycosides philinopside E and intercedenside B from sea cucumbers. Figure 14. Triterpenoid glycosides philinopside E and intercedenside B from sea cucumb Moreover, the sulfated triterpene glycosides, philinopsides A (122) and B (123) showed significant
cytotoxicity (IC50 = 0.75–3.50 µg/mL) against ten tumor cell lines (CAKI, HOS, KB-VIN, KB,
SM-MEL-2, U87-MG, HCT-8, IA9, A549, and PC3) [51] (Figure 15). Figure 15. Triterpenoid glycosides philinopsides A and B from sea cucumbers. Collectively, all these tripterpene glycosides of sea cucumber are very potent cytotoxic agents
towards a wide array of cancer types and the structural properties such as the composition of the sugar
moiety and the sulfation in the glycon unit are affecting directly to their cytotoxic potential. Even though a number of saponin compounds have been isolated and identified as potent cytotoxic
agents only few of them have been studied to unravel the mode of their cytotoxicity. Among them
detailed cytotoxic mechanisms of frondoside A (124), cucumarioside A2-2 (125), echinoside A (126)
and ds-echinoside A (127) have been reported against several cancer types in vitro and in vivo 7900 Molecules 2013, 18 Molecules 2013, 18 (Figures 16 and 17). All four compounds have shown their cytotoxicity towards cancer cells by
arresting the cell cycle progression via activating the apoptosis pathways which leads to the cell death. Frondoside A has shown potent apoptotic inducing properties against breast cancer, pancreatic cancer
and leukemia, cucumarioside A2-2 has studies against leukemia and echinoside A and ds-echinoside A
has been characterized against lever cancer [52–54]. These compounds activate the intrinsic apoptotic
pathway via suppressing the tumor suppressor gene p53. With the suppression of p53, apoptosis pathways
are induced and the caspases 3, 7, 8 and 9, the enzymes regulate the cell death process are activated. Interestingly in vivo studies have confirmed that frondoside A (100 µg/kg/day) effectively decreased
the growth of breast cancer xenografts in athymic mice without exerting any side-effects [52]. Moreover, frondoside A is also capable of inhibiting the cancer cell migration and invasion which will
ultimately reduce the progression of cancer to the other parts of the body. Similarly echinoside A and
ds-echinoside A treatment (2.5 mg kg−1) to the mice bearing H22 hepatocarcinoma tumors has reduced
the tumor weight by 49.8% and 55% respectively [54]. These studies evidently prove the higher
potential of these compounds as novel natural pharmacological agents against tumor growth and
cancer progression. Figure 16. Triterpenoid glycosides frondoside A, cucumarioside A2-2 from sea cucumbers. 7901 Molecules 2013, 18 Figure 17. Triterpenoid glycosides echinoside A and ds-echinoside A from sea cucumbers. Figure 17. Triterpenoid glycosides echinoside A and ds-echinoside A from sea cucumbers. 4. Triterpenoids from Marine Algae Molecules 2013, 18 7903 Two cycloartane-type triterpenoids, 3-epicyclomusalenol (145), and cyclosadol (146) were isolated
from brown algae Kjellmaniella crassifolia, two compounds were obtained from this species for the
first time. 3-epicyclomusalenol (145), and cyclosadol (146) have been reported to have moderate
chemo preventive effects [63,64] (Figure 19). Figure 19. Cycloartane-type triterpenoids from marine algae. Figure 19. Cycloartane-type triterpenoids from marine algae. 5. Triterpenoids from Marine-Derived Fungi 5. Triterpenoids from Marine-Derived Fungi 4. Triterpenoids from Marine Algae Two cytotoxic squalenoid-derived triterpenoids, laurenmariannol (128) and (21a)-21-hydroxythyrsiferol
(129) were isolated and identified from the marine red alga Laurencia mariannensis, which was
collected off the coast of Hainan and Weizhou Islands of China. Both compounds have displayed
significant cytotoxic activity against P-388 tumor cells with IC50 values of 0.6 and 6.6 mg/mL,
respectively [55]. The red seaweed Laurencia viridis is a rich source of squalene derived secondary metabolites. Three squalene-derived brominated triterpenes, dehydrothyrsiferol (130), isodehydrothyrsiferol (131)
and 10-epidehydrothyrisiferol (132), isolated from Laurencia viridis, have shown potent cytotoxic
activities against a number of cancer cell lines [56,57]. Polyethers, iubol (133), 22-hydroxy-15(28)-dehydrovenustatriol (134), 1,2-dehydropseudo-
dehydrothyrsiferol (135), and secodehydrothyrsiferol (136) exhibited significant cytotoxic activity
against a panel of cancer cell lines [58]. Two compounds, 16-hydroxydehydrothyrsiferol (137),
thyresenol A (138) and thyrsenol B (139) were also isolated from Laurencia viridis, and these 7902 Molecules 2013, 18 compounds have exhibited significant inhibitory action on protein phosphatase at a concentration of
10 mM. Moreover, they have shown potent cytotoxic activity against P388 cell line [59,60]. Moreover, they have shown potent cytotoxic activity against P388 cell line [59,60]. Five cytotoxic triterpenoids 28-anhy-drothyrsiferyl diacetate (140), l5-anhy-drothyrsiferyl diacetate
diacetate (141), magireol-A (142), magireol B (143) and magireol C (144) were isolated from Japanese
red alga Laurencia obtuse [61,62] (Figure 18). Figure 18. Squalenoid-derived triterpenoids from marine algae. Figure 18. Squalenoid-derived triterpenoids from marine algae. Figure 18. Squalenoid-derived triterpenoids from marine algae. β
α 5. Triterpenoids from Marine-Derived Fungi Triterpenoids are frequently found in marine source, but have rarely been reported from marine-
derived fungi. Rainer Ebel have reviewed 7 triterpenoids from marine-derived fungi [65]. In 2011,
three triterpenoids xylariacins A-C (147–149) were isolated from the fermented extract of Xylarialean
sp. A45, an endophytic fungus of Annona squamosa L., and their structures were determined by NMR
spectroscopy. These compounds have shown modest cytotoxic activities against human tumor cell line
HepG2 [66] (Figure 20). Figure 20. Triterpenoids xylariacins A-C from marine-derived fungi. 6. Structure Activity Relationships Figure 20. Triterpenoids xylariacins A-C from marine-derived fungi. 6. Structure Activity Relationships 6. Structure Activity Relationships Even though not much research in this area has been carried out, the anti-cancer activity of
triterpenes is believed to be directly correlated to their structural features. As suggested by many
authors the bioactivity of the triterpenes is a result of its strong membranolytic activity, and this
membranolytic activity is a function of the structural features of the glycosides [67]. In triterpene 7904 Molecules 2013, 18 glycosides the presence of an 18(20)-lactone as the aglycon with at least one oxygen group near it has
critical significance for biological activity of glycosides bearing 9(11) double bonds. In the case of
glycosides with a 7(8)-double bonds in their aglycon structure, those lacking a 16-keto group are more
active than those with a 16-keto group [68]. The characteristics of the attached glycon structure are
also critical for the bioactivities of the triterpene glycosides. It has been found that for the actions
leading to modification of the cellular membrane, presence of a linear tetrasaccharide chain is
significant [69]. And also Maltsev et al. [70] have reported that glycosides having quinovose as a
second monosaccharide unit are more active over others. The sulfation of the sugar chain is also a
significant factor related to bioactivity. A sulfate group at C-4 of the first xylose residue increases the
effect against membranes. The absence of a sulfate group at C-4 of the xylose residue in biocides
decreases its activity by more than one fold magnitude. On the other hand the presence of a sulfate at
C-4 of the first xylose in branched pentanosides having 3-O-methyl groups as a terminal monosaccharide
increases activity. However, the same sulfate can decrease the activity of branched pentanosides which
have glucose as the terminal residue. Sulfate groups attached to a C-6 position of terminal glucose and
3-O-methylglucose residues impart a great reduction in the activity [67]. 7. Addressing the Limitations of Using Anti-Cancer Tritepenoids as Therapeutics Up to date the vast chemodiversity in the oceans has paved the way for natural product chemists to
mine for new bioactive compounds. Among them triterpenoids are one of the most studied classes of
compounds. Due to the extreme environments in the oceans, survival demands have resulted in the
evolution of these sophisticated toxic compounds and this fact is confirmed by the proven toxicity of
these compounds in biochemical studies. Triterpenoids derived from sea cucumbers, sponges and algae
have been used as ingredients in Traditional Chinese Medicine for years. Even though there are many
lead compounds with promising potential to be used as drugs for cancer therapy, the cytotoxicity itself
would be a constraint for this purpose, because most of the compounds could be cytotoxic towards
normal cells in addition to the cancerous cells. In identification of therapeutics from natural products
the preference is given to the compounds having high specificity towards the cancer cells in their
cytotoxic action, while minimizing the damage to normal cells [71]. Therefore considerable
cytotoxicity studies should be conducted employing the lead compounds before introducing them to
the drug development phase. However, advances in drug delivery systems could be applied effectively
in specific delivery of therapeutics. Cancer cells carry specific surface receptors that are expressed at
higher levels than their normal counterparts. Often these receptors have binding affinity towards
specific proteins or peptides [72]. This could be used for the direct targeting of cancer cells which is
effectively applicable to develop targeted delivery systems. Nano-drug carriers coated with cancer cell
receptor binding factors are a novel and effective approach for the delivery of drugs [73,74]. This
method could be used to deliver anti-cancer triterpene glycosides to the cancer tissues and thereby
protecting the adjacent normal tissue cells. Moreover, the possibility of continuous supply of the product and the ecological importance of the
triterpene sources such as sponges, sea cucumbers and algae are factors of importance before entering
to the drug development phase. Sustainable production of compound through chemical synthesis or
culturing of these marine organisms should be ensured. The structural complexities have challenged Molecules 2013, 18 7905 the chemical synthesis and thus it would limit the entering of these compounds to drug development
phase. However, with the advances in synthetic chemistry and understanding of triterpene biosynthetic
processes, new opportunities for exploitation of these compounds as drug leads are opening up. Acknowledgments This work was supported by a grant from Marine Bioprocess Research Center of the Marine
Biotechnology Program funded by the Ministry of Land, Transport and Maritime, Republic of Korea. Conflict of Interest The authors declare no conflict of interest. The authors declare no conflict of interest. References Molecules 2013, 18 7906 13. Narang, A.S.; Desai, D.S. Anticancer drug development. In Pharmaceutical Perspectives of
Cancer Therapeutics; Springer: Berlin, Germany, 2009; pp. 49–92. 14. Petronelli, A.; Pannitteri, G.; Testa, U. Triterpenoids as new promising anticancer drugs. Anticancer Drugs 2009, 20, 880–892. 15. Tang, S.A.; Deng, Z.W.; Li, J.; Fu, H.Z.; Pei, Y.H.; Zhang. S.; Lin, W.H. A new isomalabaricane
triterpenoid from sponge Jaspis sp. Chin. Chem. Lett. 2005, 16, 353–355. 16. McCormick, J.L., McKee, T.C., Cardellina, J.H.; Leid, M.; Boyd, M.R. Cytotoxic triterpenes
from a marine sponge, Stelletta sp. J. Nat. Prod. 1996, 59, 1047–1050. 17. Ryu, G.; Matsunaga, S.; Fusetani, N. Globostellatic acids A-D, new cytotoxic isomalabaricane
triterpenes from the marine sponge Stelletta globostellata. J. Nat. Prod. 1996, 59, 512–514. 18. Lin, H.W.; Wang, Z.L.; Wu, J.H; Shi, N.; Zhang, H.J; Chen, W.S.; Morris-Natschke, S.L.;
Lin, A.S. Stellettins L and M, Cytotoxic isomalabaricane-type triterpenes, and sterols from the
marine sponge Stelletta tenuis. J. Nat. Prod. 2007, 70, 1114–1117. 19. Lv, F.; Deng, Z.W.; Li, J.; Fu, H.Z.; Soest, R.W.M.; Proksch, P.; Lin, W.H. Isomalabaricane-Type
Compounds from the Marine Sponge Rhabdastrella aff. distincta. J. Nat. Prod. 2004, 67,
2033–2036. 20. Su, J.Y.; Meng, Y.H.; Zen, L.M. Stelletttin A, a new triterpenoid pigment from the marine sponge
Stelletta tenuzs. J. Nat. Prod. 1994, 57, 1450–1451. 21. Essack, M.; Bajic, V.B.; Archer, J.A.C. Recently confirmed apoptosis-inducing lead compounds
isolated from marine sponge of potential relevance in cancer treatment. Mar. Drugs 2011, 9,
1580–1606. 22. Ebada, S.S.; Lin, W.H.; Proksch, P. Bioactive sesterterpenes and triterpenes from marine sponges:
Occurrence and pharmacological significance. Mar. Drugs 2010, 8, 313–346. 23. Tasdemir, D.; Mangalindan, G.C.; Concepcion, G.P.; Verbitski, S.M., Rabindran, S.;
Miranda, M.; Greenstein, M.; Hooper, J.N.A.; Harper, M.K.; Ireland, C.M. Bioactive isomalabaricane
triterpenes from the marine sponge Rhabdastrella globostellata. J. Nat. Prod. 2002, 65, 210–214. 24. Clement, J.A.; Li, M.; Hecht, S.M., Kingston, D.G. I. Bioactive isomalabaricane triterpenoids
from Rhabdastrella globostellata that stabilize the binding of DNA polymerase β to DNA. J. Nat. Prod. 2006, 69, 373–376. 25. Tsuda, M.; Ishibashi, M.; Agemi, K.; Sasaki, T.; Kobayashi, J. Stelliferins A–F, new
antineoplastic isomalabaricane triterpenes from the Okinawan marine sponge Jaspis stellifera. Tetrahedron 1991, 47, 2181–2194. 26. Meragelman, K.M.; McKee, T.C., Boyd, M.R. New cytotoxic isomalabaricane triterpenes from
the sponge Jaspis Species. J. Nat. Prod. 2001, 64, 389–392. 27. Liu, W.K.; Ho, J.C.K.; Che, C.T. References 1. Connolly, J.D.; Hill, R.A. Triterpenoids. Nat. Prod. Rep. 1997, 14, 661–679. 2. Connolly, J.D.; Hill, R.A. Triterpenoids. Nat. Prod. Rep. 2002, 19, 494–513. 3. Ravi, B.N.; Wells, R.J. Malabaricane triterpenes from a fijian collection of the sponge Jaspis
stellifera. J. Org. Chem. 1981, 46, 1998–2001. 3. Ravi, B.N.; Wells, R.J. Malabaricane triterpenes from a fijian collection of the sponge Jaspis
stellifera. J. Org. Chem. 1981, 46, 1998–2001. 4. Rao, Z.G.; Deng, S.Z.; Wu, H.M.; Jiang, S.K. Rhabdastrellic acid-A, a novel triterpenoid from the
marine sponge Rhabdastrella globostellata. J. Nat. Prod. 1997, 60, 1163–1164. 4. Rao, Z.G.; Deng, S.Z.; Wu, H.M.; Jiang, S.K. Rhabdastrellic acid-A, a novel triterpenoid from the
marine sponge Rhabdastrella globostellata. J. Nat. Prod. 1997, 60, 1163–1164. 5. Aminin, D.L.; Koy, C., Dmitrenok, P.S.; Müller-Hilke B.; Koczan, D.; Arbogast, B.; Silchenko,
A.A.; Kalinin, V.I..; Avilov, S.A.; Stonik, V.A.; et al. Immunomodulatory effects of holothurian
triterpene glycosides on mammalian splenocytes determined by mass spectrometric proteome
analysis. J. Proteomics. 2009, 72, 886–906. 5. Aminin, D.L.; Koy, C., Dmitrenok, P.S.; Müller-Hilke B.; Koczan, D.; Arbogast, B.; Silchenko,
A.A.; Kalinin, V.I..; Avilov, S.A.; Stonik, V.A.; et al. Immunomodulatory effects of holothurian
triterpene glycosides on mammalian splenocytes determined by mass spectrometric proteome
analysis. J. Proteomics. 2009, 72, 886–906. 6. Cen-Pacheco, F.; Nordström, L.; Souto, M.L.; Martín, M.N.; Fernández, J.J.; Daranas, A.H. Studies on polyethers produced by red algae. Mar. Drugs 2010, 8, 1178–1188. 7. Zhang, M.; Wang,W.L.; Fang,Y.C.; Zhu,T.J.; Gu, Q.Q. Zhu, W.M. Cytotoxic alkaloids and
antibiotic nordammarane triterpenoids from the marine-derived fungus Aspergillus sydowi. J. Nat. Prod. 2008, 71, 985–989. 8. Bishayee, A.; Ahmed, S.; Brankov, N.; Perloff, M. Triterpenoids as potential agents for the
chemoprevention and therapy of breast cancer. Front Biosci. 2011, 16, 980–996. 9. Drag, M.; Surowiak, P.; Drag-Zalesinska, M.; Dietel, M.; Lage, H.; Oleksyszyn, J. Comparision
of the cytotoxic effects of birch bark extract, betulin and betulinic acid towards human gastric
carcinoma and pancreatic carcinoma drug-sensitive and drug-resistant cell lines. Molecules 2009,
14, 1639–1651. 10. Liby, K.T.; Yore, M.M.; Sporn, M.B. Triterpenoids and rexinoids as multifunctional agents for
the prevention and treatment of cancer. Nat. Rev. Cancer 2007, 7, 357–369. 11. Hill, R.A.; Connolly, J.D. Triterpenoids. Nat. Prod. Rep. 2012, 29, 780–818. 12. Naveen Kumar, D.R.; Shikha, S.; Cijo George, V.; Suresh, P.K.; Ashok Kumar, R. Anticancer
and anti-metastatic activities of RHEUM EMODI rhizome chloroform extracts. Asian J. Pharm. Clin. Res. 2012, 3, 189–194. References Apoptotic activity of isomalabaricane triterpenes on human
promyelocytic leukemia HL60 cells. Cancer Letters 2005, 230, 102–110. 28. Antonov, A.S.; Kalinovsky, A.I.; Stonik, V.A.; Afiyatullov, S.S.; Aminin, D.L.; Dmitrenok, P.S.;
Mollo, E.; Cimino, G. Isolation and structures of erylosides from the Carribean sponge Erylus
formosus. J. Nat. Prod. 2007, 70, 169–178. 29. Rudi, A.; Kashman, Y. Sodwanones A-F, new triterpenoids from the marine sponge Axinella
weltneri. J. Nat. Prod. 1994, 57, 1416–1423. Molecules 2013, 18 7907 30. Rudi, A.; Goldberg, I.; Stein, Z.; Kashman, Y. Sodwanones G, H, and I, new cytotoxic triterpenes
from a marine sponge. J. Nat. Prod. 1995, 58, 1702–1712. 31. Rudi, A.; Aknin, M.; Gaydou, E.M.; Kashman, Y. Sodwanones K, L, and M; new triterpenes from
the marine sponge Axinella weltneri. J. Nat. Prod. 1997, 60, 700–703. 32. Bon, C.F.; Berrué, F.; Thomas, O.P.; Reyes, F.; Amade, P. Sodwanone S, a triterpene from the
marine sponge Axinella weltneri. J. Nat. Prod. 2005, 68, 1284–1287. 33. Dai, J.Q.; Fishback, J.A.; Zhou,Y.D.; Nagle, D.G. Sodwanone and yardenone triterpenes from a
South African species of the marine sponge Axinella inhibit hypoxia-inducible factor-1 (HIF-1)
activation in both breast and prostate tumor cells. J. Nat. Prod. 2006, 69, 1715–1720. 34. Cimino, G.; Madaio, A.; Trivellone, E. Minor triterpenoids from the mediterranean sponge,
Raspaciona aculeata. J. Nat. Prod. 1994, 57, 784–790. 35. Cimino, G.; Crispino, A.; Madaio, A.; Trivellone, E. Raspacionin B, a further triterpenoid from
the mediterranean sponge Raspaciona aculeata. J. Nat. Prod. 1993, 56, 534–538. 36. Cimino, G.; Epifanio, R. D. A.; Madaio, A.; Puliti, R.; Trivellone, E. Absolute stereochemistry of
raspacionin, the main triterpenoid from the marine sponge Raspaciona aculeata. J. Nat. Prod. 1993, 56, 1622–1626. 37. Jain, S.; Abraham, I.; Carvalho, P.; Kuang, Y.H.; Shaala, L.; Youssef, D.T.A.; Avery, M.A.,
Chen, Z.S.; Sayed, K.E. Sipholane triterpenoids: Chemistry, reversal of ABCB1/P-glycoprotein-
mediated multidrug resistance, and pharmacophore modeling. J. Nat. Prod. 2009, 72, 1291–1298. 38. Jain, S.; Laphookhieo, S.; Shi, Z.; Fu, L.W.; Akiyama, S.; Chen, Z.S.; Youssef, D.T. A.;
Soest, R.W.M.; Sayed, K.E. Reversal of P-glycoprotein-mediated multidrug resistance by
sipholane triterpenoids. J. Nat. Prod. 2007, 70, 928–931. 39. Shi, Z.; Jain, S.; Kim, I.W.; Peng, X.X.; Abraham, I.; Youssef, D.T.A.; Fu, L.W.; Sayed, K.E.;
Ambudkar, S.V.; Chen, Z.S. Sipholenol A, a marine-derived sipholane triterpene, potently
reverses P-glycoprotein (ABCB1)-mediated multidrug resistance in cancer cells. Cancer Sci. 2007, 98, 1373–380. 40. References Abraham, I.; Jain, S.; Wu, C.P.; Khanfar, M.; Kuang, Y.H.; Dai, C.L.; Shi, Z.; Chen, X.;
Fu, L.W.; Ambudkar, S.V.; Sayed, K.E.; Chen, Z.S. Marine sponge-derived sipholane triterpenoids
reverse P-glycoprotein (ABCB1)-mediated multidrug resistance in cancer cells. Biochem. Pharmacol. 2010, 80, 1497–1506. 41. Zhang, S.; Yi, Y.; Tang, H.; Li, L.; Sun, P.; Wu, J. Two new bioactive triterpene glycosides from
the sea cucumber Pseudocolochirus violaceus. J. Asian Nat. Prod. Res. 2006, 8, 1–8. 42. Dang, N.H.; Thanh, N.V.; Kiem, P.V.; Huong, L.M.; Minh, C.V.; Kim, Y.H. Two new triterpene
glycosides from the vietnamese sea cucumber Holothuria scabra. Arch. Pharm. Res. 2007, 30,
1387–1391. 43. Liu, B.S.; Yi, Y.H.; Li, L. Arguside A: A new cytotoxic triterpene glycoside from the sea
cucumber Bohadschia argus Jaeger. Chem. Biodiver. 2007, 4, 2845–2851. 44. Liu, B.S.; Yi, Y.H.; Li, L. Argusides B and C, two new cytotoxic triterpene glycosides from the
sea cucumber Bohadschia argus Jaeger. Chem. Biodiver. 2008, 5, 1288–1297. 45. Liu, B.S., Yi, Y.H., Li, L., et al. Argusides D and E, two new cytotoxic triterpene glycosides from
the sea cucumber Bohadschia argus Jaeger. Chem. Biodiver. 2008, 5, 1425–1433. Molecules 2013, 18 7908 46. Sun. P.; Liu, B.S.; Yi, Y.H. A new cytotoxic lanostane-type triterpene glycoside from the sea
cucumber Holothuria impatiens. Chem. Biodiver. 2007, 4, 450–457. 47. Sun, G.Q.; Li, L.; Yi, Y.H.; Yuan, W.H.; Liua, B.S.; Weng, Y.Y.; Zhang, S.L.; Sun, P.;
Wang, Z.L. Two new cytotoxic nonsulfated pentasaccharide holostane (=20-Hydroxylanostan-18-
oic Acid γ-Lactone) glycosides from the sea cucumber Holothuria grisea. Helv. Chim. Acta 2008,
91, 1453–1460. 48. Wu, J.; Yi, Y.H.; Tang, H.F.; Zou, Z.R.; Wu. H.M. Structure and cytotoxicity of a new
lanostane-type triterpene glycoside from the sea cucumber Holothuria hilla. Chem. Biodivers
2006, 3, 1249–1254. 49. Zou, Z.R.; Yi, Y.H.; Wu, H.M.; Wu, J.H.; Liaw, C.C.; Lee, K.H. Intercedensides A−C, three new
cytotoxic triterpene glycosides from the sea cucumber Mensamaria intercedens Lampert. J. Nat. Prod. 2003, 66, 1055–1060. 50. Zhang, S.Y.; Tang, H.F.; Yi, Y.H. Cytotoxic triterpene glycosides from the sea cucumber
Pseudocolochirus violaceus. Fitoterapia 2007, 78, 283–287. 51. Yi, Y.H.; Xu, Q.Z.; Li, L.; Zhang, S.L.; Wu, H.M.; Ding, J.; Tong, Y.G.; Tan, W.F.; Li, M.H.;
Tian, F.; Wu, J.H.; Liaw, C.C.; Bastow, K.F.; Lee, K.H. Philinopsides A and B, two new sulfated
triterpene glycosides from the sea cucumber Pentacta quadrangularis. Helv. Chim. Acta 2006, 89,
54–63. 52. Marzouqi, N.A.; Iratni, R.; Nemmar, A. References Frondoside A inhibits human breast cancer cell survival,
migration, invasion and the growth of breast tumor xenografts. Eur. J. Pharmacol. 2011, 668, 25–34. 53. Jin, J.O.; Shastina, V.V.; Shin, S.W. Differential effects of triterpene glycosides, frondoside A and
cucumarioside A2–2 isolated from sea cucumbers on caspase activation and apoptosis of human
leukemia cells. FEBS Lett. 2009, 583, 697–702. 54. Zhao, Q.; Xue, Y.; Wang, J.F.; In vitro and in vivo anti-tumour activities of echinoside A and
ds-echinoside A from Pearsonothuria graeffei. J. Sci. Food Agric. 2012, 9, 965–974. 55. Ji, N.Y.; Li, X.M.; Xie, H.; Ding, J.; Li, K.; Ding, L.P.; Wang, B.G. Highly oxygenated
triterpenoids from the marine red alga Laurencia mariannensis (Rhodomelaceae). Helv. Chim. Acta
2008, 91, 1940–1946. 56. Norte, M.; Fernandez, J.J.; Souto, M.L.; Garcia-Gravalos, M.D. Two new antitumoral polyether
squalene derivatives. Tetrahedron Lett. 1996, 15, 2671–2674. 57. Pec, M.K.; Aguirre, A.; Moser-Their, K.; Fernandez, J.J.; Souto, M.L.; Dota, J.; Diaz-Gonzalez, F.;
Villar, J. Induction of apoptosis in estrogen dependent and independent breast cancer cells by the
marine terpenoid dehydrothyrsiferol. Biochem. Pharmacol. 2003, 65, 1451–1461. 58. Pacheco, F.C.; Villa-Pulgarin, J.A.; Mollinedo, F.; Martín, M.N.; Fernández, J.J.; Daranas, A.H. New polyether triterpenoids from laurencia viridis and their biological evaluation. Mar. Drugs
2011, 9, 2220–2235. 59. Souto, M.L.; Manrı´quez, C. P.; Norte, M.; Leira, F.; Ferna´ndez, J. The inhibitory effects of
squalene-derived triterpenes on protein phosphatase PP2A. Bioorg. Med. Chem. Lett. 2003, 13,
1261–1264. 60. Norte, M.; Fernandez, J.J.; Saouto, M.L.; Gavin, J.A.; GarciaGravalos, M.D. Thyrsenols A and B
two unusual polyether squalene derivatives. Tetrahedron 1997, 53, 3173–3178. Molecules 2013, 18 7909 61. Suzuki, T.; Takeda, S.; Suzuki, M.; Kurosawa, E.; Kato, A.; Imanaka, Y. Constituents of marine
plants. Part 67. Cytotoxic squalene-derived polyethers from the marine red alga Laurencia obtusa
(Hudson) Lamouroux. Chem. Lett. 1987, 361–364. 61. Suzuki, T.; Takeda, S.; Suzuki, M.; Kurosawa, E.; Kato, A.; Imanaka, Y. Constituents of marine
plants. Part 67. Cytotoxic squalene-derived polyethers from the marine red alga Laurencia obtusa
(Hudson) Lamouroux. Chem. Lett. 1987, 361–364. 62. Gamal, A. A.E. Biological importance of marine algae. Saudi Pharm. J. 2010, 18 63. Wu, Z.H.; liu, T.; Gu, C.X.; Shao, C.L.; Zhou, J.; Wang, C.Y. Steroids and triterpenoids from the
brown alga Kjellmaniella crassifolia. Chem. Nat. Compd. 2012, 48, 158–160. 64. Kikuchi, T.; Akihisa, T.; Tokuda, H.; Ukiya, M.; Watanabe, K.; Nishino, H. © 2013 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article
distributed under the terms and conditions of the Creative Commons Attribution license
(http://creativecommons.org/licenses/by/3.0/). References Cancer
chemopreventive effects of cycloartane-type and related triterpenoids in in vitro and in vivo
models. J. Nat. Prod. 2007, 70, 918–922. 65. Ebel, R. Terpenes from marine-derived fungi. Mar. Drugs 2010, 8, 2340–2368. 66. Lin, T.; Lin, X.; Lu, C.H.; Shen, Y.M. Three new triterpenes from Xylarialean sp. A45, an
endophytic fungus from Annona squamosa L. Helv. Chim. Acta 2011, 94, 301–305. 67. Kalinin, V.I. System-theoretical (Holistic) Approach to the modeling of structural functional
relationships of biomolecules and their evolution: An example of triterpene glycosides from sea
cucumbers (Echinodermata, Holothurioidea). J. Theor. Biol. 2000, 206, 151–168. 68. Kalinin, V.I.; Prkofieva, N. G.; Likhatskaya, G. N. Hemolytic activities of triterpene glycosides
from the holothurian order Dendrochirotida: Some trends in the evolution of this group of toxins. Toxicon 1996, 34, 475–483. 69. Kalinin, V.I.; Volkova, O.V.; Likhatskaya, G. N. Hemolytic activity of triterpene glycosides from
Cucumariidae family holothurians and evolution of this group of toxins. J. Nat. Toxins 1992, 1,
17–30. 70. Maltsev, I.I.; Stekhova, S.I.; Schentsova, E.B.; Anisimov, M.M.; Stonik, V.A. Antimicrobial
activities of glycosides from the sea cucumbers of family. Stichopodidae. Khim-Pharm. Zhurn
1985, 1985, 54–56. 71. Kim, S.K.; Himaya, S.W.A. Triterpene Glycosides from Sea Cucumbers and Their Biological
Activities. Adv. Food Nutr. Res. 2012, 65, 297–319. 72. Allen, T.M.; Cullis, P.R. Drug Delivery Systems: Entering the Mainstream. Science 2004, 303,
1818–1822. 73. Ferrari, M. Cancer nanotechnology: Opportunities and challenges. Nat. Rev. Cancer 2005, 5,
161–171. 74. Cho, K.; Wang, X.; Nie, S.; Chen, Z.; Shin, D.M. Therapeutic Nanoparticles for Drug Delivery in
Cancer. Clin. Cancer Res. 2008. 14, 1310–1316. © 2013 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article
distributed under the terms and conditions of the Creative Commons Attribution license
(http://creativecommons.org/licenses/by/3.0/). © 2013 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article
distributed under the terms and conditions of the Creative Commons Attribution license
(http://creativecommons.org/licenses/by/3.0/). © 2013 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article
distributed under the terms and conditions of the Creative Commons Attribution license
(http://creativecommons.org/licenses/by/3.0/).
|
https://openalex.org/W2148110196
|
https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0044012&type=printable
|
English
| null |
Abdominal Contents from Two Large Early Cretaceous Compsognathids (Dinosauria: Theropoda) Demonstrate Feeding on Confuciusornithids and Dromaeosaurids
|
PloS one
| 2,012
|
cc-by
| 8,611
|
Results Sinocalliopteryx (CAGS-IG-T1) Description and
Comparison Materials Abdominal contents provide the most reliable record of diet in
extinct vertebrates although preservation of such remains is rare
and frequently difficult to demonstrate unequivocally. The Lower
Cretaceous Yixian Formation in northeastern China preserves a
remarkably diverse terrestrial fauna in fine-grained volcaniclastic-
lacustrine deposits [1,2]. Such lagersta¨tten preserve remarkable
anatomical features, including integumentary structures, organic
compounds (such as proteins responsible for pigmentation), and
abdominal contents in exquisite detail [3]. To date, gut contents
have been found in a wide range of Jehol taxa (Table 1),
demonstrating clear trophic relationships within the Jehol ecosys-
tem. The holotype of Sinocalliopteryx gigas (JMP-V-05-8-01) is a
complete, articulated, and exceptionally well-preserved skull and
skeleton with long filamentous integument (Figure 1; [4]). A new
specimen of Sinocalliopteryx sp. (CAGS-IG-T1) is a partially
articulated but incomplete skeleton lacking the cervical vertebrae,
parts of the dorsal and caudal series, both pectoral and pelvic
girdles and the proximal parts of both fore- and hindlimbs. Both
specimens are from the Jianshangou Beds of the Yixian Formation
(Neocomian; [5,6]), Beipiao, western Liaoning Province, north-
eastern China. Abstract Two skeletons of the large compsognathid Sinocalliopteryx gigas include intact abdominal contents. Both specimens come
from the Jianshangou Beds of the lower Yixian Formation (Neocomian), Liaoning, China. The holotype of S. gigas preserves a
partial dromaeosaurid leg in the abdominal cavity, here attributed to Sinornithosaurus. A second, newly-discovered
specimen preserves the remains of at least two individuals of the primitive avialan, Confuciusornis sanctus, in addition to
acid-etched bones from a possible ornithischian. Although it cannot be stated whether such prey items were scavenged or
actively hunted, the presence of two Confuciusornis in a grossly similar state of digestion suggests they were consumed in
rapid succession. Given the lack of clear arboreal adaptations in Sinocalliopteryx, we suggest it may have been an adept
stealth hunter. Citation: Xing L, Bell PR, Persons WS IV, Ji S, Miyashita T, et al. (2012) Abdominal Contents from Two Large Early Cretaceous Co
Theropoda) Demonstrate Feeding on Confuciusornithids and Dromaeosaurids. PLoS ONE 7(8): e44012. doi:10.1371/journal.pone.004 Editor: Andrew A. Farke, Raymond M. Alf Museum of Paleontology, United States of America Editor: Andrew A. Farke, Raymond M. Alf Museum of Paleontology, United States of America Editor: Andrew A. Farke, Raymond M. Alf Museum of Paleontology, United States of America Received April 18, 2012; Accepted July 27, 2012; Published August 29, 2012 Received April 18, 2012; Accepted July 27, 2012; Published August 29, 2012 Copyright: 2012 Xing et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author and source are credited. Copyright: 2012 Xing et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author and source are credited. Competing Interests: The authors have declared that no competing interests exist. * E-mail: philbyb@gmail.com Competing Interests: The authors have declared that no competing interests exist. * E-mail: philbyb@gmail.com Abdominal Contents from Two Large Early Cretaceous
Compsognathids (Dinosauria: Theropoda) Demonstrate
Feeding on Confuciusornithids and Dromaeosaurids
Lida Xing1, Phil R. Bell2*, W. Scott Persons IV1, Shuan Ji3, Tetsuto Miyashita1, Michael E. Burns1, Lida Xing1, Phil R. Bell2*, W. Scott Persons IV1, Shuan Ji3, Tetsuto Miyashita1, Michael E. Burns1,
Qiang Ji3, Philip J. Currie1 1 Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada, 2 Pipestone Creek Dinosaur Initiative, Clairmont, Alberta, Canada, 3 Institute of
Geology, Chinese Academy of Geological Sciences, Beijing, China Institutional Abbreviations Compsognathidae, typified by the type species Compsognathus
longipes, includes a group of relatively small (approximately 1 m
long) basal coelurosaurs from the Late Jurassic to Early Cretaceous
of Europe and Asia. However, Jehol compsognathids such as
Huaxiagnathus and Sinocalliopteryx attained significantly larger sizes,
reaching lengths of up to 2.3 m in the latter [4]. In their original
description of Sinocalliopteryx gigas, Ji et al. [4] commented on the
partial leg of an unidentified dromaeosaurid in the posterior region
of the abdominal cavity, which they cite as evidence of a highly
predaceous lifestyle in Sinocalliopteryx. The purpose of this paper is
to describe and reassess the abdominal contents of S. gigas based on
the holotype and a second specimen that indicates wider dietary
preferences with implications for feeding strategies of Compsog-
nathidae. CAGS-IG, Institute of Geology, Chinese Academy of Geolog-
ical Sciences, Beijing; China; JMP, Jinzhou Museum of Paleon-
tology, Jinzhou, Liaoning Province, China; NIGP, Nanjing
Institute of Geology and Paleontology, Nanjing, China; GMV,
China National Geological Museum, Beijing, China. ation: Xing L, Bell PR, Persons WS IV, Ji S, Miyashita T, et al. (2012) Abdominal Contents from Two Large Early Cretaceous Comps
eropoda) Demonstrate Feeding on Confuciusornithids and Dromaeosaurids. PLoS ONE 7(8): e44012. doi:10.1371/journal.pone.0044012 August 2012 | Volume 7 | Issue 8 | e44012 Sinocalliopteryx (CAGS-IG-T1) Description and
Comparison The skull of CAGS-IG-T1 includes both maxillae, right nasal,
right lacrimal, right prefrontal, right jugal, left palatine and vomer,
and fragmentary right dentary (Figure 2A, B). The left maxilla,
shown in medial view, has at least ten alveoli, six of which hold
teeth. Given that the anterior ramus is incomplete, the maxillary PLOS ONE | www.plosone.org August 2012 | Volume 7 | Issue 8 | e44012 1 Abdominal Contents from Chinese Compsognathids Table 1. Reported abdominal contents of Jehol taxa. Taxon
Abdominal contents
Reference
Mammalia
Repenomamus
Psittacosaurus
Hu et al. [67]
Dinosauria
Sinosauropteryx
unidentified mammal
Chen et al. [15]; Currie and Chen [10]
Sinocalliopteryx
Sinornithosaurus, Confuciusornis, unidentified
ornithiscian
Ji et al. [4]; this study
Microraptor
enantiornithine bird
O’Connor et al. [54]
Aves
Confuciusornis
cf. Jinanichthys
Dalsa¨tt et al. [68]
Hongshanornis
Plant seeds
Zheng et al. [69]
Jeholornis
Plant seeds
Zhou and Zhang [70]
Jianchangornis
Fish remains
Zhou et al. [71]
Sapeornis
Plant seeds
Zheng et al. [69]
Yanornis
Fish remains
Yuan [72]; Zhou et al. [73]
Choristodera
Hyphalosaurus
Fish remains
UALVP 54043, Unpublished
Squamata
Yabeinosaurus
Fish remains
Zhou and Wang [74]
doi:10.1371/journal.pone.0044012.t001 doi:10.1371/journal.pone.0044012.t001 tooth count probably exceeded ten by one or two. The most
posterior maxillary alveolus is ventral to or slightly anterior to the
anterior end of the maxillary-lacrimal contact. The anterior ramus
of the maxilla is demarcated by the transition to the poster-
odorsally-oriented ascending ramus. The posterodorsal process of the maxilla is dorsoventrally deeper than the horizontal ramus. Near the posterior end, the posterodorsal process splits into the
larger and longer lateral prong and the smaller and shorter medial
prong, between which the lacrimal was clasped. The medial
surface of the maxilla above the palatal shelf is smooth and not Figure 1. Holotype of Sinocalliopteryx gigas (JMP-V-05-8-01).
doi:10.1371/journal.pone.0044012.g001
PLOS ONE | www.plosone.org
2
August 2012 | Volume 7 | Issue 8 | e44012 Figure 1. Holotype of Sinocalliopteryx gigas (JMP-V-05-8-01). doi:10.1371/journal.pone.0044012.g001 Figure 1. Holotype of Sinocalliopteryx gigas (JMP-V-05-8-01). doi:10.1371/journal.pone.0044012.g001 Figure 1. Holotype of Sinocalliopteryx gigas (JMP-V-05-8-01). doi:10.1371/journal.pone.0044012.g001 August 2012 | Volume 7 | Issue 8 | e44012 PLOS ONE | www.plosone.org PLOS ONE | www.plosone.org 2 Figure 2. Sinocalliopteryx gigas (CAGS-IG-T1), partial skeleton. A, B; skull; C, dorsal vertebrae and ribs. Arrow points to partly covered
Confuciusornis humerus; D, associated feet; E, associated pedal phalanges and unguals; F, articulated tail with filamentous integument. Abbreviations:
C, centrum; Ch, chevron; Dr, dorsal rib; In, integument; Ju; jugal; La, lacrimal; Mt, metatarsal; Mx, maxilla; Na, nasal; Pal, palatine; Pfr, prefrontal; Ph,
phalanx; Pm, premaxilla; Po, postorbital; Tc, tooth crown; Un, ungual; V, vomer. Scale bars in C–F equal 10 cm in 1 cm increments. doi:10 1371/journal pone 0044012 g002
Abdominal Contents from Chinese Compsognathids Abdominal Contents from Chinese Compsognathids Figure 2. Sinocalliopteryx gigas (CAGS-IG-T1), partial skeleton. A, B; skull; C, dorsal vertebrae and ribs. Arrow points to partly covered
Confuciusornis humerus; D, associated feet; E, associated pedal phalanges and unguals; F, articulated tail with filamentous integument. Abbreviations:
C, centrum; Ch, chevron; Dr, dorsal rib; In, integument; Ju; jugal; La, lacrimal; Mt, metatarsal; Mx, maxilla; Na, nasal; Pal, palatine; Pfr, prefrontal; Ph,
phalanx; Pm, premaxilla; Po, postorbital; Tc, tooth crown; Un, ungual; V, vomer. Scale bars in C–F equal 10 cm in 1 cm increments. doi:10.1371/journal.pone.0044012.g002 excavated into the conspicuous maxillary antrum and promax-
illary recess as in other theropods [7]. The antorbital fossa has a
distinct margin. The maxillary fenestra is absent. maxilla posterior to the maxillary tooth row and posterior to the
anterior margin of the antorbital fenestra. The dentary is
preserved for the anteroposterior length of three tooth positions. maxilla posterior to the maxillary tooth row and posterior to the
anterior margin of the antorbital fenestra. The dentary is
preserved for the anteroposterior length of three tooth positions. Six dorsal vertebrae are preserved in one of the slabs
(Figure 2C). The neurocentral sutures are visible in all of these
vertebrae, but the sutures are not completely open because the
neural arches and the centra are tightly knit. All of the dorsal
vertebrae lack pleurocoels as in other compsognathids [8–14]. Six
left dorsal ribs are preserved with the vertebral series. In a separate
slab, two dorsal ribs, nine medial gastralia, and at least ten lateral
gastralia surround the abdominal contents of this specimen. August 2012 | Volume 7 | Issue 8 | e44012 Abdominal Contents from Chinese Compsognathids and partially overlapped by these elements (Figure 3A–F). The
right and left ischia as well as the abdomen of Sinocalliopteryx
CAGS-IG-T1 (along with the gastralia and the abdominal
contents) have all shifted posteriorly relative to their position in
life. Two elements of the abdominal contents (scapulocoracoid and
sternum) lie on a horizontal plane between the left and right ischia
(Figure 3C). Two slabs contain caudal vertebrae (Figure 2F). One
of the two slabs contains the 11th to 15th caudal vertebrae with L-
shaped haemal arches. The other slab contains an articulated
series of 13 mid- to distal-caudal vertebrae, of which 11 are
entirely preserved. In that slab, only the first two vertebrae have
dorsoventrally low neural spines. In comparison with the holotype
of Sinocalliopteryx [4], the most anterior vertebra in the series
represents the 16th caudal vertebra. All but the last two of the
vertebrae are associated with L-shaped haemal arches. In the same
slab, filamentous integument is preserved along both the dorsal
and ventral margins of the tail (Figure 2F). The qualities of
preservation and preparation on the specimen do not permit
microscopic comparison of the integument. The neurocentral
sutures are closed in all mid- to distal-caudal vertebrae. as long as or slightly longer than manual phalanx I-1, as in
Compsognathus, Huaxiagnathus, Juravenator, Scipionyx, and the holotype
of Sinocalliopteryx [4,9,13,14,16,17] but not as in Sinosauropteryx in
which manual phalanx I-1 is substantially longer [10]. Manual
phalanges II-1 and II-2 and the ungual for the digit are preserved
near the metatarsals in a separate slab. Both right and left metatarsals are preserved in a single slab. All
of the metatarsals are present for the left foot, whereas the right
foot is represented by only metatarsals II–IV (Figure 2D). In the
left foot, metatarsal I is 24% of the length of metatarsal III. Metatarsal V is reduced to a curved splint less than half the length
of metatarsal IV. Metatarsals II, III, and IV are cylindrical and
straight. Distal to the metatarsals is a complete digit III, and two
phalanges of digit I. Additional pedal phalanges are in the distal
foot slab. The right pedal phalanges II-1, II-2, III-1, III-2, III-3,
IV-3, IV-4, and pedal unguals II–IV are present (Figure 2E). Abdominal Contents from Chinese Compsognathids CAGS-IG-T1 clearly represents a compsognathid, distinguished
by the nasal excluded from the antorbital fenestra by the maxilla
and lacrimal, the absence of pleurocoelus in the dorsal vertebrae,
and the manual phalanx I-1 nearly as long as metacarpal II
[7,10,12,14,18]. CAGS-IG-T1 is anatomically almost identical to
JMP-V-05-8-01 (the holotype of Sinocalliopteryx gigas; [4]) and
therefore referable to Sinocalliopteryx gigas. CAGS-IG-T1 is larger
based on the postcranial measurements (Table 2). The size
difference between the two specimens is relatively greater in the
length of the metatarsals than in the radius or height of the
maxilla, presumably due to allometric growth. Although the
original diagnosis of Sinocallipteryx does not include any characters
preserved in CAGS-IG-T1, this specimen and the holotype of
Sinocalliopteryx gigas can be distinguished from the similarly-sized, The forelimb elements are scattered across two slabs. The main
forelimb slab has the partially articulated right forearm and hand. The radius, metacarpal II, metacarpal III, and manual phalanx I-
1 are complete, whereas other manual elements are overlain on
one another such that identification is difficult. Metacarpal III is
less than half as wide transversely as metacarpal II. This is the case
in Compsognathus and Sinocalliopteryx [4,14,16], but differs from
Huaxiagnathus, Nqwebasaurus, and Sinosauropteryx, in which metacar-
pal II is half as wide transversely as metacarpal III [10,11,13]. Although the full length of metacarpal II cannot be measured, it is Figure 3. Abdominal contents of Sinocalliopteryx gigas (CAGS-IG-T1). A, B; block containing Confuciusornis (blue) and unidentified
ornithischian (red) remains. C, Close up of Confuciusornis sternal and pectoral elements (small box in B); D, E; associated skeleton of Confuciusornis
(large box in B); F, proximal Confuciusornis humerus (arrow in Figure 2). Abbreviations: C, carpal; Dr, dorsal rib; Fu, furcula; Gs, gastralia; H, humerus; Il,
ilium; Is, ischium; Mu, manual ungula; Ms, miscellaneous ornithischian bone; Ph, phalanx; Pu, pubis; Ra, radius; Sc, scapulocoracoid; St, sternum. Scale
bars in A, B equal 10 cm in 1 cm increments; C, F in 1 mm increments. doi:10.1371/journal.pone.0044012.g003 Figure 3. Abdominal contents of Sinocalliopteryx gigas (CAGS-IG-T1). A, B; block containing Confuciusornis (blue) and unidentified
ornithischian (red) remains. C, Close up of Confuciusornis sternal and pectoral elements (small box in B); D, E; associated skeleton of Confuciusornis
(large box in B); F, proximal Confuciusornis humerus (arrow in Figure 2). Figure 1. Holotype of Sinocalliopteryx gigas (JMP-V-05-8-01).
doi:10.1371/journal.pone.0044012.g001
PLOS ONE | www.plosone.org
2
August 2012 | Volume 7 | Issue 8 | e44012 The
abdominal contents are between the dorsal ribs and the gastralia Six dorsal vertebrae are preserved in one of the slabs
(Figure 2C). The neurocentral sutures are visible in all of these
vertebrae, but the sutures are not completely open because the
neural arches and the centra are tightly knit. All of the dorsal
vertebrae lack pleurocoels as in other compsognathids [8–14]. Six
left dorsal ribs are preserved with the vertebral series. In a separate
slab, two dorsal ribs, nine medial gastralia, and at least ten lateral
gastralia surround the abdominal contents of this specimen. The
abdominal contents are between the dorsal ribs and the gastralia A lacrimal duct is present dorsal to the anterior margin of the
preorbital bar. The dorsal edge of the lacrimal is inflated into a
longitudinal, dorsally low cornual process. The prefrontal is as
long as the anterior ramus of the lacrimal. The postorbital process
of the right jugal was anteriorly displaced and now sits on the right
and left maxillae. The process retains a groove along its anterior
margin that would have received the postorbital. The vomer is
dorsoventrally deeper than the horizontal ramus of the maxilla
and has a dorsally convex margin. The palatine contacts the August 2012 | Volume 7 | Issue 8 | e44012 August 2012 | Volume 7 | Issue 8 | e44012 PLOS ONE | www.plosone.org PLOS ONE | www.plosone.org 3 Abdominal Contents from Chinese Compsognathids Abdominal Contents of CAGS-IG-T1 A disarticulated but associated skeleton of a confuciusornithine
bird is preserved within the posterior part of the Sinocalliopteryx
abdominal cavityin the vicinity of the distal ends of the ischia and
dorsal to the gastral basket (Figure 3A, B). Other remains are
scattered throughout the rest of the block. The associated elements
include the furcula, left and right scapulocoracoids, right humerus,
both radii, a metacarpal II, several manual phalanges, pelvis, a
pedal phalanx, possibly part of the femoral shaft, and several
unidentifiable bone fragments (Figure 3D, E). The proximal end of
a second humerus is preserved in the block containing the dorsal
vertebrae, whereas the humerus is partly covered by the dorsal ribs
(Figures 2C, 3F). Several additional elements reside in the region
between the ischia of Sinocalliopteryx, including a dorsal vertebra, at
least one phalanx, and part of the shaft of the ?tibiotarsus. The
sternum, a ?carpal, and a second right scapulocoracoid are
preserved close to the left ischium (Figure 3C). A fourth
scapulocoracoid overlies the left ischium. In addition to confuciusornithid remains, parts of two large
bones are also present within the Sinocalliopteryx abdominal cavity
(Figure 3A, B). Both are platy and incomplete with significant
surficial modification; the entire surfaces of both elements are
deeply pockmarked, resulting in the almost total loss of the original
external bone surfaces (Figure 4). The largest bone, tentatively
identified as a scapula, measures 13.5 cm in maximum length and
is 6.5 cm wide. The proximal end is expanded dorsoventrally, but
the acromion process is incomplete. The anteroventral expansion
is larger than the acromion and retroverted such that the posterior
margin of the expansion forms an acute angle with the scapula
blade. In its short, robust morphology, the scapula resembles the
scapulae of Psittacosaurus [26] and the basal ornithopod Yueosaurus
[27]; however, the element is so heavily modified that assignment
to any one taxon is contentious. The furcula, visible in posterior view, is robust and U-shaped. A
groove on this posterior surface is typical of C. santus [18]. Four
scapulocoracoids indicate the presence of at least two individuals. The scapula and coracoid are fused, a condition restricted among
Mesozoic Aves to Confuciusornis sanctus and Archaeopteryx lithographica
[18,19] but also present in some nonavian theropods such as
Velociraptor [20]. The sternum is damaged, presumably as a result
of digestive processes, but retains a median carina as in C. Abdominal Contents from Chinese Compsognathids brevis fossa, which is present in maniraptoran theropods [19]. The
distal end of the ischium is missing, but the proximal portion
retains a dorsal process that extends towards but does not contact
the postacetabular blade (Figure 3D, E). This feature is present in
C. sanctus and some enantiornithine birds, but is less developed in
Archaeopteryx [18,19]. contemporaneous compsognathid Huaxiagnathus [13] based on
several features of the maxilla: 1) The maxilla not as tall
dorsoventrally in both specimens of Sinocalliopteryx as it is in
Huaxiagnathus, in which the maxilla is two thirds taller at maximum
than the anterior ramus; 2) The dorsal margin of the posterodorsal
process of the maxilla forms an acute angle with the alveolar
margin in Sinocalliopteryx whereas in Huaxiagnathus, the dorsal
margin of the process is subparallel to the alveolar margin; 3) The
maxillary fenestra is absent in Sinocalliopteryx, whereas the fenestra
appears to be present in Huaxiagnathus [13]; 4) The anterior margin
of the antorbital fenestra is dorsal to the seventh or eighth
maxillary tooth position in Huaxiagnathus [13], whereas the anterior
margin of the fenestra is dorsal to at least the ninth or possibly the
tenth tooth position in Sinocalliopteryx. Based on the aforementioned shared features, the avian remains
in CAGS-IG-T1 are unequivocally assignable to Confuciusornis. Confuciusornis has had a troubled taxonomic history because the
description of the type species was inadequate, and as many as five
species have been assigned to that genus. Recent studies, however,
have demonstrated that all of these specimens fall within the range
of variation for the type species and are therefore synonymous
with C. sanctus [19,24]. Moreover, a second genus of confuciu-
sornithid, Jinzhouornis, and its two constituent species, has also been
shown to be qualitatively and quantitatively indistinguishable from
C. sanctus [25]. In light of this and the morphological consider-
ations already discussed, the associated confuciusornithid remains
in CAGS-IG-T1 are assigned to C. sanctus. Abdominal Contents of CAGS-IG-T1 sanctus,
whereas a carina is absent in Changchengornis [19]. The sternum of
Eoconfuciusornis apparently did not ossify [21]. Abdominal Contents of JMP-V-05-8-01 The abdominal contents of JMP-V-05-8-01 resembles an
inverted C-shape. Forming the upper part of the ‘C’ is a large
oval mass (approx. 10 cm long), centrally and dorsally positioned
within the abdominal cavity (Figure 5). This mass is composed of a
dense accumulation of filamentous feather-like structures up to
(and possibly exceeding) 22 mm in length. Where they are less
densely gathered, the feather-like structures show fibers that
branch off from a central filament (Figure 6A, B). In one area, a
single ‘tuft’ shows individual filaments that converge at their bases
(Figure 6C, D) in the same arrangement as the tufted integument
described for Sinornithosaurus [28]. The dromaeosaurid pes and
distal part of the leg transects this mass posteriorly to form the
vertical part of the ‘C’. A collection of feather-like structures occurs
along the length of, but apparently is not connected to, the
dromaeosaurid tibiotarsus. A central filament, or rachis, is visible
in each of these structures. A discretely arcing arrangement of
filaments has a striking resemblance to asymmetrical avian
contour feathers (Figure 5E, F). Ventrally, two small, circular
masses (approx. 3 cm in diameter) associated with gastroliths [4]
are present anterior to the pubic boot. The proximal end of the
dromaeosaurid tibiotarsus coincides with the more posterior of the
two smaller masses (Figure 5). The two circular masses within the
gastral basket are made up of fine, indeterminate matter, with no
indication of the filamentous structures seen elsewhere in the gut. The humerus is characteristically confuciusornithine, having an
expanded proximal end and a triangular deltopectoral crest that
constitutes more than one-third of the length of the humerus [19]. An oval foramen pierces the deltopectoral crest, which is an
autapomorphic feature of Confuciusornis [19,22]. A deltopectoral
foramen is absent in all other confuciusornithids including
Eoconfuciusornis [21] and Changchengornis [23]. The postacetabular process of the ilium is shorter than the
preacetabular process and tapers distally. There is no evidence of a Table 2. Select measurements (mm) for Sinocalliopteryx
gigas. Table 2. Select measurements (mm) for Sinocalliopteryx
gigas. Element
JMP-V-05-8-01
CAGS-IG-T1
Maxillary height (max)
40.8
44.5
Metatarsal III length
147.3
206.3
Radius length
100.7
118.64
doi:10.1371/journal.pone.0044012.t002 The dromaeosaurid hindlimb is from the right side of the body
and is preserved with its right lateral side exposed (Figure 5). Abdominal Contents from Chinese Compsognathids Abbreviations: C, carpal; Dr, dorsal rib; Fu, furcula; Gs, gastralia; H, humerus; Il,
ilium; Is, ischium; Mu, manual ungula; Ms, miscellaneous ornithischian bone; Ph, phalanx; Pu, pubis; Ra, radius; Sc, scapulocoracoid; St, sternum. Scale
bars in A, B equal 10 cm in 1 cm increments; C, F in 1 mm increments. doi:10.1371/journal.pone.0044012.g003 August 2012 | Volume 7 | Issue 8 | e44012 PLOS ONE | www.plosone.org 4 Abdominal Contents from Chinese Compsognathids August 2012 | Volume 7 | Issue 8 | e44012 Abdominal Contents of JMP-V-05-8-01 It is
overlain by the left gastralia and the left dorsal ribs and overlays a
number of right gastralia and one of the dorsal vertebrae. August 2012 | Volume 7 | Issue 8 | e44012 August 2012 | Volume 7 | Issue 8 | e44012 PLOS ONE | www.plosone.org 5 Abdominal Contents from Chinese Compsognathids Abdominal Contents from Chinese Compsognathids Figure 4. Possible ornithischian scapula (central red element in Figure 3B) within the abdomen of Sinocalliopteryx gigas (CAGS-IG-
T1). Note disorganized bone texture as a result of corrosion by gastric juices. doi:10.1371/journal.pone.0044012.g004 Figure 4. Possible ornithischian scapula (central red element in Figure 3B) within the abdomen of Sinocalliopteryx gigas (CAGS-IG-
T1). Note disorganized bone texture as a result of corrosion by gastric juices. doi:10.1371/journal.pone.0044012.g004 the length of the tibia and fibula [29–31], and the shaft of phalanx
II-2 is not strongly constricted between the articular facets [29]. Metatarsal IV displays a prominent ventral flange. The metatar-
sals are semi-arctometarsalian to a greater extent than in
Tianyuraptor [31]. Unfortunately, the metatarsals are crushed and
obscure one another, making other potentially diagnostic charac-
ters difficult or impossible to observe. Phalanx III-1 is not
exceptionally elongate or slender as it is in Graciliraptor [32]. The
limb is distinguishable from Microraptor based on its overall larger
size (tibial length 15.5 cm) and its lower stratigraphic position;
however, other, more diagnostic characters of the femur and pedal
unguals are missing in JMP-V-05-8-01. The preserved elements
are similar to those of Sinornithosaurus [29], and it is to this genus
that the limb is tentatively referred. Therefore, it can be conclusively identified as positioned within the
abdominal cavity. The visible limb elements include the tibia,
fibula, metatarsals III and IV, and numerous phalanges. Meta-
tarsals I and II are also likely present, but their positions are
obscured by matrix and the other limb bones. Some of the
phalangeal elements remain articulated, and nearly all the pedal
elements are near their articulated positions (Figure 7). The
metatarsals lie parallel to the tibia and fibula, with theirproximal
ends adjacent to the distal end of the tibia. The phalanges are
positioned in a clinched arrangement (Figure 7). Phalanx II-3 is hypertrophied, which is diagnostic of Dromaeo-
sauridae (Figure 7). As is common among dromaeosaurids from
the Jehol Group (including Graciliraptor, Microraptor, Sinornithosaurus,
and Tianyuraptor), the metatarsals are greatly elongate relative to Figure 5. Abdominal contents of Sinocalliopteryx gigas (JMP-V-05-8-01). Abdominal Contents of JMP-V-05-8-01 Blue, undigested feather-like structures; Red, dromaeosaurid
hindlimb; Yellow, gastroliths. Greek numerals (i–iii) denote enlargements in Figure 5. Abbreviations: Dr, dorsal rib; F, femur; H, humerus; Pu, pubis. doi:10.1371/journal.pone.0044012.g005 Figure 5. Abdominal contents of Sinocalliopteryx gigas (JMP-V-05-8-01). Blue, undigested feather-like structures; Red, dromaeosaurid
hindlimb; Yellow, gastroliths. Greek numerals (i–iii) denote enlargements in Figure 5. Abbreviations: Dr, dorsal rib; F, femur; H, humerus; Pu, pubis. doi:10.1371/journal.pone.0044012.g005 August 2012 | Volume 7 | Issue 8 | e44012 PLOS ONE | www.plosone.org PLOS ONE | www.plosone.org 6 Abdominal Contents from Chinese Compsognathids Figure 6. Undigested filamentous integument within the abdominal cavity of Sinocalliopteryx gigas (JMP-V-05-8-01). A, B; Enlargement
of i in Figure 4. Filamentous integument showing central vein (arrow in A; grey lines in B). Black lines are margins between adjacent filaments. Note
the stomach contents overlie the right dorsal rib (DrR) but are overlain by the left dorsal rib (DrL). C, D; Enlargement of ii in Figure 4. Tuft of filaments
showing single point of origin (arrow). E; Enlargement of iii in Figure 4 showing scattered filamentous structures. Boxed area and interpretive
illustration (F) shows a discrete association of parallel filaments similar to the barbs of an avian feather. Scale = 5 mm. doi:10.1371/journal.pone.0044012.g006 Figure 6. Undigested filamentous integument within the abdominal cavity of Sinocalliopteryx gigas (JMP-V-05-8-01). A, B; Enlargement
of i in Figure 4. Filamentous integument showing central vein (arrow in A; grey lines in B). Black lines are margins between adjacent filaments. Note
the stomach contents overlie the right dorsal rib (DrR) but are overlain by the left dorsal rib (DrL). C, D; Enlargement of ii in Figure 4. Tuft of filaments
showing single point of origin (arrow). E; Enlargement of iii in Figure 4 showing scattered filamentous structures. Boxed area and interpretive
illustration (F) shows a discrete association of parallel filaments similar to the barbs of an avian feather. Scale = 5 mm. doi:10.1371/journal.pone.0044012.g006 Discussion disarticulated and often broken, still retain relatively smooth
(uncorroded) bone surfaces, indicating minimal impact from
gastric acids. In contrast, the larger ornithischian bones show
considerable corrosive effects and the near total loss of smooth
periosteal bone. The marked disparity in digestion (corrosion)
between remains indicates a hiatus between the consumption of
the ornithischian and subsequent feeding on confuciusornithines. A wide variety of prey items have been identified within the
abdominal cavities of compsognathids. Fish and lepidosaurian
reptiles were identified within the exceptionally well-preserved
digestive tract of Scipionyx samniticus [33], the remains of a
lepidosaur (Bavarisaurus cf. macrodactylus) were found within the
holotype of Compsognathus longipes [8], and bones of an unidentified
small mammal within the holotype of Sinosauropteryx [10,15]. A
second Sinosauropteryx specimen (GMV 2124) preserves the jaws of
triconodont (Sinobaatar) and symmetrodont (Zhangheotherium) mam-
mals [34]. Miscellaneous, partially-digested bones were also
observed within the holotype specimen of Huaxiagnathus [13]. Based on our identification, the Sinornithosaurus limb in Sinocalliop-
teryx (CAGS-IG-T1) corresponds to an individual that can be
estimated at roughly one meter in total length [29]. If the
Sinornithosaurus was predated upon (rather than scavenged), this
would imply Sinocalliopteryx was capable of tackling carnivorous
prey more than a third its own size. The addition of at least two
confuciusornithines and an unidentified ornithischian within the
abdominal cavity of CAGS-IG-T1 demonstrates a diverse diet in
Sinocalliopteryx. Among the abdominal contents, several confuciusornithine
skeletal elements are notably absent (e.g. skull, ribs, vertebrae,
synsacrum, tarsometatarsus). It is unclear if these missing body
parts were: 1) never consumed by the Sinocalliopteryx; 2) were
consumed but were then preferentially dissolved/digested/egested;
3) consumed and preserved but are obscured by matrix and other
elements; or 4) are preserved in another block that was not
recovered. Barring further preparation and the successful recovery
of additional components of the specimen, these competing
explanations remain untestable. Inferences about the Digestive System of Sinocalliopteryx Inferences about the Digestive System of Sinocalliopteryx
Information regarding the organs and internal anatomy of
dinosaurs is exceptionally rare. Undoubtedly the best example of
preserved internal anatomy is that of the juvenile compsognathid,
Scipionyx samniticus (SBA-SA 163760), which preserves vestiges of
many of the major organs in exquisite detail [33,35]. In addition, CAGS-IG-T1 possesses abdominal contents in different stages
of digestion. The remains of the confuciusornithines, although PLOS ONE | www.plosone.org August 2012 | Volume 7 | Issue 8 | e44012 7 Abdominal Contents from Chinese Compsognathids Figure 7. Close up of Sinornithosaurus right hindlimb within the abdominal cavity of Sinocalliopteryx (JMP-V-05-8-01). Photograph and
interpretive illustration. Gastralia (Gs) and dorsal rib (Dr) belong to Sinocalliopteryx. Note the similar lengths of metatarsals III and IV. Abbreviations: Fi,
fibula; Mt, metatarsal; T, tibia. doi:10.1371/journal.pone.0044012.g007 Figure 7. Close up of Sinornithosaurus right hindlimb within the abdominal cavity of Sinocalliopteryx (JMP-V-05-8-01). Photograph and
interpretive illustration. Gastralia (Gs) and dorsal rib (Dr) belong to Sinocalliopteryx. Note the similar lengths of metatarsals III and IV. Abbreviations: Fi,
fibula; Mt, metatarsal; T, tibia. doi:10.1371/journal.pone.0044012.g007 There is evidence that crocodilians can increase secretion of
stomach acids by shunting deoxygenated blood to the stomach (by
increasing levels of PCO2; [37]), giving them the most acidic
foregut yet measured in any animal. Gastric pH may drop as low
as 1.2 in crocodilians [38], whereas it is generally always above 2.6
in birds [39]. The increase in acidity in crocodilians may also be
an adaptation to deal with large meals (A. mississippiensis will
voluntarily consume 23% of its body mass at one time; [40]). Because of low acidity, most modern birds are unable to digest
bone and instead will compact and orally egest this indigestible
material [39]. Given the presence of acid-etched bones in the gut
of Sinocalliopteryx and Scipionyx, as well as the presence of undigested
bone and muscle fibers in theropod coprolites [41,42], it is known
that at least some carnivorous dinosaurs possessed highly acidic
foreguts conducive to digestive processing of bone [35]. Preserved
theropod feces from Late Cretaceous tyrannosaurids retain
modified bone fragments [41,42], implying that some ingested
bone was not regurgitated in at least some non-avian theropods. However, undigested muscle tissue from a tyrannosaurid coprolite
[42] suggests that some non-avian theropod digestive tracts were
not as destructive as those of extant crocodilians. Therefore,
modern crocodilians do not necessarily provide ideal analogues for
Sinocalliopteryx digestion. Inferences about the Digestive System of Sinocalliopteryx Based on the digestive efficacy of Alligator
mississippiensis [37], a predicted minimum gastric residence time of
13 days would be required to reach the level of corrosion observed
in the presumed ornithischian bones. By comparison, the gastric
residence time for birds is generally less than 12 hours [36]. remnants of the articular cartilages, ligaments, and muscle tissues
are also preserved providing unsurpassed insight into the soft tissue
anatomy of a theropod [33]. Moreover, as a compsognathid,
Scipionyx serves as a useful model for interpreting the abdominal
contents and the presumed digestive tract of Sinocalliopteryx. The C-shaped abdominal contents in JMP-V-05-8-01 appear
reflective of the original contour of the digestive tract [33,35]. Furthermore, the contents become smaller and less identifiable
along the length of the inferred gut, presumably as a result of more
advanced digestion. The largest mass within the abdomen of JMP-
V-05-8-01 contains discernible feather-like structures and the
partial leg of an ingested dromaeosaurid. These remains most
likely represent a cololite that delimits the stomach. Further along
the length of the C-shaped digestive tract, the two smaller food
masses are composed of amorphous material suggestive of longer
residence times within the digestive tract. Their proximity to the
stomach suggests they may have been contained within the
duodenal loop, which is distinct in Scipionyx [33], and modern birds
[36]. In Scipionyx, the anterior part of the descending loop of the
duodenum (i.e. behind the pyloric sphincter) is dorsoventrally
oriented. Further along its length, the duodenum turns posteriorly,
becoming parallel with the gastral basket in precisely the same way
as the abdominal contents of JMP-V-05-8-01. The duodenum of
Scipionyx also contains incompletely digested elements (lizard-like
squamae and a possible fish vertebra [33]), which is consistent with
the progression of ingested remains in Sinocalliopteryx. August 2012 | Volume 7 | Issue 8 | e44012 August 2012 | Volume 7 | Issue 8 | e44012 PLOS ONE | www.plosone.org PLOS ONE | www.plosone.org 8 Abdominal Contents from Chinese Compsognathids Figure 8. Sinocalliopteryx as a stealth hunter feeding on the dromaeosaur Sinornithosaurus (left) and the primitive bird Confuciusornis
(right). Illustration by Cheung Chungtat. doi:10.1371/journal.pone.0044012.g008 Figure 8. Sinocalliopteryx as a stealth hunter feeding on the dromaeosaur Sinornithosaurus (left) and the primitive bird Confuciusornis
(right). Illustration by Cheung Chungtat. doi:10.1371/journal.pone.0044012.g008 Geo-gastroliths are swallowed sediment particles such as pebbles
and grit irrespective of function or deliberate/accidental origin
[43]. Inferences about the Digestive System of Sinocalliopteryx Such stones are known from a wide variety of theropods
including Allosaurus [44], Baryonyx [45], Caudipteryx [46], Lourinha-
nosaurus [47], Nqwebasaurus [48], Sinornithomimus [49], Sinosauropteryx
[50], Syntarsus [51], and possibly Tarbosaurus [52]. In a recent
review of geo-gastrolith function, Wings [43] found aid in
digestion (trituration, food mixing, stomach cleaning, and mineral
supplement) as the most plausible reason for the deliberate
ingestion of stomach stones. However, accidental ingestion (e.g. consumption of gastrolith-containing prey) was found to be a
major factor in extant carnivores, including crocodilians. The
apparent absence of geo-gastroliths in CAGS-IG-T1 suggests such
stones were not a critical part of Sinocalliopteryx digestion. In fact,
Wings [43] argued that low numbers of stomach stones, such as
those found in Allosaurus [43] and Baryonyx [45] are likely the result
of accidental ingestion. Discrete accumulations of dozens or
hundreds of stones in many individuals, such as those in the
ornithomimid Sinornithomimus [49], are almost certainly digestion
aids. It is therefore likely that the few stones found in the holotype
of Sinocalliopteryx were a result of accidental ingestion. It is notable
that geo-gastroliths in Sinocalliopteryx occur in the posterior
abdomen rather than the stomach where they occur in extant
crocodilians [38]. In crocodilians, a particularly strong pyloric
sphincter prevents the passage of geo-gastroliths into the midgut
[38,53]. In birds, geo-gastroliths are held within the muscular
ventriculus, or gizzard, which functions as the primary trituration
site [53]. In Sinocalliopteryx, the association between geo-gastroliths
and the highly processed food masses in the midgut region negate
the possibility of a gizzard. Had the animal lived, it is likely that
these stones would have been passed in the faeces. August 2012 | Volume 7 | Issue 8 | e44012 References Memorie della Societa` Italiana de Scienze Naturali e del Museo Civico di Storia
Naturale di Milano 37: 1–281. 1. Chen P, Wang Q, Zhang H, Cao M, Li W, et al. (2005) Jianshangou Bed of the
Yixian Formation in west Liaoning, China. Sci China Ser D - Earth Sci 48: 298–
312. 18. Martin LD, Zhou Z (1998) Confuciusornis sanctus compared to Archaeopteryx
lithographica. Naturwissenchaften 85: 286–289. 2. Fu¨rsich FT, Jingeng S, Baoyu J, Yanhong P (2007) High resolution
palaeoecological and taphonomic analysis of Early Cretaceous lake biota,
western Liaoning (NE-China). Pal Pal Pal 253: 434–457. 19. Chiappe LM, Ji S, Ji Q, Norell MA (1999) Anatomy and systematics of the
Confuciusornithidae (Aves) from the late Mesozoic of northeastern China. Bull
Am Mus Nat Hist 242: 1–89. g (
)
3. Li Q, Gao K-Q, Vinther J, Shawkey MD, Clarke JA, et al. (2010) Plumage color
patterns of an extinct dinosaur. Science 327: 1369–1372. 20. Norell MA, Makovicky PJ (1999) Important features of the dromaeosaurid
skeleton II: information from newly collected specimens of Velociraptor mongoliensis. Am Mus Nov 3282: 1–48. 4. Ji S, Ji Q, Lu J, Yuan C (2007) A new giant compsognathid dinosaur with long
filamentous integuments from Lower Cretaceous of Northeastern China. Acta
Geol Sin 81: 8–15. 21. Zhang FC, Zhou ZH, Benton MJ (2008) A primitive confuciusornithid bird from
China and its implications for early avian flight. Sci China Ser D - Earth Sci 51:
625–639. 5. Swisher CC, Wang XL, Zhou ZZ, Wang YQ, Jin F, et al. (2002) Further support
for a Cretaceous age for the feathered-dinosaur beds of Liaoning, China: New
40Ar/39Ar dating of the Yixian and Tuchengzi formations. Chinese Sci Bull 47:
135–138. 22. Hou L-H, Zhao Z-H, Martin LD, Feduccia A (1995) A beaked bird from the
Jurassic of China. Nature 377: 616–618. 6. Swisher CC, Wang YQ, Wang XL, Xu X, Wang Y (1999) Cretaceous age for
the feathered dinosaurs of Liaoning, China. Nature 400: 58–61. J
23. Ji Q, Chiappe LM, Ji S (1999) A new late Mesozoic confuciusornithid bird from
China. J Vert Pal 19: 1–7. the feathered dinosaurs of Liaoning, China. Nature 400: 58–61. 7. Witmer LM (1997) The evolution of the antorbital cavity of Archosaurs: a study
in soft-tissue reconstruction in the fossil record with an analysis of the function of
pneumaticity. J Vert Pal Mem 3: 1–76. 24. Acknowledgments The authors thank Fucheng Zhang and Xing Xu (The Institute of
Vertebrate Paleontology and Paleoanthropology, Chinese Academy of
Sciences, China) for critical comments and suggestions on this paper. Thanks to Cheung Chuntat for the exceptional illustrations in figure 8. All
other illustrations by PRB. Detailed revisions by Cristiano Dal Sasso, an
anonymous reviewer, and the handling editor Andrew Farke greatly
improved the final version and we are appreciative of their efforts. Active hunting of flight-capable prey by a land-bound predator
may seem intrinsically implausible, but there are abundant extant
examples, wild felids among the most famous. The back-footed cat
(Felis nigripes) of southern Africa routinely ambushes and chases
down cursorial birds before they are able to become airborne [56]. Servals (Leptailurus serval) are long-legged and adept at pouncing on
alighted birds, and at snagging fleeing birds midair [57–59]. Avian
prey is known to constitute nearly half the diet of some leopard
cats (Prionailurus bengalensis) [60], which both climb trees to prey on
roosting birds and ambush foraging birds on the ground. Among
canids, foxes are expert bird hunters, commonly taking anser- Author Contributions Analyzed the data: LX PB WSP MB TM PC SJ QJ. Wrote the paper: LX
PB WSP TM MB. Abdominal Contents from Chinese Compsognathids O’Connor et al. [54] reported on a specimen of Microraptor with
the remains of an enantiornithine bird within its abdominal cavity,
and argued that such presumed predation on a bird with clear
arboreal perching adaptations was evidence supporting a highly
arboreal/aerial lifestyle in Microraptor. Based on various other lines
of evidence, we agree with this ultimate conclusion; however, that
Jehol birds were evidently on the menu of Sinocalliopteryx must be
regarded as a strong contradiction to the necessity of O’Connor
et al’s [54] ecological inference. Confuciusornis was not as well
adapted to perching as enantiornithine birds, but does nonetheless
possess long curved pedal claws and a posteriorly-facing hallux,
and was capable of powered flight. While Sinocalliopteryx does have
proportionately longer arms than most compsognathids and may
have been capable of tree climbing, it lacks any definitive arboreal
adaptations; at over two meters in length, is best regarded as a
predominantly terrestrial animal. iforme, galliforme, and passeriforme game [61,62]. Among extant
reptiles, monitor lizards and various snakes consume birds in both
arboreal and terrestrial contexts [63–66]. In a majority of these examples, what is required to successfully
apprehend avian prey is not climbing prowess, but stealth, such
that the predator can reach its striking distance before the prey
takes flight. It should be remembered that Confuciusornis and other
Jehol birds were not as well adapted for flight as most modern
aves, and, therefore, likely required greater time to mount an
aerial takeoff and escape. Nevertheless, the evidence of bird
predation in Sinocalliopteryx suggests that it was a highly capable
stealth hunter (Figure 8). References Maruga´n-Lobo´n J, Chiappe LM, Ji S, Zhao Z, Chunling G, et al. (2011)
Quantitative patterns of morphological variation in the appendicular skeleton of
the Early Cretaceous bird Confuciusornis. J Syst Pal 9: 91–101. the Early Cretaceous bird Confuciusornis. J Syst Pal 9: 91–101. 8. Ostrom JH (1978) The osteology of Compsognathus longipes Wagner. Zitteliana 4:
73–118. 25. Chiappe LM, Maruga´n-Lobo´n J, Ji S, Zhou Z (2008) Life history of a basal bird:
morphometrics of the Early Cretaceous Confuciusornis. Biol Letters 4: 719–723. 9. Chiappe LM, Go¨lich U (2010) Anatomy of Juravenator starki (Theropoda:
Coelurosauria) from the Late Jurassic of Germany. Neues Jahrbuch fu¨r Geologie
und Pala¨ontologie Abhandlungen 258: 257–296. 26. Xu X, Wang Y (1998) New Psittacosaurus occurrences in the Early Cretaceous
Yixian Formation (Liaoning, China) and its stratigraphic significance. Vert
Palasiatica 36: 147–158. 10. Currie PJ, Chen PJ (2001) Anatomy of Sinosauropteryx prima from Liaoning,
northeastern China. Can J Earth Sci 38: 1705–1727. 27. Zheng W, Jin X, Shibata M, Azuma Y, Yu F (2012) A new ornithischian
dinosaur from the Cretaceous Liangtoutanf Formation of Tiantai, Zhejiang
Province, China. Cret Res 34: 208–219. 11. de Klerk WJ, Forster CA, Sampson SD, Chinsamy A, Ross CF (2000). A new
coelurosaurian dinosaur from the Early Cretaceous of South Africa. J Vert Pal
20: 324–332. 28. Xu X (2001) Branched integumental structures in Sinornithosaurus and the origin
of feathers. Nature 410: 200–204. 12. Holtz TR, Molnar RE, Currie PJ (2004) Basal Tetanurae. In D. Weishampel, P. Dodson, and H. Osmo´lska (eds.) The Dinosauria (2nd edition). The University of
California Press, Berkeley. 71–110. 29. Xu X, Wang X-L (2000) Troodontid-like pes in the dromaeosaurid
Sinornithosaurus. Paleont Soc Korea Sp Publ 4: 179–188. 30. Xu X, Norell MA (2006) Non-avian dinosaur fossils from the Lower Cretaceous
Jehol Group of western Liaoning, China. Geol J 41: 419–437. y
13. Hwang SH, Norell MA, Ji Q, Gao K-Q (2004) A large compsognathid from the
Early Cretaceous Yixian Formation of China. J Syst Pal 2: 13–30. 31. Zheng X, Xu X, You H, Zhao Q, Dong Z (2010) A short-armed dromaeosaurid
from the Jehol Group China with implications for early dromaeosaurid
evolution. Proc Roy Soc B 22: 211–217. 14. Peyer K (2006) A reconsideration of Compsognathus from the upper Tithonian of
Canjuers, southeastern France. J Vert Pal 26: 879–896. 14. Peyer K (2006) A reconsideration of Compsognathus from th 15. Predation on Flying Prey and Ecological Implications Remains as delicate as small bird bones have presumably short
digestion periods, and the multiple Confuciusornis within the
abdominal cavity of CAGS-IG-T1 must have been consumed in
fairly rapid succession, in order for the first individual not to have
had time to be digested noticeably beyond that of the second. Moreover, levels of corrosion on all the confuciusornithine
elements are similar on a macroscopic level, again suggesting that
the birds were consumed in rapid succession. Such short durations
between meals provides anecdotal evidence for high metabolic rate
in Sinocalliopteryx. In both CAGS-IG-T1 and JMP-V-05-8-01, scavenging cannot
be definitively ruled out as an alternative to predation. However,
as argued by O’Connor et al. [54], a high degree of articulation
among gut contents shows that, when ingested, the carcasses were
at least fresh enough not to have disarticulated. The association of
two or more birds is perhaps more easily explained by selective
hunting than by the chance discovery of multiple C. sanctus
carcasses; however, this is speculative. In the case of CAGS-IG-
T1, it is improbable that every individual organism represented
within the gut contents was consumed exclusively as a result of
scavenging, as true obligate tetrapod scavengers are rare [55]. The presence of at least two confuciusornithine birds within the
abdominal cavity of Sinocalliopteryx (CAGS-IG-T1) argues against
circumstantial consumption (i.e. the coincidental scavenging of
two or more carcasses of the same species), and suggests a
behavioral proficiency for predating on flying prey. It is not known
if the dromaeosaurid Sinornithosaurus possessed elongate hind and
forelimb feathers, as in the closely related Microraptor. If it did, the
Sinornithosaurus remains within JMP-V-05-8-01 may constitute an
additional example of a flight-capable maniraptoran eaten by a
Sinocalliopteryx. August 2012 | Volume 7 | Issue 8 | e44012 PLOS ONE | www.plosone.org PLOS ONE | www.plosone.org 9 Abdominal Contents from Chinese Compsognathids Abdominal Contents from Chinese Compsognathids Acta Pal Pol 52: 1–16. 43. Wings O (2007) A review of gastrolith function with implica 64. Losos JB, Greene HW (1988) Ecological and evolutionary implications of diet in
monitor lizards. Biol J Linn Soc 35: 379–407. vertebrates and a revised classification. Acta Pal Pol 52: 1–16. (2000) The Howe Ranch dinosaurs. Aathal: Sauriermuseum Aathal. 96 44. Ayer J (2000) The Howe Ranch dinosaurs. Aathal: Sauriermuseum J
65. Rodrı´guez-Robles JA, Bell CJ, Greene HW (1999) Gape size and evolution of
diet in snakes: feeding ecology of erycine boas. J Zool 248: 49–58. 45. Charig AJ, Milner AC (1997) Baryonyx walkeri, a fish-eating dinosaur from the
Wealden of Surrey. Bull Nat Hist Mus Geol Ser 53: 11–70. Wealden of Surrey. Bull Nat Hist Mus Geol Ser 53: 11–70. y
46. Ji Q, Currie PJ, Norell MA, Ji S-A (1998) Two feathered dinosaurs from
northeastern China. Nature 393: 753–761. g
gy
y
J
66. Slip DJ, Shine R (1988) Feeding habits of the diamond python, Morelia s. spilota:
ambush predation by a boid snake. J Herp 22: 323–330. 47. Mateus I, Mateus H, Antunes MT, Mateus O, Taquet P, et al. (1998) Upper
Jurassic theropod dinosaur embryos from Lourinha˜, Portugal. Mem Acad
Cie´ncias Lisboa 37: 101–109. 67. Hu YM, Meng J, Wang YQ (2005) Large Mesozoic mammals fed on young
dinosaurs. Nature 433: 149–152. 68. Dalsa¨tt J, Zhou ZH, Zhang FC (2006) Food remains in Confuciusornis sanctus
suggest a fish diet. Naturwissenschaften 93: 444–446. 48. De Klerk WJ, Forster CA, Sampson SD, Chinsamy A, Ross CF (2000) A new
coelurosaurian dinosaur from the Early Cretaceous of South Africa. J Vert Pal
20: 324–332. 69. Zheng X, Martin LD, Zhou Z, Burnham DA, Zhang F, et al (2011) Fossil
evidence of avian crops from the Early Cretaceous of China. Proc Nat Acad Sci
108: 15904–15907. 49. Kobayashi Y, Lu¨ J-C (2003) A new ornithomimid dinosaur with gregarious
habits from the Late Cretaceous of China. Acta Pal Pol 48 : 235–259. 70. Zhou ZH, Zhang FC (2002) A long-tailed, seed-eating bird from the Early
Cretaceous of China. Nature 418: 405–409. 50. Dong Z, Chen P (2000) A tiny fossil lizard in the stomach content of the
feathered dinosaur Sinosauropteryx from northeastern China. Vert PalAsiatica 38
(supplement): 10. 71. Zhou ZH, Zhang FC, Li Z H (2009) A new basal ornithurine bird (Jianchangornis
microdonta gen. et sp. nov.) from the Lower Cretaceous of China. Abdominal Contents from Chinese Compsognathids 34. Hurum JH, Luo Z2X, Kielan2Jaworowska Z (2006) Were mammals originally
venomous? Acta Pal Pol 51: 1–11. 54. O’Connor J, Zhou Z, Xu X (2011). Additional specimen of Microraptor provides
unique evidence of dinosaurs preying on birds. PNAS doi: 10.1073/
pnas.1117727108. 35. Dal Sasso C, Signore M (1998) Exceptional soft tissue preservation in a theropod
dinosaur from Italy. Nature 392: 383–387. p
55. DeVault TL, Rhodes OE, Shivik JA (2003) Scavenging by vertebrates:
behavioral, ecological, and evolutionary perspectives on an important energy
transfer pathway. Oikos 102: 225–234. y
36. Whittow GC (2000) Sturkie’s avian physiology. 5th ed. Academic Press: San
Diego. 704 p. p
y
56. Smithers RHN (1971) Mammals of Botswana. Mem Nat Mus Rhod 4: 1–340. 37. Farmer CG, Uriona TJ, Olsen DB, Steenblik M, Sanders K (2008) The right-to-
left shunt of crocodilians serves digestion. Phys Biochem Zool 81: 125–137. 57. Kingdon JS (1977) East African mammals, Vol. 3A (carnivores). London and
New York: Academic Press. 38. Huchzermeyer FW (2003) Crocodiles: Biology, husbandry and diseases. Cambridge: CAB International. 352 p. h 58. Sunquist M, Sunquist F (2002) Wild cats of the world. Chicago: University of
Chicago Press. 145 p. g
p
39. Sturkie PD (2000) Avian physiology, 5th Ed. Ithaca: Cornell University Press. 685 p. 59. Hunter L (2005) Cats of Africa: behaviour, ecology, and conservation. New
Holland Publishers. 76 p. 40. Uriona TJ, Farmer CG (2006) Contribution of the diaphragmaticus muscle to
vital capacity in post-prandial American alligators (Alligator mississippiensis). J Exp
Biol 208: 3047–3053. 60. Sunquist M, Sunquist F (2002) Wild cats of the world. Chicago: University of
Chicago Press. 227 p. 41. Chin K, Tokaryk TT, Erickson GM, Calk LC (1998) A king-sized theropod
coprolite. Nature 393: 680–682. 61. Richards DF (1977) Observations on the diet of the Red fox (Vulpes vulpes) in
South Devon. J Zool 183: 495–504. 62. Sargeant AB, Allen SH, Eberhardt RT (1984) Red fox predation on breeding
ducks in midcontinent North America. Wildlife Monographs 89: 3–41. 42. Chin K, Eberth DA, Schweitzer MH, Rando TA, Sloboda WJ, et al. (2003)
Remarkable preservation of undigested muscle tissue within a Late Cretaceous
tyrannosaurid coprolite from Alberta, Canada. Palaios 18: 286–294. g
p
63. Vestjens WJM (1976) Reptilian predation on birds and eggs at Lake Cowal,
NSW. Emu 77: 36–37. y
p
43. Wings O (2007) A review of gastrolith function with implications for fossil
vertebrates and a revised classification. References Chen P-J, Dong D-M, Zheng SN (1998) An exceptionally well-preserved
theropod dinosaur from the Yixian Formation of China. Nature 391: 147–152. 32. Xu X, Wang X-L (2004) A new dromaeosaur (Dinosauria: Theropoda) from the
Early Cretaceous Yixian Formation of western Liaoning. Vert PalAsiatica 42:
111–119. 16. Gishlick AD, Gauthier JA (2007) On the manual morphology of Compsognathus
longipes and its bearing on the diagnosis of Compsognathidae. Zool J Linn Soc
149: 569–581. 33. Dal Sasso C, Maganuco S (2011) Scipionyx samniticus (Theropoda: Compsog-
nathidae) from the Lower Cretaceous of Italy; osteology, ontogenetic assessment,
phylogeny, soft tissue anatomy taphonomy and palaeobiology. Memorie 37: 1–
281. 17. Dal Sasso C, Maganuco S (2011) Scipionyx samniticus (Theropoda: Compsog-
nathidae) from the Lower Cretaceous of Italy. Osteology, ontogenetic
assessment, phylogeny, soft tissue anatomy, taphonomy and palaeobiology. August 2012 | Volume 7 | Issue 8 | e44012 August 2012 | Volume 7 | Issue 8 | e44012 PLOS ONE | www.plosone.org 10 Abdominal Contents from Chinese Compsognathids Abdominal Contents from Chinese Compsognathids Vert PalAsiatica
47: 299–310. (
pp
)
51. Whittle CH, Everhart MJ (2000) Apparent and implied evolutionary trends in
lithophagic vertebrates from New Mexico and elsewhere. New Mex Mus Nat
Hist Sci Bull 17: 75–82. 72. Yuan C (2004) Further study of Yanornis martini (Ornithurae) from the Mesozoic
Jehol Biota in western Liaoning, China. Acta Geol Sin 78: 464–467. hol Biota in western Liaoning, China. Acta Geol Sin 78: 464–467. 52. Suzuki S, Watabe M (2000) Report on the Japan-Mongolia joint paleontological
expedition to the Gobi Desert, 1998. Hayashibara Mus Nat Sci Res Bull 1: 83–
98. 73. Zhou ZH, Clarke J, Zhang FC (2004). Gastroliths in Yanornis: An indicaiton of
the earliest radical diet-switching and gizzard plasticity in the lineage leading to
living birds? Naturwissenschaften 91: 571–574. 53. Stevens CE, Hume ID (1995) Comparative physiology of the vertebrate digestive
system. Cambridge: Cambridge University Press. 420 p. 74. Zhou ZH, Wang Y. (2010). Vertebrate diversity of the Jehol Biota as compared
with other lagersta¨tten. Sci China Earth Sci, 53: 1894–1907. August 2012 | Volume 7 | Issue 8 | e44012 PLOS ONE | www.plosone.org 11
|
https://openalex.org/W2591972960
|
http://www.scielo.br/pdf/rep/v37n1/1809-4538-rep-37-01-00108.pdf
|
English
| null |
Capital account regulation in Brazil: An assessment of the 2009-2013 period
|
Brazilian Journal of Political Economy
| 2,017
|
cc-by
| 10,613
|
Brazilian Journal of Political Economy, vol. 37, nº 1 (146), pp. 108-129 January-March/2017 Brazilian Journal of Political Economy, vol. 37, nº 1 (146), pp. 108-129 January-March/2017 * Professor of Economics da Universidade Estadual de Campinas (UNICAMP) and CNPq researcher.
Email: prates@unicamp.br; Professor of Economics da Universidade do Estado do Rio de Janeiro (UERJ)
and CNPq researcher. Email: luizfpaula@terra.com.br. Submitted: 6/August/2015; Approved: 15/
March/2016. 1* We are grateful for the research assistance of Aline Gomes and the comments of the anonymous
referee. All remaining errors are the authors’ responsibility.
* Professor of Economics da Universidade Estadual de Campinas (UNICAMP) and CNPq researcher.
Email: prates@unicamp.br; Professor of Economics da Universidade do Estado do Rio de Janeiro (UERJ)
and CNPq researcher. Email: luizfpaula@terra.com.br. Submitted: 6/August/2015; Approved: 15/
March/2016. Capital account regulation in Brazil:
An assessment of the 2009-2013 period1*
Regulação dos fluxos de capitais no Brasil:
Uma avaliação do período 2009-2013 Daniela Magalhães Prates
Luiz Fernando de Paula* 1* We are grateful for the research assistance of Aline Gomes and the comments of the anonymous
referee. All remaining errors are the authors’ responsibility.
* P
f
f E
i
d
U i
id d E
d
l d C
i
(UNICAMP)
d CNP
h 108 • Revista de Economia Política 37 (1), 2017 Introduction Introduction A new wave of capital flows to emerging economies, boosted by the post-cri-
sis circumstances (quantitative easing policies and historical low interest rates in
advanced countries, etc.), took place from the 2nd quarter of 2009 to the first
quarter of 2013. In May 2013, when the Federal Reserve (Fed) merely indicated
that it might begin tapering this policy toward the end of the calendar year, global
investors set into motion a portfolio adjustment that caused a temporary but sig-
nificant reversal in capital flows to U.S., putting upward pressures on the exchange
rates of many emerging economies (BIS, 2014). During the boom phase of the recent cycle, once more, “emerging-market as-
sets” became objects of desire on the part of global investors, resuming policy di-
lemmas to emerging countries stemming from the combination of high growth
rates, accelerating inflation (associated with the post-crisis commodity prices
boom), excessive currency appreciation and/or asset price overshooting. In order to deal with these policy dilemmas, some of these countries resorted
to capital account regulation (CAR), which encompasses capital controls and fi-
nancial prudential regulation (Gallagher et al, 2012). This regulation aimed at
halting the trend of currency appreciation and of speculative bubbles in asset pric-
es as well as at reducing the risks of rising current account deficits and the costs of
excessive Foreign Exchange (FX) reserve accumulation. Thus, unlike the pre-crisis
context, most of these countries did not adopt a hands-off approach to capital
inflows. Brazil was one of the emerging countries that had a stronger trend of currency
appreciation until July 2011 due to the combination of huge capital inflows, in-
creasing commodity prices, high domestic interest rate, and a sophisticated and
deep FX derivatives market completely open to foreign investors that provides
room for speculation on the exchange rate through operations with FX derivatives
(the so-called derivatives carry-trade). Under this context that can be understood
the recent implementation of the CAR, which was complemented with another
kind of regulation, called here FX Derivatives Regulation (FXDR), whose target is
FX derivatives operations of all agents, being them financial and non-financial as
well as non-residents and residents. As the experience of implementation of capital account regulation (CAR) in
2009-2013 in Brazil is still recent there are few empirical works that assessed its
effectiveness. Revista de Economia Política 37 (1), 2017 • pp. 108-128 Daniela Magalhães Prates
Luiz Fernando de Paula* resumo: O Brasil foi um dos países emergentes que enfrentou fortes pressões de
apreciação cambial entre o 2º trimestre de 2009 e julho de 2011. É sob este contexto que
pode ser entendida a aplicação da Regulação da Conta Capital (CAR) depois de 2009, que
foi complementada com outro tipo de regulação, a Regulação dos Derivativos Cambiais
(FXDR). Este trabalho mostra que, somente quando o governo brasileiro adotou estes dois
tipos de regulação simultaneamente, houve um aumento da eficácia das políticas em deter
essas pressões. A experiência brasileira também revela que não é possível estabelecer uma
hierarquia entre os instrumentos temporários para gerenciar fluxos de capital e medidas
prudenciais permanentes, tal como defende a abordagem atual do FMI. Palavras-chave: liberalização da conta de capital; controles de capitais; economia
brasileira. abstract: Brazil was one of the emerging countries that had a stronger trend of currency
appreciation from the 2nd quarter of 2009 to July 2011. Under this context that can be
understood the implementation of capital account regulation (CAR) after 2009, which was
complemented with another kind of regulation, the so-called FX Derivatives Regulation
(FXDR). This paper shows that only when Brazilian government adopted these two kinds
of regulations simultaneously, the policy effectiveness increased in terms of protecting the
Brazilian currency from upward pressures. Brazilian experience also highlights that it is not
possible to establish a hierarchy between temporary instruments to manage capital flows
and permanent prudential measures, as supported by the IMF current approach. Keywords: capital account liberalization; capital controls; Brazilian economy. JEL classification: F31; F32; F53. 1* We are grateful for the research assistance of Aline Gomes and the comments of the anonymous
referee. All remaining errors are the authors’ responsibility. * Professor of Economics da Universidade Estadual de Campinas (UNICAMP) and CNPq researcher. Email: prates@unicamp.br; Professor of Economics da Universidade do Estado do Rio de Janeiro (UERJ)
and CNPq researcher. Email: luizfpaula@terra.com.br. Submitted: 6/August/2015; Approved: 15/
March/2016. http://dx.doi.org/10.1590/0101-31572016v37n01a06 108 • Revista de Economia Política 37 (1), 2017 Introduction Using a GARCH regression, Baumann and Gallagher (2012) found
that the introduction of CAR in Brazil between October 2009 and December 2012
had a small but significant impact on shifting the composition of capital inflows
toward longer-term investment, on curbing the level and volatility of exchange
rate, and on modestly increasing Brazilian monetary policy autonomy. Another
empirical study (Chamon and Garcia, 2016) analyzed the impact of the capital
controls adopted in Brazil from late 2009 until 2012, and concluded that while Revista de Economia Política 37 (1), 2017 • pp. 108-128 109 the first several measures (until mid-2011) had very limited success in containing
the currency appreciation, the Brazilian exchange rate seems to respond strongly
just after the adoption of a financial tax on the notional amount of derivatives. On
the other hand, Klein (2012), by analyzing the pattern of controls on capital in-
flows in a set of 44 advanced and emerging economies (including Brazil), con-
cluded that the Brazilian tax on financial operations (IOF) was an episodic control
on the capital inflows that did not temper the appreciation of the Brazilian cur-
rency; however, the period covered in his study ended in 2010, therefore before the
adoption of a broader FXDR. Differently from these empirical assessments based on econometric methods,
this paper applies qualitative methods and descriptive statistics to analyze the
measures of CAR and FXDR implemented in Brazil over 2009-2013. This ap-
proach was chosen due to the overlap and interdependency of factors affecting the
Brazilian exchange rate path in this period: i) external shocks with high frequency,
given the unstable global environment and the even high volatility of international
capital flows after the global financial crisis; ii) macroeconomic policy shifts, espe-
cially in monetary and foreign exchange policies; iii) frequent changes in regula-
tion at the domestic level. In line with the empirical findings of Baumann and Gallagher (2012) and
Chamon and Garcia (2016), our assessment suggest that the regulations adopted
in Brazil in the aforementioned period were effective in affecting the exchange rate
path, yet only when the three kinds of measures (capital controls, prudential finan-
cial regulation and FXDR) were simultaneously in place the policy effectiveness
increased in terms of protecting the Brazilian currency from upward pressures. 2 This augmentation is related to the notion that capital flows from capital-rich countries to the capital-
poor countries due to a comparatively higher marginal productivity capital in the former ones. International financial integration, economic performance and macroeconomic
instability Prasad et al. (2003) sum up the conventional view that gives support to inter-
national financial integration, pointing out that the potential benefits of financial
liberalization for emerging market countries can be divided in two channels: direct
and indirect. Direct channels include augmentation of domestic savings2, reduc-
tion in the cost of capital due to better global allocation of risk, reduction of con-
sumption volatility, transfer of technological and managerial know-how, and stim-
ulation of domestic financial sector development. Indirect channels include
promotion of specialization, commitment to better economic policies, and signal-
ing the practice of more friendly policies. Therefore, according to this view finan-
cial liberalization results in market discipline that shall stimulate more consistent
macroeconomic policy (understood as sound fiscal and monetary policies, etc.) as
market force (“rational foreign investors”) can penalize bad policies. A lot of empirical works, most of them using panel data and measuring the
international financial integration with the use of different de jure and de facto
indexes, seek to evaluate the relationship between capital account liberalization,
on one hand, and economic growth, financial crises and/or macroeconomic vola-
tility, on the other hand. Some surveys conclude that empirical evidences in gen-
eral do not present a robust relationship between financial liberalization and eco-
nomic growth (Prasad et al., 2003; Einchengreen, 2004, Ch 3). Some IMF’s economists works have acknowledged the potential risks and
costs related to international financial integration and specifically to the volatility
of capital flows in emerging economies (Prasad et al., 2003; Kose et al., 2006; IMF,
2008), as the surge of capital inflows can have negative effects on emerging econo-
mies, including the appreciation of the domestic currency beyond the equilibrium
level, fiscal costs of sterilization related to international reserves accumulation, in-
flationary pressures can result from incomplete sterilization and/or credit boom,
and possible bubbles in certain sectors as equity markets. However, IMF (2008)
sustains that financial globalization leads to better macroeconomic outcomes when
certain threshold conditions of financial and institutional developments are met
(financial market development, institutional quality, sound macroeconomic policies,
trade integration), but some analysts have argued that such conditions are almost
the same factors pointed out as collateral benefits of financial globalization, gener-
ating a logical contradiction between consequences and causes (Biancareli, 2008). Introduction Therefore, our main contribution is to show that there are important feedbacks
between capital controls, prudential financial regulation and FXDR, as much as
between these measures and macroeconomic policy, so that it is not possible to
establish a hierarchy between instruments to manage capital flows and curb their
undesirable outcomes (such as currency appreciation and financial fragility) as
supported by the current IMF approach. The arguments are organized in four sections, besides this introduction. Second section focuses on international financial integration, capital flows and
CAR in emerging economies, including a brief discussion on the new institutional
view of the International Monetary Fund (IMF) on capital controls. Third section
provides a brief overview of capital flows and capital account liberalization in
Brazil. Fourth section details the specificity of the Brazilian experience and the key
features of CAR and FXDR after the global financial crisis. Some lessons from the
Brazilian experience are presented in the final section. 110 Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 Revista de Economia Política 37 (1), 2017 • pp. 108-128 International financial integration, economic performance and macroeconomic
instability Yet, since the breakdown of the Bretton Woods system, some prominent main-
stream economists have stressed that with the financial liberalization and the Revista de Economia Política 37 (1), 2017 • pp. 108-128 111 emergence and spread of new financial instruments (such as derivatives), the likeli-
hood of occurrence of speculative financial operations increases substantially. Tobin (1978) was one of the first economists to state that the main macroeco-
nomic problem related to integrated financial markets is not the choice of the ap-
propriate exchange rate regime but the excessive short-run capital mobility that
reduces the autonomy of national governments to pursue domestic objectives with
respect to employment, output and inflation. Stepping forward Stiglitz (2000)
points out that capital flows to emerging countries are markedly pro-cyclical and
exacerbate economic booms, and that financial liberalization exposes countries to
the vicissitudes associated with changes in economic circumstances outside the
country; so that such economies are exposed to sudden change in lenders’ and in-
vestors’ perceptions. As the Post Keynesian literature (Schulmeister, 1988; Harvey, 2009) highlight-
ed, in this setting, featured by floating exchange rates and free capital mobility,
short-term capital flows constitute the chief determinant of nominal exchange
rates, which are highly volatile. In this perspective, the speculative feature of these
flows, subordinate to financial investors’ risk aversion/appetite, is the main cause
of the volatility of exchange rates. However, exchange rate volatility in general is
higher in emerging economies than in developed ones due to the monetary asym-
metry of the international monetary and financial system that refers to the hierar-
chical dimension of the international monetary system. In other words, national
currencies are hierarchically positioned according to their degree of liquidity,
which relates to their ability to perform internationally the three functions of mon-
ey: medium of exchange, unit of account and denomination of contracts, and store
of value (international reserve currency). The key currency (currently, the fiduciary
US dollar) is placed at the top of the hierarchy because it has the highest degree of
liquidity. The currencies issued by the other core countries are in intermediate po-
sition, as they have a high liquidity premium, but not as high as the dollar. At the
opposite end are the currencies issued by the emerging economies, which are non-
liquid currencies, for they are incapable of performing those functions. International financial integration, economic performance and macroeconomic
instability Consequently, these currencies, priced with a lower liquidity premium, are espe-
cially vulnerable to the inherent volatility of capital flows in the post-Bretton
Woods system. The smaller and less liquid foreign exchange and financial markets
of emerging economies make them more vulnerable to one-way expectations and
herd behavior (Andrade and Prates, 2013). Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 Policy space and capital account regulation In order to enhance the possibility of a successful management of exchange
rate regime in emerging economies some measures to reduce the volatility of cap-
ital flows and the likelihood of speculation attack against the domestic currency
can be necessary. One possibility is the use of official intervention in the foreign
exchange market, which may exert direct influence on nominal exchange rate as
it alters the relative supply of domestic and foreign currencies. On the one hand, 112 Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 the countries’ ability to resist currency depreciation is limited by its stock of for-
eign exchange reserves and its access to potential credit lines. Thus, reserve accu-
mulation can be seen as an insurance against future negative shocks and specula-
tion against the domestic currency, as emerging economies have limited access to
the international capital market during times of high risk aversion of foreign in-
vestors. On the other hand, the ability to avoid currency appreciation may require
the use of sterilized intervention. Monetary authorities have often sought to ster-
ilize the impact of foreign exchange intervention through open market operations
and other measures, such as increasing bank reserve requirements. Moreover,
sterilization often implies quasi-fiscal costs, as it in general involves the central
bank exchanging high-yield domestic assets for low-yield foreign reserves
(Cardarelli et al., 2009). Another possibility to enhance the management of the exchange rate regime
(that is not excluding official intervention in the currency markets) in emerging
economies is the use of ‘capital account regulation’ (CAR) (Gallagher et al., 2012)
or ‘capital management techniques’ which include capital controls, that is mea-
sures that manage volume, composition, and/or allocation of international private
capital flows, and/or ‘prudential domestic financial regulations’, which refer to
policies, such as capital-adequacy standards, reporting requirements, or restric-
tions on the ability and terms under which domestic financial institutions can
provide to certain types of projects (Epstein et al., 2003: p.6-7). Revista de Economia Política 37 (1), 2017 • pp. 108-128 Policy space and capital account regulation Capital controls
can be used for different though related objectives, such as: (i) to reduce the vul-
nerability of a country to financial crises, including capital flight during currency
crisis; (ii) to drive a wedge between onshore and offshore interest rates in order to
provide monetary authorities with some policy autonomy at least in the short-run;
and (iii) to maintain some short-term stability of nominal exchange rate and to
reduce exchange rate pressures derived from excessive capital inflows. Regarding the effectiveness of CAR, Magud and Reinhart (2006) review more
than 30 papers that evaluated capital controls either on inflows or outflows
around the world, making use of a capital controls effectiveness index in order to
standardize the results of the empirical studies. They conclude that “capital con-
trols on inflows seem to make monetary policy more independent; alter the com-
position of capital flow; reduce real exchange rate pressures (although the evi-
dence is more controversial)”, but “seem not to reduce the volume of net flows
(and hence, the current account balance)”, while “limiting private external bor-
rowing in the ‘good times’ plays an important prudential role because more often
than not countries that are ‘debt intolerant’” (Magud and Reinhart, 2006: 26-27). Finally, Magud and Reinhart (2006) argue that enhancing the effectiveness of cap-
ital controls is necessary to take into account country-specific characteristics in
their design. Concerned with the amount and volatility of capital flows to emerging econo-
mies after the contagious of the 2008 crisis, IMF revised its official position re-
garding the evaluation of capital controls (IMF, 2010, 2011, 2012b; Ostry et al. Revista de Economia Política 37 (1), 2017 • pp. 108-128 113 2010, 2011). This finally resulted in a new institutional view endorsed by the IMF
(IMF, 2012), especially with regards to the regulation of capital inflows. 2010, 2011). This finally resulted in a new institutional view endorsed by the IMF
(IMF, 2012), especially with regards to the regulation of capital inflows. In fact, the IMF in its definitive policy framework (IMF 2012) has made rel-
evant progress compared not only to its traditional rejection of capital controls
but also in comparison to its preliminary approaches of 2010 and 2011 to tolerate
capital controls under highly specific circumstances (Ostry et al. 2010 to IMF,
2011). Policy space and capital account regulation By introducing the new term “capital flow management measures” (CFMs),
IMF gives more policy space to emerging economies subject to major capital in-
flow surges inasmuch it loosened the clear-cut hierarchy between instruments to
manage capital flows which cover the whole range of macroeconomic policies,
prudential regulations and capital controls (defined in a jurisdictional manner). p
g
p
j
Yet, at the same time, this new policy framework contains two main short-
comings. First, by defining CFMs as a temporary instrument embedded in an over-
all strategy of financial opening, the organization insists on the general advantages
of financial liberalization, which set serious limits to developing and emerging
economies’ policy space. Second, the Fund keeps defending the separation of a
permanent prudential financial regulation (referred to as MPMs) and only tempo-
rary CFMs. In our view, this discrimination is not feasible especially in emerging
and developing economies inasmuch their currencies are characterized by a limit-
ed acceptance at the international level which increases the potential harmful ef-
fects of international capital flows in terms of financial fragility and macroeco-
nomic management.3 3 For a more details on the new IMF institutional view on capital controls, see Gallagher (2012) and
Fritz and Prates (2014). Capital flows and capital account
liberalization in Brazil: A brief overview Capital account liberalization in Brazil began in the 1990s and was most time
incremental, marked by key rules that, given their strong impact on capital inflows
and outflows, can be considered as landmarks. This was the case with the ap-
proval, in 1991, of Annex IV of Central Bank of Brazil Resolution no. 1,289, per-
mitting foreign institutional investors to participate directly in the Brazilian capi-
tal market and, in 1992, the redesign of CC5 accounts, permitting residents and
non-residents to make capital transfers abroad from Brazil. So, both capital in-
flows and capital outflows were liberalized in Brazil. The process of financial
opening gained momentum in January 2000, when the Resolution CMN no. 2,689
allowed the unrestricted access of non-resident (i.e. foreign) investors to all the
segments of the domestic financial market, including the derivatives market. Afterwards, during the 2000s there was in course a process of consolidation of the
foreign exchange rules (Paula, 2011). Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 114 Domestic norms on foreign exchange transactions allow the implementa-
tion of capital controls at any time – there is no formal restriction on this con-
cern. Law 4,321/1961, which allows the adoption of controls on capital out-
flows by foreign investors and transnational enterprises, has not been repealed. However, there are some limits to the efficacy of CCR due to two specificities of
the Brazilian economy. The first one is the high degree of financial openness of the Brazilian econo-
my. The country had a broad and deep experience of external financial liberal-
ization. Although Brazil adopted a more gradual style of Washington Consensus
policies compared other Latin American countries, capital account liberalization
was relatively fast and widespread. The second specificity is the huge differential
between internal and external interest rates, which attracted dramatically capital
inflows, mainly portfolio ones, and stimulated private agents to find loopholes
to circumvent the regulations (regulatory arbitrage), as detailed in the next sec-
tion. Despite the reduction of the policy rate (Selic) by the BCB from August
2011 to February 2012, the differential between the internal and external inter-
est rates has been still high compared to other emerging countries (see Figure 1). Since middle of 2013, when BCB began to increase the Selic rate seeking to re-
duce the inflation rate, interest rate differential increased dramatically from
5.9% to 7.8% in December 2013. 15
Revista de Economia Política 37 (1), 2017 • pp. 108-128 Capital flows and capital account
liberalization in Brazil: A brief overview Figure 1: Interest rate differential (%), 2010-2013
11,1%
5,9%
7,8%
-5%
-3%
-1%
1%
3%
5%
7%
9%
11%
Jan-10
Feb-10
Mar-10
Apr-10
May-10
Jun-10
Jul-10
Aug-10
Sep-10
Oct-10
Nov-10
Dec-10
Jan-11
Feb-11
Mar-11
Apr-11
May-11
Jun-11
Jul-11
Aug-11
Sep-11
Oct-11
Nov-11
Dec-11
Jan-12
Feb-12
Mar-12
Apr-12
May-12
Jun-12
Jul-12
Aug-12
Sep-12
Oct-12
Nov-12
Dec-12
Jan-13
Feb-13
Mar-13
Apr-13
May-13
Jun-13
Jul-13
Aug-13
Sep-13
Oct-13
Nov-13
Dec-13
Brazil
South Korea
Thailand
Turkey
Indonesia
South Africa
Mexico
Peru
Note: (1) Interest rate differencial = Country policy rate minus Fed Fund Rate plus country-risk . Source: Authors’ elaboration with data from Central Bank of Brazil. Figure 1: Interest rate differential (%), 2010-2013 Source: Authors’ elaboration with data from Central Bank of Brazil. As we can see in the Figure 2, a new surge of capital inflows to Latin America –
except to Argentina and Venezuela - started in the middle of 2009, with a quick re- 15
Revista de Economia Política 37 (1), 2017 • pp. 108-128 115 covery of capital inflows after the contagious of the global financial crisis4. The main
drivers behind of this wave to emerging economies were: (i) loosening monetary
policy in advanced economies due to the “quantitative easing” policy of the FED, and
later of the ECB, widening the interest rate differentials and creating abundant glob-
al market liquidity; (ii) better economic performance of the emerging economies and
the slow recovery of the developed countries; (iii) sound fiscal and debt position of
the emerging economies relative to advance economies; and (iv) and particularly im-
portant to Latin American countries, quick and continuous recovered of commodity
prices until May 2011, when prices started a decline trend. This new surge of capital
inflows, however, has lost momentum (though, not sharply) and become more vola-
tile since 2013. As we have already mentioned, in May 2013 Fed announced a gra-
dual reduction in the monetary stimulus which resulted in a greater turbulence in the
world financial market, with particular impact on emerging economies. 250
300
4 For an analysis of the causes and consequences of the capital inflows boom to Latin America in 2009-
2011, see Paula (2013). Capital flows and capital account
liberalization in Brazil: A brief overview Figure 2: Financial account net balance (USD million), 1994-01/2014-01
-30000
-20000
-10000
0
10000
20000
30000
40000
50000
1994 Q1
1994 Q3
1995 Q1
1995 Q3
1996 Q1
1996 Q3
1997 Q1
1997 Q3
1998 Q1
1998 Q3
1999 Q1
1999 Q3
2000 Q1
2000 Q3
2001 Q1
2001 Q3
2002 Q1
2002 Q3
2003 Q1
2003 Q3
2004 Q1
2004 Q3
2005 Q1
2005 Q3
2006 Q1
2006 Q3
2007 Q1
2007 Q3
2008 Q1
2008 Q3
2009 Q1
2009 Q3
2010 Q1
2010 Q3
2011 Q1
2011 Q3
2012 Q1
2012 Q3
2013 Q1
2013 Q3
2014 Q1
Argentina
Brazil
Chile
Colombia
Mexico
Peru
Venezuela
Source: MF (2014) - International Financial Statistics. Figure 2: Financial account net balance (USD million), 1994-01/2014-01 Source: MF (2014) - International Financial Statistics. 350
400
Brazil
Argentina
Chile
Colombia
All these factors, along with improved global risk appetite, attracted capital
inflows, especially portfolio debt capital flows. The episode of capital inflows after
the Lehman Brother’s contagion was characterized by a predominance of volatile
portfolio inflows, much more than previous wave, with a sharp and unprecedent-
ed increase in the flows (net flows of more than USD 50 billion in some quarters),
followed by the direct investments that have increased in 2011. Note that Brazil Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 116 had records of capital flows in the recent wave, followed far above by Mexico,
Colombia, Peru and Chile. This movement can be attributed to the improvement
in the country-risk (Brazil got the degree of “investment grade” in the second
quarter of 2008) and the interest rate huge differentials. 11,1%
11% Figure 3 shows the BCB interventions in the foreign exchange market (spot
market) since 1999, when Brazil adopted a floating exchange rate regime, where
negative values means that it is selling foreign currencies and positive values means
that it is buying them. Capital flows and capital account
liberalization in Brazil: A brief overview A general outlook shows broadly speaking two distinguish
BCB pattern of intervention: (i) from 1999 to September 2005 exchange rate pol-
icy was of “free float” type, in which BCB did only eventual and few interventions,
mainly in periods of instability in the foreign exchange market (such as during the
2002 confidence crisis), selling dollars in order to avoid further devaluation of the
domestic currency; (ii) since October 2005 BCB began to buy foreign currencies in
a continuous but uneven way, mainly as part of the international reserves accumu-
lation policy, without having any commitment with a certain level of exchange
rate, that continued to have an appreciation trend. Some more aggressive interven-
tion was done eventually in order to reduce greater exchange rate volatility in the
sense of appreciation or depreciation. So, there was a more typical dirty floating
behavior. In the end of August 2013, however, BCB launched a program of selling
US$ 2 billion of FX swaps weekly, besides loans of US$ 1 billion to the banks in
the spot market every week, to smoothen the capital flight due to the announce-
ment of the ‘tapering’ by Fed. As a result of such policy, BCB sold USD in the sec-
ond semester of 2013 (see Figure 3). 5,9%
7,8%
-5%
-3%
-1%
1%
3%
5%
7%
9%
Jan-10
Feb-10
Mar-10
Apr-10
May-10
Jun-10
Jul-10
Aug-10
Sep-10
Oct-10
Nov-10
Dec-10
Jan-11
Feb-11
Mar-11
Apr-11
May-11
Jun-11
Jul-11
Aug-11
Sep-11
Oct-11
Nov-11
Dec-11
Jan-12
Feb-12
Mar-12
Apr-12
May-12
Jun-12
Jul-12
Aug-12
Sep-12
Oct-12
Nov-12
Dec-12
Jan-13
Feb-13
Mar-13
Apr-13
May-13
Jun-13
Jul-13
Aug-13
Sep-13
Oct-13
Nov-13
Dec-13
Brazil
South Korea
Thailand
Turkey
Indonesia
South Africa
Mexico
Peru
Note: (1) Interest rate differencial = Country policy rate minus Fed Fund Rate plus country-risk . 11
Figure 3: BCB intervention in the foreign exchange market (US$ billion), 1994 –2013
-10
-5
0
5
10
15
1999.01
1999.06
1999.11
2000.04
2000.09
2001.02
2001.07
2001.12
2002.05
2002.10
2003.03
2003.08
2004.01
2004.06
2004.11
2005.04
2005.09
2006.02
2006.07
2006.12
2007.05
2007.10
2008.03
2008.08
2009.01
2009.06
2009.11
2010.04
2010.09
2011.02
2011.07
2011.12
2012.05
2012.10
2013.03
2013.08
October 2005
Lehman Brothers
Contagion
Confidence
crisis
140
Source: Authors’ elaboration with data from Central Bank of Brazil. Capital flows and capital account
liberalization in Brazil: A brief overview Note: (+) purchase (-) selling
Revista de Economia Política 37 (1) 2017
pp 108 128 Figure 3: BCB intervention in the foreign exchange market (US$ billion), 1994 –2013 117 130
Revista de Economia Política 37 (1), 2017 • pp. 108-128 130
Revista de Economia Política 37 (1), 2017 • pp. 108-128 Since the 1990s, CCR have been mainly endogenous in Brazil, in the sense
that they have been adopted and tightened during periods of boom of capital
flows, and have been loosened during periods of capital flight (Cardoso and
Goldfajn, 1998; Paula, 2011). The exception occurred during Lula da Silva’s
government when the Brazilian economy faced a capital flows boom in 2005-
2008 without adopting CCR (instead BCB accumulated FX reserves with very
high fiscal costs). During the 2000s financial liberalization was integral part of
the ‘model’ of economic policy inspired in the New Consensus on
Macroeconomics (floating exchange regime, inflation target regime and primary
fiscal surplus). Likely the only important change was the policy of foreign ex-
change reserves accumulation that aimed at having a cushion of safety against
currency speculation and reducing exchange rate volatility. It is worth to men-
tion that Brazil compared to other major Latin American economies did a more
aggressive FX reserve accumulation policy (see Figure 4), that however did not
avoid the general trend for exchange rate appreciation. This trend was somehow
tolerated by BCB as essential to the attainment of the inflation target in Brazil
(Arestis et al., 2010). Capital flows and capital account
liberalization in Brazil: A brief overview 1994 Q1
1994 Q3
1995 Q1
1995 Q3
1996 Q1
1996 Q3
1997 Q1
1997 Q3
1998 Q1
1998 Q3
1999 Q1
1999 Q3
2000 Q1
2000 Q3
2001 Q1
2001 Q3
2002 Q1
2002 Q3
2003 Q1
2003 Q3
2004 Q1
2004 Q3
2005 Q1
2005 Q3
2006 Q1
2006 Q3
2007 Q1
2007 Q3
2008 Q1
2008 Q3
2009 Q1
2009 Q3
2010 Q1
2010 Q3
2011 Q1
2011 Q3
2012 Q1
2012 Q3
2013 Q1
2013 Q3
2014 Q1
Venezuela Figure 4: Foreign exchange reserves (US$ billion), 2000-2013
0
50
100
150
200
250
300
350
400
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Brazil
Argentina
Chile
Colombia
Mexico
Peru
Venezuela
Source: IMF (2014) Figure 4: Foreign exchange reserves (US$ billion), 2000-2013 Source: IMF (2014) 2,00
2,10
2,20
2,30
2,40
PR = Prudential Regulation
CC= Capital control
FXDR = Derivatives market regulation
60 CC
90 CC
100 CC
30 PR
Indeed, intervention in the currency markets, including accumulation of FX
reserves, has been massive in Argentina, Brazil and Chile and very high in
Colombia and Peru (Figure 4). However, for some economies there was a gradual
trend of real appreciation of the domestic currencies due to massive capital in-
flows. This was the case of Brazil and Colombia that, together with Venezuela,
are the countries whose currencies had a strong upward pressure in real terms,
which means a downward trend of the real effective exchange rate (as the ex- 10 PR
Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 118 change rate is the price of the USD) (Figure 5). Brazil was one of the emerging
countries that had a stronger trend of currency appreciation until February 2012. In 2012-2013 there was some trend of currency depreciation in the major Latin
American economies. -10
1999.01
1999.06
1999.11
2000.04
2000.09
2001.02
2001.07
2001.12
2002.05
2002.10
2003.03
2003.08
2004.01
2004.06
2004.11
2005.04
2005.09
2006.02
2006.07
2006.12
2007.05
2007.10
2008.03
2008.08
2009.01
2009.06
2009.11
2010.04
2010.09
2011.02
2011.07
2011.12
2012.05
2012.10
2013.03
2013.08 Figure 5: Real Effective Exchange Rate, 2002-2013 (2005 = 100)
40
50
60
70
80
90
100
110
120
130
140
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Argentina
Brazil
Chile
Colombia
Mexico
Peru
Venezuela
Source: ECLAC - Data Bases and Statistical Publications. Note: Annual averages. Revista de Economia Política 37 (1), 2017 • pp. 108-128 Capital flows and capital account
liberalization in Brazil: A brief overview A country’s overall real effective exchange rate index is calculated by weighting its
real bilateral exchange rate indices with each of its trading partners by each partner’s share in the country’s
total trade flows in terms of exports and imports. A currency depreciates in real effective terms when this
index rises and appreciates when it falls. Figure 5: Real Effective Exchange Rate, 2002-2013 (2005 = 100) Source: ECLAC - Data Bases and Statistical Publications. Note: Annual averages. A country’s overall real effective exchange rate index is calculated by weighting its
real bilateral exchange rate indices with each of its trading partners by each partner’s share in the country’s
total trade flows in terms of exports and imports. A currency depreciates in real effective terms when this
index rises and appreciates when it falls. Source: ECLAC - Data Bases and Statistical Publications. Note: Annual averages. A country’s overall real effective exchange rate index is calculated by weighting its
real bilateral exchange rate indices with each of its trading partners by each partner’s share in the country’s
total trade flows in terms of exports and imports. A currency depreciates in real effective terms when this
index rises and appreciates when it falls. In the post-global financial crisis context, CAR was also predominantly endog-
enous in Brazil. After implementing some slight capital controls in 2009 and 2010, it
was only after January 2011 (when the first prudential financial regulation tool was
implemented) and, mainly, after July 2011 (when the Brazilian government adopted
a broader regulation of the FX derivatives operations) a more comprehensive regu-
lation has been launched, encompassing both CAR (capital controls and prudential
financial regulation) and FX derivatives market regulation (FXDR). This new regu-
latory approach was an integral part of a broader change in the conduction of eco-
nomic policy during the first Dilma Roussef’s government (2011-2014). On the one
hand, BCB adopted a more flexible monetary policy with the use of broader tools of
monetary policy, including macro-prudential measures, and since August 2011 be-
gan a gradual and continuous reduction in the interest rates (from 12.5% in August
2011 to 8.5% in June 2012) so adopting a more “forward looking” behavior. The
BCB policy changed by mid-2013, when the policy rate (Selic) target began to in-
crease again in order to face inflationary pressures. Revista de Economia Política 37 (1), 2017 • pp. Capital flows and capital account
liberalization in Brazil: A brief overview 108-128 119 With the adoption of CAR and FXDR over 2010-2012, the Brazilian govern-
ment increased its policy space and was able to manage the level of the exchange
rate, curbing the currency appreciation trend caused by huge capital flows and
derivatives carry trade, as we will analyze in the next section. 5 Predominantly an appreciation trend, i.e., a fall of the BRL/USD exchange rate which is the price of
USD Capital account regulation and FX derivatives regulation Before detailing CAR and FXDR in Brazil after the global financial crisis, it
is worth to clarify that this last type of regulation is key in Brazil due to the speci-
ficities of the FX derivatives market in Brazil which gave rise to a central role of
this market in the trend of the country´s exchange rate (BRL)5 . This central role
stems from the much higher liquidity and depth of the FX futures market, in com-
parison with the FX spot market. The predominance of the organized segment in
the FX derivatives markets (i.e., futures traded in BM&F Bovespa) is a specificity
of Brazil´s currency market. According to Avdjiev et al. (2010), the BRL was the
second most traded currency worldwide in the organized derivatives markets in
2010. A major distinction of the Brazilian FX derivatives (futures and OTC) market
is that these operations are non-deliverable. This means that gains or losses in
these operations are liquidated in domestic (BRL – Brazilian real), and not in for-
eign currency (USD). Due to their non-deliverable legislation, the margin require-
ments of FX futures transactions can be fulfilled in BRL. Along with the unre-
stricted access of non-residents to the FX futures market in the context of financial
liberalization, this specific norm has contributed to its higher liquidity in compari-
son with the FX spot market as FX futures operations can be carried out without
any effective foreign currency flows. Both before (2003 to mid-2008) and after (since 2009) the global financial
crisis, during periods of low risk aversion, foreign institutional investors have
become the most important investor group in the FX Futures market, fostering a
real appreciation trend through derivative carry trade. This is a different kind of
currency speculation strategy, compared to the canonical carry trade through
spot market operations - when an investor borrows money in a currency with a
low interest rate and uses it to take long positions in currencies backed by high
interest rate (Gagnon and Chaboud, 2007). This strategy presents advantages
because of their inherent high degree of leverage (as in order to be carried out,
financial derivatives operations require only the payment of a margin require-
ment). Capital account regulation and FX derivatives regulation In derivatives markets carry trade expresses itself as a bet which results in a
short position in the funding currency and a long position in the target currency minantly an appreciation trend, i.e., a fall of the BRL/USD exchange rate which is the price of Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 120 (Idem, 2007). In the case of Brazil, due to the huge differential between the inter-
nal and external interest rates, since 2003 foreign investors have taken, predomi-
nantly, one way bets on the Brazilian currency appreciation through short posi-
tions in the FX futures market (selling USD dollars and buying BRL), which has
resulted in downward pressure on the USD price and, thus, in upward pressure on
the BRL price. FX future and spot markets are linked by the arbitrage carried out mainly by
banks as the dealers in the FX spot market. In front of the downward trend of the
USD futures price, these agents took the contrary position of foreign investors in
the FX futures market (long position in USD and short in BRL). With this strategy,
banks have earned arbitrage profits and, at the same time, caused additional ap-
preciation of the Brazilian currency. The derivatives carry trade turns out to be even more attractive in Brazil due
to the non-deliverable feature of the FX futures market. In the case of Brazil, until
October 2010, foreign and domestic agents could engage in derivatives carry trade
not even investing on the margin, as usual with derivatives operations, but without
disbursing one USD. More than that, this carry trade strategy could also be per-
formed without the expenditure of one single BRL because investors could meet
their margin requirements in BRL via domestic borrowed securities or guarantees
from local banks. Despite the leadership of foreign investors, profit-seeking do-
mestic agents, such as institutional investors and companies, have also engaged in
derivatives carry trade (Fritz and Prates, 2014). Therefore, while other countries only face a problem of low efficacy of capital
controls to deal with FX derivatives operations (due to its high degree of leverage),
Brazilian authorities were dealing with an even greater challenge, as these opera-
tions could simulate the impact of capital flows on the exchange rate without any
effective foreign currency flows. Consequently, capital controls, which focus only
on foreign capital flows, have proven to be ineffective in restraining them. 6 For a detailed explanation of the differences among the three kinds of regulations, see Prates (2014). Revista de Economia Política 37 (1), 2017 • pp. 108-128 Capital account regulation and FX derivatives regulation At the
same time, prudential financial regulation was also insufficient as it reaches only
financial institutions. Therefore, the two kind of CAR do not reach operations car-
ried out by non-resident investors and non-financial resident agents in the FX fu-
ture market6. The Brazilian regulatory authorities after some time realized this constraint. Since October 2010, they have launched, along with CAR, specific measures to tap
FX derivatives operations, the already mentioned “FX Derivatives Regulation”
(FXDR). This new kind of regulation has revealed to be key in restraining the BRL
appreciation trend and, in turn, mitigating the economic policy dilemma faced by
the Brazilian government, mainly, containing inflationary pressures without rein-
forcing the exchange rate misalignment (Figure 6). Revista de Economia Política 37 (1), 2017 • pp. 108-128 121 Figure 6: BRL/USD exchange rate, Jan.2009-December 2013 1,40
1,50
1,60
1,70
1,80
1,90
2,00
2,10
2,20
2,30
2,40
7/1/09
8/4/09
9/7/09
10/11/09
11/14/09
12/18/09
1/21/10
2/24/10
3/30/10
5/3/10
6/6/10
7/10/10
8/13/10
9/16/10
10/20/10
11/23/10
12/27/10
1/30/11
3/5/11
4/8/11
5/12/11
6/15/11
7/19/11
8/22/11
9/25/11
10/29/11
12/2/11
1/5/12
2/8/12
3/13/12
4/16/12
5/20/12
6/23/12
7/27/12
8/30/12
10/3/12
11/6/12
12/10/12
1/13/13
2/16/13
3/22/13
4/25/13
5/29/13
7/2/13
8/5/13
9/8/13
10/12/13
11/15/13
12/19/13
10 CC
20 and 30 CC
10 PR
40 and 50 CC
20 PR
30 FXDR
10 and 20
FXDR
70CC
80 CC
PR = Prudential Regulation
CC= Capital control
FXDR = Derivatives market regulation
60 CC
90 CC
100 CC
30 PR
50 FXDR
4
0
FXDR
Source: Authors’ elaboration with data from Central Bank of Brazil. Source: Authors’ elaboration with data from Central Bank of Brazil. In October 2010, a price-based capital control (a financial tax on inflows,
called Imposto de Operações Financeiras, IOF), already adopted at a low level in
2009, was increased to curb the undesirable effects on financial and macroeco-
nomic stability of one important kind of capital flows outside the scope of pruden-
tial financial regulation: portfolio investment in equity and fixed income. Brazilian
government a few days later also closed a loophole that allowed foreign investors
to avoid the higher tax on fixed income investments established before. Moreover,
the first FXDR was implemented: the IOF on margin requirements on FX deriva-
tives transactions was increased from 0.38 per cent to 6 per cent and some loop-
holes for IOF on margin requirements were closed (Table 1, in annex). Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 Capital account regulation and FX derivatives regulation However, the first rounds of CAR and FXDR showed to be insufficient, as the
IOF was too low to stem the derivatives carry trade due to its high leverage degree. Moreover, private agents found loopholes to circumvent the regulations (Figure 1). One of the main channels of circumvention after October 2010 was the increase in
bank´s short dollar positions in the spot currency market. In fact, the IOF on port-
folio inflows encouraged the build-up of long real/short dollar positions in the
on-shore derivatives market, that is, the derivatives carry trade supported by resi-
dent banks. To close this loophole, the Central Bank of Brazil adopted a non-interest re-
serve requirement on these positions in January 2011, which is a prudential finan-
cial regulation tool. Nevertheless, banks found another channel of regulatory arbi-
trage by switching to short term foreign borrowings which also allow them to
obtain arbitrage gains between the internal and external interest rates. The regula-
tory response was the IOF on this kind of capital flows adopted in March 2011. Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 122 However, private agents were able to make longer term loans giving the excess of
liquidity and searching for yield in the international financial market. Then, in
April the government extended the IOF to these loans. Consequently, until the first
semester of 2011, the impact of the CAR was mainly on the composition of in-
flows rather than on their volume. Concerning the currency appreciation trend, this could be curbed only after
the launch of a broader FXDR in the end of July 2011. At that time, the govern-
ment imposed a financial tax of 1 per cent on excessive long positions on BRL in
the FX derivatives market. These measures at least had a longer lasting effect as
they reach not only the marginal requirements, but the notional value of the carry
trade operations at the FX derivatives market. The exchange rate BRL/US$ in-
creased from 1.70 in 02/28/2012 to 2.00 in 18/05/2012, a nominal devaluation of
17.6% (Figure 1). An additional reason for such exchange rate behavior is some reduction in the
net capital flows to Brazil since mid-2011 due to both the BCB policy determina-
tion to reduce short-term interest rate and the increase of risk aversion of foreign
investors due to the higher likelihood of the imminence of a euro crisis. Revista de Economia Política 37 (1), 2017 • pp. 108-128 Capital account regulation and FX derivatives regulation Therefore,
besides shaping the composition of capital flows, the CAR launched by Brazil be-
gan to affect the size of flows in the new internal and external setting. This new
setting, in turn, has allowed the loosening of capital controls since December 2012
(Table 1). Yet, in the second quarter of 2013, it turned out that the effectiveness of CAR
on capital inflows and FXDR on investors´ long positions in FX derivatives de-
pended on the phase of the capital flows cycle; in other words, this effectiveness
was highly asymmetric in its boom and bust phases. As mentioned in the Introduction, in May 2013, when Fed indicated that it
might begin tapering its quantitative easing policy toward the end of the calendar
year, global investors launched a portfolio adjustment that caused a temporary but
significant reversal in capital flows to the US, putting upward pressures on the
exchange rates of many emerging economies (BIS, 2014), the most vulnerable to
global investor due to the monetary asymmetries of the international monetary
system (see section 2.1). The BRL was one of the most affected, mainly due to the higher liquidity and
deepness of the Brazilian currency and financial markets and the huge net short
positions of foreign investors in the FX future market – which bet on the BRL ap-
preciation in that moment (Prates, 2014). In order to mitigate the currency depre-
ciation, the government withdrew in June and July virtually all CAR and FXDR. Only the IOF for new and renewed foreign loans with maturities of up to 1 year
remained in force (Table 1). Then, in face of the cycle downturn, the regulations were counter-cyclically
removed. However, the quick response of Brazilian policy makers was insufficient
to curb the currency depreciation. In a setting of flight to quality (i.e, to U.S
Treasury bonds) and high risk aversion, the removal of the broader regulatory
framework which only penalizes bets in favor of the BRL was almost harmless to Revista de Economia Política 37 (1), 2017 • pp. 108-128 123 stem the currency depreciation. Its only impact was on the foreign portfolio in-
vestment in the public bonds market, which were stimulated by the greater returns
after the IOF withdrew and the policy rate rise since March 2013 (Prates, 2014). However, theses inflows were insufficient to stop the currency depreciation
due to the very dynamic of the Brazilian currency market. Capital account regulation and FX derivatives regulation As during the boom, in
the bust phase changes in investors’ positions in the FX derivatives market were
the main determinant of the BRL trend. Indeed, the withdrawal of the IOF on long
positions made easier the portfolio adjustment to short positions, which means
bets on the BRL depreciation. In this setting, only a financial tax on excessive
short positions (i.e, a FX derivatives regulation which penalize bets on the BRL
depreciation) could restrain this process. Even if a capital outflow regulation were
in force, although useful, it would be insufficient for the same reason (i.e., the cur-
rency market features). Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 Some lessons from the Brazilian experience Financial globalization has been featured by broader instability because both
the higher frequency of currency crises and speculative attacks and the smaller
domestic economic policy autonomy in this setting. Cyclical capital swings have
strong effect on major macroeconomic variables, such as exchange rates, interest
rates, domestic credit, and asset prices. In order to increase the room of maneuver
of the exchange rate regime in emerging economies some measures to reduce the
volatility of capital flows and the likelihood of speculative attacks on their curren-
cies are necessary. As we have seen in this paper, one possibility to enhance the
management of exchange rates in emerging economies is the use of CAR that in-
cludes capital controls and/or prudential financial regulations. Concerned with that amount and volatility of capital flows to emerging econ-
omies after the contagious of the 2008 crisis, IMF revised its official position to-
wards capital controls. By introducing the new term “capital flow management
measures” (CFMs), the Fund gives more policy space to emerging economies sub-
ject to major capital inflow surges inasmuch it loosened the clear-cut hierarchy
between instruments to manage capital flows which cover the whole range of mac-
roeconomic policies, macroprudential measures (MPM) and capital controls. Yet,
by labeling CFMs as a temporary instrument, the IMF still supports financial lib-
eralization in these economies as a final goal, keeps discriminating between CFMs
and MPMs and set bounds to emerging economies’ policy space and their coun-
try-specific needs. Concerning the Brazilian recent experience, some lessons can be learned in
dealing with capital flows and agents FX positions: (i) In countries with open, depth and non-deliverable FX derivatives markets,
a third type of regulation, the FX derivatives regulation, needs to be adopted along
with CAR (capital controls and prudential financial regulation). While other coun-
tries faced only a problem of low efficacy of these two regulations due to the high Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 124 leverage degree of derivatives operations, in Brazil they turned out to be ineffective
as these operations are liquidated in domestic currency. This means that they are
likely to have an impact on the exchange rate with very low or even without any
foreign capital inflows or outflows taking place. Some lessons from the Brazilian experience Besides that, most of these syn-
thetic operations are carried out in the Brazilian futures exchange – the derivatives
organized market – by a wider set of agents beyond non-financial institutions,
such as resident companies and non-resident investors. Thus, neither prudential
financial regulation measures nor capital controls are sufficient to curb these op-
erations. (ii) In Brazil, CAR and FXDR have had two goals: (a) a macroprudential one,
namely, limiting the financial fragility associated with capital flow reversals and
FX derivatives exposures; (b) increasing the policy space to control the key macro-
economic prices (exchange rate and interest rate). There are important feedbacks
between these two goals: as the Brazilian experience before the global financial
crisis showed, currency appreciation stimulates speculative positions in FX deriva-
tives, threatening financial stability. Therefore, the capacity of maintaining the
exchange rate in a competitive level (second goal) contributes to financial stability
(first goal). (iii) There are important feedbacks and complementarities among capital con-
trols, prudential financial regulation and FXDR, as much as between these mea-
sures and macroeconomic policy Capital controls need to be adopted to cover
particular types of capital flows that are outside the scope of prudential regulation
(for instance, foreign loans by non-financial companies). Moreover, FXDR may be
needed to curb currency appreciation and/or financial risks in emerging economies
with open and sophisticated FX derivatives markets, depending on the institu-
tional features of these markets (in Brazil, the non-deliverable feature of FX de-
rivatives and the predominance of futures traded in BM&F Bovespa). In turn, as
Fritz and Prates (2014) highlighted, in the case of Korea the adoption of pruden-
tial financial regulation in the aftermath of the global financial crisis has been able
to reach FX derivatives operations as they are carried on OTC markets where
banks perform the role of counterparts in all transactions and are liquidated in US
dollars. Thus prudential regulation has been able to cover all operations. Therefore,
it is not possible to establish a hierarchy among instruments to manage capital
flows and prevent their undesirable consequences, as supported by the current
IMF approach. (iv) A wider interest rate differential stimulates regulatory arbitrage, mainly in
case of countries with sophisticated financial markets. In this context, CAR and
FXDR have to be even more dynamic, flexible and adjustable, involving a steady
“fine-tuning” to close loopholes found by private agents through spot and FX de-
rivatives transactions. Revista de Economia Política 37 (1), 2017 • pp. 108-128 References Andrade, R. and Prates, D.M. (2013). “Exchange rate dynamics in a peripheral monetary economy: A
Keynesian perspective”. Journal of Post-Keynesian Economics, 35(3): 399-416. Arestis, P., Paula, L.F. and Ferrari-Filho, F. (2011). “Inflation targeting in Brazil”. International Review
of Applied Economics, 25(2): 127–148. Avdjiev, S., Upper, C. and von Kleist, K. (2010). “Highlights of international banking and financial
market activity”. BIS Quarterly Review, December. Avdjiev, S., Upper, C. and von Kleist, K. (2010). “Highlig
market activity”. BIS Quarterly Review, December. Baumann, B. A., and Gallagher, K.P. (2012). “Navigating capital flows in Brazil and Chile”. Initiative
for Policy Dialogue Working Paper Series, June. Baumann, B. A., and Gallagher, K.P. (2012). “Navigatin
for Policy Dialogue Working Paper Series, June. Biancareli, A. (2008). “A visão convencional sobre a abertura financeira e suas mutações recentes: uma
resenha crítica”. Texto para Discussão IE/UNICAMP n. 143, June. BIS (2014) BIS 84th Annual Report. Basel: Bank for International Settlements, June. Cardoso, E. and Goldfajn, I. (1998). “Capital flows to Brazil: The endogeneity of capital controls”. IMF Staff Papers 45(1): 161-202 Central Bank of Brazil (2014), http://www.bcb.gov.br, access in June 2014. Chamon, M. and Garcia, M. (2016). “Capital controls in Brazil: Effective?” Journal of International
Money and Finance 61: 163-187. Dodd, R. and Griffith-Jones, S. (2007). Brazil’s Derivatives Markets: Hedging, Central Bank Interven-
tion and Regulation. Santiago: ECLAC. Economic Commission for Latin America – ECLAC (2014), Data Bases and Statistical Publications,
http://website.eclac.cl/infest/ajax/cepalstat.asp?carpeta=estadisticas&idioma=i, access in July. p
j
p
p
p
Eichengreen, B. (2004), Capital Flows and Crises. Cambridge: The MIT Press. Epstein, G., Grabel, I. and Jomo, K.S. (2003). “Capital management techniques in developing coun-
tries”. Working Paper Series nº 56, University of Massachusetts, April. Fritz, B. and Prates, D.M. (2014). “The new IMF approach to capital account management and its
blind spots: Lessons from Brazil and South Korea”. International Review of Applied Economics
28(2): 210-239. Gagnon, J.E. and Chaboud, A.P. (2007). “What can the data tell us about carry trades in Japanese Yen?”
International Finance Discussion Papers, No 899, Board of Governors of the Federal Reserve
System, July. Gallagher, K. (2012). The IMF’s New View on Financial Globalization: A Critical Assessment, Par-
dee Center, Issues in Brief No. 2. Gallagher, K., Griffith-Jones, S. and Ocampo, J.A. (2012). “Capital account regulations for stability
and development: A new approach”. In Regulating Global Capital Flows for Long-Run Develo-
pment. Boston University/Pardee Centre Task Force Report, March. Klein, M.W. (2012). Some lessons from the Brazilian experience Only when Brazilian government adopted the three kinds of
regulations simultaneously (capital controls, prudential financial regulation and
FXDR), the policy effectiveness increased in terms of protecting the Brazilian cur-
rency from upward pressures. (v) CAR and FXDR need to be a permanent, yet flexible part of the policy Revista de Economia Política 37 (1), 2017 • pp. 108-128 125 toolkit of emerging economies to increase their policy space and reduce the risks
associated with liability structures towards capital flow reversals. Therefore, CAR
and FXDR are integral part of the macroeconomic policy, as they can help eco-
nomic authorities to face and eventually solve some macroeconomic trade-offs, as
it is the case of a situation in which the central bank wants to intervene in the FX
market in order to affect nominal exchange rate, and at the same time would like
to avoid the fiscal costs to sterilize such operations. Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 References “Capital controls: Gates versus walls”. NBER Working Paper No. 18526, Novem-
ber. 126 Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 Harvey, J.T. (2009) Currencies, Capital Flows and Crises: A Post Keynesian Analysis of Exchange Rate
Determination. London: Routledge. He, D. and McCauley, R. (2010). “Offshore markets for the domestic currency: monetary and financial
stability issues”. BIS Working Papers no. 320. Basle: BIS, September. IEDI (2012), “A maior eficácia dos controles sobre captações externas associada a taxa de juros baixos
â
bi
á
l” C
IEDI
547 N
b IEDI (2012), “A maior eficácia dos controles sobre captações externas associada a taxa de juros baixos
e câmbio estável”. Carta IEDI no. 547, November. International Monetary Fund – IMF (2008). “Reaping the benefits of financial globalization”. IMF
Occasional Paper 264, December. International Monetary Fund – IMF (2012). The Liberalization and Management of Capital Flows –
An Institutional View. Washington D.C: International Monetary Fund, November. International Monetary Fund – IMF (2013). International Financial Statistics, http://elibrary-data.imf. org/finddatareports.aspx?d=33061&e=169393, access in July 2014. Kose, M., Prasad, E., Rogoff, K. and Wei, S-J. (2006). “Financial globalization: A reappraisal”. IMF
Working Paper 06/189, August. Magud, N. and Reinhart, C. (2006). “Capital controls: An evaluation”, NBER Working Paper nº
11973, January. Ostry, J., Ghosh, A., Habermeier, K., Chamon, M., Qureshi, M. and Reinhardt, D. (2010), “‘Capital
inflows: The role of controls”. IMF Staff Position Note 10/04, February. Ostry, J., Ghosh, A., Habermeier, K., Laeven, L., Chamon, M. Qureshi, M. and Kokenyne, A. (2011). “Managing capital inflows: What tools to use?” IMF Staff Discussion Note 11/06, April. Paula, L.F. (2011). Financial Liberalization and Economic Performance: Brazil at Crossroads. London:
Routledge. Paula, L.F., Ferrari-Filho, F. and Gomes, A.M. (2013). “Capital flows, international imbalances and
economic policies in Latin America”. In Arestis, P. and Sawyer, M. (ed.). Economic Policies, Go-
vernance and the New Economics. Basingstoke: Palgrave Macmillan. Prates, D. M . (2014). “How to evaluate financial regulation of Brazil”. In: Bresser Pereira, L.C., Kregel,
J. and Burlamaqui, L. (ed.). Financial Stability and Growth - Perspectives on Financial Regulation
and New Developmentalism. London & New York: Routledge. Prates, D. M. (2015). A Gestão do Regime de Câmbio Flutuante no Brasil: Especificidades e Dilemas. Brasília: IPEA. Prasad, E., Rogoff, K., Wei, S. and Kose, M. (2003). “Effects of financial globalization on developing
countries: Some empirical evidence”. Mimeo. Washington D.C.: International Monetary Fund,
March. Schulmeister, S. (1988). Revista de Economia Política 37 (1), 2017 • pp. 108-128 References “Currency speculation and dollar fluctuations”. Banca Nazionale Del Lavoro
Quarterly Review, 167: 343-365. Stiglitz, J. (2000). “Capital market liberalization, economic growth, and instability”, World Develop-
ment, 28(6): 1075-1086. Tobin, J. (1978). “A proposal for international monetary reform”, Eastern Economic Journal, 4: 153-159. Revista de Economia Política 37 (1), 2017 • pp. 108-128 127 Annex 128
Table 1. Brazil: Capital account regulation (capital controls – CC and prudential
financial regulation - PR) and FX derivatives regulation (FXDR) – 2009/2013
Data
Number
and Kind
Measure
Agents
Oct./2009
10 CC
The Ministry of Finance implemented a 2% financial tran-
saction tax (IOF) on non-resident equity and fixed income
portfolio inflows. Non-resident
investors
Oct./2010
20 and
3 CC
(i) IOF increased from 2 to 4 percent for fixed income
portfolio investments and equity funds. (ii) IOF increased to 6 percent for fixed income
investments
(iii) Limitations were also introduced on the ability of
foreign investors to shift investment from equity to fixed
income investment
Non-resident
investors
Oct./2010
10 and 20
FXDR
(i) IOF on margin requirements on FX derivatives transac-
tions increased from 0.38 percent to 6 per cent
(ii) Loopholes for IOF on margin requirements were
closed: foreign investors in the futures markets were
no longer allowed to meet their margin requirements
via locally borrowed securities or guarantees from local
banks, which allowed them to avoid payment of the tax
Resident banks,
institutional
investors and
companies and
non-residents
investors
Jan./2011
10 PR
Non-interest reserve requirement equivalent to 60
percent of bank´s short dollar positions in the FX spot
market that exceed US$ 3 billion or their capital base,
whichever is smaller (to be implemented over 90 days)
Resident banks
Mar./2011
40 CC
Increased to 6 percent the IOF on new foreign loans
(banking loans and securities issued abroad) with
maturities of up a year. Companies and banks previously
only paid a 5.38 percent IOF on loans up to 90 days. Resident banks
and companies
April/2011
50 CC
(i) 6 percent IOF extended for the renewal of foreign
loans with maturities of up a year
(ii) 6 percent IOF extended for both new and renewed
foreign loans with maturities of up to 2 years
Resident banks
and companies
July/2011
20 PR
The Non-interest reserve requirement became
mandatory for amounts over USD 1 billion or their capital
base (whichever is smaller). Revista de Economia Política 37 (1), 2017 • pp. 108-128 References Resident banks
July/2011
30 FXDR
Excessive long positions on BRL off all agents pay a
financial tax of 1 percent. This tax can be increased up to
25 per cent
Resident banks,
institutional
investors and
companies and
non-residents
investors
Dec/2011
60 CC
IOF on equity and fixed income (linked with
infrastructure projects) portfolio inflows reduced to 0%. Non-resident
investors
Brazilian Journal of Political Economy 37 (1) 2017 • pp 108-128 128
Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 Brazilian Journal of Political Economy 37 (1), 2017 • pp. 108-128 128 Mar./2012
70 CC
(i) 6 percent IOF extended for both new and renewed
foreign loans with maturities of up to 3 years
(ii)Export advanced payment transactions with maturities
of more than a year prohibited
(iii) 6 percent IOF extended for both new and renewed
foreign loans with maturities of up to 5 years
Resident banks
and companies
Mar./2012
40 FXDR
Exporters hedge operations (up to 1,2 times the exports
of the previous year) exempted from the IOF. Resident expor-
ters
June/2012
80 CC
6 percent IOF only for new and renewed foreign loans
with maturities of up to 2 years (namely, the changes
adopted in March were reversed)
Resident banks
and companies
Dec./2012
90 CC
(i) 6 percent IOF for foreign loans with maturities of up
to 1 year
(ii)Export advanced payment transactions maturity
extended from 1 for 5 years. Resident banks
and companies
Jun./2013
100 CC
IOF on fixed income portfolio inflows reduced to 0
percent. Non-resident
investors
Jun./2013
50 FXDR
IOF of 1 percent on excessive long net positions of FX
derivatives of all agents reduced to 0 percent. Resident banks,
institutional
investors and
companies and
non-residents
investors
Jul./2013
30 PR
Non-interest reserve requirement on bank´s short dollar
positions in the FX spot market reduced from 60 percent
to 0 percent
Resident banks
Source: Authors’ elaboration based on Central Bank of Brazil’s and Minister of Finance´s websites. Revista de Economia Política 37 (1), 2017 • pp. 108-128 129
|
https://openalex.org/W3170595717
|
https://www.biorxiv.org/content/biorxiv/early/2021/06/10/2021.06.10.447902.full.pdf
|
English
| null |
CREB repressor in mushroom body enhances <i>Drosophila</i> LTM formation
|
bioRxiv (Cold Spring Harbor Laboratory)
| 2,021
|
cc-by
| 25,785
|
.
CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021.
;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint CREB repressor in mushroom body enhances Drosophila LTM formation
1
Authors
2
Chun-Chao Chen,1,* Hsuan-Wen Lin,1 Kuan-Lin Feng,2 Ruei-Yu Jhang,1,3 Linyi Chen,1,3 J. Ste
3
de Belle4,5,6, Tim Tully,1,2* and Ann-Shyn Chiang,1,2,7,8,9,10*
4
5
Affiliations
6
1Brain Research Center, National Tsing Hua University, Hsinchu 30013, Taiwan. 7
2Institute of Systems Neuroscience and Department of Life Science, National Tsing Hua
8
University, Hsinchu 30013, Taiwan. 9
3Department of Medical Science and Institute of Molecular Medicine, National Tsing Hua
10
University, Hsinchu 30013, Taiwan. 11
4Department of Psychological Sciences, University of San Diego, San Diego, CA, USA. 12
5School of Life Sciences, University of Nevada, Las Vegas, NV, USA. 13
6Acrovirt LLC, Las Vegas, NV, USA. 14
7Kaohsiung Medical University, Kaohsiung 80708, Taiwan. 15
8National Health Research Institutes, Zhunan 35053, Taiwan. 16
9China Medical University, TaiChung 40402, Taiwan. 17
10Kavli Institute for Brain and Mind, University of California at San Diego, La Jolla, CA 9209
18
0526, USA. 19
20
21
*Correspondence should be addressed to C.C.C (chenchunchao@gapp.nthu.edu.tw), T.T. 22
(ttully@gapp.nthu.edu.tw) or A.S.C. (aschiang@life.nthu.edu.tw). 23
24
25
26
27
28
29
30 Affiliations
6 3Department of Medical Science and Institute of Molecular Medicine, National Tsing Hua
10
University, Hsinchu 30013, Taiwan. 11 3Department of Medical Science and Institute of Molecular Medicine, National Tsing Hua
10
University, Hsinchu 30013, Taiwan. 11 4Department of Psychological Sciences, University of San Diego, San Diego, CA, USA. 12 5School of Life Sciences, University of Nevada, Las Vegas, NV, USA. 13 5School of Life Sciences, University of Nevada, Las Vegas, NV, USA. 13 6Acrovirt LLC, Las Vegas, NV, USA. 14 6Acrovirt LLC, Las Vegas, NV, USA. 14 7Kaohsiung Medical University, Kaohsiung 80708, Taiwan. 15 7Kaohsiung Medical University, Kaohsiung 80708, Taiwan. 15 7Kaohsiung Medical University, Kaohsiung 80708, Taiwan. 15 8National Health Research Institutes, Zhunan 35053, Taiwan. 16 National Health Research Institutes, Zhunan 35053, Taiwan. 9China Medical University, TaiChung 40402, Taiwan. 17 9China Medical University, TaiChung 40402, Taiwan. 17 Page 1 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Abstract
36 Long-term memory (LTM) requires learning-induced synthesis of new proteins allocated to specific
37
neurons and synapses in a neural circuit. Not all learned information, however, becomes permanent
38
memory. How the brain gates relevant information into LTM remains unclear. In Drosophila adults,
39
a single training session in an olfactory aversive task is not sufficient to induce protein synthesis-
40
dependent LTM. Instead, multiple spaced training sessions are required. Here, we report that initial
41
learning induces neural activity in the early subset of Kenyon cells of the mushroom body (MB),
42
and output from these neurons inhibits LTM formation. Specifically in response to spaced training,
43
Schnurri activates CREBB expression which then appears to suppress the inhibitory output from
44
MB. One training session can enhance LTM formation when this inhibitory effect is relieved. We
45
propose that learning-induced protein synthesis and spaced training-induced CREBB act
46
antagonistically to modulate output from early MB neurons during LTM formation. 47 60 Page 2 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint .
CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021.
;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Introduction
62 ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Introduction
62 Drosophila continues to demonstrate its utility as a model system to study memory, more
63
than four decades after the first mutant was described (Dudai et al., 1976). Genetic dissection of
64
olfactory aversive memory formation using various single-gene mutants has revealed at the
65
behavioral level several distinct temporal phases, including short-term memory (STM), middle-
66
term memory (MTM), anesthesia-resistant memory (ARM) and long-term memory (LTM) (Tully
67
et al., 1994; Tully et al., 1990; Tully, 1996; Quinn and Dudai, 1976). The initial learning event
68
(acquisition) after a single training session (1x) appears to induce STM, MTM and ARM, while
69
spaced training (10 training sessions with 15 min rest intervals between each, 10xS) appears
70
uniquely required to induce LTM consolidation. Manipulations of several of these “memory genes”
71
also have established cases where memory formation is either impaired or enhanced, revealing bi-
72
directional biochemical modulation of memory formation (Yin et al., 1994; Yin et al., 1995a; Ge et
73
al., 2004; Presente et al., 2004; Wu et al., 2007; Pavlopoulos et al., 2008; Huang et al., 2012; Tubon
74
et al., 2013; Fropf et al., 2014; Lee et al., 2018; Scheunemann et al., 2018). 75 As the neural substrates of olfactory memory formation are elucidated in flies, a remarkable
76
“memory circuit” is emerging. Olfactory information delivered from the antennal lobe (AL) by
77
projection neurons (PN) and foot shock reinforcement delivered by dopaminergic neurons (DAN)
78
both converge on mushroom body (MB) neurons in the central brain where their coincidence
79
triggers cascading cellular events that underlie learning (Dubnau and Chiang 2013; Perisse et al.,
80
2013; Davis, 2015; Cognigni et al., 2018). MBs play a predominant role in subsequent memory
81
formation, together with several groups of extrinsic MB neurons. Sequential genetically-defined
82
memory phases map onto distinct subpopulations of these neurons. STM involves
83
′′ and neurons and two classes of MB output neurons (MBON: MB-M4, MB-M6) (Blum
84
et al., 2009; Scheunemann et al., 2012; Bouzaiane et al., 2015). MTM involves neural activity in
85 Page 3 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. and MB-V2 neurons (Blum et al., 2009; Scheunemann et al., 2012; Bouzaiane et al., 2015).
86 and MB-V2 neurons (Blum et al., 2009; Scheunemann et al., 2012; Bouzaiane et al., 2015). 86 and MB-V2 neurons (Blum et al., 2009; Scheunemann et al., 2012; Bouzaiane et al., 2015). 86
ARM requires neural activity in MB , ′′, neurons, dorsal paired medial (DPM) neurons,
87
anterior paired lateral (APL) neurons, DAN and four different MB output neurons (MB-M4, MB-
88
M6, MB-V2, MBON-2′2a) (Lee et al., 2011; Knapek et al., 2011; Placais et al., 2012; Wu et al.,
89
2013; Bouzaiane et al., 2015; Yang et al., 2016; Scholz-Kornehl and Schwärzel, 2016; Kotoula et
90
al., 2017; Shyu et al., 2019). LTM involves neural activity in late MB neurons with output from
91
DPM, serotonergic projection neurons (SPN) and three classes of MBONs (MB-V3, MB-M4,
92
MBON-3,3′1). Cyclic AMP response element binding protein (CREB)-dependent consolidation
93
of LTM also requires activity in dorsal anterior lateral (DAL) neurons (Chen et al., 2012; Pai et al.,
94
2013; Tonoki and Davis 2015; Bouzaiane et al., 2015; Wu et al., 2017; Scheunemann et al., 2018). 95
Finally, memory retrieval depends on neural activity in DAL, pioneer neurons and four classes
96
of MBONs (MB-V2, MB-V3, MB-M4, MBON-3,3′1) (Séjourné et al., 2011; Chen et al., 2012;
97
Pai et al., 2013; Bouzaiane et al., 2015; Wu et al., 2017). 98 ARM requires neural activity in MB , ′′, neurons, dorsal paired medial (DPM) neurons,
87
anterior paired lateral (APL) neurons, DAN and four different MB output neurons (MB-M4, MB-
88
M6, MB-V2, MBON-2′2a) (Lee et al., 2011; Knapek et al., 2011; Placais et al., 2012; Wu et al.,
89
2013; Bouzaiane et al., 2015; Yang et al., 2016; Scholz-Kornehl and Schwärzel, 2016; Kotoula et
90
al., 2017; Shyu et al., 2019). LTM involves neural activity in late MB neurons with output from
91
DPM, serotonergic projection neurons (SPN) and three classes of MBONs (MB-V3, MB-M4,
92
MBON-3,3′1). Cyclic AMP response element binding protein (CREB)-dependent consolidation
93
of LTM also requires activity in dorsal anterior lateral (DAL) neurons (Chen et al., 2012; Pai et al.,
94
2013; Tonoki and Davis 2015; Bouzaiane et al., 2015; Wu et al., 2017; Scheunemann et al., 2018). and MB-V2 neurons (Blum et al., 2009; Scheunemann et al., 2012; Bouzaiane et al., 2015).
86 95
Finally, memory retrieval depends on neural activity in DAL, pioneer neurons and four classes
96
of MBONs (MB-V2, MB-V3, MB-M4, MBON-3,3′1) (Séjourné et al., 2011; Chen et al., 2012;
97
Pai et al., 2013; Bouzaiane et al., 2015; Wu et al., 2017). 98 Here, we describe another enlightening property of olfactory memory in Drosophila: inhibition
99
of LTM formation at the circuit level. Output from the early subpopulation of MB neurons
100
appears initially to inhibit LTM formation, but with spaced training, transcription of crebB (dcreb2,
101
repressor) is induced therein, apparently reducing neural output therefrom and thereby enabling
102
LTM formation. Thus, persistent olfactory memory formation appears modulated or “gated” at the
103
level of neural activity in early MB neurons. These observations presage the need for a more
104
general deconvolution of biochemical mechanisms into distinct neuronal subtypes within a memory
105
circuit. 106
107
Results
108
Learning inhibits LTM
109 Here, we describe another enlightening property of olfactory memory in Drosophila: inhibition
99
of LTM formation at the circuit level. Output from the early subpopulation of MB neurons
100
appears initially to inhibit LTM formation, but with spaced training, transcription of crebB (dcreb2,
101
repressor) is induced therein, apparently reducing neural output therefrom and thereby enabling
102
LTM formation. Thus, persistent olfactory memory formation appears modulated or “gated” at the
103
level of neural activity in early MB neurons. These observations presage the need for a more
104
general deconvolution of biochemical mechanisms into distinct neuronal subtypes within a memory
105
circuit. 106 Page 4 of 55 Page 4 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Sixty of the single-gene mutants mentioned above were generated using transposon mutagenesis
110
and were screened for impairments of memory one day after 10xS training (Dubnau et al., 2003). and MB-V2 neurons (Blum et al., 2009; Scheunemann et al., 2012; Bouzaiane et al., 2015).
86 111
Twenty-two of these lines carried P-Gal4 enhancer traps, which enabled us to drive targeted
112
inducible expression of a temperature-sensitive RicinCS transgene and then block protein synthesis
113
after 10xS (Chen et al., 2012; Pai et al., 2013; Wu et al., 2017). With protein synthesis inhibited in
114
this manner, we found impairments of 1-day memory in nine of these lines (figure supplement 1A-
115
C; table supplement 1) (Tully et al., 1994; Yin et al., 1994). Remarkably, GFP was expressed in
116
DAL neurons in all nine cases (figure supplement 1B), an observation that contributed to our
117
characterization of DAL neurons extrinsic to the MB as bona fide “LTM neurons” (Chen et al.,
118
2012). Two of the enhancer-trap memory mutants that we screened were particularly informative. 119
In umnitza flies, GFP was expressed in DAL neurons but very weakly in MB, and 1-day memory
120
after 10xS was impaired. Conversely in norka flies, GFP was expressed in MB but not in DAL
121
neurons and 1-day memory after 10xS was normal. 122 What then might be going on in the seven enhancer-trap mutants with normal memory and with
123
GFP expression in both MB & DAL neurons (Figure 1A; figure supplement 1A)? First, we
124
confirmed that active RicinCS inhibition of protein synthesis in different subsets of MB neurons did
125
not impair 1-day memory after 10xS (figure supplement 2A-B and 3). In contrast, 1-day memory
126
after 10xS was impaired by active RicinCS in DAL or MB-V3 neurons (Chen et al., 2012; Pai et al.,
127
2013; Wu et al., 2017) but was normal after massed training (10x training sessions with no rest
128
intervals; 10xM) or in control flies with inactive RicinCS (18 °C) (figure supplement 3). 129 We next tested the hypothesis that inhibition of protein synthesis in MB might enhance LTM
130
formation, thereby off-setting the impairment produced by blocking protein synthesis in DAL
131
neurons. Using cry-Gal80 or MB-Gal80, we blocked transgenic RicinCS expression outside (i.e. 132
DAL neurons) or inside of MB, respectively, in these seven enhancer-trap memory lines (Figure
133 We next tested the hypothesis that inhibition of protein synthesis in MB might enhance LTM
130
formation, thereby off-setting the impairment produced by blocking protein synthesis in DAL
131
neurons. Using cry-Gal80 or MB-Gal80, we blocked transgenic RicinCS expression outside (i.e. and MB-V2 neurons (Blum et al., 2009; Scheunemann et al., 2012; Bouzaiane et al., 2015).
86 132
DAL neurons) or inside of MB, respectively, in these seven enhancer-trap memory lines (Figure
133 Page 5 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 1B-C) and then subjected them to suboptimal 3xS training. Surprisingly, LTM formation was
134
enhanced in all seven lines by cry-Gal80 subtraction (Figure 1D-F). 135 enhanced in all seven lines by cry-Gal80 subtraction (Figure 1D-F). 135 The MB is composed of approximately 2,500 intrinsic neurons (Kenyon cells; KCs)
136
developmentally derived from four neuroblasts and distinguished by their projections that form the
137
, ′′ and lobes (Ito et al., 1997; Zhu et al., 2003; Lin et al., 2007). We looked among these
138
neuronal subpopulations to identify where LTM enhancement might reside (Figure 2; figure
139
supplement 2A). Active Ricin was expressed in all KCs or in , ′′, or neurons separately. 140
Enhanced LTM was observed after 3xS only in neurons, which were previously shown to have
141
a role in LTM formation (Blum et al., 2009; Yu et al., 2006) (Figure 2A). 142 The neurons are subdivided further into three types: pioneer , early and late
143
neurons based on their birth sequences (Zhu et al., 2003; Lin et al., 2007; Tanaka et al., 2008; Aso
144
et al., 2014). When active Ricin was expressed in these three subpopulations separately, we
145
observed enhanced LTM after 3xS only when protein synthesis was blocked in early neurons
146
(Figure 2A, left). Enhanced LTM lasted for at least 4 days (Figure 2A, right) and was not observed
147
in control flies (inactive RicinCS) after 3xS training or in flies with active RicinCS after 3xM (figure
148
supplement 2C). Blocking protein synthesis in early neurons enhanced 1- and 4-day memories
149
even after only 1x training (Figure 2B). and MB-V2 neurons (Blum et al., 2009; Scheunemann et al., 2012; Bouzaiane et al., 2015).
86 We confirmed these results using two additional
169
Gal4 drivers expressing specifically in early neurons (Figure 4C; figure supplement 2A) and
170
observed normal 1-day memory after 3xM (Figure 4B, middle) as in control flies carrying UAS-
171
shits alone, Gal4 alone (Figure 4B, right) or at the permissive temperature for shits (Figure 4A, left
172
and 4C, right). Finally, we established that LTM after 10xS training was not impaired when neural
173
output from early neurons was blocked (figure supplement 4B). Together, these results indicate
174
that, in the absence of spaced training, neural output from early neurons inhibits LTM formation. 175
To examine the influence of early neuron membrane excitability on memory, we used ectopic
176
expression of either transgenic hyperexcitors (UAS-ShawDN, a dominant-negative Shaw potassium
177
channel and UAS-NaChBac, a sodium channel) or hypoexcitors (UAS-Shaw, a Shaw potassium
178
channel and UAS-Kir2.1::GFP, an inward-rectifying potassium channel Venken et al., 2011). 179
Temporal control of these transgenes was enabled using a tub-Gal80ts transgene (conditional
180
expression of Gal80 suppresses Gal4 expression at 18 °C but not at 30°C McGuire et al., 2003). 181
We found that increasing membrane excitability of early neurons impaired 1 day memory after
182 and MB-V2 neurons (Blum et al., 2009; Scheunemann et al., 2012; Bouzaiane et al., 2015).
86 Notably, LTM enhancement after 1x training required two
150
copies of transgenic RicinCS. Together, these results indicate that inhibition of protein synthesis in
151
early neurons yields a bona fide enhancement of LTM formation
152 We next inquired about the most effective time after training when inhibition of protein synthesis
153
would enhance LTM. RicinCS in early neurons was activated for 12 h in a series of time
154
windows staggered by 2 h during the first 24 h after 1x training (Chen et al., 2012; Wu et al., 2017). 155
LTM was enhanced when protein synthesis was blocked beginning from 0- to 4-h but not from 6-
156
to 12-h after training (Figure 3A). Shortening the inhibition period to 3 h, we resolved the window
157
of protein-synthesis-dependent LTM inhibition to the first 6 h after training (Figure 3B). 158 Page 6 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 159 159
Output from early neurons inhibits LTM
160
To address whether the inhibitory effect on LTM in early lobes depends on their neural output,
161
we blocked synaptic transmission from pioneer , late or early neurons using UAS-shits
162
(Dubnau et al., 2001; McGuire et al., 2001) In these experiments we found that 1- and 4-day
163
memory after 3xS training were enhanced by this manipulation in early (Figure 4A-B), but not
164
in pioneer or in late neurons (figure supplement 4A). Interestingly, we also established that
165
LTM after 3xS was enhanced when synaptic transmission was blocked from early during the
166
first 8 h period after training but not 9-24 h after training or during 1-day memory retrieval (Figure
167
4A). Thus, this temporal requirement for synaptic transmission from early neurons corresponds
168
to the requirement for protein synthesis (Figure 3). Output from early neurons inhibits LTM
160 It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 10xS training (Figure 5A, left) without affecting (1) memory after 1x training (fig. S5A), (2) 1-day
183
memory after 10xM training (figure supplement 5B) or (3) 1-day memory after 10xS training when
184
flies were kept at permissive temperature (18°C) (figure supplement 5C). This inhibitory effect
185
appears to be complete, because inhibition of protein synthesis by feeding flies cycloheximide
186
(CXM) did not further reduce 1-day LTM after 10xS training (Figure 5A, right). Decreasing
187
membrane excitability of early neurons with ectopic expression of hypoexciter transgenes, on
188
the other hand, enhanced both 1- and 4-day memory after 1x training (Figure 5B, left and middle),
189
whereas enhancement of 1-day memory was not observed in control transgenic flies kept at the
190
permissive temperature (18°C) (Figure 5B, right). We also found normal 1-day memory in these
191
transgenic flies after 10xS training (Figure 5C). Thus, sufficient spaced training appeared to occlude
192
the enhancing effects on LTM formation of decreased membrane excitability in early neurons. 193
These data further support the notion that neural activity from early neurons inhibits LTM
194
formation downstream (Dubnau and Chiang 2013; Pai, et al., 2013; Wu et al., 2017). 195 10xS training (Figure 5A, left) without affecting (1) memory after 1x training (fig. S5A), (2) 1-day
183 Output from early neurons inhibits LTM
160 To address whether the inhibitory effect on LTM in early lobes depends on their neural output,
161
we blocked synaptic transmission from pioneer , late or early neurons using UAS-shits
162
(Dubnau et al., 2001; McGuire et al., 2001) In these experiments we found that 1- and 4-day
163
memory after 3xS training were enhanced by this manipulation in early (Figure 4A-B), but not
164
in pioneer or in late neurons (figure supplement 4A). Interestingly, we also established that
165
LTM after 3xS was enhanced when synaptic transmission was blocked from early during the
166
first 8 h period after training but not 9-24 h after training or during 1-day memory retrieval (Figure
167
4A). Thus, this temporal requirement for synaptic transmission from early neurons corresponds
168
to the requirement for protein synthesis (Figure 3). We confirmed these results using two additional
169
Gal4 drivers expressing specifically in early neurons (Figure 4C; figure supplement 2A) and
170
observed normal 1-day memory after 3xM (Figure 4B, middle) as in control flies carrying UAS-
171
shits alone, Gal4 alone (Figure 4B, right) or at the permissive temperature for shits (Figure 4A, left
172
and 4C, right). Finally, we established that LTM after 10xS training was not impaired when neural
173
output from early neurons was blocked (figure supplement 4B). Together, these results indicate
174
that, in the absence of spaced training, neural output from early neurons inhibits LTM formation. 175
To examine the influence of early neuron membrane excitability on memory, we used ectopic
176
expression of either transgenic hyperexcitors (UAS-ShawDN, a dominant-negative Shaw potassium
177
channel and UAS-NaChBac, a sodium channel) or hypoexcitors (UAS-Shaw, a Shaw potassium
178
channel and UAS-Kir2.1::GFP, an inward-rectifying potassium channel Venken et al., 2011). 179
Temporal control of these transgenes was enabled using a tub-Gal80ts transgene (conditional
180 Page 7 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. cAMP signaling in early neurons enhances LTM
197 Neural excitability is suggested to be modulated by cAMP signaling (Davis et al., 1998; Baines,
198
2003), which in MB is also involved in LTM formation (Blum et al., 2009). Accordingly, we
199
inducibly overexpressed rutabaga+ (rut+) adenylyl cyclase (AC) or constitutively active cAMP-
200
dependent protein kinase (Pkaact1) transgenes in early neurons and found that 1-day memory
201
after 1x training was enhanced to levels normally seen after 10xS in both cases (Figure 6A and 7A). 202
Moreover, inducible RNAi knockdowns of these genes impaired 1-day memory after 10xS (Figure
203
6b and 7b; figure supplement 6). Together, these results suggest that LTM formation is also
204
modulated by cAMP in early neurons. 205 Page 8 of 55 Page 8 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Protein synthesis-dependent LTM formation also depends on CREBB transcription factors, and
208
expression of CREBB protein is thought to be dependent on the expression level of protein kinases
209
involved in cAMP signaling (Lee et al., 2018). Consistent with our Rutabaga and PKA knockdown
210
results, we induced RNAi knockdown of CREBB in early neurons and observed impairment
211
of 1-day memory after 10xS training. Further impairment was not seen in combination with
212
systemic protein synthesis inhibition by feeding CXM (Figure 8C). Because inhibition of protein
213
synthesis in early neurons instead produced an enhancing effect on LTM formation, we
214
inducibly expressed crebB repressor transgenes (Zhang et al., 2019) in these neurons only, with the
215
expectation that the manipulations would lead to enhanced memory. Indeed, expressing crebB
216
enhanced 1-day memory after 1x training to levels normally seen after 10xS training (Figure 8A-
217
B; figure supplement 6). 218 Spaced training induces crebB transcription
220 CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint protein synthesis in early neurons enhanced LTM formation (Figure 1-3) suggests that crebB
232
gene products function to repress protein synthesis. 233
234
Schnurri (Shn) regulates CREBB-dependent LTM formation
235
How does spaced training induce CREBB expression? To address this question, we sought to
236
identify positive regulators of crebB transcription during LTM formation. Yeast two-hybrid and
237
chromatin immunoprecipitation experiments previously revealed several such candidates that bind
238
in the crebB promoter region (data not shown). Prominent among these were (1) CREB family
239
protein CREBA, a leucine-zipper transcription factor (Smolik et al., 1992) and (2) Shn, a zinc finger
240
C2H2 transcription factor encoded by the shn gene (Marty et al., 2000) which also was identified
241
in a transposon mutagenesis screen for impairment of 1-day memory after 10xS as the umnitza
242
mutant (Dubnau et al., 2003) (described above, see figure supplement 1). 243 protein synthesis in early neurons enhanced LTM formation (Figure 1-3) suggests that crebB
232
gene products function to repress protein synthesis. 233 protein synthesis in early neurons enhanced LTM formation (Figure 1-3) suggests that crebB
232
gene products function to repress protein synthesis. 233 Spaced training induces crebB transcription
220 We next generated a crebB promoter-driven Gal4 transgene containing an 11-kb 5′ genomic
221
sequence just upstream of CREBB (see Methods) (Yin et. al., 1995b). This crebB-Gal4 drives GFP
222
expression in most glia cells and brain neurons, including most MB neurons, though higher levels
223
of expression can be seen in compared to ′′ or neurons (Figure 9A). By photo converting
224
pre-existing green KAEDE to red prior to training (Chen et. al., 2012), we measured significantly
225
more newly synthesized crebB-Gal4 green KAEDE in the MB -lobe during 24-h intervals after
226
5xS or 10xS training, but not after 1x training or 10xM training in comparison with naïve control
227
flies (Figure 9B-C). This training-induced increase in crebB KAEDE appeared specific to the MB
228
neurons because spaced training did not significantly change the levels of new crebB KAEDE in
229
ellipsoid body (EB) or glia (Figure 9B, right). These results demonstrate that multiple sessions of
230
spaced training increases CREBB expression in early neurons. Our finding that inhibition of
231 Page 9 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this prep
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
as not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
v preprint protein synthesis in early neurons enhanced LTM formation (Figure 1-3) suggests that crebB
gene products function to repress protein synthesis. . Schnurri (Shn) regulates CREBB-dependent LTM formation
235 How does spaced training induce CREBB expression? To address this question, we sought to
236
identify positive regulators of crebB transcription during LTM formation. Yeast two-hybrid and
237
chromatin immunoprecipitation experiments previously revealed several such candidates that bind
238
in the crebB promoter region (data not shown). Prominent among these were (1) CREB family
239
protein CREBA, a leucine-zipper transcription factor (Smolik et al., 1992) and (2) Shn, a zinc finger
240
C2H2 transcription factor encoded by the shn gene (Marty et al., 2000) which also was identified
241
in a transposon mutagenesis screen for impairment of 1-day memory after 10xS as the umnitza
242
mutant (Dubnau et al., 2003) (described above, see figure supplement 1). 243 Together, these findings implicated CREBA and Shn as candidate regulators of LTM formation
244
through transcriptional activation of crebB. Interestingly, we inducibly overexpressed each in early
245
neurons and found enhanced 1-day memory after 1x or 3xS training with this manipulation of
246
the shn+ transgene but not crebA+ (Figure 10A; figure supplement 7A-B). We also observed
247
strongly elevated CREBB protein expression in transgenic shn+ flies, but not in transgenic rut+ or
248
Pkaact1 flies (figure supplement 8). Moreover, inducible RNAi knockdowns of shn but not crebA
249
impaired 1-day memory after 10xS (Figure 10B; figure supplement 7C), and further impairment
250
was not seen with systemic protein synthesis inhibition after CXM feeding (Figure 10B, right). 251
Memory after 10xS was fully rescued in shn knockdown flies by crebB co-expression (Figure 10C)
252
and was enhanced relative to controls after 1x (Figure 10D). Taken together, these results show that
253
CREBB expression in early neurons in response to spaced training is Shn-dependent. 254 Page 10 of 55 Page 10 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. Schnurri (Shn) regulates CREBB-dependent LTM formation
235 ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Our data suggest that MB neurons provide a compelling cellular gating mechanism for LTM
257
formation. A single training session is sufficient to increase early neuronal excitability, the
258
output from which produces a downstream inhibitory effect on LTM formation. After spaced
259
training, cAMP signaling regulates neural excitability and/or Shn increases CREBB expression, the
260
net effects of which we suggest then represses further protein synthesis, thereby reducing early
261
output and relieving the inhibitory effect on LTM formation. Remarkably, our observations
262
emerged from a screen of enhancer trap memory mutants using RicinCS protein synthesis inhibition
263
(figure supplement 1). 1-day memory after 10xS was impaired in nine lines, eight of which showed
264
expression in both MB & DAL neurons. Curiously, another seven lines were not impaired in 1-day
265
memory after 10xS training – but they, too, showed enhancer expression patterns in MB & DAL
266
neurons. We hypothesized that blocking protein synthesis in DAL neurons impaired LTM but doing
267
so in (some) MB neurons might actually enhance LTM, negating the inhibitory effects in DAL
268
neurons. 269 We tested this idea by maintaining RicinCS expression in MB while blocking RicinCS expression
270
outside of MB using cry-Gal80 (Figure 1). Surprisingly, LTM now was enhanced in all seven of
271
these enhancer trap lines (Figure 1). We then identified early as the subset of MB neurons
272
responsible for this enhancing effect (Figure 2). Inhibition protein synthesis in early neurons
273
during the first 6 h, or blocking synaptic transmission from early neurons during the first 8 h
274
after training was sufficient to enhance LTM (Figure 3 and 4). Increasing excitability of early
275
neurons impaired LTM, but decreasing excitability again enhanced LTM (Figure 5). We next asked
276
whether these neural excitability-dependent effects were also cAMP dependent. RNAi mediated
277
knockdown of Rutabaga or PKA in early impaired LTM, while overexpression of a rut+ or
278
Pkaact1 transgene enhanced LTM (Figure 6 and 7). CREBB expression is suggest to be
279
synergistically and post-transcriptionally regulated by protein kinases responding to cAMP
280
signaling (15) and accordingly, our RNAi mediated knockdown of CREBB in early impaired
281 Page 11 of 55 . Schnurri (Shn) regulates CREBB-dependent LTM formation
235 CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint LTM, while overexpression of a crebB transgene enhanced LTM (Figure 8). Finally, using a crebB
282
promoter driven Gal4 transgene, we show that CREBB transcription increases after 5xS or 10xS
283
spaced training but not after 1x training (Figure 9). Thus, spaced training-dependent expression of
284
CREBB repressor proteins in early neurons blocks this inhibitory output from early
285
neurons, thereby allowing LTM formation (downstream) to proceed. 286 LTM, while overexpression of a crebB transgene enhanced LTM (Figure 8). Finally, using a crebB
282
promoter driven Gal4 transgene, we show that CREBB transcription increases after 5xS or 10xS
283
spaced training but not after 1x training (Figure 9). Thus, spaced training-dependent expression of
284
CREBB repressor proteins in early neurons blocks this inhibitory output from early
285
neurons, thereby allowing LTM formation (downstream) to proceed. 286 LTM, while overexpression of a crebB transgene enhanced LTM (Figure 8). Finally, using a crebB
282
promoter driven Gal4 transgene, we show that CREBB transcription increases after 5xS or 10xS
283
spaced training but not after 1x training (Figure 9). Thus, spaced training-dependent expression of
284
CREBB repressor proteins in early neurons blocks this inhibitory output from early
285
neurons, thereby allowing LTM formation (downstream) to proceed. 286 CREBB repressor proteins in early neurons blocks this inhibitory output from early
285
neurons, thereby allowing LTM formation (downstream) to proceed. 286
An enhancing role associated with Shn-induced expression of CREBB repressor is a novel aspect
287
of this LTM gating mechanism (Figure 10 and figure supplement 8). Previous reports have claimed
288
that chronic expression of a CREBB repressor or RNAi transgenes in all neurons impaired 1-
289
day memory after spaced training (Yu et al., 2006; Lee et al., 2018). Schnurri (Shn) regulates CREBB-dependent LTM formation
235 ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint decreased (OK107 expressing in all MB neurons; c739 expressing in all neurons; 1471
307
expressing in neurons), in others expression increased (c747 and c772 expressing variably in all
308
MB neurons) or in some no changes were detected (c320 expressing variably in ′′ and
309
subpopulation, 17d expressing primarily in late and in early neurons). Indeed, these authors
310
point out that, because CRE-luciferase was expressed in more than one subpopulation of MB
311
neurons, only net effects of CREB function could be quantified. Obviously, such a conclusion must
312
be drawn from any behavioral data collected after CREBB manipulations in multiple
313
subpopulations of MB neurons. Our study provides a dramatic example of this point. By restricting
314
our manipulation only to the early neurons and only in adult stage animals, we show that acute
315
overexpression or knockdown of CREBB enhances or impairs LTM formation, respectively (Figure
316
8) and that spaced training serves to increase the expression of CREBB in these neurons (Figure 9). 317
Of particular relevance to our future studies is the curious discovery that output from early
318
neurons specifically inhibits LTM formation. We find no evidence of inhibitory transmitter (i.e.,
319
GABA) synthesis or signaling in early neurons, however, and others have suggested that
320
memory-relevant MB output synapses are cholinergic (Barnstedt et al.,2016). Thus, we presume
321
that inhibition of LTM lies somewhere downstream in the memory circuit. Furthermore, we note
322
that ARM appears to involve neurons (Lee et al.,2011; Knapek et al.,2011; Scholz-Kornehl and
323
Schwärzel, 2016; Kotoula et al., 2017; Shyu et al., 2019) and to inhibit LTM formation (Isabel et
324
al., 2004; Placais et al., 2012). Thus, a molecular link between ARM and LTM may reside in early
325 decreased (OK107 expressing in all MB neurons; c739 expressing in all neurons; 1471
307
expressing in neurons), in others expression increased (c747 and c772 expressing variably in all
308
MB neurons) or in some no changes were detected (c320 expressing variably in ′′ and
309
subpopulation, 17d expressing primarily in late and in early neurons). Indeed, these authors
310
point out that, because CRE-luciferase was expressed in more than one subpopulation of MB
311
neurons, only net effects of CREB function could be quantified. Schnurri (Shn) regulates CREBB-dependent LTM formation
235 Chen et al., (2012) documented,
290
however, that these chronic disruptions of CREBB produced developmental abnormalities in MB
291
structure. In contrast, acute induced expression of active RicinCS or CREBB repressor only in adult
292
neurons did not impair 1-day memory after spaced training (and did not produce structural
293
defects). Using a different inducible system (MB247-Switch) to acutely expresses CREBB in and
294
neurons, Hirano et al., (2016) showed a mild impairment of 1-day memory after spaced training. 295
More interestingly, they used various molecular genetic tools to show that interactions among
296
CREBB, CREB Binding Protein (CBP) and CREB Regulated Transcription Coactivator (CRTC)
297
in MB clearly were involved in LTM formation or maintenance, respectively. Using the same
298
inducible gene switch tool, Miyashita et al., (2018) showed a fascinating positive regulatory loop
299
between Fos and CREBB in MB during LTM formation – but they did not show behavioral data
300
pertaining to manipulation of CREBB per se- and they did not restrict their experiments to early
301
neurons. 302 A recent study that features cyclic AMP-response element (CRE)-driven transgenes is pertinent
303
to this report. Zhang et al., (2015) expressed a CRE-luciferase transgene in different subpopulations
304
of MB neurons and then monitored luciferase activity in live flies at various times after spaced
305
training. Immediately after spaced training, they showed in some cases luciferase expression
306 P
12 f 55
A recent study that features cyclic AMP-response element (CRE)-driven transgenes is pertinent
303
to this report. Zhang et al., (2015) expressed a CRE-luciferase transgene in different subpopulations
304
of MB neurons and then monitored luciferase activity in live flies at various times after spaced
305
training. Immediately after spaced training, they showed in some cases luciferase expression
306 Page 12 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. Schnurri (Shn) regulates CREBB-dependent LTM formation
235 Obviously, such a conclusion must
312
be drawn from any behavioral data collected after CREBB manipulations in multiple
313
subpopulations of MB neurons. Our study provides a dramatic example of this point. By restricting
314
our manipulation only to the early neurons and only in adult stage animals, we show that acute
315
overexpression or knockdown of CREBB enhances or impairs LTM formation, respectively (Figure
316
8) and that spaced training serves to increase the expression of CREBB in these neurons (Figure 9). 317
Of particular relevance to our future studies is the curious discovery that output from early
318
neurons specifically inhibits LTM formation. We find no evidence of inhibitory transmitter (i.e.,
319
GABA) synthesis or signaling in early neurons, however, and others have suggested that
320
memory-relevant MB output synapses are cholinergic (Barnstedt et al.,2016). Thus, we presume
321
that inhibition of LTM lies somewhere downstream in the memory circuit. Furthermore, we note
322
that ARM appears to involve neurons (Lee et al.,2011; Knapek et al.,2011; Scholz-Kornehl and
323
Schwärzel, 2016; Kotoula et al., 2017; Shyu et al., 2019) and to inhibit LTM formation (Isabel et
324
al., 2004; Placais et al., 2012). Thus, a molecular link between ARM and LTM may reside in early
325 More generally, our results underscore the need to study behavior-genetic relations in each of the
327
seven MB neuronal subpopulations (Aso et al., 2014) separately before drawing firm conclusions
328
about a role for MB in specific memory phases or in the dynamics of a larger memory circuit
329
involving neurons intrinsic and extrinsic to MB. With the more complex circuitries in vertebrate
330 More generally, our results underscore the need to study behavior-genetic relations in each of the
327
seven MB neuronal subpopulations (Aso et al., 2014) separately before drawing firm conclusions
328
about a role for MB in specific memory phases or in the dynamics of a larger memory circuit
329
involving neurons intrinsic and extrinsic to MB. With the more complex circuitries in vertebrate
330 More generally, our results underscore the need to study behavior-genetic relations in each of the
327
seven MB neuronal subpopulations (Aso et al., 2014) separately before drawing firm conclusions
328
about a role for MB in specific memory phases or in the dynamics of a larger memory circuit
329
involving neurons intrinsic and extrinsic to MB. Schnurri (Shn) regulates CREBB-dependent LTM formation
235 With the more complex circuitries in vertebrate
330 Page 13 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint models, such deconstruction of memory formation into specific neuronal subtypes will be
. CC-BY 4.0 International license
available under a
as not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
v preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It
The copyright holder for this pre
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint mal models, such deconstruction of memory formation into specific neuronal subtypes will be
n more critical and enlightening. . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint animal models, such deconstruction of memory formation into specific neuronal subtypes will be
331
even more critical and enlightening. 332
333
Materials and Methods
334
A collection of Drosophila P-Gal4 transposon insertions were previously selected in an enhancer
335
trap mutagenesis screen for long-term memory phenotypes (Dubnau et al., 2003). The resultant
336
Gal4 expression patterns in seven of these mutants were leveraged to drive and temporally control
337
cold-sensitive RicinCS activity to block protein synthesis in the identified neuron subsets. In addition,
338
we spatially restricted RicinCS activity by inhibiting Gal4 with MB or DAL neuron-specific
339
expression of Gal80. Schnurri (Shn) regulates CREBB-dependent LTM formation
235 We used an automated olfactory aversive learning task (Tully et al., 1994) and
340
assessed LTM after blocking protein synthesis, inhibiting consolidation in these temporally and
341
spatially restricted domains to identify the subsets of neurons critical for this task. Blocking
342
transmission from these neurons with Gal4-targetted temperature-sensitive Dynamints after training
343
was used to test the implicated roles of these neurons in LTM consolidation (Dubnau et al., 2001;
344
McGuire et al., 2001). Spatial and temporal regulation of K+ and Na+ channel activity with
345
transgene overexpression and RNAi knockdown within these neurons was used to assess the
346
downstream impacts of signaling valence on LTM. Similarly, restricted expression of transgenes
347
was used to examine the training-responsive effects on LTM. We evaluated training-responsive
348
CREBB expression with confocal microscopy using a Gal4-targeted UV-sensitive KAEDE reporter
349
system (Chen et al 2012) In various experiments flies were fed CXM to provide a systemic level
350 animal models, such deconstruction of memory formation into specific neuronal subtypes will be
331
even more critical and enlightening. 332 A collection of Drosophila P-Gal4 transposon insertions were previously selected in an enhancer
335
trap mutagenesis screen for long-term memory phenotypes (Dubnau et al., 2003). The resultant
336
Gal4 expression patterns in seven of these mutants were leveraged to drive and temporally control
337
cold-sensitive RicinCS activity to block protein synthesis in the identified neuron subsets. In addition,
338
we spatially restricted RicinCS activity by inhibiting Gal4 with MB or DAL neuron-specific
339
expression of Gal80. We used an automated olfactory aversive learning task (Tully et al., 1994) and
340
assessed LTM after blocking protein synthesis, inhibiting consolidation in these temporally and
341
spatially restricted domains to identify the subsets of neurons critical for this task. Blocking
342
transmission from these neurons with Gal4-targetted temperature-sensitive Dynamints after training
343
was used to test the implicated roles of these neurons in LTM consolidation (Dubnau et al., 2001;
344
McGuire et al., 2001). Spatial and temporal regulation of K+ and Na+ channel activity with
345
transgene overexpression and RNAi knockdown within these neurons was used to assess the
346
downstream impacts of signaling valence on LTM. Similarly, restricted expression of transgenes
347
was used to examine the training-responsive effects on LTM. We evaluated training-responsive
348
CREBB expression with confocal microscopy using a Gal4-targeted UV-sensitive KAEDE reporter
349
system (Chen et al., 2012). Schnurri (Shn) regulates CREBB-dependent LTM formation
235 In various experiments, flies were fed CXM to provide a systemic level
350
of protein synthesis inhibition. Detailed procedures for all methods are described in the
351
supplementary materials. 352
353 Page 14 of 55 Page 14 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is
The copyright holder for this prep
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Fly stocks were maintained on standard corn meal/yeast/agar medium at 25 ± 1 °C or 18 ± 1 °C and
70% relative humidity on a 12:12-h light:dark cycle. All genotypes and sources are listed in table
supplement 2. . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Fly stocks were maintained on standard corn meal/yeast/agar medium at 25 ± 1 °C or 18 ± 1 °C and
355
70% relative humidity on a 12:12-h light:dark cycle. All genotypes and sources are listed in table
356
supplement 2. 357
358
Behaviour
359
Olfactory associative learning was evaluated by training 6- to 7-day-old flies in a T-maze apparatus
360
with a Pavlovian olfactory conditioning procedure (Tully and Quinn, 1985) as described previously
361
(Chen et al., 2012; Pai et al., 2013; Wu et al., 2017). All experiments were conducted in the dark
362
in an environment-controlled room at the required temperatures and 70% relative humidity. The
363
odours used were 3-octanol (OCT) and 4-methylcyclohexanol (MCH). Schnurri (Shn) regulates CREBB-dependent LTM formation
235 Each experiment consisted
364
of two groups of approximately 100 flies, each of which was conditioned with one of the two odours. 365
Flies were exposed sequentially to two odours that were carried through the training chamber in a
366
current of air (odours were bubbled at 750 ml/min). In a single training session, flies first were
367
exposed for 60 s to the conditioned stimulus (CS ＋), during which time they received the
368
unconditioned stimulus (US), which consisted of 12 1.5-s pulses of 60 V dc electric shock presented
369
at 5-s interpulse intervals. After the presentation of the CS+ condition, the chamber was flushed
370
with fresh air for 45 s. Then flies were exposed for 60 s to the unpaired CS–. To evaluate memory
371
retention immediately after single-session training (acquisition), flies were gently tapped into an
372
elevator-like compartment immediately after training. After 90 s, the flies were transported to the
373
choice point of a T-maze, in which they were exposed to two converging currents of air (one
374 Fly stocks were maintained on standard corn meal/yeast/agar medium at 25 ± 1 °C or 18 ± 1 °C and
355
70% relative humidity on a 12:12-h light:dark cycle. All genotypes and sources are listed in table
356
supplement 2. 357 Page 15 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint (NCS － – NCS+)/(NCS － + NCS+) was calculated and averaged over these two complementary
380
experiments, with the final PI = (PI1 + PI2)/2. Averaging of the two reciprocal scores eliminated
381
any potential biases originating from the machine, naïve odour preferences, or non-associative
382
changes in olfaction. For 24-h memory experiments, flies were subjected to single-session training,
383
training massed together without rest, or training spaced out with 15-min rest intervals. Schnurri (Shn) regulates CREBB-dependent LTM formation
235 For these
384
training protocols, robotic trainers were used. All genotypes were trained and tested in parallel and
385
rotated among all of the robotic trainers to ensure a balanced experiment. The genetic backgrounds
386
of all fly strains were equilibrated to the “Canton” wild-type background by five or more
387
generations of backcrossing. In tub-Gal80ts experiments, flies raised at 18 °C were transferred to
388
30 °C for at least five days before the experiments. 389 any potential biases originating from the machine, naïve odour preferences, or non-associative
382
changes in olfaction. For 24-h memory experiments, flies were subjected to single-session training,
383
training massed together without rest, or training spaced out with 15-min rest intervals. For these
384
training protocols, robotic trainers were used. All genotypes were trained and tested in parallel and
385
rotated among all of the robotic trainers to ensure a balanced experiment. The genetic backgrounds
386
of all fly strains were equilibrated to the “Canton” wild-type background by five or more
387
generations of backcrossing. In tub-Gal80ts experiments, flies raised at 18 °C were transferred to
388
30 °C for at least five days before the experiments. 389
390
Pharmacological treatment
391
To block protein synthesis, flies were fed 35 mM cycloheximide (Sigma) in 5% glucose 1 day
392
before training until immediately before the test (Tully et al., 1994). 393
394
crebB promoter construct
395
To engineer the crebB promoter construct, polymerase chain reaction (PCR) was performed using
396
genomic DNA from the wild-type Canton-S w1118 (iso1CJ) fly line as the template together with
397
the forward primer 5′GAAAAGTGCCACCTGCTGCATGTCTACCAACAGTTCGAG 3′ and the
398
reverse primer 5′CCGGATCTGCTAGCGGTTCCAGCTGCTGTCTGTATGAC 3′. A 11.6-kb
399
PCR product was generated and inserted into the pBPGAL4.2Uw-2 vector, was digested with AatII
400
and KpnI using In-Fusion® cloning system (Clontech). The promoter construct was injected into
401
attP40-containing fly strains to obtain the transgenic fly lines. 402
403
KAEDE measurement
404 Pharmacological treatment
391 Briefly, pre-existing KAEDE proteins were photoconverted into red fluorescent
412
proteins by 365–395 nm UV irradiation generated from a 120-W mercury lamp. For behavioural
413
testing, approximately 15–20 flies kept in a clear plastic syringe were directly exposed to UV light
414
at a distance of 5 cm for 1 h. Individual neurons expressing KAEDE were directly visualised
415
through an open window in the fly’s head capsule. Living samples were used because the signal-
416
to-noise ratio of green to red KAEDE is greatly reduced after chemical fixation. KAEDE neurons
417
were located in less than 5 s by a fast pre-scanning of red KAEDE excited by a 561-nm laser, to
418
avoid unnecessary fluorescence quenching of green KAEDE during repeated scanning. A single
419
optical slice through the MB -lobe tip was imaged at a resolution of 1024×1024 pixels under a
420
confocal microscope with a 40× C-Apochromat water-immersion objective lens (N.A. value 1.2,
421
working distance 220 μm). All brain samples in the experiment were imaged with the same optical
422
settings maximised for green and red KAEDE immediately before and after photoconversion,
423
respectively. In all cases, both green KAEDE (excited by a 488-nm laser) and red KAEDE (excited
424
by a 561-nm laser) were measured. By using the amount of red KAEDE as an internal standard to
425
calibrate individual variation, we calculated the rate of increase in green KAEDE synthesis after
426
photoconversion with the formula (ΔF) = %(Ft1 – average Ft0)/average Ft0, where Ft1 and Ft0 are
427 KAEDE is a photoconvertible green fluorescent protein, irreversibly changing its structure to a red
405
fluorescent protein upon ultraviolet irradiation (Ando et al., 2002). Taking advantage of circadian
406
transcription and protein synthesis in the lateral clock neurons, we previously validated de novo
407
KAEDE synthesis in per-Gal4>UAS-kaede flies, in which it faithfully reports the cyclic
408
transcriptions of the period gene. Feeding cycloheximide also suppressed green KAEDE synthesis,
409
while not affecting the already-converted red KAEDE (Chen et al., 2012). To measure the amount
410
of newly synthesised KAEDE in MB neurons, we used procedures adapted from a previous study
411
(Chen et al., 2012). Briefly, pre-existing KAEDE proteins were photoconverted into red fluorescent
412
proteins by 365–395 nm UV irradiation generated from a 120-W mercury lamp. Pharmacological treatment
391 To block protein synthesis, flies were fed 35 mM cycloheximide (Sigma) in 5% glucose 1 day
392
before training until immediately before the test (Tully et al., 1994). 393 To block protein synthesis, flies were fed 35 mM cycloheximide (Sigma) in 5% glucose 1 day
392
before training until immediately before the test (Tully et al., 1994). 393
394
bB To block protein synthesis, flies were fed 35 mM cycloheximide (Sigma) in 5% glucose 1 day
392
before training until immediately before the test (Tully et al., 1994). 393 To engineer the crebB promoter construct, polymerase chain reaction (PCR) was performed using
396
genomic DNA from the wild-type Canton-S w1118 (iso1CJ) fly line as the template together with
397
the forward primer 5′GAAAAGTGCCACCTGCTGCATGTCTACCAACAGTTCGAG 3′ and the
398
reverse primer 5′CCGGATCTGCTAGCGGTTCCAGCTGCTGTCTGTATGAC 3′. A 11.6-kb
399
PCR product was generated and inserted into the pBPGAL4.2Uw-2 vector, was digested with AatII
400
and KpnI using In-Fusion® cloning system (Clontech). The promoter construct was injected into
401
attP40-containing fly strains to obtain the transgenic fly lines. 402 Page 16 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint KAEDE is a photoconvertible green fluorescent protein, irreversibly changing its structure to a red
405
fluorescent protein upon ultraviolet irradiation (Ando et al., 2002). Taking advantage of circadian
406
transcription and protein synthesis in the lateral clock neurons, we previously validated de novo
407
KAEDE synthesis in per-Gal4>UAS-kaede flies, in which it faithfully reports the cyclic
408
transcriptions of the period gene. Feeding cycloheximide also suppressed green KAEDE synthesis,
409
while not affecting the already-converted red KAEDE (Chen et al., 2012). To measure the amount
410
of newly synthesised KAEDE in MB neurons, we used procedures adapted from a previous study
411
(Chen et al., 2012). Spatiotemporal inhibition of protein synthesis
431 RicinCS, a mutated Ricin A chain, inactivates eukaryotic ribosomes by hydrolytically cleaving the
432
N-glycosidic bond (A4324) of the 28S ribosomal RNA subunit at high temperatures (30°C), but
433
not at low temperatures (18°C) (Endo et al., 1987; Endo and Tsurugi, 1987; Moffat et al., 1992;
434
Allen et al., 2002). We previously validated the spatiotemporal effect of RicinCS inhibition in the
435
Drosophila brain using lateral clock neurons. We found that RicinCS can effectively inhibit ~80%
436
of protein synthesis at a permissive temperature (30°C), which is quickly reversed to normal levels
437
after shifting to a restrictive temperature (18°C) (Chen et al., 2012). This suggests a quick
438
restoration of ribosomal synthesis once RicinCS becomes inactive. While active RicinCS is a potent
439
cytotoxin for inhibiting protein synthesis, it tends not to be lethal, as RicinCS eventually inhibits its
440
own synthesis (Chen et al., 2012, Moffat et al., 1992; Allen et al., 2002). In the current experiments,
441
two copies of RicinCS was used to block protein synthesis. All flies were raised at 18°C to keep
442
RicinCS inactive. Before or after training at 18°C, the Gal4> UAS-ricinCS;UAS-ricinCS flies were
443
transferred to 30°C for 24 h to activate RicinCS, and then shifted back to 18°C for 1 h to inactivate
444
RicinCS before the experiments. Temporal control of RicinCS activation is indicated in the figures
445
for the relevant experiments. 446 Pharmacological treatment
391 For behavioural
413
testing, approximately 15–20 flies kept in a clear plastic syringe were directly exposed to UV light
414
at a distance of 5 cm for 1 h. Individual neurons expressing KAEDE were directly visualised
415
through an open window in the fly’s head capsule. Living samples were used because the signal-
416
to-noise ratio of green to red KAEDE is greatly reduced after chemical fixation. KAEDE neurons
417
were located in less than 5 s by a fast pre-scanning of red KAEDE excited by a 561-nm laser, to
418
avoid unnecessary fluorescence quenching of green KAEDE during repeated scanning. A single
419
optical slice through the MB -lobe tip was imaged at a resolution of 1024×1024 pixels under a
420
confocal microscope with a 40× C-Apochromat water-immersion objective lens (N.A. value 1.2,
421
working distance 220 μm). All brain samples in the experiment were imaged with the same optical
422
settings maximised for green and red KAEDE immediately before and after photoconversion,
423
respectively. In all cases, both green KAEDE (excited by a 488-nm laser) and red KAEDE (excited
424
by a 561-nm laser) were measured. By using the amount of red KAEDE as an internal standard to
425 Page 17 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Immunohistochemistry
448 Brains were dissected in phosphate-buffered saline (PBS), fixed with a commercial microwave
449
oven (2,450 MHz, 1100 Watts) in 4% paraformaldehyde on ice for 60 s three times, and then
450
immersed in 4% paraformaldehyde with 0.25% Triton X-100 for 60 s three times. After being
451
washed in PBS for 10 min at room temperature, brain samples were incubated in PBS containing
452
2% Triton X-100 (PBS-T) and 10% normal goat serum, and then degassed in a vacuum chamber to
453
expel tracheal air for four cycles (depressurizing to –70 mmHg and then holding for 10 min). Next,
454 Page 18 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint brain samples were blocked and penetrated in PBS-T at 4 °C overnight, and then incubated in PBS-
455
T containing (1) 1:40 mouse 4F3 anti-DLG antibody (Developmental Studies Hybridoma Bank,
456
University of Iowa) to label Disc large proteins, and (2) 1:500 mouse anti-CREBB 657 antibody
457
(from Jerry Yin (Tubon et al., 2013)) at 4 °C for 1 day. Samples were subsequently washed in PBS-
458
T three times and incubated in PBS-T containing 1:200 biotinylated goat anti-mouse IgG
459
(Molecular Probes) as the secondary antibody at 25 °C for 1 day. Brain samples were then washed
460
and incubated with 1:500 Alexa Fluor 635 streptavidin (Molecular Probes) at 25 °C for 1 day. 461
Finally, after extensive washing, immunolabeled brain samples were directly cleared for 5 min in
462
FocusClear, an aqueous solution that renders biological tissue transparent (Chiang et al., 2001),
463
and mounted between two cover slips separated by a spacer ring with a thickness of ~200 μm. 464
Sample brains were imaged under a Zeiss LSM 780 or 880 confocal microscope with a 40× C-
465
Apochromat water-immersion objective lens (N.A. value 1.2, working distance 220 μm). Immunohistochemistry
448 466
467
Statistics
468
All raw data were analysed parametrically with SigmaPlot 10.0 and SigmaStat 3.5 statistical
469
software. All the data including the behaviour Performance Index (PI) or KAEDE image (ΔF) were
470
evaluated via unpaired t-test (two groups) or one-way analysis of variance (ANOVA) (> two
471
groups). Data were evaluated with the Mann-Whitney Rank Sum Test in cases of unequal variances. 472
Data in all figures are presented as the mean SE. Experiments were replicated using multiple Gal4
473
drivers with equivalent expression patterns, and multiple effector genes and reagents that impact
474
shared cellular functions. 475
476
. CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. The copyright holder for this p
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint brain samples were blocked and penetrated in PBS-T at 4 °C overnight, and then incubated in PBS-
455
T containing (1) 1:40 mouse 4F3 anti-DLG antibody (Developmental Studies Hybridoma Bank,
456
University of Iowa) to label Disc large proteins, and (2) 1:500 mouse anti-CREBB 657 antibody
457
(from Jerry Yin (Tubon et al., 2013)) at 4 °C for 1 day. Samples were subsequently washed in PBS-
458
T three times and incubated in PBS-T containing 1:200 biotinylated goat anti-mouse IgG
459
(Molecular Probes) as the secondary antibody at 25 °C for 1 day. Brain samples were then washed
460
and incubated with 1:500 Alexa Fluor 635 streptavidin (Molecular Probes) at 25 °C for 1 day. 461
Finally, after extensive washing, immunolabeled brain samples were directly cleared for 5 min in
462
FocusClear, an aqueous solution that renders biological tissue transparent (Chiang et al., 2001),
463
and mounted between two cover slips separated by a spacer ring with a thickness of ~200 μm. 464
Sample brains were imaged under a Zeiss LSM 780 or 880 confocal microscope with a 40× C-
465
Apochromat water-immersion objective lens (N.A. value 1.2, working distance 220 μm). 466 brain samples were blocked and penetrated in PBS-T at 4 °C overnight, and then incubated in PBS-
455 All raw data were analysed parametrically with SigmaPlot 10.0 and SigmaStat 3.5 statistical
469
software. e thank the Bloomington Drosophila stock center, Vienna Drosophila RNAi Center (VDRC) and Kyoto Drosophila Genomics Resource Centers (DGRC) for fly stocks. We
also thank the Developmental Studies Hybridoma Bank for the antibodies. Funding: This work was financially supported by Funding: This work was financially supported by Funding: This work was financially supported by •
The Brain Research Center under the Higher Education Sprout Project co-funded by
the Ministry of Education and the Ministry of Science and Technology in Taiwan •
Yushan Scholar Program from the Ministry of Education in Taiwan •
Dart NeuroScience LLC in U.S.A. Immunohistochemistry
448 All the data including the behaviour Performance Index (PI) or KAEDE image (ΔF) were
470
evaluated via unpaired t-test (two groups) or one-way analysis of variance (ANOVA) (> two
471
groups). Data were evaluated with the Mann-Whitney Rank Sum Test in cases of unequal variances. 472
Data in all figures are presented as the mean SE. Experiments were replicated using multiple Gal4
473
drivers with equivalent expression patterns, and multiple effector genes and reagents that impact
474
shared cellular functions. 475 Page 19 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint We thank the Bloomington Drosophila stock center, Vienna Drosophila RNAi Center
478
(VDRC) and Kyoto Drosophila Genomics Resource Centers (DGRC) for fly stocks. We
479
also thank the Developmental Studies Hybridoma Bank for the antibodies. 480
Funding: This work was financially supported by
481
•
The Brain Research Center under the Higher Education Sprout Project co-funded by
482
the Ministry of Education and the Ministry of Science and Technology in Taiwan
483
•
Yushan Scholar Program from the Ministry of Education in Taiwan
484
•
Dart NeuroScience LLC in U.S.A. 485
Author contributions:
486
Conceived the project, analysed the data, and wrote the manuscript: C.C.C., J.S.D.,
487
T.T. and A.S.C. 488
Imaging experiments: H.W.L. 489
Behavioural experiments: C.C.C. and F.K.L. 490
Generated creb2-Gal4 transgenic flies: R.Y.J. and L.C. 491
492
Competing interests: All other authors declare they have no competing interests. 493
494
Data and materials availability: All data are available in the main text or the supplementary
495
materials. 496
497
References
498
Allen, M.J., O'Kane, C.J., and Moffat, K.G. (2002). Cell ablation using wild-type and cold-
499
sensitive ricin-A chain in Drosophila embryonic mesoderm. Genesis 34, 132-134. 500
. Immunohistochemistry
448 CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity Allen, M.J., O'Kane, C.J., and Moffat, K.G. (2002). Cell ablation using wild-type and cold-
499
sensitive ricin-A chain in Drosophila embryonic mesoderm. Genesis 34, 132-134.
500 Author contributions: Generated creb2-Gal4 transgenic flies: R.Y.J. and L.C. Allen, M.J., O'Kane, C.J., and Moffat, K.G. (2002). Cell ablation using wild-type and cold-
499
sensitive ricin-A chain in Drosophila embryonic mesoderm. Genesis 34, 132-134. 500 Page 20 of 55 Page 20 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Ando, R., Hama, H., Yamamoto-Hino, M., Mizuno, H., and Miyawaki, A. (2002). An optical
501 Ando, R., Hama, H., Yamamoto-Hino, M., Mizuno, H., and Miyawaki, A. (2002). An optical
501 marker based on the UV-induced green-to-red photoconversion of a fluorescent protein. Proc Natl
502
Acad Sci U S A 99, 12651-12656. 503 Aso, Y., Hattori, D., Yu, Y., Johnston, R.M., Iyer, N.A., Ngo, T.T., Dionne, H., Abbott, L.F.,
504 Axel, R., Tanimoto, H., et al. (2014). The neuronal architecture of the mushroom body provides a
505
logic for associative learning. Elife 3, e04577. 506 Baines, R.A. (2003). Postsynaptic protein kinase A reduces neuronal excitability in response to
507
increased synaptic excitation in the Drosophila CNS. J Neurosci 23, 8664-8672. 508 Barnstedt, O., Owald, D., Felsenberg, J., Brain, R., Moszynski, J.P., Talbot, C.B., Perrat, P.N.,
509
and Waddell S (2016) Memory-Relevant Mushroom Body Output Synapses Are Cholinergic
510 Barnstedt, O., Owald, D., Felsenberg, J., Brain, R., Moszynski, J.P., Talbot, C.B., Perrat, P.N.,
509 Barnstedt, O., Owald, D., Felsenberg, J., Brain, R., Moszynski, J.P., Talbot,
509 and Waddell, S. (2016). Memory-Relevant Mushroom Body Output Synapses Are Cholinergic. 510
Neuron 89, 1237-1247. 511 Blum, A.L., Li, W., Cressy, M., and Dubnau, J. (2009). Short- and long-term memory in
512 Bouzaiane, E., Trannoy, S., Scheunemann, L., Placais, P.Y., and Preat, T. (2015). Two
514
independent mushroom body output circuits retrieve the six discrete components of Drosophila
515
aversive memory. Cell Rep 11, 1280-1292. 516 Bouzaiane, E., Trannoy, S., Scheunemann, L., Placais, P.Y., and Preat, T. (2015). Author contributions: Two
514 independent mushroom body output circuits retrieve the six discrete components of Drosophila
515
aversive memory. Cell Rep 11, 1280-1292. 516 Chen, C.C., Wu, J.K., Lin, H.W., Pai, T.P., Fu, T.F., Wu, C.L., Tully, T., and Chiang, A.S. 517 Chen, C.C., Wu, J.K., Lin, H.W., Pai, T.P., Fu, T.F., Wu, C.L., Tully, T., an
517 (2012). Visualizing long-term memory formation in two neurons of the Drosophila brain. Science
518
335, 678-685. 519 Chiang, A.S., Liu, Y.C., Chiu, S.L., Hu, S.H., Huang, C.Y., and Hsieh, C.H. (2001). Three-
520
dimensional mapping of brain neuropils in the cockroach, Diploptera punctata. J Comp Neurol
521 Chiang, A.S., Liu, Y.C., Chiu, S.L., Hu, S.H., Huang, C.Y., and Hsieh, C.H. 520 dimensional mapping of brain neuropils in the cockroach, Diploptera punctata. J Comp Neurol
521
440, 1-11. 522 dimensional mapping of brain neuropils in the cockroach, Diploptera punctata. J Comp Neurol
521
440 1 11
522 440, 1-11. 522 Cognigni, P., Felsenberg, J., and Waddell, S. (2018). Do the right thing: neural network
523 mechanisms of memory formation, expression and update in Drosophila. Curr Opin Neurobiol
524 mechanisms of memory formation, expression and update in Drosophila. Curr Opin Neurobiol
524
49 51-58
525 mechanisms of memory formation, expression and update in Drosophila. Curr Opin Neurobiol
524
49, 51-58. 525 49, 51-58. 525 Page 21 of 55 Page 21 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is mad
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Davis, G.W., DiAntonio, A., Petersen, S.A., and Goodman, C.S. (1998). Postsynaptic PKA
526 Davis, G.W., DiAntonio, A., Petersen, S.A., and Goodman, C.S. (1998). Postsynaptic PKA
526 controls quantal size and reveals a retrograde signal that regulates presynaptic transmitter release
527
in Drosophila. Neuron 20, 305-315. 528 q
g
g
g
p
y
p
in Drosophila. Neuron 20, 305-315. 528 Davis, R.L. (2015). SnapShot: Olfactory Classical Conditioning of Drosophila. Cell 163, 524-524
529
e521. Author contributions: 530 Dubnau, J., and Chiang, A.S. (2013). Systems memory consolidation in Drosophila. Curr Opin
531
Neurobiol 23, 84-91. 532 Dubnau, J., Chiang, A.S., Grady, L., Barditch, J., Gossweiler, S., McNeil, J., Smith, P., Buldoc,
533
F., Scott, R., Certa, U., et al. (2003). The staufen/pumilio pathway is involved in Drosophila long-
534
term memory. Curr Biol 13, 286-296. 535 Dubnau, J., Chiang, A.S., Grady, L., Barditch, J., Gossweiler, S., McNeil, J., Smith, P., Buldoc,
533 F., Scott, R., Certa, U., et al. (2003). The staufen/pumilio pathway is involved in Drosophila long-
534
term memory. Curr Biol 13, 286-296. 535 Dubnau, J., Grady, L., Kitamoto, T., and Tully, T. (2001). Disruption of neurotransmission in
536
Drosophila mushroom body blocks retrieval but not acquisition of memory. Nature 411, 476-480. 537
Dudai, Y., Jan, Y.N., Byers, D., Quinn, W.G., and Benzer, S. (1976). dunce, a mutant of
538 Dubnau, J., Grady, L., Kitamoto, T., and Tully, T. (2001). Disruption of neurotransmission in
536
D
hil
h
b d bl
k
t i
l b t
t
i iti
f
N t
411 476 480
537 p
y
q
y
Dudai, Y., Jan, Y.N., Byers, D., Quinn, W.G., and Benzer, S. (1976). dunce, a mutant of
538 Dudai, Y., Jan, Y.N., Byers, D., Quinn, W.G., and Benzer, S. (1976). dunce, a mutant of
538 Drosophila deficient in learning. Proc Natl Acad Sci U S A 73, 1684-1688. 539 Drosophila deficient in learning. Proc Natl Acad Sci U S A 73, 1684-1688. 539 Endo, Y., Mitsui, K., Motizuki, M., and Tsurugi, K. (1987). The mechanism of action of ricin and
540
related toxic lectins on eukaryotic ribosomes. The site and the characteristics of the modification
541
in 28 S ribosomal RNA caused by the toxins. J Biol Chem 262, 5908-5912. 542 related toxic lectins on eukaryotic ribosomes. The site and the characteristics of the modification
541
in 28 S ribosomal RNA caused by the toxins. J Biol Chem 262, 5908-5912. 542 Endo, Y., and Tsurugi, K. (1987). RNA N-glycosidase activity of ricin A-chain. Mechanism of
543
action of the toxic lectin ricin on eukaryotic ribosomes. J Biol Chem 262, 8128-8130. 544 on of the toxic lectin ricin on eukaryotic ribosomes. J Biol Chem 262, 8128-8130. Fropf, R., Zhang, J., Tanenhaus, A.K., Fropf, W.J., Siefkes, E., and Yin, J.C. (2014). Time of day
545 influences memory formation and dCREB2 proteins in Drosophila. Author contributions: Front Syst Neurosci 8, 43. 546 Ge, X., Hannan, F., Xie, Z., Feng, C., Tully, T., Zhou, H., Xie, Z., and Zhong, Y. (2004). Notch
547 signaling in Drosophila long-term memory formation. Proc Natl Acad Sci U S A 101, 10172-
548
10176. 549 10176. 549 Page 22 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is m
The copyright holder for this prepri
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Hirano, Y., Ihara, K., Masuda, T., Yamamoto, T., Iwata, I., Takahashi, A., Awata, H., Nakamura,
550 Hirano, Y., Ihara, K., Masuda, T., Yamamoto, T., Iwata, I., Takahashi, A., Awata, H., Nakamura,
550 N., Takakura, M., Suzuki, Y., et al. (2016). Shifting transcriptional machinery is required for
551 N., Takakura, M., Suzuki, Y., et al. (2016). Shifting transcriptional machinery is required for
551 long-term memory maintenance and modification in Drosophila mushroom bodies. Nat Commun
552
7, 13471. 553 Hodge, J.J., Choi, J.C., O'Kane, C.J., and Griffith, L.C. (2005). Shaw potassium channel genes in
554
Drosophila. J Neurobiol 63, 235-254. 555 Hodge, J.J., Choi, J.C., O'Kane, C.J., and Griffith, L.C. (2005). Shaw potassium channel genes in
554 Drosophila. J Neurobiol 63, 235-254. 555 Huang, C., Zheng, X., Zhao, H., Li, M., Wang, P., Xie, Z., Wang, L., and Zhong, Y. (2012). A
556
permissive role of mushroom body alpha/beta core neurons in long-term memory consolidation in
557 Huang, C., Zheng, X., Zhao, H., Li, M., Wang, P., Xie, Z., Wang, L., and Zhong, Y. (2012). A
556 Huang, C., Zheng, X., Zhao, H., Li, M., Wang, P., Xie, Z., Wang, L., and Zh
556 permissive role of mushroom body alpha/beta core neurons in long-term memory consolidation in
557
Drosophila. Curr Biol 22, 1981-1989. 558 Isabel, G., Pascual, A., and Preat, T. (2004). Exclusive consolidated memory phases in
559
Drosophila. Science 304, 1024-1027. 560 Ito, K., Awano, W., Suzuki, K., Hiromi, Y., and Yamamoto, D. (1997). Author contributions: Proc Natl Acad Sci U S A 108, 13794-13799. 578 memory in Drosophila. Proc Natl Acad Sci U S A 108, 13794-13799. 578 Lin, H.H., Lai, J.S., Chin, A.L., Chen, Y.C., and Chiang, A.S. (2007). A map of olfactory
579 Lin, H.H., Lai, J.S., Chin, A.L., Chen, Y.C., and Chiang, A.S. (2007). A map of olfactory
579 representation in the Drosophila mushroom body. Cell 128, 1205-1217. 580 representation in the Drosophila mushroom body. Cell 128, 1205-1217. 580 Marty, T., Muller, B., Basler, K., and Affolter, M. (2000). Schnurri mediates Dpp-dependent
581
repression of brinker transcription. Nat Cell Biol 2, 745-749. 582 repression of brinker transcription. Nat Cell Biol 2, 745-749. 582 repression of brinker transcription. Nat Cell Biol 2, 745-749. 582 McGuire, S.E., Le, P.T., and Davis, R.L. (2001). The role of Drosophila mushroom body
583
i
li
i
lf
t
S i
293 1330 1333
584 McGuire, S.E., Le, P.T., and Davis, R.L. (2001). The role of Drosophila mushroom body
583 McGuire, S.E., Le, P.T., and Davis, R.L. (2001). The role of Drosophila mushroom body
583
signaling in olfactory memory. Science 293, 1330-1333. 584 signaling in olfactory memory. Science 293, 1330-1333. 584 signaling in olfactory memory. Science 293, 1330-1333. signaling in olfactory memory. Science 293, 1330-1333. 584 McGuire, S.E., Le, P.T., Osborn, A.J., Matsumoto, K., and Davis, R.L. (2003). Spatiotemporal
585
rescue of memory dysfunction in Drosophila. Science 302, 1765-1768. 586 rescue of memory dysfunction in Drosophila. Science 302, 1765-1768. 586 Miyashita, T., Kikuchi, E., Horiuchi, J., and Saitoe, M. (2018). Long-Term Memory Engram
587 Miyashita, T., Kikuchi, E., Horiuchi, J., and Saitoe, M. (2018). Long-Term Memory Engram
587
Cells Are Established by c-Fos/CREB Transcriptional Cycling. Cell Rep 25, 2716-2728 e2713. 588 Miyashita, T., Kikuchi, E., Horiuchi, J., and Saitoe, M. (2018). Long Term Memory Engram
587
Cells Are Established by c-Fos/CREB Transcriptional Cycling. Cell Rep 25, 2716-2728 e2713. 588 Cells Are Established by c-Fos/CREB Transcriptional Cycling. Cell Rep 25, 2716-2728 e2713. 588 Moffat, K.G., Gould, J.H., Smith, H.K., and O'Kane, C.J. (1992). Inducible cell ablation in
589
Drosophila by cold-sensitive ricin A chain. Development 114, 681-687. 590 Moffat, K.G., Gould, J.H., Smith, H.K., and O'Kane, C.J. (1992). Inducible
589 Drosophila by cold-sensitive ricin A chain. Development 114, 681-687. 590 Pai, T.P., Chen, C.C., Lin, H.H., Chin, A.L., Lai, J.S., Lee, P.T., Tully, T., a
591 (2013). Author contributions: The Drosophila
561
mushroom body is a quadruple structure of clonal units each of which contains a virtually
562
identical set of neurones and glial cells. Development 124, 761-771. 563 Ito, K., Awano, W., Suzuki, K., Hiromi, Y., and Yamamoto, D. (1997). The Drosophila
561 mushroom body is a quadruple structure of clonal units each of which contains a virtually
562
identical set of neurones and glial cells. Development 124, 761-771. 563 identical set of neurones and glial cells. Development 124, 761-771. 563 Joiner, W.J., Crocker, A., White, B.H., and Sehgal, A. (2006). Sleep in Drosophila is regulated by
564
adult mushroom bodies. Nature 441, 757-760. 565 Knapek, S., Sigrist, S., and Tanimoto, H. (2011). Bruchpilot, a synaptic active zone protein for
566
anesthesia-resistant memory. J Neurosci 31, 3453-3458. 567 Kotoula, V., Moressis, A., Semelidou, O., and Skoulakis, E.M.C. (2017). Drk-mediated signaling
568
to Rho kinase is required for anesthesia-resistant memory in Drosophila. Proc Natl Acad Sci U S
569
A 114, 10984-10989. 570 Krashes, M.J., Keene, A.C., Leung, B., Armstrong, J.D., and Waddell, S. (2007). Sequential use
571 Krashes, M.J., Keene, A.C., Leung, B., Armstrong, J.D., and Waddell, S. (2007). Sequential use
571
of mushroom body neuron subsets during Drosophila odor memory processing. Neuron 53, 103-
572
115. 573 115. 573 Page 23 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is m
The copyright holder for this prepr
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Lee, P.T., Lin, G., Lin, W.W., Diao, F., White, B.H., and Bellen, H.J. (2018). A kinase-dependent
574 feedforward loop affects CREBB stability and long term memory formation. Elife 7. 575 Lee, P.T., Lin, H.W., Chang, Y.H., Fu, T.F., Dubnau, J., Hirsh, J., Lee, T., and Chiang, A.S. 576 (2011). Serotonin-mushroom body circuit modulating the formation of anesthesia-resistant
577 (2011). Serotonin-mushroom body circuit modulating the formation of anesthesia-resistant
577
memory in Drosophila. Proc Natl Acad Sci U S A 108, 13794-13799. 578 memory in Drosophila. Author contributions: 604
Scheunemann, L., Jost, E., Richlitzki, A., Day, J.P., Sebastian, S., Thum, A.S., Efetova, M.,
605
Davies, S.A., and Schwarzel, M. (2012). Consolidated and labile odor memory are separately
606
encoded within the Drosophila brain. J Neurosci 32, 17163-17171. 607 Davies, S.A., and Schwarzel, M. (2012). Consolidated and labile odor memory are separately
606
encoded within the Drosophila brain. J Neurosci 32, 17163-17171. 607 encoded within the Drosophila brain. J Neurosci 32, 17163-17171. 607 Scheunemann, L., Placais, P.Y., Dromard, Y., Schwarzel, M., and Preat, T. (2018). Dunce
608
Phosphodiesterase Acts as a Checkpoint for Drosophila Long-Term Memory in a Pair of
609
Serotonergic Neurons. Neuron 98, 350-365 e355. 610 Scheunemann, L., Placais, P.Y., Dromard, Y., Schwarzel, M., and Preat, T. (2018). Dunce
608
Phosphodiesterase Acts as a Checkpoint for Drosophila Long-Term Memory in a Pair of
609 Scheunemann, L., Placais, P.Y., Dromard, Y., Schwarzel, M., and Preat, T. (2018). Dunce
608 Phosphodiesterase Acts as a Checkpoint for Drosophila Long-Term Memory in a Pair of
609
Serotonergic Neurons. Neuron 98, 350-365 e355. 610 Scholz-Kornehl, S., and Schwarzel, M. (2016). Circuit Analysis of a Drosophila Dopamine Type
611
2 Receptor That Supports Anesthesia-Resistant Memory. J Neurosci 36, 7936-7945. 612 Sejourne, J., Placais, P.Y., Aso, Y., Siwanowicz, I., Trannoy, S., Thoma, V., Tedjakumala, S.R.,
613 Rubin, G.M., Tchenio, P., Ito, K., et al. (2011). Mushroom body efferent neurons responsible for
614
aversive olfactory memory retrieval in Drosophila. Nat Neurosci 14, 903-910. 615 aversive olfactory memory retrieval in Drosophila. Nat Neurosci 14, 903-910. Shyu, W.H., Lee, W.P., Chiang, M.H., Chang, C.C., Fu, T.F., Chiang, H.C.,
616 C.L. (2019). Electrical synapses between mushroom body neurons are critical for consolidated
617
memory retrieval in Drosophila. PLoS Genet 15, e1008153. 618 Smolik, S.M., Rose, R.E., and Goodman, R.H. (1992). A cyclic AMP-responsive element-binding
619 transcriptional activator in Drosophila melanogaster, dCREB-A, is a member of the leucine
620
i
f
il
M l C ll Bi l 12 4123 4131 transcriptional activator in Drosophila melanogaster, dCREB-A, is a member of the leucine
620
zipper family. Mol Cell Biol 12, 4123-4131. 621 p
p
g
,
,
zipper family. Mol Cell Biol 12, 4123-4131. 621 zipper family. Mol Cell Biol 12, 4123-4131. 621 Stoleru, D., Peng, Y., Agosto, J., and Rosbash, M. (2004). Coupled oscillators control morning
622
and evening locomotor behaviour of Drosophila Nature 431 862 868
623 Stoleru, D., Peng, Y., Agosto, J., and Rosbash, M. (2004). Author contributions: Drosophila ORB protein in two mushroom body output neurons is necessary for long-
592
term memory formation. Proc Natl Acad Sci U S A 110, 7898-7903. 593 (2013). Drosophila ORB protein in two mushroom body output neurons i
592 term memory formation. Proc Natl Acad Sci U S A 110, 7898-7903. 593 Pavlopoulos, E., Anezaki, M., and Skoulakis, E.M. (2008). Neuralized is expressed in the
594 Pavlopoulos, E., Anezaki, M., and Skoulakis, E.M. (2008). Neuralized is expressed in the
594
alpha/beta lobes of adult Drosophila mushroom bodies and facilitates olfactory long-term
595 alpha/beta lobes of adult Drosophila mushroom bodies and facilitates olfactory long-term
595
memory formation. Proc Natl Acad Sci U S A 105, 14674-14679. 596 Perisse, E., Burke, C., Huetteroth, W., and Waddell, S. (2013). Shocking revelations and
597 Perisse, E., Burke, C., Huetteroth, W., and Waddell, S. (2013). Shocking revelations and
597
h i
i
h
d
f D
hil
lf
C
Bi l 23 R752 763 saccharin sweetness in the study of Drosophila olfactory memory. Curr Biol 23, R752-763. 598 Page 24 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Placais, P.Y., Trannoy, S., Isabel, G., Aso, Y., Siwanowicz, I., Belliart-Guerin, G., Vernier, P.,
599 Placais, P.Y., Trannoy, S., Isabel, G., Aso, Y., Siwanowicz, I., Belliart-Guerin, G., Vernier, P.,
599 Birman, S., Tanimoto, H., and Preat, T. (2012). Slow oscillations in two pairs of dopaminergic
600
neurons gate long-term memory formation in Drosophila. Nat Neurosci 15, 592-599. 601 Presente, A., Boyles, R.S., Serway, C.N., de Belle, J.S., and Andres, A.J. (2004). Notch is
602
required for long-term memory in Drosophila. Proc Natl Acad Sci U S A 101, 1764-1768. 603
Quinn, W.G., and Dudai, Y. (1976). Memory phases in Drosophila. Nature 262, 576-577. 604
Scheunemann L Jost E Richlitzki A Day J P Sebastian S Thum A S Efetova M
605 Presente, A., Boyles, R.S., Serway, C.N., de Belle, J.S., and Andres, A.J. (2004). Notch is
602
required for long-term memory in Drosophila. Proc Natl Acad Sci U S A 101, 1764-1768. 603
Quinn, W.G., and Dudai, Y. (1976). Memory phases in Drosophila. Nature 262, 576-577. aka, N.K., Tanimoto, H., and Ito, K. (2008). Neuronal assemblies of the Drosophila Tanaka, N.K., Tanimoto, H., and Ito, K. (2008). Neuronal assemblies of the Drosophila
624 mushroom body. J Comp Neurol 508, 711-755. 625 mushroom body. J Comp Neurol 508, 711-755. 625 Tonoki, A., and Davis, R.L. (2015). Aging impairs protein-synthesis-dependent long-term
626
memory in Drosophila. J Neurosci 35, 1173-1180. 627 memory in Drosophila. J Neurosci 35, 1173-1180. 627 Tubon, T.C., Jr., Zhang, J., Friedman, E.L., Jin, H., Gonzales, E.D., Zhou, H., Drier, D., Gerstner,
628
J.R., Paulson, E.A., Fropf, R., et al. (2013). dCREB2-mediated enhancement of memory
629
formation. J Neurosci 33, 7475-7487. 630 Tubon, T.C., Jr., Zhang, J., Friedman, E.L., Jin, H., Gonzales, E.D., Zhou, H., Drier, D., Gerstner,
628
J.R., Paulson, E.A., Fropf, R., et al. (2013). dCREB2-mediated enhancement of memory
629
formation. J Neurosci 33, 7475-7487. 630 Tully, T. (1996). Discovery of genes involved with learning and memory: an experimental
631
synthesis of Hirschian and Benzerian perspectives. Proc Natl Acad Sci U S A 93, 13460-13467. 632
Tully, T., Boynton, S., Brandes, C., Dura, J.M., Mihalek, R., Preat, T., and Villella, A. (1990). 633
Genetic dissection of memory formation in Drosophila melanogaster. Cold Spring Harb Symp
634
Quant Biol 55, 203-211. 635 Tully, T. (1996). Discovery of genes involved with learning and memory: an experimental
631
synthesis of Hirschian and Benzerian perspectives. Proc Natl Acad Sci U S A 93, 13460-13467. 632
Tully, T., Boynton, S., Brandes, C., Dura, J.M., Mihalek, R., Preat, T., and Villella, A. (1990). 633
Genetic dissection of memory formation in Drosophila melanogaster. Cold Spring Harb Symp
634
Quant Biol 55, 203-211. 635 Tully, T., Preat, T., Boynton, S.C., and Del Vecchio, M. (1994). Genetic dissection of
636
consolidated memory in Drosophila. Cell 79, 35-47. 637 Tully, T., Preat, T., Boynton, S.C., and Del Vecchio, M. (1994). Genetic dissection of
636
lid t d
i D
hil
C ll 79 35 47 consolidated memory in Drosophila. Cell 79, 35-47. 637 Tully, T., and Quinn, W.G. (1985). Classical conditioning and retention in normal and mutant
638
Drosophila melanogaster. J Comp Physiol A 157, 263-277. 639 Drosophila melanogaster. J Comp Physiol A 157, 263-277. 639 Venken, K.J., Simpson, J.H., and Bellen, H.J. (2011). Genetic manipulation of genes and cells in
640
the nervous system of the fruit fly. Neuron 72, 202-230. 641 Wu, C.L., Shih, M.F., Lee, P.T., and Chiang, A.S. (2013). An octopamine-mushroom body circuit
642
modulates the formation of anesthesia-resistant memory in Drosophila. Author contributions: Coupled oscillators control morning
622 and evening locomotor behaviour of Drosophila. Nature 431, 862-868. 623 and evening locomotor behaviour of Drosophila. Nature 431, 862-868. 623 Page 25 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Tanaka, N.K., Tanimoto, H., and Ito, K. (2008). Neuronal assemblies of the Drosophila
624 aka, N.K., Tanimoto, H., and Ito, K. (2008). Neuronal assemblies of the Drosophila Curr Biol 23, 2346-2354. 643
Wu, C.L., Xia, S., Fu, T.F., Wang, H., Chen, Y.H., Leong, D., Chiang, A.S., and Tully, T. (2007). 644
Specific requirement of NMDA receptors for long-term memory consolidation in Drosophila
645
ellipsoid body Nat Neurosci 10 1578-1586
646 Wu, C.L., Shih, M.F., Lee, P.T., and Chiang, A.S. (2013). An octopamine-mushroom body circuit
642
modulates the formation of anesthesia-resistant memory in Drosophila. Curr Biol 23, 2346-2354. 643
Wu, C.L., Xia, S., Fu, T.F., Wang, H., Chen, Y.H., Leong, D., Chiang, A.S., and Tully, T. (2007). 644 Specific requirement of NMDA receptors for long-term memory consolidation in Drosophila
645 Specific requirement of NMDA receptors for long-term memory consolidation in Drosophila
645
ellipsoid body. Nat Neurosci 10, 1578-1586. 646 ellipsoid body. Nat Neurosci 10, 1578-1586. 646 Page 26 of 55 Page 26 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. I
The copyright holder for this p
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Wu, J.K., Tai, C.Y., Feng, K.L., Chen, S.L., Chen, C.C., and Chiang, A.S. (2017). Long-term
647 Wu, J.K., Tai, C.Y., Feng, K.L., Chen, S.L., Chen, C.C., and Chiang, A.S. (2017). Long-term
647 memory requires sequential protein synthesis in three subsets of mushroom body output neurons
648
in Drosophila. Sci Rep 7, 7112. 649 Yang, C.H., Shih, M.F., Chang, C.C., Chiang, M.H., Shih, H.W., Tsai, Y.L., Chiang, A.S., Fu,
650 T.F., and Wu, C.L. (2016). Additive Expression of Consolidated Memory through Drosophila
651
Mushroom Body Subsets. PLoS Genet 12, e1006061. 652 Mushroom Body Subsets. PLoS Genet 12, e1006061. 652 Yin, J.C., Del Vecchio, M., Zhou, H., and Tully, T. (1995a). CREB as a memory modulator:
653 in, J.C., Del Vecchio, M., Zhou, H., and Tully, T. (1995a). CREB as a memory mo induced expression of a dCREB2 activator isoform enhances long-term memory in Drosophila. 654
Cell 81, 107-115. 655 Yin, J.C., Wallach, J.S., Del Vecchio, M., Wilder, E.L., Zhou, H., Quinn, W.G., and Tully, T. 656 (1994). aka, N.K., Tanimoto, H., and Ito, K. (2008). Neuronal assemblies of the Drosophila Induction of a dominant negative CREB transgene specifically blocks long-term memory
657
in Drosophila. Cell 79, 49-58. 658 in Drosophila. Cell 79, 49-58. 658 Yin, J.C., Wallach, J.S., Wilder, E.L., Klingensmith, J., Dang, D., Perrimon, N., Zhou, H., Tully,
659
T., and Quinn, W.G. (1995b). A Drosophila CREB/CREM homolog encodes multiple isoforms,
660 Yin, J.C., Wallach, J.S., Wilder, E.L., Klingensmith, J., Dang, D., Perrimon, N., Zhou, H., Tully,
659
T and Quinn W G (1995b) A Drosophila CREB/CREM homolog encodes multiple isoforms
660 Yin, J.C., Wallach, J.S., Wilder, E.L., Klingensmith, J., Dang, D., Perrimon, N., Zhou, H., Tully,
659 Yin, J.C., Wallach, J.S., Wilder, E.L., Klingensmith, J., Dang, D., Perrimon, N., Zhou, H., Tully,
659
T., and Quinn, W.G. (1995b). A Drosophila CREB/CREM homolog encodes multiple isoforms,
660
including a cyclic AMP-dependent protein kinase-responsive transcriptional activator and
661
antagonist. Mol Cell Biol 15, 5123-5130. 662 T., and Quinn, W.G. (1995b). A Drosophila CREB/CREM homolog encodes multiple isoforms,
660
including a cyclic AMP-dependent protein kinase-responsive transcriptional activator and
661
antagonist. Mol Cell Biol 15, 5123-5130. 662 Yu, D., Akalal, D.B., and Davis, R.L. (2006). Drosophila alpha/beta mushroom body neurons
663
form a branch-specific, long-term cellular memory trace after spaced olfactory conditioning. 664
Neuron 52, 845-855. 665 Zhang, J., Tanenhaus, A.K., Davis, J.C., Hanlon, B.M., and Yin, J.C. (2015). Spatio-temporal in
666
vivo recording of dCREB2 dynamics in Drosophila long-term memory processing. Neurobiol
667
Learn Mem 118, 80-88. 668 vivo recording of dCREB2 dynamics in Drosophila long-term memory processing. Neurobiol
667 Learn Mem 118, 80-88. 668 Zhang, S.X., Rogulja, D., and Crickmore, M.A. (2019). Recurrent Circuitry Sustains Drosophila
669
Courtship Drive While Priming Itself for Satiety Curr Biol 29 3216 3228 e3219
670 Zhang, S.X., Rogulja, D., and Crickmore, M.A. (2019). Recurrent Circuitry Sustains Drosophila
669 Zhang, S.X., Rogulja, D., and Crickmore, M.A. (2019). Recurrent Circuitry Sustains Drosophila
669
Courtship Drive While Priming Itself for Satiety. Curr Biol 29, 3216-3228 e3219. 670 Courtship Drive While Priming Itself for Satiety. Curr Biol 29, 3216-3228 e3219. 670 Courtship Drive While Priming Itself for Satiety. Curr Biol 29, 3216-3228 e3219. 670 Page 27 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. aka, N.K., Tanimoto, H., and Ito, K. (2008). Neuronal assemblies of the Drosophila ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Zhu, S., Chiang, A.S., and Lee, T. (2003). Development of the Drosophila mushroom bodies:
elaboration, remodeling and spatial organization of dendrites in the calyx. Development 130,
2603-2610. . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Zhu, S., Chiang, A.S., and Lee, T. (2003). Development of the Drosophila mushroom bodies:
671
elaboration, remodeling and spatial organization of dendrites in the calyx. Development 130,
672
2603-2610. 673
674
675
676
677
678
679
680
681
Figures
682 Zhu, S., Chiang, A.S., and Lee, T. (2003). Development of the Drosophila mushroom bodies:
1 Figures Page 28 of 55
Figures
Figure 1 Page 28 of 55 Page 28 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 685 685
Figure 2
686 Figure 2 Page 29 of 55 Page 29 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. aka, N.K., Tanimoto, H., and Ito, K. (2008). Neuronal assemblies of the Drosophila It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint .
CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021.
;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 687
Figure 3
688 687
Figure 3
688 87
Figure 3
88 Figure 3 Page 30 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 689
Figure 4
690 689
Figure 4
690 689 Page 31 of 55 Page 31 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Figure 5 691
Figure 5
692 691 Page 32 of 55 Page 32 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Figure 6 693
Figure 6
694 693 Figure 6 Page 33 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. 687
Figure 3
688 It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 695
Figure 7
696 Figure 7 695
Figure 7
696 5
Figure 7
6 Figure 7 Page 34 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 7 697 Figure 8 698 Page 35 of 55 Page 35 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Figure 9 699 Figure 9 Page 36 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. Figure 1. CSHL memory Gal4 patterns in which protein synthesis inhibition has no net
708 effect on LTM formation. (A) Full expression patterns that include MB and DAL neurons (top). 709
(B) Expression restricted from DAL neurons by cry-Gal80 inhibition of Gal4 (center). (C)
710
Expression restricted from MB neurons by MB-Gal80 inhibition of Gal4 (bottom). DAL (arrow)
711
and MB (arrowhead). (D-F) Protein synthesis inhibition in MB neurons enhances LTM
712
formation. Effects of memory circuit Gal4-targeted RicinCS on 1-day memory after three spaced
713
training cycles (3xS), compared with wild-type (+/+) controls. Cold-sensitive RicinCS blocks
714
protein synthesis at the permissive temperature (30 °C) between training and testing. (D) Protein
715
synthesis inhibition in all Gal4-expressing elements of the memory circuit has no net effect on
716
LTM. (E) RicinCS expression and blocking of protein synthesis in MB but not in DAL neurons
717
(where cry-Gal80 inhibits Gal4), enhances 1-day memory in all seven Gal4 genotypes. (F) By
718
contrast, RicinCS expression and blocking of protein synthesis in DAL but not in MB neurons
719
(where MB-Gal80 inhibits Gal4) has no effect on LTM. In all figures, temperature control
720
schedules are indicated (top). Memory performance indices are calculated as the normalised
721
percent avoidance of shock-paired odour. Bars represent mean ± SE, n = 8/bar unless stated
722
otherwise. *, P < 0.05; **, P < 0.01, ***, P < 0.001. All genotypes are listed in table supplement
723
2. 724 687
Figure 3
688 It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 701
Figure 10
702
. CC-BY 4.0 International license
available under a Figure 10 701 706 Page 37 of 55 Page 37 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Figure 4. Blocking early neuron signaling enhances LTM formation. Effects of early
740 Gal4-targeted Dynamints (shits) on 1-day memory after 3xS training. (A) Blocking signaling
741
in the first 8-h after training enhances 1-day memory, whereas blocking signaling during
742
subsequent 8-h windows has no effect, compared with the unexpressed shits/+ control (also see
743
figure supplement 4). (B) Similarly, blocking signaling in the first 8-h after 3xS training enhances
744
4-day memory (left), but not after 3xM training (center) or after 3xS training of the Gal4 driver or
745
shits transgene alone (right). (C) Blocking early signaling using two additional Gal4 patterns
746
confirms this inhibitory effect and enhancement of 1-day memory (left). Memory is unaffected in
747
flies held at the permissive temperature (18 °C) after 3xS training (right). 748 Figure 3. Protein synthesis in early neurons after subthreshold training antagonizes
733 LTM formation. Blocking protein synthesis in all MB (OK107), all (c739) and early
734
(VT26665) neurons 0-6 h after 1x training enhances 1-day memory compared with wild-type
735
(+/+) controls (related to Figure 2; also compare with Figure 4A). (A) Memory enhancement is
736
significant in all groups with RicinCS activity 0-12 h, 2-14 h and 4-16 h after training but not for
737
later time windows. (B) Memory enhancement is significant in all groups with RicinCS activity 0-
738
3 h and 3-6 h but not 6-9 h after training. 739 copy of RicinCS is insufficient. (C) Expression of inactive RicinCS in neurons at the restrictive
731
temperature (18 °C) has no effect on memory after 1x training.
732 copy of RicinCS is insufficient. (C) Expression of inactive RicinCS in neurons at the restrictive
731
temperature (18 °C) has no effect on memory after 1x training. 732 copy of RicinCS is insufficient. (C) Expression of inactive RicinCS in neurons at the restrictive
731
temperature (18 °C) has no effect on memory after 1x training. 732 Figure 2. Protein synthesis inhibition in early MB neurons enhances
725 Effects of MB Gal4-targeted RicinCS on memory after sub-threshold training compared with wild-
726
type controls (see MB Gal4 expression patterns in figure supplement 2A) (A) Protein synthesis
727
inhibition in MB Gal4 patterns that include early neurons enhance 1-day (left) and 4-day
728
(right) memory after 3xS training. (B) Two copies of RicinCS expressed in early neurons are
729
necessary to enhance 1-day (left) and 4-day (right) memory after 1x training, whereas only one
730 Page 38 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Figure 5. Neural membrane excitability in early neurons bi-directionally regulates LTM
749 Page 39 of 55
formation. MB early Gal4-targeted expression of K+ and Na+ channel proteins is induced at
750
the restrictive temperature for the tub-Gal80ts inhibitor (30 °C) from five days before training
751
until testing. (A) ShawDN and NaChBac overexpression increase neural activity and impair 1-day
752
memory after 10xS training (left). The same LTM is similarly blocked by systemic protein
753
synthesis inhibition induced by CXM feeding (right) (also see figure supplement 5). (B) In
754 Page 39 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint contrast, Shaw and Kir2.1::GFP overexpression decrease neural activity and enhance1-day
755
memory after only 1x training (left), which endures for at least 4 days (center). Memory is
756
unaffected in these flies held at the permissive temperature for tub-Gal80ts (18 °C) after 1x
757
training (right). (C) Decreasing neural activity as in (B) does not affect 1-day memory after 10xS
758
training. 759 Figure 6. Modulation of Rutabaga (AC) in early neurons bi-directionally regulates LTM
760 Figure 6. Modulation of Rutabaga (AC) in early neurons bi directionally regulates LTM
760
formation. (A) Overexpressing AC in early neurons enhances 1-day memory after only 1x
761
training (left) and lasts at least four days (left center). Memory is unaffected in these flies held at
762
the permissive temperature for tub-Gal80ts (18 °C) after 1x training (right center). One-day
763
memory is unaffected in these flies held at 30 °C after 10xS training (right). Gal4-targeted rut+
764
overexpression is induced at the restrictive temperature for tub-Gal80ts (30 °C) from five days
765
prior to training until testing. (B) By contrast, adult-stage specific RNAi down-regulation of AC
766
in early neurons (two independent RNAi lines) impairs 1-day memory after 10xS training
767
(left) (also see figure supplement 6). The same LTM is similarly blocked by systemic protein
768
synthesis inhibition induced by cycloheximide (CXM) feeding (right). 769 formation. (A) Overexpressing AC in early neurons enhances 1-day memory after only 1x
761
training (left) and lasts at least four days (left center). Memory is unaffected in these flies held at
762
the permissive temperature for tub-Gal80ts (18 °C) after 1x training (right center). One-day
763
memory is unaffected in these flies held at 30 °C after 10xS training (right). Gal4-targeted rut+
764
overexpression is induced at the restrictive temperature for tub-Gal80ts (30 °C) from five days
765
prior to training until testing. (B) By contrast, adult-stage specific RNAi down-regulation of AC
766
in early neurons (two independent RNAi lines) impairs 1-day memory after 10xS training
767
(left) (also see figure supplement 6). The same LTM is similarly blocked by systemic protein
768
synthesis inhibition induced by cycloheximide (CXM) feeding (right). 769 blocked by systemic protein synthesis inhibition induced by cycloheximide (CXM) feeding
779
(right).
780 blocked by systemic protein synthesis inhibition induced by cycloheximide (CXM) feeding
779
(right). 780 Figure 8. Modulation of CREBB protein in early neurons bi-directionally regulates
781 Figure 8. Modulation of CREBB protein in early neurons bi-directionally regulates
781 Figure 8. Modulation of CREBB protein in early neurons bi-directionally regulates
781
LTM formation. (A) Overexpressing CREBB proteins in early neurons enhances 1-day
782
memory after only 1x training (left) and lasts at least four days (center). Gal4-targeted crebB
783
overexpression is induced at the restrictive temperature for tub-Gal80ts (30 °C) from five days
784
before training until testing. Memory is unaffected in these flies held at the permissive
785
temperature for tub-Gal80ts (18 °C) after 1x training (right). One-day memory is also unaffected
786
in these flies held at 30 °C after 10xS training. (B) By contrast, adult-stage specific RNAi down-
787
regulation of CREBB proteins in early neurons (with two independent RNAi constructs)
788
impairs 1-day memory after 10xS training (left) (also see figure supplement 6). The same LTM is
789
similarly blocked by systemic protein synthesis inhibition induced by CXM feeding (right). 790 LTM formation. (A) Overexpressing CREBB proteins in early neurons enhances 1-day
782
memory after only 1x training (left) and lasts at least four days (center). Gal4-targeted crebB
783
overexpression is induced at the restrictive temperature for tub-Gal80ts (30 °C) from five days
784
before training until testing. Memory is unaffected in these flies held at the permissive
785
temperature for tub-Gal80ts (18 °C) after 1x training (right). One-day memory is also unaffected
786
in these flies held at 30 °C after 10xS training. (B) By contrast, adult-stage specific RNAi down-
787
regulation of CREBB proteins in early neurons (with two independent RNAi constructs)
788
impairs 1-day memory after 10xS training (left) (also see figure supplement 6). The same LTM i
789
similarly blocked by systemic protein synthesis inhibition induced by CXM feeding (right). 790 ure 7. Modulation of PKA in early neurons bi-directionally regulates LTM formation. (A) Overexpressing constitutively active PKAact1 in early neurons enhances 1-day
771
memory after only 1x training (left) and lasts at least four days (left center). Memory is unaffected
772
in these flies held at the permissive temperature for tub-Gal80ts (18 °C) after 1x training (right
773
center). One-day memory is unaffected in these flies held at 30 °C after 10xS training (right). 774
Gal4-targeted PKAact1 overexpression is induced at the restrictive temperature for tub-Gal80ts (30
775
°C) from five days prior to training until testing. (B) By contrast, adult-stage specific RNAi
776
down-regulation of PKA in early neurons (two independent RNAi lines) impairs 1-day
777
memory after 10xS training (left) (also see figure supplement 6). The same LTM is similarly
778 formation. (A) Overexpressing constitutively active PKAact1 in early neurons enhances 1-day
771
memory after only 1x training (left) and lasts at least four days (left center). Memory is unaffected
772
in these flies held at the permissive temperature for tub-Gal80ts (18 °C) after 1x training (right
773
center). One-day memory is unaffected in these flies held at 30 °C after 10xS training (right). 774
Gal4-targeted PKAact1 overexpression is induced at the restrictive temperature for tub-Gal80ts (30
775
°C) from five days prior to training until testing. (B) By contrast, adult-stage specific RNAi
776
down-regulation of PKA in early neurons (two independent RNAi lines) impairs 1-day
777
memory after 10xS training (left) (also see figure supplement 6). The same LTM is similarly
778 Page 40 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint blocked by systemic protein synthesis inhibition induced by cycloheximide (CXM) feeding
779
(right). 780 Figure 9. Spaced training activates crebB transcription. (A) CREBB expression visualized in
791 Figure 9. Spaced training activates crebB transcription. (A) CREBB expression visualized in
791
dissected brains with crebB-Gal4 driven UAS-mCD8::GFP (green), counterstained with DLG-
792
antibody immunostaining (magenta), and viewed under a confocal microscope. Cross sections of
793
vertical and horizontal MB lobes (center) show more prominent expression in neurons than in
794
′′ and neurons (labeled). Scale bar = 10 m. (B-C) Promotor activation of crebB 24 h after
795
training reported by de novo Kaede synthesis, estimated by the ratio of new (green, 488 nm) and
796
preexisting (red, 561 nm) protein (% ∆ F/F0). For each brain, single optical slices through the
797
MB -lobe tip or ellipsoid body (EB) were imaged under identical conditions. (B) Spaced
798
training stimulates crebB activity preferentially in the -lobe, in comparison with EB controls. 799
(C) A minimum of 5xS training cycles are necessary to observe Kaede synthesis reflecting crebB
800
activity. Bars represent mean ± SE, n 8. 801 Page 41 of 55 Page 41 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Figure 10. Shn in early neurons regulates CREBB dependent LTM formation. (A)
802 Overexpressing Shn proteins in early neurons enhances 1-day memory after only 1x training
803
(left) and lasts at least four days ((left center). Gal4-targeted shn+ overexpression is induced at the
804
restrictive temperature for tub-Gal80ts (30 °C) from five days before training until testing. 805
Memory is unaffected in these flies held at the permissive temperature for tub-Gal80ts (18 °C)
806
after 1x training (right center) (also see figure supplement 7). One-day memory is also unaffected
807
in these flies held at 30 °C after 10xS training (right). (B) In contrast, adult-stage specific RNAi
808
down-regulation of Shn in early neurons (with two independent RNAi constructs) impairs 1-
809
day memory after 10xS training (left) (also see figure supplement 6). The same LTM is similarly
810
blocked by systemic protein synthesis inhibition induced by CXM feeding (right). (C) Down-
811
regulating Shn but co-overexpressing CREBB in early neurons does not impair 1-day
812
memory after 10xS training. (D) One-day memory is enhanced after only 1x training (left) and
813
lasts at least four days (center). Memory is unaffected in flies held at the permissive temperature
814
for tub-Gal80ts (18 °C) after 1x training (right). 815 816
817
818
819
820
821
822
823
824
825
826
827
828
829
830 Page 42 of 55 Page 42 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Page 43 of 55
e supplement 1. Protein synthesis inhibition in MB and DAL neurons blocks LTM formation. Figure 10. Shn in early neurons regulates CREBB dependent LTM formation. (A)
802 (A) Effec
mory circuit Gal4-targeted RicinCS on 1-day memory after ten spaced training cycles (10xS), compared with
ype (+/+) controls. Cold-sensitive RicinCS blocks protein synthesis at the permissive temperature (30 °C) Figure supplement 1. Protein synthesis inhibition in MB and DAL neurons blocks LTM formation. (A) Effects
832
of memory circuit Gal4-targeted RicinCS on 1-day memory after ten spaced training cycles (10xS), compared with
833
wild-type (+/+) controls. Cold-sensitive RicinCS blocks protein synthesis at the permissive temperature (30 °C)
834 Page 43 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint between training and testing. Inhibition of protein synthesis in eight of 22 Gal4-expressing patterns that include both
835
MB and DAL neurons and one that includes DAL but not MB neurons (Umnitza) blocks 1-day memory. 836
Interestingly, inhibition of protein synthesis in seven other Gal4-expressing patterns that include both MB and DAL
837
neurons (related to Figure 1) and one that includes MB but not DAL neurons (Norka) have no net effect on LTM. (B)
838
Nine Gal4 expression patterns in which protein synthesis inhibition blocks 1-day memory (A). (C) We observe no
839
LTM effects after RicinCS expression in patterns shown above (B) at the restrictive temperature (18 °C) or in flies
840
expressing the nine Gal4 drivers or ricinCS transgene alone (left). RicinCS expression in all nine patterns has no effect
841
on 1-day memory after 10xM training (right). 842
In all Extended Data figures, temperature control schedules for behavior experiments are indicated (top). Memory
843
performance indices are calculated as the normalized percent avoidance of shock-paired odor. Bars represent mean ±
844
SE, n = 8/bar unless otherwise noted. *, P < 0.05; **, P < 0.01, ***, P < 0.001. All images of dissected brains show
845
Gal4-driven UAS-mCD8::GFP (green), counterstained with DLG-antibody immunostaining (magenta), as viewed
846
under a confocal microscope. Figure 10. Shn in early neurons regulates CREBB dependent LTM formation. (A)
802 Scale bar = 50 m unless otherwise noted. All fly genotypes are listed in table
847
supplement 1. 848 between training and testing. Inhibition of protein synthesis in eight of 22 Gal4-expressing patterns that include both
835
MB and DAL neurons and one that includes DAL but not MB neurons (Umnitza) blocks 1-day memory. 836
Interestingly, inhibition of protein synthesis in seven other Gal4-expressing patterns that include both MB and DAL
837
neurons (related to Figure 1) and one that includes MB but not DAL neurons (Norka) have no net effect on LTM. (B)
838
Nine Gal4 expression patterns in which protein synthesis inhibition blocks 1-day memory (A). (C) We observe no
839
LTM effects after RicinCS expression in patterns shown above (B) at the restrictive temperature (18 °C) or in flies
840
expressing the nine Gal4 drivers or ricinCS transgene alone (left). RicinCS expression in all nine patterns has no effect
841
on 1-day memory after 10xM training (right). 842 In all Extended Data figures, temperature control schedules for behavior experiments are indicated (top). Memory
843
performance indices are calculated as the normalized percent avoidance of shock-paired odor. Bars represent mean ±
844
SE, n = 8/bar unless otherwise noted. *, P < 0.05; **, P < 0.01, ***, P < 0.001. All images of dissected brains show
845
Gal4-driven UAS-mCD8::GFP (green), counterstained with DLG-antibody immunostaining (magenta), as viewed
846
under a confocal microscope. Scale bar = 50 m unless otherwise noted. All fly genotypes are listed in table
847
supplement 1. 848 In all Extended Data figures, temperature control schedules for behavior experiments are indicated (top). Memory
843
performance indices are calculated as the normalized percent avoidance of shock-paired odor. Bars represent mean ±
844
SE, n = 8/bar unless otherwise noted. *, P < 0.05; **, P < 0.01, ***, P < 0.001. All images of dissected brains show
845
Gal4-driven UAS-mCD8::GFP (green), counterstained with DLG-antibody immunostaining (magenta), as viewed
846
under a confocal microscope. Scale bar = 50 m unless otherwise noted. All fly genotypes are listed in table
847
supplement 1. 848 Page 44 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. subtypes. Spatial distributions of three neuron subtypes shown in a schematic representation (right) and in cross
852
section at the vertical lobes (inset). MB split-Gal4 MB477B shows specific expression in early (surface) neurons.
853
Scale bar (inset) = 10 m. (B) Effects of MB Gal4-targeted RicinCS on memory compared with wild-type controls
854
(related to Figure 2). Blocking protein synthesis in MB neurons after 10xS training has no effect on LTM (compare
855
with the memory enhancing effects of protein synthesis inhibition in early neurons after 3xS and 1x training,
856
Figure 2). (C) RicinCS expression in MBs neurons has no effect on memory at the restrictive temperature (18 °C) after
857
3xS training (left), or at the persmissive temperature (30 °C) after 3xM training (right).
858 Figure 10. Shn in early neurons regulates CREBB dependent LTM formation. (A)
802 ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Page 45 of 55
Figure supplement 2. Protein synthesis inhibition in MB neurons does not affect LTM formation aft
training. (A) Gal4 expression patterns that delineate five genetically and developmentally distinct MB ne Figure supplement 2. Protein synthesis inhibition in MB neurons does not affect LTM formation after 10xS
850
training. (A) Gal4 expression patterns that delineate five genetically and developmentally distinct MB neuron
851 Page 45 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint subtypes. Spatial distributions of three neuron subtypes shown in a schematic representation (right) and in cross
852
section at the vertical lobes (inset). MB split-Gal4 MB477B shows specific expression in early (surface) neurons. 853
Scale bar (inset) = 10 m. (B) Effects of MB Gal4-targeted RicinCS on memory compared with wild-type controls
854
(related to Figure 2). Blocking protein synthesis in MB neurons after 10xS training has no effect on LTM (compare
855
with the memory enhancing effects of protein synthesis inhibition in early neurons after 3xS and 1x training,
856
Figure 2). (C) RicinCS expression in MBs neurons has no effect on memory at the restrictive temperature (18 °C) after
857
3xS training (left), or at the persmissive temperature (30 °C) after 3xM training (right). 858 859 Page 46 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. Figure 10. Shn in early neurons regulates CREBB dependent LTM formation. (A)
802 It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 860
Figure supplement 3. Protein synthesis inhibition in specific extrinsic MB neurons blocks LTM formation. (A)
861
Gal4 expression patterns that delineate intrinsic MB neurons and extrinsic MB-V3 and DAL neurons. (B) Blocking
862
protein synthesis in MB neurons with two copies of active RicinCS has no effect on LTM after 10xS training (left). By
863
contrast, blocking protein synthesis in MB-V3 or DAL neurons inhibits LTM after 10xS training (right), but not after
864
10xM training or after 10xS training when flies are held at the restrictive temperature (18 °C). 865 Figure supplement 3. Protein synthesis inhibition in specific extrinsic MB neurons blocks LTM formation. (A)
861
Gal4 expression patterns that delineate intrinsic MB neurons and extrinsic MB-V3 and DAL neurons. (B) Blocking
862
protein synthesis in MB neurons with two copies of active RicinCS has no effect on LTM after 10xS training (left). By
863
contrast, blocking protein synthesis in MB-V3 or DAL neurons inhibits LTM after 10xS training (right), but not after
864
10xM training or after 10xS training when flies are held at the restrictive temperature (18 °C). 865 gure supplement 3. Protein synthesis inhibition in specific extrinsic MB neurons blocks LTM Figure supplement 3. Protein synthesis inhibition in specific extrinsic MB neurons blocks LTM formation. (A)
861
Gal4 expression patterns that delineate intrinsic MB neurons and extrinsic MB-V3 and DAL neurons. (B) Blocking
862
protein synthesis in MB neurons with two copies of active RicinCS has no effect on LTM after 10xS training (left). By
863
contrast, blocking protein synthesis in MB-V3 or DAL neurons inhibits LTM after 10xS training (right), but not after
864
10xM training or after 10xS training when flies are held at the restrictive temperature (18 °C). 865 866 Page 47 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. Figure 10. Shn in early neurons regulates CREBB dependent LTM formation. (A)
802 It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 867
Figure supplement 4. Blocking transmission from MB neurons and effects on LTM formation. Effects of
868
Gal4-targeted Dynamints (shits) on 1-day memory. (A) Blocking neural transmission from pioneer and late
869
neurons during sequential 8-h time windows after subthreshold 3xS training has no effect on memory. (B) Blocking
870
neural transmission from early neurons during sequential 8-h time windows after 10xS training has no effect on
871
memory. By comparison, blocking transmission from early neurons after 3xS training enhances memory (Figure
872
4). 873 Figure supplement 4. Blocking transmission from MB neurons and effects on LTM formation. Effects of
868 Figure supplement 4. Blocking transmission from MB neurons and effects on LTM formation. Effects of
868
Gal4-targeted Dynamints (shits) on 1-day memory. (A) Blocking neural transmission from pioneer and late
869
neurons during sequential 8-h time windows after subthreshold 3xS training has no effect on memory. (B) Blocking
870
l
i i
f
l
d
i
i l 8 h i
i d
f
10 S
i i
h
ff
871 . Blocking transmission from MB neurons and effects on LTM formation. Effects of Figure supplement 4. Blocking transmission from MB neurons and effects on LTM formation. Effects of
868
Gal4-targeted Dynamints (shits) on 1-day memory. (A) Blocking neural transmission from pioneer and late
869
neurons during sequential 8-h time windows after subthreshold 3xS training has no effect on memory. (B) Blocking
870
neural transmission from early neurons during sequential 8-h time windows after 10xS training has no effect on
871
memory. By comparison, blocking transmission from early neurons after 3xS training enhances memory (Figure
872
4). 873 Gal4-targeted Dynamints (shits) on 1-day memory. (A) Blocking neural transmission from pioneer and late
869
neurons during sequential 8-h time windows after subthreshold 3xS training has no effect on memory. (B) Blocking
870
neural transmission from early neurons during sequential 8-h time windows after 10xS training has no effect on
871
memory. By comparison, blocking transmission from early neurons after 3xS training enhances memory (Figure
872
4). 873 Page 48 of 55 Page 48 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. .
CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021.
;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Figure supplement 5. Increasing membrane excitability in early neurons and effects on memory formation. Early Gal4-targeted expression of K+ and Na+ channel proteins is induced at the restrictive temperature for the
tub-Gal80ts inhibitor (30 °C) from five days prior to training until testing. Experimental groups are compared with
wild-type (+/+) controls. (A) ShawDN and NaChBac overexpression increase neural activity in early neurons but
have no effects on immediate memory after 1x training. (B) Similarly, elevated neural activity has no effect on 1-day Figure supplement 5. Increasing membrane excitability in early neurons and effects on memory formation.
875 Figure supplement 5. Increasing membrane excitability in early neurons and effects on memory formation. 875 igure supplement 5. Increasing membrane excitability in early neurons and effects on mem Early Gal4-targeted expression of K+ and Na+ channel proteins is induced at the restrictive temperature for the
876
tub-Gal80ts inhibitor (30 °C) from five days prior to training until testing. Experimental groups are compared with
877
wild-type (+/+) controls. (A) ShawDN and NaChBac overexpression increase neural activity in early neurons but
878
have no effects on immediate memory after 1x training. (B) Similarly, elevated neural activity has no effect on 1-day
879
memory after 10xM training. By comparison, elevated neural activity impairs 1-day memory after 10xS training
880
(Figure 5). (C) Channel proteins are not induced in flies maintained at the permissive temperature for the tub-Gal80ts
881
inhibitor (18 °C) and show no differences in memory after 10xS training. 882 Page 49 of 55 Page 49 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Figure supplement 6. Down-regulation of CREBB, Rutabaga (AC), PKA and Schnurri (Shn) in early
4
neurons and effects on memory formation. (A) Adult-stage specific RNAi down-regulation of these proteins
5
(encoded by rut, Pka, crebB and shn) in early neurons (two RNAi constructs each) have no effects on 1-day
6
memory after 10xM training. Gal4-targeted RNAi specific for each gene is induced at the restrictive temperature fo
7
tub-Gal80ts (30 °C) from five days prior to training until testing. Experimental groups are compared with wild-type
8
(+/+) controls. Figure supplement 5. Increasing membrane excitability in early neurons and effects on memory formation.
875 (B) Proteins are not down-regulated in flies maintained at the permissive temperature for the tub-
9
Gal80ts inhibitor (18 °C) and there are no differences in memory in comparison with control flies after 10xS trainin
0
(C) Down regulation of all proteins in early neurons has no effect on immediate memory after 1x training. By
comparison, overexpressing AC, PKA, CREBB and Shn in early neurons enhances 1-day memory after only 1x
2
training (Figure 6-8 and 10). Figure supplement 6. Down-regulation of CREBB, Rutabaga (AC), PKA and Schnurri (Shn) in early
neurons and effects on memory formation. (A) Adult-stage specific RNAi down-regulation of these proteins Figure supplement 6. Down-regulation of CREBB, Rutabaga (AC), PKA and Schnurri (Shn) in early
884
neurons and effects on memory formation. (A) Adult-stage specific RNAi down-regulation of these prot
885
(encoded by rut, Pka, crebB and shn) in early neurons (two RNAi constructs each) have no effects on 1-day
886
f
10
i i
G l4
d
Ai
ifi f
h
i i d
d
h
i i
88 Figure supplement 6. Down-regulation of CREBB, Rutabaga (AC), PKA and Schnurri (Shn) in early
884 Figure supplement 6. Down-regulation of CREBB, Rutabaga (AC), PKA and Schnurri (Shn) in early
884
neurons and effects on memory formation. (A) Adult-stage specific RNAi down-regulation of these proteins
885
(encoded by rut, Pka, crebB and shn) in early neurons (two RNAi constructs each) have no effects on 1-day
886
memory after 10xM training. Gal4-targeted RNAi specific for each gene is induced at the restrictive temperature for
887
tub-Gal80ts (30 °C) from five days prior to training until testing. Experimental groups are compared with wild-type
888
(+/+) controls. (B) Proteins are not down-regulated in flies maintained at the permissive temperature for the tub-
889
Gal80ts inhibitor (18 °C) and there are no differences in memory in comparison with control flies after 10xS training. 890
(C) Down regulation of all proteins in early neurons has no effect on immediate memory after 1x training. By
891
comparison, overexpressing AC, PKA, CREBB and Shn in early neurons enhances 1-day memory after only 1x
892
training (Figure 6-8 and 10). 893 neurons and effects on memory formation. Figure supplement 5. Increasing membrane excitability in early neurons and effects on memory formation.
875 (A) Adult-stage specific RNAi down-regulation of these proteins
885
(encoded by rut, Pka, crebB and shn) in early neurons (two RNAi constructs each) have no effects on 1-day
886
memory after 10xM training. Gal4-targeted RNAi specific for each gene is induced at the restrictive temperature for
887
tub-Gal80ts (30 °C) from five days prior to training until testing. Experimental groups are compared with wild-type
888
(+/+) controls. (B) Proteins are not down-regulated in flies maintained at the permissive temperature for the tub-
889
Gal80ts inhibitor (18 °C) and there are no differences in memory in comparison with control flies after 10xS training. 890
(C) Down regulation of all proteins in early neurons has no effect on immediate memory after 1x training. By
891
comparison, overexpressing AC, PKA, CREBB and Shn in early neurons enhances 1-day memory after only 1x
892
training (Figure 6-8 and 10). 893 894 Page 50 of 55 Page 50 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 895
Figure supplement 7. Overexpression of Shn but not CREBA protein in early neurons enhances LTM
896
formation. (A) Overexpressing Shn but not CREBA in early neurons enhances 1-day memory after 3xS training. 897
Gal4-targeted crebA or shn overexpression is induced at the restrictive temperature for tub-Gal80ts (30 °C) from five
898
days prior to training until testing. (B) Enhanced memory lasts at least four days (left). Figure supplement 5. Increasing membrane excitability in early neurons and effects on memory formation.
875 Memory is unaffected in these
899
flies held at the permissive temperature for tub-Gal80ts (18 °C) after 3xS training (center), or at the restrictive
900
temperature (30 °C) after 3xM training (right). (C) Adult-stage specific down-regulation of CREBA protein in early
901
neurons has no effect on 1-day memory after 10xS training. 902 Figure supplement 7. Overexpression of Shn but not CREBA protein in early neurons enhances LTM
896 formation. (A) Overexpressing Shn but not CREBA in early neurons enhances 1-day memory after 3xS training. 897
Gal4-targeted crebA or shn overexpression is induced at the restrictive temperature for tub-Gal80ts (30 °C) from five
898
days prior to training until testing. (B) Enhanced memory lasts at least four days (left). Memory is unaffected in these
899
flies held at the permissive temperature for tub-Gal80ts (18 °C) after 3xS training (center), or at the restrictive
900
temperature (30 °C) after 3xM training (right). (C) Adult-stage specific down-regulation of CREBA protein in early
901
neurons has no effect on 1-day memory after 10xS training. 902 903 Page 51 of 55 Page 51 of 55 . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 904
Figure supplement 8. Transgene over-expression and CREBB protein levels in MB Kenyon cells. Gal4-OK107-
905
driven shn over-expression led to elevated CREBB protein levels (second from top) in comparison with the control
906
(top). Over-expression of rut and PKA had only minor impact on CREBB expression (second from bottom and
907
bottom, respectively). Figure supplement 5. Increasing membrane excitability in early neurons and effects on memory formation.
875 Confocal images of Drosophila MB Kenyon cell bodies (encircled by the white dotted line)
908
show the intensity of CREBB immunostaining in a jet colormap. 909 904 Figure supplement 8. Transgene over-expression and CREBB protein levels in MB Kenyon cells. Gal4-OK107-
905
driven shn over-expression led to elevated CREBB protein levels (second from top) in comparison with the control
906
(top). Over-expression of rut and PKA had only minor impact on CREBB expression (second from bottom and
907
bottom, respectively). Confocal images of Drosophila MB Kenyon cell bodies (encircled by the white dotted line)
908
show the intensity of CREBB immunostaining in a jet colormap. 909 Figure supplement 8. Transgene over-expression and CREBB protein levels in MB Kenyon cells. Gal4-OK107-
905
driven shn over-expression led to elevated CREBB protein levels (second from top) in comparison with the control
906
(top). Over-expression of rut and PKA had only minor impact on CREBB expression (second from bottom and
907
bottom, respectively). Confocal images of Drosophila MB Kenyon cell bodies (encircled by the white dotted line)
908
show the intensity of CREBB immunostaining in a jet colormap. 909 Page 52 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint d
m
’’ap
’’m
pioneer
(posterior)
early
(surface)
late
(core)
DAL
Avgust
-
+
-
-
-
++
+++
+++
-
Barbos-1
-
+
-
-
-
+++
+++
+++
+
Ikar
-
+
+
+
-
+++
+++
+++
+++
Laska
-
+
+
+
-
+++
+++
+++
++
Novichok
-
+
-
-
-
+++
+++
+++
+
Rafael
-
++
-
-
-
+++
+++
+++
-
Toi
-
+
-
-
-
+++
+++
+++
+
Arleekin
-
-
++
++
-
++
+++
+++
+
Beck-1
-
-
-
-
-
-
+
+++
+++
Beck-2
-
-
-
-
-
-
+
+++
+++
Krasavietz
-
+
+
+
-
+++
+++
+++
+
Pastrel
-
+
-
-
-
+++
+++
+++
-
Rogdi
N.D. Figure supplement 5. Increasing membrane excitability in early neurons and effects on memory formation.
875 +++
+++
+++
-
+
-
+++
+++
Tungus
++
+
-
-
-
+++
-
+++
+++
Umnitza
+
+
-
-
-
+
-
+++
+
Valiet-2
-
-
-
-
-
-
++
+++
-
Table supplement 1. CSHL memory Gal4 patterns. Table shows Gal4 targeted GFP intensity was graded as strong
915
(+++), intermediate (++), weak (+), absence (-) or non-distinguishable (N.D.). DAL, dorsal anterior lateral neuron;
916
EB, ellipsoid body. 917 t 1. CSHL memory Gal4 patterns. Table shows Gal4 targeted GFP intensity was graded as strong Table supplement 1. CSHL memory Gal4 patterns. Table shows Gal4 targeted GFP intensity was graded as strong
915
(+++), intermediate (++), weak (+), absence (-) or non-distinguishable (N.D.). DAL, dorsal anterior lateral neuron;
916
EB, ellipsoid body. 917 Table supplement 1. CSHL memory Gal4 patterns. Table shows Gal4 targeted GFP intensity was graded as strong
915
(+++), intermediate (++), weak (+), absence (-) or non-distinguishable (N.D.). DAL, dorsal anterior lateral neuron;
916
EB, ellipsoid body. 917 918 Page 53 of 55 Page 53 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint 919 Page 54 of 55 . CC-BY 4.0 International license
available under a
(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made
The copyright holder for this preprint
this version posted June 10, 2021. ;
https://doi.org/10.1101/2021.06.10.447902
doi:
bioRxiv preprint Table supplement 2. Fly genotypes used in this study. Figure supplement 5. Increasing membrane excitability in early neurons and effects on memory formation.
875 Table shows all fly genotypes, descriptions and sources us
in this study, grouped by distinct types of transgenes. References are listed in Methods. Table supplement 2. Fly genotypes used in this study. Table shows all fly genotypes, descriptions and sources used
921
in this study, grouped by distinct types of transgenes. References are listed in Methods. 922 Table supplement 2. Fly genotypes used in this study. Table shows all fly genotypes, descriptions and sources used
921
in this study, grouped by distinct types of transgenes. References are listed in Methods. 922 in this study, grouped by distinct types of transgenes. References are listed in Methods. 923 Page 55 of 55 Page 55 of 55
|
https://openalex.org/W2331404222
|
https://www.nature.com/articles/cddis201676.pdf
|
English
| null |
Cathepsin B pulls the emergency brake on cellular necrosis
|
Cell death and disease
| 2,016
|
cc-by
| 1,831
|
Cathepsin B pulls the emergency brake on cellular
necrosis FS Wouters*,1,2 and G Bunt1,3 FS Wouters*,1,2 and G Bunt1,3 FS Wouters*,1,2 and G Bunt1,3 Cell Death and Disease (2016) 7, e2170; doi:10.1038/cddis.2016.76; published online 31 March 2016 Cell death can take many shapes. Programmed ‘apoptotic’
and uncontrolled ‘necrotic’ cell death mark the extremes of the
spectrum of possibilities.1 Whereas the induction of apoptosis
follows distinct signaling steps and cleanly removes a cell from
its tissue, necrosis represents the cell’s wholesale disintegra-
tion and is in general harmful to the organism. Massive cellular
insults lead to necrosis. It appears, however, that cells have
some choice in the mode of their demise depending on the
severity of the insult. Hypoxic cores in solid tumors2 or
perfusion-impaired brain tissue in stroke,3 for instance are
typically necrotic, but possess an apoptotic shell. occurs rapidly and is completed within minutes. Among the
observed bcl-2 members, anti-apoptotic bcl-xl is degraded
before pro-apoptotic bid and bax. Cells thus appear to
accumulate pro-apoptotic reactants in the early stages of
lysosomal rupture. The pro-apoptotic bid showed a conspicuous behavior. Before the overall sigmoidal degradation, a pronounced and
very rapid cleavage was observed in the first minutes of
lysosomal lysis. Fast proteolytic processing of bid is used
in apoptosis and might serve the same role in lysosomal lysis. In apoptosis, the caspase 8 protease activates bid by its
truncation, directly connecting the extrinsic and intrinsic
apoptotic signaling pathways. This mechanism is called the
‘bid shunt’ and processing of bid by cathepsin B has been
suggested to operate similarly.11 Our work places the bid
shunt very early in lysosomal cell death signaling and this is
the first time this event is caught ‘on film’. The
same
can
be
seen
for
lysosomal
damage.4–6
Lysosomes can rupture during ischemic or traumatic cell
injury,7 giving rise to both apoptotic and necrotic outcomes. Lysosomes are cellular organelles with a proteolytic
function
and
possess
an
impressive
set
of
protease
enzymes8 dedicated to the degradation of both intracellular
material, such as damaged or old organelles, and extracellular
components such as matrix proteins. The release of the
lysosomal proteases into the cytosol sentences the cell to
death by autodigestion. In order to gain a better understanding of the regulation of
bcl-2 protein levels by cathepsins and the involvement in
programmatic steps in necrosis and apoptosis, we investi-
gated the interplay of both systems in more detail. 1Laboratory for Molecular and Cellular Systems, Institute of Neuropathology, University Medical Center Göttingen, Göttingen, Germany; 2Centre for Nanoscale Molecular
Physiology of the Brain, Göttingen, Germany and 3Clinical Optical Microscopy, Institute of Neuropathology, University Medical Center Göttingen, Göttingen, Germany
*Corresponding author: FS Wouters, Laboratory for Molecular and Cellular Systems, Institute of Neuropathology, University Medical Center Göttingen, Waldweg 33,
D-37073 Göttingen, Germany. Tel: +49 551 3912368; Fax: +49 551 396031; E-mail: fred.wouters@gwdg.de Citation: Cell Death and Disease (2016) 7, e2170; doi:10.1038/cddis.2016.76
& 2016 Macmillan Publishers Limited
All rights reserved 2041-4889/16 Citation: Cell Death and Disease (2016) 7, e2170; doi:10.1038/cddis.2016.76
& 2016 Macmillan Publishers Limited
All rights reserved 2041-4889/16 OPEN www.nature.com/cddis News and Commentary Conflict of Interest
The authors declare no conflict of interest. general inhibition of the thiol-cathepsin classes resulted in the
complete inhibition of the proteolysis of these same sensors,
we concluded that their accelerated degradation upon
cathepsin B inhibition is most likely caused by another,
unidentified, thiol-cathepsin. This proteolytic cascade pre-
vents the degradation of apoptosis-supporting proteins,
shifting the balance toward apotosis. The identity of this
cathepsin requires further research. The authors declare no conflict of interest. Acknowledgements. This work was supported by a grant from the Federal
Ministry of Science and Education, BMBF (13N9243) and the Open Access
Publication Fund of the Göttingen University. 1. Yuan J et al. Genes Dev 2010; 24: 2592–2602. 2. Riva C et al. Anticancer Res 1998; 18: 4729–4736. 3. Memezawa H et al. Stroke 1992; 23: 552–559. 4. Kilinc M et al. Neurobiol Dis 2010; 40: 293–302. 5. Qin AP et al. Neurosci Bull 2008; 24: 117–123. 6. Yamashima T et al. Prog Neurobiol 2009; 89: 343–358. 7. Boya P et al. Oncogene 2008; 27: 6434–6451. 8. Müller S et al. Biochim Biophys Acta 2012; 1824: 34–43. 9. de Castro MAG et al. Cell Death Discov 2016; 2: e16012. 10. Cirman T et al. J Biol Chem 2004; 279: 3578–3587. 11. Guicciardi ME et al. J Clin Invest 2000; 106: 1127–11378. Cells thus appear to be able to pull an ‘emergency brake’
upon lethal injury with damaging of lysosomes. Starting with
cathepsin B, an early programmatic protease cascade is
initiated that appears to avoid necrosis in favor of apoptosis. In
this apoptotic exit effort, cathepsin B preferentially degrades
anti-apoptotic bcl-xl, activates bid by its rapid and selective
truncation, and degrades a thiol-cathepsin to avoid the
removal of apoptotic bid and bax (Figure 1, right). Up to a
certain level of lysosomal damage, this three-pronged rescue
program steers dying cells away from necrosis. When massive
damage overwhelms the cell with thiol-cathepsins that
eat away at essential cellular proteins, necrosis becomes
unavoidable. Our research contributes to the understanding of
cell death on a molecular mechanistic level by uncovering
programmatic steps at the interface between necrosis and
apoptosis. Knowledge on cell death decision signaling is
therapeutically relevant, for example, in the formulation of
more effective chemotherapy and in combatting toxic drug
side effects. Cell Death and Disease is an open-access journal
published by Nature Publishing Group. This work is
licensed under a Creative Commons Attribution 4.0 International
License. Cathepsin B pulls the emergency brake on cellular
necrosis g
y
y
Given the relatively low amount of lysosomes that were
disrupted in the early stages after lysosomal disruption, we
selected the neutral-active cathepsin B as likely candidate
for most of the early proteolytic actions. Hence, we repeated
the experiments in the presence of a selective cathepsin B
inhibitor and an inhibitor for the thiol-cathepsin class, to which
cathepsin B belongs. Inhibition of the thiol-cathepsins con-
firmed their predominant role in the degradation of bcl-2 family
proteins, as the proteolysis of all three bcl-2 constructs was
inhibited. Selective cathepsin B inhibition helped uncover its
role in steering cell death. We found that early bid truncation
was abolished, confirming cathepsin B as the responsible
protease. Furthermore, when we limited lysosomal protease
action to the time window of the early bid truncation, we
detected late-apoptotic caspase 3/7 activation, showing that
cells had chosen an apoptotic exit. In a recent Cell Death Discovery article,9 we looked at the
fate of the major regulators of apoptosis, the bcl-2 protein
family, in order to understand how cell fate upon lysosomal
rupture is directed toward necrosis or apoptosis. This protein
family contains both pro- and anti-apoptotic members and has
been shown to be part of the proteolytic spectrum of cathepsin
lysosomal
proteases.10
We
reasoned
that
a
regulated
degradation of these proteins, by affecting the equilibrium of
pro- and anti-apoptotic signaling, might offer a means for
switching between cell death forms. To this end, we imaged the proteolytic degradation of the
major bcl-2 family representatives bcl-xl, bid and bax upon
induced lysosomal disruption in real time using förster
resonance energy transfer (FRET) microscopy of single cells. We created proteolytic FRET sensors by sandwiching full-
length bcl-2 proteins between cyan and yellow fluorescent
proteins. A strong FRET signal is detected in the intact sensor,
which is lost upon cleavage (Figure 1, left). Unexpectedly and paradoxically, bax and also bid exhibited
a new accelerated proteolytic behavior upon selective inhibi-
tion of cathepsin B. The delay phase before the onset of
cleavage was completely abolished, in particular for bax. As a The bcl-2 sensors show a common sigmoidal ‘delay-snap’
cleavage behavior: after a delay of tens of minutes, cleavage 2 News and Commentary 2 2 2 Figure 1
Left: upon lysosomal damage, cathepsin proteases are released in the cytosol, finally resulting in cell death. The apoptosis-regulating bcl-2 family proteins bcl-xl, bid
and bax are cleaved by cathepsins. Cathepsin B pulls the emergency brake on cellular
necrosis Their degradation was followed by FRET microscopy on single cells in real time. Cleavage of FRET sensors consisting of full-length bcl-2
proteins sandwiched between cyan and yellow fluorescent protein was imaged. The intact sensors show FRET, the cleaved sensors do not. Right: cathepsin B initiates an
apoptotic exit program by (i) proteolytic removal and consequent inactivation of anti-apoptotic bcl-xl, (ii) the rapid and controlled proteolytic activation (t-bid) of
pro-apoptotic bid, and (iii) the proteolytic inactivation of an unknown thiol-cathepsin that would otherwise have removed pro-apoptotic bax and bid. Even in the background of a
strong necrotic stimulus as lysosomal lysis, cathepsin B thus appears to launch an early apoptotic exit program. FRET, förster resonance energy transfer; t-bid, truncated bid. Figure 1
Left: upon lysosomal damage, cathepsin proteases are released in the cytosol, finally resulting in cell death. The apoptosis-regulating bcl-2 family proteins bcl-xl, bid
and bax are cleaved by cathepsins. Their degradation was followed by FRET microscopy on single cells in real time. Cleavage of FRET sensors consisting of full-length bcl-2
proteins sandwiched between cyan and yellow fluorescent protein was imaged. The intact sensors show FRET, the cleaved sensors do not. Right: cathepsin B initiates an
apoptotic exit program by (i) proteolytic removal and consequent inactivation of anti-apoptotic bcl-xl, (ii) the rapid and controlled proteolytic activation (t-bid) of
pro-apoptotic bid, and (iii) the proteolytic inactivation of an unknown thiol-cathepsin that would otherwise have removed pro-apoptotic bax and bid. Even in the background of a
strong necrotic stimulus as lysosomal lysis, cathepsin B thus appears to launch an early apoptotic exit program. FRET, förster resonance energy transfer; t-bid, truncated bid. Figure 1
Left: upon lysosomal damage, cathepsin proteases are released in the cytosol, finally resulting in cell death. The apoptosis-regulating bcl-2 family proteins bcl-xl, bid
and bax are cleaved by cathepsins. Their degradation was followed by FRET microscopy on single cells in real time. Cleavage of FRET sensors consisting of full-length bcl-2
proteins sandwiched between cyan and yellow fluorescent protein was imaged. The intact sensors show FRET, the cleaved sensors do not. Right: cathepsin B initiates an
apoptotic exit program by (i) proteolytic removal and consequent inactivation of anti-apoptotic bcl-xl, (ii) the rapid and controlled proteolytic activation (t-bid) of
pro-apoptotic bid, and (iii) the proteolytic inactivation of an unknown thiol-cathepsin that would otherwise have removed pro-apoptotic bax and bid. Cathepsin B pulls the emergency brake on cellular
necrosis Even in the background of a
strong necrotic stimulus as lysosomal lysis, cathepsin B thus appears to launch an early apoptotic exit program. FRET, förster resonance energy transfer; t-bid, truncated bid. Cell Death and Disease Conflict of Interest
The authors declare no conflict of interest. The images or other third party material in this article are
included in the article’s Creative Commons license, unless indicated
otherwise in the credit line; if the material is not included under the
Creative Commons license, users will need to obtain permission from
the license holder to reproduce the material. To view a copy of this
license, visit http://creativecommons.org/licenses/by/4.0/ Cell Death and Disease
|
https://openalex.org/W4382983239
|
https://peerj.com/articles/15541v0.2/submission
|
English
| null |
Peer Review #1 of "Multi-benthic size approach to unveil different environmental conditions in a Mediterranean harbor area (Ancona, Adriatic Sea, Italy) (v0.2)"
| null | 2,023
|
cc-by
| 29,305
|
Multi-benthic size approach to unveil different environmental
conditions in a Mediterranean harbor area (Ancona, Adriatic
Sea, Italy) Elisa Baldrighi
Corresp., 1, 2 , Sarah Pizzini
3, 4 , Elisa Punzo
1 , Angela Santelli
3 , Pierluigi Strafella
3 , Tommaso Scirocco
5 ,
Elena Manini
3 , Daniele Fattorini
6, 7 , Claudio Vasapollo
Corresp. 3 Elisa Baldrighi
Corresp., 1, 2 , Sarah Pizzini
3, 4 , Elisa Punzo
1 , Angela Santelli
3 , Pierluigi Stra
Elena Manini
3 , Daniele Fattorini
6, 7 , Claudio Vasapollo
Corresp. 3 1 1Institute for Biological Resources and Marine Biotechnologies—IRBIM, National Research Council—CNR, Italy, ANCONA, MARCHE, Italy
2 2Department of Biology, University of Nevada-Reno, Reno, Nevada, United States of America, Reno, Nevada, USA
3 1Institute for Biological Resources and Marine Biotechnologies—IRBIM, National Research Council—CNR, Italy, Ancona, Italy
4 3Fano Marine Center, The Inter-Institute Center for Research on Marine Biodiversity, Resources and Biotechnologies, Fano, Italy, Fano, Italy
5 1Institute for Biological Resources and Marine Biotechnologies—IRBIM, National Research Council—CNR, Italy, Lesina, Italy
6 4Dipartimento di Scienze della Vita e dell’Ambiente (Disva), Università Politecnica delle Marche (Univpm), Ancona, Italy, Ancona, Italy
7 5Consorzio Nazionale Interuniversitario per le Scienze del Mare (Conisma), Unità di Ricerca di Ancona (Italy)., ANCONA, MARCHE, Italy
Corresponding Authors: Elisa Baldrighi, Claudio Vasapollo
Email address: elisa.baldrighi@irbim.cnr.it, claudio.vasapollo@cnr.it Harbors are hubs of human activity and are subject to the continuous discharge and
release of industrial, agricultural, and municipal waste and contaminants. Benthic
organisms are largely known to reflect environmental conditions they live in. Despite meio-
and macrofauna interacting within the benthic system, they are ecologically distinct
components of the benthos and as such may not necessarily respond to environmental
conditions and/or disturbances in the same way. However, in a few field studies the spatial
patterns of meio- and macrofauna have been simultaneously compared. In the present
study, we assess the response and patterns in the abundance, diversity, and distribution of
the two benthic size classes to the different environmental conditions they live in (i.e.,
sediment concentrations of selected trace metals and Polycyclic aromatic hydrocarbons
(PAHs); organic matter contents and grain size) characterizing the Ancona harbor (Adriatic
Sea). Meio- and macrofauna provided partially similar types of information depending on
the indices used (univariate measures or community structure/ species composition) and
the different ‘response-to-stress’. The community structure (i.e., taxa composition) of both
benthic size components clearly showed differences among sampling stations located from
inside to outside the harbor, reflecting the marked environmental heterogeneity and
disturbance typically characterizing these systems. Manuscript to be reviewed Multi-benthic size approach to unveil different environmental
conditions in a Mediterranean harbor area (Ancona, Adriatic
Sea, Italy) Notwithstanding, the univariate
measures (i.e., meio- and macrofauna total abundance, diversity indices and equitability)
didn’t show similar spatial patterns. Meiofauna were likely to be more sensitive to the
effects of environmental features and contaminants than macrofauna. Overall, trace
PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) 1
Multi-benthic size approach to unveil different environmental conditions in a
2
Mediterranean harbor area (Ancona, Adriatic Sea, Italy) 1
Multi-benthic size approach to unveil d
2
Mediterranean harbor area (Ancona, A
3
4
Elisa Baldrighi1,2*, Sarah Pizzini1,3, Elisa Punzo
5
Tommaso Scirocco1, Elena Manini1, Daniele Fa
6
7
1Institute for Marine Biological Resources and Biot
8
CNR, Italy
9
2Department of Biology, University of Nevada-Reno,
10
3Fano Marine Center, The Inter-Institute Center for
11
Biotechnologies, Fano, Italy
12
4Dipartimento di Scienze della Vita e dell’Ambie
13
(Univpm), Ancona, Italy
14
5Consorzio Nazionale Interuniversitario per le Scienz
15
(Italy). 16
17
Corresponding Author:
18
Elisa Baldrighi1,2
19
Largo Fiera della Pesca 2, Ancona, 60125, Italy
20
Email address: elisa.baldrighi@irbim.cnr.it
21
22
Corresponding Author:
23
Claudio Vasapollo1
24
Largo Fiera della Pesca 2, Ancona, 60125, Italy
25
Email address: claudio.vasapollo@cnr.it
26
27
Abstract
28
Harbors are hubs of human activity and are subje
29
industrial, agricultural, and municipal waste and c
30
known to reflect environmental conditions they li
31
interacting within the benthic system, they are eco
32
and as such may not necessarily respond to enviro
33
same way. However, in a few field studies the spa
34
been simultaneously compared. In the present stu
35
abundance, diversity, and distribution of the two b
36
environmental conditions they live in (i.e., sedim
37
Polycyclic aromatic hydrocarbons (PAHs); organ
38
the Ancona harbor (Adriatic Sea). 39
Meio- and macrofauna provided partially similar
40
used (univariate measures or community structure
PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) 3
4
Elisa Baldrighi1,2*, Sarah Pizzini1,3, Elisa Punzo1, Angela Santelli1, Pierluigi Strafella1,
5
Tommaso Scirocco1, Elena Manini1, Daniele Fattorini4,5, Claudio Vasapollo1*
6
7
1Institute for Marine Biological Resources and Biotechnology—IRBIM, National Research Coun
8
CNR, Italy
9
2Department of Biology, University of Nevada-Reno, Reno, Nevada, United States of America
10
3Fano Marine Center, The Inter-Institute Center for Research on Marine Biodiversity, Resources
11
Biotechnologies, Fano, Italy
12
4Dipartimento di Scienze della Vita e dell’Ambiente (Disva), Università Politecnica delle Ma
13
(Univpm), Ancona, Italy
14
5Consorzio Nazionale Interuniversitario per le Scienze del Mare (Conisma), Unità di Ricerca di An
15
(Italy). Manuscript to be reviewed Manuscript to be reviewed metals and PAHs affected the community composition of the two benthic components, but
only the meiofauna abundance and diversity were related to the environmental variables
considered (i.e., quantity and quality of organic matter). Our results pinpoint the
importance of studying both meio- and macrofauna communities, which could provide
greater insight into the processes affecting the investigated area and reveal different
aspects of the benthic ecosystems in response to harbor conditions. metals and PAHs affected the community composition of the two benthic components, but
only the meiofauna abundance and diversity were related to the environmental variables
considered (i.e., quantity and quality of organic matter). Our results pinpoint the
importance of studying both meio- and macrofauna communities, which could provide
greater insight into the processes affecting the investigated area and reveal different
aspects of the benthic ecosystems in response to harbor conditions. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 1
Multi-benthic size approach to unveil different environmental conditions in a
2
Mediterranean harbor area (Ancona, Adriatic Sea, Italy)
3
4
Elisa Baldrighi1,2*, Sarah Pizzini1,3, Elisa Punzo1, Angela Santelli1, Pierluigi Strafella1,
5
Tommaso Scirocco1, Elena Manini1, Daniele Fattorini4,5, Claudio Vasapollo1*
6
7
1Institute for Marine Biological Resources and Biotechnology—IRBIM, National Research Council—
8
CNR, Italy
9
2Department of Biology, University of Nevada-Reno, Reno, Nevada, United States of America
10
3Fano Marine Center, The Inter-Institute Center for Research on Marine Biodiversity, Resources and
11
Biotechnologies, Fano, Italy
12
4Dipartimento di Scienze della Vita e dell’Ambiente (Disva), Università Politecnica delle Marche
13
(Univpm), Ancona, Italy
14
5Consorzio Nazionale Interuniversitario per le Scienze del Mare (Conisma), Unità di Ricerca di Ancona
15
(Italy). 16
17
Corresponding Author:
18
Elisa Baldrighi1,2
19
Largo Fiera della Pesca 2, Ancona, 60125, Italy
20
Email address: elisa.baldrighi@irbim.cnr.it
21
22
Corresponding Author:
23
Claudio Vasapollo1
24
Largo Fiera della Pesca 2, Ancona, 60125, Italy
25
Email address: claudio.vasapollo@cnr.it
26 1
Multi-benthic size approach to unveil different environmental conditions in a
2
Mediterranean harbor area (Ancona, Adriatic Sea, Italy) 16
17
Corresponding Author:
18
Elisa Baldrighi1,2
19
Largo Fiera della Pesca 2, Ancona, 60125, Italy
20
Email address: elisa.baldrighi@irbim.cnr.it
21
22
Corresponding Author:
23
Claudio Vasapollo1
24
Largo Fiera della Pesca 2, Ancona, 60125, Italy
25
Email address: claudio.vasapollo@cnr.it
26
27
Abstract
28
Harbors are hubs of human activity and are subject to the continuous discharge and release of
29
industrial, agricultural, and municipal waste and contaminants. Benthic organisms are largely
30
known to reflect environmental conditions they live in. Despite meio- and macrofauna
31
interacting within the benthic system, they are ecologically distinct components of the bentho
32
and as such may not necessarily respond to environmental conditions and/or disturbances in t
33
same way. However, in a few field studies the spatial patterns of meio- and macrofauna have
34
been simultaneously compared. In the present study, we assess the response and patterns in th
35
abundance, diversity, and distribution of the two benthic size classes to the different
36
environmental conditions they live in (i.e., sediment concentrations of selected trace metals a
37
Polycyclic aromatic hydrocarbons (PAHs); organic matter contents and grain size) characteri
38
the Ancona harbor (Adriatic Sea). 39
Meio- and macrofauna provided partially similar types of information depending on the indic
40
used (univariate measures or community structure/ species composition) and the different 4Dipartimento di Scienze della Vita e dell’Ambiente (Disva), Università Politecnica delle Marche
(Univpm), Ancona, Italy Manuscript to be reviewed 41
‘response-to-stress’. The community structure (i.e., taxa composition) of both benthic size
42
components clearly showed differences among sampling stations located from inside to outside
43
the harbor, reflecting the marked environmental heterogeneity and disturbance typically
44
characterizing these systems. Notwithstanding, the univariate measures (i.e., meio- and
45
macrofauna total abundance, diversity indices and equitability) didn’t show similar spatial
46
patterns. Meiofauna were likely to be more sensitive to the effects of environmental features and
47
contaminants than macrofauna. Overall, trace metals and PAHs affected the community
48
composition of the two benthic components, but only the meiofauna abundance and diversity
49
were related to the environmental variables considered (i.e., quantity and quality of organic
50
matter). Our results pinpoint the importance of studying both meio- and macrofauna
51
communities, which could provide greater insight into the processes affecting the investigated
52
area and reveal different aspects of the benthic ecosystems in response to harbor conditions. 53
54
55
Introduction
56
Coastal waters are widely recognized as marine areas of high ecological and economic value, but
57
also as highly threatened zone, exposed to multiple human activities (e.g., harbors) and their
58
negative impacts (Travizi et al., 2019). 59
Harbors are essential to the economic growth of coastal regions, where maritime traffic,
60
shipping, international trade, and fishing are continuously increasing (Simonini et al., 2005;
61
Franzo et al., 2022). Harbors are usually characterized by high sediment pollution levels due to
62
heavy metals and hydrocarbons caused by intense maritime traffic and huge organic matter loads
63
(Covazzi Harriague et al., 2007; Baldrighi et al., 2019). The high concentrations of contaminants
64
and the relevant inputs of organic matter represent a persistent and ongoing threat, especially for
65
the biota living in the sediment (Travizi et al., 2019; Franzo et al., 2022). In addition, the
66
exposure of the innermost part of harbors to both wind and waves is limited and may create
67
conditions of reduced water circulation, favoring sedimentation processes, anoxia, and trapping
68
pollutants (Guerra-García & García-Gómez, 2004a; Spagnolo et al., 2011). The EU Water
69
Framework Directive (WFD; Directive 2000/60/EC) considers harbors as ‘heavily modified
70
water bodies’, which cannot meet the common criteria of good ecological quality status. Manuscript to be reviewed 71
Therefore, their effective management is crucial for the sustainable use of these maritime spaces
72
and for the protection of the adjacent coastal habitats (Chatzinikolaou et al., 2018; Franzo et al.,
73
2022). 74
Benthic organisms are largely known to reflect environmental conditions they live in. Among
75
benthic components, species of meio- and macrofauna are widely recognized as good ecological
76
indicators (Schratzberger et al., 2003; Patricio et al., 2012). Benthic meiofaunal and macrofaunal
77
communities are regularly utilized in impact assessment, but very few studies are carried out
78
taking into account both communities (Whomersley et al., 2009; Frontalini et al., 2011; Covazzi-
79
Harriague et al., 2012; Xu et al., 2014). The meio- and macrofauna are ecologically distinct 41
‘response-to-stress’. The community structure (i.e., taxa composition) of both benthic size
42
components clearly showed differences among sampling stations located from inside to outside
43
the harbor, reflecting the marked environmental heterogeneity and disturbance typically
44
characterizing these systems. Notwithstanding, the univariate measures (i.e., meio- and
45
macrofauna total abundance, diversity indices and equitability) didn’t show similar spatial
46
patterns. Meiofauna were likely to be more sensitive to the effects of environmental features and
47
contaminants than macrofauna. Overall, trace metals and PAHs affected the community
48
composition of the two benthic components, but only the meiofauna abundance and diversity
49
were related to the environmental variables considered (i.e., quantity and quality of organic
50
matter). Our results pinpoint the importance of studying both meio- and macrofauna
51
communities, which could provide greater insight into the processes affecting the investigated
52
area and reveal different aspects of the benthic ecosystems in response to harbor conditions. 27
Abstract 39
Meio- and macrofauna provided partially similar types of information depending on the indices
40
used (univariate measures or community structure/ species composition) and the different PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 80
components of the benthos and as such may not necessarily respond to environmental conditions
81
and/or disturbances in the same way (Schratzberger et al., 2003; Patrício et al., 2012). 89
or temporary (animals that start off as meiofauna but grow into macrofauna) (Giere, 2009). In
90
some environments (e.g., open slope systems), temporary meiofauna (e.g., Bivalvia, Oligochaeta,
91
Amphipoda, Nemertea, Priapulida, Holothuroidea, Ascidiacea Cnidaria and Decapoda larvae)
92
have been found to constitute a high percentage of the meiobenthic community (Bianchelli et al.,
93
2010). All these organisms are expected to grow into macrofauna and become part of the
94
macrobenthic population of the slope systems. According to modern phylogenetic approaches,
95
seven phyla belong exclusively to meiofauna, while seventeen phyla are accommodated between
96
meio- and macrofauna (Giere and Schratzberger, 2023). (
g
)
97
Meiobenthic organisms cover different ecological roles according to their trophic group, living
98
mode, locomotion adaptation to move and digging in different kind of sediment grains and they
99
comprise both unicellular (e.g., Foraminifera, Ciliata) a metazoan organism (Giere, 2009). 100
Moreover, because of their high density even small volumes of samples can be analyzed to
101
assess meiofaunal changes over different spatial scales and environmental conditions (e.g.,
102
Schratzberger & Ingels, 2018; Ridall & Ingels, 2021). Consequently, meiofaunal communities
103
have generated considerable interest as potential indicators of anthropogenic disturbances in
104
aquatic ecosystems (e.g., Ridall & Ingels, 2021; Semprucci et al., 2022). 97
Meiobenthic organisms cover different ecological roles according to their trophic group, living
98
mode, locomotion adaptation to move and digging in different kind of sediment grains and they
99
comprise both unicellular (e.g., Foraminifera, Ciliata) a metazoan organism (Giere, 2009). 100
Moreover, because of their high density even small volumes of samples can be analyzed to
101
assess meiofaunal changes over different spatial scales and environmental conditions (e.g.,
102
Schratzberger & Ingels, 2018; Ridall & Ingels, 2021). Consequently, meiofaunal communities
103
have generated considerable interest as potential indicators of anthropogenic disturbances in
104
aquatic ecosystems (e.g., Ridall & Ingels, 2021; Semprucci et al., 2022). 55
Introduction 55
Introduction
56
Coastal waters are widely recognized as marine areas of high ecological and economic value, but
57
also as highly threatened zone, exposed to multiple human activities (e.g., harbors) and their
58
negative impacts (Travizi et al., 2019). 59
Harbors are essential to the economic growth of coastal regions, where maritime traffic,
60
shipping, international trade, and fishing are continuously increasing (Simonini et al., 2005;
61
Franzo et al., 2022). Harbors are usually characterized by high sediment pollution levels due to
62
heavy metals and hydrocarbons caused by intense maritime traffic and huge organic matter loads
63
(Covazzi Harriague et al., 2007; Baldrighi et al., 2019). The high concentrations of contaminants
64
and the relevant inputs of organic matter represent a persistent and ongoing threat, especially for
65
the biota living in the sediment (Travizi et al., 2019; Franzo et al., 2022). In addition, the
66
exposure of the innermost part of harbors to both wind and waves is limited and may create
67
conditions of reduced water circulation, favoring sedimentation processes, anoxia, and trapping
68
pollutants (Guerra-García & García-Gómez, 2004a; Spagnolo et al., 2011). The EU Water
69
Framework Directive (WFD; Directive 2000/60/EC) considers harbors as ‘heavily modified
70
water bodies’, which cannot meet the common criteria of good ecological quality status. 71
Therefore, their effective management is crucial for the sustainable use of these maritime spaces
72
and for the protection of the adjacent coastal habitats (Chatzinikolaou et al., 2018; Franzo et al.,
73
2022). 74
Benthic organisms are largely known to reflect environmental conditions they live in. Among
75
benthic components, species of meio- and macrofauna are widely recognized as good ecological
76
indicators (Schratzberger et al., 2003; Patricio et al., 2012). Benthic meiofaunal and macrofaunal
77
communities are regularly utilized in impact assessment, but very few studies are carried out
78
taking into account both communities (Whomersley et al., 2009; Frontalini et al., 2011; Covazzi-
79
Harriague et al., 2012; Xu et al., 2014). The meio- and macrofauna are ecologically distinct PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed q
y
( g ,
g
,
;
p
,
)
105
There are some advantages in using macrofauna, strictly linked to the functional traits of this
106
group: 1) relative longevity, with many species having life spans in excess of 2 yr, allows the
107
macrofauna to integrate responses to environmental pressures over extended time periods
108
(Simboura & Zenetos, 2002); 2) sedentary or sessile mode of life, therefore organisms are not
109
able to escape stress and integrate the environmental quality in a given area; 3) relatively easy to
110
sample quantitatively, even if larger amount of sediment are needed compared to the meiofauna;
111
4) relatively easy taxonomic identification and available taxonomic keys for most groups and 5)
112
well-documented and predictive response to a number of environmental stressors (thus,
113
community changes can be interpreted with a degree of confidence) (Grey et al., 1988;
114
Somerfield et al., 2006; Todorova et al., 2020). Meio- and macrofauna are ecologically distinct
115
components of the benthos, however both benthic size classes are highly influenced by some
116
main sediment features such as the grain size (Moreno et al., 2008b; Pereira et al., 2018), the
117
quantity and quality of organic matter (Vezzulli et al., 2003; Papageorgiou et al., 2010) and
118
presence of pollutants (McCready et al., 2006; Dauvin et al., 2017). Macrobenthic invertebrates
119
have been identified by the EU WFD as key biological components to assess the ecological PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 143
In the present study, we characterized the meio- and macrofaunal communities (i.e., abundance,
144
diversity, and distribution patterns) in relation to the environmental conditions they live in (i.e.,
145
concentration of selected Heavy Metals (HMs) and Polycyclic aromatic hydrocarbons (PAHs);
146
organic matter contents into the sediment and grain size) characterizing the Ancona harbor. We
147
hypothesized that the analysis of two different benthic components may give us more
148
comprehensive information on the environmental features of the harbor, instead of a single
149
benthic size component as is usually done (Xu et al., 2012). In a future perspective, this multi-
150
size approach should be highly recommended in coastal monitoring and management plans. 143
In the present study, we characterized the meio- and macrofaunal communities (i.e., abundance,
144
diversity, and distribution patterns) in relation to the environmental conditions they live in (i.e.,
145
concentration of selected Heavy Metals (HMs) and Polycyclic aromatic hydrocarbons (PAHs);
146
organic matter contents into the sediment and grain size) characterizing the Ancona harbor. We
147
hypothesized that the analysis of two different benthic components may give us more
148
comprehensive information on the environmental features of the harbor, instead of a single
149
benthic size component as is usually done (Xu et al., 2012). In a future perspective, this multi-
150
size approach should be highly recommended in coastal monitoring and management plans. 143
In the present study, we characterized the meio- and macrofaunal communities (i.e., abundance,
144
diversity, and distribution patterns) in relation to the environmental conditions they live in (i.e.,
145
concentration of selected Heavy Metals (HMs) and Polycyclic aromatic hydrocarbons (PAHs);
146
organic matter contents into the sediment and grain size) characterizing the Ancona harbor. We
147
hypothesized that the analysis of two different benthic components may give us more
148
comprehensive information on the environmental features of the harbor, instead of a single
149
benthic size component as is usually done (Xu et al., 2012). In a future perspective, this multi-
150
size approach should be highly recommended in coastal monitoring and management plans. Manuscript to be reviewed In the regional
133
perspective the basin is highly positioned on the list of ‘Priority issues in the Mediterranean
134
environment’ drawn up by the European Environment Agency (EEA), with 20 (15%) out of the
135
131 hotspot pollution sites identified along the Mediterranean coastline in the frame of the
136
Strategic Action Programme (SAP) of the United Nations Environment Programme (UNEP;
137
EEA, 2006), including many harbors such as the Ancona one (Travizi et al., 2019). 138
Since previous works published on Ancona harbor have always treated the different benthic 120
status of aquatic ecosystems, due to their important role in ecosystem functioning and to their
121
involvement in food-web nutrient recycling (Punzo et al., 2017). 122
The complementary use of two sets of faunistic groups with contrasting ecological characteristics
123
could provide greater insight into the processes affecting such an area. 122
The complementary use of two sets of faunistic groups with contrasting ecological characteristics
123
could provide greater insight into the processes affecting such an area. 130
Adriatic Sea ecosystem is negatively affected by many kinds of biological and ecological threats
131
e.g., eutrophication, pollution, fragmentation of benthic habitats, invasion of alien species
132
(Katsanevakis et al., 2011; .M M et al., 2013; Corriero et al., 2016). In the regional
133
perspective the basin is highly positioned on the list of ‘Priority issues in the Mediterranean
134
environment’ drawn up by the European Environment Agency (EEA), with 20 (15%) out of the
135
131 hotspot pollution sites identified along the Mediterranean coastline in the frame of the
136
Strategic Action Programme (SAP) of the United Nations Environment Programme (UNEP;
137
EEA, 2006), including many harbors such as the Ancona one (Travizi et al., 2019). 138
Since previous works published on Ancona harbor have always treated the different benthic
139
components separately (Mirto & Danovaro 2004; Spagnolo et al., 2011; 2019; Baldrighi et al.,
140
2019; Travizi et al., 2019; Franzo et al., 2022), the aim of the present work is to test whether
141
meio- and macrofaunal assemblages could provide a comparable and/or complementary
142
assessment of its ecological conditions. Manuscript to be reviewed These investigations have demonstrated the fundamental advantage of a multi-
128
species approach, with the inclusion of many taxonomic and functional groups that have a broad
129
range of sensitivities to any given environmental regime (e.g., Frontalini et al., 2011). 130
Adriatic Sea ecosystem is negatively affected by many kinds of biological and ecological threats
131
e.g., eutrophication, pollution, fragmentation of benthic habitats, invasion of alien species
132
(Katsanevakis et al., 2011; .M M et al., 2013; Corriero et al., 2016). In the regional
133
perspective the basin is highly positioned on the list of ‘Priority issues in the Mediterranean
134
environment’ drawn up by the European Environment Agency (EEA), with 20 (15%) out of the
135
131 hotspot pollution sites identified along the Mediterranean coastline in the frame of the
136
Strategic Action Programme (SAP) of the United Nations Environment Programme (UNEP;
137
EEA, 2006), including many harbors such as the Ancona one (Travizi et al., 2019). 138
Since previous works published on Ancona harbor have always treated the different benthic
139
components separately (Mirto & Danovaro 2004; Spagnolo et al., 2011; 2019; Baldrighi et al.,
140
2019; Tra i i t l 2019; Fran o t l 2022) the aim of the present
ork is to test
hether 120
status of aquatic ecosystems, due to their important role in ecosystem functioning and to their
121
involvement in food-web nutrient recycling (Punzo et al., 2017). 122
The complementary use of two sets of faunistic groups with contrasting ecological characteristics
123
could provide greater insight into the processes affecting such an area. 124
Up to now, field studies where the spatial patterns of meio- and macrofauna have been
125
simultaneously compared, changes in both assemblages as a response to natural gradients were
126
found to be scattered across habitats (Somerfield et al., 2006; Papageorgiou et al., 2010; Patrício
127
et al., 2012). These investigations have demonstrated the fundamental advantage of a multi-
128
species approach, with the inclusion of many taxonomic and functional groups that have a broad
129
range of sensitivities to any given environmental regime (e.g., Frontalini et al., 2011). 130
Adriatic Sea ecosystem is negatively affected by many kinds of biological and ecological threats
131
e.g., eutrophication, pollution, fragmentation of benthic habitats, invasion of alien species
132
(Katsanevakis et al., 2011; .M M et al., 2013; Corriero et al., 2016). Manuscript to be reviewed 120
status of aquatic ecosystems, due to their important role in ecosystem functioning and to their
121
involvement in food-web nutrient recycling (Punzo et al., 2017). 122
The complementary use of two sets of faunistic groups with contrasting ecological characteristics
123
could provide greater insight into the processes affecting such an area. 124
Up to now, field studies where the spatial patterns of meio- and macrofauna have been
125
simultaneously compared, changes in both assemblages as a response to natural gradients were
126
found to be scattered across habitats (Somerfield et al., 2006; Papageorgiou et al., 2010; Patrício
127
et al., 2012). These investigations have demonstrated the fundamental advantage of a multi-
128
species approach, with the inclusion of many taxonomic and functional groups that have a broad
129
range of sensitivities to any given environmental regime (e.g., Frontalini et al., 2011). 130
Adriatic Sea ecosystem is negatively affected by many kinds of biological and ecological threats
131
e.g., eutrophication, pollution, fragmentation of benthic habitats, invasion of alien species
132
(Katsanevakis et al., 2011; .M M et al., 2013; Corriero et al., 2016). In the regional
133
perspective the basin is highly positioned on the list of ‘Priority issues in the Mediterranean
134
environment’ drawn up by the European Environment Agency (EEA), with 20 (15%) out of the
135
131 hotspot pollution sites identified along the Mediterranean coastline in the frame of the
136
Strategic Action Programme (SAP) of the United Nations Environment Programme (UNEP;
137
EEA, 2006), including many harbors such as the Ancona one (Travizi et al., 2019). 138
Since previous works published on Ancona harbor have always treated the different benthic 120
status of aquatic ecosystems, due to their important role in ecosystem functioning and to their
121
involvement in food-web nutrient recycling (Punzo et al., 2017). 122
The complementary use of two sets of faunistic groups with contrasting ecological characteristics
123
could provide greater insight into the processes affecting such an area. 124
Up to now, field studies where the spatial patterns of meio- and macrofauna have been
125
simultaneously compared, changes in both assemblages as a response to natural gradients were
126
found to be scattered across habitats (Somerfield et al., 2006; Papageorgiou et al., 2010; Patrício
127
et al., 2012). Manuscript to be reviewed 159
waste dumping derived from fishing boats and is also affected by a strong industrial pollution
160
due to the presence of shipyards. Consequently, a huge organic matter load and high heavy metal
161
and hydrocarbon concentrations are present inside the harbor area (Mirto & Danovaro, 2004;
162
Bianchelli et al., 2016). 170
At each station, sediment samples for characterizing the benthic fauna (meio- and macrofauna)
171
and environmental features were collected with a box-corer (40×30 cm wide and 50 cm high) in
172
three independent replicates (i.e., box-corer deployments), processed and preserved differently
173
according to the analysis to be performed (see below). At all stations, the temperature and
174
salinity at the sea bottom were measured using CTD (Conductivity, Temperature, and Depth)
175
probe equipped with previously calibrated sensors. 152
Materials & Methods 155
The Ancona harbor (water depth range, 4-15 m) is located in the western coast of central
156
Adriatic Sea (Fig. 1), it has a water sheet of 700,000 m2 and 5,400 m of docks (Spagnolo et al.,
157
2011). The harbor is one of the most important of the Adriatic Sea, with intense ferryboat and
158
merchant ship activity. Previous investigations reported that the area is subjected to organic 155
The Ancona harbor (water depth range, 4-15 m) is located in the western coast of central
156
Adriatic Sea (Fig. 1), it has a water sheet of 700,000 m2 and 5,400 m of docks (Spagnolo et al.,
157
2011). The harbor is one of the most important of the Adriatic Sea, with intense ferryboat and
158
merchant ship activity. Previous investigations reported that the area is subjected to organic PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed Extracts were concentrated using a RC 10.09 Vacuum
212
Concentration System (Jouan SA, Saint-Herblain, France) and purified by J.T.Baker™
213
BAKERBOND™ Octadecyl (C18) Solid Phase Extraction (SPE) cartridges (Avantor Inc.,
214
Radnor, PA, USA). PAHs were analyzed by High-Performance Liquid Chromatography (HPLC)
215
using both Fluorimetric (FLD) and (UV) Diode Array Detection (DAD) (Infinity 1260 Series,
216
Agilent Technologies, Santa Clara, CA, USA). 210
PAHs were determined after KOH-methanol extraction with a Mars 6 Microwave Digestion
211
System (CEM Corporation). Extracts were concentrated using a RC 10.09 Vacuum
212
Concentration System (Jouan SA, Saint-Herblain, France) and purified by J.T.Baker™
213
BAKERBOND™ Octadecyl (C18) Solid Phase Extraction (SPE) cartridges (Avantor Inc.,
214
Radnor, PA, USA). PAHs were analyzed by High-Performance Liquid Chromatography (HPLC)
215
using both Fluorimetric (FLD) and (UV) Diode Array Detection (DAD) (Infinity 1260 Series,
216
Agilent Technologies, Santa Clara, CA, USA). 212
Concentration System (Jouan SA, Saint-Herblain, France) and purified by J.T.Baker™
213
BAKERBOND™ Octadecyl (C18) Solid Phase Extraction (SPE) cartridges (Avantor Inc.,
214
Radnor, PA, USA). PAHs were analyzed by High-Performance Liquid Chromatography (HPLC)
215
using both Fluorimetric (FLD) and (UV) Diode Array Detection (DAD) (Infinity 1260 Series,
216
Agilent Technologies, Santa Clara, CA, USA). 217
For both HMs and PAHs, appropriated blank solutions and the Standard Reference Material
218
(SRM) 1944: New York/New Jersey Waterway Sediment (NIST – National Institute of Standards
219
and Technology), digested as samples, were used to check for accuracy, precision, and
220
recoveries of the employed analytical methodologies; concentrations obtained from SRM
221
analyses were always within the 95% confidence intervals of the NIST certified values. 222
The results obtained in this study were compared with threshold values for chemicals specified in
223
the Ministerial Decree 173/2016, the Italian normative that rules the management of dredged
224
sediments and sets out their quality. Only those values exceeding the upper thresholds values
225
(L2) are defined as alerting values (Table 2.5, Ministerial Decree 173/2016). 226
227
Meiofauna and macrofauna
228
229
For meiofauna samples, the content of each box-corer (three independent replicates) was sub-
230
sampled with PVC corers (inner diameter, 4.5 cm). The top 3 cm of sediment, where meiofaunal
231
organisms are typically more abundant, were preserved in 4% buffered formaldehyde. For
232
meiofaunal extraction, sediment samples were sieved through a 500 R mesh; a 32 R mesh
233
was used to retain the smallest metazoan organisms. Manuscript to be reviewed 199
Concentration of selected HMs and PAHs was determined in the surface sediment (0-3 cm)
200
collected at each sampling station and frozen at -20°C. Analyses were conducted following
201
previous validated methods, fully described in Benedetti et al. (2014) and Etiope et al. (2014). 202
In brief, HMs were determined after digestion under pressure with nitric acid and hydrogen
203
peroxide (7:1), using a Mars 6 Microwave Digestion System (CEM Corporation, Charlotte, NC,
204
USA). As, Cd, Cr, Cu, Fe, Mn, Ni, Pb, V, and Zn were analyzed by Atomic Absorption
205
Spectroscopy (AAS), with flame (SpectrAA 220FS Spectrometer, Varian Inc., Palo Alto, CA,
206
USA) and flameless atomization (240Z AA Spectrometer, Agilent Technologies Inc., Santa
207
Clara, CA, USA), while the Hg content was quantified by Cold Vapor Atomic Absorption
208
Spectroscopy (CVAAS; QuickTrace M-6100 Mercury Analyzer, Teledyne CETAC
209
Technologies Ltd., Omaha, NE, USA). 199
Concentration of selected HMs and PAHs was determined in the surface sediment (0-3 cm)
200
collected at each sampling station and frozen at -20°C. Analyses were conducted following
201
previous validated methods, fully described in Benedetti et al. (2014) and Etiope et al. (2014). 202
In brief, HMs were determined after digestion under pressure with nitric acid and hydrogen
203
peroxide (7:1), using a Mars 6 Microwave Digestion System (CEM Corporation, Charlotte, NC,
204
USA). As, Cd, Cr, Cu, Fe, Mn, Ni, Pb, V, and Zn were analyzed by Atomic Absorption
205
Spectroscopy (AAS), with flame (SpectrAA 220FS Spectrometer, Varian Inc., Palo Alto, CA,
206
USA) and flameless atomization (240Z AA Spectrometer, Agilent Technologies Inc., Santa
207
Clara, CA, USA), while the Hg content was quantified by Cold Vapor Atomic Absorption
208
Spectroscopy (CVAAS; QuickTrace M-6100 Mercury Analyzer, Teledyne CETAC
209
Technologies Ltd., Omaha, NE, USA). 210
PAHs were determined after KOH-methanol extraction with a Mars 6 Microwave Digestion
211
System (CEM Corporation). Extracts were concentrated using a RC 10.09 Vacuum
212
Concentration System (Jouan SA, Saint-Herblain, France) and purified by J.T.Baker™
213
BAKERBOND™ Octadecyl (C18) Solid Phase Extraction (SPE) cartridges (Avantor Inc.,
214
Radnor, PA, USA). PAHs were analyzed by High-Performance Liquid Chromatography (HPLC)
215
using both Fluorimetric (FLD) and (UV) Diode Array Detection (DAD) (Infinity 1260 Series,
216
Agilent Technologies, Santa Clara, CA, USA). 210
PAHs were determined after KOH-methanol extraction with a Mars 6 Microwave Digestion
211
System (CEM Corporation). 177
Environmental variables and microbial component
178 179
The content of each box-corer was sub-sampled with PVC corers (inner diameter, 4.5 cm) to
180
assess the biochemical composition of the organic matter and grain size. The top 3 cm of
181
sediment from three independent replicates for each parameter were frozen at -20°C, except for
182
the grain size determination, for which samples were kept at in situ temperature in single
183
replicate until brought to the laboratory. The biochemical composition of the organic matter
184
(total protein, carbohydrate, and lipid concentration) and chloroplastic pigments (chlorophyll-a
185
(Chl-a) and phaeopigment (Phaeo) concentration) were determined by standard techniques
186
(Danovaro, 2010). Concentrations were calculated using standard curves and normalized to
187
sediment dry weight after desiccation (60°C, 24 h). Biopolymeric organic carbon (BPC) was
188
calculated as the sum of the carbon equivalents of carbohydrates, proteins, and lipids (Fabiano et
189
al., 1995) and was used as a proxy for the available trophic resources. The value of the protein to
190
carbohydrate ratio (PRT/CHO) was utilized as descriptor of the nutritional quality of organic
191
matter in the sediment, with a PRT/CHO ratio >1.0 indicating relatively high quality and high
192
food availability (Pusceddu et al., 2010). 193
For grain size determination, aliquots of fresh sediment were sieved over a 63 R mesh. The two
194
fractions (>63 R sand; <63 R silt/mud) were dried in an oven at 60°C and weighed. Data
195
were expressed as a percentage of sediment total dry weight (Pusceddu et al., 2010). 198 PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 238
For macrofauna samples, the first 20 cm of three independent box-corer deployments were
239
sieved in situ using a 500 µm mesh and all organisms retained were preserved in 5% buffered
240
formaldehyde. Macrofauna was sorted in laboratory using a Leica S8APO stereomicroscope an
241
a Leica DM2500 binocular microscope, identified and classified to the lowest possible
242
taxonomic level, and quantified (no. of individuals m-2). The data collected was subjected to a
243
control and validation process and organized in a dedicated database. The nomenclature of the
244
species was verified and validated using the web portal https://www.marinespecies.org/ and the
245
‘WoRMS Taxon Match Tool’ (World Register of Marine Species Editorial Board, 2022). 246
Taxon richness (i.e., Orders or Classes and species for meio- and macrofauna, respectively; S),
247
total number of individuals per taxon/species (N), Shannon’s diversity index (H’, based on log2
248
Shannon & Weaver, 1949), and Pielou’s equitability index (J’; Pielou, 1969) of benthic
249
communities were calculated. 250
251
Statistical analysis
252
253
To assess differences in benthic communities among the sampling stations a Permutational
254
Multivariate Analysis of Variance (PERMANOVA; 9999, number of random unrestricted
255
permutations of raw data) was used (Anderson, 2001). The design included one factor: the
256
sampling station (five levels, fixed). The analysis was based on Bray-Curtis’ similarity of
257
previously fourth root transformed meio- and macrofaunal data. In case of significant difference
258
obtained by the main test, the pairwise test was performed and, as there was a limited number o
259
unique permutations, the p values were obtained from Monte Carlo tests (Anderson & Robinson
260
2003). Permutational Multivariate Analysis of Dispersion (PERMDISP) test was applied to
261
assess if differences among the sampling stations (between-group) were due to real differences
262
benthic community composition and not to differences in the multivariate dispersion of replicat
263
(within-group) among their respective centroids. A non-metric Multidimensional Scaling
264
(nMDS) ordination was carried out to visualize similarities among the sampling stations. 265
Similarity Percentages (SIMPER) analysis (cut-off, 90%) was used to identify the meio- and
266
macrofaunal taxa that contributed to the dissimilarity among the sampling stations. 267
One-way Analysis of Variance (ANOVA) was used to explore differences among the sampling
268
stations for organic matter content and microbial component, total abundance, diversity, and
269
equitability indices in meio- and macrofaunal benthic communities. Manuscript to be reviewed The latter fraction was centrifuged 3 times
234
with LUDOX® HS-40 colloidal silica (diluted with water to a final density of 1.18 g cmY#) and
235
stained with Rose Bengal (0.5 g LY5; Heip et al., 1985). Meiofaunal organisms were counted (no. 236
of individuals 10 cmY) and identified to the higher taxonomic level (i.e., Order and Class) under
237
a Leica S8APO stereomicroscope. 217
For both HMs and PAHs, appropriated blank solutions and the Standard Reference Material
218
(SRM) 1944: New York/New Jersey Waterway Sediment (NIST – National Institute of Standards
219
and Technology), digested as samples, were used to check for accuracy, precision, and
220
recoveries of the employed analytical methodologies; concentrations obtained from SRM
221
analyses were always within the 95% confidence intervals of the NIST certified values. 222
The results obtained in this study were compared with threshold values for chemicals specified in
223
the Ministerial Decree 173/2016, the Italian normative that rules the management of dredged
224
sediments and sets out their quality. Only those values exceeding the upper thresholds values
225
(L2) are defined as alerting values (Table 2.5, Ministerial Decree 173/2016). 226 217
For both HMs and PAHs, appropriated blank solutions and the Standard Reference Material
218
(SRM) 1944: New York/New Jersey Waterway Sediment (NIST – National Institute of Standards
219
and Technology), digested as samples, were used to check for accuracy, precision, and
220
recoveries of the employed analytical methodologies; concentrations obtained from SRM
221
analyses were always within the 95% confidence intervals of the NIST certified values. 222
The results obtained in this study were compared with threshold values for chemicals specified in
223
the Ministerial Decree 173/2016, the Italian normative that rules the management of dredged
224
sediments and sets out their quality. Only those values exceeding the upper thresholds values
225
(L2) are defined as alerting values (Table 2.5, Ministerial Decree 173/2016). PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed The ANOVA assumptions
270
were tested graphically plotting residuals vs. fitted values, normality of residuals and residuals
271
vs. covariate (factor station) to assess the variance homogeneity (Zuur et al., 2016). When
272
ANOVA showed significant differences, the Tukey's Honestly Significant Difference (HSD) te
273
was performed to find significant effects between different levels. 274
Since environmental variables have been processed in single replicates, a value of 1
275
(corresponding to the 70% of data around the mean in a normal standard curve; Quinn &
276
Keough, 2002) has been subtracted and added to create the missing replicates for which the mea 238
For macrofauna samples, the first 20 cm of three independent box-corer deployments were
239
sieved in situ using a 500 µm mesh and all organisms retained were preserved in 5% buffered
240
formaldehyde. Macrofauna was sorted in laboratory using a Leica S8APO stereomicroscope and
241
a Leica DM2500 binocular microscope, identified and classified to the lowest possible
242
taxonomic level, and quantified (no. of individuals m-2). The data collected was subjected to a
243
control and validation process and organized in a dedicated database. The nomenclature of the
244
species was verified and validated using the web portal https://www.marinespecies.org/ and the
245
‘WoRMS Taxon Match Tool’ (World Register of Marine Species Editorial Board, 2022). 246
Taxon richness (i.e., Orders or Classes and species for meio- and macrofauna, respectively; S),
247
total number of individuals per taxon/species (N), Shannon’s diversity index (H’, based on log2;
248
Shannon & Weaver, 1949), and Pielou’s equitability index (J’; Pielou, 1969) of benthic
249
communities were calculated. 250
251
Statistical analysis
252
253
To assess differences in benthic communities among the sampling stations a Permutational
254
Multivariate Analysis of Variance (PERMANOVA; 9999, number of random unrestricted
255
permutations of raw data) was used (Anderson, 2001). The design included one factor: the
256
sampling station (five levels, fixed). The analysis was based on Bray-Curtis’ similarity of
257
previously fourth root transformed meio- and macrofaunal data. In case of significant differences
258
obtained by the main test, the pairwise test was performed and, as there was a limited number of
259
unique permutations, the p values were obtained from Monte Carlo tests (Anderson & Robinson,
260
2003). Manuscript to be reviewed 253
To assess differences in benthic communities among the sampling stations a Permutational
254
Multivariate Analysis of Variance (PERMANOVA; 9999, number of random unrestricted
255
permutations of raw data) was used (Anderson, 2001). The design included one factor: the
256
sampling station (five levels, fixed). The analysis was based on Bray-Curtis’ similarity of
257
previously fourth root transformed meio- and macrofaunal data. In case of significant differences
258
obtained by the main test, the pairwise test was performed and, as there was a limited number of
259
unique permutations, the p values were obtained from Monte Carlo tests (Anderson & Robinson,
260
2003). Permutational Multivariate Analysis of Dispersion (PERMDISP) test was applied to
261
assess if differences among the sampling stations (between-group) were due to real differences in
262
benthic community composition and not to differences in the multivariate dispersion of replicates
263
(within-group) among their respective centroids. A non-metric Multidimensional Scaling
264
(nMDS) ordination was carried out to visualize similarities among the sampling stations. 265
Similarity Percentages (SIMPER) analysis (cut-off, 90%) was used to identify the meio- and
266
macrofaunal taxa that contributed to the dissimilarity among the sampling stations. 267
One-way Analysis of Variance (ANOVA) was used to explore differences among the sampling
268
stations for organic matter content and microbial component, total abundance, diversity, and
269
equitability indices in meio- and macrofaunal benthic communities. The ANOVA assumptions
270
were tested graphically plotting residuals vs. fitted values, normality of residuals and residuals
271
vs. covariate (factor station) to assess the variance homogeneity (Zuur et al., 2016). When
272
ANOVA showed significant differences, the Tukey's Honestly Significant Difference (HSD) test
273
was performed to find significant effects between different levels. 274
Since environmental variables have been processed in single replicates, a value of 1
275
(corresponding to the 70% of data around the mean in a normal standard curve; Quinn &
276
Keough, 2002) has been subtracted and added to create the missing replicates for which the mean 267
One-way Analysis of Variance (ANOVA) was used to explore differences among the sampling
268
stations for organic matter content and microbial component, total abundance, diversity, and
269
equitability indices in meio- and macrofaunal benthic communities. The ANOVA assumptions
270
were tested graphically plotting residuals vs. fitted values, normality of residuals and residuals
271
vs. covariate (factor station) to assess the variance homogeneity (Zuur et al., 2016). Manuscript to be reviewed Permutational Multivariate Analysis of Dispersion (PERMDISP) test was applied to
261
assess if differences among the sampling stations (between-group) were due to real differences in
262
benthic community composition and not to differences in the multivariate dispersion of replicates
263
(within-group) among their respective centroids. A non-metric Multidimensional Scaling
264
(nMDS) ordination was carried out to visualize similarities among the sampling stations. 265
Similarity Percentages (SIMPER) analysis (cut-off, 90%) was used to identify the meio- and
266
macrofaunal taxa that contributed to the dissimilarity among the sampling stations. 267
One-way Analysis of Variance (ANOVA) was used to explore differences among the sampling
268
stations for organic matter content and microbial component, total abundance, diversity, and
269
equitability indices in meio- and macrofaunal benthic communities. The ANOVA assumptions
270
were tested graphically plotting residuals vs. fitted values, normality of residuals and residuals
271
vs. covariate (factor station) to assess the variance homogeneity (Zuur et al., 2016). When
272
ANOVA showed significant differences, the Tukey's Honestly Significant Difference (HSD) test
273
was performed to find significant effects between different levels. 274
Since environmental variables have been processed in single replicates, a value of 1
275
(corresponding to the 70% of data around the mean in a normal standard curve; Quinn &
276
K
h 2002) h
b
bt
t d
d dd d t
t th
i i
li
t
f
hi h th 238
For macrofauna samples, the first 20 cm of three independent box-corer deployments were
239
sieved in situ using a 500 µm mesh and all organisms retained were preserved in 5% buffered
240
formaldehyde. Macrofauna was sorted in laboratory using a Leica S8APO stereomicroscope and
241
a Leica DM2500 binocular microscope, identified and classified to the lowest possible
242
taxonomic level, and quantified (no. of individuals m-2). The data collected was subjected to a
243
control and validation process and organized in a dedicated database. The nomenclature of the
244
species was verified and validated using the web portal https://www.marinespecies.org/ and the
245
‘WoRMS Taxon Match Tool’ (World Register of Marine Species Editorial Board, 2022). 246
Taxon richness (i.e., Orders or Classes and species for meio- and macrofauna, respectively; S),
247
total number of individuals per taxon/species (N), Shannon’s diversity index (H’, based on log2;
248
Shannon & Weaver, 1949), and Pielou’s equitability index (J’; Pielou, 1969) of benthic
249
communities were calculated. Manuscript to be reviewed By doing so, groups of variables were created each of which
303
represented by a single PCA score, and all the scores obtained were used as covariates in the
304
above DistLM analysis to assess the relations between faunal communities and the
305
environmental features. A distance-based Redundancy Analysis (dbRDA) was then applied to
306
visually investigate the relationship between the community assemblages and the environmental
307
data (Anderson et al., 2008). All the above statistical tests were considered significant at ^ =
308
0.05. Moreover, the RELATE routine (a Mantel test like analysis) was used to search for cross-
309
taxon correlations between meio- and macrofaunal similarity matrices. 310
The PERMANOVA, PERMDISP, nMDS, SIMPER, PCA, RELATE and DistLM procedures
311
were performed with PRIMER™ and PERMANOVA+ ecological software (Clarke & Gorley,
312
2006; Anderson et al., 2008); the ANOVA, the Tukey’s test and the VARIMAX procedures
313
were performed using the free software R (R Core Team, 2018; v. 4.1.3). 314 277
correspond to the original value. Two more values have been created, one lower and one higher
278
than the actual value, and the average of the three values given exactly the observed value. 279
The number of environmental and pollutant-related variables were separated in two groups:
280
pollutants (comprising HMs and PAHs) and environmental variables (comprising the organic
281
matter compounds and silt/mud percentage). Since many compounds had linked each other,
282
multicollinearity (cut-off at Spearman’s correlation = |0.8|) among variables was assessed to
283
reduce the dataset and avoid problems in the analysis algorithms. In case of multicollinearity, the
284
variables with a more stringent biological or environmental value were retained acting as proxy
285
for the omitted collinear ones. A summary of the collinear variable and which ones have been
286
chosen is given in Table S1. After normalization, both type of variables was used to characterize
287
the study area by means of a Principal Component Analysis (PCA). 305
environmental features. A distance-based Redundancy Analysis (dbRDA) was then applied to
306
visually investigate the relationship between the community assemblages and the environmental
307
data (Anderson et al., 2008). All the above statistical tests were considered significant at ^ =
308
0.05. Moreover, the RELATE routine (a Mantel test like analysis) was used to search for cross-
309
taxon correlations between meio- and macrofaunal similarity matrices. Manuscript to be reviewed When
272
ANOVA showed significant differences, the Tukey's Honestly Significant Difference (HSD) test
273
was performed to find significant effects between different levels. 274
Since environmental variables have been processed in single replicates, a value of 1
275
(corresponding to the 70% of data around the mean in a normal standard curve; Quinn & PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 277
correspond to the original value. Two more values have been created, one lower and one higher
278
than the actual value, and the average of the three values given exactly the observed value. 279
The number of environmental and pollutant-related variables were separated in two groups:
280
pollutants (comprising HMs and PAHs) and environmental variables (comprising the organic
281
matter compounds and silt/mud percentage). Since many compounds had linked each other,
282
multicollinearity (cut-off at Spearman’s correlation = |0.8|) among variables was assessed to
283
reduce the dataset and avoid problems in the analysis algorithms. In case of multicollinearity, the
284
variables with a more stringent biological or environmental value were retained acting as proxy
285
for the omitted collinear ones. A summary of the collinear variable and which ones have been
286
chosen is given in Table S1. After normalization, both type of variables was used to characterize
287
the study area by means of a Principal Component Analysis (PCA). 288
Spearman’s rank correlation analysis was performed to test relationships between meio- and
289
macrofauna total abundance, taxon richness, Shannon’s diversity (Shannon & Weaver, 1949) and
290
Pieolu’s equitability (Pielou, 1966) indices, and the environmental features considered. 291
In order to verify the existence of a significant relation between the benthos data matrix and the
292
environmental data (pollutants and environmental variables), the distance-based Linear Modeling
293
(DistLM) procedure was utilized with a backward selection of the variables, and each model
294
assessed by means of the Akaike’s Information Criterion corrected for small samples (AICc;
295
Anderson et al., 2008). Since the environmental and pollutant variables were too numerous (even
296
after multicollinearity assessment) respect to the meio- and macrofaunal ones, to further reduce
297
their number, both type of variables was analyzed (separately) by means of another (PCA) to
298
identify groups with similar variability. Only those principal components showing eigenvalues
299
>1 (Kaiser-Guttman criterion; Zwick & Velicer, 1986) were considered. To obtain a better
300
insight into the output loadings, the orthogonal varimax rotation of extracted PCA components
301
was performed. After a varimax rotation, each original variable tends to be associated with one
302
(or a small number) of PCA axis. All the sediment samples showed a
340
distinct predominance of Low Molecular Weight PAHs, mainly driven by the Naphthalene
341
concentration (exceeding the legislative thresholds at MAN, DS, and LR stations), which
342
averagely accounted for 37% of the a19 PAHs, followed by its methylated isomers: 1- and 2-
343
Methylnaphtalene (24 and 22%, respectively). A prevalence of volatile with 2-3 aromatic rings
344
was reported at all stations (Fig. 2a) (for wind direction and speed on Ancona harbor area during
345
February 2015, please see Fig. S1). Excluding Naphthalene and its related compounds, the PAH
346
residual contamination was mainly ascribable to Phenanthrene and Fluoranthene (Table S4), and
347
the principal PAH diagnostic ratios (Tobiszewski & b; 2012), commonly used as a tool
348
to discriminate the analyte origin and sources (Giuliani et al., 2019; Pizzini et al., 2021),
349
indicated a petrogenic origin (Fig. 2b). 350
The results from PCA plot considering the environmental variables, summarized the differences
351
among the sampling stations (Fig. 3). In details, MAN station clearly separated from the other Manuscript to be reviewed 310
The PERMANOVA, PERMDISP, nMDS, SIMPER, PCA, RELATE and DistLM procedures
311
were performed with PRIMER™ and PERMANOVA+ ecological software (Clarke & Gorley,
312
2006; Anderson et al., 2008); the ANOVA, the Tukey’s test and the VARIMAX procedures
313
were performed using the free software R (R Core Team, 2018; v. 4.1.3). 310
The PERMANOVA, PERMDISP, nMDS, SIMPER, PCA, RELATE and DistLM procedures
311
were performed with PRIMER™ and PERMANOVA+ ecological software (Clarke & Gorley,
312
2006; Anderson et al., 2008); the ANOVA, the Tukey’s test and the VARIMAX procedures
313
were performed using the free software R (R Core Team, 2018; v. 4.1.3). PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 316
317
Environmental features and contamination levels in the Ancona harbor
318
319
Bottom water temperature reported a mean value of 10.3±0.2°C and salinity ranged from a
320
minimum of 31.3 (at LR station) to a maximum of 37.6 PSU (at API station). The sandy fraction
321
(>63 R , characterized stations located outside the harbor (DS, LR, and API stations), while the
322
silty muddy fraction (<63 R , was predominant in the inner stations MAN and PORT (Table
323
S2). Chl-a and organic matter contents into the sediments showed significantly (F4, 10 = 12.97, p
324
= 0.001) higher values of ‘fresh’ material (i.e., Chl-a) at MAN and DS stations, while the BPC
325
significantly (F4,10 = 22.21, p = 0.001) decreased from inside to outside the harbor (Table S2). 326
The significant (F4,10 = 12.61, p = 0.001) highest value in the quality of organic matter (i.e. 327
PRT/CHO) was reported at PORT station (Table S2), due to a particularly low concentration of
328
CHO into the sediment of this station (Table S3). 329
Analyses aimed at determining HM concentrations in the sediment of the Ancona harbor have
330
highlighted a general decrease of contamination values moving from the innermost sampling
331
station (MAN) to those one more external (Table S4). MAN station turned out to be affected by
332
pollution levels 3-4 times higher than those detected at API station, reaching even Cu and Zn
333
concentrations 13 and 6 times higher, respectively. Just these two elements, Cu and Zn, were the
334
only exceeding the threshold limits imposed by the Ministerial Decree 173/2016 (46.31 µg g-1 of
335
Cu detected at MAN station; 297.2 and 114.9 µg g-1 of Zn detected at MAN and PORT stations,
336
respectively). In the remaining sampling stations, there were no overruns (Table S4). 337
The concentration levels of a19 PAHs ranged from 73.38 to 213.4 µg g-1 following, although not
338
linearly, the same pattern highlighted for HMs, with higher values in the innermost sampling
339
station (MAN), decreasing towards outer ones (Table S4). 317
Environmental features and contamination levels in the Ancona harbo 319
Bottom water temperature reported a mean value of 10.3±0.2°C and salinity ranged from a
320
minimum of 31.3 (at LR station) to a maximum of 37.6 PSU (at API station). The sandy fraction
321
(>63 R , characterized stations located outside the harbor (DS, LR, and API stations), while the
322
silty muddy fraction (<63 R , was predominant in the inner stations MAN and PORT (Table
323
S2). Chl-a and organic matter contents into the sediments showed significantly (F4, 10 = 12.97, p
324
= 0.001) higher values of ‘fresh’ material (i.e., Chl-a) at MAN and DS stations, while the BPC
325
significantly (F4,10 = 22.21, p = 0.001) decreased from inside to outside the harbor (Table S2). 326
The significant (F4,10 = 12.61, p = 0.001) highest value in the quality of organic matter (i.e. 327
PRT/CHO) was reported at PORT station (Table S2), due to a particularly low concentration of
328
CHO into the sediment of this station (Table S3). 329
Analyses aimed at determining HM concentrations in the sediment of the Ancona harbor have
330
highlighted a general decrease of contamination values moving from the innermost sampling
331
station (MAN) to those one more external (Table S4). MAN station turned out to be affected by
332
pollution levels 3-4 times higher than those detected at API station, reaching even Cu and Zn
333
concentrations 13 and 6 times higher, respectively. Just these two elements, Cu and Zn, were the
334
only exceeding the threshold limits imposed by the Ministerial Decree 173/2016 (46.31 µg g-1 of
335
Cu detected at MAN station; 297.2 and 114.9 µg g-1 of Zn detected at MAN and PORT stations,
336
respectively). In the remaining sampling stations, there were no overruns (Table S4). 329
Analyses aimed at determining HM concentrations in the sediment of the Ancona harbor have
330
highlighted a general decrease of contamination values moving from the innermost sampling
331
station (MAN) to those one more external (Table S4). MAN station turned out to be affected by
332
pollution levels 3-4 times higher than those detected at API station, reaching even Cu and Zn
333
concentrations 13 and 6 times higher, respectively. Manuscript to be reviewed Manuscript to be reviewed 356
and Fluoranthene into the sediments and by intermediate values between the innermost stations
357
and the outermost station for most of the pollutants and environmental variables considered. 358
359
Meiobenthic assemblages
360
361
A total of 12 taxa were identified with Nematoda representing the dominant taxon at all sampling
362
stations, with a percentage ranging from 75 to 95% (Fig. 4a). The second most represented taxon
363
was Copepoda with their nauplii; among less represented taxa (i.e., others) Bivalvia, Ciliata,
364
Foraminifera, Kinorhyncha, Oligochaeta, Ostracoda, Platyhelminthes, Polychaeta, Sipuncula,
365
and Tardigrada constituted from 3 to 15% of the meiobenthic community (Fig. 4b). Meiofauna
366
abundance and values of its diversity indices are reported in Table S5 and represented in Figure
367
5. The ANOVA tests detected significant differences among all the sampling stations for taxa
368
richness (N; Table 1), with the highest value reported at DS station (Fig. 5a). The ANOVA tests
369
reported significant differences among the sampling stations also for all diversity and equitability
370
indices (Table 1 and Table S5). Regarding the number of taxa, PORT station showed a
371
significantly lower value on average compared to all the other sampling stations (Fig. 5b). 372
Shannon’s diversity and Pielou’s equitability indices showed both similar patterns, with MAN
373
station presenting the highest value (Fig. 5c and Fig. 5d). 374
PERMANOVA analysis with pairwise test reported significant differences (Table 2) in
375
meiofaunal community composition among all the sampling stations; PERMDISP test did not
376
show any significant dispersion around centroids, confirming that the differences among the
377
sampling stations were due to a real difference in meiobenthic composition (Pseudo-F2,4 = 5.38,
378
P(perm) = 0.088). 379
The nMDS plot (Fig. 6a) shows the separation among the sampling stations, as well as a low
380
inter-replica variability. In particular, the innermost MAN and PORT stations were separated
381
from to the outermost ones. SIMPER analysis detected a dissimilarity percentage from 15 (LR
382
vs. DS stations) to 54% (PORT vs. LR stations). The highest values of dissimilarity percentage
383
l
i
d
i h PORT
i
d
i l d
l
b
d
(i
i d 356
and Fluoranthene into the sediments and by intermediate values between the innermost stations
357
and the outermost station for most of the pollutants and environmental variables considered. Manuscript to be reviewed 358 361
A total of 12 taxa were identified with Nematoda representing the dominant taxon at all sampling
362
stations, with a percentage ranging from 75 to 95% (Fig. 4a). The second most represented taxon
363
was Copepoda with their nauplii; among less represented taxa (i.e., others) Bivalvia, Ciliata,
364
Foraminifera, Kinorhyncha, Oligochaeta, Ostracoda, Platyhelminthes, Polychaeta, Sipuncula,
365
and Tardigrada constituted from 3 to 15% of the meiobenthic community (Fig. 4b). Meiofauna
366
abundance and values of its diversity indices are reported in Table S5 and represented in Figure
367
5. The ANOVA tests detected significant differences among all the sampling stations for taxa
368
richness (N; Table 1), with the highest value reported at DS station (Fig. 5a). The ANOVA tests
369
reported significant differences among the sampling stations also for all diversity and equitability
370
indices (Table 1 and Table S5). Regarding the number of taxa, PORT station showed a
371
significantly lower value on average compared to all the other sampling stations (Fig. 5b). 372
Shannon’s diversity and Pielou’s equitability indices showed both similar patterns, with MAN
373
station presenting the highest value (Fig. 5c and Fig. 5d). 372
Shannon’s diversity and Pielou’s equitability indices showed both similar patterns, with MAN
373
station presenting the highest value (Fig. 5c and Fig. 5d). 374
PERMANOVA analysis with pairwise test reported significant differences (Table 2) in
375
meiofaunal community composition among all the sampling stations; PERMDISP test did not
376
show any significant dispersion around centroids, confirming that the differences among the
377
sampling stations were due to a real difference in meiobenthic composition (Pseudo-F2,4 = 5.38,
378
P(perm) = 0.088). 374
PERMANOVA analysis with pairwise test reported significant differences (Table 2) in
375
meiofaunal community composition among all the sampling stations; PERMDISP test did not
376
show any significant dispersion around centroids, confirming that the differences among the
377
sampling stations were due to a real difference in meiobenthic composition (Pseudo-F2,4 = 5.38,
378
P(perm) = 0.088). 379
The nMDS plot (Fig. 6a) shows the separation among the sampling stations, as well as a low
380
inter-replica variability. In particular, the innermost MAN and PORT stations were separated
381
from to the outermost ones. SIMPER analysis detected a dissimilarity percentage from 15 (LR
382
vs. DS stations) to 54% (PORT vs. LR stations). 317
Environmental features and contamination levels in the Ancona harbo Just these two elements, Cu and Zn, were the
334
only exceeding the threshold limits imposed by the Ministerial Decree 173/2016 (46.31 µg g-1 of
335
Cu detected at MAN station; 297.2 and 114.9 µg g-1 of Zn detected at MAN and PORT stations,
336
respectively). In the remaining sampling stations, there were no overruns (Table S4). 337
The concentration levels of a19 PAHs ranged from 73.38 to 213.4 µg g-1 following, although not
338
linearly, the same pattern highlighted for HMs, with higher values in the innermost sampling
339
station (MAN), decreasing towards outer ones (Table S4). All the sediment samples showed a
340
distinct predominance of Low Molecular Weight PAHs, mainly driven by the Naphthalene
341
concentration (exceeding the legislative thresholds at MAN, DS, and LR stations), which
342
averagely accounted for 37% of the a19 PAHs, followed by its methylated isomers: 1- and 2-
343
Methylnaphtalene (24 and 22%, respectively). A prevalence of volatile with 2-3 aromatic rings
344
was reported at all stations (Fig. 2a) (for wind direction and speed on Ancona harbor area during
345
February 2015, please see Fig. S1). Excluding Naphthalene and its related compounds, the PAH
346
residual contamination was mainly ascribable to Phenanthrene and Fluoranthene (Table S4), and
347
the principal PAH diagnostic ratios (Tobiszewski & b; 2012), commonly used as a tool
348
to discriminate the analyte origin and sources (Giuliani et al., 2019; Pizzini et al., 2021),
349
indicated a petrogenic origin (Fig. 2b). 350
The results from PCA plot considering the environmental variables, summarized the differences
351
among the sampling stations (Fig. 3). In details, MAN station clearly separated from the other
352
stations being characterized by higher values in pollutants, food sources and mud content into the
353
sediment. This explained the variability along the first axis (41.8% of variation). The variability
354
along the second axis (23% of variation) was mainly explained by the contrast between the DS
355
and LR stations and all the others, the latter characterized by the lowest values in mud content PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed The highest values of dissimilarity percentage
383
were always associated with PORT station and mainly due to very low abundances (i.e., ind. 384
10cm-2 50 ± 8) in some of the most represented taxa such as: Copepoda, Foraminifera, Nematoda
385
and Polychaeta (Table S6). For the other sampling stations, the dissimilarity was mainly due to
386
the presence/absence or differences in the abundances of the taxa others: Bivalvia, Ciliata,
387
Kinorhyncha, Oligochaeta, Ostracoda, Platyhelminthes and Sipuncula (Table S6). Several
388
positive correlations emerged between meiofauna descriptors and quantity of organic matter;
389
moreover, meiofauna abundance and its diversity were positively correlated to some HMs (i.e.,
390
Cd, Cu, Hg, Pb and Zn) and to silt/mud content into the sediment (Table S7a). The best model
391
selected by the DistLM analysis, reported in Table 3, comprised only the pollutant compounds. 392
The resulting dbRDA showed that the first two axes explained 83.4% of the variance of
393
meiofaunal community composition, corresponding to the combination of the following
394
environmental factors: V/Ni ratio (commonly used as a diagnostic marker of maritime traffic; PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) 398
Macrofauna assemblages 400
A total of 93 taxa were identified, these included: 43 Annelida (42 Polychaeta and 1
401
Oligochaeta), 29 Mollusca (22 Bivalvia, 6 Gastropoda, and 1 Scaphopoda), 10 Crustacea (6
402
Amphipoda, 2 Cumacea, 1 Isopoda, and 1 Tanaidacea), 5 Nematoda, 2 Bryozoa, 2 Cnidaria (1
403
Anthozoa and 1 Hydrozoa), 1 Nemertea, and 1 Ophiuroidea (Table S8). Annelida was the most
404
represented group (from 68 to 90% at LR and MAN stations, respectively), followed by
405
Mollusca (from 4 to 24% at MAN and LR stations, respectively) at all the sampling stations. 406
Other less represented groups such as Cnidaria, Isopoda, Ophiuroidea, and Tanaidacea were
407
found only at one or two sampling stations (Fig. 7). Macrofauna abundance ranged from 610 ±
408
241 to 3455 ± 425 individuals m-2 at MAN and LR stations, respectively; values of its diversity
409
indices are reported in Table S5 are represented in Figure 8. The ANOVA tests reported
410
significant higher abundance value at LR station (Fig. 8a) compared to all the other stations
411
(Table 1). LR station was also characterized by the highest number of taxa (Fig. 8b) but also by
412
lowest values (Fig. 8c and Fig. 8d) of Shannon’s diversity and Pielou’s equitability indices
413
(Table 1 and Table S5). PERMANOVA analysis with pairwise test reported significant
414
differences (Table 4) in macrofaunal community composition among the majority of the
415
sampling stations; PERMDISP test did not show any significant dispersion around centroids,
416
confirming that the differences among the sampling stations were due to a real difference in
417
macrobenthic composition (Pseudo-F2,4 = 0.99, P(perm) = 0.767). 418
The nMDS plot (Fig. 6b) shows the separation among the sampling stations. In detail, the
419
innermost MAN station and for a lesser extend the outermost API station were distinguished
420
from the others. SIMPER analysis reported high percentages of dissimilarity ranging from 63.4
421
(LR vs. DS stations) to 88.3% (MAN vs. API stations) between all pairs of sampling stations. 422
The great dissimilarity is mainly due to the presence/absence of species or to a particularly
423
abundant presence of them in one station compared to the others (Table S9). Manuscript to be reviewed 395
Viana et al., 2014), Naphthalene, and Benzo[a]pyrene (ARC1) and Zn, Fluorene, and percentage
396
of silt/mud (ARC4; Fig. S2a). 395
Viana et al., 2014), Naphthalene, and Benzo[a]pyrene (ARC1) and Zn, Fluorene, and percentage
396
of silt/mud (ARC4; Fig. S2a). 396
of silt/mud (ARC4; Fig. S2a). Manuscript to be reviewed 435
axes explained the 42.9% of the variance in the macrofaunal community composition (Fig. S2b). 436
This latter low percentage indicates that the residual variance associated to the community was
437
not captured by the graph, and it is likely that an unobserved variable should have had improved
438
the general plot. In any case, along the first axis of the dbRDA there was a clear separation
439
between the innermost stations MAN and PORT, characterized by high values of Zn and
440
percentage of silt/mud compared to the outermost stations LR, DS, and API; while along the
441
second axis LR and DS stations were separated from API, MAN, and PORT stations. 442
443
444
Discussion
445
446
Environmental features of Ancona harbor
447
448
The presence of muddy sediments in the innermost stations and sand in the outermost ones was
449
clearly due to a reduced exposure to hydrological factors (wind, waves, and currents) which
450
create conditions of poor water renewal inside the Ancona harbor, favoring the presence of fine
451
sediments (Spagnolo et al., 2011). The sediment particle size can influence sediment organic
452
matter load and pollutant content, with fine-grained components commonly showing a high
453
content in organic matter and pollutants (Papageorgiou et al., 2010). This might have facilitated
454
an overall accumulation of HMs and PAHs inside the harbor, as reported in other harbor areas
455
(e.g., McCready et al., 2006; Losi et al., 2021). In particular, the high values of Cu and Zn
456
detected at MAN and PORT stations were likely correlated to the shipyard activities present
457
within the harbor, especially with the use of new generation antifouling paints (Costa et al.,
458
2016; Pereira et al., 2018). Furthermore, the pronounced prevalence of volatile, easily
459
transportable PAHs (e.g., Naphthalene; Fig. 2a) pointed out that fuel combustion linked to
460
maritime traffic was the major source of these organic contaminants in the harbor basin. The
461
petrogenic origin of the PAH residual contamination (e.g., Phenanthrene and Fluoranthene; Fig. 462
2b) was supported by the detection in the sediment samples of V, Ni, and Pb, commonly
463
considered as tracers of accidental oil spills and/or marine fuels (El Nemr et al., 2006), as well as
464
by the values of the V/Ni ratio, marker of an intense maritime traffic (Viana et al., 2014). 398
Macrofauna assemblages Species like
424
Ampelisca diadema, Aponuphis bilineata, Jasmineira caudata, Spiophanes bombyx, Kurtiella
425
bidentata, and Euclymene oerstedii characterized mainly the outermost API station, while some
426
other species such as Tubificoides swirencoides, Streblospio sp., Heteromastus filiformis, and
427
Chaetozone caputesocis were found inhabiting the innermost stations (Table S9). Four
428
significant correlations (three out of four were negative) were detected between macrofauna
429
descriptors and environmental variables; only macrofauna species richness was (negatively)
430
correlated to three TEs and to the percentage of finest sediment fraction (Table S7b). 431
The best model selected by the DistLM analysis, reported in Table 5, comprised only ARC2
432
(Benzo[ghi]perylene, and Anthracene/(Anthracene + Phenanthrene) and
433
Fl
h
/(Fl
h
P
) di
i
i
T bi
ki
d
b ; 2012)
d 433
Fluoranthene/(Fluoranthene + Pyrene) diagnostic ratios; Tobiszewski and b; 2012) and
434
ARC4 (Zn, Fluorene, and percentage of silt/mud).The resulting dbRDA showed that the first two 433
Fluoranthene/(Fluoranthene + Pyrene) diagnostic ratios; Tobiszewski and b; 2012) and
434
ARC4 (Zn, Fluorene, and percentage of silt/mud).The resulting dbRDA showed that the first two 434
ARC4 (Zn, Fluorene, and percentage of silt/mud).The resulting dbRDA showed that the first two 434
ARC4 (Zn, Fluorene, and percentage of silt/mud).The resulting dbRDA showed that the first two PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 465
Considering the threshold values for chemicals specified in the Ministerial Decree 173/2016, few
466
values were reported exceeding the established alerting thresholds and always from the
467
innermost stations. 468
Not only contaminants, but also natural and/or anthropogenic changes in the benthic trophic
469
status (i.e., organic matter quantity and sediment biochemical composition) may affect the
470
benthic communities (Pusceddu et al., 2011; Foti et al., 2014). The protein, carbohydrate, lipid,
471
and BPC content in the sediments have been proposed and utilized to assess the benthic trophic
472
status of marine coastal environments, including the Adriatic Sea (Vezzulli & Fabiano, 2006). In
473
Dell’Anno et al. (2002) PRT and CHO sedimentary contents were suggested as proxies for
474
organic matter quality and threshold values were established for ranking the trophic status and 435
axes explained the 42.9% of the variance in the macrofaunal community composition (Fig. S2b). 436
This latter low percentage indicates that the residual variance associated to the community was
437
not captured by the graph, and it is likely that an unobserved variable should have had improved
438
the general plot. In any case, along the first axis of the dbRDA there was a clear separation
439
between the innermost stations MAN and PORT, characterized by high values of Zn and
440
percentage of silt/mud compared to the outermost stations LR, DS, and API; while along the
441
second axis LR and DS stations were separated from API, MAN, and PORT stations. 442 448
The presence of muddy sediments in the innermost stations and sand in the outermost ones was
449
clearly due to a reduced exposure to hydrological factors (wind, waves, and currents) which
450
create conditions of poor water renewal inside the Ancona harbor, favoring the presence of fine
451
sediments (Spagnolo et al., 2011). The sediment particle size can influence sediment organic
452
matter load and pollutant content, with fine-grained components commonly showing a high
453
content in organic matter and pollutants (Papageorgiou et al., 2010). This might have facilitated
454
an overall accumulation of HMs and PAHs inside the harbor, as reported in other harbor areas
455
(e.g., McCready et al., 2006; Losi et al., 2021). Manuscript to be reviewed 475
the environmental quality of coastal marine ecosystems. Applying those thresholds to the
476
investigated sediments, the trophic status of Ancona harbor could be ranked as hyper-trophic
477
(PRT >4 mg g-1) with the exception of the API station, ranked as eutrophic (PRT e 4 mg g-1). 478
However, in terms of CHO content, Ancona harbor should be ranked as meso-oligotrophic
479
system (CHO <5 mg g-1). Pusceddu et al. (2009; 2011) identified as eutrophic systems those
480
characterized by BPC concentration >3 mgC g-1, as found at MAN, PORT, LR, and DS stations,
481
and as mesotrophic systems those characterized by BPC concentration in the range 1-3 mgC g-1,
482
as in the case of the API station. In Ancona harbor PRT/CHO ratio resulted always >1, indicating
483
a great input of recent production’s material and highlighting the good trophic quality of the
484
organic matter (Pusceddu et al., 2009). Increasing of organic loads in the sediment have been
485
usually reported from other harbor areas, especially from the innermost zones (Danulat et al.,
486
2002; Covazzi-Harriague et al., 2012; Xu et al., 2014; Rebai et al., 2022). Concentrations of Chl-
487
a here reported were extremely high if compared to those reported in February in a previous
488
study conducted along the Adriatic coasts (0.11-0.23 µg g-1; Bianchelli et al., 2016), indicating
489
the presence of ‘fresh’ primary organic matter. However, similar results in Chl-a concentrations
490
were reported from Tunisian harbors (Rebai et al., 2022) along with high level of organic matter. 491
The PCA on measured environmental variables indicated the presence of a clear spatial
492
heterogeneity among the sampling stations and a separation between innermost stations and
493
outermost ones due to higher organic matter and contaminant loads inside the harbor basin, as
494
previously reported in the same study area (Spagnolo et al., 2011; Baldrighi et al., 2019) and in
495
other enclosed systems (Guerra-Garcìa et al., 2003; Vezzulli et al., 2003; Losi et al., 2013;
496
Dauvin et al., 2017; Mehlhorn et al., 2021). This marked environmental variability in harbors is
497
a common feature. Manuscript to be reviewed Indeed, environmental disturbance within harbors may change rapidly over
498
spatial scales of a few meters, depending on various factors like the localization and magnitude
499
of pollution sources, allochthonous inputs of different nature, tidal regime, water circulation,
500
harbor position, shape, and size (McCready et al., 2006; Vassallo et al., 2006; Xu et al., 2014). 501 499
of pollution sources, allochthonous inputs of different nature, tidal regime, water circulation,
500
harbor position, shape, and size (McCready et al., 2006; Vassallo et al., 2006; Xu et al., 2014). 501
502
Meiofaunal response to harbor environmental conditions
503
504
In Ancona harbor, the meiofaunal total abundance, community structure and, for a lesser extent,
505
univariate measures (i.e., diversity and equitability indices) reflected the marked spatial
506
heterogeneity showed by the PCA and the clear separation both between inner and outer
507
sampling stations and among the sampling stations themselves (MAN vs. PORT vs. DS + LR vs. 508
API). Meiofaunal abundance was in the range of values reported by Baldrighi et al. (2019) for
509
Ancona harbor (i.e. from 912±404 to 2512±717 ind.10cm-2) with lower values characterizing the
510
outermost station and for other harbors and coastal areas affected by pollution and/or high
511
organic matter loads (Vezzulli et al., 2003; Veiga et al., 2009; Pusceddu et al., 2011; Dal Zotto
512
et al., 2016; Semprucci et al., 2016; Sedano et al., 2014; Kulakova, 2022). The only exception
513
was represented by the paucity of meiofaunal organisms found at PORT station. Considering that
514
total meiofaunal abundance was positively linked to products derived from primary production Manuscript to be reviewed In particular, the high values of Cu and Zn
456
detected at MAN and PORT stations were likely correlated to the shipyard activities present
457
within the harbor, especially with the use of new generation antifouling paints (Costa et al.,
458
2016; Pereira et al., 2018). Furthermore, the pronounced prevalence of volatile, easily
459
transportable PAHs (e.g., Naphthalene; Fig. 2a) pointed out that fuel combustion linked to
460
maritime traffic was the major source of these organic contaminants in the harbor basin. The
461
petrogenic origin of the PAH residual contamination (e.g., Phenanthrene and Fluoranthene; Fig. 462
2b) was supported by the detection in the sediment samples of V, Ni, and Pb, commonly
463
considered as tracers of accidental oil spills and/or marine fuels (El Nemr et al., 2006), as well as
464
by the values of the V/Ni ratio, marker of an intense maritime traffic (Viana et al., 2014). 465
Considering the threshold values for chemicals specified in the Ministerial Decree 173/2016, few
466
values were reported exceeding the established alerting thresholds and always from the
467
innermost stations. 468
Not only contaminants, but also natural and/or anthropogenic changes in the benthic trophic
469
status (i.e., organic matter quantity and sediment biochemical composition) may affect the
470
benthic communities (Pusceddu et al., 2011; Foti et al., 2014). The protein, carbohydrate, lipid,
471
and BPC content in the sediments have been proposed and utilized to assess the benthic trophic
472
status of marine coastal environments, including the Adriatic Sea (Vezzulli & Fabiano, 2006). In
473
Dell’Anno et al. (2002) PRT and CHO sedimentary contents were suggested as proxies for
474
organic matter quality and threshold values were established for ranking the trophic status and PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 515
(Chl-a, Phaeo, and chloroplast pigment equivalents - CPE), its low abundance at PORT station
516
could be partially justified by the lowest detected value of ‘fresh’ material (Chl-a) and/or other
517
factors known to regulate the coexistence of different organisms such as competition for
518
resources and predation (Schratzberger et al., 2003; Schratzberger et al., 2008). Moreover, given
519
its small size, low mobility, and lack of dispersive life stages (Giere, 2009), meiofauna is more
520
susceptible to within-habitat physical variability and environmental disturbances than larger,
521
more mobile, and potentially more highly dispersed members of the macrofauna (Austen &
522
Widdicombe, 2006; Schratzberger et al., 2008). This would explain the drop in meiofauna
523
abundance, not reported for the macrofauna, at that sampling station. A total of 12 meiofauna
524
taxa (Orders or Classes) were identified and in four sampling stations out of five the majority of
525
taxa were represented. The measures of diversity (i.e., S, H’, and J’) were comparable to the
526
values reported in harbor areas (e.g., Moreno et al., 2008a; Moreno et al., 2008b; Sedano et al.,
527
2014; Kulakova, 2022) and in enclosed/transitional systems in the Adriatic Sea (e.g., Pusceddu et
528
al., 2007; Pusceddu et al., 2011; Frontalini et al., 2014; Covazzi-Harriague et al., 2012). 529
However, the strong dominance of Nematoda justified the low values reported for the Pielou’s
530
equitability index, particularly at LR and DS stations. The dominance of the most resistant and
531
adaptable group is a peculiarity of more stressed, less stable environments and characterized by
532
fine sediment fraction, such as harbors (Semprucci et al., 2015). Abundance and diversity indices
533
were correlated to different proxies of food sources (quantity and quality) into the sediment and
534
to its grain size, confirming the effect of these environmental variables on meiofaunal
535
populations (Balsamo et al., 2010). 515
(Chl-a, Phaeo, and chloroplast pigment equivalents - CPE), its low abundance at PORT station
516
could be partially justified by the lowest detected value of ‘fresh’ material (Chl-a) and/or other
517
factors known to regulate the coexistence of different organisms such as competition for
518
resources and predation (Schratzberger et al., 2003; Schratzberger et al., 2008). Manuscript to be reviewed The dominance of the most resistant and
531
adaptable group is a peculiarity of more stressed, less stable environments and characterized by
532
fine sediment fraction, such as harbors (Semprucci et al., 2015). Abundance and diversity indices
533
were correlated to different proxies of food sources (quantity and quality) into the sediment and
534
to its grain size, confirming the effect of these environmental variables on meiofaunal
535
populations (Balsamo et al 2010) 529
However, the strong dominance of Nematoda justified the low values reported for the Pielou’s
530
equitability index, particularly at LR and DS stations. The dominance of the most resistant and
531
adaptable group is a peculiarity of more stressed, less stable environments and characterized by
532
fine sediment fraction, such as harbors (Semprucci et al., 2015). Abundance and diversity indices
533
were correlated to different proxies of food sources (quantity and quality) into the sediment and
534
to its grain size, confirming the effect of these environmental variables on meiofaunal
535
populations (Balsamo et al., 2010). 536
The dissimilarity among the sampling stations was mainly due to some less abundant and more
537
sensitive taxa, not present in the innermost stations. Usually, organisms that can cope with
538
unfavorable conditions take over (e.g., Nematoda), whereas more sensitive taxa disappear or
539
become rare (Mirto et al., 2014; Zeppilli et al., 2015). In the case of Ancona harbor, Bivalvia,
540
Kinorhyncha, Platyhelminthes, and Tardigrada were found only at the outermost stations (LR,
541
DS, and API) being identified as less tolerant taxa (Baldrighi et al., 2019 and literature therein). 542
Conversely, the more tolerant and widespread groups of Ciliata, Oligochaeta, and Polychaeta
543
(Pusceddu et al., 2007; Moreno et al., 2008a; Moreno et al., 2008b; Semprucci et al., 2015)
544
characterized the innermost stations, along with a non-negligible presence (from 34% to 80%
545
among other taxa, Fig.4b) of soft-shelled Foraminifera inhabiting the sediment at all the
546
investigated sampling stations. Soft-shelled monothalamous Foraminifera are an important
547
component living in the sediment and populating the Adriatic Sea (Sabbatini et al., 2013), but
548
most of the time this component is overlooked in meiofauna studies. The high presence of this
549
group has been found to be associate to high values of Chl-a, eutrophic conditions, and high
550
variability of environmental parameters (e.g., organic matter loads, salinity, temperature, oxygen
551
content; Sabbatini et al., 2013). Manuscript to be reviewed Moreover, given
519
its small size, low mobility, and lack of dispersive life stages (Giere, 2009), meiofauna is more
520
susceptible to within-habitat physical variability and environmental disturbances than larger,
521
more mobile, and potentially more highly dispersed members of the macrofauna (Austen &
522
Widdicombe, 2006; Schratzberger et al., 2008). This would explain the drop in meiofauna
523
abundance, not reported for the macrofauna, at that sampling station. A total of 12 meiofauna
524
taxa (Orders or Classes) were identified and in four sampling stations out of five the majority of
525
taxa were represented. The measures of diversity (i.e., S, H’, and J’) were comparable to the
526
values reported in harbor areas (e.g., Moreno et al., 2008a; Moreno et al., 2008b; Sedano et al.,
527
2014; Kulakova, 2022) and in enclosed/transitional systems in the Adriatic Sea (e.g., Pusceddu et
528
al., 2007; Pusceddu et al., 2011; Frontalini et al., 2014; Covazzi-Harriague et al., 2012). 515
(Chl-a, Phaeo, and chloroplast pigment equivalents - CPE), its low abundance at PORT station
516
could be partially justified by the lowest detected value of ‘fresh’ material (Chl-a) and/or other
517
factors known to regulate the coexistence of different organisms such as competition for
518
resources and predation (Schratzberger et al., 2003; Schratzberger et al., 2008). Moreover, given
519
its small size, low mobility, and lack of dispersive life stages (Giere, 2009), meiofauna is more
520
susceptible to within-habitat physical variability and environmental disturbances than larger,
521
more mobile, and potentially more highly dispersed members of the macrofauna (Austen &
522
Widdicombe, 2006; Schratzberger et al., 2008). This would explain the drop in meiofauna
523
abundance, not reported for the macrofauna, at that sampling station. A total of 12 meiofauna
524
taxa (Orders or Classes) were identified and in four sampling stations out of five the majority of
525
taxa were represented. The measures of diversity (i.e., S, H’, and J’) were comparable to the
526
values reported in harbor areas (e.g., Moreno et al., 2008a; Moreno et al., 2008b; Sedano et al.,
527
2014; Kulakova, 2022) and in enclosed/transitional systems in the Adriatic Sea (e.g., Pusceddu et
528
al., 2007; Pusceddu et al., 2011; Frontalini et al., 2014; Covazzi-Harriague et al., 2012). 529
However, the strong dominance of Nematoda justified the low values reported for the Pielou’s
530
equitability index, particularly at LR and DS stations. 502
Meiofaunal response to harbor environmental conditions 504
In Ancona harbor, the meiofaunal total abundance, community structure and, for a lesser extent,
505
univariate measures (i.e., diversity and equitability indices) reflected the marked spatial
506
heterogeneity showed by the PCA and the clear separation both between inner and outer
507
sampling stations and among the sampling stations themselves (MAN vs. PORT vs. DS + LR vs. 508
API). Meiofaunal abundance was in the range of values reported by Baldrighi et al. (2019) for
509
Ancona harbor (i.e. from 912±404 to 2512±717 ind.10cm-2) with lower values characterizing the
510
outermost station and for other harbors and coastal areas affected by pollution and/or high
511
organic matter loads (Vezzulli et al., 2003; Veiga et al., 2009; Pusceddu et al., 2011; Dal Zotto
512
et al., 2016; Semprucci et al., 2016; Sedano et al., 2014; Kulakova, 2022). The only exception
513
was represented by the paucity of meiofaunal organisms found at PORT station. Considering that
514
total meiofaunal abundance was positively linked to products derived from primary production PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed The strong tolerance and positive response of some
552
monothalamous species to environmental stress (Sabbatini et al., 2010) may justify their
553
presence in the Ancona harbor. A further identification of the species characterizing these
554
sediments will be necessary, in any case, to confirm our hypothesis. Changes in the community 536
The dissimilarity among the sampling stations was mainly due to some less abundant and more
537
sensitive taxa, not present in the innermost stations. Usually, organisms that can cope with
538
unfavorable conditions take over (e.g., Nematoda), whereas more sensitive taxa disappear or
539
become rare (Mirto et al., 2014; Zeppilli et al., 2015). In the case of Ancona harbor, Bivalvia,
540
Ki
h
h
Pl t h l i th
d T di
d
f
d
l
t th
t
t t ti
(LR 536
The dissimilarity among the sampling stations was mainly due to some less abundant and more
537
sensitive taxa, not present in the innermost stations. Usually, organisms that can cope with
538
unfavorable conditions take over (e.g., Nematoda), whereas more sensitive taxa disappear or
539
become rare (Mirto et al., 2014; Zeppilli et al., 2015). In the case of Ancona harbor, Bivalvia, 538
unfavorable conditions take over (e.g., Nematoda), whereas more sensitive taxa disappear or
539
become rare (Mirto et al., 2014; Zeppilli et al., 2015). In the case of Ancona harbor, Bivalvia,
540
Kinorhyncha, Platyhelminthes, and Tardigrada were found only at the outermost stations (LR,
541
DS, and API) being identified as less tolerant taxa (Baldrighi et al., 2019 and literature therein). 542
Conversely, the more tolerant and widespread groups of Ciliata, Oligochaeta, and Polychaeta
543
(Pusceddu et al., 2007; Moreno et al., 2008a; Moreno et al., 2008b; Semprucci et al., 2015)
544
characterized the innermost stations, along with a non-negligible presence (from 34% to 80%
545
among other taxa, Fig.4b) of soft-shelled Foraminifera inhabiting the sediment at all the
546
investigated sampling stations. Soft-shelled monothalamous Foraminifera are an important
547
component living in the sediment and populating the Adriatic Sea (Sabbatini et al., 2013), but
548
most of the time this component is overlooked in meiofauna studies. The high presence of this
549
group has been found to be associate to high values of Chl-a, eutrophic conditions, and high
550
variability of environmental parameters (e.g., organic matter loads, salinity, temperature, oxygen
551
content; Sabbatini et al., 2013). Manuscript to be reviewed 555
structure were supported by DistLM analysis, which revealed that pollutants and, secondly, the
556
grain size could explain the variability in the meiofaunal composition. 555
structure were supported by DistLM analysis, which revealed that pollutants and, secondly, the
556
grain size could explain the variability in the meiofaunal composition. 555
structure were supported by DistLM analysis, which revealed that pollutants and, secondly, the
556
grain size could explain the variability in the meiofaunal composition. 557
Food sources did not have any effect on the meiobenthic community. According to Dell’Anno et
558
al. (2002) and Pusceddu et al. (2009), the system of Ancona harbor can be ranked as eutrophic
559
(inside) - mesotrophic (outside) with high quality of organic matter. Thus, food sources did not
560
constitute a limiting factor for the meiofaunal community (Muniz et al., 2005; Covazzi
561
Harriague et al., 2012; Dal Zotto et al., 2016), as reported instead for oligotrophic systems (e.g.,
562
Covazzi Harriague et al., 2013). Same results were reported in Franzo et al. (2022) analyzing the
563
nematode communities inhabiting different Adriatic harbors, including that of Ancona. Authors
564
showed that the main environmental factor that shaped the nematode assemblages in all harbors
565
were the PAH concentration levels, while food sources and the grain size were much less
566
relevant. Interestingly, some positive correlations between HMs and meiofaunal abundance and
567
its related univariate measures were reported in the present study, as elsewhere (Schratzberger et
568
al., 2003; Xu et al., 2014). In the study conducted by Cibic et al. (2017), authors pinpointed as
569
heavy metal content may influence meiofaunal abundance and its composition. The positive
570
nature of the correlation could be the result of a meiobenthic community well adapted to
571
permanent stress conditions (Cibic et al., 2017). Manuscript to be reviewed The strong tolerance and positive response of some
552
monothalamous species to environmental stress (Sabbatini et al., 2010) may justify their
553
presence in the Ancona harbor. A further identification of the species characterizing these
554
sediments will be necessary, in any case, to confirm our hypothesis. Changes in the community 542
Conversely, the more tolerant and widespread groups of Ciliata, Oligochaeta, and Polychaeta
543
(Pusceddu et al., 2007; Moreno et al., 2008a; Moreno et al., 2008b; Semprucci et al., 2015)
544
characterized the innermost stations, along with a non-negligible presence (from 34% to 80%
545
among other taxa, Fig.4b) of soft-shelled Foraminifera inhabiting the sediment at all the
546
investigated sampling stations. Soft-shelled monothalamous Foraminifera are an important
547
component living in the sediment and populating the Adriatic Sea (Sabbatini et al., 2013), but
548
most of the time this component is overlooked in meiofauna studies. The high presence of this
549
group has been found to be associate to high values of Chl-a, eutrophic conditions, and high
550
variability of environmental parameters (e.g., organic matter loads, salinity, temperature, oxygen
551
content; Sabbatini et al., 2013). The strong tolerance and positive response of some
552
monothalamous species to environmental stress (Sabbatini et al., 2010) may justify their
553
presence in the Ancona harbor. A further identification of the species characterizing these
554
sediments will be necessary, in any case, to confirm our hypothesis. Changes in the community PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 595
the difference between the innermost sampling stations and the outermost one (SIMPER
596
analysis, Table S9). Species composition may be affected by pollutant concentrations and high
597
levels of organic matter, through a decrease in diversity and abundance of sensitive species
598
(Callier et al., 2009). Most of the species found inhabiting the Ancona harbor sediments were
599
typically soft-bottom species and belonged to the ecological groups of disturbance-tolerant,
600
second- and first-order opportunistic species (Borja et al., 2000). Polychaeta, along with the
601
tubificid oligochaeta T. swirencoides, were the taxa most represented and diversified. Many
602
Polychaeta species have a high level of tolerance to adverse effects such as pollution and natural
603
perturbations (Borja et al., 2000), and for this reason they usually constitute the majority of
604
benthic organisms living in harbor systems (Guerra-García & García-Gómez, 2004b). Usually,
605
Polychaeta species richness and their diversity inside of harbor areas are low because of high
606
pollution levels and the lack of oxygen in the water column (Estacio et al., 1997; Dhainaut-
607
Courtois et al., 2000). The same trend of increasing Polychaeta diversity moving outside the
608
harbor area was also reported in the present study, with tolerant and opportunistic species such as
609
Capitella capitata, C. caputesocis, H. filiformis, Sternaspis scutata and Streblospio sp. (Borja et
610
al., 2000) particularly abundant in the innermost sampling stations. The species Prionospio
611
cirrifera, mainly recorded outside the harbor (API station), is traditionally identified as an
612
opportunistic spionid living in silty-clay sediments with high organic content (Borja et al., 2000;
613
Simonini et al., 2004). Spagnolo et al. (2011) reported the same finding and authors explained
614
this as a result of a scarce tolerance of P. cirrifera to copper, detected at a concentration 9 times
615
higher at the innermost MAN station compared to the average concentration detected at all the
616
other sampling stations. A similar consideration could arise for the high abundance of the
617
opportunistic species S. bombyx at API station. Polychaeta species ranked as disturbance-
618
sensitive (Borja et al., 2000), such as Aricidea fragilis, Glycera capitata, Paradoneis armata,
619
and Paraonis fulgens have also been found inhabiting the most impacted sampling stations inside
620
the harbor but in lower abundances compared to the opportunistic species. Manuscript to be reviewed This leads us to think
621
about a certain kind of adaptation of these species to cope with less favorable conditions, but this
622
aspect needs further investigation. Due to their economic and ecological importance, as well as
623
their sedentary life, Mollusca has assumed a major role in monitoring contaminants worldwide
624
(Pizzini et al., 2015; Pizzini et al., 2017; Grotti et al., 2016). Kurtiella bidentata, M. 625
galloprovincialis, and Nucula nitidosa are defined as disturbance-tolerant species and they
626
tended to dominate the innermost sampling stations in Ancona harbor, with the only exception
627
for K. bidentata, particularly abundant at API station. Abra alba characterized LR and DS
628
stations, confirming its preference for sandy sediments with medium-high levels of organic
629
matter quantity and quality (Guerra-García & García-Gómez, 2004b). This species has been
630
reported to abound in harbors affected by heavy metal pollution (Dhainaut-Courtois et al., 2000). 631
LR station was characterized by a conspicuous presence of the Nonindigenous species (NIS)
632
bivalve Anadara transversa bivalve of Indo Pacific origin (Streftaris & Zenetos 2006) 595
the difference between the innermost sampling stations and the outermost one (SIMPER
596
analysis, Table S9). Species composition may be affected by pollutant concentrations and high
597
levels of organic matter, through a decrease in diversity and abundance of sensitive species
598
(Callier et al., 2009). Most of the species found inhabiting the Ancona harbor sediments were
599
typically soft-bottom species and belonged to the ecological groups of disturbance-tolerant,
600
second- and first-order opportunistic species (Borja et al., 2000). Polychaeta, along with the
601
tubificid oligochaeta T. swirencoides, were the taxa most represented and diversified. Many
602
Polychaeta species have a high level of tolerance to adverse effects such as pollution and natural
603
perturbations (Borja et al., 2000), and for this reason they usually constitute the majority of
604
benthic organisms living in harbor systems (Guerra-García & García-Gómez, 2004b). Usually,
605
Polychaeta species richness and their diversity inside of harbor areas are low because of high
606
pollution levels and the lack of oxygen in the water column (Estacio et al., 1997; Dhainaut-
607
Courtois et al., 2000). The same trend of increasing Polychaeta diversity moving outside the
608
harbor area was also reported in the present study, with tolerant and opportunistic species such as
609
Capitella capitata, C. caputesocis, H. filiformis, Sternaspis scutata and Streblospio sp. 573
Macrofaunal response to harbor environmental conditions 575
In the present study, total macrofaunal abundance and its measures of diversity (i.e., S, H’, and
576
J’), fell within the range of values reported by Spagnolo et al. (2011) and Travizi et al. (2019) for
577
Ancona harbor and from other ports worldwide (e.g., Gusmao et al., 2016; Dauvin et al., 2017;
578
Li et al., 2017; Rebai et al., 2022). Results here reported showed an overall increasing trend in
579
macrofaunal abundance and species richness from inside to outside the study area, however, LR
580
station significantly differed from the other sampling stations when univariate measures were
581
considered. The low values of the Shannon’s diversity and of the Pielou's equitability indices
582
detected at LR station pointed to the presence of few dominant species such as Mytilus
583
galloprovincialis and Capitella capitata. Regarding the community structure, species
584
composition differed moving from inside to outside the harbor, overall reporting high
585
percentages of dissimilarity among stations (SIMPER analysis, Table S9). 586
The macrobenthic community was mostly composed by the dominant groups of Annelida,
587
Mollusca, and Crustacea, as usually reported from enclosed environments impacted by pollutants
588
and characterized by high organic matter loads (Guerra-García & García-Gómez, 2004b;
589
Spagnolo et al., 2011; Travizi et al., 2019). Among the group of Annelida, the Oligochaeta
590
species T. swirencoides was identified for the first time in Ancona harbor and it was found to be
591
particularly abundant in all the sampling stations and even dominant at LR station. Only at API
592
station the species was absent. Tubificid oligochaetes, also called sludge worms, are very
593
common in high polluted areas (Brusca & Brusca, 2003) and they are recognized as a pollution-
594
tolerant taxon (Pelletier et al., 2010). Thus, T. swirencoides was the species most responsible for PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 634
consistent level of hemoglobin in their bodies, allowing them to colonize habitats with low
635
oxygen concentrations (e.g., Zenetos, 1994; Morello et al., 2004). Manuscript to be reviewed 634
consistent level of hemoglobin in their bodies, allowing them to colonize habitats with low
635
oxygen concentrations (e.g., Zenetos, 1994; Morello et al., 2004). 636
A large number of crustaceans (Amphipoda, Isopoda, Tanaidacea) have been categorized as
637
pollution-sensitive taxa, especially compared to Polychaeta (Pelletier et al., 2010). Crustacean
638
communities have been considered to be among the most sensitive to changes in environmental
639
variables (Gómez-Gesteira & Dauvin, 2000), and for this reason crustacean species richness and
640
diversity inside harbors are generally considerably low (Estacio et al., 1997; Dhainaut-Courtois
641
et al., 2000). Three abundant species characterized the innermost sampling stations in Ancona
642
harbor: the Amphipoda Leptocheirus pilosus, the Caprellida Phtisica marina, and the Tanaidacea
643
Apseudopsis latreillii. P. marina and A. latreillii have been reported in high number in sediments
644
containing less sand and high concentrations of N, P, Cu, and organic matter (Guerra-García &
645
García-Gómez, 2004a). A. latreillii belongs to the group of species that may occur under normal
646
conditions, but whose populations are stimulated by organic enrichment, while P. marina
647
belongs to the group of species very sensitive to organic enrichment. According to the present
648
study and the results of other previous investigations (Conradi et al., 2000; Guerra-García &
649
García-Gómez, 2001; Guerra-García & García-Gómez, 2004a) this species is able to live even in
650
impacted habitats with moderate-high levels of heavy metals and PAHs. Conversely, the more
651
sensitive Amphipoda Ampelisca diadema dominated the crustacean assemblages at API station. 652
As for the meiofauna, pollutant content and the different sediment texture inside and outside the
653
Ancona harbor affected the macrofaunal composition, as previously reported (Guerra-García and
654
García-Gómez, 2001; Guerra-García & García-Gómez, 2004a; Guerra-García & García-Gómez,
655
2004b; Spagnolo et al., 2011; Travizi et al., 2019). Univariate descriptors as well as the analysis
656
of species characterizing the benthic communities indicated the presence of modified, but quite
657
diverse and presumably well-established soft-bottom communities in all the investigated
658
sampling stations. This might reflect the successful adaptation of many pollution-tolerant species
659
to the long-term pollution and unstable environmental conditions of Ancona harbor (Travizi et
660
al., 2019). 634
consistent level of hemoglobin in their bodies, allowing them to colonize habitats with low
635
oxygen concentrations (e.g., Zenetos, 1994; Morello et al., 2004). 661
Meiofauna and Macrofauna comparison 661
Meiofauna and Macrofauna comparison
662
663
Meio- and macrofauna can co-vary, even if the contrasting ecology of these two benthic
664
components can give us different information on the environment where they live in (Austen et
665
al., 1989; Schratzberger et al., 2003; Patrìcio et al., 2012). When compared together, both
666
taxonomical groups inhabiting Ancona harbor showed weak but significant relationship (Table
667
S10), as also evident from the nMDS (Figure 6). The weakness was likely due to the different
668
response of meiofauna to the PORT conditions, characterized by an extreme paucity in this
669
smaller benthic component. 670
However, with the only exception for PORT station, all the other stations followed the same
671
distributions for both benthic component, notwithstanding the different intra-replica variability. 672
This suggest that, even if the two faunal categories respond differently to the environmental and
673
pollutant variables (or at least with different intensity), both capture and highlight the 663
Meio- and macrofauna can co-vary, even if the contrasting ecology of these two benthic
664
components can give us different information on the environment where they live in (Austen et
665
al., 1989; Schratzberger et al., 2003; Patrìcio et al., 2012). When compared together, both
666
taxonomical groups inhabiting Ancona harbor showed weak but significant relationship (Table
667
S10), as also evident from the nMDS (Figure 6). The weakness was likely due to the different
668
response of meiofauna to the PORT conditions, characterized by an extreme paucity in this
669
smaller benthic component. 670
However, with the only exception for PORT station, all the other stations followed the same
671
distributions for both benthic component, notwithstanding the different intra-replica variability. 672
This suggest that, even if the two faunal categories respond differently to the environmental and
673
pollutant variables (or at least with different intensity), both capture and highlight the 663
Meio- and macrofauna can co-vary, even if the contrasting ecology of these two benthic
664
components can give us different information on the environment where they live in (Austen et
665
al., 1989; Schratzberger et al., 2003; Patrìcio et al., 2012). When compared together, both
666
taxonomical groups inhabiting Ancona harbor showed weak but significant relationship (Table
667
S10), as also evident from the nMDS (Figure 6). Manuscript to be reviewed (Borja et
610
al., 2000) particularly abundant in the innermost sampling stations. The species Prionospio
611
cirrifera, mainly recorded outside the harbor (API station), is traditionally identified as an
612
opportunistic spionid living in silty-clay sediments with high organic content (Borja et al., 2000;
613
Simonini et al., 2004). Spagnolo et al. (2011) reported the same finding and authors explained
614
this as a result of a scarce tolerance of P. cirrifera to copper, detected at a concentration 9 times
615
higher at the innermost MAN station compared to the average concentration detected at all the
616
other sampling stations. A similar consideration could arise for the high abundance of the
617
opportunistic species S. bombyx at API station. Polychaeta species ranked as disturbance-
618
sensitive (Borja et al., 2000), such as Aricidea fragilis, Glycera capitata, Paradoneis armata,
619
and Paraonis fulgens have also been found inhabiting the most impacted sampling stations inside
620
the harbor but in lower abundances compared to the opportunistic species. This leads us to think
621
about a certain kind of adaptation of these species to cope with less favorable conditions, but this
622
aspect needs further investigation. Due to their economic and ecological importance, as well as
623
their sedentary life, Mollusca has assumed a major role in monitoring contaminants worldwide
624
(Pizzini et al., 2015; Pizzini et al., 2017; Grotti et al., 2016). Kurtiella bidentata, M. 625
galloprovincialis, and Nucula nitidosa are defined as disturbance-tolerant species and they
626
tended to dominate the innermost sampling stations in Ancona harbor, with the only exception
627
for K. bidentata, particularly abundant at API station. Abra alba characterized LR and DS
628
stations, confirming its preference for sandy sediments with medium-high levels of organic
629
matter quantity and quality (Guerra-García & García-Gómez, 2004b). This species has been
630
reported to abound in harbors affected by heavy metal pollution (Dhainaut-Courtois et al., 2000). 631
LR station was characterized by a conspicuous presence of the Nonindigenous species (NIS)
632
bivalve Anadara transversa, bivalve of Indo-Pacific origin (Streftaris & Zenetos, 2006) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed The authors are grateful to the crews
703
of the boat Tecnopesca that was employed in sampling operations. 704
705
706
References
707
Anderson MJ. 2001. Permutation tests for univariate or multivariate analysis of variance and
708
regression. Canadian journal of fisheries and aquatic sciences 58(3): 626-639. 709
710
Anderson MJ, Robinson J. 2003. Generalized discriminant analysis based on distances. Australian & New
711
Zealand Journal of Statistics 45(3): 301-318. 712 684
In considering two benthic size components at the same time (meio- and macrofauna), we were
685
provided by a broader response to environmental conditions in the Ancona harbor. The following
686
considerations emerged: 687
-
The meiofauna was affected by the quality and quantity of organic matter, suggesting that
688
meiobenthic assemblages were more receptive to within-habitat food variability than
689
macrofauna. 690
-
Both invertebrate groups were characterized by distinctive assemblages across the harbor,
691
particularly evident for the meiofauna, consistent with changes detected for environmental
692
features. 693
-
The present investigation confirmed the fundamental advantage of a multi-benthic size
694
approach, with the inclusion of different taxonomical groups considered to cover a broader
695
range of functions into the ecosystem. Optimally, both groups should be used in marine
696
pollution monitoring programs included in the EU Marine Strategy Framework Directive
697
(MSFD; Directive 2008/56/EC) in the context of its Descriptor 1 ‘maintenance of
698
biodiversity’ and Descriptor 6 ‘sea floor integrity’. Manuscript to be reviewed 674
disturbances affecting the harbor area. This result, enforce the concept of cross-taxon congruence
675
(i.e., an interoperability among groups of organisms that respond as a unique community,
676
although both conserving their peculiarities) (Su et al., 2004). Similar congruence between
677
meiofauna and macrofauna groups was described by, e.g., Corte et al. (2017); Cronin-O’Reilly et
678
al. (2018). Notwithstanding, this result should be taken with caution as the relations between
679
different groups of organisms could be weak, as it is in the present case, to provide reliable
680
predictions of biodiversity in impacted areas (Heino, 2010), thus more data are needed to
681
confirm or definitely discard the usability of cross-taxon congruence. 682 676
although both conserving their peculiarities) (Su et al., 2004). Similar congruence between
677
meiofauna and macrofauna groups was described by, e.g., Corte et al. (2017); Cronin-O’Reilly e
678
al. (2018). Notwithstanding, this result should be taken with caution as the relations between
679
different groups of organisms could be weak, as it is in the present case, to provide reliable
680
predictions of biodiversity in impacted areas (Heino, 2010), thus more data are needed to
681
confirm or definitely discard the usability of cross-taxon congruence. 682
683
Conclusions
684
In considering two benthic size components at the same time (meio- and macrofauna), we were
685
provided by a broader response to environmental conditions in the Ancona harbor. The followin
686
considerations emerged:
687
-
The meiofauna was affected by the quality and quantity of organic matter, suggesting that
688
meiobenthic assemblages were more receptive to within-habitat food variability than
689
macrofauna. 690
-
Both invertebrate groups were characterized by distinctive assemblages across the harbor,
691
particularly evident for the meiofauna, consistent with changes detected for environmental
692
features. 693
-
The present investigation confirmed the fundamental advantage of a multi-benthic size
694
approach, with the inclusion of different taxonomical groups considered to cover a broader
695
range of functions into the ecosystem. Optimally, both groups should be used in marine
696
pollution monitoring programs included in the EU Marine Strategy Framework Directive
697
(MSFD; Directive 2008/56/EC) in the context of its Descriptor 1 ‘maintenance of
698
biodiversity’ and Descriptor 6 ‘sea floor integrity’. 699
700
Acknowledgements
701
The first author is very grateful to Dr. Jaques Grall and Vincent Le Garrec (UBO) for the long
702
time spent on the identification of macrofaunal organisms. 661
Meiofauna and Macrofauna comparison The weakness was likely due to the different
668
response of meiofauna to the PORT conditions, characterized by an extreme paucity in this
669
smaller benthic component. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) 700
Acknowledgements g
701
The first author is very grateful to Dr. Jaques Grall and Vincent Le Garrec (UBO) for the long
702
time spent on the identification of macrofaunal organisms. The authors are grateful to the crews
703
of the boat Tecnopesca that was employed in sampling operations. 704
705
706
References
707
Anderson MJ. 2001. Permutation tests for univariate or multivariate analysis of variance and
708
regression. Canadian journal of fisheries and aquatic sciences 58(3): 626-639. 709
710
Anderson MJ, Robinson J. 2003. Generalized discriminant analysis based on distances. Australian & New
711
Zealand Journal of Statistics 45(3): 301-318. 712 701
The first author is very grateful to Dr. Jaques Grall and Vincent Le Garrec (UBO) for the long
702
time spent on the identification of macrofaunal organisms. The authors are grateful to the crews
703
of the boat Tecnopesca that was employed in sampling operations. 710
Anderson MJ, Robinson J. 2003. Generalized discriminant analysis based on distances. Australian & New
711
Zealand Journal of Statistics 45(3): 301-318. 712 710
Anderson MJ, Robinson J. 2003. Generalized discriminant analysis based on distances. Australian & New
711
Zealand Journal of Statistics 45(3): 301-318. 712 PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 713
Anderson M, Gorley R, Clarke K. 2008. PERMANOVA+ for PRIMER: Guide to software and statistical
714
methods. Plymouth, UK: PRIMER-e. 715 716
Austen MC, Warwick RM, Rosado MC. 1989. Meiobenthic and macrobenthic community structure along
717
a putative pollution gradient in southern Portugal. Marine Pollution Bulletin 20: 398-405. 718 719
Austen MC, Widdicombe S. 2006. Comparison of the response of meio-and macrobenthos to disturbance
720
and organic enrichment. Journal of Experimental Marine Biology and Ecology 330: 96-104. 721 722
Baldrighi E, Semprucci F, Franzo A, Cvitkovic I, Bogner D, Despalatovic M, Berto D, Malgozata
723
Formalewicz M, Scarpato A, Frapiccini E, Marini M, Grego M. 2019. Meiofaunal communities in four
724
adriatic ports: baseline data for risk assessment in ballast water management. Marine Pollution Bulletin
725
147: 171–184. https://doi.org/10.1016/j.marpolbul.2018.06.056. 727
Balsamo M, Albertelli G, Ceccherelli VU, Coccioni R, Colangelo MA, Curini- Galletti M, Danovaro R,
728
D'Addabbo R, De Leonardis C, Fabiano M, Frontalini F, Gallo M, Gambi C, Guidi L, Moreno M,
729
Pusceddu A, Sandulli R, Semprucci F, Todaro MA, Tongiorgi P. 2010. Meiofauna of the Adriatic Sea:
730
current state of knowledge and future perspective. Chemistry and Ecology 26: 45–63. 731 728
D Addabbo R, De Leonardis C, Fabiano M, Frontalini F, Gallo M, Gambi C, Guidi L, Moreno M,
729
Pusceddu A, Sandulli R, Semprucci F, Todaro MA, Tongiorgi P. 2010. Meiofauna of the Adriatic Sea:
730
current state of knowledge and future perspective. Chemistry and Ecology 26: 45–63. 731 732
Balsamo M, Semprucci F, Frontalini F, Coccioni R. 2012. Meiofauna as a tool for marine ecosystem
733
biomonitoring. Marine Ecosystems 4: 77–104. 734 735
Barnett BE. 1983. Oligochaetes as indicators of pollution in the humber estuary, with special reference to
736
Tubificoides benedeni, Environmental Pollution Series A, Ecological and Biological, Volume 30, Issue 4,
737
Pages 277-291. https://doi.org/10.1016/0143-1471(83)90055-7. 738 735
Barnett BE. 1983. Oligochaetes as indicators of pollution in the humber estuary, with special reference to
736
Tubificoides benedeni, Environmental Pollution Series A, Ecological and Biological, Volume 30, Issue 4,
737
Pages 277-291. https://doi.org/10.1016/0143-1471(83)90055-7. g
p
y,
p
736
Tubificoides benedeni, Environmental Pollution Series A, Ecological and Biological, Volume 30, Issue 4,
737
P
277 291 htt
//d i
/10 1016/0143 1471(83)90055 7 739
Benedetti M, Gorbi S, Fattorini D, D'Errico G, Piva F, Pacitti D, Regoli F. 2014. Manuscript to be reviewed Environmental hazards
740
from natural hydrocarbons seepage: Integrated classification of risk from sediment chemistry, 741
bioavailability and biomarkers responses in sentinel species. Environmental Pollution 185: 116-126. DOI:
742
10.1016/j.envpol.2013.10.023. 743 744
Bianchelli S, Gambi C, Zeppilli D, Danovaro R. 2010. Metazoan meiofauna in deep-sea canyons and
745
adjacent open slopes: a large-scale comparison with focus on the rare taxa. Deep Sea Research Part I:
746
Oceanographic Research Papers 57: 420-433. 747 748
Bianchelli S, Pusceddu A, Buschi E, Danovaro R. 2016. Trophic status and meiofauna biodiversity in the
749
Northern Adriatic Sea: insights for the assessment of good environmental status. Marine Environmental
750
Research 113: 18-30. 751 752
Borja A, Franco J, Perez V. 2000. A marine biotic index to establish the ecological quality of soft-bottom
753
benthos within European estuarine and coastal environments. Marine Pollution Bulletin 40: 1100–1114. 754 755
Borja A, Muxika I, Franco J. 2003. The application of a marine biotic index to different impact sources
756
affecting soft-bottom benthic communities along European coasts. Marine Pollution Bulletin 46: 835–
757
845. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 81
Bernardi Aubry F, Boero F, Buia MC, Cabrini M, Camatti E, Cardone F, Cataletto B 781
Bernardi Aubry F, Boero F, Buia MC, Cabrini M, Camatti E, Cardone F, Cataletto B, Cattaneo Vietti R,
782
Cecere E Cibic T Colangelo P De Olazabal A D'onghia G Finotto S Fiore N Fornasaro D Fraschetti 781
Bernardi Aubry F, Boero F, Buia MC, Cabrini M, Camatti E, Cardone F, Cataletto B, Cattaneo Vietti R, 781
Bernardi Aubry F, Boero F, Buia MC, Cabrini M, Camatti E, Cardone F, Cataletto B, Cattaneo Vietti R,
782
Cecere E, Cibic T, Colangelo P, De Olazabal A, D'onghia G, Finotto S, Fiore N, Fornasaro D, Fraschetti
783
S, Gambi MC, Giangrande A, Gravili C, Guglielmo R, Longo C, Lorenti M, Luglie A, Maiorano P,
784
Mazzocchi MG, Mercurio M, Mastrototaro F, Mistri M, Monti M, Munari C, Musco L, Nonnis-Marzano
785
C, Padedda BM, Patti FP, Petrocelli A, Piraino S, Portacci G, Pugnetti A, Pulina S, Romagnoli T, Rosati
786
I, Sarno D, Satta CT, Sechi N, Schiaparelli S, Scipione B, Sion L, Terlizzi A, Tirelli V, Totti C, Tursi A, 82
Cecere E, Cibic T, Colangelo P, De Olazabal A, D'onghia G, Finotto S, Fiore N, Forn 783
S, Gambi MC, Giangrande A, Gravili C, Guglielmo R, Longo C, Lorenti M, Luglie A, Maiorano P, 784
Mazzocchi MG, Mercurio M, Mastrototaro F, Mistri M, Monti M, Munari C, Musco L, Nonnis-Marzano 85
C, Padedda BM, Patti FP, Petrocelli A, Piraino S, Portacci G, Pugnetti A, Pulina S, R 786
I, Sarno D, Satta CT, Sechi N, Schiaparelli S, Scipione B, Sion L, Terlizzi A, Tirelli V, Totti C, Tursi A, 786
I, Sarno D, Satta CT, Sechi N, Schiaparelli S, Scipione B, Sion L, Terlizzi A, Tirelli V, Totti C, Tursi A,
787
Ungaro N, Zingone A, Zupo V, Basset A. 2016. Ecosystem vulnerability to alien and invasive species: a
788
case study on marine habitats along the Italian coast Aquatic Conservation: Marine and Freshwater 786
I, Sarno D, Satta CT, Sechi N, Schiaparelli S, Scipione B, Sion L, Terlizzi A, Tirelli V, Totti C, Tursi A,
787
Ungaro N, Zingone A, Zupo V, Basset A. 2016. Ecosystem vulnerability to alien and invasive species: a 787
Ungaro N, Zingone A, Zupo V, Basset A. 2016. Ecosystem vulnerability to alien and invasive species: a
788
case study on marine habitats along the Italian coast. Manuscript to be reviewed 80
Corriero G, Pierri C, Accoroni S, Alabiso G, Bavestrello G, Barbone E, Bastianini M 780
Corriero G, Pierri C, Accoroni S, Alabiso G, Bavestrello G, Barbone E, Bastianini M, Bazzoni AM,
781
Bernardi Aubry F, Boero F, Buia MC, Cabrini M, Camatti E, Cardone F, Cataletto B, Cattaneo Vietti R,
782
Cecere E, Cibic T, Colangelo P, De Olazabal A, D'onghia G, Finotto S, Fiore N, Fornasaro D, Fraschetti
783
S, Gambi MC, Giangrande A, Gravili C, Guglielmo R, Longo C, Lorenti M, Luglie A, Maiorano P,
784
Mazzocchi MG, Mercurio M, Mastrototaro F, Mistri M, Monti M, Munari C, Musco L, Nonnis-Marzano
785
C, Padedda BM, Patti FP, Petrocelli A, Piraino S, Portacci G, Pugnetti A, Pulina S, Romagnoli T, Rosati
786
I, Sarno D, Satta CT, Sechi N, Schiaparelli S, Scipione B, Sion L, Terlizzi A, Tirelli V, Totti C, Tursi A,
787
Ungaro N, Zingone A, Zupo V, Basset A. 2016. Ecosystem vulnerability to alien and invasive species: a
788
case study on marine habitats along the Italian coast. Aquatic Conservation: Marine and Freshwater
789
Ecosystems 26: 392–409. Manuscript to be reviewed 759
Brusca RC, Brusca GJ. 2003. Invertebrates. Sinauer Associates, Sunderland, MA, 936 pp. 760 761
Callier MD, Fletcher RL, Thorp CH, Fichet D. 2009. Macrofaunal community responses to marina-
762
related pollution on the South coast of England and west coast of France. Journal of the Marine Biological
763
Association of the United Kingdom 89: 19-29. 764 765
Chatzinikolaou E, Mandalakis M, Damianidis P, Dailianis T, Gambineri S, Rossano C, Scapini F, Carucci
766
A, Arvanitidis C. 2018. Spatio-temporal benthic biodiversity patterns and pollution pressure in three
767
Mediterranean touristic ports. Science of the Total Environment 624: 648–660. 768
https://doi.org/10.1016/j.scitotenv.2017.12.111. 769 765
Chatzinikolaou E, Mandalakis M, Damianidis P, Dailianis T, Gambineri S, Rossano C, Scapini F, Carucci
766
A, Arvanitidis C. 2018. Spatio-temporal benthic biodiversity patterns and pollution pressure in three
767
M dit
t
i ti
t
S i
f th T t l E
i
t 624 648 660 p
766
A, Arvanitidis C. 2018. Spatio-temporal benthic biodiversity patterns and pollution pressure in three 69
770
Cibic T, Franzo A, Nasi F, Auriemma R, Del Negro P. 2017. The port of Trieste (northern Adriatic
771
Sea)— a case study of the “ecosystem approach to management”. Frontiers in Marine Science 4: 336. 772
https://doi.org/10.3389/fmars.2017.00336. 774
Clarke KR, Gorley RN. 2006. Primer. PRIMER-e, Plymouth, 866. 776
Conradi M, Lòpez-González PJ, Cervera JL, García-Gómez JC. 2000. Seasonality and spatial distribution
777
of peracarids associated with the bryozoan Bugula neritina in Algeciras Bay, Spain. Journal of Crustacean
778
Biology 20: 126-141. 779 776
Conradi M, Lòpez-González PJ, Cervera JL, García-Gómez JC. 2000. Seasonality and spatial distribution
777
of peracarids associated with the bryozoan Bugula neritina in Algeciras Bay, Spain. Journal of Crustacean
778
Biology 20: 126-141. Manuscript to be reviewed 802
Covazzi Harriague A, Albertelli G, Misic C. 2012. Macro-and meiofaunal community features in the
803
critical environmental system of a tourist harbour (Rapallo, Ligurian Sea, NW Mediterranean). Marine
804
environmental research, 74, 64-72. 805 802
Covazzi Harriague A, Albertelli G, Misic C. 2012. Macro-and meiofaunal community features in
803
critical environmental system of a tourist harbour (Rapallo, Ligurian Sea, NW Mediterranean). M
804
environmental research, 74, 64-72. 805
806
Covazzi Harriague A, Misic C, Valentini I, Polidori E, Albertelli G, Pusceddu A. 2013. Meio- and
807
macrofauna communities in three sandy beaches of the northern Adriatic Sea protected by artificia
808
Chemistry and Ecology 29:181-195. DOI: 10.1080/02757540.2012.704911. 809
810
Cronin-O’Reilly S, Taylor JD, Jermyn I, Allcock AL, Cunliffe M, Johnson MP. 2018. Limited
811
congruence exhibited across microbial, meiofaunal and macrofaunal benthic assemblages in a
812
heterogeneous coastal environment. Scientific Reports 8. DOI: 10.1038/s41598-018-33799-9. 813
814
Dal Zotto M, Santulli A, Simonini R, Todaro MA. 2016. Organic enrichment effects on a marine
815
meiofauna community, with focus on Kinorhyncha. Zoologischer Anzeiger 265: 127–140. 816
817
Danovaro R. 2010. Methods for the Study of Deep-sea Sediments, Their Functioning and Biodive
818
CRC Press Taylor & Francis Group. 819
820
Danulat E, Muniz P, Garcia-Alonso J, Yannicelli B. 2002. First assessment of the highly contamin
821
harbour of Montevideo, Uruguay. Marine pollution bulletin 44: 551-576. 822
823
Dauvin JC, Bakalem A, Baffreau A, Grimes S. 2017. Benthic ecological status of Algerian
824
harbours. Marine pollution bulletin 125: 378-388. 825
826
Dell'Anno A, Mei ML, Pusceddu A, Danovaro R. 2002. Assessing the trophic state and eutrophica
827
coastal marine systems: a new approach based on the biochemical composition of sediment organ
828
matter. Marine Pollution Bulletin 44 : 611-622. 829
830
Dhainaut-Courtois NC, Pruvot A, Baudet EK. 2000. Les peuplements macrozoobenthiques indica
831
des qualités physico-chimiques et chimiques des sédiments portuaires– Exemple du Port de Boulo
832
sur-mer (Manche). Bulletin de la Société Zoologique de France, 125: 49-62. 833 806
Covazzi Harriague A, Misic C, Valentini I, Polidori E, Albertelli G, Pusceddu A. 2013. Meio- and
807
macrofauna communities in three sandy beaches of the northern Adriatic Sea protected by artificial reefs,
808
Chemistry and Ecology 29:181-195. DOI: 10.1080/02757540.2012.704911. 809 810
Cronin-O’Reilly S, Taylor JD, Jermyn I, Allcock AL, Cunliffe M, Johnson MP. 2018. Limited
811
congruence exhibited across microbial, meiofaunal and macrofaunal benthic assemblages in a
812
heterogeneous coastal environment. Manuscript to be reviewed Aquatic Conservation: Marine and Freshwater
789
Ecosystems 26: 392–409. 791
Corte GN, Checon HH, Fonseca G, Vieira DC, Gallucci F, Domenico M Di, Amaral ACZ. 2017. Cross-
792
taxon congruence in benthic communities: Searching for surrogates in marine sediments. Ecological
793
Indicators 78:173–182. DOI: 10.1016/j.ecolind.2017.03.031. 94 795
Costa LDF, Mirlean N, Wasserman JC, Wallner-Kersanach M. 2016. Variability of labile metals in
796
estuarine sediments in areas under the influence of antifouling paints, southern Brazil. Environmental
797
Earth Sciences 75: 580. https://doi.org/10.1007/s12665-016-5355-5. 98 799
Covazzi Harriague A, Misic C, Petrillo M, Albertelli G. 2007. Stressors affecting the macrobenthic
800
community in Rapallo harbour (Ligurian Sea, Italy). Scientia Marina 71: 705-714. 801 PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed Scientific Reports 8. DOI: 10.1038/s41598-018-33799-9. 813 814
Dal Zotto M, Santulli A, Simonini R, Todaro MA. 2016. Organic enrichment effects on a marine
815
meiofauna community, with focus on Kinorhyncha. Zoologischer Anzeiger 265: 127–140. 816 817
Danovaro R. 2010. Methods for the Study of Deep-sea Sediments, Their Functioning and Biodiversity. 818
CRC Press Taylor & Francis Group. 819 8 9
820
Danulat E, Muniz P, Garcia-Alonso J, Yannicelli B. 2002. First assessment of the highly contaminated
821
harbour of Montevideo, Uruguay. Marine pollution bulletin 44: 551-576. 822 823
Dauvin JC, Bakalem A, Baffreau A, Grimes S. 2017. Benthic ecological status of Algerian
824
harbours. Marine pollution bulletin 125: 378-388. 825 826
Dell'Anno A, Mei ML, Pusceddu A, Danovaro R. 2002. Assessing the trophic state and eutrophication of
827
coastal marine systems: a new approach based on the biochemical composition of sediment organic
828
matter. Marine Pollution Bulletin 44 : 611-622. 829 830
Dhainaut-Courtois NC, Pruvot A, Baudet EK. 2000. Les peuplements macrozoobenthiques indicateurs
831
des qualités physico-chimiques et chimiques des sédiments portuaires– Exemple du Port de Boulogne-
832
sur-mer (Manche). Bulletin de la Société Zoologique de France, 125: 49-62. 833 834
EC, 2000. Directive of the European Parliament and of the Council 2000/60/EC establishing a
835
Framework for Community Action in the Field of Water Policy. Available at: https://eur-
836
lex.europa.eu/resource.html?uri=cellar:5c835afb-2ec6 -4577-bdf8-
837
756d3d694eeb.0004.02/DOC_1&format=PDF. 838 839
EEA, UNEP, 2006. Priority Issues in the Mediterranean Environment. European Environment Agency,
840
Copenhagen. https://www.eea.europa.eu/publications/eea_report_2006_4. 842
El Nemr A, Khaled A, El Sikaily A. 2006. Distribution and Statistical Analysis of Leachable and Total
843
Heavy Metals in the Sediments of the Suez Gulf. Environmental Monitoring and Assessment 118: 89-
844
112. https://doi.org/10.1007/s10661-006-0985-9. 845 842
El Nemr A, Khaled A, El Sikaily A. 2006. Distribution and Statistical Analysis of Leachable and Total
843
Heavy Metals in the Sediments of the Suez Gulf. Environmental Monitoring and Assessment 118: 89-
844
112. https://doi.org/10.1007/s10661-006-0985-9. 845 PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 846
Estacio FJ, García-Adiego EM, Fa DA, García-Gómez JC, Daza JL, Hortas F, Gómez-Ariza JL. 1997. 847
Ecological analysis in a polluted area of Algeciras Bay (southern Spain): external ‘versus’ internal
848
outfalls and environmental implications. Marine Pollution Bulletin 34: 780-793. 849 846
Estacio FJ, García-Adiego EM, Fa DA, García-Gómez JC, Daza JL, Hortas F, Gó 8 6
stac o
J, Ga c a
d ego
,
a
, Ga c a Gó e JC,
a a J ,
o tas
, Gó e
a J . 997. 847
Ecological analysis in a polluted area of Algeciras Bay (southern Spain): external ‘versus’ internal
848
outfalls and environmental implications. Marine Pollution Bulletin 34: 780-793. 849 850
Etiope G, Panieri G, Fattorini D, Regoli F, Vannoli P, Italiano F, Locritani M, Carmisciano CA. 2014. 851
Thermogenic hydrocarbon seep in shallow Adriatic Sea (Italy): Gas origin, sediment contamination and
852
benthic foraminifera. Marine and Petroleum Geology 57:283-293. DOI:
853
10.1016/j.marpetgeo.2014.06.006. 855
Foti A, Fenzi GA, Di Pippo F, Gravina MF, Magni P. 2014. Testing the saprobity hypothesis in a
856
Mediterranean lagoon: effects of confinement and organic enrichment on benthic communities. Marine
857
Environmental Research 99: 85-94. 858 859
Franzo A, Baldrighi E, Grassi E, Grego M, Balsamo M, Basili M, Semprucci F. 2022. Free-living
860
nematodes of Mediterranean ports: A mandatory contribution for their use in ecological quality
861
assessment. Marine Pollution Bulletin 180: 113814. https://doi.org/10.1016/j.marpolbul.2022.113814
862 863
Frontalini F, Semprucci F, Coccioni R, Balsamo M, Bittoni P, Covazzi-Harriague A. 2011. On the
864
quantitative distribution and community structure of the meio and macrofaunal communities in the coastal
865
area of the Central Adriatic Sea (Italy). Environmental Monitoring and Assessment 180: 325-344. 866 867
Frontalini F, Semprucci F, Armynot du Châtelet E, Francescangeli F, Margaritelli 867
Frontalini F, Semprucci F, Armynot du Châtelet E, Francescangeli F, Margaritelli G, Rettori R, Spagnoli
868
F, Balsamo M, Coccioni R. 2014. Biodiversity trends of the meiofauna and foraminifera assemblages of
869
Lake Varano (southern Italy). Proceedings of the Biological Society of Washington 127: 7–22. 870 867
Frontalini F, Semprucci F, Armynot du Châtelet E, Francescangeli F, Margaritelli G, Rettori R, Spagnoli
868
F, Balsamo M, Coccioni R. 2014. Manuscript to be reviewed 890
Grotti M, Pizzini S, Abelmoschi ML, Cozzi G, Piazza R, Soggia F. 2016. Retrospective biomonitoring of
891
chemical contamination in the marine coastal environment of Terra Nova Bay (Ross Sea, Antarctica) by
892
environmental specimen banking. Chemosphere 165: 418-426. 893 894
Guerra-García JM, García-Gómez JC. 2001. The spatial distribution of Caprellidea (Crustacea:
895
Amphipoda): a stress bioindicator in Ceuta (North Africa, Gibraltar area). Pubbl. Staz. zool. Napoli,
896
(Marine Ecology) 22: 357-367. 897 895
Amphipoda): a stress bioindicator in Ceuta (North Africa, Gibraltar area). Pubbl. Staz. zool. Napoli,
896
(Marine Ecology) 22: 357-367. 897 898
Guerra-García JM, García-Gómez JC. 2004a. Polychaete assemblages and sediment pollution in a
899
harbour with two opposing entrances. Helgoland Marine Research 58:183–191. DOI 10.1007/s10152-
900
004-0184-4
901 902
Guerra-García JM, García-Gomez JC 2004b. Soft bottom mollusc assemblages and pollution in a harbour
903
with two opposing entrances. Estuarine, Coastal and Shelf Marine Science 60: 273-283. 904 904
905
Gusmao JB, Brauko KM, Eriksson BK, Lana PC. 2016. Functional diversity of macrobenthic
906
assemblages decreases in response to sewage discharges. Ecological Indicators 66: 65-75. 907 908
Heino J. 2010. Are indicator groups and cross-taxon congruence useful for predicting biodiversity in
909
aquatic ecosystems? Ecological Indicators 10:112–117. doi: 10.1016/j.ecolind.2009.04.013. 910 911
Heip C, Vincx M, Vranken G. 1985. The ecology of marine nematodes. Oceanography and Marine
912
Biology - An Annual Review 23: 399–489. 913 9 3
914
Katsanevakis S, Zenetos A, Mačić V, Beqiraj S, Poursanidis D, Kashta L. 2011. Invading the Adriatic:
915
spatial patterns of marine alien species across the Ionian- Adriatic boundary. Aquatic Biology 13: 107–
916
118. 917 918
KulaKova I. 2022. Assessment of the ecologicAl stAte of the DAnube DeltA coAstAl AreA
919
(northwestern pArt of the blAck seA) bAseD on meiobenthos AnD nemAtoDe AssemblAges. Geo-Eco-
920
Marina 28: 21-39. 921 922
Li B., Li X., Bouma T. J., Soissons L. M., Cozzoli F., Wang Q., Zhou Z., Chen L. 2017. Analysis of
923
macrobenthic assemblages and ecological health of Yellow River Delta, China, using AMBI & M-AMBI
924
assessment method. Marine Pollution Bulletin, 119: 23-32. 925 926
Losi V, Ferrero TJ, Moreno M, Gaozza L, Rovere A, Firpo M, Marques JC, Albertelli G. 2013. The use
927
of nematodes in assessing ecological conditions in shallow waters surrounding a Mediterranean harbour
928
facility. Estuarine, Coastal and Shelf Science 130: 209-221. Manuscript to be reviewed Biodiversity trends of the meiofauna and foraminifera assemblages of
869
L k V
(
h
I l ) P
di
f h Bi l
i
l S
i
f W
hi
127 7 22 869
Lake Varano (southern Italy). Proceedings of the Biological Society of Washington 127: 7–22. 870 871
Giere O. 2009. Meiobenthology: The Microscopic Motile Fauna of Aquatic Sediments, Second ed. 872
Springer-Verlag (527 pp.). 874
Giere O, Schratzberger M. 2023. New Horizons in Meiobenthos Research Profiles, Patterns and
875
Potentials. Cham: Springer International Publishing (407 pp.). 876 877
Giuliani S, Romanelli M, Piazza R, Vecchiato M, Pizzini S, Tranchida G, D’Agostino F, Romano S,
878
Bellucci LG. 2019. When research meets NGOs: The GVC-UCODEP project in the p Giang Province
879
and q River (Northern Vietnam) and its feedback on national monitoring programs. Environmental
880
Science & Policy 101: 279-290. https://doi.org/10.1016/j.envsci.2019.09.004. 882
Gómez-Gesteira JL, Dauvin JC. 2000. Amphipods are good bioindicators of the impact
883
of oil spills on soft-bottom macrobenthic communities. Marine Pollution Bulletin 40: 1017-
884
1027. 885 886
Gray J. S., Aschan M., Carr M. R., Clarke K. R., Green R. H., Pearson T. H., Rosenberg R., Warwick
887
R. M. 1988. Analysis of community attributes of the benthic macrofauna of Frierfjord/Langesundfjord
888
and in a mesocosm experiment. Marine Ecology Progress Series, 46: 151-165
889 886
Gray J. S., Aschan M., Carr M. R., Clarke K. R., Green R. H., Pearson T. H., Rosenberg R., Warwick
887
R. M. 1988. Analysis of community attributes of the benthic macrofauna of Frierfjord/Langesundfjord
888
and in a mesocosm experiment. Marine Ecology Progress Series, 46: 151-165
889 PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 929 930
Losi V, Grassi E, Balsamo M, Rocchi M, Gaozza L, Semprucci F. 2021. Changes in taxonomic structure
931
and functional traits of nematodes as tools in the assessment of port impact. Estuarine, Coastal and Shelf
932
Science 260: 107524. 933 PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed Marine Strategy Framework Directive 2008/56/EC of the European Parliament and of the
975
Council of 17 June 2008 Establishing a Framework for Community Action in the Field of Marine
976
Environmental Policy. pp. 1–22. 977 941
Mehlhorn P, Viehberg F, Kirsten K, Newman B, Frenzel P, Gildeeva O, Green A, Hahn A, Haberzett T. 942
2021. Spatial distribution and consequences of contaminants in harbour sediments–A case study from
943
Richards Bay Harbour, South Africa. Marine Pollution Bulletin 172: 112764. 944 941
Mehlhorn P, Viehberg F, Kirsten K, Newman B, Frenzel P, Gildeeva O, Green A, Hahn A, Haberzett T. 942
2021. Spatial distribution and consequences of contaminants in harbour sediments–A case study from
943
Ri h d B
H b
S
th Af i
M i
P ll ti
B ll ti 172 112764 945
Mestres M, Sierra JP, Mösso C, Sánchez-Arcilla A. 2010. Sources of contamination and modelled
946
pollutant trajectories in a Mediterranean harbour (Tarragona, Spain). Marine pollution bulletin 60: 898-
947
907. 948 945
Mestres M, Sierra JP, Mösso C, Sánchez-Arcilla A. 2010. Sources of contamination and modelled
946
pollutant trajectories in a Mediterranean harbour (Tarragona, Spain). Marine pollution bulletin 60: 898-
947
907. 948 953
954
Mirto S, Danovaro R. 2004. Meiofaunal colonisation on artificial substrates: a tool for
955
biomonitoring the environmental quality on coastal marine systems. Marine Pollution Bulletin 48: 919–
956
926. 957
958
Mirto S, Arigò C, Genovese L, Pusceddu A, Gambi C, Danovaro R. 2014. Nematode assemblage
959
response to fish-farm impact in vegetated (Posidonia oceanica) and non vegetated habitats. Aquaculture
960
Environment Interaction 5: 17–28. 961 954
Mirto S, Danovaro R. 2004. Meiofaunal colonisation on artificial substrates: a tool for
955
biomonitoring the environmental quality on coastal marine systems. Marine Pollution Bulletin 48: 919–
956
926. 957 958
Mirto S, Arigò C, Genovese L, Pusceddu A, Gambi C, Danovaro R. 2014. Nematode assemblage
959
response to fish-farm impact in vegetated (Posidonia oceanica) and non vegetated habitats. Aquaculture
960
Environment Interaction 5: 17–28. 961 958
Mirto S, Arigò C, Genovese L, Pusceddu A, Gambi C, Danovaro R. 2014. Nematode assemblage
959
response to fish-farm impact in vegetated (Posidonia oceanica) and non vegetated habitats. Aquaculture
960
Environment Interaction 5: 17–28. 961 962
Morello E, Solustri C, Antolini B, Froglia C. 2004. Manuscript to be reviewed 934
Luna GM, Dell’Anno A, Danovaro R. 2006. DNA extraction procedure: a critical issue for bacterial
935
diversity assessment in marine sediments. Environmental Microbiology 8 : 308-320. 936 937
McCready S, Birch GF, Long ER. 2006. Metallic and organic contaminants in sediments of Sydney
938
Harbour, Australia and vicinity—a chemical dataset for evaluating sediment quality
939
guidelines. Environment International 32: 455-465. 940 938
Harbour, Australia and vicinity—a chemical dataset for evaluating sediment quality
939
guidelines. Environment International 32: 455-465. 940
941
Mehlhorn P, Viehberg F, Kirsten K, Newman B, Frenzel P, Gildeeva O, Green A, Hahn A, Haberzett T. 942
2021. Spatial distribution and consequences of contaminants in harbour sediments–A case study from
943
Richards Bay Harbour, South Africa. Marine Pollution Bulletin 172: 112764. 944
945
Mestres M, Sierra JP, Mösso C, Sánchez-Arcilla A. 2010. Sources of contamination and modelled
946
pollutant trajectories in a Mediterranean harbour (Tarragona, Spain). Marine pollution bulletin 60: 898-
947
907. 948
949
Ministerial Decree of the Environment and Protection of the Territory and the Sea of 21
950
September 2016 n.173 “Regolamento recante modalità e criteri tecnici per l'autorizzazione
951
all'immersione in mare dei materiali di escavo di fondali marini” and related technical attachments. 952
https://www.gazzettaufficiale.it/eli/id/2016/09/06/16G00184/sg
953
954
Mirto S, Danovaro R. 2004. Meiofaunal colonisation on artificial substrates: a tool for
955
biomonitoring the environmental quality on coastal marine systems. Marine Pollution Bulletin 48: 919–
956
926. 957
958
Mirto S, Arigò C, Genovese L, Pusceddu A, Gambi C, Danovaro R. 2014. Nematode assemblage
959
response to fish-farm impact in vegetated (Posidonia oceanica) and non vegetated habitats. Aquaculture
960
Environment Interaction 5: 17–28. 961
962
Morello E, Solustri C, Antolini B, Froglia C. 2004. On the distribution of the allochthonous bivalves
963
Anadara inaequivalvis (Bruguière, 1789), Anadara demiri (Piani, 1981) and Musculista senhousia
964
(Benson in Cantor, 1842) in the Adriatic Sea, Italy. Biogeographia–The Journal of Integrative
965
Biogeography 25(1). 966
967
Moreno M, Ferrero TJ, Gallizia I, Vezzulli L, Albertelli G, Fabiano M. 2008a. An assessment of the
968
spatial heterogeneity of environmental disturbance within an enclosed harbour through the analysis of
969
meiofauna and nematode assemblages. Estuarine, Coastal and Shelf Science 77: 565-576. 970
971
Moreno M, Vezzulli L, Marin V, Laconi P, Albertelli G, Fabiano M. 2008b. The use of meiofauna
972
diversity as an indicator of pollution in harbours. ICES Journal of Marine Science 65: 1428–1435. 973
974
MSFD, 2008. Manuscript to be reviewed On the distribution of the allochthonous bivalves
963
Anadara inaequivalvis (Bruguière, 1789), Anadara demiri (Piani, 1981) and Musculista senhousia
964
(Benson in Cantor, 1842) in the Adriatic Sea, Italy. Biogeographia–The Journal of Integrative
965
Biogeography 25(1). 966 967
Moreno M, Ferrero TJ, Gallizia I, Vezzulli L, Albertelli G, Fabiano M. 2008a. An assessment of the
968
spatial heterogeneity of environmental disturbance within an enclosed harbour through the analysis of
969
meiofauna and nematode assemblages. Estuarine, Coastal and Shelf Science 77: 565-576. 970 971
Moreno M, Vezzulli L, Marin V, Laconi P, Albertelli G, Fabiano M. 2008b. The use of meiofauna
972
diversity as an indicator of pollution in harbours. ICES Journal of Marine Science 65: 1428–1435. 973 974
MSFD, 2008. Marine Strategy Framework Directive 2008/56/EC of the European Parliament and of the
975
Council of 17 June 2008 Establishing a Framework for Community Action in the Field of Marine
976
Environmental Policy. pp. 1–22. 977 974
MSFD, 2008. Marine Strategy Framework Directive 2008/56/EC of the European Parliament and of the
975
Council of 17 June 2008 Establishing a Framework for Community Action in the Field of Marine
976
E
i
t l P li
1 22 PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 978
Muniz P, Pires-Vanin AM. 2005. More about taxonomic sufficiency: a case study 978
Muniz P, Pires-Vanin AM. 2005. More about taxonomic sufficiency: a case study using polychaete
979
communities in a subtropical bay moderately affected by urban sewage. Ocean Science Journal 40: 17-33. 980 979
communities in a subtropical bay moderately affected by urban sewage. Ocean Science Journal 40: 17-33. 980 981
Papageorgiou N, Kalantzi I, Karakassis I. 2010. Effects of fish farming on the biological and geochemical
982
properties of muddy and sandy sediments in the Mediterranean Sea. Marine environmental research 69:
983
326-336. 984 985
Patrício J, Adão H, Neto JM, Alves AS, Traunspurger W, Marques JC. 2012. Do nematode and
986
macrofauna assemblages provide similar ecological assessment information?. Ecological Indicators 14:
987
124-137. 988 989
Paul JF, Scott KJ, Campbell DE, Gentile JH, Strobel CS, Valente RM, Weisberg S 989
Paul JF, Scott KJ, Campbell DE, Gentile JH, Strobel CS, Valente RM, Weisberg SB, Holland AF,
990
Ranasinghe JA. 2001. Developing and applying a benthic index of estuarine condition for the Virginian
991
Biogeographic Province. Ecological Indicators 1: 83-99. 990
Ranasinghe JA. 2001. Developing and applying a benthic index of estuarine condition for the Virginian
991
Biogeographic Province. Ecological Indicators 1: 83-99. 992 990
Ranasinghe JA. 2001. Developing and applying a benthic index of estuarine condi 990
Ranasinghe JA. 2001. Developing and applying a benthic index of estuarine condition for the Virginian
991
Biogeographic Province Ecological Indicators 1: 83 99 993
Pećarević M, Mikuš J, Bratoš-Cetinić A, Dulčić J, Čalić M. 2013. Introduced marine species in Croatian
994
waters (Eastern Adriatic Sea). Mediterranean Marine Science 14: 224–237. 995 996
Pelletier MC, Gold AJ, Heltshe JF, Buffum HW. 2010. A method to identify estuarine macroinvertebrate
997
pollution indicator species in the Virginian Biogeographic Province. Ecological Indicators 10: 1037-1048. 998 999
Pereira TJ, Gingold R, Villegas ADM, Rocha-Olivares A. 2018. Patterns of spatial variation of
1000
meiofauna in sandy beaches of northwestern Mexico with contrasting levels of disturbance. Thalassas: An
1001
International Journal of Marine Science 34: 53-63. 1002 1003
Pereira TL, Wallner-Kersanach M, Costa LDF, Costa DP, Baisch PRM. 2018. Nickel, vanadium, and lead
1004
as indicators of sediment contamination of marina, refinery, and shipyard areas. Environmental Science
1005
and Pollution Research 25: 1719-1730. https://doi.org/10.1007/s11356-017-0503-3. 1006 Pielou EC. 1969. An Introduction to Mathematical Ecology. Wiley-Interscience, New York 1007
Pielou EC. 1969. An Introduction to Mathematical Ecology. Manuscript to be reviewed 1020
comparison between Antarctic and Mediterranean bivalves. Chemosphere 175: 28 35. 1021
DOI.10.1016/j.chemosphere.2017.02.023 1020
comparison between Antarctic and Mediterranean bivalves. Chemosphere 175: 28 1022
Pizzini S, Morabito E, Gregoris E, Vecchiato M, Corami F, Piazza R, Gambaro A. 2021. Occurrence and
1023
source apportionment of organic pollutants in deep sediment cores of the Venice Lagoon. Marine
1024
Pollution Bulletin 164: 112053. https://doi.org/10.1016/j.marpolbul.2021.112053. 102 1026
Punzo E, Gomiero A, Tassetti AN, Strafella P, Santelli A, Salvalaggio V, Spagnolo A, Scarcella G, De
1027
Biasi AM, Kozinkova L, Fabi G. 2017. Environmental impact of offshore gas activities on the benthic
1028
environment: a case study. Environmental management 60: 340-356. 1029 1030
Pusceddu A, Gambi C, Manini E, Danovaro R. 2007. Trophic state, ecosystem efficiency and biodiversity
1031
of transitional aquatic ecosystems: analysis of environmental quality based on different benthic indicators. 1032
Chemical Ecology 23: 505–515. 1034
Pusceddu A, Dell’Anno A, Fabiano M, Danovaro R. 2009. Quantity and bioavailability of sediment
1035
organic matter as signatures of benthic trophic status. Marine Ecology Progress Series 375: 41-52. 1036 1037
Pusceddu A, Bianchelli S, Sanchez Vidal A, Canals M, Durrieu De Madron X, Heussner S, Lykousis V,
1038
de Stigter H, Trincardi F, Danovaro R. 2010. Organic matter in sediments of canyons and open slopes of
1039
the Portuguese, Catalan, Southern Adriatic and Cretan Sea margins. Deep Sea Research. I 57: 441-457. 1040 1041
Pusceddu A, Bianchelli S, Gambi C, Danovaro R. 2011. Assessment of benthic trophic status of marine
1042
coastal ecosystems: significance of meiofaunal rare taxa. Estuarine, Coastal and Shelf Science 93: 420-
1043
430. 1044 1044
1045
Quinn G, Keough M. 2002. Experimental design and data analysis for biologists. Cambridge University
1046
Press. 1047 1048
R Core Team 2018. R: A language and environment for statistical computing. R Foundation for
1049
Statistical Computing. doi:3-900051-14-3. 1050 1050
1051
Rebai N, Mosbahi N, Dauvin JC, Neifar L. 2022. Ecological Risk Assessment of Heavy Metals and
1052
Environmental Quality of Tunisian Harbours. Journal of Marine Science and Engineering 10: 1625. 1053 1054
Ridall A, Ingels J. 2021. Suitability of free-living marine nematodes as bioindicators: Status and future
1055
considerations. Frontiers in Marine Science 8: 685327. 1056 1057
Sabbatini A, Bonatto S, Gooday AJ, Morigi C, Pancotti I, Pucci F, Negri A. 2010. Modern benthic
1058
foraminifers at Northern shallow sites of Adriatic Sea and soft-walled, monothalamous taxa: a brief
1059
overview. Micropaleontology 359-376. Manuscript to be reviewed Wiley-Interscience, New York (286 pp.). 1008 1009
Pearson TH, Rosenberg R. 1978. Macrobenthic succession in relation to organic enrichment and pollution
1010
of the marine environment. Oceanography and Marine Biology 16: 229-311. 1011 1012
Pielou E. 1966. Shannon’s formula as a measure of specific diversity: its use and misuse. Am. Nat. 118,
1013
463–465. https://doi.org/10.1086/282439. 1014 1015
Pizzini S, Marchiori E, Piazza R, Cozzi G, Barbante C. 2015. Determination by HRGC/HRMS of PBDE
1016
levels in edible Mediterranean bivalves collected from north-western Adriatic coasts. Microchemical
1017
Journal 121:184-191. DOI.10.1016/j.microc.2015.03.010 1018
Pizzini S, Sbicego C, Corami F, Grotti M, Magi E, Bonato T, Cozzi G, Barbante C, Piazza R. 2017. 3, #u! 1019
dichlorobiphenyl (non-Aroclor PCB-11) as a marker of non-legacy PCB contamination in marine species: PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 1064
Schratzberger M, Daniel F, Wall CM, Kilbride R, Macnaughton SJ, Boyd SE, Rees HL, Lee K, Swannell
1065
RPJ. 2003. Response of estuarine meio- and macrofauna to in situ bioremediation of oil-contaminated
1066
sediment. Marine Pollution Bulletin 46: 430–443. 1067 1064
Schratzberger M, Daniel F, Wall CM, Kilbride R, Macnaughton SJ, Boyd SE, Rees HL, Lee K, Swannell
1065
RPJ. 2003. Response of estuarine meio- and macrofauna to in situ bioremediation of oil-contaminated 1068
Schratzberger M, Maxwell TAD, Warr K, Ellis JR, Rogers SI. 2008. Spatial variability of infaunal
1069
nematode and polychaete assemblages in two muddy subtidal habitats. Marine Biology 153: 621-642. 1070 1071
Schratzberger M, Ingels J. 2018. Meiofauna matters: the roles of meiofauna in benthic
1072
ecosystems. Journal of Experimental Marine Biology and Ecology 502: 12-25. 1073 1074
Sedano Vera F, Marquina D, Espinosa Torre F. 2014. Usefulness of meiofauna at high taxonomic levels
1075
as a tool to assess harbor quality status. Marina del Este Harbor (Granada, Spain) as a case study. Revista
1076
ciencias marinas y costeras 6: 103-113. 1077 1078
Semprucci F, Frontalini F, Sbrocca C, Armynot du Châtelet E, Bout-Roumazeilles V, Coccioni R,
1079
Balsamo M. 2015. Meiobenthos and free-living nematodes as tools for biomonitoring environments
1080
affected by riverine impact. Environmental Monitoring and Assessment 187, 251. 1081 1082
Semprucci F, Balsamo M, Sandulli R. 2016. Assessment of the Ecological quality (EcoQ) of the Venice
1083
lagoon using the structure and biodiversity of the meiofaunal assemblages. Ecological Indicators 67: 451–
1084
457. 1085 1086
Semprucci F, Grassi E, Balsamo M. 2022. Simple Is the Best: An Alternative Method for the Analysis of
1087
Free-Living Nematode Assemblage Structure. Water 14: 1114. https://doi.org/10.3390/w14071114
1088 1089
Shannon C, Weaver W. 1949. The Mathematical Theory of Communication. University of Illinois Press,
1090
Urbana, US. 1091 1092
Simboura N, Zenetos A. 2002. Benthic indicators to use in ecological quality classification of
1093
Mediterranean soft bottom marine ecosystems, including a new biotic index. Mediterranean Marine
1094
Science 3: 77–111. 1095 1096
Simonini R, Ansaloni I, Bonvicini Pagliai AM, Prevedelli D. 2004. Organic enrichment and structure of
1097
the macrozoobenthic community in the northern Adriatic Sea in an area facing Adige and Po mouths. 1098
ICES Journal of Marine Science 61: 871-881. 1099 1100
Simonini R, Ansaloni I, Cavallini F, Graziosi F, Iotti M, N’siala GM, Mauri M, Montanari G, Preti M,
1101
Prevedelli D. 2005. Manuscript to be reviewed 1060 1061
Sabbatini A, Nardelli MP, Morigi C, Negri A. 2013. Contribution of soft-shelled monothalamous taxa to
1062
foraminiferal assemblages in the Adriatic Sea. Acta Protozoologica 52(3). PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 1107
1108
Spagnolo A, Scarcella G, Sarappa A. 2011. Benthic community response to sediment features in Ancona
1109
harbour (Northern Adriatic Sea, Italy). Vie Milieu 61: 119–128. 1110 1111
Spagnolo A, Cuicchi C, De Biasi A M, Ferrà C, Montagnini L, Punzo E, Salvalaggio V, Santelli A,
1112
Strafella P, Fabi G. 2019. Effects of the installation of offshore pipelines on macrozoobenthic
1113
communities (northern and central Adriatic Sea). Marine pollution bulletin 138: 534-544. 1114 1115
Streftaris N, Zenetos A. 2006. Alien marine species in the Mediterranean-the 100 ‘Worst Invasives’ and
1116
their impact. Mediterranean Marine Science 7: 87-118. 111 1118
Su CJ, Debinski DM, Jakubauskas ME, Kindscher K. 2004. Beyond Species Richness: Community
1119
Similarity as a Measure of Cross-Taxon Congruence for Coarse-Filter Conservation. Conservation
1120
Biology 18:167–173. DOI: 10.1111/j.1523-1739.2004.00337.x. 1121 1122
Travizi A, ; M I, Bacci T, Bertasi F, Cuicchi C, Flander-Putrle V, Grati F, Grossi L, Jaklin A, Lipej
1123
L,
v B, Mikac B, Marusso V, Montagnini L, M V, Penna P, Salvalaggio V, Santelli A,
1124
Scirocco T, Spagnolo A, Trabucco B, Vani D. 2019. Macrozoobenthos in the Adriatic Sea ports: Soft-
1125
bottom communities with an overview of non-indigenous species. Marine Pollution Bulletin 147:159–
1126
170. https://doi.org/10.1016/j.marpolbul.2019.01.016
1127 1122
Travizi A, ; M I, Bacci T, Bertasi F, Cuicchi C, Flander-Putrle V, Grati F, Grossi L, Jaklin A, Lipej
1123
L,
v B, Mikac B, Marusso V, Montagnini L, M V, Penna P, Salvalaggio V, Santelli A,
1124
Scirocco T, Spagnolo A, Trabucco B, Vani D. 2019. Macrozoobenthos in the Adriatic Sea ports: Soft-
1125
bottom communities with an overview of non-indigenous species Marine Pollution Bulletin 147:159– 1122
Travizi A, ; M I, Bacci T, Bertasi F, Cuicchi C, Flander-Putrle V, Grati F, Grossi L, Jaklin A, Lipej 1122
Travizi A, ; M I, Bacci T, Bertasi F, Cuicchi C, Flander-Putrle V, Grati F, G 1123
L,
v B, Mikac B, Marusso V, Montagnini L, M V, Penna P, Salvalagg 1123
L,
v B, Mikac B, Marusso V, Montagnini L, M V, Penna P, Salvalaggio V, Santelli A,
1124
Scirocco T, Spagnolo A, Trabucco B, Vani D. 2019. Macrozoobenthos in the Adriatic Sea ports: Soft-
1125
bottom communities with an overview of non-indigenous species. Marine Pollution Bulletin 147:159–
1126
170. https://doi.org/10.1016/j.marpolbul.2019.01.016
1127 1128
Tobiszewski M, b; J. 2012. Manuscript to be reviewed Effects of long-term dumping of harbor-dredged material on macrozoobenthos at four
1102
disposal sites along the Emilia-Romagna coast (Northern Adriatic Sea, Italy). Marine Pollution
1103
Bulletin 50: 1595-1605. https://doi.org/10.1016/j.marpolbul.2005.06.031 1105
Somerfield PJ, Cochrane SJ, Dahle S, Pearson TH. 2006. Free-living nematodes and macrobenthos in a
1106
high-latitude glacial fjord. Journal of Experimental Marine Biology and Ecology 330: 284-296. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed PAH diagnostic ratios for the identification of pollution emission
1129
sources. Environmental Pollution 162 : 110-119. https://doi.org/10.1016/j.envpol.2011.10.025. 1130 1131
Todorova VR, Doncheva VG, Trifonova EV. 2020. First Implementation of Marine Strategy Framework
1132
Directive for Benthic Habitats Assessment in the Bulgarian Black Sea. Ecologia Balkanica. 1133 1134
Vassallo P, Fabiano M, Vezzulli L, Sandulli R, Marques JC, Jørgensen SE. 2006. Assessing the health of
1135
coastal marine ecosystems: a holistic approach based on sediment micro and meio-benthic
1136
measures. Ecological indicators 6: 525-542. 1135
coastal marine ecosystems: a holistic approach based on sediment micro and meio-benthic
1136
measures. Ecological indicators 6: 525-542. 1137
1138
Veiga P, Rubal M, Besteiro C. 2009. Shallow sublittoral meiofauna communities and sediment polycycli
1139
aromatic hydrocarbons (PAHs) content on the Galician coast (NW Spain), six months after the prestige
1140
oil spill. Marine Pollution Bulletin 58: 581–588. https://doi.org/10.1016/j.marpolbul.2008.11.002. 1141
1142
Vezzulli L, Marrale D, Moreno M, Fabiano M. 2003. Sediment organic matter and meiofauna community
1143
response to long-term fish-farm impact in the Ligurian Sea (Western Mediterranean). Chemistry and
1144
Ecology 19: 431-440. 1145
1146
Viana M, Hammingh P, Colette A, Querol X, Degraeuwe B, de Vlieger I, van Aardenne J. 2014. Impact
1147
of maritime transport emissions on coastal air quality in Europe. Atmospheric Environment 90: 96-105. 1148
https://dx.doi.org/10.1016/j.atmosenv.2014.03.046. 1149 1138
Veiga P, Rubal M, Besteiro C. 2009. Shallow sublittoral meiofauna communities and sediment polycyclic
1139
aromatic hydrocarbons (PAHs) content on the Galician coast (NW Spain), six months after the prestige
1140
oil spill. Marine Pollution Bulletin 58: 581–588. https://doi.org/10.1016/j.marpolbul.2008.11.002. 1141 1140
oil spill. Marine Pollution Bulletin 58: 581–588. https://doi.org/10.1016/j.marpolbul.2008.11.002. 1141
1142
Vezzulli L, Marrale D, Moreno M, Fabiano M. 2003. Sediment organic matter and meiofauna community
1143
response to long-term fish-farm impact in the Ligurian Sea (Western Mediterranean). Chemistry and
1144
Ecology 19: 431-440. 1145
1146
Viana M, Hammingh P, Colette A, Querol X, Degraeuwe B, de Vlieger I, van Aardenne J. 2014. Impact
1147
of maritime transport emissions on coastal air quality in Europe. Atmospheric Environment 90: 96-105. 1148
https://dx.doi.org/10.1016/j.atmosenv.2014.03.046. 1149 1142
Vezzulli L, Marrale D, Moreno M, Fabiano M. 2003. Sediment organic matter and meiofauna community
1143
response to long-term fish-farm impact in the Ligurian Sea (Western Mediterranean). Chemistry and
1144
Ecology 19: 431-440. 1145 PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 1150
Warwick RM, Dashfield SL, Somerfield PJ. 2006. The integral structure of a benthic infaunal
1151
assemblage. Journal of Experimental Marine Biology and Ecology 330: 12-18. 1152 1153
Weisberg SB, Ranasinghe JA, Dauer DM, Schaffner LC, Diaz RJ, Frithsen JB. 1997. An estuarine
1154
benthic index of biotic integrity (B-IBI) for Chesapeake Bay. Estuaries 20: 149-158. 1155 1156
Whomersley P, Huxham M, Schratzberger M, Bolam S. 2009. Differential response of meio-and
1157
macrofauna to in situ burial. Journal of the Marine Biological Association of the United Kingdom 89:
1158
1091-1098. 1160
WoRMS Editorial Board, 2022. World Register of Marine Species. Available from
1161
https://www.marinespecies.org at VLIZ. Accessed 2022. doi:10.14284/170
1162 1163
Zenetos A. 1994. Scapharca demiri (Piani, 1981): Primo ritrovamento nel nord Egeo. La Conchiglia 271:
1164
37-38. 1165 1166
Zeppilli D, Sarrazin J, Leduc D, Arbizu PM, Fontaneto D, Fontanier C, Gooday A 1166
Zeppilli D, Sarrazin J, Leduc D, Arbizu PM, Fontaneto D, Fontanier C, Gooday AJ, Kristensen RM,
1167
Ivanenko VN, Sorensen MV, Vanreusel A, Thébault J, Mea M, Allio N, Andro T, Arvigo A, Castrec J,
1168
Danielo M, Foulon V, Fumeron R, Hermabessiere L, Hulot V, James T, Langonne-Augen R, Le Bot T,
1169
Long M, Mahabror D, Morel Q, Pantalos M, Pouplard E, Raimondeau L, Rio- Cabello A, Seite S,
1170
Traisnel G, Urvoy K, Van Der Stegen T, Weyand M, Fernandes D. 2015. Is the meiofauna a good
1171
indicator for climate change and anthropogenic impacts? Marine Biodiversity 45: 505–535. 1172 pp
,
,
,
,
,
,
y
,
,
1167
Ivanenko VN, Sorensen MV, Vanreusel A, Thébault J, Mea M, Allio N, Andro T, Arvigo A, Castrec J,
1168
Danielo M, Foulon V, Fumeron R, Hermabessiere L, Hulot V, James T, Langonne-Augen R, Le Bot T, 1167
Ivanenko VN, Sorensen MV, Vanreusel A, Thébault J, Mea M, Allio N, Andro T, Arvigo A, Castrec J, 1167
Ivanenko VN, Sorensen MV, Vanreusel A, Thébault J, Mea M, Allio N, Andro T, Arvigo A, Castrec J,
1168
Danielo M, Foulon V, Fumeron R, Hermabessiere L, Hulot V, James T, Langonne-Augen R, Le Bot T,
1169
Long M, Mahabror D, Morel Q, Pantalos M, Pouplard E, Raimondeau L, Rio- Cabello A, Seite S,
1170
Traisnel G, Urvoy K, Van Der Stegen T, Weyand M, Fernandes D. 2015. Is the meiofauna a good
1171
indicator for climate change and anthropogenic impacts? Marine Biodiversity 45: 505–535. Manuscript to be reviewed Figure 1 Manuscript to be reviewed 1169
Long M, Mahabror D, Morel Q, Pantalos M, Pouplard E, Raimondeau L, Rio- Cabello A, Seite S,
1170
Traisnel G, Urvoy K, Van Der Stegen T, Weyand M, Fernandes D. 2015. Is the meiofauna a good 1169
Long M, Mahabror D, Morel Q, Pantalos M, Pouplard E, Raimondeau L, Rio- Cabello A, Seite S, 1173
Todorova V. and Konsulova T., 2005. Manual for quantitative sampling and sample treatment of marine-
1174
soft bottom macrozoobenthic. 38pp. 1175
Zuur AF, Ieno EN. 2016. A protocol for conducting and presenting results of regression-type analyses. 1176
Methods in Ecology and Evolution 7: 636–645. doi:10.1111/2041-210X.12577. 1177 1175
Zuur AF, Ieno EN. 2016. A protocol for conducting and presenting results of regression-type analyses. 1176
Methods in Ecology and Evolution 7: 636–645. doi:10.1111/2041-210X.12577. 1177 1178
Zwick WR, Velicer WF. 1986. Comparison of five rules for determining the number of components to
1179
retain. Psychological Bulletin 99: 432–442. https://doi.org/10.1037/0033-2909.99.3.432. 1178
Zwick WR, Velicer WF. 1986. Comparison of five rules for determining the number of components to
1179
retain. Psychological Bulletin 99: 432–442. https://doi.org/10.1037/0033-2909.99.3.432. 1178
Zwick WR, Velicer WF. 1986. Comparison of five rules for determining the number of components to
1179
retain. Psychological Bulletin 99: 432–442. https://doi.org/10.1037/0033-2909.99.3.432. 1180
Xu W Z, Cheung SG, Shin PK. 2014. Structure and taxonomic composition of free-living nematode and
1181
macrofaunal assemblages in a eutrophic subtropical harbour, Hong Kong. Marine pollution bulletin 85:
1182
764-773 1180
Xu W Z, Cheung SG, Shin PK. 2014. Structure and taxonomic composition of free-living nematode and 1181
macrofaunal assemblages in a eutrophic subtropical harbour, Hong Kong. Marine pollution bulletin 85:
1182
764-773. 1182
764-773. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed Figure 1 Map of the sampling area (Ancona harbor) and location of the five sampling stations. The pink rectangle indicates the geographical position of the Ancona harbor, Italy. Map of the sampling area (Ancona harbor) and location of the five sampling stations. The pink rectangle indicates the geographical position of the Ancon The pink rectangle indicates the geographical position of the Ancona harbor, Italy. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed Manuscript to be reviewed Figure 2 (a) Distribution pattern and (b) origin of Polycyclic aromatic hydrocarbons (PAHs)
characterizing the sediment at the investigated sampling stations. Σ LMW = sum of Low Molecular Weight compounds (Naphthalene, 1-Methylnaphtalene, 2-
Methylnaphtalene, Acenaphthylene, Acenaphthene, Fluorene, Phenanthrene, Anthracene); Σ
HMW = sum of High Molecular Weight compounds (Fluoranthene, Pyrene, Σ LMW = sum of Low Molecular Weight compounds (Naphthalene, 1-Methylnaphtalene, 2-
Methylnaphtalene, Acenaphthylene, Acenaphthene, Fluorene, Phenanthrene, Anthracene); Σ
HMW = sum of High Molecular Weight compounds (Fluoranthene, Pyrene,
Benz[a]anthracene, Chrysene, Benzo[b]fluoranthene, Benzo[k]fluoranthene, Benzo[a]pyrene,
7,12-Dimethylbenz[a]anthracene, Benzo[ghi]perylene, Indeno[1,2,3-cd]pyrene,
Dibenz[a,h]anthracene); Σ COMB = sum of 9 non-alkylated PAHs (Fluoranthene, Pyrene,
Benz[a]anthracene, Chrysene, Benzo[b]fluoranthene, Benzo[k]fluoranthene, Benzo[a]pyrene,
Benzo[ghi]perylene, Indeno[1,2,3-cd]pyrene). Σ LMW = sum of Low Molecular Weight compounds (Naphthalene, 1-Methylnaphtalene, 2-
Methylnaphtalene, Acenaphthylene, Acenaphthene, Fluorene, Phenanthrene, Anthracene); Σ
HMW = sum of High Molecular Weight compounds (Fluoranthene, Pyrene,
Benz[a]anthracene, Chrysene, Benzo[b]fluoranthene, Benzo[k]fluoranthene, Benzo[a]pyrene,
7,12-Dimethylbenz[a]anthracene, Benzo[ghi]perylene, Indeno[1,2,3-cd]pyrene,
Dibenz[a,h]anthracene); Σ COMB = sum of 9 non-alkylated PAHs (Fluoranthene, Pyrene,
Benz[a]anthracene, Chrysene, Benzo[b]fluoranthene, Benzo[k]fluoranthene, Benzo[a]pyrene,
Benzo[ghi]perylene, Indeno[1,2,3-cd]pyrene). Benz[a]anthracene, Chrysene, Benzo[b]fluoranthene, Benzo[k]fluoranthene, Benzo[a]pyrene,
Benzo[ghi]perylene, Indeno[1,2,3-cd]pyrene). PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed Manuscript to be reviewe PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023)
Manuscript to be reviewed PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Figure 3 Principal Component Analysis (PCA) with environmental variables. Principal Component Analysis (PCA) with environmental variables. BPC = Biopolymeric organic carbon; CPE = Chloroplastic pigment equivalent; BPER =
Benzo[ghi]perylene; BAP = Benzo[a]pyrene; FLU = Fluorene; ANT/PHE = Anthracene/
Phenanthrene; FLT/PYR = Fluoranthene/Pyrene. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed Manuscript to be reviewed Manuscript to be reviewed Figure 4 (a) Meiofaunal community structure and (b) contribution of taxa others at each
sampling station. Mean values of replicated samples (n =3) are shown. Bold values reported inside the bars are
total abundances (ind.10cm
-2) of meiofauna (4a) and others (4b). PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed Manuscript to be reviewed Figure 6 Non-metric Multidimensional Scaling (nMDS) of benthic communities. Figure 5 Meiofaunal univariate measures. (a) Meiofauna abundance (N = no. of individuals 10 cm
-2) and its diversity indices: (b)
meiofauna taxa richness (S), (c) Shannon’s diversity index (H’, based on log2), and (d)
Pielou’s equitability index (J’). Bars represent 95% confidence intervals. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed Manuscript to be reviewed Non-metric Multidimensional Scaling (nMDS) of benthic communities. Non-metric Multidimensional Scaling (nMDS) plots on (a) meiobenthic and (b) macrobenthic
community structures characterizing the sediment at the investigated sampling stations. Data presented were fourth root scale transformed prior to analysis. Data presented were fourth root scale transformed prior to analysis. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Figure 7 Macrofaunal community structure at each sampling station. Mean values of replicated samples (n =3) are shown. Manuscript to be reviewed Manuscript to be reviewed PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed Manuscript to be reviewed Table 1(on next page) Table 1(on next page) Figure 8 Macrofaunal univariate measures. (a) Macrofauna abundance (N = no. of individuals m
-2) and its diversity indices: (b)
macrofauna taxa richness (S), (c) Shannon’s diversity index (H’, based on log2), and (d)
Pielou’s equitability index (J’). Bars represent 95% confidence intervals. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) 5
d.f. = degrees of freedom; HSD = Honestly Significant Difference.
6 5
d.f. = degrees of freedom; HSD = Honestly Significant Difference.
6 Manuscript to be reviewed Table 1. One-way Analysis of Variance (ANOVA) results on total meiofauna and macrofauna
abundance (N) and their diversity indices. Number of taxa (S), Shannon’s diversity index (H’, based on
log2), and Pielou’s equitability index (J’), characterizing the sediment at the investigated sampling
stations. In bold significant p values. Meiofauna
Index
d.f. F value
p value
Tukey's HSD post hoc test
Stations
4
714.30
<0.001
DS>LR>MAN>API>PORT
N
Residuals
10
Stations
4
11.30
<0.001
PORT<API=DS=LR=MAN
S
Residuals
10
Stations
4
44.10
<0.001
MAN>PORT>API=LR(=DS)>DS
H’
Residuals
10
Stations
4
41.60
<0.001
MAN>PORT=API=LR(=DS)>DS
J’
Residuals
10
Macrofauna
Index
d.f. F value
p value
Tukey's HSD post hoc test
Stations
4
10.60
0.002
LR>API=DS=MAN=PORT
N
Residuals
10
Stations
4
3.80
0.041
MAN<LR
S
Residuals
10
Stations
4
8.20
0.003
LR<API=DS=MAN=PORT
H’
Residuals
10
Stations
4
3.90
0.036
LR<API
J’
Residuals
10 1
Table 1. One-way Analysis of Variance (ANOVA) results on total meiofauna and macrofauna
2
abundance (N) and their diversity indices. Number of taxa (S), Shannon’s diversity index (H’, based on
3
log2), and Pielou’s equitability index (J’), characterizing the sediment at the investigated sampling
4
stations. In bold significant p values. 1
Table 1. One-way Analysis of Variance (ANOVA) results on total meiofauna and macrofauna
2
abundance (N) and their diversity indices. Number of taxa (S), Shannon’s diversity index (H’, based on
3
log2), and Pielou’s equitability index (J’), characterizing the sediment at the investigated sampling
4
stations. In bold significant p values. Table 1(on next page) One-way Analysis of Variance (ANOVA) results on total meiofauna and macrofauna
abundance (N) and their diversity indices. Number of taxa (S), Shannon’s diversity index (H’, based on log2), and Pielou’s equitability
index (J’), characterizing the sediment at the investigated sampling stations. In bold
significant p values. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed Table 2(on next page) Table 2(on next page) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) 3
d.f. = degrees of freedom; SS = Sum of Squares; MS = Mean Square; P(perm) = Permutation p-value; P(MC) =
4
Monte Carlo p-value. 3
d.f. = degrees of freedom; SS = Sum of Squares; MS = Mean Square; P(perm) = Permutation p-value; P(MC) =
4
Monte Carlo p value Table 2(on next page) Permutational Multivariate Analysis of Variance (PERMANOVA) and pairwise test and
results on total meiofaunal community composition. In bold significant values are reported. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 3
d.f. = degrees of freedom; SS = Sum of Squares; MS = Mean Square; P(perm) = Perm Manuscript to be reviewed Table 2. Permutational Multivariate Analysis of Variance (PERMANOVA) and pairwise test and
results on total meiofaunal community composition. In bold significant values are reported. PERMANOVA
Source
d.f. SS
MS
Pseudo-F value
P(perm)
Unique permutations
Stations
4
7315.80
1829
38.0437
0.001
9915
Residuals
10
480.84
48.084
Total
14
7796.70
PAIRWISE TEST
Groups
t
value
P(perm)
P(MC)
MAN vs. PORT
7.39
0.103
<0.001
MAN vs. LR
8.21
0.099
<0.001
MAN vs. DS
4.86
0.099
0.004
MAN vs. API
6.12
0.099
0.002
PORT vs. LR
8.23
0.095
<0.001
PORT vs. DS
6.44
0.101
<0.001
PORT vs. API
5.35
0.098
0.002
LR vs. DS
2.95
0.101
0.016
LR vs. API
5.20
0.101
0.002
DS vs. API
3.98
0.098
0.005 1
Table 2. Permutational Multivariate Analysis of Variance (PERMANOVA) and pairwise test and
2
results on total meiofaunal community composition. In bold significant values are reported. 4
Monte Carlo p-value. Manuscript to be reviewed Manuscript to be reviewed Table 3(on next page) Table 3(on next page) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 1
Table 3. Distance-based
L ModelinM (( L analysis on meiofaunal community composition
2
characterizing the sediment at the investigated sampling stations. V
are
c
after
V
3
Rotated PCA A
as follof ARC
= A
A
c varia
asso
c
A
to t
t
rotated a
(
indi
c
A
t
t
4
num
of t
t
a
a
B =
B
A
c varia
asso
c
A
to t
t
rotated a
F
a
c
A
list of t
t
5
varia
refers to t
t
te
A
START SO
L S
AICc
R2
RSS
No. of variables
Selections
80.92
0.94
447.00
6
All
SEQUENTIAL TESTS
Variable
AICc
SS
Pseudo-F value
p value
Prop. Cumul. Res. d.f. ! 76.25
92.83
1.66
0.224
0.01
0.93
9
"
74.60
173.62
2.89
0.53
0.02
0.91
10
BEST SOLUTION
AICc
R2
RSS
No. of variables
Selections
74.59
0.91
713.44
4
ARC1-
ARC4
6
AIC
# = A
$%& $')* Information Criterion
#+, ,'#- '. for small samples; R2 = Coeffi
#& 'C
- of determination; RSS =
7
Residuals Sums of Squares; SS = Sums of Squares; Prop. = Proportion of variation e
e/0 %
& C
'. b1 t
2
' varia
b
0'v
8
Cumul. = Cumulative total of Prop; Res. d.f. = Residual degrees of freedom. 9 1
Table 3. Distance-based
L ModelinM (( L analysis on meiofaunal community composition
2
characterizing the sediment at the investigated sampling stations. V
are
c
after
V
3
Rotated PCA A
as follo ARC
= A
varia
asso
to t
rotated a
(
indi
t 6
AIC
# = A
$%& $')* Information Criterion
#+, ,'#- '. for small samples; R2 = Coeffi
#& 'C
- of determination; RSS =
7
Residuals Sums of Squares; SS = Sums of Squares; Prop. = Proportion of variation e
e/0 %
& C
'. b1 t
2
' varia
b
0'v
8
Cumul. = Cumulative total of Prop; Res. d.f. = Residual degrees of freedom. 9 6
AIC
# = A
$%& $')* Information Criterion
#+, ,'#- '. Table 3(on next page) Distance-based Linear Modeling (DistLM) analysis on meiofaunal community
composition characterizing the sediment at the investigated sampling stations. Variables are coded after Varimax Rotated PCA Axis as follow: ARCx = Abiotic variables
associated to the rotated axis (x indicates the number of the axis); BRCx = Biotic variables
associated to the rotated axis. For a complete list of the variables refers to the text. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed Table 4(on next page) Table 4(on next page) Manuscript to be reviewed for small samples; R2 = Coeffi
#& 'C
- of determination; RSS =
7
Residuals Sums of Squares; SS = Sums of Squares; Prop. = Proportion of variation e
e/0 %
& C
'. b1 t
2
' varia
b
0'v
8
Cumul. = Cumulative total of Prop; Res. d.f. = Residual degrees of freedom. 9 PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed 24 PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) 3
d.f. = degrees of freedom; SS = Sum of Squares; MS = Mean Square; P(perm) = Permutation p-value; P(MC) =
4
Monte Carlo p-value. 3
d.f. = degrees of freedom; SS = Sum of Squares; MS = Mean Square; P(perm) = Permutation p-value; P(MC) =
4
Monte Carlo p-value. Manuscript to be reviewed Manuscript to be reviewed Table 4. Permutational
3
4
56 789: 796 ; Analysis of Variance (PER3
< => ?<@ and pair
p 7D ; test and
results on total macrofaunal community composition. In bold significant values are reported. PER 3
<
=
>
? <
Source
d.f. SS
3
E
Pseudo-F value
P(perm)
Uni
U
4; permutations
Stations
4
23534
5883.5
4.55
GHGGI
9915
Residuals
10
12936
1293.6
Total
14
36470
PAIRPJ
E
K TEST
N
:
O
4
Q
D
t
value
P(perm)
P(3
R
@
MAN vs. PORT
2.03
0.12
GHGTW
MAN vs. LR
2.67
0.10
GHGII
MAN vs. DS
2.41
0.09
GHGI
X
MAN vs. API
3.09
0.10
GHGGY
PORT vs. LR
1.57
0.10
0.100
PORT vs. DS
1.71
0.10
0.069
PORT vs. API
2.03
0.10
GHG
Z
I
LR vs. DS
1.57
0.10
0.106
LR vs. API
2.23
0.10
GHGT
[
DS vs. API
2.03
0.10
GHG
Z
\ 1
Table 4. Permutational
3
4
56 789: 796 ; Analysis of Variance (PER3
< => ?<@ and pair
p 7D ; test and
2
results on total macrofaunal community composition. In bold significant values are reported. 4
Monte Carlo p-value. Table 4(on next page) Permutational Multivariate Analysis of Variance (PERMANOVA) and pairwise test and
results on total macrofaunal community composition. In bold significant values are reported. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) 3
d.f. = degrees of freedom; SS = Sum of Squares; MS = Mean Square; P(p Manuscript to be reviewed Manuscript to be reviewed Manuscript to be reviewed Table 5(on next page) Table 5(on next page) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) q
;
q
;
p
p
8
Cumul. = Cumulative total of Prop; Res. d.f. = Residual degrees of freedom. p
;
7
Residuals Sums of Squares; SS = Sums of Squares; Prop. = Proportion of variation e
6
AIC
= A
Information Criterion
for small samples; R2 = Coeffi
of Table 5(on next page) Distance-based Linear Modeling (DistLM) analysis on macrofaunal community
composition characterizing the sediment at the investigated sampling stations. Variables are coded after Varimax Rotated PCA Axis as follow: ARCx = Abiotic variables
associated to the rotated axis (x indicates the number of the axis); BRCx = Biotic variables
associated to the rotated axis. For a complete list of the variables refers to the text. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) Manuscript to be reviewed rs u ws xyz { are
| }~
z
~
3
after
rs u w
s
Rotated PCA A
w
{ as follo
ARC
= A
xw}
w| varia
xy
z { asso
| w
s
z ~ to t
z rotated a
w
{ (
4
indi
|
s
z { t
z num
x
z
u of t
z a
w{
=
w}
w| varia
xy
z { asso
|
ws
z
~ to t
z rotated a
w
{
}u a
| }
y
z
z
5
list of t
z varia
xyz { refers to t
z te
1
Table 5. Distance-based Linear
]^_` d ghi (DistL
]j analysis on macrofaunal community
2
composition c
klm ln
o
`
m
g
q
ghi t
k
` sediment at t
k
` investigated sampling stations. rs u ws xyz { are
| }~
z
~
3
after
Rotated PCA A
as follo
ARC
= A
varia
asso
to t
rotated a
( 6
AIC
= A
Information Criterion
for small samples; R2 = Coeffi
of determination; RSS =
7
Residuals Sums of Squares; SS = Sums of Squares; Prop. = Proportion of variation e
¡¢ t
£
varia
¡
¤
8
Cumul. = Cumulative total of Prop; Res. d.f. = Residual degrees of freedom. 6
AIC
= A
Information Criterion
for small samples; R2 = Coeffi
of determination; RSS =
7
Residuals Sums of Squares; SS = Sums of Squares; Prop. = Proportion of variation e
¡¢ t
£
varia
¡
¤
8
Cumul. = Cumulative total of Prop; Res. d.f. = Residual degrees of freedom. 6
AIC
= A
Information Criterion
for small samples; R2 = Coeffi
of determination; RSS =
7
Residuals Sums of Squares; SS = Sums of Squares; Prop. = Proportion of variation e
¡¢ t
£
varia
¡
¤
8
Cumul. = Cumulative total of Prop; Res. d.f. = Residual degrees of freedom. Manuscript to be reviewed 1
Table 5. Distance-based Linear
]^_` d ghi (DistL
]j analysis on macrofaunal community
2
composition c
klm ln
o
`
m
g
q
ghi t
k
` sediment at t
k
` investigated sampling stations. rs u ws xyz { are
| }~
z
~
3
after
rs u w
s
Rotated PCA A
w
{ as follo
ARC
= A
xw}
w| varia
xy
z { asso
| w
s
z ~ to t
z rotated a
w
{ (
4
indi
|
s
z { t
z num
x
z
u of t
z a
w{
=
w}
w| varia
xy
z { asso
|
ws
z
~ to t
z rotated a
w
{
}u a
| }
y
z
z
5
list of t
z varia
xyz { refers to t
z te
START SOLUTION
AICc
R2
RSS
No. of variables
Selections
28.74
0.70
11856
6
All
SEQUENTIAL TESTS
Variable
AICc
SS
Pseudo-F value
p value
Prop. Cumul. Res. d.f.
123.48
1746.70
1.23
0.251
0.048
0.65
9
ARC3
120.00
2136.70
1.53
0.131
0.059
0.59
10
118.28
3193.80
2.17
0.022
0.000
0.51
11
ARC1
116.72
2906.30
1.78
0.049
0.080
0.43
12
BEST SOLUTION
AICc
R2
RSS
No. of variables
Selections
116.72
0.43
20821
2
ARC2;
ARC4
6
AIC
= A
Information Criterion
for small samples; R2 = Coeffi
of determination; RSS =
7
Residuals Sums of Squares; SS = Sums of Squares; Prop. = Proportion of variation e
¡¢ t
£
varia
¡
¤
8
Cumul. = Cumulative total of Prop; Res. d.f. = Residual degrees of freedom. 1
Table 5. Distance-based Linear
]^_` d ghi (DistL
]j analysis on macrofaunal community
2
composition c
klm ln
o
`
m
g
q
ghi t
k
` sediment at t
k
` investigated sampling stations. PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023) PeerJ reviewing PDF | (2023:01:81322:1:1:NEW 2 May 2023)
|
https://openalex.org/W3042955872
|
https://link.springer.com/content/pdf/10.1007/s00415-020-10068-2.pdf
|
English
| null |
Use of a symptom-based questionnaire to screen for the presence of significant voiding dysfunction in patients with multiple sclerosis and lower urinary tract symptoms: a pilot study
|
Journal of neurology
| 2,020
|
cc-by
| 5,720
|
* Jalesh N. Panicker
j.panicker@ucl.ac.uk Abstract Introduction Lower urinary tract dysfunction is common in people with multiple sclerosis, leading to overactive bladder
symptoms, voiding difficulties or a combination. First-line medications for overactive bladder symptoms are effective. Current
guidelines recommend measuring post-void residual volume (PVR) before commencing these treatments, as they can poten-
tially exacerbate voiding difficulties in those with significant underlying voiding dysfunction (pre-treatment PVR > 100 ml). However, facilities to do so are not readily available to all clinicians, potentially delaying effective therapy. Aims To conduct a pilot study investigating the association between lower urinary tract symptoms and PVR volume in peo-
ple with multiple sclerosis using a validated questionnaire and to determine if questionnaire scores can be used to exclude
a significantly elevated (> 100 ml) PVR volume. Introduction Lower urinary tract dysfunction is common in people with multiple sclerosis, leading to overactive bladder
symptoms, voiding difficulties or a combination. First-line medications for overactive bladder symptoms are effective. Current
guidelines recommend measuring post-void residual volume (PVR) before commencing these treatments, as they can poten-
tially exacerbate voiding difficulties in those with significant underlying voiding dysfunction (pre-treatment PVR > 100 ml). However, facilities to do so are not readily available to all clinicians, potentially delaying effective therapy. Ai
T
d
il
d i
i
i
h
i i
b
l
i
d PVR
l
i guidelines recommend measuring post-void residual volume (PVR) before commencing these treatments, as they can poten-
tially exacerbate voiding difficulties in those with significant underlying voiding dysfunction (pre-treatment PVR > 100 ml). However, facilities to do so are not readily available to all clinicians, potentially delaying effective therapy. Aims To conduct a pilot study investigating the association between lower urinary tract symptoms and PVR volume in peo-
ple with multiple sclerosis using a validated questionnaire and to determine if questionnaire scores can be used to exclude
a significantly elevated (> 100 ml) PVR volume. gfi
gi
y g
g y
p
lities to do so are not readily available to all clinicians, potentially delaying effective therapy. f
Aims To conduct a pilot study investigating the association between lower urinary tract symptoms and PVR volume in peo-
ple with multiple sclerosis using a validated questionnaire and to determine if questionnaire scores can be used to exclude
a significantly elevated (> 100 ml) PVR volume. Journal of Neurology (2020) 267:3683–3688
https://doi.org/10.1007/s00415-020-10068-2 Journal of Neurology (2020) 267:3683–3688
https://doi.org/10.1007/s00415-020-10068-2 ORIGINAL COMMUNICATION Abstract i
Methods Patients with multiple sclerosis referred to a tertiary hospital uro-neurology service completed the Urinary Symp-
tom Profile questionnaire and underwent PVR measurement by bladder ultrasound. A ratio of the questionnaire low stream
score/total score was calculated to standardise the relative degree of voiding symptoms compared to overall lower urinary
tract symptoms. Results Of 40 patients (29 females, mean age 50 years), 30% had an elevated PVR volume. PVR volume was correlated with
low stream score and ratio of low stream/total score. A cut-off of > 0.15 for low stream/total score ratio had 92% sensitivity
and 71% specificity in predicting an elevated PVR volume. i
Conclusion A symptom-based questionnaire maybe a useful screening tool to distinguish patients in whom PVR measure-
ment is required from those who could safely start on treatment for overactive bladder symptoms. Keywords Multiple sclerosis · Lower urinary tract symptoms · Management · Questionnaire 1
Queen Square Multiple Sclerosis Centre, Department
of Neuroinflammation, UCL Queen Square Institute
of Neurology, Faculty of Brain Sciences, University College
London, London, UK Vivien Li1,2 · Jalesh N. Panicker2,3 · Collette Haslam2 · Jeremy Chataway1,4 Vivien Li1,2 · Jalesh N. Panicker2,3 · Collette Haslam2 · Jeremy Chataway1,4 Received: 12 May 2020 / Revised: 8 July 2020 / Accepted: 9 July 2020 / Published online: 16 July 2020
© The Author(s) 2020 * Jalesh N. Panicker
j.panicker@ucl.ac.uk 1
Queen Square Multiple Sclerosis Centre, Department
of Neuroinflammation, UCL Queen Square Institute
of Neurology, Faculty of Brain Sciences, University College
London, London, UK
2
Department of Uro‑Neurology, The National Hospital
for Neurology and Neurosurgery, Queen Square, London, UK
3
Faculty of Brain Sciences, UCL Queen Square Institute
of Neurology, University College London, London, UK
4
National Institute for Health Research, Biomedical Research
Centre, University College London Hospitals, London, UK Introduction Lower urinary tract (LUT) dysfunction is common in mul-
tiple sclerosis (MS) affecting up to 75% of patients. Patients
may experience different LUT symptoms depending on the
site of demyelinating lesions in the central nervous system. Lesions of the suprapontine or spinal micturition pathways
result in symptoms such as urinary urgency, urgency uri-
nary incontinence, increased daytime frequency and noc-
turia, collectively known as storage or overactive bladder
(OAB) symptom [1]. Urodynamic studies typically dem-
onstrate detrusor overactivity. Dysfunction of the voiding
phase can result from spinal cord lesions producing detru-
sor-sphincter dyssynergia where there is loss of coordinated
activity between the detrusor and urethral sphincters. Less
commonly, voiding dysfunction can arise due to detrusor 2
Department of Uro‑Neurology, The National Hospital
for Neurology and Neurosurgery, Queen Square, London, UK 3
Faculty of Brain Sciences, UCL Queen Square Institute
of Neurology, University College London, London, UK 4
National Institute for Health Research, Biomedical Research
Centre, University College London Hospitals, London, UK :(012345
1 3 (0121 3456789)
3 Journal of Neurology (2020) 267:3683–3688 3684 α-receptor antagonists, required a urinary catheter (including
intermittent self-catheterisation, urethral indwelling catheter
or suprapubic catheter) and had other urological conditions,
including urinary tract infection (based on urine dipstick
performed on the sample produced for PVR measurement),
or another neurological condition. underactivity. In both situations, symptoms include hesi-
tancy, straining, slow and interrupted stream and incomplete
bladder emptying [2]. Based on the UK consensus on the optimal management
of LUT dysfunction in MS, all patients with MS presenting
with new bladder symptoms should undergo measurement
of the post-void residual (PVR) volume prior to starting oral
agents for OAB symptoms, namely anti-muscarinic agents or
β3-receptor agonist [3]. This is because subjective voiding
symptoms are often unreliable [4]. Furthermore, the risks of
missing an elevated PVR include increased risk for urinary
tract infections and exacerbation of storage symptoms as
the undetected voiding dysfunction could worsen following
treatment commencement. However, facilities and equip-
ment to measure PVR are not readily accessible to many
neurologists, which may lead to delay in or never commenc-
ing highly effective treatment. One study surveying almost
10,000 patients with MS found 65% experienced at least one
moderate LUT symptom, but only half received treatment
with an anti-muscarinic medication [5]. Data collected included patient demographics, MS sub-
type, duration of MS, use of disease-modifying therapies
and duration of LUT symptoms. Introduction Patients completed the Uri-
nary Symptom Profile (USP) (Box 1), a standardised vali-
dated 13-item questionnaire of LUT symptoms addressing
three domains: stress urinary incontinence (SUI) (maximum
score 9), overactive bladder (OAB) (maximum score 21)
and low stream (LS) (maximum score 9) symptoms, with a
greater score indicating worse symptoms [7]. To standardise
the contribution of voiding symptoms represented by the LS
score to the overall degree of LUT symptoms, the ratio of
USP LS/total score was calculated. Overactive bladder (OAB) The aims of this pilot study are to investigate the associa-
tion between LUT symptoms amongst patients with MS and
their PVR measurements using a validated questionnaire
and to determine if questionnaire scores could be used to
exclude a significantly elevated (> 100 ml) PVR. This would
potentially allow the questionnaire to be used to identify
patients with an elevated PVR if bladder ultrasound was
not available. 2. How many times a week have you had to rush
to the toilet to urinate because you urgently needed to
go? (0 = never, 1 = less than once a week, 2 = several
times a week, 3 = several times a day)
3. When you have had an urgent need to urinate,
for how many minutes on average have you been
able to hold on? (0 = more than 15 min, 1 = from 6 to
15 min, 2 = from 1 to 5 min, 3 = less than 1 min)
4. How many times a week have you experienced
a urine leak preceded by an urgent need to urinate
that you were unable to control? (0 = never, 1 = less
than once a week, 2 = several times a week, 3 = several
times a day)
(a)
In the above case, what kind of leaks did
you have? (0 = no leaks in this case, 1 = a few drops,
2 = light leaks, 3 = heavy leaks) 2. How many times a week have you had to rush
to the toilet to urinate because you urgently needed to
go? (0 = never, 1 = less than once a week, 2 = several
times a week, 3 = several times a day) 3. When you have had an urgent need to urinate,
for how many minutes on average have you been
able to hold on? (0 = more than 15 min, 1 = from 6 to
15 min, 2 = from 1 to 5 min, 3 = less than 1 min) 4. How many times a week have you experienced
a urine leak preceded by an urgent need to urinate
that you were unable to control? (0 = never, 1 = less
than once a week, 2 = several times a week, 3 = several
times a day) Stress urinary incontinence (SUI) 1. Over the past 4 weeks, please specify the num-
ber of times a week you had leaks during physical
effort: (0 = no urine leaks, 1 = less than one urine leak
a week, 2 = several urine leaks a week, 3 = several
urine leaks a day) (a)
During strenuous physical effortf f
(b) During moderate physical effortf f
(c)
During light physical effort Box 1 Urinary symptom profile [7] A previous study by Milleman et al. found that clinical
factors associated with an increased risk of elevated PVR
in women with OAB symptoms include age older than
55 years, history of incontinence surgery, history of MS and
stage 2 or greater vaginal prolapse [6]. However, there is a
paucity of research into the association between LUT symp-
toms experienced by MS patients in particular and their PVR
measurements. A simple tool that can be easily administered
to patients with MS experiencing LUT symptoms to help
distinguish those who are at high versus low likelihood of
having an elevated PVR would be clinically useful. Results 6. How many times on average have you woken
up during the night by a need to urinate? (0 = never
or once, 1 = twice, 2 = 3 or 4 times, 3 = more than 4
times) Methods (0 = 2 h or more, 1 = between 1
and 2 h, 2 = between 30 min and 1 h, 3 = less than
30 min)
6. How many times on average have you woken
up during the night by a need to urinate? (0 = never
or once, 1 = twice, 2 = 3 or 4 times, 3 = more than 4
times)
7. How many times a week have you had a
urine leak while asleep or have you woken up wet? (0 = never, 1 = less than once a week, 2 = several times
a week, 3 = several times a day)
Low stream (LS)
8. How would you describe your usual urination over
these past 4 weeks? (0 = normal, 1 = needed to push
with abdominal (stomach) muscles or lean forward (or
required a change of position) to urinate, 2 = needed to
press on the lower stomach with my hands, 3 = used a
catheter)
9. In general, how would you describe your urine
flow? (0 = normal, 1 = weak, 2 = drop by drop,
3 = used a catheter)
10. In general, how has your urination been? (0 = nor-
mal and quick, 1 = difficult to start then normal,
1 = easy at first but slow to finish, 2 = very slow from
start to finish, 3 = used a catheter) Over the last 4 weeks and under everyday conditions
of social, professional or family life: i
Assessments were conducted as part of routine clinical
management. Patients provided written informed consent for
anonymised questionnaire data to be used. 5. During the day, in general, how long elapsed
between urinating? (0 = 2 h or more, 1 = between 1
and 2 h, 2 = between 30 min and 1 h, 3 = less than
30 min) 5. During the day, in general, how long elapsed
between urinating? (0 = 2 h or more, 1 = between 1
and 2 h, 2 = between 30 min and 1 h, 3 = less than
30 min) Low stream (LS) 8. How would you describe your usual urination over
these past 4 weeks? (0 = normal, 1 = needed to push
with abdominal (stomach) muscles or lean forward (or
required a change of position) to urinate, 2 = needed to
press on the lower stomach with my hands, 3 = used a
catheter) 9. In general, how would you describe your urine
flow? (0 = normal, 1 = weak, 2 = drop by drop,
3 = used a catheter) 9. In general, how would you describe your urine
flow? (0 = normal, 1 = weak, 2 = drop by drop,
3 = used a catheter) 10. In general, how has your urination been? (0 = nor-
mal and quick, 1 = difficult to start then normal,
1 = easy at first but slow to finish, 2 = very slow from
start to finish, 3 = used a catheter) USP score and PVR measurement All patients underwent non-invasive uroflowme-
try with determination of maximum urinary flow rate
(Qmax) and measurement of PVR in clinic via ultra-
sound by trained nursing staff (Albyn Medical Smartflow,
Bardscan Portable Ultrasound). PVR was classified as
elevated if it exceeded 100 ml [3]. Mean USP questionnaire SUI, OAB, LS and total scores
were 0.78 (range 0–7), 6.4 (range 0–18) and 1.6 (range 0–4),
8.8 (range 0–22) respectively, indicating a mix of storage
and voiding LUT symptoms. Mean (SD, range) PVR was 88 (103, 0–387) ml and was
greater than 100 ml in 12 patients (30%). There was a signifi-
cant difference in mean PVR between the group of patients
with PVR of below or above 100 ml (33 ml versus 218 ml,
mean difference 185 ml, 95% confidence interval [CI] 145-
225 ml, p < 0.001). Mean (SD) Qmax was 20 (16) ml/s and
there was no significant correlation between PVR and Qmax
(r2 = 0.12, p = 0.08). Characteristics of patients with PVR
below or over 100 ml is shown in Table 1. Only mean USP
LS score and USP LS/total score demonstrated significant
differences between the two groups. There was trend towards
a significant difference in sex distribution (p = 0.056). Demographics A total of 40 patients (29 female) completed both the USP
questionnaire and PVR measurement. Of these 30 (75%)
were Caucasian. Mean (SD) and median (IQR) age were 50
(13) years and (41–60) respectively, with no significant dif-
ference between females (mean 51, SD 12 years) and males
(mean 50, SD 16 years). Mean (SD) and median (IQR)
duration since MS diagnosis was 18 (13) and 14.5 years
(9–24) respectively. 28 (70%) had relapsing remitting MS,
7 (17.5%) had secondary progressive MS and 5 (12.5%)
had primary progressive MS. 17 (43%) were currently on
disease-modifying therapy. Mean (SD) and median (IQR)
duration of LUT symptoms was 7.8 (5.9) and 6.5 (3–12)
years respectively. Mean age of patients with RRMS was
significantly lower than those with progressive MS [47 vs. 61 years, p = 0.002, 95% confidence interval (CI) 5.2–22]. However, there was no statistically significant difference in
duration since MS diagnosis (15 vs. 23 years) nor duration of
LUT symptoms (7.8 vs. 8.0 years). There were no significant
sex differences in proportion with relapsing remitting versus
progressive MS and duration of MS and LUT symptoms. 7. How many times a week have you had a
urine leak while asleep or have you woken up wet? (0 = never, 1 = less than once a week, 2 = several times
a week, 3 = several times a day) Methods The target population was patients aged 18 and over with a
confirmed diagnosis of MS referred to the Uro-Neurology
department at National Hospital for Neurology and Neu-
rosurgery from April 2019 to January 2020. Patients were
excluded if they were on medications to treat LUT symp-
toms, such as anti-muscarinic agents, β3-receptor agonists or (a)
In the above case, what kind of leaks did
you have? (0 = no leaks in this case, 1 = a few drops,
2 = light leaks, 3 = heavy leaks) 3 Journal of Neurology (2020) 267:3683–3688 3685 Over the last 4 weeks and under everyday conditions
of social, professional or family life:
5. During the day, in general, how long elapsed
between urinating? (0 = 2 h or more, 1 = between 1
and 2 h, 2 = between 30 min and 1 h, 3 = less than
30 min)
6. How many times on average have you woken
up during the night by a need to urinate? (0 = never
or once, 1 = twice, 2 = 3 or 4 times, 3 = more than 4
times)
7. How many times a week have you had a
urine leak while asleep or have you woken up wet? (0 = never, 1 = less than once a week, 2 = several times
a week, 3 = several times a day)
Low stream (LS)
8. How would you describe your usual urination over
these past 4 weeks? (0 = normal, 1 = needed to push
with abdominal (stomach) muscles or lean forward (or
required a change of position) to urinate, 2 = needed to
press on the lower stomach with my hands, 3 = used a
catheter)
9. In general, how would you describe your urine
flow? (0 = normal, 1 = weak, 2 = drop by drop,
3 = used a catheter)
10. In general, how has your urination been? (0 = nor-
mal and quick, 1 = difficult to start then normal,
1 = easy at first but slow to finish, 2 = very slow from
start to finish, 3 = used a catheter) high (> 100 ml) or normal (< 100 ml)]. A level of p < 0.05
was considered statistically significant. Over the last 4 weeks and under everyday conditions
of social, professional or family life:
5. During the day, in general, how long elapsed
between urinating? Statistical analysis 1) and USP LS/total score
(r = 0.422, p = 0.007), but not SUI or OAB scores. Males
had significantly higher PVR (mean 165 ml vs. 60 ml, mean
difference 105 ml, 95% CI 39–171, p = 0.003) and USP LS
score (mean 2.36 vs. 1.34, mean difference 1.02, 95% CI
0.03–2.01, p = 0.045) than females. In order to determine the ability of the USP question-
naire, specifically the USP LS/total score ratio, to predict
receiver operating characteristics (ROC) curve was cal-
culated (Fig. 2). AUC was 0.77 (95% confidence interval
0.62–0.93, p = 0.007). A cut-off value of 0.15 produced a
sensitivity of 92% and false positive rate of 29%. Amongst
this cohort, only one patient with a PVR of over 100 ml had
a ratio below 0.15 and was misclassified (i.e., false negative). Fig. 1 Correlation between USP low stream score and post-void
residual volume
Fig. 2 Receiver operating characteristics curve for USP low stream
(LS)/total score ratio in predicting a post-void residual volume of
over 100 ml Fig. 2 Receiver operating characteristics curve for USP low stream
(LS)/total score ratio in predicting a post-void residual volume of
over 100 ml Fig. 1 Correlation between USP low stream score and post-void
residual volume Fig. 1 Correlation between USP low stream score and post-void
residual volume 1 3 Statistical analysis Fisher’s exact test was used to analyse the relationship
between categorical variables, including PVR [classified as 1 3 Journal of Neurology (2020) 267:3683–3688 3686 Table 1 Characteristics of patients (n = 40) with PVR below or over 100 ml
PVR post-void residual volume, CI confidence interval, SD standard deviation, RRMS relapsing–remitting multiple sclerosis, SPMS secondary
progressive multiple sclerosis, PPMS primary progressive multiple sclerosis, LUT lower urinary tract, USP urinary symptom profile, SUI stress
urinary incontinence, OAB overactive bladder, LS low stream
PVR < 100 ml
PVR > 100 ml
p value, mean difference, 95% CI
Sex [females (%)]
23 (58%)
6 (15%)
Age (mean, SD; median, IQR)
52 (11), 55 (43–60)
50 (17), 47 (34–62)
MS subtype [RRMS (%), SPMS (%), PPMS (%)]
20 (50%)
4 (10%)
4 (10%)
8 (20%)
3 (7.5%)
1 (2.5%)
MS duration, years (mean, SD; median, IQR)
18 (12), 16 (9–25)
17 (14),12 (7–16)
LUT symptom duration, years (mean, SD; median, IQR)
7.3 (6.4), 5 (2.8–12)
8.9 (4.9),
10 (4–11)
USP SUI score (mean, SD; median, IQR)
0.79 (1.5), 0 (0–1)
0.75 (1.2), 0 (0–1.3)
USP OAB score (mean, SD; median, IQR)
6.4 (3.5), 6 (4–8)
6.4 (4.1), 6 (3.3–9)
USP LS score (mean, SD; median, IQR)
1.1 (1.2), 1 (0–2)
2.8 (1.2), 3 (2–4)
p < 0.001, 1.7, 0.88–2.6
USP total score (mean, SD; median, IQR)
8.3 (4.6), 8 (5–10)
10 (5.4), 9 (7–11)
USP LS/total score (mean, SD; median, IQR)
0.15 (0.20), 0 (0–0.23)
0.32 (0.21), 0 (0.17–0.39)
p = 0.019, 0.17, 0.029–0.31 Table 1 Characteristics of patients (n = 40) with PVR below or over 100 ml PVR post-void residual volume, CI confidence interval, SD standard deviation, RRMS relapsing–remitting multiple sclerosis, SPMS secondary
progressive multiple sclerosis, PPMS primary progressive multiple sclerosis, LUT lower urinary tract, USP urinary symptom profile, SUI stress
urinary incontinence, OAB overactive bladder, LS low stream PVR post-void residual volume, CI confidence interval, SD standard deviation, RRMS relapsing–remitting multiple sclerosis, SPMS secondary
progressive multiple sclerosis, PPMS primary progressive multiple sclerosis, LUT lower urinary tract, USP urinary symptom profile, SUI stress
urinary incontinence, OAB overactive bladder, LS low stream Predictors of PVR
Based on univariate analyses, patient age, duration of MS,
MS subtype and duration of LUT symptoms were not asso-
ciated with PVR. PVR was correlated with USP LS score
(r = 0.573, p < 0.001) (Fig. Statistical analysis Statistical analysis was performed in IBM SPSS Statistics
version 25. T test was used to compare means between
groups. Pearson test was used to assess correlation between
variables (PVR and questionnaire scores, PVR and Qmax). Linear regression was used to analyse the relationship
between independent variables to the dependent variable of
PVR. Discussion Invasive urodynamics
data was only available in a small proportion of our cohort,
so we were not able draw conclusions on the relationship
between these findings, questionnaire responses and PVR. In the context of MS, the aetiology of an elevated USP LS
score was assumed to be neurogenic LUT dysfunction such
as detrusor-sphincter dyssynergia or detrusor underactivity. However, given the mean age of this population of 51 years,
other non-neurological causes such as benign prostatic
hypertrophy may also contribute to a raised PVR, which
was found to be higher in the male patients. Traditionally it has been thought that whilst storage LUT
symptoms are generally a good guide as to the presence of
detrusor overactivity, they are less reliable when there is
additional voiding dysfunction and incomplete emptying
[4]. Interestingly our study found that USP LS score was
moderately correlated with PVR volume. Previous studies
assessing the relationship between neurological and uro-
logical symptoms as well as subjective urinary symptoms
versus objective urological measurements have produced
mixed findings. One study which assessed EDSS and func-
tional system scores (FSS), clinical LUT symptoms and fill-
ing cystometry in patients with MS found that the degree
of lower limb pyramidal weakness correlated with severity
of urological symptoms in general, as both are thought to
reflect the extent of spinal involvement. However, only 47%
of patients with a significantly raised PVR reported a sensa-
tion of incomplete bladder emptying, suggesting that lack
of voiding LUT symptoms cannot sufficiently exclude the
presence of a high PVR accurately [8]. In another study, an EDSS pyramidal or bladder/bowel
FSS of ≤ 1 was found to be not associated with incomplete
bladder emptying, suggesting that it may not be necessary
to measure the PVR in such patients [9]. However, the
opposite findings have also been reported in another larger
study which found no association between EDSS bladder/
bowel FSS and measured PVR [10]. One weakness of using
the EDSS is low intra-rater and inter-rater reproducibility,
especially for patients with mild to moderate disability [11]. Furthermore, as the bladder/bowel FSS only ranges from
0–3, it does not sufficiently delineate the severity nor fully
encapsulate the nature of symptoms. Severity of LUT symp-
toms as assessed by unstructured interview also did not pre-
dict objective measures of LUT dysfunction including PVR,
Qmax and the presence of detrusor-sphincter dyssynergia
[12]. Discussion Whilst numerous questionnaires on bladder symptoms are
available, the USP was chosen as it is a validated question-
naire that incorporates different domains of LUT symptoms
and better describes the frequently mixed symptoms expe-
rienced by patients with MS. Although not specifically vali-
dated in neurological patients, it has been recommended as
a screening tool for LUT symptoms in MS [13] and used as
an outcome measure in previous studies of patients with MS
[14–16] and other neurological conditions [17, 18]. A ratio
of the USP LS/total scores was calculated to account for the
degree of voiding LUT symptoms compared to a patient’s
overall profile of LUT symptoms. Our study found that using a USP LS/total score ratio of
0.15 as a cut-off has good sensitivity in predicting patients
with MS who have a PVR exceeding 100 ml. For healthcare
professionals encountering the large number of patients with
MS who have OAB symptoms and are considering an anti-
muscarinic agent or β3-receptor agonist, but do not have
direct or immediate access to facilities to measure the PVR,
this short symptom-based patient questionnaire can provide
a way to distinguish between patients in whom treatment
can be initiated without delay from those who require PVR
measurement first. Whilst the false positive rate of 29%
is relatively high and these patients may be subjected to a
degree of inconvenience as a result of waiting for a formal
PVR measurement to be carried out, this is a non-invasive
and essentially risk-free investigation. i
The limitations of our study are the relatively small num-
ber of individuals who underwent assessment, but the demo-
graphic characteristics of our patient population was within
the range of the previously reported studies [8–10, 12]. Patients were also taken from those referred to a tertiary uro-
neurology service, so may have more severe LUT symptoms
than the overall population of patients with MS. However,
as referrals are generally made for those with significant
symptoms in whom more advanced investigation and treat-
ment is being considered pending evaluation such as PVR
measurement, these may also be the individuals in whom
such a tool is most valuable. Our study only used one PVR
measurement and this can fluctuate over time and on differ-
ent occasions. Therefore, repeated measurements would be
useful to further validate this finding. Predictors of PVR Fig. 2 Receiver operating characteristics curve for USP low stream
(LS)/total score ratio in predicting a post-void residual volume of
over 100 ml Based on univariate analyses, patient age, duration of MS,
MS subtype and duration of LUT symptoms were not asso-
ciated with PVR. PVR was correlated with USP LS score
(r = 0.573, p < 0.001) (Fig. 1) and USP LS/total score
(r = 0.422, p = 0.007), but not SUI or OAB scores. Males
had significantly higher PVR (mean 165 ml vs. 60 ml, mean
difference 105 ml, 95% CI 39–171, p = 0.003) and USP LS
score (mean 2.36 vs. 1.34, mean difference 1.02, 95% CI
0.03–2.01, p = 0.045) than females. receiver operating characteristics (ROC) curve was cal-
culated (Fig. 2). AUC was 0.77 (95% confidence interval
0.62–0.93, p = 0.007). A cut-off value of 0.15 produced a
sensitivity of 92% and false positive rate of 29%. Amongst
this cohort, only one patient with a PVR of over 100 ml had
a ratio below 0.15 and was misclassified (i.e., false negative). In order to determine the ability of the USP question-
naire, specifically the USP LS/total score ratio, to predict
a PVR of over 100 ml, area under the curve (AUC) of the 1 3 1 3 Journal of Neurology (2020) 267:3683–3688 3687 Discussion As this is a pilot study, further validation in a larger
cohort, including patients of different ages and levels of
disability, and different settings, such as in primary care or
general neurology and MS clinics, would be required before
this tool could be used in practice. All patients commenced
on an anti-muscarinic medication or β3-receptor agonist
for OAB symptoms do still require regular monitoring and
prompt review is advised if patients experience symptoms
suggestive of a deterioration in voiding function, such as a
paradoxical worsening of OAB symptoms or recurrent uri-
nary tract infections. While a ratio of the USP LS/total score
has not been used in previous studies, the pilot data from our
study raise the possibility of using a simple validated ques-
tionnaire as a screening tool to distinguish patients in whom
further urological investigations including PVR measure-
ment is required from those that could be safely started on 1 3 Journal of Neurology (2020) 267:3683–3688 3688 com/retrieve/pii/B9780444529015000174 [Accessed January 26,
2020]. com/retrieve/pii/B9780444529015000174 [Accessed January 26,
2020]. treatment for OAB symptoms. This could reduce the delay
in initiating first-line therapy for these frequently disabling
but treatable symptoms. 3. Fowler CJ, Panicker JN, Drake M et al (2009) A UK consensus
on the management of the bladder in multiple sclerosis. J Neurol
Neurosurg Psychiatry 80(5):470 Acknowledgements JNP undertook this work at UCLH/UCL Institute
of Neurology and is supported in part by funding from the United
Kingdom’s Department of Health NIHR Biomedical Research Centres
consultant PA funding scheme. g
y
y
4. Fowler CJ (2003) Investigation and management of neuro-
genic bladder dysfunction. J Neurol Neurosurg Psychiatry
74(90004):27iv–2731 5. Mahajan ST, Patel PB, Marrie RA (2010) Under treatment of
overactive bladder symptoms in patients with multiple sclerosis:
an ancillary analysis of the NARCOMS patient registry. J Urol
183(4):1432–1437 Funding In the last 3 years, JC has received support from the Effi-
cacy and Evaluation (EME) Programme, a Medical Research Council
(MRC) and National Institute for Health Research (NIHR) partnership
and the Health Technology Assessment (HTA) Programme (NIHR), the
UK MS Society, the US National MS Society and the Rosetrees Trust. He is supported in part by the National Institute for Health Research,
University College London Hospitals, Biomedical Research Centre,
London, UK. 6. Milleman M, Langenstroer P, Guralnick ML (2004) post-void
residual urine volume in women with overactive bladder symp-
toms. J Urol 172(5):1911–1914 7. Discussion Haab F, Richard F, Amarenco G et al (2008) Comprehensive eval-
uation of bladder and urethral dysfunction symptoms: develop-
ment and psychometric validation of the urinary symptom profile
(USP) questionnaire. Urology 71(4):646–656 8. Betts CD, D’Mellow MT, Fowler CJ (1993) Urinary symptoms
and the neurological features of bladder dysfunction in multiple
sclerosis. J Neurol Neurosurg Psychiatry 56(3):245–250 Data availability Anonymised data will be shared by request from
qualified investigators. 9. Kirchhof K, Fowler C (2000) The value of the kurtzke functional
systems scales in predicting incomplete bladder emptying. Spinal
Cord 38(7):409–413 Conflicts of interest The authors declared that they have no conflict
of interest. Conflicts of interest The authors declared that they have no conflict
of interest. 10. Kragt JJ, Hoogervorst ELJ, Uitdehaag BMJ, Polman CH (2004)
Relation between objective and subjective measures of bladder
dysfunction in multiple sclerosis. Neurology 63(9):1716–1718 Ethical standard statement Assessments were conducted as part of
routine clinical management. Patients provided written informed con-
sent for anonymised questionnaire data to be used. 11. Ontaneda D, Fox RJ, Chataway J (2015) Clinical trials in progres-
sive multiple sclerosis: lessons learned and future perspectives. Lancet Neurol 14(2):208–223 12. Haggiag S, Bolla G, Picconi O, Galgani S, Gasperini C (2017)
Discrepancies between urinary symptoms assessment and objec-
tive bladder dysfunctions in multiple sclerosis. Mult Scler Demy-
elinating Disord 2(1):11i Open Access This article is licensed under a Creative Commons Attri-
bution 4.0 International License, which permits use, sharing, adapta-
tion, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons licence, and indicate if changes
were made. The images or other third party material in this article are
included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in
the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a
copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. 13. Coyle PK (2016) Symptom management and lifestyle modifica-
tions in multiple sclerosis: CONTINUUM. Lifelong Learn Neurol
22(3):815–836 14. Mauruc E, Guinet-Lacoste A, Falcou L et al (2017) Nocturnal
urinary disorders and multiple sclerosis: clinical and urodynamic
study of 309 patients. J Urol 197(2):432–437 15. Motavasseli D, Chesnel C, Charlanes A et al (2018) Adherence to
anticholinergic therapy and clean intermittent self-catheterization
in patients with multiple sclerosis. Int Neurourol J 22(2):133–141 16. Chesnel C, Charlanes A, Hentzen C et al (2018) Lower urinary
tract symptoms in elderly population with multiple sclerosis. Int
Neurourol J 22(1):58–64 17. Batla A, Pareés I, Edwards MJ et al (2016) Lower urinary tract
dysfunction in patients with functional movement disorders. J
Neurol Sci 361:192–194 References 1. Gajewski JB, Schurch B, Hamid R et al (2018) An international
continence society (ICS) report on the terminology for adult neu-
rogenic lower urinary tract dysfunction (ANLUTD). Neurourol
Urodyn 37(3):1152–1161 18. Lad M, Parkinson MH, Rai M et al (2017) Urinary, bowel and
sexual symptoms in a cohort of patients with Friedreich’s ataxia. Orphanet J Rare Dis 12(1):158 2. Panicker JN, De Sèze M, Fowler CJ. Neurogenic lower urinary
tract dysfunction. In: Handbook of Clinical Neurology. Vol 110. Elsevier; 2013:209–220. Available at: https://linkinghub.elsevier. 1 3
|
https://openalex.org/W2793791904
|
https://tos.org/oceanography/assets/docs/30-4_briscoe.pdf
|
English
| null |
Thank You, Walter Munk, for Being There at the Beginning
|
Oceanography
| 2,017
|
cc-by
| 673
|
CITATION Briscoe, M. 2017. Thank you, Walter Munk, for being there at the beginning. Oceanography 30(4):8, https://doi.org/10.5670/oceanog.2017.402. https://doi.org/10.5670/oceanog.2017.402 COPYRIGHT COPYRIGHT
This article has been published in Oceanography, Volume 30, Number 4, a quarterly
journal of The Oceanography Society. Copyright 2017 by The Oceanography Society. All rights reserved. USAGE Permission is granted to copy this article for use in teaching and research. Republication, systematic reproduction, or collective redistribution of any portion of
this article by photocopy machine, reposting, or other means is permitted only with the
approval of The Oceanography Society. Send all correspondence to: info@tos.org or
The Oceanography Society, PO Box 1931, Rockville, MD 20849-1931, USA. Thank You, Walter Munk,
for Being There at the Beginning By Mel Briscoe Photo by Erik Jepsen/UC San Diego Publications Photo by Erik Jepsen/UC San Diego Publications Photo by Erik Jepsen/UC San Diego Publications Walter Munk turned 100 on October 19, 2017. I first met Walter
almost half a century ago, when I was a young scientist work-
ing in Europe, and was in the process of changing my focus
from fluid mechanics to physical oceanography. At that time,
an older European colleague told me that there were just two
real oceanographers in the United States: Henry Stommel and
Walter Munk. Over the past 50 years, I’ve had the satisfac-
tion of working down the hallway from Henry (at Woods Hole
Oceanographic Institution), and of having contact with Walter
on a fair number of occasions. In addition to being a gregari-
ous and engaging gentleman, Walter has been an inspiring and
enthusiastic colleague to many of us, and a formative presence
to more than a few. is a long story in and of itself. (I worked closely with Walter
through the Heard Island years, and we both learned a lot about
whales.) The 1989 article was Walter’s second in Oceanography;
his first was in volume 1, number 1, on ocean acoustic tomogra-
phy (https://doi.org/10.5670/oceanog.1988.31). Over the years,
Walter has authored 14 submissions to Oceanography and has
been mentioned in the magazine over 200 times. Articles writ-
ten by or mentioning Walter now have more than 350,000 hits
on Google. And he is still working, on wind waves as it turns out,
a fitting return to one of his early successes: predicting the wave
conditions for the D-Day invasion of Normandy. Walter is a strikingly gracious gentleman who has had an
enormous positive impact on our science and on the people
he has encountered, myself included. A recent biography and
tribute that appeared in 2016 in Acoustics Today (Volume 12,
pages 36–42) provides more context for these statements; no
need to repeat them here. Those of us working in oceanogra-
phy today are fortunate to have his personality and his intel-
lect as part of our culture. And TOS is fortunate to have had
his sincere and continuing interest in the success of our Society. Carry on, Walter! Walter played an important role in the creation of The
Oceanography Society. Oceanography | Vol.30, No.4
8 AUTHOR
Mel Briscoe (mel@oceangeeks.com) is President, OceanGeeks LLC,
Alexandria, VA, USA. DOWNLOADED FROM HTTP://TOS.ORG/OCEANOGRAPHY MILESTONES Thank You, Walter Munk,
for Being There at the Beginning By Mel Briscoe His eclectic approach to oceanogra-
phy and his penchant for speaking and writing simply, directly,
and clearly so that many, rather than just a few, could under-
stand him were already legendary—and inspired the mission of
Oceanography magazine. In our initial inquiries to colleagues
in 1987 prior to forming TOS, Walter was encouraging about
the Society’s goals and structure. In his closing remarks at the
Inaugural Meeting of TOS in 1989, he commented, “It is time for
the oceanographic disciplines to come together. It is time for an
Oceanography Society.” His full remarks are in Oceanography,
volume 2, number 2 (https://doi.org/10.5670/oceanog.1989.19);
that issue also contains his initial article about the Heard Island
Experiment (https://doi.org/10.5670/oceanog.1989.10), which AUTHOR
Mel Briscoe (mel@oceangeeks.com) is President, OceanGeeks LLC,
Alexandria, VA, USA. ARTICLE CITATION
Briscoe, M. 2017. Thank you, Walter Munk, for being there at the beginning. Oceanography 30(4):8, https://doi.org/10.5670/oceanog.2017.402. 8
|
https://openalex.org/W2344927793
|
https://www.mdpi.com/2075-5309/6/2/20/pdf?version=1461922172
|
English
| null |
An Analysis of BIM Web Service Requirements and Design to Support Energy Efficient Building Lifecycle
|
Buildings
| 2,016
|
cc-by
| 13,877
|
Yufei Jiang 1,*, Xiao Liu 1, Fangxiao Liu 2, Dinghao Wu 1 and Chimay J. Anumba 2 1
College of Information Sciences and Technology, The Pennsylvania State University, University Park,
PA 16802, USA; xvl5190@ist.psu.edu (X.L.); dinghao@psu.edu (D.W.) p
g
p
2
Department of Architectural Engineering, The Pennsylvania State University, University Park,
PA 16802, USA; fzl116@psu.edu (F.L.); anumba@engr.psu.edu (C.J.A.) p
(
)
g p
*
Correspondence: yzj107@psu.edu; Tel.: +1-814-880-2305 p
y j
p
†
This paper is an extended version of our paper published in Proceedings of the 2012 ASCE International
Conference on Computing in Civil Engineering, Special Session on BIM Computer Science Fundamentals,
2012 [1]. †
This paper is an extended version of our paper published in Proceedings of the 2012 ASCE International
Conference on Computing in Civil Engineering, Special Session on BIM Computer Science Fundamentals,
2012 [1]. Academic Editor: Tanyel Bulbul Received: 24 September 2015; Accepted: 25 April 2016; Published: 29 April 2016 Received: 24 September 2015; Accepted: 25 April 2016 Abstract:
Energy Efficient Building (EEB) design, construction, and operations require the
development and sharing of building information among different individuals, organizations,
and computer applications. The Representational State Transfer (RESTful) Building Information
Modeling (BIM) web service is a solution to enable an effective exchange of data. This paper
presents an investigation into the core RESTful web service requirements needed to effectively
support the EEB project lifecycle. The requirements include information exchange requirements,
distributed collaboration requirements, internal data storage requirements, and partial model query
requirements. We also propose a RESTful web service design model on different abstraction layers
to enhance the BIM lifecycle in energy efficient building design. We have implemented a RESTful
Application Program Interface (API) prototype on a mock BIMserver to demonstrate our idea. We evaluate our design by conducting a user study based on the Technology Acceptance Model
(TAM). The results show that our design can enhance the efficiency of data exchange in EEB
design scenarios. Keywords: BIM; web service; energy efficient building lifecycle 1. Introduction p
y
The third method is a set of web interfaces based on the REST style architecture, or so called
RESTful interface. Implementing web services based on the RESTful interface is most suitable for
BIM. The strength of the RESTful interface is its simplicity. In this architecture, the representation of
all resources would be abstracted as a Uniform Resource Identifier (URI). The resources here do not
just refer to the files or other tangible objects; it also could be a composition of several files, a table,
a query, a result set of queries, and any concepts. The web interface is essential to the usability of
a web service. As previously mentioned, simulations for different scenes and purposes are highly
domain-specific. RESTful web interfaces can effectively isolate and hide internal implementation and
domain knowledge from end users. Thus, the domain experts can fully focus on their areas. They can
invoke other services and retrieve data from other parties as utilities, without knowing their details. Remote data models could be manipulated by invoking RESTful web interfaces. Clients are loosely
coupled with the web service provider, which makes the procedures like data model merging, partial
model query, and events notification flexible. BIM, which is an emerging methodology to facilitate information exchange among the whole
AEC community, is not just designed for EEB projects. However, in this paper, we focus in particular
on the EEB lifecycle. An EEB project consists of tasks ranging from building constructions to energy
analysis. There are many existing simulation tools provided by different parties that assist experts in
various domains to perform analysis, such as EnergyPlus and DOE-2 [11]. However, to effectively
use them, different data have to be taken as inputs. Figure 1 shows some examples of simulation
tools and their required data models [12]. To get the required data, we need to trace back to some
source software. Assume an expert wants to model the energy consumptions in EnergyPlus, she will
need the data in EnergyPlus Input Data Files (IDF). To extract these data, she is supposed to find the
IFC (Industry Foundation Classes) file which is generated by the ArchiCAD. However, the process
depends on a number of systems: the EnergyPlus software, an adapter for data format transform,
and the source software ArchiCAD. 1. Introduction The topic of facilitating data exchange between different design and simulation tools in the
Architecture, Engineering, and Construction (AEC) industry is gaining increasing interest. One
notable fact is that AEC has its specific ways to adopt the assistance from computer and information
sciences in the whole building design life cycle [2]. Instead of taking advantage of a web-service-based
solution, the workstation-based Computer-Aided Design (CAD) approach, which dates back to the
1980s, is still widely used by the AEC industry [2]. In current AEC workflows, simulation processes
are not fully integrated into the design and modeling process [3]. In most cases, engineers need
to manually build models for simulation, which duplicates work that has already been performed. It was reported that approximately 80% of the effort required to run a simulation is spent on building
models [4]. With Building Information Modeling (BIM), attributes and data can be attached to
a model, which potentially allows integrated analysis and simulation, especially for Energy Efficient
Building (EEB) projects [5,6]. However, most of the analysis and simulation software are developed www.mdpi.com/journal/buildings Buildings 2016, 6, 20; doi:10.3390/buildings6020020 www.mdpi.com/journal/buildings 2 of 24 Buildings 2016, 6, 20 Buildings 2016, 6, 20 by experts in different domains, and their data representations are different in nature, which puts
a constraint on model reuse and data exchange among tools. by experts in different domains, and their data representations are different in nature, which puts
a constraint on model reuse and data exchange among tools. Web service, as a software system to launch functions and exchange data over the web, is
a powerful approach to solving the interoperability problem of BIM. There are three kinds of
commonly used web service interfaces: remote procedure call (RPC), service-oriented architecture
(SOA), and representational state transfer (REST) style architecture [7]. RPC allows two distributed
remote heterogeneous systems to call the functions or methods of each other. However, their
interactions and function calls are environmentally sensitive, which relate to the implementations
in a specific programming language. This design pattern violates the principle of loose coupling
in software engineering. To overcome the design defect, SOA was proposed and now SOA
has been widely adopted by many web service providers [8]. Instead of relying on a specific
implementation, SOA is driven by messages (events) [9]. One of the disadvantages of this architecture
is its complexity [10]. 1. Introduction p
p
yp
•
We compare and discuss our design with existing technologies and implementations by
conducting a user study under the guidance of TAM. Figure 1. Data exchange in simulation tools of Energy Efficient Building (EEB) projects. Figure 1. Data exchange in simulation tools of Energy Efficient Building (EEB) projects. This paper is organized as follows. First, we introduce previous efforts on BIM-based building
life cycles and web-service-facilitated BIM in Section 2. Then, we present the details of requirements
for RESTful BIM web service to support EEB in Section 3, and propose our designs in Section 4. Section 5 introduces our implementation. We evaluate and discuss our approach in Section 6. We
conclude the paper in Section 7. 1. Introduction It is quite challenging, without an automated workflow, to figure
out the whole process and then acquire the permission of these systems to get the data originated
and transformed. To make this easier, if we adopt the RESTful design, the IDF file can be generated
and stored when the IFC file is created with the ArchiCAD. It will be provided as a resource online
directly for the expert to retrieve whenever he make an HTTP request. The same idea can be applied
to the use cases with other simulation tools. Previous research has presented some broad overview on possible applications of web service
in the AEC community, the building lifecycle, and BIM. However, the study that focuses on
RESTful web service, which is adopted by more and more leading web service providers, is still
inadequate. Moreover, as an emerging area in the AEC area, the EEB lifecycle support gradually
involves more participants and digitalized data, which requires a new collaboration pattern than
ever before. Accordingly, a study on the requirements and the design of RESTful BIM web service
to support EEB lifecycle is necessary. Our goal is to use the proposed requirements and designs to 3 of 24 Buildings 2016, 6, 20 Buildings 2016, 6, 20 investigate existing platforms and the technology adoption status within the EEB lifecycle supporting
area. In addition, we also implement an EEB BIM data RESTful Application Program Interface
(API) prototype and conduct a comparative study on some existing platforms and systems. We
take advantage of the Technology Acceptance Model (TAM) as our methodology to perform this
comparison and evaluation. More specifically, in this paper, we make the following contributions: •
We identify the requirements of RESTful BIM web services to support the EEB lifecycle. We identify the requirements of RESTful BIM web services to support the EEB lifecycle. We propose a set of design models of RESTful BIM web services on different abstraction lev We identify the requirements of RESTful BIM web services to support the EEB lifecycle. We propose a set of design models of RESTful BIM web services on different abstraction levels. y
q
pp
•
We propose a set of design models of RESTful BIM web services on different p
p
g
We implement a set of RESTful API on a mock BIMserver as a prototype. 2.1. Building Information Modeling Data and information fragmentation, redundancies and inefficiencies have been challenging
the AEC community in recent years, and BIM is being widely adopted in the industry to foster
communication and collaboration [13]. According to the National Building Information Modeling
Standard [14], Building Information Modeling is defined as “a digital representation of physical and
functional characteristics of a facility”, and it serves as “a shared knowledge resource for information
about a facility forming a reliable basis for decisions during its lifecycle from inception onward”. BIM acts as an integrated platform for team members to share and exchange project information
through comprehensive object-oriented building models [15,16]. It utilizes CAD that ties all building
components together with rich information embedded [17], and can be implemented in the lifecycle
of a project representing the plan, design, construct and operate phases [18]. 4 of 24 Buildings 2016, 6, 20 2.2.1. BIM to Support Building Life Cycle The traditional AEC industry relies on two-dimensional (2D) printed drawings and CAD
files that store 2D geometry information of building elements [19,20]. With the adoption of 3D
graphic prototyping technologies, 3D CAD emerged and was quickly introduced to the construction
industry [20]. BIM is the latest generation of object-oriented CAD systems (OOCAD) and is capable
of assigning both graphic and non-graphic attributes to building elements
[19]. In addition to
supporting collaborative work processes, BIM has been successfully implemented in other areas
in the AEC industry. The use of BIM goes beyond the planning and design stage of the project
throughout the building life cycle: it can be utilized as a tool to facilitate sustainable design and
analysis [21], automatic safety checking and hazard identification [22], and automated construction
process measurement [23]. It bridges the building information across the design teams to construction
teams and the operators by allowing each party to retrieve, append and modify information during
their period of contribution. Figure 2 shows 25 BIM applications throughout the lifecycle of a building
(plan, design, construct and operate) which are developed by the Computer Integrated Construction
Research program at the Pennsylvania State University [24]. Figure 2. Twenty-five Building Information Modeling (BIM) applications. Figure 2. Twenty-five Building Information Modeling (BIM) applications. 3.1. Information Exchange Requirements In EEB design and maintenance, there are a substantial number of participants and stakeholders. Accordingly, scalability is the first concern in the information exchange requirements. Simple Object
Access Protocol (SOAP) is one of the existing designs for communication; however, the performance
of the server drops when the whole system scales up [35]. A typical BIM-assisted project needs
collaboration between architects, engineers, contractors, and project managers. It is possible that
conflict checks and clash detections will be performed round after round whenever updates happen. The BIM web service should take advantage of the stateless feature of REST to avoid repeated
construction and improve the performance. “Stateless” means that a BIM web service does not store
its clients’ states. The stateless nature of RESTful service improves the scalability of the server and
allows application-aware caching intermediaries, which make the client-side access more efficient. pp
g
The second requirement of facilitating information exchange in EEB lifecycle is resource
accessibility. A key to building a RESTful BIM web service is to identify and abstract appropriate
resources that will be exposed to the end users or their clients. A resource is an object with a type,
associated data, relationships to other resources, and a set of methods applied to it. BIM has its
data model to organize data from different parties. In EEB lifecycle supporting scenarios, the data
models do not only contain the geometry information of a building, but also have airflow simulation
information, lighting information, Heating, Ventilating, and Air Conditioning (HVAC) information,
and geographic information. By adopting the RESTful design pattern, each single resource should
map to a Universal Resource Identifier (URI). Resources in the same category should be grouped
into collections. Each collection is also a resource, which should also be available by mapping to a
URI. This mapping can be done in a recursive bottom-up manner to make all resources accessible. In the current AEC community, there are already some general designs of BIM data models following
OO principles. The RESTful interface can be abstracted from these “objects” and the relationship
between objects. Additionally, since the exposed interfaces can be accessed by anyone along the
EEB lifecycle, authentication should also be considered in the Application Program Interface (API)
engineering. The notable point is that authentication implies client’s state tracking, which may
contradict RESTful stateless principle. 3. Requirements In this section, we discuss the requirements of the RESTful BIM web service to support EEB
lifecycle from end user perspectives. 2.2.2. BIM to Support Energy Efficient Building Design Several previous efforts have tried to integrate building information models with energy
simulation tools. Bazjanac [25] first disclosed the impact of a National Building Information
Model Standard (NBIMS) on building energy performance simulation and analysis. He [26] further
proposed a new methodology to semi-automate Building Energy Performance (BEP) simulation. This
methodology demonstrates benefits when applied to simulations with EnergyPlus. Young et al. [27]
surveyed that the lack of software interoperability and functionality is rated as one of the greatest
obstacles to improving the business value of BIM. Moon et al. [28] conducted case studies to evaluate
the interoperability of a BIM based architecture model and performance simulation programs (e.g.,
EnergyPlus). Jiang et al. [1] discussed a set of core BIMserver requirements to effectively support
information sharing in energy efficient retrofit projects. These requirements have incorporated the 5 of 24 Buildings 2016, 6, 20 BIM functionalities into efficient energy building designs and filled in the gap between the original
BIM standard and other platforms. In addition, they developed a tool called QueryGenerator to
automatically generate BIMserver Java query code [29]. Yu et al. [30] investigated a methodology
to integrate BIMserver with OpenStudio to efficiently exchange energy efficient building data in a
uniformed manner which provides a convenient tool for data transformation and make it easier for
information exchange. Liu et al. [31] proposed a framework to embed change management (CM) in
BIM. A prototype system platform is implemented based on this proposed framework. BIM has become a critical realm in the AEC industry, and its implementation will benefit owners,
contractors, designers and operators [32]. A well-designed BIM system can help save lifecycle cost
and cycle time, improve sustainable performance, and reduce human resources [33,34]. 3.2. Distributed Collaboration Requirements The topic of supporting distributed collaboration has been studied for two decades since the
modern enterprise has more and more “dynamic boundaries” and consists of a significant number of
“autonomous business units” [38]. Those large engineering and manufacturing enterprises found out
that the paper-based collaboration pattern was dramatically bogging down the whole workflow. An
important outcome of the research in this area is Product Data Management (PDM), which could be
dated back to the mid 90s [38]. The core idea of PDM is to track and manage “meta-knowledge”
(e.g., design release management, change management, and impact analysis) through the whole
product lifecycle [39]. To support distributed collaboration, a RESTful BIM web service should at
least play a role of PDM first to support distribution collaboration. p y
pp
More specifically, distributed collaboration requires a well designed mechanism to synchronize
contributions made by all teams. For example, it is highly possible that the geometry information
is updated after clash detection or energy evaluation. Under this circumstance, the geometry needs
revision on the basis of the supporting report. However, malfunctions may exist due to desynchrony
in other remote on-going development version branches. It would be problematic for geometry
engineers to tell upon which version the clash detection report is based, if more than one commitment
of updates in the geometry is submitted into the BIMserver. To tackle the malfunction caused by desynchrony, the BIM web service should retain a version
control system to enable each side to get informed on the change logs and the current version status. By indexing all the changes in the data model, a RESTful BIM web service should support the
roll-back feature. With a version control system, a BIM web service can adopt incremental backup
which is faster than a full backup. More advanced features, such as making the version control
system decentralized, will make a project iterate faster. However, since this advanced feature needs
every participant maintain their branches of a full copy of the BIM data, more training is required. Thus, a decentralized version control system is an optional requirement. Beyond that, a BIM web service must overcome those traditional PDM system shortfalls that
might be exposed by emerging challenges in EEB lifecycle supporting area. First, a traditional
PDM still uses a file or a document as a basic unit of querying and storing. 3.1. Information Exchange Requirements A solution is to adopt some open standard authentication
mechanisms like OAuth [36], which tracks client’s state in a different layer [37]. In practice, OAuth
has been reported to work well with REST [37]. In addition to that, all communication should take
place over certified Hypertext Transfer Protocol Secure (HTTPS) protocol. Different access control
policies should be applied to people as the roles they play vary in the whole lifecycle. Buildings 2016, 6, 20 6 of 24 3.2. Distributed Collaboration Requirements However, in EEB
design and analysis, a class or an abstract entity (e.g., a space, a boundary, or a facade) is a basic
unit to work on, which is hard to map to a specific document. A BIM web service must use an
additional abstract layer to make files and documents hidden from its clients. Second, a traditional
PDM integrates and manages the product data in a static way which is like a warehouse. In the
scenario of EEB design, integrated BIM models make some global analysis and optimization possible. A BIM web service host should have the capability of generating new data and knowledge in the
process of merging fragmented data from different departments. At last, traditional PDM targets
intra-enterprise communication and collaboration, instead of inter-enterprise collaboration. The
nature of EEB design is inter-enterprise based. As we have mentioned, different enterprises use
heterogeneous systems and data schema. Using the web service with simple and unified interfaces
to ease the data exchange between the distributed participants who work on heterogeneous local
systems is a must. 3.4.1. Automated Query Generation from MVD Automated Query Generation should take advantage of some existing documents that describe
resource data schema as input. Those documents could be written in natural languages or a domain
language that is familiar to domain experts. MVD meets these criteria, which makes it a potential
candidate of query generation input. An MVD is defined as a subset of the Industry Foundation
Classes (IFC) [42] schema to satisfy one or several specific data exchange requirements. For example,
the “Concept Design MVD” has been developed to facilitate selective extraction or importing of
relevant building information [26]. Whereas an MVD is independent of a particular IFC release, it is
implementation dependent. It is critical to automate the data exchange process by generating queries
from MVDs automatically. The automated generation of queries from MVDs will facilitate, simplify,
and streamline information exchanges between tools built around BIMservers, and enable flexible
queries to BIMserver. This functionality will also enable a BIMserver to selectively export relevant
information for one or multiple tools. 3.3. Internal Data Storage Requirements Internal data handling can be addressed using file-based and model-based approaches. These
two approaches have different strengths and shortcomings. A file server stores files statically to
provide data persistence and retrieval service. A model-driven server essentially is a web service
server. To be more specific, it is a Data-As-a-Service server. In this context, data refers to models. It does have a data persistence layer. However, it will not directly expose the information in the
data persistence layer as “data” to its clients. It parses the data in the file first and uses these data to
construct its own pre-defined data structure and maintain a comprehensive model as its internal data Buildings 2016, 6, 20 7 of 24 Buildings 2016, 6, 20 storage. This process is similar to a marsh up, which takes different web services as sources to render
them into a single new service [40]. Comparing web service servers and file servers, we can find out that a file server’s structure
can be quite clear and simple, which is easy to implement. The underlying database could be a
classical relational database. Adopting such a file server will not impact current energy-efficient
retrofit workflows. A file server does not necessarily need a unified file format or a format translator. It stores files and facilitates the collaboration of sections of design and simulation teams that share
the same file format. The basic unit of a file server is a file, while sometimes users have to query
data that is distributed among different files. For example, a simulation tool may need the height
and width of all windows. However, a file server may not be able to parse the information it stores
semantically. It also cannot support such data export or other advanced functions such as model
merging and partial model query [5]. A model-driven server can compensate for the file server’s
drawbacks. The advantage of this method is that some advanced functionalities (e.g., clash detection
and model merging) could be performed on the stored model on the fly [41]. 3.4. Partial Model Query Requirements To query data from a RESTful web service, a request is sent out by an HTTP call. This HTTP
call usually carries a payload that contains the data schema of interest in some lightweight formats
like JavaScript Object Notation (JSON). The JSON files could be generated by applications like
OpenStudio automatically. JSON files are easy to be parsed by a machine. However, it is not an
easy task for an end user to compose a long piece of JSON data schema to perform a query, especially
a partial model query. Thus, there are two major requirements on the issue of partial model query
composing. The first one is to automatically generate queries from Model View Definition (MVD). The other one is to have a domain query language. 4. Design In this section, we present three design patterns of RESTful BIM web service for developers
from three different levels. More specifically, from the low level to the high level, they are library
component as a service, building information model as a service, and application as a service. 3.4.2. Query Language •
Hidden from the users behind GUI and support transaction queries with ACID (atomicity,
consistency, isolation, and durability) [48], and able to reuse routine queries. 3.4.2. Query Language The other perspective to address the problems in partial model query is to develop a
domain-specific high-level declarative query language. Some general query languages have already
been available (e.g., Structured Query Language SQL). However, it is difficult to achieve the partial
model query of BIM data models directly by using a general query language [43]. Thus, one
of the requirements is to move closer to a seamless system integration using the fundamental
concepts behind the ontological engineering which helps to define domain-specific concepts. Current
integration efforts are usually based on how information are interpreted which is the syntax to
describe the information. Under this circumstance, an AEC industry query language should be
designed for such a purpose. However, with the increasing complexity of the information, we need
to completely and correctly exchange information among heterogeneous systems, and the language 8 of 24 Buildings 2016, 6, 20 itself should be unambiguous. For the areas that have special requirements on query and design,
adopting a domain-specific language (DSL) has been proven as a viable solution. An example is a
Very high-speed integrated circuit Hardware Description Language (VHDL), a hardware description
language [44]. To design an AEC domain specific query language, one should start from a query
language that can execute partial queries on an IFC/MVD compliant data model. Dating back to
2003, Predictive Modeling Query Language (PMQL) is an effort that tried to achieve partial model
query of IFC files [43]. Generalized Model Subset Definition (GMSD) goes one step further to support
better “partial model query on specific model view,” and less “request-response cycles” [45]. Recent
years have seen other designs like Building Information Model Query Language (BimQL) [46] and
Query Language for 4D Building Information Models (QL4BIM) [47]. In summary, a domain specific query language is needed and should have at least some of the
following characteristics: •
Compatible with MVDs: Queries should be able to be (semi-)automatically composed based on
the existing MVDs,
C
tibl
ith th RESTf l i t
f •
Compatible with MVDs: Queries should be able to be (semi-)automatically composed based on
the existing MVDs, g
•
Compatible with the RESTful interface, •
Easy for normal users to construct queries, and y
q
•
Hidden from the users behind GUI and support transaction queries with ACID (atomicity,
consistency, isolation, and durability) [48], and able to reuse routine queries. 4.1. Library Component as a Service A BIM service system consists of many different components. In this design pattern, we create
each component inside a BIM service as a standalone RESTful web service. Such components
could be authentication service, security plugin, or query engine. The benefits brought by such
a design style are multiple-fold. First, RESTful style architecture decouples each component of a
system. Each component interacts with each other via a set of RESTful web interfaces, which provide
better encapsulation than library calls [49]. Additionally, the dependency between components is
minimized. Therefore, the whole system is easy to decompose, integrate, refactor and migrate. When necessary, it is easy to change web interfaces into library calls. However, on the reverse side,
a system built based on library calls is hard to be split into several standalone RESTful web services. Third, because each component is designed to be standalone RESTful web service, functional tests
could be done earlier. Thus, it is feasible to launch part of the service ahead of the completion of
other components, which meets the requirements of agile development and quick delivery strategy. Fourth, because components interact with each other by HTTP calls, they can be deployed on
different distributed clusters, which enhances the scalability and portability of the whole service. EEB lifecycles involve considerable participants and stakeholders at different stages, which is a
distributed workflow in nature. A RESTful-web-service-enabled distributed system facilitates such a
distributed workflow. Figure 3 illustrates a library component as a service design. In this example, a RESTful BIM web
service consists of a web server, a data server, and other two functional components (component 1
and component 2). From the figure, we can see that the RESTful API is not only used for exposing
the available service for different clients, it is also the communication channel of each component
inside the box of the BIM web service system. For example, the web server receives a request to
fetch the temperature data of the first floor of a building in degrees Celsius. After parsing this
request, the web server will first send out an HTTP request to hit the data server to retrieve the
temperature data. The data server then replies an HTTP response that contains temperature data in 9 of 24 9 of 24 Buildings 2016, 6, 20 degrees Fahrenheit. 4.1. Library Component as a Service In this case, the web server sends out another HTTP request to component 2 for
calling the Fahrenheit-to-Celsius conversion service. At last, the web service returns the correct data
to the client. Figure 3. Library component as a service. Figure 3. Library component as a service. 4.2. Building Information Model as a Service 4.2. Building Information Model as a Service 4.2. Building Information Model as a Service In the design pattern of building information model as a service, there is a BIMserver playing the
role of central data hub. Each application is built surrounding this BIM data hub. Each application
interacts with central BIMserver via RESTful APIs. Figure 4 illustrates the design of the building
information model as a service. The BIMserver is not only a data host but also a data pipe among all
applications built upon it. Taking advantage of this model, a workflow with N applications merely
connects these applications with the central BIMserver, rather than building N(N −1) directed data
pipes. Because of the adoption of HTTP-protocol-based RESTful API, the way of connecting each
application and BIMserver is unified. All applications can rely on a set of fixed HTTP methods such as
POST, GET, DELETE, and PUT to perform state transition on the resources in the BIMserver without
any prior knowledge. It also eases exception handling. A set of fixed HTTP status codes tells a client
how a server responds to a request. A client without any prior knowledge of the server can still take
appropriate operations by parsing returned HTTP status code. For example, in any RESTful web
service, error code 404 means resource not found. Based on this, a client could be programmed to
have error handlers to each type of error code in advance. Second, each application can subscribe to
the specific service exposed by the BIMserver via a RESTful API. Whenever an event is triggered by
the operation launched by one application, other applications who subscribe to such type of events
will be notified and automatically perform corresponding routines. 10 of 24 10 of 24 Buildings 2016, 6, 20 Figure 4. BIM as a service. Figure 4. BIM as a service. 5. Implementation We implemented a proof-of-concept RESTful API prototype to demonstrate how a RESTful web
service can better streamline the data exchange in multizone airflow analysis. We chose multizone
airflow analysis for our demonstration because it is an important scenario in EEB lifecycle supporting
process. To save users from repeatedly creating models for airflow analysis, DeGraw et al. propose a
methodology to generate multizone airflow models from energy analysis BIM [50]. Multizone airflow
analysis models used by CONTAM, a multizone airflow and contaminant transport analysis software,
are simpler than energy analysis BIM used by EnergyPlus [50]. By simplifying and customizing
energy analysis models, it is possible to generate airflow analysis models. They describe the detailed
object translation process and the technical feasibility in the paper. 4.3. Application as a Service In the model of application as a service, there is no central data server. Each service has its
BIM data storage infrastructure, a local copy of BIM data, the application built upon it, and a set of
RESTful APIs. Each service could be a service provider and a service requester at the same time. An information exchange session could be established easily among arbitrary nodes of service in the
network. Service initial requests and response payload could be sent over HTTP protocol. BIM data
could be queried, pushed, and synchronized between any two services. Such collaboration pattern
allows each service has highly heterogeneous internal implementation and data models, which helps
architecture designers and analysts take advantage of their existing software assets. Additionally,
project data is hard to be compromised by a participant, since there is no central data server and
every application works on its data copy. Thus, procedures of authentication and reading/writing
permission granting are simplified. Even an untrusted third party can join the collaboration any time. If an EEB project requires a complete project BIM data warehouse that contains all most-up-to-date
updates for newcomers to copy from, a node of service could be assigned to be “master” node. Project maintainers can “select” useful updates from all participants. Please note that “master” is
just a naming convention. From a technical perspective, all nodes of service in a network are equally
important, or, we can say the node called “master” just offers the service of integration. Figure 5
illustrates this design pattern. 11 of 24 Buildings 2016, 6, 20 Figure 5. Application as a service. Figure 5. Application as a service. 5.2. Server Side On the server side, we build a mock BIMserver. To simplify the internal implementation and
better focus on web service and API design, we stimulate the on-the-fly energy model generation
process in a real BIMserver. We store some pre-defined and previously generated energy models
in JSON format in the server. There is no difference from an end user’s perspective. The whole
set of RESTful API and data schema was defined in the YAML (YAML is a “human friendly data
serialization standard for all programming languages [52]”. YAML is a recursive acronym for “YAML
Ain’t Markup Language”.)
language. Table 1 lists the details of each RESTful API. The server
host domain is restfulbimserver.mockable.io. Then a valid URL is domain name + API. For example,
Figure 7 shows a valid curl (curl is an “open source command line tool and library for transferring
data with a URL syntax [53]”) request that accesses a node named OS:Space2 in a RESTful approach. In the code listing, “GET” is the HTTP method. A user can use it to hit the endpoint (the URL) of
the nodes in an energy analysis model. The parameter, object name “OS%3ASpace2”, sets constraints
to this query. The returned result of this curl request is shown in Figure 8. Figure 9 shows a bit
more advanced request. The expression after the question mark (“?”) in the URI is the filter to select
the data objects “paths” whose elevation is greater than or equal to three among all available paths. Multiple filters can be connected by “&” in a single request. All attributes in a data object can be used
as filters in a query. The returned result of this curl request is shown in Figure 10. Table 1. A summary of RESTful designs and their benefits. EndPoint
Method
Resource Description
/buildings/{name}
GET
The whole building airflow network model
/nodes
GET, PUT, POST
Enclosed areas of the building are the nodes in the model. by hitting endpoints nodes,
users will retrieve all the available nodes in the model. /nodes/{name}
GET, PUT, POST
This endpoint refers to the nodes specified by their names. /paths/{name}
GET, PUT, POST
A path is a potential avenue of airflow. Each wall or component may have leakage,
then there is a path defined upon it. This end point refers to the paths specified
by their names. refers to the paths specified. 5.1. Overall Architecture The overall workflow of multizone airflow analysis and the role of our prototype is shown in
Figure 6. Airflow analysis needs three kinds of data. The first one is the 2D airflow zone which
can be generated out of energy models. This part consists of most data of airflow models. This
data retrieving process is enclosed by the dotted lines in Figure 6, which is implemented by our
prototype. Our prototype was implemented under such a client-server architecture. Our server that
publishes the APIs play a role of a BIMserver in the figure. The client plays a role of CONTAM in the
figure. The second part is the airtightness of all kinds of materials. This part of data can be fetched
from the National Institute of Standards and Technology (NIST) commercial buildings airtightness
database [51]. Airtightness will be assigned to every “boundary” of airflow zones according to
its attributes. The third part is the weather data, which can be obtained from other sources. The
airtightness data and weather data can be explored, accessed and updated by a set of RESTful API in
the same way as what we have done on energy models, which fits our application as a service design. 12 of 24 Buildings 2016, 6, 20 Figure 6. Example to show RESTful application as service. Figure 6. Example to show RESTful application as service. Figure 10. The path models returned from the server. 5.2. Server Side /thermalZones
GET, PUT, POST
Thermal zones are not airflow zones. A thermal zone contains all the space which
have similar thermal loads for an HVAC system. This endpoint refers to
all the thermal zones in the model. /thermalZones/{id}
GET, PUT, POST
This endpoint refers to the thermal zones specified by their names. /users/{username}
GET, PUT, POST, DELETE
This endpoint makes users’ information accessible. curl -X GET
--header "Accept: application /json"
"http :// restfulbimserver .mockable.io/nodes/OS%3 ASpace2" Table 1. A summary of RESTful designs and their benefits. Figure 7. An HTTP GET request to fetch node information in RESTful (Representational State
Transfer) style. 13 of 24 13 of 24 Buildings 2016, 6, 20 {
"name ": "OS:Space2",
"area ": 36,
"level ": "OS: BuildingStory
1",
"returnAir ": [
{
"system ": "OS: ThermalZone
1"
}
],
"volume ": 108,
"supplyAir ": [
{
"system ": "OS: ThermalZone
1"
}
],
"y": 6,
"x": 0,
"z": 0
} {
"name ": "OS:Space2",
"area ": 36,
"level ": "OS: BuildingStory
1",
"returnAir ": [
{
"system ": "OS: ThermalZone
1"
}
],
"volume ": 108,
"supplyAir ": [
{
"system ": "OS: ThermalZone
1"
}
],
"y": 6,
"x": 0,
"z": 0
} Figure 8. Node OS: Space2 data in Java Script Object Notation (JSON). curl -X GET
--header "Accept: application /json"
"http :// restfulbimserver .mockable.io/paths?elevation =>3" Figure 9. An HTTP GET request to retrieve paths whose elevation is greater than or equal to 3. {
"paths ": [
{
"elevation ": 3,
"name ": "OS:Surface
19",
"area ": 72,
"tilt ": 0,
"level ": "OS: BuildingStory
1",
"azimuth ": 90,
"nodes ": [
"OS:Space 1"
],
"type ": "roof"
},
{
"elevation ": 3,
"name ": "OS:Surface
12",
"area ": 36,
"tilt ": 0,
"level ": "OS: BuildingStory
1",
"azimuth ": 90,
"nodes ": [
"OS:Space 3"
],
"type ": "roof"
},
{
"elevation ": 3,
"name ": "OS:Surface 6",
"area ": 36,
"tilt ": 0,
"level ": "OS: BuildingStory
1",
"azimuth ": 90,
"nodes ": [
"OS:Space 2"
],
"type ": "roof"
}
]
}
Figure 10. The path models returned from the server. 5.2. Server Side {
"paths ": [
{
"elevation ": 3,
"name ": "OS:Surface
19",
"area ": 72,
"tilt ": 0,
"level ": "OS: BuildingStory
1",
"azimuth ": 90,
"nodes ": [
"OS:Space 1"
],
"type ": "roof"
},
{
"elevation ": 3,
"name ": "OS:Surface
12",
"area ": 36,
"tilt ": 0,
"level ": "OS: BuildingStory
1",
"azimuth ": 90,
"nodes ": [
"OS:Space 3"
],
"type ": "roof"
},
{
"elevation ": 3,
"name ": "OS:Surface 6",
"area ": 36,
"tilt ": 0,
"level ": "OS: BuildingStory
1",
"azimuth ": 90,
"nodes ": [
"OS:Space 2"
],
"type ": "roof"
}
]
}
Figure 10. The path models returned from the server. Figure 10. The path models returned from the server. 14 of 24 Buildings 2016, 6, 20 14 of 24 5.3. Client Side We build a client as a web service requester to play the role of data consumer just like CONTAM,
without implementing its airflow analysis logics. The implementation is based on the node.js
framework and Swagger framework [54]. It has a web-based user interface. It visualizes all the
resources and HTTP methods available on the server side. Our users can interact with the BIMserver
through this client which will be used in our evaluation and user study. Figures 11 and 12 show the screenshots of the client user interfaces. Figure 11a is the original
status of the web page when users visit the client. After clicking on a specific resource name such
as node, the page will display what kind of operations can be performed against this resource, which
is shown in Figure 11b. In addition, by clicking a method such as GET, the data schema and more
details of this API will be displayed, which is shown in Figure 12. At last, by clicking the “try it out”
button on a specific API, the client will send a request to the server and display the returned models
or the error code. An example is shown in Figure 12. Users with some programming skills can use a
shell command or write a script to hit the endpoints in the server side. (a)
(b)
Figure 11. Client user interface screenshots I. (a)
(b) (b) Figure 11. Client user interface screenshots I. Figure 12. Cont. Figure 12. Cont. 15 of 24 Buildings 2016, 6, 20 Figure 12. Client user interface screenshots II. Figure 12. Client user interface screenshots II. 6. Evaluation 6.1. Theoretical Framework and Study Measurements 6.1.2. Measurements and Study Design More specifically, BIMserver has two ways to query data. The first approach is to
write a piece of query code in Java. Figure 14 shows a code snippet from the BIMserver.org advanced
query demo. It queries the data of doors whose height is over 2 feet from the first floor. The other
way is also based on web interfaces. They integrate BIM Service interface exchange (BIMSie) [63] into
their implementation, which accepts SOAP and JSON as web API call payload. However, BIMserver
itself does not offer RESTful style web services. The BIMSie interfaces abstract library calls rather
than resources. In addition, all requests are sent via HTTP GET method, no matter if it is adding a
project or it is deleting a user. Table 2 shows a list of BIMsie APIs that are related to user operation. In the mini project, we first give each user a short tutorial of our API prototype and how to use
our web-based client to fetch data. Then, we show them a target data schema that CONTAM accepts
and ask users to try their best to compose a JSON format file that meets that data schema. During this
mini project, users are allowed to access any resources and help they need. They also can directly ask
questions to us. They must fetch data through our APIs, but not necessarily through our web-based
client. Some users may want to directly use curl commands or write their own scripts to hit our APIs. After a user finishes the mini project, we will conduct a one-to-one interview with that user. We first ask them questions regarding their subjective thoughts on our prototype. These questions
are designed in the guidance of TAM. Then, we let them compare our design with two counterparts
in the BIMserver. More specifically, BIMserver has two ways to query data. The first approach is to
write a piece of query code in Java. Figure 14 shows a code snippet from the BIMserver.org advanced
query demo. It queries the data of doors whose height is over 2 feet from the first floor. The other
way is also based on web interfaces. They integrate BIM Service interface exchange (BIMSie) [63] into
their implementation, which accepts SOAP and JSON as web API call payload. However, BIMserver
itself does not offer RESTful style web services. The BIMSie interfaces abstract library calls rather
than resources. 6.1.1. Theoretical Basis We use the Technology Acceptance Model (TAM) as our theoretical framework [55], which is
one of the most widely used models to check the users’ acceptance to a new information technology. It has been successfully applied to examine how users adopt an information system in many different
areas [56–60]. The validity of the model itself also has been examined quantitatively in a rigorous
approach [61]. The model of TAM and its elements are shown in Figure 13. We can see that there are multiple
variables in this model, ranging from users’ first impressions of a new information technology to
the phase of actual usage. Those variables affect each other and form a directed graph. The two
most important variables of TAM are Perceived Usefulness (PU) and Perceived Ease Of Use (PEOU). They are two major drivers to affect users’ actual usage of the new technology. According to Ajzen
and Fishbein, perceived usefulness is defined as “the degree to which an individual believes that
using a particular system would enhance his or her job performance” [62], and perceived ease of use
is defined as “the degree to which an individual believes that using a particular system would be
free of physical and mental effort” [62]. Different design features and different users’ demographic
backgrounds may cause different degrees of perceived usefulness and ease of use. Sometimes we
categorize those factors as external variables. From Figure 13, we can see that the perceived ease of
use is a very important variable since it affects perceived usefulness. If a user feels a new technology is
easy to use, then he or she is more likely to believe this new technology is useful. Perceived usefulness
and perceived ease of use will together influence users’ overall attitude towards the new information
technology. Users’ intention to use is decided by their perceived usefulness and attitudes. Eventually,
users’ intention to use will lead to their actual usage and acceptance of new technology. 16 of 24 16 of 24 Buildings 2016, 6, 20 External
Variables
Perceived
Usefulness
Perceived
Ease of
Use
Attitude
Towards
Using
Behavioral
Intention
Actual
System
Use
Figure 13. Technology Acceptance Model (TAM) overview. Perceived
Usefulness Attitude
Towards
Using Actual
System
Use External
Variables Figure 13. Technology Acceptance Model (TAM) overview. 6.1.2. Measurements and Study Design Participants. This user study was conducted in the Consortium for Building Energy Innovation
(CBEI) project. We invited the researchers and engineers who are involved in this project via emails or
in person to participate in our user study. These selected experts share some common characteristics. They all work on the some areas related to EEB design and have experience on the BIMserver,
which is an existing open source platform developed by BIMserver.org that facilitates information
exchange in building projects. Some of them have experiences with multiple BIM data management
systems. However, they also have different expertise and roles. Therefore, we use two tags, “analyst”
and “developer”, to denote our participants. Analysts can compose analysis logics, operate some
domain software, build models, and interpret analysis results. Developers can implement new
applications to fulfill those analysis logics, build plugins based on existing platforms (e.g., BIMserver
and OpenStudio), and parse data. In CBEI, they usually work together on a research topic. Two tags
are not exclusive: if a participant does both jobs in their daily work, she or he can have both tags. There are 10 participants in total. Among them, three are analysts, three are developers, and four are
both analysts and developers. Tasks. Our study contains two parts:
user experiments and interviews. First, we ask
users to perform a mini project on our prototype. Second, we interview these experts in a
semi-structured manner. In the mini project, we first give each user a short tutorial of our API prototype and how to use
our web-based client to fetch data. Then, we show them a target data schema that CONTAM accepts
and ask users to try their best to compose a JSON format file that meets that data schema. During this
mini project, users are allowed to access any resources and help they need. They also can directly ask
questions to us. They must fetch data through our APIs, but not necessarily through our web-based
client. Some users may want to directly use curl commands or write their own scripts to hit our APIs. After a user finishes the mini project, we will conduct a one-to-one interview with that user. We first ask them questions regarding their subjective thoughts on our prototype. These questions
are designed in the guidance of TAM. Then, we let them compare our design with two counterparts
in the BIMserver. 6.2. Findings We analyzed the participants’ reactions to the RESTful BIM web service design after experiencing
our prototype. Overall, we found out that a majority of participants (eight out of 10) hold positive
points of view on our design based on the comparison with their previous experience on BIMserver
and other existing BIM data storage systems. They also described many ideas about how this design
can be implemented and applied to other sub-domains of EEB analysis and simulation. We have
categorized their ideas into these TAM concepts: perceived ease of use, perceived usefulness, altitude
towards using, and behavior intention. We also investigated external variables by combining users’
own thoughts and our analysis. 1
@Override
2
public
void
query( IfcModelInterface
model , PrintWriter
out) {
3
out.println("Running␣doors␣example");
4
List <IfcBuildingStorey > stories = model.getAll( IfcBuildingStorey .class );
5
Map <Double , IfcBuildingStorey > orderedStories = new
TreeMap <Double , IfcBuildingStorey >();
6
for ( IfcBuildingStorey
storey : stories) {
7
orderedStories .put(storey. getElevation (), storey );
8
}
9
if ( orderedStories .size () > 1) {
10
IfcBuildingStorey
firstFloor = stories.get (1);
11
for ( IfcRelContainedInSpatialStructure
rel : firstFloor. getContainsElements ()) {
12
for (IfcProduct
product : rel. getRelatedElements ()) {
13
if (product
instanceof
IfcDoor) {
14
ifcDoor
ifcDoor = (IfcDoor)product;
15
if (ifcDoor. getOverallHeight () > 2) {
16
out.println(ifcDoor.getName () + "␣" +
17
ifcDoor. getOverallHeight ());
18 }}}}}} 6.1.2. Measurements and Study Design In addition, all requests are sent via HTTP GET method, no matter if it is adding a
project or it is deleting a user. Table 2 shows a list of BIMsie APIs that are related to user operation. Buildings 2016, 6, 20 17 of 24 17 of 24 1
@Override
2
public
void
query( IfcModelInterface
model , PrintWriter
out) {
3
out.println("Running␣doors␣example");
4
List <IfcBuildingStorey > stories = model.getAll( IfcBuildingStorey .class );
5
Map <Double , IfcBuildingStorey > orderedStories = new
TreeMap <Double , IfcBuildingStorey >();
6
for ( IfcBuildingStorey
storey : stories) {
7
orderedStories .put(storey. getElevation (), storey );
8
}
9
if ( orderedStories .size () > 1) {
10
IfcBuildingStorey
firstFloor = stories.get (1);
11
for ( IfcRelContainedInSpatialStructure
rel : firstFloor. getContainsElements ()) {
12
for (IfcProduct
product : rel. getRelatedElements ()) {
13
if (product
instanceof
IfcDoor) {
14
ifcDoor
ifcDoor = (IfcDoor)product;
15
if (ifcDoor. getOverallHeight () > 2) {
16
out.println(ifcDoor.getName () + "␣" +
17
ifcDoor. getOverallHeight ());
18 }}}}}} Figure 14. An advanced query sample from BIMserver.org (version 1.1 beta, 2012). Table 2. User related methods in the BIMSie design. Table 2. User related methods in the BIMSie design. Table 2. User related methods in the BIMSie design. addUser
setAllowUsersToCreateTopLevelProjects
addUserToExtendedDataSchema
undeleteUser
addUserToProject
removeUserFromProject
changeUserType
userHasCheckinRights
loginUserToken
userHasCheckinRights
getAllAuthorizedUsersOfProject
newUser
getAllCheckoutsByUser
userHasRights
getAllNonAuthorizedProjectsOfUser
deleteUser
getAllRevisionsByUser
getLoggedInUser
getUserByUoid
isAllowUsersToCreateTopLevelProjects
getUserByUserName
isHideUserListForNonAdmin
getUserRelatedLogs
setHideUserListForNonAdmin
getUserSettings
getUsersProjects
getAllUsers
getAllNonAuthorizedUsersOfProject
registerNewUserHandler
unregisterNewUserHandler setAllowUsersToCreateTopLevelProjects
undeleteUser
removeUserFromProject
userHasCheckinRights
userHasCheckinRights
newUser
userHasRights
deleteUser
getLoggedInUser
isAllowUsersToCreateTopLevelProjects
isHideUserListForNonAdmin
setHideUserListForNonAdmin
getUsersProjects
getAllNonAuthorizedUsersOfProject
unregisterNewUserHandler addUser
addUserToExtendedDataSchema
addUserToProject
changeUserType
loginUserToken
getAllAuthorizedUsersOfProject
getAllCheckoutsByUser
getAllNonAuthorizedProjectsOfUser
getAllRevisionsByUser
getUserByUoid
getUserByUserName
getUserRelatedLogs
getUserSettings
getAllUsers
registerNewUserHandler Every interview is recorded and transcribed to texts. The transcripts were analyzed to explore
users’ thoughts related to concepts in the TAM framework. More specifically, we coded a selected
range of key phrases. (e.g., “useful”, “can be used”, “simple”, and “easy to use”). We merge those
phrases and map them into higher-level concepts in the TAM framework. 6.2.2. Perceived Usefulness In general, eight of 10 users indicated that they felt our design is useful. The other two users
held negative opinions on perceived usefulness. Among the people who said RESTful style design is useful, many of them expressed that its
usefulness is brought by its ease of use: I used BIMserver quite a lot, I am very familiar with BIMserver. I have many big project
experiences on BIMserver. I definitely know the difficulties when there are so many APIs which are
just subtly different and so many data objects. If BIMserver or other BIM platforms integrate your
APIs, I prefer to use yours. (Interviewee 2. BIM expert and developer). I used BIMserver quite a lot, I am very familiar with BIMserver. I have many big project
experiences on BIMserver. I definitely know the difficulties when there are so many APIs which are
just subtly different and so many data objects. If BIMserver or other BIM platforms integrate your
APIs, I prefer to use yours. (Interviewee 2. BIM expert and developer). I think its usefulness is two-fold. First, you build it based on some general concepts, not IFC. Second, compared with writing advanced query code (in Java), this method allows you to write
script in any language and just embed a web request. This feature is useful to me. (Interviewee 8. analyst and developer). “It is like we achieve the same goal with less effort, so it is useful and I prefer it (RESTful web
service).” (Interviewee 3. analyst and developer). A user also uses the examples from other domains to support her opinion: A user also uses the examples from other domains to support her opinion: “I believe taking advantage of RESTful web service is a trend. If you see all leading Internet
companies like Facebook or Twitter and check their API documents, you will find that all of their
web services are based on REST. This design style is also suitable for EEB design.” (Interviewee 1,
analyst and developer). “I believe taking advantage of RESTful web service is a trend. If you see all leading Internet
companies like Facebook or Twitter and check their API documents, you will find that all of their
web services are based on REST. This design style is also suitable for EEB design.” (Interviewee 1,
analyst and developer). However, a user also gives a negative comment: “I feel confused when I was seeing your design, BIMserver interface and those query code. I don’t
understand them and don’t know how to use them. I have been doing EEB structure analysis for
several years, I haven’t encountered a case that I need web services or RESTful web services. All I
rely on is GUI (graphical user interface), like BIMserver’s GUI or Revit GUI. I don’t know if it
(the prototype) is easy to others, it is hard to me” (Interviewee 6, analyst) “I feel confused when I was seeing your design, BIMserver interface and those query code. I don’t
understand them and don’t know how to use them. I have been doing EEB structure analysis for
several years, I haven’t encountered a case that I need web services or RESTful web services. All I
rely on is GUI (graphical user interface), like BIMserver’s GUI or Revit GUI. I don’t know if it
(the prototype) is easy to others, it is hard to me” (Interviewee 6, analyst) 6.2.1. Perceived Ease of Use Most of the participants (eight of 10) believe our design is easy to use. Only one participant gives
a negative comment, and another participant feels neutral. Some users believe the most important
advantage of our design is its simplicity: 18 of 24 Buildings 2016, 6, 20 “It is really easy to find the operation I want. There are many duplicated API in BIMsie.”
(Interviewee 2, analyst and developer). “It is really easy to find the operation I want. There are many duplicated API in BIMsie.”
(Interviewee 2, analyst and developer). “I didn’t need to know details of those internal implementation when I was using an
API.”(Interviewee 3, analyst and developer) “It is easier to infer all APIs you need, all the designs stick to one design paradigm. Its readability
and learnability is better.” (Interviewee 9, developer) “It is easier to infer all APIs you need, all the designs stick to one design paradigm. Its readability
and learnability is better.” (Interviewee 9, developer) Some users expressed their perceived ease of use from different perspectives: “RESTful API is not an ad hoc design, there are so many existing open source software, libraries,
and packages libraries to help you handle RESTful things. I might even not need to write my own
code.” (Interviewee 10, developer) “RESTful API is not an ad hoc design, there are so many existing open source software, libraries,
and packages libraries to help you handle RESTful things. I might even not need to write my own
code.” (Interviewee 10, developer) “All solutions for me need additional learning. However, this design is more general, which means
you can find more resources, tutorials, and examples to learn. Even if I need an IT guy to help me,
I don’t need to spend time to teach him too much civil engineering knowledge since RESTful APIs
hide those details” (Interviewee 5, analyst). However, a user also gives a negative comment: 6.2.3. Attitude Towards Using When the users were using our prototypes, they gave various comments and raised some
questions. Generally, eight of 10 users had positive attitudes towards using them. Two users had
neutral attitudes and no one had a negative attitude. Some users said they realized some flaws of existing tools when they were using our prototypes: “All BIMsie interfaces use HTTP GET method, I felt GET is not supposed to be used to update
server side data.” (Interviewee 2, analyst and developer). “I like how your design organizes everything and hides the internal implementation
details. Directly submitting some Java code on a web page may cause some security
risks.”(Interviewee 9, developer) A user indicated that during his usage of our prototype, he can quickly identify where the
problem is according to the HTTP standard status codes, which is inspiring: A user indicated that during his usage of our prototype, he can quickly identify where the
problem is according to the HTTP standard status codes, which is inspiring: “Error codes are clear, you even don’t need to read error messages. I can easily parse them and have
routines to respond”(Interviewee 5, analyst) “Error codes are clear, you even don’t need to read error messages. I can easily parse them and have
routines to respond”(Interviewee 5, analyst) A user who gave a negative comment regarding perceived usefulness held a neutral attitude
here and said the using experience itself is acceptable: A user who gave a negative comment regarding perceived usefulness held a neutral attitude
here and said the using experience itself is acceptable: The
experience
of
using
your
prototype
is
OK. I
don’t
see
many
problems
there. (Interviewee 7, analyst). The
experience
of
using
your
prototype
is
OK. I
don’t
see
many
problems
there. (Interviewee 7, analyst). 6.2.2. Perceived Usefulness For those who felt neutral or negative, they also gave their reasons: “My requirements and workflow are quite fixed. My daily challenges mainly are analysis logic
issues, not related to data exchange. I am familiar with the tools that I am using. I am not sure if I
can get benefits from a new tool” (Interviewee 7, analyst) 19 of 24 19 of 24 Buildings 2016, 6, 20 6.2.4. Behavior Intention When being asked if they would like to use RESTful BIM web service in their future projects if
all functionalities have been fully implemented, seven users said yes, one user said not sure, and two
users said no. The behavior intention to use of some users is based on their perceived usefulness and
attitude towards using them. They gave similar reasons here to support their decisions: “Yes, I think I will use it. And I don’t need to read IFC documents every time.” (Interviewee 8,
analyst and developer) “Yes, I think I will use it. And I don’t need to read IFC documents every time.” (Interviewee 8,
analyst and developer) “If BIMserver integrated RESTful web interfaces, then I would definitely use it.” (Interviewee 3,
analyst and developer). “If BIMserver integrated RESTful web interfaces, then I would definitely use it.” (Interviewee 3,
analyst and developer). A user gave a reason from a more comprehensive perspective: “I think it is developer-friendly and user-friendly at the same time. I will use it to request a service,
but also use it to launch a service. I know there are many great applications in many labs. Few
people know them and many efforts are wasted. They have no incentives to maintain it. If there is a
simpler way to deliver your application over web, then more and more researchers can publish their
works and they can get more feedback and encouragement. The whole community will benefit from
this.” (Interviewee 2, analyst and developer) “I think it is developer-friendly and user-friendly at the same time. I will use it to request a service,
but also use it to launch a service. I know there are many great applications in many labs. Few
people know them and many efforts are wasted. They have no incentives to maintain it. If there is a
simpler way to deliver your application over web, then more and more researchers can publish their
works and they can get more feedback and encouragement. The whole community will benefit from
this.” (Interviewee 2, analyst and developer) One user who had no strong intention to use said he was hesitated because he needs to conside
more factors to make a decision: One user who had no strong intention to use said he was hesitated because he needs to consider
more factors to make a decision: “I got your idea, and it is great. 6.3. Discussion In this subsection, we first analyze the external variables that affect users’ preference towards
our proposal. Then, we discuss our reflections on this study. 6.2.4. Behavior Intention However, a tool that was built based on a great idea might not be
t t
l I
t
if it d
t i
ll
itt
I
t
if it i
t
d t bl
it “I got your idea, and it is great. However, a tool that was built based on a great idea might not be
a great tool. I must see if its document is well-written. I must see if it is mature and stable or it
is still fast changing. Considering if I should adopt a tool is different than talking about an idea”
(Interviewee 7, analyst). The user who had no intention to use it because she believes she does not need it in her
daily work: The user who had no intention to use it because she believes she does not need it in her
aily work: I think it depends on what kind of projects you are doing. I never need a partial model query and
web services. (interviewee 6, analyst). I think it depends on what kind of projects you are doing. I never need a partial model query and
web services. (interviewee 6, analyst). 20 of 24 20 of 24 Buildings 2016, 6, 20 6.3.1. External Variables Table 3 shows users’ feedback and their expertise. We group the rows by user tags. From
the table, we can find that two analysts did not show strong interests in this idea while the other
eight generally support it. By analyzing this preference pattern and interview scripts, we identify
two major external variables. The first variable is the role an expert played in a project. Different roles cause those users to
consider the same issue from different perspectives. The analysts’ focus is more on analysis logics. Their work is to refine their analysis algorithm to better abstract knowledge from existing data. They usually interact with service provider’s APIs indirectly. For those developers and analysts who
also do development jobs, publishing their applications or requesting remote data are challenges they
are facing in their daily work. They understand our motivation and get the point of our idea more
quickly. This is like Google Maps releasing a set of better web services and APIs. Those mobile app
developers will start to use it to refine their applications while the end users might just feel indifferent
to this news. However, those end users still get better user experience in the end. They get benefits in
an indirect way. This result helps us identify our contribution as a developer-oriented solution rather
than an end-user-oriented solution. Instead of additionally lowering the bar of using our design to
make it accessible to more people, making it more developer friendly might be the right direction
to go. The second external variable is the nature of a project. The number of data objects needed by
a project is various. Some projects need a few kinds of discrete data objects which do not relate to
each other. In this case, even if the data size is huge, the query itself is simple. In this scenario,
some interviewees pointed out that every approach works similarly in these kinds of problems. Users cannot gain much benefit by importing a new workflow and design. However, some projects’
data models are very complex including both data schema and the relationship between those data
schema. In this case, a resource-modeling-centric RESTful web service helps both service developers
and application builders tackle the problem by following a top-down approach. If an EEB design
sub-domain happens to have these features, then the users who work in these areas recognize the
benefits of our method. Table 3. 6.3.1. External Variables Interview feedback. Table 3. Interview feedback. User Number
Tags
PEOU
PU
ATU
BIU
1
Analyst & Programmer
2
Analyst & Programmer
3
Analyst & Programmer
8
Analyst & Programmer
–
4
Programmer
9
Programmer
10
Programmer
5
Analyst
6
Analyst
–
7
Analyst
–
–
6.3.3. Limitations and Future Work This study presents some preliminary results collected from 10 interviewees. This research can
be regarded as a pilot study on this direction. Being different to some information systems that can be
operated by normal users, due to the fact that our proposed requirements, designs, and implemented
prototypes are targeting some expert users who work on BIM-enabled EEB design, it is difficult to
recruit a large number of qualified interviewees in a short time. In addition, our current prototype
RESTful APIs just cover the essential resources for multi-zone airflow analysis. There are many other
areas in EEB design to be implemented. In the future, to perform larger and deeper user studies, we need to build plugins or integrate
our API to some existing BIM data platforms and cover more functionalities. Thus, we can expand
our user studies from two directions. First, we can collaborate with some groups and work with
them on a real-world project by using our applications. By attending their daily work, we can use the
ethnography methodology to observe and record how our idea affects their work efficiency. Second,
we can invite more researchers to try our application. Then, we can use questionnaires to collect some
quantitative data to perform an analysis. 6.3.2. Reflections on the Study 6.3.2. Reflections on the Study 6.3.2. Reflections on the Study There is a long debate on when to apply web services and how to choose a proper design style
from those co-existing design paradigms. It should be clear that there is no universal best design,
but relatively more appropriate design to a specific case. Regarding the case we present in this paper, 21 of 24 Buildings 2016, 6, 20 BIM-enabled EEB design has some unique challenges, such as interdisciplinary, multiple participants
and stakeholders, fast evolving community, and deep involvement of researchers, which make its
loose coupling collaboration pattern different from some traditional domains. Those challenges are
exactly what RESTful web services are good at handling. By communicating with the researchers, we
found most of them agree with the points presented in our paper. Even the BIMserver official blog
described the current challenge and future direction of the community: “the commercial use and
increasing complexity of applications that use BIMserver as their base, demand query capabilities
beyond the current plugins. We’ve always waited for the industry (or academic world) to develop
a rich BIM Query language, but time is running out. That is why we decided to remove the
Query-plugin type and build one new query language...and we plan to gradually migrate the current
API to a new one” [64]. Our research is orthogonal to the BIMserver and other BIM data platforms. They can integrate our approach into their implementations. An interface is a contract between service providers and users. The web service interfaces that
look like method calls in a programming language such as “addUserToProject” are actually modeling
on logics. It works better when all parties share a similar background or work closely. However, if
participants belong to different parties and have heterogeneous knowledge backgrounds, the best
contract between them should be the data schema itself, or more precisely “resources”, which is
more stable and more understandable. That is the reason we use RESTful web service which models
interfaces based on resources. Acknowledgments: This research is supported in part by the Consortium for Building Energy Innovation (CBEI),
sponsored by the Department of Energy under the grant DE-EE0004261. 7. Conclusions In this paper, we investigate the advantages of using BIM RESTful web services to support
the EEB lifecycle. We identify the requirements to support the EEB lifecycle from four different
perspectives: information exchange requirements, distributed collaboration requirements, internal
data storage requirements, and partial model query requirements. Three RESTful style designs at
different abstraction levels are proposed and detailed with rationales. We have implemented a set
of RESTful APIs and conducted a pilot user study. In the user study, by comparing with the current
BIMserver partial model query methods, users generally perceive more usefulness, ease of use, and
behavioral intention to use in performing an EEB design task. We conclude that the RESTful BIM web
service is a feasible method to improve the efficacy and efficiency of the EEB lifecycle support and
enhance the BIM data accessibility. Acknowledgments: This research is supported in part by the Consortium for Building Energy Innovation (CBEI),
sponsored by the Department of Energy under the grant DE-EE0004261. 22 of 24 22 of 24 Buildings 2016, 6, 20 Author Contributions: Yufei Jiang and Dinghao Wu conceived the concept and design; Yufei Jiang and Xiao
Liu performed the experiments; Yufei Jiang and Xiao Liu analyzed the data; Yufei Jiang, Xiao Liu, Fangxiao Liu,
Dinghao Wu, and Chimay J. Anumba wrote the manuscript. Conflicts of Interest: The authors declare no conflict of interest. References 1. Jiang, Y.; Ming, J.; Wu, D.; Yen, J.; Mitra, P.; Messner, J.I.; Leicht, R. BIM server requirements to support
the energy efficient building lifecycle. In Proceedings of the 2012 ASCE International Conference on
Computing in Civil Engineering, Clearwater Beach, FL, USA, 17–20 June 2012. 1. Jiang, Y.; Ming, J.; Wu, D.; Yen, J.; Mitra, P.; Messner, J.I.; Leicht, R. BIM server requirements to support
the energy efficient building lifecycle. In Proceedings of the 2012 ASCE International Conference on
Computing in Civil Engineering, Clearwater Beach, FL, USA, 17–20 June 2012. 2. Watson, A. BIM-a driver for change. In Proceedings of the International Conference on Computing in Civil
and Building Engineering, Nottingham University Press, Nottingham, UK, 30 June–2 July 2010. 2. Watson, A. BIM-a driver for change. In Proceedings of the International Conference on Computing in Civil
and Building Engineering, Nottingham University Press, Nottingham, UK, 30 June–2 July 2010. 3. Hirsch, A.; Pless, S.; Guglielmetti, R.; Torcellini, P.A.; Okada, D.; Antia, P. The Role of Modeling When
Designing for Absolute Energy Use Intensity Requirements in a Design-Build Framework; NREL/CP-5500-49067;
National Renewable Energy Laboratory: Golden, CO, USA, 2011. 3. Hirsch, A.; Pless, S.; Guglielmetti, R.; Torcellini, P.A.; Okada, D.; Antia, P. The Role of Modeling When
Designing for Absolute Energy Use Intensity Requirements in a Design-Build Framework; NREL/CP-5500-49067;
National Renewable Energy Laboratory: Golden, CO, USA, 2011. 4. Bazjanac, V. Acquisition of building geometry in the simulation of energy performance. In Proceedings of
the 2001 Building Simulation Conference, Rio de Janeiro, Brazil, 13–15 August 2001. 4. Bazjanac, V. Acquisition of building geometry in the simulation of energy performance. In Proceedings of
the 2001 Building Simulation Conference, Rio de Janeiro, Brazil, 13–15 August 2001. 5. Beetz, J.; van Berlo, L.; de Laat, R.; van den Helm, P. BIMserver.org—An open source IFC model server. In Proceedings of the CIP W78 Conference, Cairo, Egypt, 16–18 November 2010. 6. Beetz, J.; Berlo, V.L.; Laat, D.R.R.; Bonsma, P. Advances in the development and application of an open
source model server for building information. In Proceedings of the CIP W78 Conference, Sophia Antipolis,
France, 26–28 October 2011. 7. Fielding, R.T. Architectural Styles and the Design of Network-Based Software Architectures. Ph.D. Thesis,
University of California, Irvine, CA, USA, 2000. 8. Legner, C.;
Heutschi, R. SOA adoption in practice—Findings from early SOA implementations. In Proceedings of the 2007 European Conference on Information Systems (ECIS 2007), St. Gallen,
Switzerland, 7–9 June 2007. References Automated construction progress measurement using a 4D building information safety: Automatic safety checking of construction models and schedules. Autom. Constr. 2013, 29, 183–195. 23. Kim, C.; Son, H.; Kim, C. Automated construction progress measurement using a 4D building information
model and 3D data. Autom. Constr. 2013, 31, 75–82. 23. Kim, C.; Son, H.; Kim, C. Automated construction progress measurement using a 4D building information
model and 3D data. Autom. Constr. 2013, 31, 75–82. 24. Messner, J.; Anumba, C.; Dubler, C.; Goodman, S.; Kasprzak, C.; Kreider, R.; Leicht, R.; Saluja, C.;
Zikic, N. BIM project execution planning guide and templates:
version 2.0. Available online:
http://bim.psu.edu/Resources/Project/BIM_PxP-V2.1/BIM_PxP_Guide_&_Templates_V2.1.zip
(accessed on 15 July 2015). 25. Maile, T.; Fischer, M.; Bazjanac, V. Building energy performance simulation tools—A life-cycle and
interoperable perspective. Center Integr. Facil. Eng. (CIFE) Working Pap. 2007, 107, 1–49. 26. Bazjanac, V. Impact of the US national building information model standard (NBIMS) on building energy
performance simulation. In Proceedings of the Building Simulation 2007 Conference, Beijing, China,
3–6 September 2007. p
27. Young, N.; Jones, S.; Bernstein, H.M.; Gudgel, J. The Business Value of Bim—Getting Building Information
Modeling to the Bottom Line; McGraw-Hill Construction: Bedford, MA, USA, 2009. 28. Moon, H.J.; Choi, M.S.; Kim, S.K.; Ryu, S.H. Case studies for the evaluation of interoperability
between a BIM based architectural model and building performance analysis programs. In Proceedings
of 12th Conference of International Building Performance Simulation Association, Sydney, Australia,
14–16 November 2011. 29. Jiang, Y.; Yu, N.; Ming, J.; Lee, S.; DeGraw, J.; Yen, J.; Messner, J.I.; Wu, D. Automatic building information
model query generation. J. Inf. Technol. Constr. (ITCon) 2015, 20, 518–535. 30. Yu, N.; Jiang, Y.; Luo, L.; Lee, S.; Jallow, A.; Wu, D.; Messner, J.I.; Leicht, R.M.; Yen, J. Integrating BIMserver
and OpenStudio for energy efficient building. In Proceedings of the 2013 ASCE International Conference
on Computing in Civil Engineering, Los Angeles, CA, USA, 23–25 June 2013. 31. Liu, F.; Jallow, A.K.; Anumba, C.J.; Wu, D. A framework for integrating change management with building
information modeling. In Proceedings of the 15th International Conference on Computing in Civil and
Building Engineering (ICCCBE 2014), ASCE, Orlando, FL, USA, June 23–25 2014. 32. Dubler, C.R.; Messner, J.; Anumba, C.J. Using lean theory to identify waste associated with information
exchanges on a building project. In Proceedings of the Construction Research Congress/ASCE Conference,
Banff, AL, Canada, 8–10 May 2010. 33. Rogers, S. References 9. Michelson, B.M. Event-driven architecture overview. Patricia Seybold Group 2006, 2, doi:10.1571/
bda2-2-06cc. 10. Exforsys. SOA Disadvantage, 2007. Available online: http://www.exforsys.com/tutorials/soa/soa-dis
advantages.html (accessed on 1 August 2015). 11. Associates, J.J.H. Welcome to DOE-2.com. Available online:
http://www.doe2.com/ (accessed on
28 April 2016). 12. O’Donnell, J. SimModel: A domain data model for whole building energy simulation. In Proceedings of
the SimBuild 2011, Sydney, Australia, 14–16 November 2011. 13. Campbell, D.A. Building information modeling: The Web3D application for AEC. In Proceedings of the
Twelfth International Conference on 3D Web Technology, ACM, Umbria, Italy, 15–18 April 2007. 14. BuildingSMARTalliance. United States National Building Information Modeling Standard, 2007. Available online: http://www.wbdg.org/pdfs/NBIMSv1_p1.pdf (accessed on 3 August 2015). 15. Eastman, C.; Eastman, C.M.; Teicholz, P.; Sacks, R. BIM Handbook: A Guide to Building Information Modeling
for Owners, Managers, Designers, Engineers and Contractors; John Wiley & Sons: Hoboken, NJ, USA, 2011. ,
;
,
;
,
;
,
g
f
g
for Owners, Managers, Designers, Engineers and Contractors; John Wiley & Sons: Hoboken, NJ, USA, 2011. 16. Liu, F.; Jallow, A.; Anumba, C.; Wu, D. Building knowledge modeling: Integrating knowledge in BIM. In Proceedings of the 30th International Conference on Applications of IT in the AEC Industry, Beijing,
China, 9–12 October 2013. 17. Smith, D.K.; Tardif, M. Building Information Modeling:
A Strategic Implementation Guide for Architects,
Engineers, Constructors, and Real Estate Asset Managers; John Wiley & Sons: Hoboken, NJ, USA, 2009. 18. Azhar, S. Building information modeling (BIM): Trends, benefits, risks, and challenges for the AEC industry. Leadersh. Manag. Eng. 2011, 11, 241–252. 19. Howell, I.; Batcheler, B. Building information modeling two years later—Huge potential, some success and
several limitations. Laiserin Lett. 2005, 22, 4. 20. Goedert, J.D.; Meadati, P. Integrating construction process documentation into building information
modeling. J. Constr. Eng. Manag. 2008, 134, 509–516. 23 of 24 23 of 24 Buildings 2016, 6, 20 21. Bynum, P.; Issa, R.R.; Olbina, S. Building information modeling in support of sustainable design and
construction. J. Constr. Eng. Manag. 2012, 139, 24–34. 22. Zhang, S.; Teizer, J.; Lee, J.K.; Eastman, C.M.; Venugopal, M. Building information modeling (BIM) and
safety: Automatic safety checking of construction models and schedules. Autom. Constr. 2013, 29, 183–195. 22. Zhang, S.; Teizer, J.; Lee, J.K.; Eastman, C.M.; Venugopal, M. Building information modeling (BIM) and
safety: Automatic safety checking of construction models and schedules. Autom. Constr. 2013, 29, 183–195. 23. Kim, C.; Son, H.; Kim, C. References BIM Implementation Strategies: Two Models or One? Available online: http://old.agc.org/
galleries/conmark/Rogers2008 (accessed on 16 July 2015). 34. Yan, H.;
Damian, P. Benefits and barriers of building information modelling. In Proceedings of
the 12th International Conference on Computing in Civil and Building Engineering, Beijing, China,
16–18 October 2008. 35. Curbera, F.; Duftler, M.; Khalaf, R.; Nagy, W.; Mukhi, N.; Weerawarana, S. Unraveling the Web services
web: An introduction to SOAP, WSDL, and UDDI. IEEE Internet Comput. 2002, 6, 86–93. 36. Hardt, D. The OAuth 2.0 Authorization Framework. Available online:
http://tools.ietf.org/html/
rfc6749.html (accessed on 17 July 2015). 37. Is OAuth Stateless? Can it Work for REST? Available online:
https://looselyconnected.wordpress. com/2010/11/12/is-oauth-stateless-can-it-work-for-rest/ (accessed on 18 July 2010). 38. Liu, D.T.; Xu, X.W. A review of web-based product data management systems. Comput. Ind. 2001,
44, 251–262. 39. Bilgic, T.; Rock, D. Product data management systems: State of the art and the future. In Proceedings
of the 1997 ASME Design Engineering Technical Conferences and Computers in Engineering Conference,
paper No. DETC97/EIM-3720; Sacramento, CA, USA, 14–17 September 1997. 40. Murugesan, S. Understanding Web 2.0. IT Prof. 2007, 9, 34–41. 40. Murugesan, S. Understanding Web 2.0. IT Prof. 2007, 9, 34–41. ugesan, S. Understanding Web 2.0. IT Prof. 2007, 9, 34– 41. Van Berlo, L. BIMServer.org Release 1.3.0 Final. Available online: http://bimserver.org/2014/04/28/
release-1-3-0-final/ (accessed on 20 July 2015). 24 of 24 24 of 24 Buildings 2016, 6, 20 42. Froese, T.; Fischer, M.; Grobler, F.; Ritzenthaler, J.; Yu, K.; Sutherland, S.; Staub, S.; Akinci, B.; Akbas, R.;
Koo, B.; et al. Industry foundation classes for project management-a trial implementation. Electr. J. Inf. Technol. Constr. 1999, 4, 17–36. 43. Adachi, Y. Overview of partial model query language. In Proceedings of the International Society
for Productivity Enhancement Concurrent Engineering Conference (ISPE CE), Madeira, Portugal,
26–30 July 2003. y
44. Camposano, R.; Saunders, L.F.; Tabet, R.M. High-level synthesis from VHDL. IEEE Des. Test Comput. 1991,
8, 43–49. 45. Weise, M.; Katranuschkov, P.; Scherer, R.J. Generalised model subset definition schema. CIB Rep. 2003,
284, 440. 46. Mazairac,
W.;
Beetz,
J. BIMQL—An
open
query
language
for
building
information
models. Adv. Eng. Inform. 2013, 27, 444–456. 47. Daum, S.; Borrmann, A.; Langenhan, C.; Petzold, F. Automated generation of building fingerprints using
a spatio-semantic query language for building information models. In eWork and eBusiness in Architecture,
Engineering and Construction: ECPPM 2014; CRC Press: Vienna, Austria, 2014. 48. Gray, J.; Reuter, A. References Transaction Processing; Morgan Kaufíann Publishers: San Francisco, CA, USA, 1993. 49
Allamaraju
S
Restful Web Services Cookbook:
Solutions for Improving Scalability and Simplicity; 8. Gray, J.; Reuter, A. Transaction Processing; Morgan Kaufíann Publishers: San Francisco, CA, USA, 1993. 9. Allamaraju,
S. Restful Web Services Cookbook:
Solutions for Improving Scalability and Simpli
O’Reilly Media, Inc.: Sebastopol, MA, USA, 2010. 50. DeGraw, J.W.; Macumber, D.; Bahnfleth, W.P. Generation of multizone airflow models from building energy
models. ASHRAE Trans. 2014, 120, 1–8. 51. Emmerich, S.J.; Persily, A.K. US commercial building airtightness requirements and measurements. In Proceedings of the AIVC Conference, Brussels, Belgium, 12–13 October 2011. 52. yaml.org. YAML 1.2. Available online: http://www.yaml.org/ (accessed on 1 March 2016). yaml.org. YAML 1.2. Available online: http://www 53. cURL project. Available online: https://curl.haxx.se/ (accessed on 1 March 2016). 54. SmartBear. Available online: http://swagger.io/ (accessed on 1 March 2016). 54. SmartBear. Available online: http://swagger.io/ (accessed on 1 March 2016). 55. Davis, F.D., Jr. A Technology Acceptance Model for Empirically Testing New End-User Information
Systems: Theory and Results. Ph.D. Thesis, Massachusetts Institute of Technology, Cambridge, MA,
USA, 1986. 56. Adams, D.A.; Nelson, R.R.; Todd, P.A. Perceived usefulness, ease of use, and usage of information
technology: A replication. MIS Q. 1992, 16, 227–247. 57. Chau, P.Y. An empirical assessment of a modified technology acceptance model. J. Manag. Inf. Syst. 1996,
13, 185–204. 58. Chau, P.Y. An empirical investigation on factors affecting the acceptance of CASE by systems developers. Inf. Manag. 1996, 30, 269–280. 59. Davis, F.D. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Q. 1989, 13, 319–340. 60. Davis, F.D.; Bagozzi, R.P.; Warshaw, P.R. User acceptance of computer technology: A comparison of
two theoretical models. Manag. Sci. 1989, 35, 982–1003. 61. Hu, P.J.; Chau, P.Y.; Sheng, O.R.L.; Tam, K.Y. Examining the technology acceptance model using physician
acceptance of telemedicine technology. J. Manag. Inf. Syst. 1999, 16, 91–112. 2. Ajzen, I.; Fishbein, M. Attitude-behavior relations: A theoretical analysis and review of empirical resea
Psychol. Bull. 1977, 84, 888. 63. Berlotti. BIMSie-API. Available online: https://github.com/BuildingSMART/BIMSie-API (accessed on
3 March 2015). 64. Berlotti. New Year’s Resolutions. Available online:
http://bimserver.org/2015/12/23/new-years-
resolutions/ (accessed on 3 March 2015). c⃝2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).
|
https://openalex.org/W1916507059
|
https://www.ntnu.no/ojs/index.php/norepid/article/download/1598/1489
|
English
| null |
Norwegian Childhood Diabetes Registry: Childhood onset diabetes in Norway 1973-2012
|
Norsk epidemiologi
| 2,013
|
cc-by
| 3,565
|
23 23 Norsk Epidemiologi 2013; 23 (1): 23-27 INTRODUCTION in evaluating metabolic control and is the only mea-
sure for which good data are available in terms of its
relationship with later microvascular and macrovascu-
lar complications (9,10). Diabetes mellitus with onset in childhood represents
one of the most frequent chronic diseases in children
and young adults. The incidence of childhood onset
type 1 diabetes in Norway is the third highest in the
world, next to Finland and Sweden (1). The incidence
has been steadily increasing since 1973 until 2009
when a peak in incidence was observed; 36 per
100,000 year (2-4). In Norway, type 1 diabetes is the
second most common chronic disease in children age
0-14 years and it represents more than 98% of all dia-
betes in children younger than 15 years. Yearly app-
roximately 290-330 children are diagnosed with type 1
diabetes in Norway. Establishment and collection of data Norwegian Childhood Diabetes Registry (NCDR) was
established in 2006, and is financed by the South-
Eastern Norway Regional Health Authority. The Wo-
men and Children’s Division, Oslo University Hospital
is responsible for the daily operation, data manage-
ment and security. An informed consent is obtained
from each patient and/or the parents, before the patient
is registered. NCDR has a prospective registration of
data; both incident cases and clinical data from yearly
examinations of all children and adolescents with
diabetes treated in paediatric departments in Norway. In addition, NCDR contains data from two earlier
diabetes registries/research projects; The Norwegian
Diabetes Registry and The Norwegian Childhood Dia-
betes and Quality project. After the merger to NCDR
these two registries stopped existing as independent
research registries/projects. Type 1 diabetes is related to a significant burden to
the patient and to the society because most cases
require lifelong treatment with insulin as well as day-
to-day monitoring and treatment of complications. The
disease has an increased risk of late complications such
as renal disease, blindness, amputations, heart disease
and stroke. Despite advances in treatment, type 1 dia-
betes is still associated with considerable premature
mortality, resulting from acute and chronic com-
plications of diabetes (5-8), even in societies with
unrestricted access to the treatment of diabetes and its
complications. ABSTRACT The Norwegian Childhood Diabetes Registry (NCDR) is a prospective, population-based, nationwide
registry which systematically register all incident cases of childhood diabetes, and systematically monitors
the outcome of diabetes care in children and adolescents. NCDR includes data on childhood onset diabetes
since 1973, and diabetes care outcome since 2001. NCDR was founded with the following objectives: To
improve the diagnostics, classifications and treatment of childhood-onset diabetes, surveillance of incidence
of diabetes in children and adolescents, surveillance of quality of diabetes care in Norwegian paediatric
departments, and to stimulate to research in diabetes. Norwegian Childhood Diabetes Registry:
Childhood onset diabetes in Norway 1973-2012 Torild Skrivarhaug
Norwegian Childhood Diabetes Registry, Department of Paediatric Medicine, Women and Children’s Division
Oslo University Hospital
Correspondence: Torild Skrivarhaug, Department of Paediatric Medicine, Oslo University Hospital, PB 4950, 0424 Oslo
E-mail: torild.skrivarhaug@medisin.uio.no Telephone: +47 22118765 / 23015648 Telefax: +47 22118663 Torild Skrivarhaug
wegian Childhood Diabetes Registry, Department of Paediatric Medicine, Women and Children’s Divisio
Oslo University Hospital Correspondence: Torild Skrivarhaug, Department of Paediatric Medicine, Oslo University Hospital, PB 4950, 0424 Oslo
E-mail: torild.skrivarhaug@medisin.uio.no Telephone: +47 22118765 / 23015648 Telefax: +47 22118663 The NCDR has four main tasks: The NCDR has four main tasks: 1) To collect standardized data from; i) incident cases,
ii) yearly examinations (starting one year after
diabetes onset). 2) To analyse and benchmark the data. 3) To report the data back to the paediatric depart-
ments and to the Health authorities. 4) To maintain and further develop the network for
quality work among paediatric departments to en-
sure equal diabetes treatment throughout Norway. 1
blood samples. Quality indicators are benchmarked. Sampling of data on diabetes care started with a few
paediatric departments in 2001, including all the
paediatric departments from 2008 (Table 1). Data are
collected on a case record form based on the World
Health Organization Basic Information Sheet for child-
ren and adolescents. HbA1c is determined for all parti-
cipants by high-performance liquid chromatography
(Tosoh G7; Tosoh Europe N.V., Belgium), at the same
central DCCT-standardized laboratory (OUS, Aker). New onset cases NCDR includes registration of all new onset cases of
childhood diabetes. Data go back to 1973, collected
retrospectively (2), and prospectively from 1989. From
2002 blood samples are collected together with clinical
data in incident cases. Norway has since 1989 been
one of the 23 EURODIAB centres reporting incidence
of childhood onset type 1 diabetes in Europe (17). The
completeness of ascertainment in NCDR is calculated
to 91% (18). The adherence rate on annual reporting has been
rising from the start in 2001 until 95% in 2010 and
2011. In 2011 all the paediatric departments parti-
cipated with a total of 2567 patients; 98% type 1
diabetes (2520/2567), 0.5% type 2 diabetes (12/2567),
1.1% MODY (28/2567), 0.3% other types of diabetes
(8/2567). The Norwegian Childhood Diabetes Study Group Year
Number of
paediatric
clinics
(n)
Number of
participants
(n)
Completeness of
participants in
eligible patients
(%)
2001
8
470
–
2002
20
963
–
2003
22
1129
–
2004
23
1417
74
2005
25
1664
75
2006
25
1721
76
2007
26
1836
82
2008
26
2059
89
2009
26
2341
93
2010
27
2457
95
2011
27
2567
95 NCDR is based on the Norwegian Childhood Diabetes
Study Group, a network of all paediatricians and dia-
betes nurses treating children and adolescents with dia-
betes in paediatric departments. Since 2008 all paediat-
ric departments in Norway are reporting to NCDR. Aims of the Norwegian Childhood Diabetes Registry Several studies provide clear evidence that better
metabolic control, as measured by a lower HbA1c
level, is associated with fewer and delayed microvas-
cular and macrovascular complications (9-13). Follow-
up data from the Diabetes Control and Complications
Trial (DCCT) indicated that 5-7 years of poor glyce-
mic control, even during adolescence and young adult-
hood, results in an increased risk of complications in
the subsequent 6-10 years (14). HbA1c reflects level
of glycemia over the preceding 4-12 weeks, weighted
toward the most recent four weeks (15). HbA1c moni-
toring has been shown to be the most useful measure NCDR was founded with the following objectives: • Surveillance of incidence of diabetes in children and
adolescents in Norway. • To improve the diagnostics, classifications and
treatment of childhood-onset diabetes. • Surveillance of quality of diabetes care in Norwe-
gian paediatric departments measured by HbA1c,
acute complications, and implementation of
screening procedures for late complications and
associated diseases according to national and
international guidelines for childhood diabetes (16). 24 T. SKRIVARHAUG 10
15
20
25
30
35
40
1973 1989 1999 2000 2006 2008 2009 2010 2011
Incidence
Year
Incidence per 100,000 person years
Figure 1. Incidence of Type 1 diabetes in children 0 -14 years in Norway, during 1973-2011. Norwegian Childhood Diabetes Registry. Incidence per 100,000 person years Figure 1. Incidence of Type 1 diabetes in children 0 -14 years in Norway, during 1973-2011. Norwegian Childhood Diabetes Registry. • To ensure equal treatment to all children and
adolescents with diabetes, and to stimulate to
research in diabetes. Table 1. Yearly registration of quality indicators in the
Norwegian Childhood Diabetes Registry. Increasing num-
bers of paediatric clinics and numbers of participants, and
increasing completeness in eligible patients 2001-2011. Diabetic ketoacidosis at presentation of childhood
onset type 1 diabetes yp
Diabetic ketoacidosis (DKA) is the leading cause of
mortality in children with type 1 diabetes worldwide,
and is associated with increased morbidity and health-
care expenditure. DKA is metabolic derangement cha-
racterised by the triad of hyperglycemia, acidosis and
ketosis that occurs in the presence of very low levels
of effective insulin action. The worldwide variation in
incidence of type 1 diabetes in children has been well
characterised, but there is less evidence concerning the
frequency of DKA at diagnosis of type 1 diabetes. A
study from Europe indicated that the frequency of
DKA at diagnosis varied from 11% to 67% (31). The
first published data from Norway on DKA at presen-
tation of type 1 diabetes, reported that 18.0% of the
children had DKA; 17.0% of the boys and 18.3% of
the girls. Divided in age groups the proportions of
patients with DKA were: age 0-4 years 16.3%, age 5-9
years 11.3%, age 10-14 years 21.9%. The severity of
DKA is categorized by the degree of acidosis (16):
41% of the patients had mild DKA, 35% had moderate
DKA and 23.5% had severe DKA (32). 1) Data at presentation of childhood onset type 1
diabetes Incidence of childhood onset type 1 diabetes
The incidence of type 1 diabetes diagnosed before 15
years has been increasing globally (1). The average in-
crease per year has been 2.5-3.0 worldwide, however,
wide variation in increase of type 1 diabetes has been
well characterized by the DIAMOND project group
(19) and from the EURODIAB study group (17,20). The increase in incidence has not been uniform; seve-
ral countries are showing periods of both less rapid
and more rapid increase (17), and some reports a re-
verse trend (21). There is indication that the incidence
is rising more steeply in some of the low prevalence
countries (1,20). Both Finland and Sweden have repor-
ted a decreasing incidence in recent years (21,22). DATA FROM THE NORWEGIAN CHILDHOOD
DIABETES REGISTRY Type 1 diabetes represents more than 98% of all diabe-
tes registered in NCDR. HbA1c The grand mean HbA1c of all clinics improved signi-
ficantly (p < 0.01) during the first 5-yr period (2001-
2005) from 8.6% to 8.2% (33). In 2011, the grand mean
HbA1c was 8.3% (range 7.6-8.9). Before bench-
marking the different paediatric departments, the mean
HbA1c in each department is adjusted for gender, age
and diabetes duration. In 2011, 97% of the patients had
a centrally analysed HbA1c. Norway has had a steadily increasing incidence from
19.0 per 100,000 years in 1973 until 2009 when a peak
in incidence was observed; 36 per 100,000 years (2-4). The increase in this period has however not been uni-
form; it includes both periods with more rapid increase
and periods with no increase at all (2004-2008). After
2009 the incidence has declined, and in 2011 it was 31
per 100,000 years (23). The decrease has to be inter-
preted with caution, since it is still possible that this is
a temporary phenomenon. Hospitalized with diabetic ketoacidosis p
Diabetic ketoacidosis (DKA) is defined as hospitalized
with pH < 7.3 or bicarbonate < 15 mmol/l. In 2011,
4.8% of the patients were hospitalized with DKA,
0.6% had more than one event. Stratified into gender;
4.7% of the males (0.5% > 1 event) and 5.1% (0.7% >
1 event) of the females were hospitalized with DKA
(34). It is indicated that the increasing incidence of child-
hood type 1 diabetes is not attributable to a global in-
crease in disease incidence, but rather to earlier mani-
festation (29,30). It is suggested that genetic suscepti-
bility to develop type 1 diabetes has decreased over
time due to an increasing environmental pressure and
that this results in a higher disease progression rate in
younger individuals. Norway has no complete registra- Nephropathy screening Annual data on diabetes care NCDR also includes prospectively collected, yearly,
standardized clinical data on diabetes care, including The Norwegian paediatric departments are anonymous
benchmarked for the following quality indicators; 25 NORWEGIAN CHILDHOOD DIABETES REGISTRY tion of incident cases with type 1 diabetes after the age
of 15 years, and cannot contribute with data to this
discussion. HbA1c, severe hypoglycemia, hospitalization for dia-
betes ketoacidosis, completed screening of late com-
plications (retinopathy, nephropathy, blood pressure,
lipids), completed screening for associated diseases
(celiac disease, hypothyroidism, hyperthyroidism). Screening for late complications and associated disea-
ses is done according to ISPAD consensus guidelines
(16). Diabetic ketoacidosis at presentation of childhood
onset type 1 diabetes Severe hypoglycemia The term severe hypoglycemia is defined as episodes
of hypoglycemia were the patient is semiconscious or
unconscious with or without seizures (16). In 2011,
severe hypoglycemia was reported in 6.3% of the pa-
tients, 1.3% had more than one event. Stratified into
gender; 6.5% of the males (1.6% > 1 event), and 6.3%
of the females (1.0% > 1 event), had experienced se-
vere hypoglycaemia during 2011 (34). In childhood onset type 1 diabetes, the male-to-
female ratio has been reported to vary with the inci-
dence level of the population under study. A slight
excess of male patients was more often noted in high
incidence countries (22,24). Also in Norway there is a
slight excess of boys. Several European studies have
suggested that the increase in incidence of childhood
onset type 1 diabetes, in relative terms, are greatest
among children aged 0-4 years (25-28). This has never
been the case in Norway, where the highest increase
always has been in the age group 10-14 years. Retinopathy screening The retinopathy screening includes fundus examina-
tion (fundoscopi) in mydriasis (after pupil dilation)
performed by experienced local ophthalmologist. In
2011, 41% of the patients had retinopathy screening. 68% were screened accordingly to ISPAD consensus
guidelines (16). No one had laser treatment. 0.3% had
non proliferative diabetes retinopathy (23). Intensified insulin treatment Intensified insulin treatment refers to either using insu-
lin pump or four or more insulin injections per day. The insulin treatment has changed during the years. In
2001, 9% of the youths in NCDR used insulin pump
and 33% were on mix insulin. In 2011, 59% used insu- ANNUAL MEETING IN NORWEGIAN
CHILDHOOD DIABETES REGISTRY Difference in outcome among treatment centres
revealed by the Benchmarking centre report, provide a
basis for discussion. Likewise has the screening for
late complications and associated diseases according to
international guidelines (16), improved following dis-
cussions at the annual NCDR meeting, leading to
changes in screening assessments. This illustrates the
value of a national quality circle in the NCDR. Smoking A significant decline in the incidence of smoking
among patients above 15 years was observed from
2001 (18%) to 2005 (5%) (p < 0.001) (33). Since 2006
the incidence has been rather stable with 6% in 2010
and 2011 (23). The topic of how important it is to
encourage young people to stop smoking or never start
smoking has been discussed in NCDR annual
meetings. Nephropathy screening Nephropathy screening includes measurement of uri-
nary albumin excretion rate in either 24h or overnight 26 T. SKRIVARHAUG T. SKRIVARHAUG lin pump, only 0.3% was on mixed insulin. Less than
5% use one to two daily insulin injections (23). urine collection or by urine albumin/creatinine ratio. Urine-albumin measurements are done at a local labo-
ratory by standard sensitive immunological methods. In 2011, urine samples were collected from 86% of the
patients. 96% were screened accordingly to ISPAD
consensus guidelines (16). Persistent microalbumin-
uria was found in 0.6% of the patients, 1.0% had
albuminuria (23). REFERENCES 27 NORWEGIAN CHILDHOOD DIABETES REGISTRY 15. Diabetes Research in Children Network Study Group. Comparison of fingerstick hemoglobin A1c levels
assayed by DCA 2000 with the DCCT/EDICT central laboratory assay: results of a Diabetes Research in
Children Network (DirecNet) Study. Pediatr Diabetes 2005; 6: 13-6. 16. Hanas R, Donaghue KC, Klingensmith G, Swift PG. ISPAD clinical practice consensus guidelines 2
pendium. Introduction. Pediatr Diabetes 2009; 10 (Suppl 12): 1-2. 17. Patterson CC, Gyurus E, Rosenbauer J, Cinek O, Neu A, Schober E, et al. Trends in childhood type 1 diabetes
incidence in Europe during 1989-2008: evidence of non-uniformity over time in rates of increase. Diabetologia 2012; 55 (8): 2142-7. 18. Skrivarhaug T, Stene LC, Strøm H, Drivvoll AK, Njølstad PR, Joner G. Increasing incidence of childhood
onset type 1 diabetes in Norway. Diabetologia 2010; 53 (Suppl 1): 142. DIAMOND Project Group. Incidence and trends of childhood Type 1 diabetes worldwide 1990-1999. et Med 2006; 23: 857-66. 20. Patterson CC, Dahlquist GG, Gyurus E, Green A, Soltesz G. Incidence trends for childhood type 1 diabetes in
Europe during 1989-2003 and predicted new cases 2005-20: a multicentre prospective registration study. Lancet 2009; 373 (9680): 2027-33. (
)
21. Berhan Y, Waernbaum I, Lind T, Mollsten A, Dahlquist G. Thirty years of prospective nationwide incidence
of childhood type 1 diabetes: the accelerating increase by time tends to level off in Sweden. Diabetes 2011;
60 (2): 577-81. ( )
22. Harjutsalo V, Sund R, Knip M, Groop PH. Changes over time in type 1 diabetes. Diabetologia 2012; 55
(Suppl 1): S105. 23. Skrivarhaug T. Annual report 2011, The Norwegian Childhood Diabetes Registry, www.barnediabetes.no. 2012. 24. Karvonen M, Pitkaniemi M, Pitkaniemi J, Kohtamaki K, Tajima N, Tuomilehto J. Sex difference in the
incidence of insulin-dependent diabetes mellitus: an analysis of the recent epidemiological data. World Health
Organization DIAMOND Project Group. Diabetes Metab Rev 1997; 13 (4): 275-91. g
j
p
( )
25. Tuomilehto J, Virtala E, Karvonen M, Lounamaa R, Pitkaniemi J, Reunanen A, et al. Increase in incidence of
insulin-dependent diabetes mellitus among children in Finland. Int J Epidemiol 1995; 24 (5): 984-92. 26. Gardner SG, Bingley PJ, Sawtell PA, Weeks S, Gale EA. Rising incidence of insulin dependent diabetes in
children aged under 5 years in the Oxford region: time trend analysis. The Bart's-Oxford Study Group. BMJ
1997; 315 (7110): 713-7. 27. Dahlquist G, Mustonen L. REFERENCES 2. Joner G, Sovik O. Increasing incidence of diabetes mellitus in Norwegian children 0-14 years of age 1973-
1982. Diabetologia 1989; 32 (2): 79-83. 3. Aamodt G, Stene LC, Njolstad PR, Sovik O, Joner G. Spatiotemporal trends and age-period-cohort modeling
of the incidence of type 1 diabetes among children aged <15 years in Norway 1973-1982 and 1989-2003. Diabetes Care 2007; 30 (4): 884-9. 4. Skrivarhaug T. Annual report 2009, The Norwegian Childhood Diabetes Registry, www.barnediabetes.no. 2010. 5. Skrivarhaug T, Bangstad HJ, Stene LC, Sandvik L, Hanssen KF, Joner G. Long-term mortality in a nation-
wide cohort of childhood-onset type 1 diabetic patients in Norway. Diabetologia 2006; 49 (2): 298-305. 6. Patterson CC, Dahlquist G, Harjutsalo V, Joner G, Feltbower RG, Svensson J, et al. Early mortality in
EURODIAB population-based cohorts of type 1 diabetes diagnosed in childhood since 1989. Diabetologia
2007; 50 (12): 2439-42. (
)
7. Groop PH, Thomas MC, Moran JL, Waden J, Thorn LM, Makinen VP, et al. The presence and severity of
chronic kidney disease predicts all-cause mortality in type 1 diabetes. Diabetes 2009; 58 (7): 1651-8. 8. Harjutsalo V, Forsblom C, Groop PH. Time trends in mortality in patients with type 1 diabetes: nationwide
population based cohort study. BMJ 2011; 343: d5364. 9. The Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes
on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N
Engl J Med 1993; 329: 977-86. 10. The Diabetes Control and Complications Trial Research Group. Effect of intensive diabetes treatment on the
development and progression of long-term complications in adolescents with insulin-dependent diabetes
mellitus. J Pediatr 1994; 125: 177-88. 11. Mohsin F, Craig ME, Cusumano J, Chan AK, Hing S, Lee JW, et al. Discordant trends in microvascular com-
plications in adolescents with type 1 diabetes from 1990 to 2002. Diabetes Care 2005; 28 (8): 1974-80. 12. Nathan DM, Cleary PA, Backlund JY, Genuth SM, Lachin JM, Orchard TJ, et al. Intensive diabetes
and cardiovascular disease in patients with type 1 diabetes. N Engl J Med 2005; 353 (25): 2643-53. p
yp
g
(
)
13. Dahl-Jorgensen K. Near-normoglycemia and late diabetic complications. The Oslo Study. Acta Endocrinol
Suppl 1987; 284: 1-38. 14. The Epidemiology of Diabetes Interventions and Complications (EDICT) Study. Sustained effect of intensive
treatment of type 1 diabetes mellitus on development and progression of diabetic nephropathy. JAMA 2003;
290: 2159-67. REFERENCES Analysis of 20 years of prospective registration of childhood onset diabetes time
trends and birth cohort effects. Swedish Childhood Diabetes Study Group. Acta Paediatr 2000; 89 (10): 1231-
7. 28. Harjutsalo V, Sjoberg L, Tuomilehto J. Time trends in the incidence of type 1 diabetes in Finnish children: a
cohort study. Lancet 2008; 371 (9626): 1777-82. 29. Weets I, De Leeuw IH, Du Caju MV, Rooman R, Keymeulen B, Mathieu C, et al. The incidence of type 1
diabetes in the age group 0-39 years has not increased in Antwerp (Belgium) between 1989 and 2000: evi-
dence for earlier disease manifestation. Diabetes Care 2002; 25 (5): 840-6. 30. Pundziute-Lycka A, Dahlquist G, Nystrom L, Arnqvist H, Bjork E, Blohme G, et al. The incidence of Type I
diabetes has not increased but shifted to a younger age at diagnosis in the 0-34 years group in Sweden 1983-
1998. Diabetologia 2002; 45 (6): 783-91. 31. Levy-Marchal C, Patterson CC, Green A. Geographical variation of presentation at diagnosis of type I
diabetes in children: the EURODIAB study. Diabetologia 2001; 44 (Suppl 3): B75-B80. 32. Skrivarhaug T, Drivvoll AK, Kummernes SJ, Joner G. Diabetic ketoacidosis at onset in children and adoles-
cents with type 1 diabetes in Norway – a nationwide population-based cohort study. Pediatr Diabetes 2012;
13 (Suppl 17): 49. 33. Margeirsdottir HD, Larsen JR, Kummernes SJ, Brunborg C, Dahl-Jorgensen K. The establishment of a new
national network leads to quality improvement in childhood diabetes: implementation of the ISPAD Guide-
lines. Pediatr Diabetes 2010; 11 (2): 88-95. 34. Skrivarhaug T, Kummernes SJ, Drivvoll AK, Dahl-Jørgensen K. Severe hypoglycemia and diabetic ketoaci-
dosis among children with type 1 diabetes in a population based, nationwide cohort. Pediatr Diabetes 2012;
13 (Suppl 17): 50.
|
https://openalex.org/W4225156892
|
https://translational-medicine.biomedcentral.com/track/pdf/10.1186/s12967-022-03394-8
|
English
| null |
Identification of invasion-metastasis associated MiRNAs in gallbladder cancer by bioinformatics and experimental validation
|
Journal of translational medicine
| 2,022
|
cc-by
| 8,534
|
Cao et al. Journal of Translational Medicine (2022) 20:188
https://doi.org/10.1186/s12967-022-03394-8 Cao et al. Journal of Translational Medicine (2022) 20:188
https://doi.org/10.1186/s12967-022-03394-8 Journal of
Translational Medicine Open Access Abstract Background: Recent studies exploring the roles of invasion-metastasis associated miRNAs in gallbladder cancer
(GBC) are limited. In the study, we aimed to identify the invasion-metastasis associated miRNAs in GBC by bioinformat-
ics and experimental validation. Methods: MiRNAs of different expression were identified by comparing GBC tumor samples with different survival
from Gene Expression Omnibus database. MiRTarBase was used for identifying the potential target genes of miRNAs. Then, we performed Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway
enrichment analysis. And miRNA-gene and protein–protein interaction (PPI) network were constructed for hub genes
evaluation. We further explored and compared miR-642a-3p and miR-145-5p expression in both The Cancer Genome
Atlas database and our hospital data. Finally, quantitative real-time PCR, wound healing assay, and Transwell assay
were conducted to validate the invasion-metastasis associated miRNAs in GBC. Results: In GSE104165 database, 25 up-regulated and 97 down-regulated miRNAs were detected with significantly
different expression in GBC tumor samples. Then, 477 potential target genes were identified from the 2 most up-
regulated miRNAs (miR-4430 and miR-642a-3p) and 268 genes from the 2 most down-regulated miRNAs (miR-451a
and miR-145-5p). After GO and KEGG analysis, mTOR and PI3K-Akt signaling pathways were found associated with the
potential target genes. Based on PPI network, the top 10 highest degree hub nodes were selected for hub genes. Fur-
thermore, the miRNA-hub gene network showed significant miR-642a-3p up-regulation and miR-145-5p down-reg-
ulation in both GBC tissues and cell lines. In the experimental validation, miR-145-5p up-regulation and miR-642a-3p
down-regulation were confirmed to suppress GBC invasion and metastasis. Conclusions: MiR-642a-3p and miR-145-5p were identified as invasion-metastasis associated miRNAs via bioinfor-
matics and experimental validation, and both up-regulation of miR-642a-3p and down-regulation of miR-145-5p
would be served as novel treatment options for GBC in the future. Keywords: Gallbladder cancer, MiRNA, Invasion, Metastasis, Bioinformatics Identification of invasion‑metastasis
associated MiRNAs in gallbladder cancer
by bioinformatics and experimental validation Jiasheng Cao1,2†, Huijiang Shao2,3†, Jiahao Hu1,2, Renan Jin1,2, Anyun Feng4, Bin Zhang1,2, Shijie Li1,2,
Tianen Chen1,2, Sarun Jeungpanich2, Win Topatana2, Yitong Tian2, Ziyi Lu2, Xiujun Cai1,2* and Mingyu Chen1,2* © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visithttp://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Background Gallbladder cancer (GBC) is the most common type of
biliary tract cancer (BTC), which has a high mortality and
a poor prognosis [1–4]. According to the 8th American
Joint Committee on Cancer (AJCC) guideline [5], surgi-
cal resection is the best potential treatment for GBC at *Correspondence: srrsh_cxj@zju.edu.cn; mychen@zju.edu.cn
†Jiasheng Cao and Huijiang Shao contributed equally
1 Department of General Surgery, Sir Run‑Run Shaw Hospital, Zhejiang
University, No. 3 East Qingchun Road, Hangzhou 310016, Zhejiang
Province, China
Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visithttp://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Page 2 of 14 Cao et al. Journal of Translational Medicine (2022) 20:188 Cao et al. Journal of Translational Medicine (2022) 20:188 Screening miRNAs of different expression
The miRNA expression data profile in txt format were
preprocessed. Limma package was applied to calculate
the miRNA of different expression from the Bioconduc-
tor package (http://www.bioconductor.org/). MiRNAs
of different expression were found by comparing GBC
tumor samples with long survival and short survival in
R Studio (Version 4.0.4). Different expression was deter-
mined as significant when P-value < 0.05 and |fold change
(FC)|> 1. Screening miRNAs of different expression an early stage, while chemotherapy/radiotherapy/immu-
notherapy/targeted therapy is recommended for GBC
at an advanced stage. Patients with advanced GBC have
limited response to treatment and poor prognosis due to
the highly invasive and metastatic characteristics of GBC,
including local tumor growth, hepatic invasion, and
lymph nodes metastasis [6, 7]. Several studies reported
the crucial role of multiple biological processes in can-
cer invasion and metastasis, such as circulating cancer
cells, immune evasion [8], epithelial-mesenchymal tran-
sition (EMT) [9, 10], and cancer stem cells [11], while the
multi-step process and the molecular mechanism of GBC
remained unclear. Therefore, it is urgent to explore the
novel biomarkers associated with invasion and metastasis
in GBC patients to improve the prognosis. Data collection Datasets of miRNA expression of GBC tissues with
survival data from the National Center for Biotech-
nology
Information
(NCBI)
(https://www.ncbi.nlm.
nih.gov/geo) online Gene Expression Omnibus (GEO)
database were included in the study. The search strat-
egy was “gallbladder cancer” [MeSH Terms] AND
“miRNA” [MeSH Terms]. After screening the datasets,
the GSE104165 dataset, which contained 40 human
GBC tumor samples and 8 normal gallbladder sam-
ples with survival data, was selected for further analy-
sis. The sample platform was GPL18402 Agilent-046064
Unrestricted_Human_miRNA_V19.0_Microarray. Target genes prediction and gene ontology (GO) analysis
MiRTarBase
(http://mirtarbase.mbc.nctu.edu.tw/php/
index.php), an online miRNA-target interactions data-
base, was used for identifying the potential target genes
of the up-regulated or down-regulated miRNAs. The
data were further validated by microarray, western blot,
reporter assay, and next-generation sequencing experi-
ments. GO analysis was extensively used for the deter-
mination of unique biological attributes of genes, gene
products, and sequences, such as molecular function,
cell components, and biological process [17]. Enrichr
(http://amp.pharm.mssm.edu/Enrichr/), a 2016 updated
webserver with a comprehensive enrichment analysis of
the gene set, was introduced to manipulate enrichment
analysis of the predicted target genes pathway and func-
tional annotation, which was further analyzed for GO
and Kyoto Encyclopedia of Genes and Genomes (KEGG)
pathways [18]. MiRNAs, the small single-stranded non-coding RNAs
(ranging from 18 to 25 bp in length), are critical in tumor
development, such as invasion, migration, and metas-
tasis [12, 13]. Different expression of miRNAs has been
detected in various cancers. For example, miR-15a/16-1
could attenuate immunosuppression and represent a
potential immunotherapy against hepatocellular carci-
noma [14], meanwhile, miR-124-3p could promote cell
proliferation in intrahepatic cholangiocarcinoma [15]. Additionally, by modulating STAT3 signaling, miR-
124-3p was identified to suppress programmed cell
death-ligand 1 (PD-L1) expression and inhibit tumo-
rigenesis of colorectal cancer [16]. Nevertheless, as a
highly aggressive cancer, studies exploring the potential
roles of invasion-metastasis associated miRNAs in GBC
are limited. MiRNA‑gene and protein–protein interaction (PPI) network
construction In the study, we aimed to conduct bioinformatic analy-
sis to identify the invasion-metastasis associated miR-
NAs in GBC, and experimental validation were further
performed to confirm the role of miRNAs. In turn, the
study would provide potential effective treatment for
improving the prognosis of GBC patients. Tools were used to determine the interactions between
genes or proteins to establish the miRNA-gene and PPI
network. The target genes were first entered to assess
functional associations in the STRING database (http://
string-db.org) [19]. Next, the connectivity degree of PPI
network was calculated in Cytoscape software (Version
3.6.0) for further hub genes evaluation. Finally, a miRNA-
hub gene regulatory network was generated. Wound healing assay and transwell assay Wound healing assay and transwell assay Wound healing assay and transwell assay The interval between surgery and death was defined as
OS. The follow-up period was defined as the duration
between surgery and recurrence or death. The final fol-
low-up was in December 2019. g
y
y
After transfection, 5 × 104 GBC cells were seeded in ibidi
Culture-Insert (ibidi GmbH, Martinsried, Germany) on
a 12-well plate. The Culture-Insert was gently removed
after appropriate cell attachment. Then the cells could
migrate into the wound area. Photomicrographs were
taken with a microscope at 0 h and 48 h after wound-
ing, respectively. The cell invasion assay was performed
by 24-well Transwell chambers (Corning, USA). Next,
1 × 105 cells were evenly suspended in a 200 µl serum-
free medium and added to the up-layer inserts after
transfection. 600 µl of medium with 20% FBS was then
added to the low-layer compartment for chemoattract-
ant. The cells on the membrane top surface were excised
using a cotton-wool bud, whereas the cells on the bottom
surface were fixed with 4% paraformaldehyde and stained
with 0.1% crystal violet after 24-h incubation. Five ran-
dom visual fields in each chamber were captured and
counted under a microscope (Zeiss, German). Validation of potential target genes using TCGA database
Based on short and long survival of GBC patients in
TCGA database, the related mRNA expression (SYK,
SH3GL1, CDKN1A, MYC, VEGFA, and EGFR) were com-
pared between two groups. Moreover, the expression of
some potential mRNA (CDKN1A, MYC, VEGFA) was
further compared based on different T stages, which rep-
resented tumor invasion. Statistical analysis Unpaired tailed student’s t tests, Kaplan–Meier sur-
vival analysis, and log-rank tests were used for statistical
analyses with GraphPad Prism 8 (GraphPad Software,
Inc., La Jolla, CA, USA). The results were shown as
mean ± standard deviation (SD). P < 0.05 was considered
statistically significant. Immunohistochemistry (IHC) Immunohistochemistry (IHC)
IHC was further conducted to indirectly validate the
potential target genes expression of miR-642a-3p and
miR-145-5p. Briefly, 4-μm sections from tissue blocks
of the included GBC patients in our hospital were taken
on coated slides. Then the slides were deparaffinized,
hydrated, and blocked with hydrogen peroxide. The
appropriate buffer for each antibody was utilized for
heat induced antigen retrieval. The slides were incubated
with primary antibody, following with secondary anti-
body (anti-SYK antibody, anti-SH3GL1 antibody, anti-
CDKN1A antibody, anti-MYC antibody, anti-VEGFA
antibody, anti-EGFR antibody). Diaminobenzidine (DAB)
chromogen and haematoxylin were used for staining and
counterstaining, respectively. RNA extraction and quantitative real‑time PCR (qRT‑PCR)
TRIzol reagent (Ambion, USA) was used to extract
total RNA from cells, and miRNeasy FFPE kits (Qiagen,
USA) were used to isolate the FFPE tissues based on the
manufacturer’s protocols. All-in-One™ miRNA qRT-
PCR Detection Kit (GeneCopoeia, USA) or Hifair® II
1st Strand cDNA Synthesis SuperMix (Yeasen, China)
were used to make complementary DNA from 1 μg of
RNA. qRT-PCR was performed using a LightCycler® 480
(Roche Molecular Systems, Inc., USA). The primers used
for miRNA qRT-PCR were shown in Additional file 1:
Table S1, and 5s small nuclear RNA was the miRNA
assays internal control. Based on the cycle thresholds
(CT), qRT-PCR results were examined and calculated as
the relative RNA levels. The FC in RNA levels between
each sample was examined by the 2−ΔΔCT method. Cell culture and tissue collection Human GBC cell line GBC-SD and human intrahepatic
biliary epithelial cells (HiBEC) were purchased from a
cell bank (Chinese Academy of Sciences). The SGC-996
cell line was provided by Prof. Anonymize (Blinded Per
Author Guidelines). The GBC-SD and SGC-996 cells
were given 25 units/ml of penicillin, 25 g/ml of strepto-
mycin, and 10% fetal bovine serum (FBS) (v/v) in a 5%
(v/v) CO2 humidified incubator at 37 °C, and they were
cultivated in DMEM and RPMI-1640, respectively. Our
hospital provided the GBC tumor tissues and matched
noncancerous tissues. The cancer genome atlas (TCGA) database and survival
data of GBC patientsh The BTC miRNA-sequencing data and clinical infor-
mation were retrieved from TCGA database (https://
portal.gdc.cancer.gov) to explore miR-642a-3p and miR-
145-5p expression levels in BTC tissues. The data of
overall survival (OS) were matched with miR-642a-3p
and miR-145-5p expression values. A total of 41 GBC
specimens were randomly selected from curative resec-
tion GBC patients retrospectively in Sir Run-Run Shaw
Hospital (SRRSH), Hangzhou, Zhejiang Province, China. Page 3 of 14 Cao et al. Journal of Translational Medicine (2022) 20:188 Cao et al. Journal of Translational Medicine (2022) 20:188 Cao et al. Journal of Translational Medicine (2022) 20:188 Fig. 1 MiRNAs clustering of different expression. A Volcano plots of differentially expressed miRNAs. Green data points indicated down-regulated
miRNAs and red data points indicated up-regulated miRNAs. The differences was set as |log FC|> 1. Black dots indicated miRNAs that were not
differentially expressed. B The heatmap of differentially expressed miRNAs. Log FC of the genes was represented by the top-right legend. FC fold
change
(See figure on next page.) MiRNA‑hub and PPI network construction MiRNA‑hub and PPI network construction
STRING online database was applied to generate a PPI
network, based on the 4 selected miRNAs (top 2 most
up-regulated and top 2 most down-regulated) with dif-
ferent expression as well as potential target genes. The
top 10 hub genes were screened out according to node
degrees (Table 3). The hub genes for the up-regulated
miRNAs were identified as MARPK1, PTEN, SYK,
TAF1, CDKN1A, POLR2E, PRKCA, CDK4, SH3GL1,
and DNAJC10. Meanwhile, the hub genes for the down-
regulated miRNAs were AKT1, VEGFA, MAPK1, EGFR,
MYC, IL6, BCL2, ACTB, SMAD4, ESR1. GO functional annotation analysis Based on the potential target genes, 3 GO functional anno-
tation analysis categories involving biological process (BP),
cellular component (CC), and molecular function (MF)
were conducted. The followings were enriched GO func-
tions for the target genes of up-regulated miRNA: leu-
kocyte degranulation, regulation of homotypic cell–cell
adhesion, and centromeric sister chromatid cohesion in
BP category; nucleolus, 90S pre-ribosome and nuclear
origin of replication recognition complex in CC category;
and phosphatidylinositol-5-phosphate binding, DNA rep-
lication origin binding and poly in MF category (Fig. 2A–
C). The enriched GO functions for the 2 down-regulated
miRNAs target genes were shown in Fig. 2D–F, involving
the transcription from RNA polymerase II (RNAP II) pro-
moter positive regulation, transcription positive reregula-
tion, DNA-templated, transforming growth factor beta2
production in BP category regulation; nuclear chromatin,
chromatin and protein kinase complex in CC category; and
RNAP II transcription regulatory region sequence-specific
binding, transcriptional activator activity, RNAP II core
promoter proximal region sequence-specific binding and
interleukin-6 receptor binding in MF category. Cytoscape software was utilized to construct the
miRNA-hub gene network. We found that 6 hub genes
(CDK4, DNAJC10, PRKCA, TAF1, POLR2E, and PTEN)
could be regulated by up-regulated miR-4430, and 3
hub genes (SYK, SH3GL1, and CDKN1A) could be reg-
ulated by miR-642a-3p (Fig. 4A). In addition, miR-451a
could potentially target 7 genes (SMAD4, ESR1, BCL2,
MARPK1, IL6, MYC, and ACTB), and 3 hub genes (MYC,
VEGFA, and EGFR) could be regulated by miR-145-5p
(Fig. 4B). Validation of potential target genes using TCGA database
Next, we evaluated the potential targets of hub genes
with the TCGA database. Based on short and long sur-
vival, SYK and SH3GL1, but CDKN1A, had a similar
trend of overall survival like miR-642a-3p (Fig. 4C). The
further analysis showed CDKN1A had a similar result
with miR-642a-3p based on different T stages, which rep-
resented tumor invasion (Additional file 1: Fig. S2A). In
the meantime, MYC and VEGFA, like miR-145-5p, were
related to T stages (invasion-related) (Additional file 1:
Fig. S2B-C; Fig. 4D). Enrichment analysis in KEGG pathway We subsequently performed a KEGG pathway enrichment
analysis to investigate the target genes enriched pathways. As shown in Fig. 3A, the enriched KEGG pathways for
up-regulation were in HTLV-1 infection, MAPK signal-
ing pathway, and viral carcinogenesis. The enriched KEGG
for down-regulated miRNAs included proteoglycans and
microRNA in cancer, and the PI3K-Akt signaling pathway
(Fig. 3B). mTOR pathway plays the central role in tumor Prognostic roles of miR‑642a‑3p and miR‑145‑5p
and validation of predicted hub genes using TCGA
database and SRRSH database Cell transfection MiRNA array GSE104165 was retrieved from the GEO
dataset (Additional file 1: Fig. S1). The volcano plot and
heatmap with miRNAs clustering of different expres-
sion were portrayed in Fig. 1. A total of 122 miRNAs in
GBC tumor samples were identified to be differentially Hsa-miRNA mimics and cognate negative control RNAs
were obtained from RiBo (Guangzhou, China) and trans-
fected into GBC cell lines at the concentration of 50 nM
using Lipofectamine™ 3000 (Invitrogen, USA) based on
the user instructions. Page 4 of 14 Cao et al. Journal of Translational Medicine (2022) 20:188 expressed, among which the expression of 25 miRNAs
was found to be significantly up-regulated and 97 miRNAs
were down-regulated. The top 10 most up-regulated and
down-regulated miRNAs were summarized in Tables 1 and
2, respectively. Then, MiRTarBase was applied to identify
the possible target genes of the up- and down-regulated
miRNAs. Notably, the 3rd up-regulated miRNA, miR-
3676-5p, did not successfully predict target genes in the
database, thus the top 2 miRNAs, including miR-4430 and
miR-642-3p, were selected for further analysis. Similarly,
miR-451a and miR-145-5p were the top 2 down-regulated
miRNAs. MiRTarBase was used to identify the 477 poten-
tial target genes from the 2 most up-regulated miRNAs and
268 genes from the 2 most down-regulated miRNAs. progression of GBC [20, 21]. Therefore, 5 genes (AKT1S1,
PTEN, MAPK1, PIK3R2, PRKCA) of up-regulated miRNAs
and 13 genes (CAB39, IRS1, BRAF, TSC1, HIF1A, VEGFA,
IKBKB, RPS6KA3, RRAGC, RPS6KB1, AKT1, MAPK1,
EIF4E) of down-regulated miRNAs, which were involved in
mTOR pathway and interacted with the selected miRNAs,
may be associated with GBC progression. Prognostic roles of miR‑642a‑3p and miR‑145‑5p
and validation of predicted hub genes using TCGA
database and SRRSH database Using SRRSH database, we found that miR-642-3p
and miR-145-5p were respectively up-regulated and
down-regulated in GBC patients with poor prognosis via
RNA extraction and qRT-PCR based on their tissue slides
(Fig. 5C, D). IHC was further conducted to indirectly
validate the potential target genes expression of miR-
642a-3p and miR-145-5p. Similar results were observed
in related target genes, such as SYK, SH3GL1, CDKN1A,
MYC, VEGFA and so on (Additional file 1: Fig. S3). Prognostic roles of miR‑642a‑3p and miR‑145‑5p
and validation of predicted hub genes using TCGA
database and SRRSH database value
adj.P.Val
B
hsa-miR-4430
1.560415197
2.69965023
13.8088015
1.38E−13
3.94E−11
21.12578004
hsa-miR-642a-3p
1.388882733
3.838242328
5.380473512
1.18E−05
6.02E−05
2.872762389
hsa-miR-3676-5p
1.328712158
4.479439939
6.483236311
6.68E−07
5.02E−06
5.740789951
hsa-miR-642b-3p
1.305137316
2.643572949
6.337901273
9.69E−07
6.77E−06
5.368095939
hsa-miR-4653-3p
1.274020361
2.896301402
10.02878504
1.71E−10
5.90E−09
14.03773895
hsa-miR-4733-5p
1.264661058
0.999736641
13.82584744
1.34E−13
3.94E−11
21.1541377
hsa-miR-4443
1.241770025
3.026350821
6.115673242
1.72E−06
1.10E−05
4.7945584
hsa-miR-4428
1.193376354
2.79367948
9.87297247
2.37E−10
7.80E−09
13.70803634
hsa-miR-4465
1.161175612
2.428359614
5.497611862
8.65E−06
4.56E−05
3.180798205
hsa-miR-3610
1.124733844
1.111994643
9.05744475
1.40E−09
3.34E−08
11.92953693 Table 1 The most up-regulated expression of miRNAs in GBC GBC gallbladder cancer, FC fold change Table 2. The most down-regulated expression of miRNAs in GBC
GBC gallbladder cancer, FC fold change
miRNA
logFC
AveExpr
t
P. value
adj.P.Val
B
hsa-miR-451a
− 2.657515961
2.162795615
− 8.922076057
1.89E−09
4.22E−08
11.62565902
hsa-miR-145-5p
− 2.530888937
2.586220072
− 8.866540584
2.14E−09
4.73E−08
11.5002749
hsa-miR-1
− 2.378360609
0.335278283
− 12.25075546
2.14E−12
2.69E−10
18.40491209
hsa-miR-99a-5p
− 2.29346529
1.634070053
− 7.94855799
1.80E−08
2.48E−07
9.367532444
hsa-miR-143-3p
− 2.113226719
1.198615156
− 8.624341644
3.72E−09
7.33E−08
10.94857798
hsa-miR-29c-3p
− 2.069281379
2.228370391
− 6.848924471
2.64E−07
2.33E−06
6.669209031
hsa-miR-125b-5p
− 2.012061309
2.869352599
− 6.861116734
2.56E−07
2.30E−06
6.699917987
hsa-miR-195-5p
− 1.962473569
1.488055322
− 6.378651302
8.72E−07
6.16E−06
5.472794062
hsa-miR-133b
− 1.917864513
0.445914537
− 14.4344557
4.87E−14
3.26E−11
22.14778776
hsa-miR-100-5p
− 1.891570713
1.864235972
− 7.865181939
2.19E−08
2.88E−07
9.168268045 Table 2. The most down-regulated expression of miRNAs in GBC GBC gallbladder cancer, FC fold change Collectively, these data indicated that miR-642-3p and
miR-145-5p were unusually expressed, which could be
prognostic biomarkers and associated with invasion and
metastasis in GBC. miR-145-5p in GBC. GBC-SD and SGC-996 cell lines,
primary GBC tissues samples were detected for the
expression of miR-642-3p and miR-145-5p. It was con-
firmed that miR-642-3p were up-regulated and miR-
145-5p were down-regulated in GBC cell lines (Fig. 5A,
B). Using SRRSH database, we found that miR-642-3p
and miR-145-5p were respectively up-regulated and
down-regulated in GBC patients with poor prognosis via
RNA extraction and qRT-PCR based on their tissue slides
(Fig. 5C, D). IHC was further conducted to indirectly
validate the potential target genes expression of miR-
642a-3p and miR-145-5p. Similar results were observed
in related target genes, such as SYK, SH3GL1, CDKN1A,
MYC, VEGFA and so on (Additional file 1: Fig. S3). miR-145-5p in GBC. GBC-SD and SGC-996 cell lines,
primary GBC tissues samples were detected for the
expression of miR-642-3p and miR-145-5p. It was con-
firmed that miR-642-3p were up-regulated and miR-
145-5p were down-regulated in GBC cell lines (Fig. 5A,
B). Prognostic roles of miR‑642a‑3p and miR‑145‑5p
and validation of predicted hub genes using TCGA
database and SRRSH database Further investigation was applied to analyze the expres-
sion and clinical significance of both miR-642-3p and Fig. 1 MiRNAs clustering of different expression. A Volcano plots of differentially expressed miRNAs. Green data points indicated down-regulated
miRNAs and red data points indicated up-regulated miRNAs. The differences was set as |log FC|> 1. Black dots indicated miRNAs that were not
differentially expressed. B The heatmap of differentially expressed miRNAs. Log FC of the genes was represented by the top-right legend. FC fold
change
(i
g
p g ) Cao et al. Journal of Translational Medicine (2022) 20:188 Page 5 of 14 Fig. 1 (See legend on previous page.) Fig. 1 (See legend on previous page.) Fig. 1 (See legend on previous page.) Cao et al. Journal of Translational Medicine (2022) 20:188 Page 6 of 14 Table 1 The most up-regulated expression of miRNAs in GBC
GBC gallbladder cancer, FC fold change
miRNA
logFC
AveExpr
t
P. value
adj.P.Val
B
hsa-miR-4430
1.560415197
2.69965023
13.8088015
1.38E−13
3.94E−11
21.12578004
hsa-miR-642a-3p
1.388882733
3.838242328
5.380473512
1.18E−05
6.02E−05
2.872762389
hsa-miR-3676-5p
1.328712158
4.479439939
6.483236311
6.68E−07
5.02E−06
5.740789951
hsa-miR-642b-3p
1.305137316
2.643572949
6.337901273
9.69E−07
6.77E−06
5.368095939
hsa-miR-4653-3p
1.274020361
2.896301402
10.02878504
1.71E−10
5.90E−09
14.03773895
hsa-miR-4733-5p
1.264661058
0.999736641
13.82584744
1.34E−13
3.94E−11
21.1541377
hsa-miR-4443
1.241770025
3.026350821
6.115673242
1.72E−06
1.10E−05
4.7945584
hsa-miR-4428
1.193376354
2.79367948
9.87297247
2.37E−10
7.80E−09
13.70803634
hsa-miR-4465
1.161175612
2.428359614
5.497611862
8.65E−06
4.56E−05
3.180798205
hsa-miR-3610
1.124733844
1.111994643
9.05744475
1.40E−09
3.34E−08
11.92953693
Table 2. The most down-regulated expression of miRNAs in GBC
GBC gallbladder cancer, FC fold change
miRNA
logFC
AveExpr
t
P. value
adj.P.Val
B
hsa-miR-451a
− 2.657515961
2.162795615
− 8.922076057
1.89E−09
4.22E−08
11.62565902
hsa-miR-145-5p
− 2.530888937
2.586220072
− 8.866540584
2.14E−09
4.73E−08
11.5002749
hsa-miR-1
− 2.378360609
0.335278283
− 12.25075546
2.14E−12
2.69E−10
18.40491209
hsa-miR-99a-5p
− 2.29346529
1.634070053
− 7.94855799
1.80E−08
2.48E−07
9.367532444
hsa-miR-143-3p
− 2.113226719
1.198615156
− 8.624341644
3.72E−09
7.33E−08
10.94857798
hsa-miR-29c-3p
− 2.069281379
2.228370391
− 6.848924471
2.64E−07
2.33E−06
6.669209031
hsa-miR-125b-5p
− 2.012061309
2.869352599
− 6.861116734
2.56E−07
2.30E−06
6.699917987
hsa-miR-195-5p
− 1.962473569
1.488055322
− 6.378651302
8.72E−07
6.16E−06
5.472794062
hsa-miR-133b
− 1.917864513
0.445914537
− 14.4344557
4.87E−14
3.26E−11
22.14778776
hsa-miR-100-5p
− 1.891570713
1.864235972
− 7.865181939
2.19E−08
2.88E−07
9.168268045 Table 1 The most up-regulated expression of miRNAs in GBC
GBC gallbladder cancer, FC fold change
miRNA
logFC
AveExpr
t
P. Validation of potential target genes of miR‑642a‑3p and
miR‑145‑5p in vitro We used miR-642a-3p inhibitor and miR-145-5p mimics
for further validation of potential target genes of miR-
642a-3p and miR-145-5p, respectively. When using miR-
642a-3p inhibitor, we found mRNA level of SH3GL1 and
CDKN1A were decreased (Additional file 1: Fig. S4A). Moreover, when using miR-145-5p mimics, we found
mRNA level of MYC, VEGFA, and EGFR were signifi-
cantly regulated (Additional file 1: Fig. S4B). Besides, the
relationship of the above-mentioned genes with tumor
invasion have been reported by previous studies [22–25]. i
Survival analysis was performed to assess the prog-
nostic value of miR-642-3p and miR-145-5p in GBC. Although miR-642-3p was not associated with worse OS
in TCGA (Fig. 5E), high miR-642-3p presented decreased
survival in our hospital data (Fig. 5G). It was implied
that miR-642-3p may be a potential oncogene. By con-
trast, miR-145-5p was indicated as a favorable prognos-
tic factor both in TCGA (Fig. 5F) and our data (Fig. 5H). MiR‑642a‑3p and miR‑145‑5p on GBC migration and invasion
in vitro MiR‑642a‑3p and miR‑145‑5p on GBC migration and invasion
in vitro in vitro To examine the role of miR-642a-3p and miR-145-5p in
GBC migration, a wound-healing assay was employed. Cao et al. Journal of Translational Medicine (2022) 20:188 Page 7 of 14 RT-qPCR revealed significantly decreased miR-642a-3p
and increased miR-145-5p levels of GBC cell lines after
transfection, respectively (Additional file 1: Fig. S5). Our results demonstrated that decreased miR-642a-3p
expression and increased miR-145-5p expression mark-
edly suppress the migration ability of GBCs compared to
Fig. 2 GO functions for the target genes of the up-regulated and down-regulated miRNAs. A Up-regulated miRNA enriched biological process. B
Up-regulated miRNAs enriched cellular component. C Up-regulated miRNAs enriched molecular function. D Down-regulated miRNAs enriched
biological process. E Down-regulated miRNAs enriched cellular component. F Down-regulated miRNAs enriched molecular function. GO gene
ontology nd down-regulated miRNAs. A Up-regulated miRNA enriched biological process. B
ated miRNAs enriched molecular function. D Down-regulated miRNAs enriched
component. F Down-regulated miRNAs enriched molecular function. GO gene target genes of the up-regulated and down-regulated miRNAs. A Up-regulated miRNA enriched biological process. B
hed cellular component. C Up-regulated miRNAs enriched molecular function. D Down-regulated miRNAs enriched
regulated miRNAs enriched cellular component. F Down-regulated miRNAs enriched molecular function. GO gene Fig. 2 GO functions for the target genes of the up-regulated and down-regulated miRNAs. A Up-regulated miRNA enriched biological process. B
Up-regulated miRNAs enriched cellular component. C Up-regulated miRNAs enriched molecular function. D Down-regulated miRNAs enriched
biological process. E Down-regulated miRNAs enriched cellular component. F Down-regulated miRNAs enriched molecular function. GO gene
ontology Our results demonstrated that decreased miR-642a-3p
expression and increased miR-145-5p expression mark-
edly suppress the migration ability of GBCs compared to RT-qPCR revealed significantly decreased miR-642a-3p
and increased miR-145-5p levels of GBC cell lines after
transfection, respectively (Additional file 1: Fig. S5). Cao et al. Journal of Translational Medicine (2022) 20:188 Page 8 of 14 Fig. 3 The distribution of target genes of four selected miRNAs of different expression in GO-enriched functions. A Up-regulated miRNAs and (B)
down-regulated miRNAs. Symbols of target genes were displayed on the left side of the graph. Gene involvement under GO terms was established
by the colored connecting lines to the right. GO gene ontology. *P < 0.05; **P < 0.01 Fig. 3 The distribution of target genes of four selected miRNAs of different expression in GO-enriched functions. A Up-regulated miRNAs and (B)
down-regulated miRNAs. Symbols of target genes were displayed on the left side of the graph. in vitro f
g
p
g
p
g
Indeed, a total of 4 miRNAs, including the top 2 up-
regulated miRNAs (miR-642a-3p and miR-4430) and the
top 2 down-regulated miRNAs (miR-145-5p and miR-
451a), were selected for experimental validation. It was
shown that miR-145-5p up-regulation and miR-642a-3p
down-regulation regulated cell invasion and migration
in GBC. Interestingly, the decrease in miR-4430 expres-
sion suppressed cell invasion and wound healing migra-
tion to enhance GBC loss-of-function, in the meantime,
up-regulation of miR-451a significantly suppressed the
invasion (Additional file 1: Fig. S6). Similar to our results,
Ueta et al. [30] found that serum EVs miR-451a were sig-
nificantly down-regulated and miR-451a inhibited GBC
cell proliferation and induced apoptosis. Taken together,
miR-451a would also be a novel therapeutic target for
GBC patients. negative control cells (Fig. 6A). Transwell invasion assay
was performed to evaluate the influence of both miRNAs
on GBC invasion. The results demonstrated the decrease
of miR-642a-3p and over-expressed miR-145-5p sup-
pressed cell invasion, respectively (P < 0.05, Fig. 6B–D). Hence, all data confirmed that both 2 miRNAs (miR-
642a-3p and miR-145-5p) were key regulators relevant to
invasion and metastasis of GBC, and approaches of tar-
geting them may represent novel directions to improve
the GBC prognosis. negative control cells (Fig. 6A). Transwell invasion assay
was performed to evaluate the influence of both miRNAs
on GBC invasion. The results demonstrated the decrease
of miR-642a-3p and over-expressed miR-145-5p sup-
pressed cell invasion, respectively (P < 0.05, Fig. 6B–D). Hence, all data confirmed that both 2 miRNAs (miR-
642a-3p and miR-145-5p) were key regulators relevant to
invasion and metastasis of GBC, and approaches of tar-
geting them may represent novel directions to improve
the GBC prognosis. In the miRNA-gene hub network, the hub genes were
potentially regulated by miR-145-5p and miR-642a-3p. Various studies demonstrated that miR-145-5p was
linked with various types of cancer, such as hepatocel-
lular carcinoma [31], gastric cancer [32, 33], and upper
tract urothelial carcinoma [34]. MiR-642a-3p was also
linked to cell migration and invasion of hepatocellu-
lar carcinoma [35]. Besides, miR-29c-3p was found to
be down-regulated (Table 2), which was in accordance
with the results that up-regulation of miR-29c-3p could
reverse EMT and decrease the metastasis ability in vitro
and in vivo [28]. Similar to our results in Table 2, Jin
et al. [36] also found that miR-143-3p suppressed tumor
angiogenesis and growth of GBC through the ITGA6/
PI3K/AKT/PLGF pathways. in vitro Gene involvement under GO terms was established
by the colored connecting lines to the right. GO gene ontology. *P < 0.05; **P < 0.01 Fig. 3 The distribution of target genes of four selected miRNAs of different expression in GO-enriched functions. A Up-regulated miRNAs and (B)
down-regulated miRNAs. Symbols of target genes were displayed on the left side of the graph. Gene involvement under GO terms was established
by the colored connecting lines to the right. GO gene ontology. *P < 0.05; **P < 0.01 Fig. 3 The distribution of target genes of four selected miRNAs of different expression in GO-enriched functions. A Up-regulated miRNAs and (B)
down-regulated miRNAs. Symbols of target genes were displayed on the left side of the graph. Gene involvement under GO terms was established
by the colored connecting lines to the right. GO gene ontology. *P < 0.05; **P < 0.01 Cao et al. Journal of Translational Medicine (2022) 20:188 Page 9 of 14 Table 3. Identification of hub genes by PPI network
PPI protein–protein interaction
Up-regulated miRNAs
Down-regulated miRNAs
Gene
Degree
Gene
Degree
MAPK1
34
AKT1
76
PTEN
27
VEGFA
68
SYK
25
MAPK1
65
TAF1
23
EGFR
63
CDKN1A
22
MYC
59
POLR2E
21
IL6
57
PRKCA
21
BCL2
52
CDK4
19
ACTB
44
SH3GL1
19
SMAD4
43
DNAJC10
18
ESR1
42 Table 3. Identification of hub genes by PPI network miR-20a/Smad7/β-catenin axis [26]. Despite the iden-
tified miRNAs, further studies would be performed to
offer more insights into improving the prognosis of GBC. Indeed, a total of 4 miRNAs, including the top 2 up-
regulated miRNAs (miR-642a-3p and miR-4430) and the
top 2 down-regulated miRNAs (miR-145-5p and miR-
451a), were selected for experimental validation. It was
shown that miR-145-5p up-regulation and miR-642a-3p
down-regulation regulated cell invasion and migration
in GBC. Interestingly, the decrease in miR-4430 expres-
sion suppressed cell invasion and wound healing migra-
tion to enhance GBC loss-of-function, in the meantime,
up-regulation of miR-451a significantly suppressed the
invasion (Additional file 1: Fig. S6). Similar to our results,
Ueta et al. [30] found that serum EVs miR-451a were sig-
nificantly down-regulated and miR-451a inhibited GBC
cell proliferation and induced apoptosis. Taken together,
miR-451a would also be a novel therapeutic target for
GBC patients. miR-20a/Smad7/β-catenin axis [26]. Despite the iden-
tified miRNAs, further studies would be performed to
offer more insights into improving the prognosis of GBC. in vitro The aberrant expression of
miRNAs have been relevant to GBC tumorigenesis and
progression. Discussion
h
d In the study, miR-642a-3p and miR-145-5p were identi-
fied as invasion-metastasis associated miRNAs according
to the bioinformatic analysis and experimental validation. As both two miRNAs were crucial in GBC invasion and
metastasis, they would be utilized as promising targets
for effective treatment to improve the prognosis of GBC
patients. qRT-PCR results indicated that miR-145-5p and miR-
642a-3p were significantly up- and down-regulated
among GBC cell lines and clinical samples, respectively. MiR-145-5p up-regulation and miR-642a-3p down-reg-
ulation could significantly suppress in vitro activation,
migration, and invasion of GBC using wound-healing and
Transwell invasion assay. To explore the possible path-
ways, we performed GO annotation and KEGG pathway
analysis for the predicted target genes of the top 2 most
up- and down-regulated miRNAs using the Enrichr tool. Several target genes were revealed to be enriched in cell–
cell adhesion, which was closely correlated to cell migra-
tion and invasion. Therefore, it would be a promising
mechanism that can relieve the impacts of miR-145-5p
and miR-642a-3p for GBC by targeting these genes. Currently, miRNAs were gradually considered as
essential regulators for tumor initiation, promotion,
and progression transcriptional dynamics in GBC [26,
27], and thus researchers should pay more attention to
systemically analyzing the invasion-metastasis associ-
ated miRNA in GBC. MiR-7-2-3p and miR-29c-3p were
identified as metastasis suppressors for GBC via high-
throughput screening, and they were also related to the
pathogenesis of GBC [28]. Meanwhile, Ma et al. [29]
demonstrated that miR-663a could regulate epithelial
membrane protein-3 to suppresses GBC progression
via interfering the MAPK/ERK pathway, indicating that
the miR-663a/EMP3/MAPK/ERK axis would be poten-
tial treatment for GBC. Besides, after being activated
by TGF-β1, miR-20a could play a crucial role in the
pathogenesis and trigger metastasis of GBC through the The study has several limitations that need to be
addressed. First, the public dataset with both miRNA and Cao et al. Journal of Translational Medicine (2022) 20:188 Page 10 of 14 Fig. 4 The regulatory network between dysregulated miRNAs and hub genes. A Up-regulated miRNAs and (B) down-regulated miRNAs. The mRNA
expression of predicted targets for miRNAs were from the TCGA database. C MiR-642a-3p: SYK expression, SH3GL1 expression, CDKN1A expression. D
MiR-145-5p: MYC expression, VEGFA expression, EGFR expression. *P < 0.05 Fig. 4 The regulatory network between dysregulated miRNAs and hub genes. A Up-regulated miRNAs and (B) down-regulated miRNAs. The mRNA
expression of predicted targets for miRNAs were from the TCGA database. Abbreviations GBC: Gallbladder cancer; BTC: Biliary tract cancer; AJCC: American Joint
Committee on Cancer; EMT: Epithelial–mesenchymal transition; PD-L1: Pro-
grammed cell death-ligand 1; NCBI: National Center for Biotechnology Infor-
mation; GEO: Gene Expression Omnibus; FC: Fold change; GO: Gene Ontology;
KEGG: Kyoto Encyclopedia of Genes and Genomes; PPI: Protein–protein
interaction; TCGA: The Cancer Genome Atlas; OS: Overall survival; HiBEC:
Human intrahepatic biliary epithelial cells; FBS: Fetal bovine serum; qRT-PCR:
Quantitative real-time PCR; CT: Cycle thresholds; IHC: Immunohistochemistry;
DAB: Diaminobenzidine; SEM: Standard error of mean; BP: Biological process;
CC: Cellular component; MF: Molecular function; RNAP II: RNA polymerase II. Discussion
h
d C MiR-642a-3p: SYK expression, SH3GL1 expression, CDKN1A expression. D
MiR-145-5p: MYC expression, VEGFA expression, EGFR expression. *P < 0.05 Cao et al. Journal of Translational Medicine (2022) 20:188 Page 11 of 14 Fig. 5 The expression and prognostic roles of miR-642a-3p and miR-145-5p in GBC. A MiR-642a-3p expression and (B) miR-145-5p expression in 2
GBC cell lines (SGC-996, GBC-SD) were compared with HiBEC. C The expression of miR-642a-3p and (D) miR-145-5p in tumor tissues of GBC patients
with different prognosis. E Kaplan–Meier curve of OS in CHOL patients with high vs. low expression of miR-642a-3p from TCGA database. F Kaplan–
Meier curve of OS in CHOL patients with high vs. low expression of miR-145-5p for patients from TCGA database. G Kaplan–Meier curve of OS with
high vs. low expression of miR-642a-3p in GBC patients from clinical data. H Kaplan–Meier curve of OS with high vs. low expression of miR-145-5p
in GBC patients from clinical data. GBC gallbladder cancer, OS overall survival, CHOL cholangiocarcinoma, TCGA the Cancer Genome Atlas. *P < 0.05;
**P < 0.01; ***P < 0.001; ****P < 0.0001 Fig. 5 The expression and prognostic roles of miR-642a-3p and miR-145-5p in GBC. A MiR-642a-3p expression and (B) miR-145-5p expression in 2
GBC cell lines (SGC-996, GBC-SD) were compared with HiBEC. C The expression of miR-642a-3p and (D) miR-145-5p in tumor tissues of GBC patients
with different prognosis. E Kaplan–Meier curve of OS in CHOL patients with high vs. low expression of miR-642a-3p from TCGA database. F Kaplan–
Meier curve of OS in CHOL patients with high vs. low expression of miR-145-5p for patients from TCGA database. G Kaplan–Meier curve of OS with
high vs. low expression of miR-642a-3p in GBC patients from clinical data. H Kaplan–Meier curve of OS with high vs. low expression of miR-145-5p
in GBC patients from clinical data. GBC gallbladder cancer, OS overall survival, CHOL cholangiocarcinoma, TCGA the Cancer Genome Atlas. *P < 0.05;
**P < 0.01; ***P < 0.001; ****P < 0.0001 mRNA profiling in GBC was limited. Therefore, we tried
to validate our results using different GBC databases
(TCGA and SRRSH databases). Unfortunately, some veri-
fication results would not be obvious because of the small
sample size. And further validation experiments should
be performed with a large sample size. Second, merely 2
most up- and down-regulated miRNAs and the relevant
target genes were included in the enrichment analysis. Supplementary Information The online version contains supplementary material available at https://doi.
org/10.1186/s12967-022-03394-8. The online version contains supplementary material available at https://doi.
org/10.1186/s12967-022-03394-8. Additional file 1: Table S1. List of the primers used for miRNA quantita-
tive real-time PCR. Figure S1. Data GSE104165 was shown. Figure S2. The
comparison of mRNA expression of hub genes based on different T stages. (A) MiR-642a-3p: CDKN1A expression. (B) MiR-145-5p: MYC expression. (C)
MiR-145-5p: VEGFA expression. *P. Figure S3. The comparison of immuno-
histochemistry of GBC tissue samples based on different survival. (A) SYK. (B) SH3GL1. (C) CDKN1A. (D) MYC. (E) VEGFA. (F) EGFR. GBC: Gallbladder
cancer. *P. Figure S4. The mRNA expression of predicted hub genes in
GBC-SD cell. (A) The expression of SH3GL1, CDKN1A, and SYK in nega-
tive control vs. miR-642a-3p inhibitor. (B) The expression of MYC, EGFR,
and VEGFA in negative control vs. miR-145-5p mimics. GBC: Gallbladder Discussion
h
d Further studies would be focused on more miRNAs of
different expression to explore more potential treatment
for GBC. Third, we selected 2 up- and down-regulated
miRNAs for in vitro experiment validation, which would
lead to the omittance of some functional miRNAs. And
in vitro and in vivo experiments should be performed to
reveal other functional miRNAs with a high validation
accuracy. Finally, the deeper molecular mechanisms of
GBC invasion and metastasis should also be explored in
the future. (up-regulation) would be served as potential therapeutic
targets for GBC in the future. Conclusions (A) SGC-996 and GBC-SD, which were
transfected with miR-4430 inhibitor and miR-451a mimic, invaded less
versus control cancer cells, respectively. (B) Quantification of SGC-996
and GBC-SD after miR-4430 inhibitor. (C) Quantification of SGC-996 and
GBC-SD after miR-451a. (D) SGC-996 and GBC-SD were transfected with
inhibitor NC, miR-4430 inhibitor, mimic NC, miR-451a mimic, respectively. Wound healing assay was performed in GBC cell with 48h of recovery. GBC: Gallbladder cancer. NC: Negative control. ***P. 4. Valle JW, Kelley RK, Nervi B, Oh DY, Zhu AX. Biliary tract cancer. Lancet. 2021;397:428–44. 5. Chun YS, Pawlik TM, Vauthey JN. 8th Edition of the AJCC cancer stag-
ing manual: pancreas and hepatobiliary cancers. Ann Surg Oncol. 2018;25:845–7. 6. Mantripragada KC, Hamid F, Shafqat H, Olszewski AJ. Adjuvant therapy
for resected gallbladder cancer: analysis of the national cancer data
base. J Natl Cancer Inst. 2017;109. 7. Chen M, Cao J, Bai Y, Tong C, Lin J, Jindal V, et al. Development and
validation of a nomogram for early detection of malignant gallbladder
lesions. Clin Transl Gastroenterol. 2019;10: e00098. 8. Tauriello DVF, Palomo-Ponce S, Stork D, Berenguer-Llergo A, Badia-
Ramentol J, Iglesias M, et al. TGFbeta drives immune evasion in geneti-
cally reconstituted colon cancer metastasis. Nature. 2018;554:538–43. Additional file 2. Ethics file. Additional file 2. Ethics file. 9. Wu MJ, Chen YS, Kim MR, Chang CC, Gampala S, Zhang Y, et al. Epithe-
lial–mesenchymal transition directs stem cell polarity via regulation of
mitofusin. Cell Metab. 2019;29(993–1002): e6. Funding
h
k 12. Bartel DP. MicroRNAs: target recognition and regulatory functions. Cell. 2009;136:215–33. This work was supported by Scientific Research Fund of Zhejiang Provincial
Education Department (Y202148325), and National Natural Science Founda-
tion of China (81827804 and 81802330). 13. Shukla GC, Singh J, Barik S. MicroRNAs: processing, maturation,
target recognition and regulatory functions. Mol Cell Pharmacol. 2011;3:83–92. Acknowledgements
Not applicable. 10. Zheng L, Xu M, Xu J, Wu K, Fang Q, Liang Y, et al. ELF3 promotes
epithelial-mesenchymal transition by protecting ZEB1 from miR-
141-3p-mediated silencing in hepatocellular carcinoma. Cell Death Dis. 2018;9:387. 2.
Rakić M, Patrlj L, Kopljar M, Kliček R, Kolovrat M, Loncar B, et al. Gall-
bladder cancer hepatobiliary. Surg Nutr. 2014;3:221–6. Received: 2 December 2021 Accepted: 15 April 2022 23. Wisnieski F, Calcagno DQ, Leal MF, Chen ES, Gigek CO, Santos LC, et al. Differential expression of histone deacetylase and acetyltransferase
genes in gastric cancer and their modulation by trichostatin A. Tumour
Biol. 2014;35:6373–81. Availability of data and materials 14. Liu N, Chang CW, Steer CJ, Wang XW, Song G. MicroRNA-15a/16-1
prevents hepatocellular carcinoma by disrupting the communica-
tion between Kupffer cells and regulatory T cells. Gastroenterology. 2021;162:575. The original data of the study are available from the corresponding authors
upon reasonable request. 15. Zhu M, Wei C, Lin J, Dong S, Gao D, Chen J, et al. UHRF1 is regulated by
miR-124-3p and promotes cell proliferation in intrahepatic cholangio-
carcinoma. J Cell Physiol. 2019;234:19875–85. Consent for publication
Not applicable. 19. Szklarczyk D, Gable AL, Lyon D, Junge A, Wyder S, Huerta-Cepas J, et al. STRING v11: protein–protein association networks with increased cov-
erage, supporting functional discovery in genome-wide experimental
datasets. Nucleic Acids Res. 2019;47:D607–13. Author contributions JSC, HJS, JHH, XJC, and MYC designed the study and collected the data; JSC,
AYF, BZ, RAJ, SJL, and TEC analyzed and interpreted the data; JSC, SJ, WT, YTT,
and ZYL wrote the manuscript; XJC and MYC revised the manuscript. All
authors read and approved the final version of the manuscript. 11. Civenni G, Bosotti R, Timpanaro A, Vazquez R, Merulla J, Pandit S,
et al. Epigenetic control of mitochondrial fission enables self-renewal
of stem-like tumor cells in human prostate cancer. Cell Metab. 2019;30(303–18): e6. 1.
Hundal R, Shaffer EA. Gallbladder cancer: epidemiology and outcome.
Clin Epidemiol. 2014;6:99–109. Conclusions In the present study, miR-642a-3p and miR-145-5p were
identified as invasion-metastasis associated miRNAs
via bioinformatic analysis and experimental validation. Both miR-642a-3p (down-regulation) and miR-145-5p Cao et al. Journal of Translational Medicine (2022) 20:188 Page 12 of 14 g. 6 MiR-642a-3p under-expression and miR-145-5p over-expression regulated GBC invasion and migration. A–C After being transfected with
iR-642a-3p inhibitor and miR-145-5p mimic, SGC-996 and GBC-SD showed weaker invasion than the control cells, respectively. D SGC-996 and
BC-SD were transfected with inhibitor NC, miR-642a-3p inhibitor, mimic NC, miR-145-5p mimic, respectively. Wound healing assay was performed
GBC cell with 48 h of recovery. GBC: Gallbladder cancer. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Error bars represented SD for n = 3 Fig. 6 MiR-642a-3p under-expression and miR-145-5p over-expression regulated GBC invasion and migration. A–C After being transfected with
miR-642a-3p inhibitor and miR-145-5p mimic, SGC-996 and GBC-SD showed weaker invasion than the control cells, respectively. D SGC-996 and
GBC-SD were transfected with inhibitor NC, miR-642a-3p inhibitor, mimic NC, miR-145-5p mimic, respectively. Wound healing assay was performed
in GBC cell with 48 h of recovery. GBC: Gallbladder cancer. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Error bars represented SD for n = 3 Fig. 6 MiR-642a-3p under-expression and miR-145-5p over-expression regulated GBC invasion and migration. A–C After being transfected with
miR-642a-3p inhibitor and miR-145-5p mimic, SGC-996 and GBC-SD showed weaker invasion than the control cells, respectively. D SGC-996 and
GBC-SD were transfected with inhibitor NC, miR-642a-3p inhibitor, mimic NC, miR-145-5p mimic, respectively. Wound healing assay was performed
in GBC cell with 48 h of recovery. GBC: Gallbladder cancer. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. Error bars represented SD for n = 3 Cao et al. Journal of Translational Medicine (2022) 20:188 Page 13 of 14 3. Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2021. CA
Cancer J Clin. 2021;71:7–33. 3. Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 20 cancer. *P. Figure S5. (A-B) After being transfected with 50 nM mimics, the
expression of miR-642a-3p and miR-145-5p were significantly elevated
in SGC-996 and GBC-SD cell lines. qRT-PCR was applied to detect the
miR-642a-3p and miR-145-5p expression levels after 48h transfection. GBC:
Gallbladder cancer; qRT-PCR: Quantitative real-time PCR. *P . Figure S6. Down-regulation of miR-4430 and up-regulation of miR-451a regulated
GBC cell invasion and metastasis. Ethics approval and consent to participate 16. Roshani Asl E, Rasmi Y, Baradaran B. MicroRNA-124-3p suppresses
PD-L1 expression and inhibits tumorigenesis of colorectal cancer cells
via modulating STAT3 signaling. J Cell Physiol. 2021;236:7071–87. The current study was approved by the institutional ethics review board of Sir
Run-Run Shaw Hospital, Zhejiang University, Hangzhou, Zhejiang Province,
China (No. 20210625-32) (Additional file 2: Ethics file). The patients provided
their written informed consent to participate in the study. 17. Gene OC. The Gene Ontology (GO) project in 2006. Nucleic Acids Res. 2006;34:D322–6. 18. Chen EY, Tan CM, Kou Y, Duan Q, Wang Z, Meirelles GV, et al. Enrichr:
interactive and collaborative HTML5 gene list enrichment analysis tool. BMC Bioinformatics. 2013;14:128. Author details 20. Mohri D, Ijichi H, Miyabayashi K, Takahashi R, Kudo Y, Sasaki T, et al. A
potent therapeutics for gallbladder cancer by combinatorial inhibi-
tion of the MAPK and mTOR signaling networks. J Gastroenterol. 2016;51:711–21. 1 Department of General Surgery, Sir Run‑Run Shaw Hospital, Zhejiang
University, No. 3 East Qingchun Road, Hangzhou 310016, Zhejiang Prov-
ince, China. 2 Zhejiang University School of Medicine, Zhejiang University,
Hangzhou 310058, Zhejiang Province, China. 3 Department of Hepatobiliary
and Pancreatic Surgery, Shaoxing People’s Hospital, Shaoxing 312000, Zheji-
ang Province, China. 4 Health Management Center, The First Affiliated Hospital,
College of Medicine, Zhejiang University, Hangzhou 310022, China. 21. Zhang Y, Wang H, Chen T, Wang H, Liang X, Zhang Y, et al. C24-cer-
amide drives gallbladder cancer progression through directly
targeting PIP4K2C to facilitate mTOR signaling activation. Hepatology. 2020;73:692. 22. Takahashi T, Shivapurkar N, Riquelme E, Shigematsu H, Reddy J, Suzuki
M, et al. Aberrant promoter hypermethylation of multiple genes in
gallbladder carcinoma and chronic cholecystitis. Clin Cancer Res. 2004;10:6126–33. Received: 2 December 2021 Accepted: 15 April 2022 Received: 2 December 2021 Accepted: 15 April 2022 Cao et al. Journal of Translational Medicine (2022) 20:188 Yu Z, Du Y, Li H, Huang J, Jiang D, Fan J, et al. miR-642 serves as a tumor
suppressor in hepatocellular carcinoma by regulating SEMA4C and p38
MAPK signaling pathway. Oncol Lett. 2020;20:74. 36. Jin YP, Hu YP, Wu XS, Wu YS, Ye YY, Li HF, et al. miR-143-3p targeting of
ITGA6 suppresses tumour growth and angiogenesis by downregulating
PLGF expression via the PI3K/AKT pathway in gallbladder carcinoma. Cell
Death Dis. 2018;9:182. Cao et al. Journal of Translational Medicine (2022) 20:188 Cao et al. Journal of Translational Medicine (2022) 20:188 Cao et al. Journal of Translational Medicine (2022) 20:188 25. Shen H, He M, Lin R, Zhan M, Xu S, Huang X, et al. PLEK2 promotes gall-
bladder cancer invasion and metastasis through EGFR/CCL2 pathway. J
Exp Clin Cancer Res. 2019;38:247. 26. Chang Y, Liu C, Yang J, Liu G, Feng F, Tang J, et al. MiR-20a triggers
metastasis of gallbladder carcinoma. J Hepatol. 2013;59:518–27. 27. Srivastava K, Srivastava A, Mittal B. Common genetic variants in pre-
microRNAs and risk of gallbladder cancer in North Indian population. J
Hum Genet. 2010;55:495–9. 28. Lu K, Feng F, Yang Y, Liu K, Duan J, Liu H, et al. High-throughput screen-
ing identified miR-7-2-3p and miR-29c-3p as metastasis suppressors in
gallbladder carcinoma. J Gastroenterol. 2020;55:51–66. 29. Ma Q, Zhang Y, Liang H, Zhang F, Liu F, Chen S, et al. EMP3, which
is regulated by miR-663a, suppresses gallbladder cancer progres-
sion via interference with the MAPK/ERK pathway. Cancer Lett. 2018;430:97–108. 30. Ueta E, Tsutsumi K, Kato H, Matsushita H, Shiraha H, Fujii M, et al. Extracellular vesicle-shuttled miRNAs as a diagnostic and prognostic
biomarker and their potential roles in gallbladder cancer patients. Sci
Rep. 2021;11:12298. 31. Wang W, Ding B, Lou W, Lin S. Promoter hypomethylation and miR-
145-5p downregulation-mediated HDAC11 overexpression promotes
sorafenib resistance and metastasis of hepatocellular carcinoma cells. Front Cell Dev Biol. 2020;8:724. 32. Zhou K, Song B, Wei M, Fang J, Xu Y. ANGPT2/MiR-145-5p suppresses
the proliferation, migration and invasion of gastric cancer epithelial
cells via the NOD_LIKE_RECEPTOR axis. Cancer Cell Int. 2020;20:416. 33. Zhou T, Chen S, Mao X. miR-145-5p affects the differentiation of gastric
cancer by targeting KLF5 directly. J Cell Physiol. 2019;234:7634–44. 34. Hsu W, Li W, Lee Y, Huang A, Chang L, Lin H, et al. MicroRNA-145
suppresses cell migration and invasion in upper tract urothelial
carcinoma by targeting ARF6. FASEB J Off Publ Fed Am Soc Exp Biol. 2020;34:5975–92. 34. Hsu W, Li W, Lee Y, Huang A, Chang L, Lin H, et al. MicroRNA-145
suppresses cell migration and invasion in upper tract urothelial
carcinoma by targeting ARF6. FASEB J Off Publ Fed Am Soc Exp Biol. 2020;34:5975–92. 35. Yu Z, Du Y, Li H, Huang J, Jiang D, Fan J, et al. miR-642 serves as a tumor
suppressor in hepatocellular carcinoma by regulating SEMA4C and p38
MAPK signaling pathway. Oncol Lett. 2020;20:74. 35. References: 24. Singh P, Jain SL, Sakhuja P, Agarwal A. Expression of VEGF-A, HER2/neu,
and KRAS in gall bladder carcinoma and their correlation with clinico-
pathological parameters. Indian J Pathol Microbiol. 2021;64:687–92. 24. Singh P, Jain SL, Sakhuja P, Agarwal A. Expression of VEGF-A, HER2/neu,
and KRAS in gall bladder carcinoma and their correlation with clinico-
pathological parameters. Indian J Pathol Microbiol. 2021;64:687–92. Page 14 of 14 Cao et al. Journal of Translational Medicine (2022) 20:188 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations. •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from:
|
https://openalex.org/W2725125753
|
https://archimer.ifremer.fr/doc/00392/50324/51016.pdf
|
English
| null |
Preface: The Oligotrophy to the UlTra-oligotrophy PACific Experiment (OUTPACE cruise, 18 February to 3 April 2015)
|
Biogeosciences
| 2,017
|
cc-by
| 12,996
|
Preface: The Oligotrophy to the UlTra-oligotrophy PACific
Experiment (OUTPACE cruise, 18 February to 3 April 2015)
Thierry Moutin1, Andrea Michelangelo Doglioli1, Alain de Verneil1, and Sophie Bonnet2
1Aix Marseille Université, CNRS, Université de Toulon, IRD, OSU Pythéas, Mediterranean Institute of Oceanograp
(MIO), UM 110, 13288, Marseille, France
2Aix Marseille Université, CNRS, Université de Toulon, IRD, OSU Pythéas, Mediterranean Institute of Oceanograp
(MIO), UM 110, 98848, Nouméa, New Caledonia Correspondence to: Thierry Moutin (thierry.moutin@mio.osupytheas.fr) Received: 16 February 2017 – Discussion started: 28 February 2017
Revised: 2 June 2017 – Accepted: 5 June 2017 – Published: 6 July 2017 plan of the cruise and water mass and climatological charac-
teristics and concludes with a general overview of the other
papers that will be published in this special issue. Abstract. The overall goal of OUTPACE (Oligotrophy to
UlTra-oligotrophy PACific Experiment) was to obtain a suc-
cessful representation of the interactions between planktonic
organisms and the cycle of biogenic elements in the western
tropical South Pacific Ocean across trophic and N2 fixation
gradients. Within the context of climate change, it is nec-
essary to better quantify the ability of the oligotrophic ocean
to sequester carbon through biological processes. OUTPACE
was organized around three main objectives, which were (1)
to perform a zonal characterization of the biogeochemistry
and biological diversity of the western tropical South Pacific
during austral summer conditions, (2) to study the production
and fate of organic matter (including carbon export) in three
contrasting trophic regimes (increasing oligotrophy) with a
particular emphasis on the role of dinitrogen fixation, and
(3) to obtain a representation of the main biogeochemical
fluxes and dynamics of the planktonic trophic network. The
international OUTPACE cruise took place between 18 Febru-
ary and 3 April 2015 aboard the RV L’Atalante and in-
volved 60 scientists (30 onboard). The west–east transect
covered ∼4000 km from the western part of the Melanesian
archipelago (New Caledonia) to the western boundary of the
South Pacific gyre (French Polynesia). Following an adap-
tive strategy, the transect initially designed along the 19◦S
parallel was adapted along-route to incorporate information
coming from satellite measurements of sea surface temper-
ature, chlorophyll a concentration, currents, and diazotroph
quantification. After providing a general context and describ-
ing previous work done in this area, this introductory paper
elucidates the objectives of OUTPACE, the implementation Biogeosciences, 14, 3207–3220, 2017
https://doi.org/10.5194/bg-14-3207-2017
© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 3.0 License. Biogeosciences, 14, 3207–3220, 2017
https://doi.org/10.5194/bg-14-3207-2017
© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 3.0 License. 1
General context The additional carbon dioxide (CO2) in the atmosphere,
mainly resulting from fossil fuel emissions linked with hu-
man activities (anthropogenic CO2), is the main cause of
global warming (Fifth Assessment Report – Climate Change
2013 – IPCC). The ocean has acted thus far as a major
sink of anthropogenic CO2 (Sabine et al., 2004), preventing
greater CO2 accumulation in the atmosphere and therefore
a greater increase in the Earth’s temperature. The biological
pump (Fig. 1), the process by which carbon (C) is transferred
from the upper to the deep ocean by biological processes,
provides the main explanation for the vertical gradient of C
in the ocean. Its strength and efficiency depends upon the
complex balance between organic matter production in the
photic zone and its remineralization in both the epipelagic
and mesopelagic zones. Before present and for tens of thou-
sands of years, the biological pump was thought to be in an
equilibrium state with an associated near-zero net exchange
of CO2 with the atmosphere (Broecker, 1991; Murnane et al.,
1999). Climate alterations then started to disrupt this equilib-
rium and the expected modification of the biological pump
will probably considerably influence oceanic C sequestration
(and therefore global warming) in future decades (Sarmiento
and Gruber, 2006). The long-term decrease in phosphate 2
The role of N2 fixation in the oligotrophic ocean and
an overview of previous cruises in the western
tropical South Pacific Ocean (WTSP) (2012) for a detailed description. availability, and the shift from primary production previously
limited by nitrogen (N) and now by phosphorus (P), associ-
ated with increasing inputs of N by dinitrogen (N2) fixation
observed at the Hawaii Ocean Time-series (HOT) station in
the north Pacific gyre (NPG) (Karl et al., 1997; Karl, 2014),
appear as a first example of biological pump alteration. The input of new N to the surface ocean through biologi-
cal N2 fixation represents a major link between the C and N
biogeochemical cycles, relating upper-ocean nutrient avail-
ability with the biological pump and ultimately the ocean
and climate. This link was recently shown to play a central
role in previous natural climate changes over long timescales
(Galbraith et al., 2013). It is nevertheless clear that expected
climate change due to anthropogenic atmospheric CO2 input
may concern shorter timescales: the increase in atmospheric
CO2 over the past 200 years equals the increase in atmo-
spheric CO2 between glacial and interglacial periods, which
took place over several thousands of years (Sarmiento and
Gruber, 2006). Furthermore, enhanced stratification in the
tropical and subtropical ocean resulting from global warm-
ing (Polovina et al., 2008) might decrease nutrient availabil-
ity for nutrients such as nitrate (NO−
3 ), potentially favoring
N2 fixation in surface waters. It may also decrease phosphate
availability (Moutin et al., 2008), and in turn N2 fixation, net
primary production, and export (Karl, 2014). McMahon et
al. (2015), covering the past 1000 years, argue that N2 fixa-
tion has increased since the industrial revolution and might
provide a negative feedback to rising CO2. Most of the surface ocean (60 %; Longhurst, 1998) is
comprised of low-nutrient, low-biomass oligotrophic areas,
which constitute the largest coherent ecosystems on our
planet. They support a large part (40 %) of the photosynthetic
C fixation in the ocean (Antoine et al., 1996). This C fixation
is mainly performed by picoplankton (smaller than 2–3 µm
in diameter) that are generally thought to represent a negli-
gible fraction of the total particulate organic C (POC) export
flux due to their small size, slow individual sinking rates, and
tight grazer control that leads to high rates of recycling in the
euphotic zone. 2
The role of N2 fixation in the oligotrophic ocean and
an overview of previous cruises in the western
tropical South Pacific Ocean (WTSP) N f ixation
2
CO2
DIC
POC
DOC
Solar radiation
Atmosphere
Ocean
Nutrient
availability
Export production
Deep layer
(Deepwater circulation)
Winter mixed layer
(Intermediate circulation)
Photic z one
Figure 1. Major C fluxes for a biological pump budget and the main
role of N2 fixation. Biological pump is the C transfer into the ocean
interior by biological processes. DIC is dissolved inorganic C, POC
is particulate organic C, and DOC is dissolved inorganic C. See
Moutin et al. (2012) for a detailed description. N f ixation
2
CO2
DIC
POC
DOC
Solar radiation
Atmosphere
Ocean
Nutrient
availability
Export production
Deep layer
(Deepwater circulation)
Winter mixed layer
(Intermediate circulation)
Photic z one The efficiency of oceanic C sequestration depends upon
many factors, among which is the availability of nutrients
to support phytoplankton growth in the photic zone (Fig. 1). Large amounts of N are required for phytoplankton growth,
as it is an essential component of proteins, nucleic acids, and
other cellular constituents. Fixed N in the form of NO−
3 or
ammonium (NH+
4 ) is directly usable for growth, but concen-
trations are low (< 0.1 µmol L−1) and often growth limiting
in most of the open-ocean euphotic zone (Falkowski et al.,
1998; Moore et al., 2013). Dissolved N2 gas concentrations
in seawater are in contrast very high in the euphotic zone (ca. 450 µmol L−1) and could constitute a nearly inexhaustible N
source for the marine biota. However, only some prokary-
otic organisms (bacteria, cyanobacteria, archaea), hereafter
referred to as N2-fixing organisms (or diazotrophs), are able
to use this gaseous N source since they possess the nitroge-
nase enzyme that breaks the triple bond between the two N
atoms of the N2 molecule and converts it into a usable form
(i.e., NH3) (Zehr et al., 1998). On the global scale, they pro-
vide 140 ± 50 Tg N per year to the surface ocean, contribut-
ing more than atmospheric and riverine N inputs (Gruber,
2004). N2-fixing organisms thus act as natural fertilizers and
contribute to sustaining life and potential C export in coastal
and oceanic environments. Figure 1. Major C fluxes for a biological pump budget and the main
role of N2 fixation. Biological pump is the C transfer into the ocean
interior by biological processes. DIC is dissolved inorganic C, POC
is particulate organic C, and DOC is dissolved inorganic C. See
Moutin et al. Published by Copernicus Publications on behalf of the European Geosciences Union. Published by Copernicus Publications on behalf of the European Geosciences Union. T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment 3208 T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment 20 to 200 µmol N m−2 d−1 in the tropical North Atlantic (Be-
navides and Voss, 2015) and Pacific (Böttjer et al., 2017;
Dore et al., 2002), they reach 30–5400 µmol N m−2 d−1 (av-
erage ∼800 µmol N m−2 d−1) in the Solomon Sea (western
part of the WTSP) (Bonnet et al., 2009, 2015), which is
in the upper range of rates reported in the global N2 fixa-
tion MAREDAT database and even surpassed its upper rates
(100–1000 µmol N m−2 d−1) (Luo et al., 2012). Very high
rates have also been recently reported in the Arabian Sea
(Gandhi et al., 2011; Kumar et al., 2017). High rates rang-
ing from 151 to 703 µmol N m−2 d−1 have also been re-
ported off New Caledonia (Garcia et al., 2007), with sea-
sonal variations closely linked with phosphate availability
(Moutin et al., 2005; Van Den Broeck et al., 2004). The
seasonal distribution of N2 fixation is corroborated by in
situ and satellite observations (Dupouy et al., 2011) of re-
current large Trichodesmium blooms during austral summer
conditions (January–March) over the 1998–2010 period in
the Melanesian archipelago (MA: New Caledonia, Vanuatu,
Fiji Islands). In addition to Trichodesmium, which dominates
the diazotroph community in the WTSP (Moutin et al., 2005;
Bonnet et al., 2015), very high abundances of UCYN-B (up
to 106–107 nifH copies L−1) have been reported (Bonnet et
al., 2015; Campbell et al., 2005; Moisander et al., 2010). The
uncultivated UCYN-A (Zehr et al., 2008) also displays high
abundances (105–106 nifH copies L−1; Bonnet et al., 2015;
Moisander et al., 2010) around the MA, but they seem to
have different ecological niches compared to Trichodesmium
and UCYN-B (Bonnet et al., 2015; Moisander et al., 2010). In addition to submicron POM export, low δ15N signa-
tures in particles from sediment traps at the HOT station sug-
gest that at least part of the production sustained by N2 fix-
ation is ultimately exported out of the photic zone (Karl et
al., 1997, 2012; Scharek et al., 1999a; Sharek et al., 1999b). This may either be direct through sinking of diazotrophs,
or indirect, through the transfer of diazotroph-derived N to
non-diazotrophic plankton, which are subsequently exported. Karl et al. T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment (2012) reported an efficient summer export flux
of C (3 times greater than the mean wintertime particle
flux) at the HOT station that was attributed to the direct ex-
port of symbiotic N2-fixing cyanobacteria associated with di-
atoms (hereafter referred to as diatom–diazotroph associa-
tions or DDAs), which have high sinking and low remineral-
ization rates during downward transit. This result is in accor-
dance with high export fluxes measured in the tropical North
Atlantic when the diazotroph community is dominated by
DDAs (Subramaniam et al., 2008; White et al., 2012). More
recently, Berthelot et al. (2015) and Bonnet et al. (2016a)
studied the fate of a bloom of unicellular diazotrophs from
Group C (UCYN-C) during a mesocosm experiment in the
New Caledonia lagoon after a phosphate enrichment and re-
vealed that ∼10 % of UCYN-C from the water column was
exported to the particle traps daily, representing as much as
22.4 ± 5.5 % of the total POC exported at the height of the
UCYN-C bloom. A δ15N budget performed in the meso-
cosms confirmed the high contribution of N2 fixation (56 %;
Knapp et al., 2016) to export compared to other tropical and
subtropical regions where active N2 fixation contributes 10
to 25 % to export production (e.g., Altabet, 1988; Knapp
et al., 2005) and exceptionally up to 92 % in the Arabian
Sea (Gandhi et al., 2011; Kumar et al., 2017). Mechanisti-
cally, the vertical downward flux was enabled by the aggre-
gation of the small (5.7 ± 0.8 µm) UCYN-C cells into large
(100–500 µm) aggregates. In addition to direct export of di-
azotrophs, the use of nanoSIMS (nanoscale secondary ion
mass spectrometry) enabled the tracking of the fate of 15N
from both Trichodesmium (Bonnet et al., 2016b) and UCYN
blooms (Berthelot et al., 2015; Bonnet et al., 2016c) and
demonstrated that ∼8 % of N originating from N2 fixation
is quickly transferred to non-diazotrophic plankton, in par-
ticular diatoms, i.e., efficient C exporters to depth (Nelson
et al., 1995) during Trichodesmium blooms (Bonnet et al.,
2016b). This reveals that N2 fixation can fuel large-size non-
diazotrophic plankton growth in the water column, suggest-
ing an indirect export pathway of organic matter sustained
by N2 fixation in the oligotrophic ocean. 2
The role of N2 fixation in the oligotrophic ocean and
an overview of previous cruises in the western
tropical South Pacific Ocean (WTSP) Consequently, the efficiency of the biological
C pump in these oligotrophic systems has long been consid-
ered to be low as the greatest proportion of fixed C is thought
to be recycled in the surface layer and rapidly reexchanged
with the atmosphere. Recent studies have challenged this view and indicate that
all primary producers, including picoplanktonic cells, con-
tribute to export from the surface layer of the ocean at rates
proportional to their production rates (Richardson and Jack-
son, 2007). Export mechanisms differ compared to larger
cells as export of picoplankton is mainly due to packaging
into larger particles via grazing and/or aggregation processes
(Jackson, 2001; Lomas et al., 2010). More recently, Close et
al. (2013) pointed out that 40–70 % of picoplanktonic cells
are small enough to escape detection under the most common
definition of suspended particulate organic matter (POM). The ecosystem changes due to climate change are complex
and it is therefore necessary to characterize in detail the inter-
actions between N2 fixation and the C cycle to obtain a pre-
cise representation of the N2 fixation process in global bio-
geochemical models, leading eventually to predictions. Even
if considerable scientific progress has been made over the last
decades (see reviews from Sohm et al., 2011, and Zehr and
Kudela, 2011), many questions remain regarding the impact
of this process on biogeochemical cycles and C export. Biogeosciences, 14, 3207–3220, 2017 www.biogeosciences.net/14/3207/2017/ 3209 T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment Most of the above-
mentioned studies were performed in microcosms and meso-
cosms and further open-ocean studies combining the set of
complementary approaches described above are needed to
better assess the fate of N2 fixation and its role in C export. The western tropical South Pacific (WTSP) is an ideal lo-
cation to study the fate of N fixed by N2 fixation, as it is
considered a hot spot of N2 fixation in the world ocean (Bon- When
going
eastward
towards
the
South
Pacific
gyre (GY), Halm et al. (2012) reported rates of 12–
190 µmol N m−2 d−1 on the western border of the gyre,
and Raimbault and Garcia (2008) and Moutin et al. (2008)
reported rates of 60 ± 30 µmol N m−2 d−1 in the central
gyre, indicating a decreasing gradient of N2 fixation from
west to east and low N2 fixation rates relative to other ocean
gyre ecosystems. The organisms responsible for these fluxes
are different from common autotrophic diazotrophs such
as Trichodesmium or UCYN-B, and are mainly affiliated
with heterotrophic proteobacteria and low abundances of
UCYN-A (Bonnet et al., 2008; Halm et al., 2012). The west to east zonal gradient of N2 fixation and the dis-
tinct diversity of N2-fixing organisms along this gradient to-
gether provide a unique opportunity to study how produc-
tion, mineralization, and export of organic matter depends
upon N2 fixation in contrasting oligotrophic regimes (from
oligotrophy to ultra-oligotrophy). Comparisons between dif-
ferent systems along a zonal gradient of trophic status and N2
fixation will provide new insights for identifying and under-
standing fundamental interactions between marine biogeo-
chemical C, N, P, silica (Si), and iron (Fe) cycles in olig-
otrophic ecosystems. The western tropical South Pacific (WTSP) is an ideal lo-
cation to study the fate of N fixed by N2 fixation, as it is
considered a hot spot of N2 fixation in the world ocean (Bon-
net et al., 2017a). While average N2 fixation rates range from Biogeosciences, 14, 3207–3220, 2017 www.biogeosciences.net/14/3207/2017/ 3
Objectives of OUTPACE images. The regions along the vessel route were first char-
acterized on a large scale through the analysis of satellite
(altimetry, SST, ocean color) data. These data were auto-
matically retrieved and processed to derive Eulerian and La-
grangian diagnostics of ocean circulation and biogeochem-
istry: Okubo–Weiss parameter maps, Lagrangian coherent
structures (LCSs), and chl a maps (d’Ovidio et al., 2015). The overall goal of OUTPACE was to obtain a precise repre-
sentation of the complex interactions between planktonic or-
ganisms and the cycle of biogenic elements (C, N, P, Si, Fe),
considering a variety of scales, from single-cell processes to
the whole WTSP Ocean. To meet this goal, the three specific
objectives of OUTPACE were The satellite data were treated in near real time on land and
the obtained data were transmitted onboard together with a
daily bulletin containing the analysis of remote sensing in-
formation, along with suggestions for LD station positions
(the complete series of the 42 bulletins is available on the
OUTPACE website at https://outpace.mio.univ-amu.fr, sec-
tion “Adaptive Strategy”). 1. to perform a zonal characterization of the biogeochem-
istry and biological diversity of the WTSP during the
strongest stratified period (austral summer). 2. to study the production and fate of organic matter (in-
cluding C export) of three contrasting environments
(from oligotrophy to ultra-oligotrophy) with a particu-
lar emphasis on N2 fixation. Two main criteria were adopted in suggesting LD station
positions: 1. The areas for the LD A and LD B (LD C) stations were
characterized by local maxima (minima) of sea surface
chl a concentration to sample MA (GY) conditions. 3. to obtain a representation of the main biogeochemical
fluxes (C, N, P, Si, Fe) and the dynamics of the plank-
tonic trophic network, both in situ and by using micro-
cosm experiments. 2. Local minima of surface current intensity for all LD sta-
tions were considered to increase the chance of sam-
pling a homogeneous water mass. The detailed study of biological production and its sub-
sequent fate at a given site implied a combination of adap-
tive and Lagrangian strategies. Indeed, as pointed out by
d’Ovidio et al. (2015), the spatiotemporal domain of an
oceanographic cruise is also one in which horizontal stir-
ring generated by ocean circulation at the mesoscale in-
duces strong variability in biogeochemical tracer distribu-
tions. T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment 3210 3
Objectives of OUTPACE Consequently, ephemeral and local gradients due to
mesoscale activity can easily mask the large-scale gradients. Following d’Ovidio et al. (2015), this problem can be over-
come by adopting an adaptive sampling strategy (described
in Sect. 4.1.) based on information on sea surface tempera-
ture (SST), chlorophyll a (chl a) concentrations, and currents
provided by satellite products analyzed in real time during
the entire cruise. Once the suggested positions for LD stations were relayed
via the daily bulletin, at one of these locations real-time anal-
ysis of water samples using quantitative polymerase chain
reaction (qPCR) was conducted to measure the abundances
of six groups of diazotrophs (Stenengren et al., 2017). In
this way, we located regions with diazotrophs, while also re-
solving the contrasting role Trichodesmium spp. or UCYN-
dominated diazotroph communities have on biogeochemical
cycles. Finally, the exact locations of the three LD stations were
then determined onboard in real time from a rapid sur-
vey using a moving vessel profiler (MVP) equipped with
conductivity–temperature–depth (CTD) and fluorimeter sen-
sors, accompanied by the hull-mounted thermosalinograph
and acoustic Doppler current profiler. 4
Implementation of the OUTPACE cruise During this rapid survey, it was planned to follow two dif-
ferent sampling routes: a cross of about 40 km each side,
followed by a zigzag route covering an area of 25 km each
side at the center of the cross. This strategy was applied as
planned for the LD A and LD C stations. The OUTPACE cruise was conducted during austral summer
conditions from 18 February to 3 April 2015 aboard the RV
L’Atalante. We performed a ∼4000 km zonal transect from
the north of New Caledonia to the western part of the GY, fi-
nally reaching French Polynesia (Fig. 2). Along this transect,
two types of stations were sampled: 15 short-duration (SD,
8 h) stations dedicated to a large-scale description of biogeo-
chemical and biodiversity gradients, and three long-duration
(LD, 7 days) stations for Lagrangian process studies. The LD A station was performed east of the northern ex-
tremity of New Caledonia in an anticyclonic recirculation
characterized by a relatively high surface chl a concentra-
tion. The LD C station was performed in a cyclonic eddy in
the most oligotrophic part of the OUTPACE transect (GY)
close to the Cook Islands. The severe meteorological conditions due to the develop-
ment of tropical cyclone Pam (a category 5 storm) that hit
Vanuatu, forced us to perform the LD B station at a more
easterly location than initially planned. Satellite imagery al-
lowed for the targeting of a large filament of high surface www.biogeosciences.net/14/3207/2017/ www.biogeosciences.net/14/3207/2017/ 4.1
Adaptive strategy Following the planned adaptive strategy, the initial tran-
sect designed to approximatively follow 19◦S was modified
along-route thanks to the information coming from satellite Biogeosciences, 14, 3207–3220, 2017 www.biogeosciences.net/14/3207/2017/ T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment 3211 (a)
(b)
160
170
180
190
200
210
-25
-20
-15
Latitude
160
170
180
190
200
210
-25
-20
-15
Latitude
Longitude
C
1 2
3 4
5
6
7
8
9 10
11
12
13
14 15
OUTPACE Cruise (18 Feb–3 Apr 2015)
Shiptrack
LD Stations
SD Stations
Reefs/Coastline
Land
C
1 2
3 4
5
6
7
8
9 10
11
12
13
14 15
1
0.32
0.1
0.03
1
0.32
0.1
0.03
Chl a (mg m- 3) Chl a (mg m- 3)
Figure 2. Transect of the OUTPACE cruise superimposed on (a) arithmetic mean surface chl a and (b) quasi-Lagrangian weighted mean
chl a of the WTSP during OUTPACE. The two types of station, short duration (X) and long (+) duration, investigated for a period longer
than 7 days, are indicated. The satellite data are weighted in time by each pixel’s distance from the ship’s average daily position for the entire
cruise. The white line shows the vessel route (data from the hull-mounted acoustic Doppler current profiler (ADCP) positioning system). Coral reefs and coastlines are shown in black, land is grey, and areas of no data are left white. The ocean color satellite products are produced
by CLS with support from CNES. (a)
(b)
160
170
180
190
200
210
-25
-20
-15
Latitude
160
170
180
190
200
210
-25
-20
-15
Latitude
Longitude
C
1 2
3 4
5
6
7
8
9 10
11
12
13
14 15
OUTPACE Cruise (18 Feb–3 Apr 2015)
Shiptrack
LD Stations
SD Stations
Reefs/Coastline
Land
C
1 2
3 4
5
6
7
8
9 10
11
12
13
14 15
1
0.32
0.1
0.03
1
0.32
0.1
0.03
Chl a (mg m- 3) Chl a (mg m- 3) (a) (b) (b) Figure 2. Transect of the OUTPACE cruise superimposed on (a) arithmetic mean surface chl a and (b) quasi-Lagrangian weighted mean
chl a of the WTSP during OUTPACE. The two types of station, short duration (X) and long (+) duration, investigated for a period longer
than 7 days, are indicated. 4.2
General program at each station Every station (Table 1) was investigated from the surface to
2000 m. 4. hauls for phytoplankton and zooplankton sampling with
specific nets; and 4.1
Adaptive strategy The satellite data are weighted in time by each pixel’s distance from the ship’s average daily position for the entire
cruise. The white line shows the vessel route (data from the hull-mounted acoustic Doppler current profiler (ADCP) positioning system). Coral reefs and coastlines are shown in black, land is grey, and areas of no data are left white. The ocean color satellite products are produced
by CLS with support from CNES. attached to the CTD rosette to quantify and visualize
suspended particulate material; chl a concentration close to Niue. Due to the circulation pat-
terns associated with the bloom, the rapid survey strategy was
adapted in order to perform four sections across the structure
(see details in de Verneil et al., 2017a). 2. one 0–500 m CTD cast and bottle sampling using the
trace metal clean SBE 9plus CTD rosette (TM-R)
equipped with 24 teflon-lined GO-FLO bottles to sam-
ple for trace metal analyses; SD station positions were chosen in relation to the LD sta-
tions, so that they were roughly equidistant from each other,
respected territorial waters, and incorporated the changing
conditions during the cruise. 3. optical sensors casts in which integrated measurements
of bio-optical properties and pigments were made with
instruments measuring hyperspectral radiometry in the
UV–visible domain with UV–VIS TriOS spectrora-
diometers, and a MicroPro free-fall profiler (Satlantic)
was used for downward irradiance measurements; 4.2.1
SD stations Each of the 15 SD stations was investigated for 8 h. Specific
operations during the SD station occupation consisted of 5. a profile of turbulence measurements using a VMP1000
equipped with microsensors for temperature and shear
that enable accurate estimates of the eddy diffusion co-
efficient Kz. 1. two 0–200 m CTD casts and Niskin bottle sampling
and one 200–2000 m CTD cast and Niskin bottle sam-
pling using the classical SBE 9plus CTD rosette (C-R)
for measurements of core parameters (dissolved oxy-
gen; dissolved inorganic carbon; total alkalinity; nutri-
ents; chl a; particulate and dissolved organic C, N, P,
and Si) and some more specific ones (for example pri-
mary production rates, N2 fixation rates, and diazotroph
abundance). An underwater vision profiler (UVP) was The specific configurations of the two CTD rosettes
are
available
here:
https://outpace.mio.univ-amu.fr/spip. php?article137. Biogeosciences, 14, 3207–3220, 2017 www.biogeosciences.net/14/3207/2017/ www.biogeosciences.net/14/3207/2017/ www.biogeosciences.net/14/3207/2017/ 3212
T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment
Table 1. Date, location, and general characteristics of the stations investigated along the OUTPACE transect. Distance in kilometers from the
first SD station (SD1). T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment 3212 3212
T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment
Table 1. Date, location, and general characteristics of the stations investigated along the OUTPACE transect. Distance in kilometers from the
first SD station (SD1). 4.2.2
LD stations The following color code was proposed to present data
from the different LD stations: LD A is green, LD B is red,
and LD C is blue. Each of the three LD stations was investigated with a drifting
array (see below) that was deployed for 7 days. A series of
CTD (C-R) casts (0–500 m) were performed every 3 h near
the actual position of the drifting array while numerous spe-
cific operations (see below) were carried out in between CTD
casts. All details regarding the OUTPACE cruise are available on
the OUTPACE website: https://outpace.mio.univ-amu.fr/. The
general
scheduled
and
realized
programs
are
available
here:
http://www.obs-vlfr.fr/proof/php/outpace/
outpace_log_and_basic_files.php. A total of 13 surface velocity program (SVP) drifters an-
chored at 15 m depth were deployed in order to study rela-
tive surface dispersion. The drifters were launched with three
at LD A, six at LD B, and four at LD C. A drifting array
(equipped with three PPS5 sediment traps, current meters,
specific oxygen sensors, and specific high-frequency temper-
ature sensors; please consult https://outpace.mio.univ-amu. fr/spip.php?article75 for details) was then deployed at the
chosen station position to start the process study. The drifting
array was recovered at the end of each LD station occupation,
immediately following the last CTD cast. An additional CTD
cast from the surface to the bottom (5000 m) was undertaken
at the LD B station. www.biogeosciences.net/14/3207/2017/ Distance in kilometers from the
first SD station (SD1). www.biogeosciences.net/14/3207/2017/ Station
CTD cast
Arrival date
Departure date
Latitude
Longitude
Latitude
Longitude
Cumulative
Bottom
(0–2000 m)
(UTC)
(UTC)
(deg min)
(deg min)
(deg)
(deg)
distance (km)
depth (m)
Nouméa
x
19/02/2015 21:00
22
14
166
28
−22,23
166,47
x
0
SD
1
007
21/02/2015 20:00
22/02/2015 09:30
18
0
S
159
54
E
−18.00
159.90
0
4068
SD
2
016
22/02/2015 21:45
23/02/2015 10:00
18
38
S
162
8
E
−18.63
162.12
210
2567
SD
3
020
24/02/2015 03:45
24/02/2015 10:00
19
0
S
164
54
E
−19.00
164.90
488
3252
LD
A
067
25/02/2015 13:00
02/03/2015 23:15
19
13
S
164
41
E
−19.21
164.69
528
3491
SD
4
071
04/03/2015 08:30
04/03/2015 15:15
20
0
S
168
0
E
−20.00
168.00
839
4995
SD
5
075
05/03/2015 05:45
05/03/2015 13:30
22
0
S
170
0
E
−22.00
170.00
1103
4405
SD
6
079
06/03/2015 03:15
06/03/2015 12:30
21
22
S
172
8
E
−21.37
172.13
1303
2509
SD
7
083
07/03/2015 00:15
07/03/2015 09:30
20
44
S
174
16
E
−20.73
174.27
1506
2451
SD
8
087
07/03/2015 21:00
08/03/2015 07:15
20
42
S
176
24
E
−20.70
176.40
1694
2028
SD
9
091
08/03/2015 22:15
09/03/2015 07:45
20
57
S
178
39
E
−20.95
178.65
1927
3864
SD
10
095
10/03/2015 00:00
10/03/2015 07:15
20
28
S
178
31
W
−20.47
181.48
2187
819
SD
11
099
10/03/2015 21:45
11/03/2015 05:15
19
59
S
175
40
W
−19.98
184.33
2449
2234
SD
12
103
11/03/2015 21:00
12/03/2015 05:00
19
29
S
172
50
W
−19.48
187.17
2711
7717
LD
B
151
15/03/2015 12:00
20/03/2015 22:30
18
14
S
170
52
W
−18.24
189.14
2985
4912
SD
13
152 (0–500 m)
21/03/2015 10:30
21/03/2015 11:00
18
12
S
169
4
W
−18.20
190.93
3096
4598
LD
C
199
23/03/2015 12:00
28/03/2015 22:00
18
25
S
165
56
W
−18.42
194.06
3371
5277
SD
14
210
30/03/2015 01:30
30/03/2015 09:15
18
25
S
163
0
W
−18.42
197.00
3640
4998
LD
15
213
31/03/2015 00:30
31/03/2015 08:30
18
16
S
160
0
W
−18.27
200.00
3916
4965
Papeete
02/04/2015 21:00
x
17
34
S
149
36
W
−17.57
210.40
x
0 Table 1. Date, location, and general characteristics of the stations investigated along the OUTPACE transect. Distance in kilometers from the
first SD station (SD1). Table 1. Date, location, and general characteristics of the stations investigated along the OUTPACE transect. 5
General characteristics of the upper water masses in
the WTSP Water characteristics (temperature, salinity, density, chl a flu-
orescence) for the upper 700 m as measured by the C-R are
presented in Fig. 3. The deep CTD casts from all SD and
LD stations are presented subsequent to post-cruise process-
ing using Sea-Bird Seasoft software, adopting the TEOS-10
standard. The upper surface layer (0–30 m) observed during
the OUTPACE transect was characterized by warm waters,
with temperatures between 26.18 and 29.93 ◦C (Fig. 3a), and
relatively low salinity, i.e., absolute salinity between 35.03
and 35.81 g kg−1 (Fig. 3b). Density anomalies of 21.72–
22.91 kg m−3 between 0 and 30 m increased gradually be-
tween 30 and 200 m to reach 24.89–25.38 kg m−3 (Fig. 3c). The salinity increase in subsurface waters (100–200 m) from
187◦W longitude (Fig. 3b) indicated the geographical border
between waters under the influence of the MA (SD1 to SD12)
and waters from the GY (SD13 to SD15, LD C). LD B was
not classified here and required a further analysis (de Verneil
et al., 2017a). The classification between MA and GY waters
will be helpful to describe general biogeochemical and bio- Specific operations during LD stations were identical to
those performed at SD stations, along with other operations
like in situ production measurements, in situ particulate ma-
terial sedimentation measurements, trace metal clean pump-
ing for process experiments, and profiles of current mea-
surements. Finally, at each LD station, a drifting PROVOR-
type Argo float (ProBio equipped with sensors to measure
chl a (fluorescence), chromophoric dissolved organic mat-
ter (CDOM) (fluorescence), PAR, irradiance at three wave-
lengths, backscattering, and dissolved oxygen (optode)) was
deployed (Table 2). Biogeosciences, 14, 3207–3220, 2017 www.biogeosciences.net/14/3207/2017/ outin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment
2. PROVOR Argo floats deployed along the OUTPACE transect. Station
Float number
Deployment location
Date of
Time (UTC)
Closest CTD cast
WMO id. 5
General characteristics of the upper water masses in
the WTSP Latitude
Longitude
deployment
LD A
ProBio075b
19◦13.00 S
164◦29.40 W
03/03/2015
22:45:00
OUT_C_067
SD 11
6901663
19◦59.56 S
175◦38.59 W
11/03/2015
05:03:00
OUT_C_099
SD 12
6901664
19◦32.61 S
172◦46.47 W
12/03/2015
05:00:00
OUT_C_103
LD B
6901666
19◦44.50 S
170◦31.31 W
13/03/2015
01:15:00
LD B
6901667
17◦38.39 S
170◦59.47 W
13/03/2015
18:49:00
LD B
6901668
18◦13.90 S
170◦44.20 W
20/03/2015
21:36:00
OUT_C_151
LD B
ProBio077b
18◦16.09 S
170◦43.80 W
20/03/2015
22:00:00
OUT_C_151
Before LD C
6901669
18◦46.92 S
168◦09.06 W
22/03/2015
07:55:00
LD C
6901670
18◦41.20 S
165◦45.18 W
22/03/2015
19:48:00
LD C
6901671
18◦28.16 S
165◦46.21 W
28/03/2015
21:25:00
OUT_C_199
LD C
ProBio079b
18◦28.16 S
165◦46.21 W
28/03/2015
21:30:00
OUT_C_151
SD 14
6901679
18◦24.26 S
162◦59.34 W
30/03/2015
09:45:00
OUT_C_210
SD 15
6901680
18◦15.29 S
159◦59.23 W
31/03/2015
08:15:00
OUT_C_213
0
100
200
300
400
500
600
700
160
170
180
190
200
210
Longitude
C
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
Chl a fluorescence (mg m )
-3
0
100
200
300
400
500
600
700
160
170
180
190
200
210
Longitude
Depth (m)
C
28
27
26
25
24
23
22
Density anomaly (kg m
-
3- 1000)
0
100
200
300
400
500
600
700
160
170
180
190
200
210
C
36.4
36.2
36
35.8
35.6
35.4
35.2
35
34.8
34.6
-1
Absolute salinity, S A(g kg )
0
100
200
300
400
500
600
700
160
170
180
190
200
210
Conservative temperature, Θ (˚C)
Depth (m)
10
15
20
25
30
C
(a)
(c)
(b)
(d)
e 3. Zonal sections of (a) conservative temperature 2, (b) absolute salinity SA, (c) density anomaly, and (d) fluorescence in th
m from the classic CTD rosette at SD and LD stations along the OUTPACE transect. The three LD stations are highlighted b
coded letter and corresponding arrow. al features. Between 200 and 700 m, a decreasing gra-
of temperature and salinity indicated the presence of
anent thermocline waters Temperature fell from 19 27
depth, considered here as the main indicator of the t
state, increased from ∼100 m depth in the MA wa
∼115 150 m depth in the GY waters which allowed T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment 3213 T. 5
General characteristics of the upper water masses in
the WTSP Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment C
0
100
200
300
400
500
600
700
160
170
180
190
200
210
Conservative temperature, Θ (˚C)
Depth (m)
10
15
20
25
30
C
(a)
(c) C
0
100
200
300
400
500
600
700
160
170
180
190
200
210
C
36.4
36.2
36
35.8
35.6
35.4
35.2
35
34.8
34.6
-1
Absolute salinity, S A(g kg )
(b)
(d) (b) (a) (d) (c) 0
100
200
300
400
500
600
700
160
170
180
190
200
210
Longitude
Depth (m)
C
28
27
26
25
24
23
22
Density anomaly (kg m
-
3- 1000)
(c) 0
100
200
300
400
500
600
700
160
170
180
190
200
210
Longitude
C
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
Chl a fluorescence (mg m )
-3
(d) Figure 3. Zonal sections of (a) conservative temperature 2, (b) absolute salinity SA, (c) density anomaly, and (d) fluorescence in the upper
700 m from the classic CTD rosette at SD and LD stations along the OUTPACE transect. The three LD stations are highlighted by their
color-coded letter and corresponding arrow. depth, considered here as the main indicator of the trophic
state, increased from ∼100 m depth in the MA waters to
∼115–150 m depth in the GY waters, which allowed us to
sample the oligotrophic to ultra-oligotrophic transition in the
WTSP for the purpose of the OUTPACE project. logical features. Between 200 and 700 m, a decreasing gra-
dient of temperature and salinity indicated the presence of
permanent thermocline waters. Temperature fell from 19.27
to 22.08 ◦C at 200 m to 5.48–6.91 ◦C at 700 m (Fig. 3a). Ab-
solute salinities of 35.76–36.21 g kg−1 at 200 m decreased
to 34.49–34.61 g kg−1 (Fig. 3b). The density anomalies of
24.89–25.38 kg m−3 at 200 m increased largely to 26.97–
27.13 kg m−3 at 700 m (Fig. 3c). The maximum fluorescence Biogeosciences, 14, 3207–3220, 2017 6
The climatological context of the campaign This resulted
in a distribution of standard deviations, the median of which
was taken to represent inherent interannual variability. The time series of SOI is presented in Fig. 4, with El Niño
(La Niña) values shaded red (blue). During OUTPACE cruise
sampling in austral summer 2015, a strong El Niño was ob-
served. The other El Niño year considered here was 2003,
chosen because of its relative strength and duration similar
to 2015; for similar reasons of intensity and duration, 2011
was designated a representative La Niña year. WTSP conditions were estimated with SST and surface
chl a concentration measured by the MODIS Aqua satel-
lite mission and available at the NASA Ocean Color Data
website (https://oceandata.sci.gsfc.nasa.gov/, downloaded 14
December 2016). Global annual and monthly (March) av-
erages of both SST and chl a were provided by NASA at
level 3 (i.e., mapped) with 4 km satellite pixel resolution, re-
sulting in four separate datasets from 2003 to present. The
data within the OUTPACE region, defined between 25◦S,
155◦E, and 15◦S, 149◦W, to envelope the cruise track as in
Fig. 2, were extracted. March was chosen as the month of
study because it was the central month of the cruise. Plots of
these four datasets can be found on the OUTPACE dataset
website (http://www.obs-vlfr.fr/proof/php/outpace/outpace_
figures.php). Probability density distributions were generated
for each dataset in its entirety. Additional probability distri-
butions were also calculated on data subsets, for years 2003
and 2011 as chosen by SOI to represent opposite ENSO
phases, along with 2015, the year of the cruise. In order to
gauge significance between probability distributions of dif-
ferent years, the temporal standard deviation of each pixel The probability density distributions for the annual and
March means of both SST and chl a are presented in Fig. 5. In
SST, the March mean showed a larger proportion of warmer
temperatures in relation to the annual mean (white lines in
Fig. 5a, c), reflecting the austral summer season. The 2011
La Niña distributions showed a greater proportion of warm
temperatures in both annual and March distributions (blue
lines in Fig. 5a, c), which is consistent with the idea of La
Niña accentuating the warm pool present in the WTSP. The
probability peak of March 2011 SST was much more local-
ized, between 26 and 30 ◦C, than 2003, 2015, and the mean
(Fig. 5c). 6
The climatological context of the campaign 2000
2005
2010
2015
2020
-4
-3
-2
-1
0
1
2
3
Year
SOI index
OUTPACE: aka the most important
cruise EVER! 2003
2011
2015 The OUTPACE cruise took place in the WTSP, a region im-
pacted by the El Niño–Southern Oscillation (ENSO), known
to be the most important mode of SST variability on inter-
annual to decadal timescales (Sarmiento and Gruber, 2006). ENSO-related SST anomalies are caused by a combina-
tion of changes in ocean circulation (mainly changes in the
strength and source of equatorial upwelling) and anomalous
local air–sea heat exchanges. The most dramatic effects of
ENSO in the surface ocean are well documented in the east-
ern tropical Pacific, where seasonal upwelling conditions can
be suppressed with severe economic consequences for fish-
eries (Chavez et al., 2003). During El Niño phases (negative
Southern Oscillation Index (SOI), defined below), the warm
pool normally positioned in the western Pacific is found far-
ther east, resulting in the aforementioned suppression of up-
welling conditions off Peru. La Niña, the opposite phase with
a positive SOI, reverses this situation. After decades of in-
tense study, ENSO is still an active field of research (Taka-
hashi et al., 2011). Figure 4. Time series of the monthly Southern Oscillation Index
(SOI) from January 2000 to October 2016. Negative and positive
values are shaded red and blue to signify El Niño and La Niña,
respectively. SOI smoothed by a lowess filter with a five-point win-
dow is shown as a solid black line. Black arrows indicate March
for the years when satellite data are available. Dashed vertical lines
indicate March 2003, 2011, and 2015. Given the known importance of ENSO for the tropical
South Pacific, it is worthwhile to determine in which cli-
matological conditions the cruise was performed. To achieve
this goal, we identified years of opposing ENSO phases and
analyzed the corresponding WTSP conditions with avail-
able satellite data. ENSO phases were identified using the
monthly time series of SOI provided by NCEP (http://www. cpc.ncep.noaa.gov/, downloaded 5 December 2016). The
SOI metric uses differences in standardized sea level air pres-
sure between Papeete (Tahiti) and Darwin (Australia) to rep-
resent the ENSO phase, as previously mentioned. The OUT-
PACE region lies between these two locations (Fig. 2), high-
lighting again ENSO’s possible influence on the cruise. was calculated for each of the four datasets. www.biogeosciences.net/14/3207/2017/ www.biogeosciences.net/14/3207/2017/ T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment 3214 Oligotrophy to the UlTra-oligotrophy PACific Experiment
2000
2005
2010
2015
2020
-4
-3
-2
-1
0
1
2
3
Year
SOI index
OUTPACE: aka the most important
cruise EVER! 2003
2011
2015
Figure 4. Time series of the monthly Southern Oscillation Index
(SOI) from January 2000 to October 2016. Negative and positive
values are shaded red and blue to signify El Niño and La Niña,
respectively. SOI smoothed by a lowess filter with a five-point win-
dow is shown as a solid black line. Black arrows indicate March
for the years when satellite data are available. Dashed vertical lines
indicate March 2003, 2011, and 2015. www.biogeosciences.net/14/3207/2017/ 6
The climatological context of the campaign The 2003 March SST (red line in Fig. 5c) was the
widest distribution, with a slight rightward skew. By contrast,
both the mean and 2015 March probabilities (white and green
lines, respectively) had a left skew. Considering the median
95 % confidence interval for the March SST time series, the
central peaks for these distributions could not be considered
distinct. www.biogeosciences.net/14/3207/2017/ Biogeosciences, 14, 3207–3220, 2017 3215 T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment 20
22
24
26
28
30
32
0
0.1
0.2
0.3
0.4
0.5
0.6
0
0.1
0.2
0.3
0.4
0.5
0.6
Probability density
SST (˚C)
Legend
Mean
2003
2011
2015
0.5
1
1.5
2
2.5
3
3.5
0
0.5
1
1.5
2
2.5
3
3.5
0
-1.5
-1
-0.5
Log 10 Chl a (mg m )
-3
Annual
March
Chl a
(a)
SST
(b)
(d)
(c)
Figure 5. Probability density estimates for MODIS Aqua data in the OUTPACE region, using mean (a) annual SST, (b) annual chl a, (c)
March SST, and (d) March chl a. Probability densities for the ensemble of all years is shown in white, while densities for 2003, 2011, and
2015 are in red, blue, and green, respectively. Also plotted are 95 % confidence intervals (2 standard deviations) for each subset, estimated
using the median variance of pixel interannual variability. Note the logarithmic scale for chl a. nd
Mean
2003
2011
2015
0.5
1
1.5
2
2.5
3
3.5
0
0.5
1
1.5
2
2.5
3
3.5
0
-1.5
-1
-0.5
Log 10 Chl a (mg m )
-3
Chl a
(b)
(d) Figure 5. Probability density estimates for MODIS Aqua data in the OUTPACE region, using mean (a) annual SST, (b) annual chl a, (c)
March SST, and (d) March chl a. Probability densities for the ensemble of all years is shown in white, while densities for 2003, 2011, and
2015 are in red, blue, and green, respectively. Also plotted are 95 % confidence intervals (2 standard deviations) for each subset, estimated
using the median variance of pixel interannual variability. Note the logarithmic scale for chl a. Satellite chl a showed distributions completely different
from SST. For all quantities considered, the annual and
March means (Fig. 6
The climatological context of the campaign 5b, d), as well as the years 2003, 2011,
and 2015, showed a bimodal distribution, where the two
peaks were distinct enough from each other to be signifi-
cant from the median pixel variability. Interestingly, 2003
and 2011 (red and blue lines, Fig. 5b) annual chl a dis-
tributions overlapped considerably more than 2015 (green
line), which had its entire probability distribution shifted to
the right, though this shift may not have been significant. The March mean for 2003 and 2015 chl a (red and green
lines, respectively, Fig. 5d), however, almost entirely over-
lapped, possibly signifying that El Niño chl a distributions
were more alike than La Niña years. The annual 2015 mean
of chl a being slightly different from other years, and yet the
March 2015 mean resembling 2003, might indicate that 2015
was indeed a slightly different year from a surface chloro-
phyll perspective, but this slight difference was concentrated
in other parts of the year than late austral summer when OUT-
PACE took place. thus, we determined that climatological effects upon the re-
sults of OUTPACE were minimized. www.biogeosciences.net/14/3207/2017/ T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment They also found discrepan-
cies between the UCYN-A and their hosts in both abundance
and distribution, which suggests that the UCYN-A could be
living freely or with a wider diversity of hosts than previ-
ously believed. Finally, the assemblage of epibiotic microor-
ganisms associated with Trichodesmium were characterized
in relation with environmental parameters (Frischkorn et al.,
2017). N2 fixation in the ocean does not only occur in tropical
sunlit surface waters but also in less obvious environments
such as temperate latitudes and aphotic waters. Here, N2 fix-
ation was also measured in the mesopelagic zone along the
OUTPACE transect and the diazotroph community present
were identified. Deep N2 fixation rates were low but mea-
surable and recurrently found along the transect, with the
exception of the easternmost stations located in the ultra-
oligotrophic subtropical Pacific gyre (Benavides et al., 2017). N2 fixation activity was presumably driven by the domi-
nating gamma-proteobacterial community and fueled by the
presence of labile organic matter compounds. Benavides et g
p
g
The dynamics of phytoplankton (Bock et al., 2017;
Leblanc et al., 2017; Lefevre et al., 2017; Guidi, 2017), het-
erotrophic bacterioplankton (Van Wambeke et al., 2017), and
zooplankton (Carlotti et al., 2017; Hunt et al., 2017) along
the zonal gradient of diazotroph diversity and activity are
described together with the composition and distribution of
dissolved organic carbon (Panagiotopoulos et al., 2017) and
the changes in inorganic carbon content along the longitu-
dinal transect (Wagener et al., 2017). Stations sampled dur-
ing the OUTPACE cruise were characterized by a highly
stratified community structure, with significant contributions
of Prochlorococcus and picophytoeukaryote populations to
biomass (Bock et al., 2017). Size-fractionated results show a
non-negligible contribution of the pico-sized fraction (< 2–
3 µm) to both Si biomass and uptake, which could confirm
the previous hypothesis of Si assimilation by Synechococcus
populations or reflect the presence of an overlooked Si group
such as Parmales (Leblanc et al., 2017). Surface DOC con-
centrations varied little (50–75 µM) across the transect, with
slightly higher values observed at LD B (78 µM), and labile
organic matter (sugars were used as a good proxy) closely
followed DOC patterns ranging from 1.5 to 3 µM with higher
values also recorded at LD B (3.5 µM). Labile organic matter
accounted for about 3–5 % of DOC, with glucose being the
dominant sugar (> 60 % of total sugars) (Panagiotopoulos et
al., 2017). T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment 3216 al. (2017) provided further evidence that N2 fixation in the
deep ocean is not negligible and likely impacts global nitro-
gen inputs in a significant manner. N2 fixation was detected at all 18 sampled stations and
the transect could be divided into two main characteris-
tic subareas (Bonnet et al., 2017b): (i) the MA waters
(160◦W to 170◦E) exhibiting very high N2 fixation rates
(631 ± 286 µmol N m−2 d−1, i.e., among the highest reported
for the global ocean; Luo et al., 2012) and dominated by Tri-
chodesmium (Stenengren et al., 2017), and (ii) the GY wa-
ters (170–160◦E) exhibiting low N2 fixation rates (average
85 ± 79 µmol N m−2 d−1) dominated by UCYN-B (Stenen-
gren et al., 2017). The differing δ15N signature of suspended
particles measured over the photic layer of MA (−0.41 ‰)
and GY waters (8.06 ‰) confirms the presence of two con-
trasting regions in terms of N2 fixation. Thanks to the La-
grangian strategy followed at the LD stations and the low dis-
persion measured, showing that we sampled the same water
masses (de Verneil et al., 2017b), N budgets were established
(Caffin et al., 2017b). N2 fixation was the major external
source of N, representing more than 90 % of new N input into
the photic layer, and the e-ratio quantifying the efficiency of
a system to export particulate organic matter was higher in
MA waters than in GY waters (Caffin et al., 2017b). Caf-
fin et al. (2017a) revealed that the diazotroph-derived nitro-
gen (DDN) was efficiently transferred from diazotrophs (Tri-
chodesmium and UCYN) to non-diazotrophic phytoplankton,
both autotrophs and heterotrophs. Hunt et al. (2017) report an
efficient transfer of DDN in zooplankton. The fate of C and
N was under the influence of programmed cell death in dia-
zotrophs (Berman-Frank et al., 2017) but diazotrophs were
poorly exported directly, and we suspect that this transfer
of DDN fueled indirect export associated with N2 fixation. By using nitrogen isotope budgets, Knapp et al. (2017) con-
firmed that > 50 % of export production was supported by
N2 fixation in MA waters. Stenegren et al. (2017) identified a
clear niche separation between a subsurface (UCYN-A1 and
A2 with their hosts) and a surface group (Trichodesmium,
UCYN-B, and the heterotrophic group, het-group) based on
a temperature–depth gradient. 7
Special issue presentation The goal of this special issue is to present the knowledge ob-
tained concerning the functioning of WTSP ecosystems and
associated biogeochemical cycles based on the datasets ac-
quired during the OUTPACE experiment. The cruise strategy
was organized to promote collaboration between physicists,
biologists, and biogeochemists with expertise including ma-
rine physics, chemistry, optics, biogeochemistry, microbiol-
ogy, molecular ecology, genetics, and modeling. Most of the
contributions to this volume have benefited from this collec-
tive effort and are presented according to the main objectives
of the OUTPACE experiment. The hydrological and dynamical context of biogeochemi-
cal sampling is described for the entire cruise route (Fume-
nia et al., 2017) and specifically at the three long-duration
stations, where low physical variability validated the quasi-
Lagrangian sampling strategy employed (de Verneil et al.,
2017b). Turbulence measurements revealed an interesting
longitudinal gradient with higher turbulence levels in the
west, i.e., the Coral Sea, compared to the eastern part
within the gyre, consistent with the increasing oligotrophy
(Bouruet-Aubertot et al., 2017). The large-scale circulation
was dominant even though the mesoscale and sub-mesoscale
circulation can have a strong influence (Rousselet et al.,
2017), in particular on the bloom observed at station LD B
(de Verneil et al., 2017a). In summary, the SOI metric of ENSO showed that 2015
was an El Niño event, and both the SST and chl a satellite
data in the WTSP partially reflected this. The March 2003
and 2015 SST distributions resembled each other, but they
also resembled the entire dataset more than the 2011 La Niña
distribution. Looking at SST, perhaps La Niña events impact
the region more than El Niño. Chl a distributions for the El
Niño years also overlapped more than with La Niña, but the
variability inherent in the time series precluded the decla-
ration of significant differences. Overall, in the WTSP both
SST and chl a were not atypical from what one would expect; An important focus of OUTPACE was on dinitrogen fix-
ation and its fate in the ecosystem (Caffin et al., 2017a, b). www.biogeosciences.net/14/3207/2017/ Biogeosciences, 14, 3207–3220, 2017 Competing interests. The authors declare that they have no conflict
of interest. Competing interests. The authors declare that they have no conflict
of interest. Special issue statement. This article is part of the special issue “In-
teractions between planktonic organisms and biogeochemical cy-
cles across trophic and N2 fixation gradients in the western tropical
South Pacific Ocean: a multidisciplinary approach (OUTPACE ex-
periment)”. It does not belong to a conference. Bock, N., Dion, M., Van Wambeke, F., and Duhamel, S.: Picophy-
toplankton Community Structure in the Western Tropical South
Pacific During Austral Summer, Biogeosciences, in preparation,
2017. Bonnet, S., Guieu, C., Bruyant, F., Prášil, O., Van Wambeke, F.,
Raimbault, P., Moutin, T., Grob, C., Gorbunov, M. Y., Zehr, J. P.,
Masquelier, S. M., Garczarek, L., and Claustre, H.: Nutrient limi-
tation of primary productivity in the Southeast Pacific (BIOSOPE
cruise), Biogeosciences, 5, 215–225, https://doi.org/10.5194/bg-
5-215-2008, 2008. Acknowledgements. This is a contribution of the OUTPACE
(Oligotrophy from Ultra-oligoTrophy PACific Experiment) project
(https://outpace.mio.univ-amu.fr/) funded by the French research
national agency (ANR-14-CE01-0007-01), the LEFE-CYBER
program (CNRS-INSU), the GOPS program (IRD), and the
CNES (BC T23, ZBC 4500048836). The OUTPACE cruise
(https://doi.org/10.17600/15000900) was managed by MIO (OSU
Institut Pythéas, AMU) from Marseilles (France). The authors
thank the crew of the RV L’Atalante for outstanding on-ship
operations. Gilles Rougier and Marc Picheral are warmly thanked
for their efficient help in CTD rosette management and data
processing, as well as Catherine Schmechtig for the LEFE-CYBER
database management. The satellite-derived data of sea surface
temperature, chl a concentrations, and currents have been provided
by CLS in the framework of the CNES funding; we warmly thank
Marie Isabelle Pujol and Guillaume Taburet for their support in
providing these data. We acknowledge NOAA, and in particular
Rick Lumpkin, for providing the SVP drifters. Bonnet, S., Biegala, I. C., Dutrieux, P., Slemons, L. O., and Capone,
D. G.: Nitrogen fixation in the western equatorial Pacific: Rates,
diazotrophic cyanobacterial size class distribution, and biogeo-
chemical significance, Global Biogeochem. Cy., 23, 1–13, 2009. Bonnet, S., Rodier, M., Turk-Kubo, K., Germineaud, C., Menkes,
C., Ganachaud, A., Cravatte, S., Raimbault, P., Campbell, E.,
Quéroué, F., Sarthou, G., Desnues, A., Maes, C., and Eldin,
G.: Contrasted geographical distribution of N2 fixation rates and
nifH phylotypes in the Coral and Solomon Seas (South-Western
Pacific) during austral winter conditions, Global Biogeochem. Cy., 29, 1874–1892, https://doi.org/10.1002/2015GB005117,
2015. References Altabet, M. A.: Variations in Nitrogen Isotopic Composition be-
tween Sinking and Suspended Particles – Implications for Ni-
trogen Cycling and Particle Transformation in the Open Ocean,
Deep-Sea Res., 35, 535–554, 1988. Bonnet, S., Berthelot, H., Turk-Kubo, K., Fawcett, S., Rahav, E.,
L’Helguen, S., and Berman-Frank, I.: Dynamics of N2 fixa-
tion and fate of diazotroph-derived nitrogen in a low-nutrient,
low-chlorophyll ecosystem: results from the VAHINE mesocosm Antoine, D., Andre, J. M., and Morel, A.: Oceanic primary pro-
duction: 2. Estimation at global scale from satellite (coastal zone Biogeosciences, 14, 3207–3220, 2017 Competing interests. The authors declare that they have no conflict
of interest. Bonnet, S., Baklouti, M., Gimenez, A., Berthelot, H., and Berman-
Frank, I.: Biogeochemical and biological impacts of diazotroph
blooms in a low-nutrient, low-chlorophyll ecosystem: synthe-
sis from the VAHINE mesocosm experiment (New Caledonia),
Biogeosciences, 13, 4461–4479, https://doi.org/10.5194/bg-13-
4461-2016, 2016a. Edited by: Emilio Marañón
Reviewed by: two anonymous referees Edited by: Emilio Marañón
Reviewed by: two anonymous referees Bonnet, S., Berthelot, H., Turk-Kubo, K., Cornet-Barthaux,
V., Fawcett, S., Berman-Frank, I., Barani, A., Gregori, G.,
Dekaezemacker, J., Benavides, M., and Capone, D. G.: Dia-
zotroph derived nitrogen supports diatom growth in the South
West Pacific: A quantitative study using nanoSIMS, Limnol. Oceanogr., 61, 1549–1562, 2016b. T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment Valdes et al. (2017) suggest that copepods can re-
tain N and P compounds obtained from feeding in the upper
layer, preventing the rapid loss of these nutrients. Copepods
were able to sustain and modify the composition of microbial
communities and could provide P for further development of
cyanobacterial blooms. Optical properties of the WTSP waters are presented, with
a focus on the cyanobacterial (diazotroph) impact upon bio-
optical properties, UV–VIS light attenuation (Dupouy et al.,
2017), and chl a algorithms (Frouin et al., 2017). Opera-
tional NASA bio-optical algorithms (OC4v6; color index,
CI) substantially underestimated surface chl a concentration,
but a normalized reflectance difference index, robust to at-
mospheric correction errors, performed well over the range
of chl a values encountered across the transect (Frouin et
al., 2017). Trichodesmium is considered the main nitrogen-
fixing species, especially in the South Pacific region. Due to
the paucity of in situ observations, alternative methods for
estimating the presence of Trichodesmium must be sought to
evaluate the global impact of these species on primary pro-
duction. Rousset et al. (2017) elaborate a new satellite-based
algorithm and use that algorithm to estimate the extent of Tri-
chodesmium surface blooms and their dynamics during the
OUTPACE experiment. Finally, the main processes control-
ling the biological carbon pump in the WTSP were investi-
gated using a unidimensional vertical (Gimenez et al., 2017)
and regional (Dutheil et al., 2017) biogeochemical–physical Biogeosciences, 14, 3207–3220, 2017 www.biogeosciences.net/14/3207/2017/ T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment 3217 color scanner) chlorophyll, Global Biogeochem. Cy., 10, 57–69,
1996. coupled models. The new knowledge gained on the interac-
tions between planktonic organisms and the cycle of biogenic
elements is then used to propose a new scheme for the biolog-
ical carbon pump function and its role, at the present time and
in the near future, in the oligotrophic Pacific Ocean (Moutin
et al., 2017). Benavides, M. and Voss, M.: Five decades of N2 fixation research
in the North Atlantic Ocean, Frontiers in Marine Science, 2, 40,
https://doi.org/10.3389/fmars.2015.00040, 2015. Benavides, M., Moisander, P. H., Dittmar, T., Berthelot, H., Grosso,
O., and Bonnet, S.: Aphotic N2 fixation is related to labile or-
ganic matter in the Western Tropical South Pacific, Biogeo-
sciences, in preparation, 2017. Data availability. All data and metadata are available at the French
INSU/CNRS LEFE CYBER database (Scientific Coordinator:
Hervé Claustre, Data Manager, Webmaster : Catherine Schmechtig)
at the following web address: http://www.obs-vlfr.fr/proof/php/
outpace/outpace.php (INSU/CNRS LEFE CYBER, 2017). Berman-Frank, I., Spungin, D., Belkin, N., Van-Wambeke, F.,
Gimenez, A., Caffin, M., Stengren, M., Foster, R., Knapp, A.,
and Bonnet, S.: Programmed cell death in diazotrophs and the
fate of C and N in the Western Tropical South Pacific, Biogeo-
sciences, in preparation, 2017. Berthelot, H., Moutin, T., L’Helguen, S., Leblanc, K., Hélias, S.,
Grosso, O., Leblond, N., Charrière, B., and Bonnet, S.: Dinitro-
gen fixation and dissolved organic nitrogen fueled primary pro-
duction and particulate export during the VAHINE mesocosm
experiment (New Caledonia lagoon), Biogeosciences, 12, 4099–
4112, https://doi.org/10.5194/bg-12-4099-2015, 2015. Edited by: Emilio Marañón
Reviewed by: two anonymous referees T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment Dutheil, C., Menkes, C., Aumont, O., Lorrain, A., Bonnet, S.,
Rodier, M., Shiozaki, T., and Dupouy, C.: Impact of Tri-
chodesmium sp. on Pacific primary production, Biogeosciences,
in preparation, 2017. Caffin, M., Foster, R., Berthelot, H., Stenegren, M., Caputo, A.,
Berntzo, L., and Bonnet, S.: Fate of N2 fixation in the Western
Tropical South Pacific Ocean: Transfert of diazotroph-derived
nitrogen to non-diazotrophic communities and export of dia-
zotrophs, Biogeosciences, in preparation, 2017a. Falkowski, P. G., Barber, R. T., and Smetacek, V.: Biogeochemical
Controls and Feedbacks on Ocean Primary Production, Science,
281, 200–206, 1998. Caffin, M., Moutin, T., Bouruet-Aubertot, P., Doglioli, A. M., Berth-
elot, H., Grosso, O., Helias-Nunige, S., Leblond, N., Gimenez,
A., de Verneil, A., and Bonnet, S.: Nitrogen budget in the photic
layer of the Western Tropical South Pacific (WTSP) Ocean: ev-
idence of high nitrogen fixation rates, Biogeosciences, in prepa-
ration, 2017b. Frischkorn, K. R., Krupke A., Dyhrman, S. T., and Van Mooy,
B. A. S.: Quorum sensing influences Trichodesmium consortia
physiology in the oligotrophic South Pacific, Biogeosciences, in
preparation, 2017. Frouin, R., Dupouy, C., and Tan, J.: Chlorophyll-a algorithms for
South Pacific oligotrophic waters, Biogeosciences, in prepara-
tion, 2017. Campbell, L., Carpenter, E. J., Montoya, J. P., Kustka, A. B.,
and Capone, D. G.: Picoplankton community structure within
and outside a Trichodesmium bloom in the southwestern Pacific
Ocean, Vie Milieu., 55, 185–195, 2005. Fumenia, A., Moutin, T., Petrenko, A., Doglioli, A. M., de Verneil,
A., and Maes, C.: Optimum multiparameter analysis of the wa-
ter mass structure in the western Tropical South Pacific Ocean
during OUTPACE, Biogeosciences, in preparation, 2017. Carlotti,F., Pagano, M., Guilloux, L., Donoso, K., and Valdes, V.:
Mesozooplankton structure and functioning across the Western
Tropical South Pacific Ocean, Biogeosciences, in preparation,
2017. Galbraith, E. D., Kienast, M., Albuquerque, A. L., Altabet, M. A.,
Batista, F., Bianchi, D., Calvert, S. E., Contreras, S., Crosta, X.,
De Pol-Holz, R., Dubois, N., Etourneau, J., Francois, R., Hsu,
T. C., Ivanochko, T., Jaccard, S. L., Kao, S. J., Kiefer, T., Kien-
ast, S., Lehmann, M. F., Martinez, P., McCarthy, M., Meckler,
A. N., Mix, A., Mobius, J., Pedersen, T. F., Pichevin, L., Quan,
T. M., Robinson, R. S., Ryabenko, E., Schmittner, A., Schneider,
R., Schneider-Mor, A., Shigemitsu, M., Sinclair, D., Somes, C.,
Studer, A. S., Tesdal, J. E., Thunell, R., Yang, J. Y. T., and Mem-
bers, N. W. T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment 3218 Duration Stations: Physical variability and context of biogeo-
chemical sampling, Biogeosciences, in preparation, 2017b. Duration Stations: Physical variability and context of biogeo-
chemical sampling, Biogeosciences, in preparation, 2017b. experiment (New Caledonia), Biogeosciences, 13, 2653–2673,
https://doi.org/10.5194/bg-13-2653-2016, 2016c. Bonnet, S., Berthelot, H., Caffin, M., and Moutin, T.: A hot
spot of N2 fixation in the western tropical South Pacific
pleads for a spatial decoupling between N2 fixation and den-
itrification, P. Natl. Acad. Sci. USA, 114, E2800–E2801,
https://doi.org/10.1073/pnas.1619514114, 2017a. Dore, J., Brium, J. R., Tupas, L. M., and Karl, D. M.: Seasonal
and interannual variability in sources of nitrogen supporting ex-
port in the oligotrophic subtropical North Pacific Ocean, Limnol. Oceanogr., 47, 1595–1607, 2002. d’Ovidio, F., Della Penna, A., Trull, T. W., Nencioli, F., Pu-
jol, M.-I., Rio, M.-H., Park, Y.-H., Cotté, C., Zhou, M., and
Blain, S.: The biogeochemical structuring role of horizon-
tal stirring: Lagrangian perspectives on iron delivery down-
stream of the Kerguelen Plateau, Biogeosciences, 12, 5567–
5581, https://doi.org/10.5194/bg-12-5567-2015, 2015. Bonnet S., Caffin, M., Berthelot, H., Grosso, O., Guieu, C., and
Moutin, T.: Contribution of dissolved and particulate fractions
to the Hot Spot of N2 fixation in the Western Tropical South Pa-
cific Ocean (OUTPACE cruise), Biogeosciences, in preparation,
2017b. Dupouy, C., Benielli-Gary, D., Neveux, J., Dandonneau, Y., and
Westberry, T. K.: An algorithm for detecting Trichodesmium
surface blooms in the South Western Tropical Pacific, Bio-
geosciences, 8, 3631–3647, https://doi.org/10.5194/bg-8-3631-
2011, 2011. Böttjer, D., Dore, J. E., Karl, D. M., Letelier, R. M., Mahaffey, C.,
Wilson, S. T., Zehr, J., and Church, M. J.: Temporal variabil-
ity of nitrogen fixation and particulate nitrogen export at Station
ALOHA, Limnol. Oceanogr., 62, 200–216, 2017. Bouruet-Aubertot, P., Cuypers, Y., Le Goff, H., Rougier, G., de
Verneuil, A., Doglioli, A. M., Picheral, M., Yohia, C., Caffin,
M., Lefèvre, D., Petrenko, A., and Moutin, T.: Longitudinal con-
trast in small scale turbulence along 20◦S in the Pacific Ocean:
origin and impact on biogeochemical fluxes, Biogeosciences, in
preparation, 2017. Dupouy, C., Frouin, R., Maillard, M., Tedetti, M., Charriere, B.,
Roettgers, R., Martias, C., Rodier, M., Pujo-Pay, M., Duhamel,
S., Guidi, L., and Sempere, R.: Influence of cyanobacteria on
Longitudinal and short-term variations in UV-Vis optical prop-
erties in the Western Tropical South Pacific Ocean, Biogeo-
sciences, in preparation, 2017. Broecker, W. S.: Keeping global change honest, Global Bio-
geochem. Cy., 1, 15–29, 1991. T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment Hunt, B., Bonnet, S., Carlotti, F., Donoso, K., Pagano, M., and
Moutin, T.: Diazotroph derived nitrogen contribution to zoo-
plankton biomass in the south west Pacific, Biogeosciences, in
preparation, 2017. C. M., Mouriño-Carballido, B., Mulholland, M. R., Needoba, J. A., Orcutt, K. M., Poulton, A. J., Rahav, E., Raimbault, P., Rees, INSU/CNRS LEFE CYBER: OUTPACE – Oligotrophy to UlTra-
oligotrophy PACific Experiment, available at: http://www. obs-vlfr.fr/proof/php/outpace/outpace.php, last access: 5 July
2017. Jackson, G. A.: Effect of coagulation on a model planktonic food
web, Deep-Sea Res. Pt. I, 48, 95–123, 2001. McMahon, K. W., McCarthy, M. D., Sherwood, O. A., Larsen, T.,
and Guilderson, T. P.: Millennial-scale plankton regime shifts in
the subtropical North Pacific Ocean, Science, 350, 1530–1533,
2015. Karl, D. M.: Microbially mediated transformations of phosphorus
in the sea: new views of an old cycle, Ann. Rev. Mar. Sci., 6,
279–337, 2014. Karl, D. M., Letelier, R., Tupas, L., Dore, J., Christian, J., and
Hebel, D.: The role of nitrogen fixation in biogeochemical cy-
cling in the subtropical North Pacific Ocean, Nature, 388, 533–
538, 1997. Moisander, P. H., Beinart, R. A., Hewson, I., White, A. E., Johnson,
K. S., Carlson, C. A., Montoya, J. P., and Zehr, J. P.: Unicellular
Cyanobacterial Distributions Broaden the Oceanic N2 Fixation
Domain, Science, 327, 1512–1514, 2010. Karl, D. M., Church, M. J., Dore, J. E., Letelier, R., and Mahaffey,
C.: Predictable and efficient carbon sequestration in the North
Pacific Ocean supported by symbiotic nitrogen fixation, P. Natl. Acad. Sci. USA, 109, 1842–1849, 2012. Moore, C. M., Mills, M. M. M., Arrigo, K. R., Berman-Frank, I.,
Bopp, L., Boyd, P. W., Galbraith, E. D., Geider, R. J., Guieu, C.,
Jaccard, S. L., Jickells, T. D., La Roche, J., Lenton, T. M., Ma-
howald, N. M., Maranon, E., Marinov, I., Moore, J. K., Nakat-
suka, T., Oschlies, A., Saito, M. A., Thingstad, T. F., Tsuda, A.,
and Ulloa, O.: Processes and patterns of oceanic nutrient limita-
tion, Nat. Geosci., 6, 701–710, 2013. Knapp, A. N., Sigman, D. M., and Lipschultz, F.: N isotopic com-
position of dissolved organic nitrogen and nitrate at the Bermuda
Atlantic Time-series Study site, Global Biogeochem. Cy., 19, 1–
15, 2005. Moutin, T., Van Den Broeck, N., Beker, B., Dupouy, C., Rim-
melin, P., and LeBouteiller, A.: Phosphate availability controls
Trichodesmium spp. biomass in the SW Pacific ocean, Mar. Ecol.-Prog. Ser., 297, 15–21, 2005. Knapp, A. N., Fawcett, S. T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment 3219 ton diazotrophy and transfer of nitrogen to picoplankton organ-
isms, Mar. Ecol-Prog. Ser., 343, 25–33, 2007. ton diazotrophy and transfer of nitrogen to picoplankton organ-
isms, Mar. Ecol-Prog. Ser., 343, 25–33, 2007. Lefevre, D., Grosso, O., Gimenez, A., Van Wambeke, F., Spungin,
D., Belkin, N., and Berman-Frank, I.: Net Community Produc-
tion accross the Western Tropical South Pacific. Implication for
the ecosystem functioning, Biogeosciences, in preparation, 2017. Gimenez, A., Baklouti, M., and Moutin, T.: Investigation of the
main processes controlling the biological carbon pump in the
Western Tropical South Pacific using a 1DV biogeochemical-
physical coupled model, Biogeosciences, in preparation, 2017. Lomas, M. W., Burke, A. L., Lomas, D. A., Bell, D. W.,
Shen, C., Dyhrman, S. T., and Ammerman, J. W.: Sargasso
Sea phosphorus biogeochemistry: an important role for dis-
solved organic phosphorus (DOP), Biogeosciences, 7, 695–710,
https://doi.org/10.5194/bg-7-695-2010, 2010. Gruber, N.: The dynamics of the marine nitrogen cycle and its in-
fluence on atmospheric CO2, in: The ocean carbon cycle and cli-
mate, edited by: Follows, M. and Oguz, T., Kluwer Academic,
Dordrecht, 2004. Longhurst, A. R.: Ecological Geography of the Sea, Academic
press, 1998. Guidi, L.: Fate of a diazotrophs bloom followed by Underwater
Video profiles in the WTSP (OUTPACE cruise), Biogeosciences,
in preparation, 2017. Luo, Y.-W., Doney, S. C., Anderson, L. A., Benavides, M., Berman-
Frank, I., Bode, A., Bonnet, S., Boström, K. H., Böttjer, D.,
Capone, D. G., Carpenter, E. J., Chen, Y. L., Church, M. J.,
Dore, J. E., Falcón, L. I., Fernández, A., Foster, R. A., Fu-
ruya, K., Gómez, F., Gundersen, K., Hynes, A. M., Karl, D. M., Kitajima, S., Langlois, R. J., LaRoche, J., Letelier, R. M.,
Marañón, E., McGillicuddy Jr., D. J., Moisander, P. H., Moore,
C. M., Mouriño-Carballido, B., Mulholland, M. R., Needoba, J. A., Orcutt, K. M., Poulton, A. J., Rahav, E., Raimbault, P., Rees,
A. P., Riemann, L., Shiozaki, T., Subramaniam, A., Tyrrell, T.,
Turk-Kubo, K. A., Varela, M., Villareal, T. A., Webb, E. A.,
White, A. E., Wu, J., and Zehr, J. P.: Database of diazotrophs
in global ocean: abundance, biomass and nitrogen fixation rates,
Earth Syst. Sci. Data, 4, 47–73, https://doi.org/10.5194/essd-4-
47-2012, 2012. Halm, H., Lam, P., Ferdelman, T. G., Lavik, G., Dittmar, T.,
LaRoche, J., D’Hondt, S., and Kuypers, M. M. M.: Heterotrophic
organisms dominate nitrogen fixation in the South Pacific Gyre,
ISME J., 6, 1238–1249, 2012. T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment G.: The acceleration of oceanic denitrification during
deglacial warming, Nat. Geosci., 6, 579–584, 2013. Chavez, F. P., Ryan, J., Lluch-Cota, S. E., and Niquen, M.: From an-
chovies to sardines and back: Multidecadal change in the Pacific
Ocean, Science, 299, 217–221, 2003. Close, H. G., Shah, S. R., Ingalls, A. E., Diefendorf, A. F., Brodie,
E. L., Hansman, R. L., Freeman, K. H., Aluwihare, L. I., and
Pearson, A.: Export of submicron particulate organic matter to
mesopelagic depth in an oligotrophic gyre, P. Natl. Acad. Sci. USA, 110, 12565–12570, 2013. A. N., Mix, A., Mobius, J., Pedersen, T. F., Pichevin, L., Quan, T. M., Robinson, R. S., Ryabenko, E., Schmittner, A., Schneider, de Verneil, A., Rousselet, L., Doglioli, A. M., Petrenko, A. A.,
and Moutin, T.: The Fate of a Southwest Pacific Bloom: Gaug-
ing the impact of submesoscale vs. mesoscale circulation on bi-
ological gradients in the subtropics, Biogeosciences Discuss.,
https://doi.org/10.5194/bg-2017-84, in review, 2017a. Gandhi, N., Singh, A., Prakash, Ramesh, R., Raman, M.,
Sheshshayee, M. S., and Shetye, S.: First direct measure-
ments of N2 fixation during a Trichodesmium bloom in the
eastern Arabian Sea, Global Biogeochem. Cy., 25, GB4014,
https://doi.org/10.1029/2010GB003970, 2011. de Verneil, A., Rousselet, L., Doglioli, A. M., Petrenko, A. A.,
Bouruet-Aubertot, P., Maes, C., and Moutin, T.: OUTPACE Long Garcia, N., Raimbault, P., and Sandroni, V.: Seasonal nitrogen fixa-
tion and primary production in the Southwest Pacific: nanoplank- Biogeosciences, 14, 3207–3220, 2017 www.biogeosciences.net/14/3207/2017/ T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment E., Martínez-Garcia, A., Leblond, N.,
Moutin, T., and Bonnet, S.: Nitrogen isotopic evidence for a
shift from nitrate- to diazotroph-fueled export production in the
VAHINE mesocosm experiments, Biogeosciences, 13, 4645–
4657, https://doi.org/10.5194/bg-13-4645-2016, 2016. Moutin, T., Karl, D. M., Duhamel, S., Rimmelin, P., Raimbault,
P., Van Mooy, B. A. S., and Claustre, H.: Phosphate availability
and the ultimate control of new nitrogen input by nitrogen fix-
ation in the tropical Pacific Ocean, Biogeosciences, 5, 95–109,
https://doi.org/10.5194/bg-5-95-2008, 2008. Knapp, A. N., Grosso, O., Leblond, N., Caffin, M., Moutin, T., and
Bonnet, S.: Zonal gradients in N2 fixation and its contribution to
export production in the Western Tropical South Pacific Ocean,
Biogeosciences, in preparation, 2017. Moutin, T., Van Wambeke, F., and Prieur, L.: Introduction to the
Biogeochemistry from the Oligotrophic to the Ultraoligotrophic
Mediterranean (BOUM) experiment, Biogeosciences, 9, 3817–
3825, https://doi.org/10.5194/bg-9-3817-2012, 2012. Kumar, P. K., Singh, A., Ramesh, R., and Nallathambi, T.:
N2 Fixation in the Eastern Arabian Sea: Probable. Role
of
Heterotrophic
Diazotrophs,
Front. Mar. Sci.,
4,
80,
https://doi.org/10.3389/fmars.2017.00080, 2017. Moutin, T., Wagener, T., Fumenia, A., Gimenez, A., Caffin, M.,
Lefevre, D., Leblanc, K., Bouruet-Aubertot, P., Helias-Nunige,
S., Rougier, G., Grosso, O., and de Verneil, A.: Phosphate avail-
ability and the ultimate control of the biological carbon pump in Leblanc K., Cornet, V., Brunet, C., Rimmelin-Maury, P., Grosso,
O., Helias-Nunige, S., and Quéguiner, B.: Siliceous plankton and
silicon biogeochemical cycle in the Tropical South Pacific. Bio-
geosciences, in preparation, 2017. Biogeosciences, 14, 3207–3220, 2017 www.biogeosciences.net/14/3207/2017/ T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment 3220 the Western Tropical South Pacific Ocean (OUTPACE cruise),
Biogeosciences, in preparation, 2017. the Western Tropical South Pacific Ocean (OUTPACE cruise),
Biogeosciences, in preparation, 2017. Stenegren, M., Caputo, A., Berg, C., Bonnet, S., and Foster, R. A.:
Distribution and drivers of symbiotic and free-living diazotrophic
cyanobacteria in the Western Tropical South Pacific, Biogeo-
sciences Discuss., https://doi.org/10.5194/bg-2017-63, in review,
2017. Murnane, R. J., Sarmiento, J. L., and Le Quere, C.: Spatial distri-
bution of air-sea CO2 fluxes and the interhemispheric transport
of carbon by the oceans, Global Biogeochem. Cy., 13, 287–305,
1999. Subramaniam, A., Yager, P. L., Carpenter, E. J., Mahaffey, C.,
Bjorkman, K. M., Cooley, S., Kustka, A. B., Montoya, J.,
Sanudo-Wilhelmy, S., Shipe, R., and Capone, D. G.: Amazon
River enhances diazotrophy and carbon sequestration in the trop-
ical North Atlantic Ocean, P. Natl. Acad. Sci. USA, 105, 10460–
10465, 2008. Nelson, D. M., Treguer, P., Brezezinski, M. A., Leynaert, A., and
Queguiner, B.: Production and dissolution of biogenic silica in
the ocean: Revised global estimates, comparison with regional
data and relationship to biogenic sedimentation, Global Bio-
geochem. Cy., 9, 359–372, 1995. Takahashi,
K.,
Montecinos,
A.,
Goubanova,
K.,
and
De-
witte,
B.:
ENSO
regimes:
Reinterpreting
the
canonical
and Modoki El Nino, Geophys. Res. Lett., 38, 5 pp.,
https://doi.org/10.1029/2011GL047364, 2011. Panagiotopoulos, C., Pujo-Pay, M., Benavides, M., and R. Sempéré. Composition and distribution of dissolved carbohydrates across
the Western Tropical South Pacific (OUTPACE cruise), Biogeo-
sciences, in preparation, 2017. Valdés, V., Donoso, K., Carlotti, F., Pagano, M., Molina, V., Escrib-
ano, R., and Fernandez, C.: Nitrogen and phosphorus recycling
mediated by copepods in western tropical south Pacific, Biogeo-
sciences, in preparation, 2017. Polovina, J. J., Howell, E. A., and Abecassis, M.: Ocean’s least pro-
ductive waters are expanding, Geophys. Res. Lett., 35, L03618,
https://doi.org/10.1029/2007GL031745, 2008. Raimbault, P. and Garcia, N.: Evidence for efficient regener-
ated production and dinitrogen fixation in nitrogen-deficient
waters of the South Pacific Ocean: impact on new and
export production estimates, Biogeosciences, 5, 323–338,
https://doi.org/10.5194/bg-5-323-2008, 2008. Van Den Broeck, N., Moutin, T., Rodier, M., and Le Bouteiller, A.:
Seasonal variations of phosphate availability in the SW Pacific
Ocean near New Caledonia, Mar. Ecol.-Prog. Ser., 268, 1–12,
2004. T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment Van Wambeke, F., Duhamel, S., Gimenez, A., Lefèvre, D., Pujo-
Pay, M., and Moutin, T.: Dynamics of phytoplankton and het-
erotrophic bacterioplankton in the Western Tropical South Pa-
cific Ocean along a gradient of diversity and activity of nitrogen
fixers, Biogeosciences, in preparation, 2017. Richardson, T. L. and Jackson, G. A.: Small phytoplankton and car-
bon export from the surface ocean, Science, 315, 838–840, 2007. Rousselet, L., de Verneil, A., Doglioli, A. M., Petrenko, A. A.,
Duhamel, S., Maes, C., and Blanke, B.: From large to sub-
mesoscale circulation influence during the OUTPACE cruise
(Western Tropical South Pacific), Biogeosciences, in preparation,
2017. Wagener, T., Metzl, N., Lo Monaco, C., Fin, J., Caffin, M., Lefevre,
D., and Moutin, T.: Anthopogenic carbon accumulation in the
South West Pacific, Biogeosciences, in preparation, 2017. Rousset G., Dupouy C., Lefèvre, J., De Boissieu, F., Ridoux, V.,
Rodier, M., and Menkes, C.: Remote sensing of Trichodesmium
in open Ocean of the South Pacific, Biogeosciences, in prepara-
tion, 2017. White, A. E., Foster, R. A., Benitez-Nelson, C. R., Masqué, P., Ver-
deny, E., Popp, B. N., Arthur, K. E., and Prahl, F. G.: Nitrogen
fixation in the Gulf of California and the Eastern Tropical North
Pacific, Prog. Oceanogr., 109, 1–17, 2012. Sabine, C. L., Feely, R. A., Gruber, N., Key, R. M., Lee, K., Bullis-
ter, J. L., Wanninkhof, R., Wong, C. S., Wallace, D. W. R.,
Tilbrook, B., Millero, F. J., Peng, T. H., Kozyr, A., Ono, T., and
Rios, A. F.: The oceanic sink for anthropogenic CO2, Science,
305, 367–371, 2004. Zehr, J. P. and Kudela, R. M.: Nitrogen Cycle of the Open Ocean:
From Genes to Ecosystems, Ann. Rev. Mar. Sci., 3, 197–225,
2011. Zehr, J. P., Mellon, M. T., and Zani, S.: New nitrogen-fixing mi-
croorganisms detected in oligotrophic oceans by amplification
of nitrogenase (nifH) genes, Appl. Environ. Microb., 64, 3444–
3450, 1998. Sarmiento, J. and Gruber, N.: Ocean Biogeochemical Dynam-
ics, Princeton University Press, Princeton, Woodstock, 503 pp.,
2006. Zehr, J. P., Bench, S. R., Carter, B. J., Hewson, I., Niazi, F., Shi,
T., Tripp, H. J., and Affourtit, J. P.: Globally Distributed Uncul-
tivated Oceanic N2-Fixing Cyanobacteria Lack Oxygenic Photo-
system II, Science, 322, 1110–1112, 2008. Scharek, R., Latasa, M., Karl, D. M., and Bidigare, R. R.: Temporal
variations in diatom abundance and downward vertical fux in the
oligotrophic North Pacific gyre, Deep-Sea Res. Pt. I, 46, 1051–
1075, 1999a. T. Moutin et al.: The Oligotrophy to the UlTra-oligotrophy PACific Experiment Sharek, R. M., Tupas, L. M., and Karl, D. M.: Diatom fluxes to
the deep sea in the oligotrophic North Pacific gyre at Station
ALOHA, Mar. Ecol.-Prog. Ser., 82, 55–67, 1999b. Sohm, J. A., Webb, E. A., and Capone, D. G.: Emerging patterns
of marine nitrogen fixation, Nat. Rev. Microbiol., 9, 499–508,
2011. Biogeosciences, 14, 3207–3220, 2017 www.biogeosciences.net/14/3207/2017/ www.biogeosciences.net/14/3207/2017/
|
https://openalex.org/W3164547793
|
https://discovery.ucl.ac.uk/10131056/1/Evidence%20for%20antibody%20as%20a%20protective%20correlate%20for%20COVID-19%20vaccines.pdf
|
English
| null |
Evidence for antibody as a protective correlate for COVID-19 vaccines
|
Vaccine
| 2,021
|
cc-by
| 4,691
|
a r t i c l e
i n f o A correlate of protection (CoP) is urgently needed to expedite development of additional COVID-19
vaccines to meet unprecedented global demand. To assess whether antibody titers may reasonably
predict efficacy and serve as the basis of a CoP, we evaluated the relationship between efficacy and
in vitro neutralizing and binding antibodies of 7 vaccines for which sufficient data have been generated. Once calibrated to titers of human convalescent sera reported in each study, a robust correlation was seen
between neutralizing titer and efficacy (q = 0.79) and binding antibody titer and efficacy (q = 0.93),
despite geographically diverse study populations subject to different forces of infection and circulating
variants, and use of different endpoints, assays, convalescent sera panels and manufacturing platforms. Together with evidence from natural history studies and animal models, these results support the use
of post-immunization antibody titers as the basis for establishing a correlate of protection for
COVID-19 vaccines. Article history:
Received 16 April 2021
Received in revised form 20 May 2021
Accepted 21 May 2021
Available online 24 May 2021 Article history:
Received 16 April 2021
Received in revised form 20 May 2021
Accepted 21 May 2021
Available online 24 May 2021
Keywords:
COVID-19
SARS-CoV-2
Vaccine
Correlate of protection 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://
creativecommons.org/licenses/by/4.0/). Since January 2020 Elsevier has created a COVID-19 resource centre with
free information in English and Mandarin on the novel coronavirus COVID- Since January 2020 Elsevier has created a COVID-19 resource centre with
free information in English and Mandarin on the novel coronavirus COVID-
19. The COVID-19 resource centre is hosted on Elsevier Connect, the
company's public news and information website. 19. The COVID-19 resource centre is hosted on Elsevier Connect, the
company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related
research that is available on the COVID-19 resource centre - including this
research content - immediately available in PubMed Central and other
publicly funded repositories, such as the WHO COVID database with rights
for unrestricted research re-use and analyses in any form or by any means Elsevier hereby grants permission to make all its COVID-19-related
research that is available on the COVID-19 resource centre - including this
research content - immediately available in PubMed Central and other
publicly funded repositories, such as the WHO COVID database with rights
for unrestricted research re-use and analyses in any form or by any means
with acknowledgement of the original source. These permissions are
granted for free by Elsevier for as long as the COVID-19 resource centre
remains active. with acknowledgement of the original source. These permissions are
granted for free by Elsevier for as long as the COVID-19 resource centre
remains active. Vaccine 39 (2021) 4423–4428 2.1. Data selection Inclusion criteria for immunogenicity and vaccine efficacy data
are described in the Supplemental Appendix. At the time of analy-
sis, seven vaccines met these criteria: Pfizer, Moderna, Gamaleya,
AstraZeneca, Sinovac, Novavax, and Johnson & Johnson. Evidence for antibody as a protective correlate for COVID-19 Kristen A. Earle a, Donna M. Ambrosino b, Andrew Fiore-Gartland c, David Goldblatt d,
Peter B. Gilbert c, George R. Siber e, Peter Dull a,⇑, Stanley A. Plotkin f a Vaccine Development & Surveillance, Bill & Melinda Gates Foundation, 500 5th Ave N, Seattle, WA 98109, USA
b Independent Advisor, Stuart, FL, USA c Vaccine and Infectious Disease Division, Fred Hutchinson Cancer Research Center, 1100 Fairview Ave N, Seattle, WA 98109, USA
d Great Ormond Street Institute of Child Health, University College London, 30 Guilford Street, London WC1N 1EH, UK
e Independent Advisor, New York, NY, USA p
,
,
,
Department of Pediatrics, Perelman School of Medicine, University of Pennsylvania, 3401 Civic Center Blvd., Philadelphia, PA 19104, USA f Department of Pediatrics, Perelman School of Medicine, University of Pennsylvania, 3401 Civic Center Blvd., Philadelphia, PA 19104, USA ⇑Corresponding author.
E-mail address: Peter.Dull@gatesfoundation.org (P. Dull). 1. Introduction evidence from non-human primate [3] and natural history [4]
studies suggests that an antibody-based correlate of protection
can be estimated for COVID-19 vaccines. We therefore assessed
the relationship between the efficacy of seven COVID-19 vaccines
in Phase III trials and the levels of both virus neutralizing antibody
(VNA) and Spike protein-binding IgG antibody to determine
whether either assay may serve as a predictor of vaccine efficacy
against COVID-19. Eleven novel COVID-19 vaccines have demonstrated efficacy
with several more undergoing Phase III clinical trials. Despite this,
to meet unprecedented global demand, additional vaccines are
needed even as placebo-controlled efficacy trials are becoming
infeasible [1]. An immunological correlate of protection (CoP) is
urgently needed not only to provide a path for regulatory approval
of new scalable, deliverable, and affordable vaccines, but for a
number of other applications, including: Phase IV studies that
enable the most efficient use of approved vaccines (i.e., heterolo-
gous priming and prime-boost regimens); serosurveys to evaluate
protection levels of populations; and to assist in predicting the
durability of protection. Many vaccines have been licensed or
had expanded indications based on a binding or functional anti-
body CoP established in multiple efficacy trials [2], but for
COVID-19 these subject level data analyses and a consensus
around the threshold of protection across multiple studies and
populations are unavailable. Meanwhile, a mounting body of https://doi.org/10.1016/j.vaccine.2021.05.063
0264-410X/ 2021 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). https://doi.org/10.1016/j.vaccine.2021.05.063
0264-410X/ 2021 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY license (http:// 2.2. Statistical analysis Vaccine efficacy (VE) was computed as one minus risk-ratio
times 100, and the risk-ratio for each study was calculated as https://doi.org/10.1016/j.vaccine.2021.05.063
0264-410X/ 2021 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). K.A. Earle, D.M. Ambrosino, A. Fiore-Gartland et al. Vaccine 39 (2021) 4423–4428 specified in the study protocol/primary publication. Correlation
was the Spearman’s rank correlation coefficient (q) between the
readouts on the x- and y-axes; both x and y data were fit using a
natural log transform. The dashed fit line was computed using
locally estimated scatterplot smoothing (LOESS) regression (all
points fit, with tricube weight function). We applied a non-
parametric Bayesian approach to evaluate the quality of VNA and
binding antibodies as trial-level surrogate endpoints [5]. Leave-
one-out cross-validation was applied to evaluate how well VE in
each held-out trial could be predicted from the observed biomar-
ker distribution and the model from the six other trials linking geo-
metric mean biomarker level to VE. developer, which vary widely in sample size and representation of
the range of disease severities. As both neutralizing and binding
antibody responses to natural infection correlate significantly with
severity of disease, the specific HCS panel selected may signifi-
cantly impact ratios generated in our analysis [8,9]; however, sen-
sitivity analysis suggested that the correlations were robust even
to randomly generated 2–10 fold shifts in the HCS geometric mean
values (Fig. S1). As expected, when the shifts are increased the
range of correlations becomes broader and approaches the uncali-
brated correlations. Third, these clinical studies were conducted on geographically
diverse study populations, which were subject to different forces
of infection and circulating variants. Variability in the timing and
location of the Phase III efficacy studies included in our analysis
suggests that the efficacy point estimates will be differentially
impacted by circulating variants of concern (VOCs) which are more
likely to escape protection afforded by vaccination [10], which is
supported by the increased correlation when isolating VE against
the wildtype strain (Fig. 2A). Finally, this analysis is subject to vari-
ability in clinical protocols, including endpoints, case definitions,
and assays, and in execution of these protocols. Our analyses sug-
gest that the correlation between efficacy and antibody biomarkers
is improved by including efficacy point estimates that most closely
correspond to the dose and schedule included in the Phase I/II
immunogenicity study (Fig. 3. Results We first evaluated peak geometric mean titers (GMT) of VNA
and binding antibodies 1–4 weeks following the recommended
vaccination regimen as reported by each manufacturer but found
low correlations with efficacy (Fig. 1A, 1B), most likely because
assays were not calibrated to a common standard. We then cali-
brated assays against an imperfect but ‘‘best available” standard,
titers of human convalescent serum (HCS) reported in each study,
to generate a vaccinated:convalescent sera ratio; this revealed high
correlation between the VNA ratio and efficacy (q = 0.79) and bind-
ing antibody titer ratio and efficacy (q = 0.93) (Fig. 1C, 1D). Neu-
tralizing or IgG binding antibody accounted for 77.5% and 94.2%,
respectively, of the variation in efficacy observed among the seven
vaccines. To assess the impact that circulating variants may have
on this relationship, we substituted primary endpoint efficacy esti-
mates with post-hoc analyses that either calculate efficacy against
the wildtype, D614G strain of SARS-CoV-2 (Novavax) or calculate
efficacy at sites without significant representation of circulating
variants (Janssen/J&J’s U.S. sites) where available. Post-hoc analysis
of Novavax vaccine efficacy against the ancestral strain (95.6%) was
determined by sequencing 56 of the 62 cases accumulated in the
UK Phase III study [6]. Vaccine efficacy for the U.S. sites of Jans-
sen/J&J’s Phase III study (72%) is included based on sequencing of
197 of the 268 cases, suggesting that strain D614G accounted for
the vast majority (96.4%) of cases [7]. Controlling for efficacy
against the ancestral strain strengthened the correlation between
VE and the VNA ratio (q = 0.96, Fig. 2A), but weakened the correla-
tion between VE and binding titer ratios (q = 0.82, Fig. 2B). Further-
more, accounting for increased antibody responses associated with
an extended interval schedule for the Oxford/AstraZeneca vaccine
improved the correlation between VE and both ratios (q = 0.86,
q = 0.93, Fig. 3). As additional data have become available, three issues chal-
lenge the interpretation and application of this model to further
vaccine development. First, trials of several vaccines (AstraZeneca,
Moderna, Pfizer) have shown substantial efficacy following the
first dose despite low neutralization titers, with many subjects
with titers from the pre-dose two sample below the lower limit
of quantification. 3. Results This suggests either relatively low neutralization
assay
sensitivity
or
non-neutralizing
antibodies
and
T
cell
responses may be functionally important, which is supported by
the evidence that binding antibody titers appear robust at the same
post-Dose 1 time points [11,12] and that Fc functional antibody
responses play a role in recovery from natural infection [13] and
correlate with protective efficacy [3]. Nonetheless, virus neutral-
ization assays remain an attractive candidate for the basis of a cor-
relate, as they represent a functional assay with high likelihood of
predicting efficacy following the full vaccination regimen. Second,
the rise in prominence of VOCs raises questions about the pro-
jected efficacy of many candidates currently in development. Pre-
dicting
efficacy
against
VOCs
will
require
measuring
neutralization titers to VOCs. While we do not yet have the data
needed to validate whether this model may be predictive for effi-
cacy against specific variants, it is encouraging that data available
to date suggest that consistent with this paradigm, reduction in
neutralization of B.1.351 by vaccinated sera corresponds to a
reduction in efficacy [10,14]. Third, our meta-analysis evaluated
prediction of vaccine efficacy over 3–5-month follow-up periods,
with data for predicting efficacy durability not yet available. Declining neutralization titers with time may correlate with or
cause falling efficacy, perhaps requiring boosters. Using a non-parametric approach, we modeled the relationship
of each of the candidate biomarkers with VE across the seven vac-
cines and attempted to predict the VE of each vaccine based on its
biomarker response (Methods). For both the VNA ratios (Fig. 4A)
and binding antibody ratios (Fig. 4B) the observed VE fell within
the 95% credible interval of the model’s prediction for all vaccines,
with the exception of Oxford/AstraZeneca and Gamaleya for VNA
and Sinovac for binding antibodies. 2.2. Statistical analysis 3), and we anticipate that this relation-
ship will be further elucidated once immunogenicity data are gen-
erated using a common assay and calibrated to the International
Standard. Finally, the observation that most of the predicted VE
95% Cis include the observed VE (Fig. 4) supports the hypothesis
that VNA and/or ELISA may be good correlates of protection. Despite these many caveats, a robust correlation between antibody
titers and efficacy is demonstrated in these results, and these
results support the utility of the antibody biomarkers for predict-
ing VE in new settings and for novel COVID-19 vaccines currently
in development. 4. Discussion In this study, a robust correlation was seen between antibody
titers and efficacy across seven different vaccines, with higher
titers correlating with higher vaccine efficacy, despite uncontrolled
variables across the studies. First, these vaccines represent four
vaccine manufacturing platforms with varying levels of engage-
ment of humoral and cell-mediated immunity: mRNA, adenoviral
vector, protein subunit, and inactivated virus vaccines. Second,
our analysis relies upon calibration to HCS panels selected by each The results reported here support the use of post-immunization
antibody levels as the basis for a CoP. We propose that the next 4424 Vaccine 39 (2021) 4423–4428 K.A. Earle, D.M. Ambrosino, A. Fiore-Gartland et al. Fig. 1. Correlation between antibody responses and efficacy rate for 7 COVID-19 vaccines. Panels A and B display correlations of antibody responses for neutralization and
ELISA assay ratios, respectively, without HCS calibration. Panels C and D display the same vaccine-induced responses, but with HCS calibration. Data included in correlation
analyses are described in Tables S1 and S2. Dot size corresponds to the number of cases reported for Phase III efficacy analyses. The y-axis is estimated log risk ratio reported
on the vaccine efficacy scale. The x-axis is ratio of the peak geometric mean neutralization titer or ELISA titer at 7–28 days post vaccination, relative to HCS. Error bars indicate
95% confidence Intervals (except for Oxford/AZ antibody responses, which represent ratios of median titers with interquartile ranges) with dashed line showing non-
parametric LOESS fit. A rank correlation value was calculated with R2 in a linear model utilized for variance explanation. K.A. Earle, D.M. Ambrosino, A. Fiore Gartland et al. Vaccine 39 (2021) 4423 4428 Fig. 1. Correlation between antibody responses and efficacy rate for 7 COVID-19 vaccines. Panels A and B display correlations of antibody responses for neutralization and
ELISA assay ratios, respectively, without HCS calibration. Panels C and D display the same vaccine-induced responses, but with HCS calibration. Data included in correlation
analyses are described in Tables S1 and S2. Dot size corresponds to the number of cases reported for Phase III efficacy analyses. The y-axis is estimated log risk ratio reported
on the vaccine efficacy scale. The x-axis is ratio of the peak geometric mean neutralization titer or ELISA titer at 7–28 days post vaccination, relative to HCS. 4. Discussion Error bars indicate
95% confidence Intervals (except for Oxford/AZ antibody responses, which represent ratios of median titers with interquartile ranges) with dashed line showing non-
parametric LOESS fit. A rank correlation value was calculated with R2 in a linear model utilized for variance explanation. venting severe disease or asymptomatic infection. Fifth, to verify
that the CoP will apply to new variants using appropriate
adapted assays as information accrues on the immune response
and efficacy of vaccines against them. steps toward achieving consensus on a CoP include the follow-
ing: First, to establish comparability of antibody measurements
among laboratories by (i) using the WHO International Standard
(NIBSC 20/136) to express VNA titers in IU/ml and binding anti-
body titers in BAU/ml and (ii) establishing a relevant proficiency
panel. Second, to agree on an assay, most likely a neutralization
assay meeting performance-based criteria, to serve as the gold
standard assay for CoP and perhaps to allow validation of sec-
ondary assays by strong correlations with the gold standard. Third, where possible, to calculate the protective threshold in
each Phase III study based either on the distribution of antibody
titers in random samples of vaccinees and controls [15] or on
case-cohort evaluations [16]. Fourth, since the CoP calculated
from different studies may differ, stakeholders should be con-
vened to arrive at a consensus on a minimum protective anti-
body level for the primary outcome of symptomatic COVID. If
possible, it would also be useful to estimate thresholds for pre- steps toward achieving consensus on a CoP include the follow-
ing: First, to establish comparability of antibody measurements
among laboratories by (i) using the WHO International Standard
(NIBSC 20/136) to express VNA titers in IU/ml and binding anti-
body titers in BAU/ml and (ii) establishing a relevant proficiency
panel. Second, to agree on an assay, most likely a neutralization
assay meeting performance-based criteria, to serve as the gold
standard assay for CoP and perhaps to allow validation of sec-
ondary assays by strong correlations with the gold standard. Third, where possible, to calculate the protective threshold in
each Phase III study based either on the distribution of antibody
titers in random samples of vaccinees and controls [15] or on
case-cohort evaluations [16]. 4. Discussion Publication of a pooled analysis of
Phase 3 trial data for Oxford/AstraZeneca suggests that interval between doses in Phase 3 studies varied from 4 to 12+ weeks, and that both vaccine efficacy and
immunogenicity data varied by dose interval. The immunogenicity data from Phase 1/2, which corresponded to a 4-week dose interval, may therefore not be representative of
immunogenicity generated in the Phase 3 study, or correspond to pooled vaccine efficacy estimates. To assess the impact on correlation, exploratory analyses of dose
intervals < 6 week and 12 weeks, and corresponding immunogenicity and efficacy, were substituted for the pooled vaccine efficacy estimate used in Fig. 1. Exploratory
analyses are denoted by blue dots. Ratios for neutralizing antibody titer (Panel A) and binding antibody titer (Panel B) were generated for Oxford/AstraZeneca using the HCS
median titer from Phase 1/2 publication, as noted in Tables S1 and S2. B A B Fig. 3. Impact of exploratory analysis of Oxford/AstraZeneca immunogenicity and efficacy by interval between doses on co alysis of Oxford/AstraZeneca immunogenicity and efficacy by interval between doses on correlation. Publication of a pooled a Fig. 3. Impact of exploratory analysis of Oxford/AstraZeneca immunogenicity and efficacy by interval between doses on correlation. Publication of a pooled analysis of
Phase 3 trial data for Oxford/AstraZeneca suggests that interval between doses in Phase 3 studies varied from 4 to 12+ weeks, and that both vaccine efficacy and
immunogenicity data varied by dose interval. The immunogenicity data from Phase 1/2, which corresponded to a 4-week dose interval, may therefore not be representative of
immunogenicity generated in the Phase 3 study, or correspond to pooled vaccine efficacy estimates. To assess the impact on correlation, exploratory analyses of dose
intervals < 6 week and 12 weeks, and corresponding immunogenicity and efficacy, were substituted for the pooled vaccine efficacy estimate used in Fig. 1. Exploratory
analyses are denoted by blue dots. Ratios for neutralizing antibody titer (Panel A) and binding antibody titer (Panel B) were generated for Oxford/AstraZeneca using the HCS
median titer from Phase 1/2 publication, as noted in Tables S1 and S2. t of exploratory analysis of Oxford/AstraZeneca immunogenicity and efficacy by interval between doses on correlation. Publica Fig. 3. Impact of exploratory analysis of Oxford/AstraZeneca immunogenicity and efficacy by interval between doses on correlation. 4. Discussion Fourth, since the CoP calculated
from different studies may differ, stakeholders should be con-
vened to arrive at a consensus on a minimum protective anti-
body level for the primary outcome of symptomatic COVID. If
possible, it would also be useful to estimate thresholds for pre- A CoP agreed to by regulators will support the demonstration of
efficacy of new vaccines using traditional criteria for establishing
non-inferiority based on the proportion of vaccinees achieving
the protective threshold and comparable GMTs, relative to an
approved comparator vaccine. Since the relationship between anti-
body responses and efficacy was consistent across all manufactur-
ing platforms evaluated to date, even though they may differ in
their induction of other antibody functions or T cell responses, an
argument can be made that any approved vaccine can serve as a
comparator for a future candidate vaccine. Even if authorized for
use based on an immunologic readout, full licensure should require
confirmatory effectiveness evaluations to closely follow any condi-
tional or accelerated approval process. 4425 Vaccine 39 (2021) 4423–4428 K.A. Earle, D.M. Ambrosino, A. Fiore-Gartland et al. Fig. 2. Impact of post-hoc analyses to assess efficacy against ancestral strain on correlation. To assess the impact that variation in circulating strains may have on
correlation between neutralizing antibody titer (Panel A) or binding antibody titer (Panel B) and efficacy, post-hoc analyses that calculate efficacy against the dominant
ancestral strain D614G (Novavax) or calculate efficacy at sites without circulating VOCs (Janssen/J&J) were substituted for primary endpoint efficacy estimates (blue dots). Binding antibody ratio for J&J (Panel B) was calculated from US-specific Phase III data, calibrated to HCS titers published with Phase I/II immunogenicity data, as noted in
Table S2. Fig. 2. Impact of post-hoc analyses to assess efficacy against ancestral strain on correlation. To assess the impact that variation in circulating strains may have on
correlation between neutralizing antibody titer (Panel A) or binding antibody titer (Panel B) and efficacy, post-hoc analyses that calculate efficacy against the dominant
ancestral strain D614G (Novavax) or calculate efficacy at sites without circulating VOCs (Janssen/J&J) were substituted for primary endpoint efficacy estimates (blue dots). Binding antibody ratio for J&J (Panel B) was calculated from US-specific Phase III data, calibrated to HCS titers published with Phase I/II immunogenicity data, as noted in
Table S2. A
B
Fig. 3. Impact of exploratory analysis of Oxford/AstraZeneca immunogenicity and efficacy by interval between doses on correlation. 4. Discussion Publication of a pooled analysis of
Phase 3 trial data for Oxford/AstraZeneca suggests that interval between doses in Phase 3 studies varied from 4 to 12+ weeks, and that both vaccine efficacy and
immunogenicity data varied by dose interval. The immunogenicity data from Phase 1/2, which corresponded to a 4-week dose interval, may therefore not be representative of
immunogenicity generated in the Phase 3 study, or correspond to pooled vaccine efficacy estimates. To assess the impact on correlation, exploratory analyses of dose
intervals < 6 week and 12 weeks, and corresponding immunogenicity and efficacy, were substituted for the pooled vaccine efficacy estimate used in Fig. 1. Exploratory
analyses are denoted by blue dots. Ratios for neutralizing antibody titer (Panel A) and binding antibody titer (Panel B) were generated for Oxford/AstraZeneca using the HCS
median titer from Phase 1/2 publication, as noted in Tables S1 and S2. 4426 K.A. Earle, D.M. Ambrosino, A. Fiore-Gartland et al. Vaccine 39 (2021) 4423–4428 g. 4. Prediction of vaccine efficacy using a cross-validation leave-one-out analysis. A non-parametric Bayesian approach was used to evaluate antibody titers as a
tential correlate. Neutralizing antibody titer (A) and binding antibody titer (B) were each evaluated using leave-one-out cross-validation: the immune response and VE
om six of the trials was used to build a model to predict the VE in the seventh trial based on the observed immune response distribution. Predicted VE is shown for each trial
ed dots and 95% confidence interval). Observed VE and immune response data are also plotted for reference (black dots). Fig. 4. Prediction of vaccine efficacy using a cross-validation leave-one-out analysis. A non-parametric Bayesian approach was used to evaluate antibody titers as a
potential correlate. Neutralizing antibody titer (A) and binding antibody titer (B) were each evaluated using leave-one-out cross-validation: the immune response and VE
from six of the trials was used to build a model to predict the VE in the seventh trial based on the observed immune response distribution. Predicted VE is shown for each trial
(red dots and 95% confidence interval). Observed VE and immune response data are also plotted for reference (black dots). 4427 K.A. Earle, D.M. Ambrosino, A. Fiore-Gartland et al. Vaccine 39 (2021) 4423–4428 [5] Gabriel EE, Sachs MC, Daniels MJ, Halloran ME. Optimizing and evaluating
biomarker
combinations
as
trial-level
general
surrogates. Stat
Med
2019;38:1135–46. https://doi.org/10.1002/sim.7996. Declaration of Competing Interest The authors declare the following financial interests/personal
relationships which may be considered as potential competing
interests: Dr. Plotkin consults for Janssen and Moderna; Dr. Siber
reports personal fees from Clover Biopharmaceuticals, other from
COVAXX, personal fees from CanSino, personal fees from CureVac,
personal fees from Valneva, personal fees and other from Affinivax,
outside the submitted work; Dr. Gilbert reports grants and non-
financial support from SanofiPasteur, outside the submitted work;
Dr. Ambrosino reports personal fees from COVAXX, personal fees
from Clover Biopharmaceuticals, outside the submitted work. [8] Chen X, Pan Z, Yue S, Yu F, Zhang J, Yang Y, et al. Disease severity dictates SARS-
CoV-2-specific
neutralizing
antibody
responses
in
COVID-19. Signal
Transduction and Targeted Therapy 2020;5:1–6. https://doi.org/10.1038/
s41392-020-00301-9. [9] Hansen CB, Jarlhelt I, Pérez-Alós L, Landsy LH, Loftager M, Rosbjerg A, et al. SARS-CoV-2 antibody responses are correlated to disease severity in COVID-19
convalescent
individuals. J
Immunol
2021;206:109–17. https://doi.org/
10.4049/jimmunol.2000898. [10] Zhou D, Dejnirattisai W, Supasa P, Liu C, Mentzer AJ, Ginn HM, et al. Evidence
of escape of SARS-CoV-2 variant B.1.351 from natural and vaccine-induced
sera. Cell 2021. https://doi.org/10.1016/j.cell.2021.02.037. [11] Walsh EE, Frenck RW, Falsey AR, Kitchin N, Absalon J, Gurtman A, et al. Safety
and immunogenicity of two RNA-based covid-19 vaccine candidates. N Engl J
Med 2020;383:2439–50. https://doi.org/10.1056/NEJMoa2027906. Funding This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors. [6] Novavax. Announcement of UK and South Africa Trial Results. 2021; published
online Jan 28. https://ir.novavax.com/events/event-details/novavax-update-
conference-call-and-webcast-0 (accessed March 8, 2021). [7] Janssen Biotech, Inc. FDA Briefing Document. Janssen Ad26.COV2.S Covid-19
Vaccine
for
the
Prevention
of
COVID-19. Sponsor
Briefing
Document
Addendum. 2021; published online Feb 26. https://www.fda.gov/media/
146218/download (accessed Feb 26, 2021). Supplementary data to this article can be found online at
https://doi.org/10.1016/j.vaccine.2021.05.063. Supplementary data to this article can be found online at
https://doi.org/10.1016/j.vaccine.2021.05.063. [13] Chen X, Rostad CA, Anderson LJ, Sun H, Lapp SA, Stephens K, et al. The
development and kinetics of functional antibody-dependent cell-mediated
cytotoxicity (ADCC) to SARS-CoV-2 spike protein. Virology 2021;559:1–9. https://doi.org/10.1016/j.virol.2021.03.009. Appendix A. Supplementary material p
//
g/
/
J
[12] Jackson LA, Anderson EJ, Rouphael NG, Roberts PC, Makhene M, Coler RN, et al. An mRNA vaccine against SARS-CoV-2 — preliminary report. N Engl J Med
2020;383:1920–31. https://doi.org/10.1056/NEJMoa2022483. [4] Lumley SF, O’Donnell D, Stoesser NE, Matthews PC, Howarth A, Hatch SB, et al.
Antibody status and incidence of SARS-CoV-2 infection in health care workers.
N Engl J Med 2021;384:533–40. https://doi.org/10.1056/NEJMoa2034545. References [14] Madhi SA, Baillie V, Cutland CL, Voysey M, Koen AL, Fairlie L, et al. Efficacy of
the ChAdOx1 nCoV-19 Covid-19 vaccine against the B.1.351 variant 0:null. N
Engl J Med 2021. https://doi.org/10.1056/NEJMoa2102214. [1] Wouters
OJ,
Shadlen
KC,
Salcher-Konrad
M,
Pollard
AJ,
Larson
HJ,
Teerawattananon Y, et al. Challenges in ensuring global access to COVID-19
vaccines: production, affordability, allocation, and deployment. The Lancet
2021;397:1023–34. https://doi.org/10.1016/S0140-6736(21)00306-8. [1] Wouters
OJ,
Shadlen
KC,
Salcher-Konrad
M,
Pollard
AJ,
Larson
HJ,
Teerawattananon Y, et al. Challenges in ensuring global access to COVID-19
vaccines: production, affordability, allocation, and deployment. The Lancet
2021;397:1023–34. https://doi.org/10.1016/S0140-6736(21)00306-8. [15] Siber GR, Chang I, Baker S, Fernsten P, O’Brien KL, Santosham M, et al. Estimating
the
protective
concentration
of
anti-pneumococcal
capsular
polysaccharide
antibodies. Vaccine
2007;25:3816–26. https://doi.org/
10.1016/j.vaccine.2007.01.119. [2] Plotkin
SA. Correlates
of
protection
induced
by
vaccination. CVI
2010;17:1055–65. https://doi.org/10.1128/CVI.00131-10. [2] Plotkin
SA. Correlates
of
protection
induced
by
vaccination. CVI
2010;17:1055–65. https://doi.org/10.1128/CVI.00131-10. [16] Qin L, Gilbert PB, Corey L, McElrath MJ, Self SG. A framework for assessing
immunological
correlates
of
protection
in
vaccine
trials. J
Infect
Dis
2007;196:1304–12. https://doi.org/10.1086/522428. [3] McMahan K, Yu J, Mercado NB, Loos C, Tostanoski LH, Chandrashekar A, et al. Correlates of protection against SARS-CoV-2 in rhesus macaques. Nature
2021;590:630–4. https://doi.org/10.1038/s41586-020-03041-6. [3] McMahan K, Yu J, Mercado NB, Loos C, Tostanoski LH, Chandrashekar A, et al. Correlates of protection against SARS-CoV-2 in rhesus macaques. Nature
2021;590:630–4. https://doi.org/10.1038/s41586-020-03041-6. [4] Lumley SF, O’Donnell D, Stoesser NE, Matthews PC, Howarth A, Hatch SB, et al. Antibody status and incidence of SARS-CoV-2 infection in health care workers. N Engl J Med 2021;384:533–40. https://doi.org/10.1056/NEJMoa2034545. 4428
|
https://openalex.org/W2738545992
|
https://repository.arizona.edu/bitstream/10150/625498/1/e005407.full.pdf
|
English
| null |
Pathogenesis of Hypertrophic Cardiomyopathy is Mutation Rather Than Disease Specific: A Comparison of the Cardiac Troponin T E163R and R92Q Mouse Models
|
Journal of the American Heart Association. Cardiovascular and cerebrovascular disease
| 2,017
|
cc-by
| 16,756
|
Pathogenesis of Hypertrophic Cardiomyopathy is Mutation
Rather Than Disease Specific: A Comparison of the
Cardiac Troponin T E163R and R92Q Mouse Models
Item Type
Article
Authors
Ferrantini, Cecilia; Coppini, Raffaele; Pioner, Josè Manuel;
Gentile, Francesca; Tosi, Benedetta; Mazzoni, Luca; Scellini,
Beatrice; Piroddi, Nicoletta; Laurino, Annunziatina; Santini,
Lorenzo; Spinelli, Valentina; Sacconi, Leonardo; De Tombe,
Pieter; Moore, Rachel; Tardiff, Jil; Mugelli, Alessandro; Olivotto,
Iacopo; Cerbai, Elisabetta; Tesi, Chiara; Poggesi, Corrado
Citation
Pathogenesis of Hypertrophic Cardiomyopathy is Mutation Rather
Than Disease Specific: A Comparison of the Cardiac Troponin T
E163R and R92Q Mouse Models 2017, 6 (7):e005407 Journal of the
American Heart Association
DOI
10.1161/JAHA.116.005407
Publisher
WILEY
Journal
Journal of the American Heart Association
Rights
© 2017 The Authors. Published on behalf of the American Heart
Association, Inc., by Wiley. This is an open access article under
the terms of the Creative Commons AttributionNonCommercial
License.
Download date
24/10/2024 04:01:19
Item License
https://creativecommons.org/licenses/by-nc/4.0/
Version
Final published version Pathogenesis of Hypertrophic Cardiomyopathy is Mutation
Rather Than Disease Specific: A Comparison of the
Cardiac Troponin T E163R and R92Q Mouse Models
Item Type
Article
Authors
Ferrantini, Cecilia; Coppini, Raffaele; Pioner, Josè Manuel;
Gentile, Francesca; Tosi, Benedetta; Mazzoni, Luca; Scellini,
Beatrice; Piroddi, Nicoletta; Laurino, Annunziatina; Santini,
Lorenzo; Spinelli, Valentina; Sacconi, Leonardo; De Tombe,
Pieter; Moore, Rachel; Tardiff, Jil; Mugelli, Alessandro; Olivotto,
Iacopo; Cerbai, Elisabetta; Tesi, Chiara; Poggesi, Corrado
Citation
Pathogenesis of Hypertrophic Cardiomyopathy is Mutation Rather
Than Disease Specific: A Comparison of the Cardiac Troponin T
E163R and R92Q Mouse Models 2017, 6 (7):e005407 Journal of the
American Heart Association
DOI
10.1161/JAHA.116.005407
Publisher
WILEY
Journal
Journal of the American Heart Association
Rights
© 2017 The Authors. Published on behalf of the American Heart
Association, Inc., by Wiley. This is an open access article under
the terms of the Creative Commons AttributionNonCommercial
License. Download date
24/10/2024 04:01:19
Item License
https://creativecommons.org/licenses/by-nc/4.0/
Version
Final published version From the Departments of Experimental and Clinical Medicine (C.F., J.M., F.G., B.T., B.S., N.P., A.M., C.T., C.P.) and NeuroFarBa (R.C., L.M., A.L., L.S., V.S., E.C.),
University of Florence, Italy; LENS, University of Florence & National Institute of Optics (INO—CNR), Florence, Italy (L.S.); Loyola University Medical Center, Department
of Physiology, Chicago, IL (P.T.); University of Arizona at Tucson, AZ (R.M., J.T.); Careggi University Hospital, Florence, Italy (I.O.).
Accompanying DataS1andFiguresS1 through S6areavailableathttp://jaha.ahajournals.org/content/6/7/e005407/DC1/embed/inline-supplementary-material-1.pdf
*Dr Ferrantini and Dr Coppini contributed equally to this work.
Correspondence to: Cecilia Ferrantini, MD, PhD, Department of Experimental and Clinical Medicine, University of Florence, Viale G. Morgagni, 63, 50141 Firenze,
Italy. E-mail: cecilia.ferrantini@unifi.it
Received March 31, 2017; accepted May 1, 2017.
ª 2017 The Authors. Published on behalf of the American Heart Association, Inc., by Wiley. This is an open access article under the terms of the Creative Commons
Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for
commercial purposes. Link to Item
http://hdl.handle.net/10150/625498 http://hdl.handle.net/10150/625498 ORIGINAL RESEARCH Pathogenesis of Hypertrophic Cardiomyopathy is Mutation Rather
Than Disease Specific: A Comparison of the Cardiac Troponin T E163R
and R92Q Mouse Models Cecilia Ferrantini, MD, PhD;* Raffaele Coppini, MD, PhD;* Jose Manuel Pioner, PhD; Francesca Gentile, PhD; Benedetta Tosi, MD, PhD;
Luca Mazzoni, PhD; Beatrice Scellini, PhD; Nicoletta Piroddi, PhD; Annunziatina Laurino, PhD; Lorenzo Santini, MS; Valentina Spinelli, PhD;
Leonardo Sacconi, PhD; Pieter De Tombe, PhD; Rachel Moore, PhD; Jil Tardiff, MD, PhD; Alessandro Mugelli, MD; Iacopo Olivotto, MD;
Elisabetta Cerbai, PhD; Chiara Tesi, PhD; Corrado Poggesi, MD Background-—In cardiomyocytes from patients with hypertrophic cardiomyopathy, mechanical dysfunction and arrhythmogenicity
are caused by mutation-driven changes in myofilament function combined with excitation-contraction (E-C) coupling abnormalities
related to adverse remodeling. Whether myofilament or E-C coupling alterations are more relevant in disease development is
unknown. Here, we aim to investigate whether the relative roles of myofilament dysfunction and E-C coupling remodeling in
determining the hypertrophic cardiomyopathy phenotype are mutation specific. ª 2017 The Authors. Published on behalf of the American Heart Association, Inc., by Wiley. This is an open access article under the terms of the Creative Commons
Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for
commercial purposes. y
Received March 31, 2017; accepted May 1, 2017.
ª 2017 The Authors. Published on behalf of the American Heart Association, Inc., by Wiley. This is an open access article under the terms of the Creative Commons
Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for
commercial purposes. Received March 31, 2017; accepted May 1, 2017. l and Clinical Medicine (C.F., J.M., F.G., B.T., B.S., N.P., A.M., C.T., C.P.) and NeuroFarBa (R.C., L.M., A.L., L.S., V.S., E.C.),
ersity of Florence & National Institute of Optics (INO—CNR), Florence, Italy (L.S.); Loyola University Medical Center, Department
rsity of Arizona at Tucson, AZ (R.M., J.T.); Careggi University Hospital, Florence, Italy (I.O.). by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from Methods and Results-—Two hypertrophic cardiomyopathy mouse models carrying the R92Q and the E163R TNNT2 mutations were
investigated. Echocardiography showed left ventricular hypertrophy, enhanced contractility, and diastolic dysfunction in both
models; however, these phenotypes were more pronounced in the R92Q mice. Both E163R and R92Q trabeculae showed
prolonged twitch relaxation and increased occurrence of premature beats. In E163R ventricular myofibrils or skinned trabeculae,
relaxation following Ca2+ removal was prolonged; resting tension and resting ATPase were higher; and isometric ATPase at maximal
Ca2+ activation, the energy cost of tension generation, and myofilament Ca2+ sensitivity were increased compared with that in wild-
type mice. No sarcomeric changes were observed in R92Q versus wild-type mice, except for a large increase in myofilament Ca2+
sensitivity. In R92Q myocardium, we found a blunted response to inotropic interventions, slower decay of Ca2+ transients, reduced
SERCA function, and increased Ca2+/calmodulin kinase II activity. Contrarily, secondary alterations of E-C coupling and signaling
were minimal in E163R myocardium. Conclusions-—In E163R models, mutation-driven myofilament abnormalities directly cause myocardial dysfunction. In R92Q,
diastolic dysfunction and arrhythmogenicity are mediated by profound cardiomyocyte signaling and E-C coupling changes. Similar
hypertrophic cardiomyopathy phenotypes can be generated through different pathways, implying different strategies for a
precision medicine approach to treatment. (J Am Heart Assoc. 2017;6:e005407. What is New? • Two hypertrophic cardiomyopathy (HCM) mouse models
with R92Q and E163R TNNT2 mutations were characterized
with echocardiographic and biophysical studies to assess
the relative roles of myofilament dysfunction and excitation-
contraction coupling remodeling in HCM pathophysiology. • Two hypertrophic cardiomyopathy (HCM) mouse models
with R92Q and E163R TNNT2 mutations were characterized
with echocardiographic and biophysical studies to assess
the relative roles of myofilament dysfunction and excitation-
contraction coupling remodeling in HCM pathophysiology. • Both models exhibited diastolic dysfunction and increased
arrhythmogenicity. • The E163R model showed severe myofilament abnormalities
including altered sarcomere energetics, with preserved
intracellular Ca2+ transients. • The R92Q model presented profound excitation-contraction
coupling alterations that slowed down Ca2+ transients and
twitch contractions in the absence of myofilament energetic
changes. by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from • Increased calcium/calmodulin-dependent protein kinase II
activity and increased myofilament Ca2+ sensitivity were
detected in both mutants, but the changes were much larger
in the R92Q model. In an era of emerging therapies that use a precision
medicine molecular approach, this notion may be directly
relevant to future management of the disease. To investigate
this complex issue, we used 2 HCM mouse models reflecting
the human phenotype but caused by different cardiac
troponin T (TnT) mutations. We assessed how primary
changes in myofilament function concur with secondary
abnormalities of cardiomyocyte excitation-contraction (E-C)
coupling to determine contractile impairment and arrhythmo-
genicity in HCM. What are the Clinical Implications? • The notion that cellular remodeling is mutation specific in
HCM may be relevant for treatment and may explain why
some patients with HCM may not respond to specific
therapies. • Drugs targeting ion channels or Ca2+ fluxes, such as late Na-
current blockers or diltiazem, may be more effective in the
presence of significant cardiomyocyte excitation-contrac-
tion coupling remodeling (eg, R92Q mutation). Methods Detailed methods are available in Data S1. • When HCM mutations have a direct impact on sarcomere
energetics (eg, E163R), sarcomere-targeting drugs may be
preferred; trials of novel myosin inhibitors in patients are
currently underway. TnT Mutant Transgenic Mouse Lines • Disease mechanisms that are common to different HCM
mutations (calcium/calmodulin-dependent protein kinase II,
myofilament Ca2+ sensitivity) should be considered as valid
targets for future therapies. All experimental protocols were performed in agreement with
current Italian and European regulations and were approved
by the local institutional review board and the animal-welfare
committee of the Italian Ministry of Health. We used a total of
sixty-seven 6- to 8-month-old male C57BL/6N transgenic
mice carrying the R92Q13,23 or E163R24 mutation in the
TNNT2 gene, as well as wild-type (WT) littermates: 22 R92Q,
24 E163R, and 21 WT mice were used for the experiments
described below. The mouse colonies were housed in the
animal facility of the University of Florence and all experi-
ments were conducted locally. The 2 transgenic lines were
generated in laboratories as previously described.23,24 minimal
echocardiographic
anomalies6
to
patients
with
extreme left ventricular (LV) remodeling and advanced heart
failure.3 Despite the vast archipelago of clinical manifestations,
LV hypertrophy (LVH), diastolic dysfunction, and an arrhythmic
propensity are considered hallmarks of the disease and occur in
most patients7,8 independently of their genetic profile. Biophysical experiments have demonstrated how HCM-
related mutations primarily alter sarcomere function, ie,
crossbridge cycling kinetics9,10 or the switched-off state of
the thin filament,11 leading to increased energy consumption
during tension generation or diastole.12–14 In addition to altered
sarcomere energetics, increased myofilament Ca2+ sensitivity
has often been reported in human HCM myocardium,9,15 both Journal of the American Heart Association
2 DOI: 10.1161/JAHA.116.005407 by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from DOI: 10.1161/JAHA.116.005407.) Key Words: excitation-contraction coupling • hypertrophic cardiomyopathy • pathophysiology • sarcomere physiology • troponin T mutations in most patients.2 HCM is a progressive disease with
different clinical stages3 and a large spectrum of clinical
phenotypes,4,5 ranging from asymptomatic individuals with mutations in most patients.2 HCM is a progressive disease with
different clinical stages3 and a large spectrum of clinical
phenotypes,4,5 ranging from asymptomatic individuals with H
ypertrophic cardiomyopathy (HCM) is the most common
Mendelian cardiac disease,1 defined as a disease of the
sarcomere because of its association with myofilament gene H
y Journal of the American Heart Association
1 Journal of the American Heart Association
1 DOI: 10.1161/JAHA.116.005407 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al ORIGINAL RESEARCH as a direct effect of the sarcomeric protein mutation15,16 and as
a consequence of secondary myofilament posttranslational
modifications.17 The complex and poorly known secondary
remodeling process that occurs in HCM involves a number of
myofilament, sarcoplasmic reticulum (SR), and sarcolemmal
alterations,18–20 reflecting adaptive/maladaptive modifica-
tions in various signaling pathways.21,22 We recently demon-
strated that in myocardial samples from patients with severe
HCM who underwent septal myectomy, changes of calcium/
calmodulin-dependent protein kinase II (CaMKII)–dependent
signaling lead to prolonged action potentials, cellular arrhyth-
mias, prolonged Ca2+ transients, and increased diastolic
intracellular [Ca2+] and [Na+].22 Such abnormalities appeared
unrelated to patients’ individual mutations.22 However, the
scarce availability of human samples does not allow us to
properly define whether HCM-related secondary remodeling is
truly independent of the underlying genotype. ORIGINAL RESEARCH ORIGINAL RESEARCH ORIGINAL RESEARCH test with Dunn’s multiple comparison test (for non-Gaussian
data sets), or Welch’s ANOVA with Games-Howell test (for
heteroscedastic groups). For variables where measurements
from an unequal number of different samples (eg, cells or
trabeculae) from each mouse were included (eg, ion fluores-
cence and myofibrils isometric force data), we used linear
mixed models to compare data groups, using the xtmixed
function of the STATA 12.0 program (StataCorp). To account
for intra-subject correlation, we included mouse-id as a
random-effect term in the model. The Tukey-Kramer post
hoc correction method was used to compute P values for all
pairwise comparisons, using the pwcompare with mcompare
(tukey) option in STATA. Overall, P<0.05 was considered
statistically significant. The range of calculated P values for
each comparison (0.05>P>0.01, 0.01>P>0.001, or P<0.001)
is indicated in the respective figure panels using symbols: red
symbols refer to R92Q versus WT comparisons, blue symbols
to E163R versus WT comparisons, and purple symbols to
R92Q versus E163R comparisons. and systolic and diastolic function using B-mode imaging and
Doppler measurements of transmitral blood flow. LV intact
trabeculae were dissected from explanted hearts26,27 and
used to record isometric force during electrical stimulation
with different pacing protocols at baseline and following b-
adrenergic activation.18 Skinned trabeculae were used to
obtain pCa-tension curves as previously described.28 Sarcom-
ere energetics was assessed in skinned trabeculae by
simultaneous measurement of isometric force and ATPase
activity with an enzyme-coupled assay.12,14 Single myofibrils
were isolated from mouse ventricular samples and used for
mechanical measurements using a fast solution switching
technique to assess maximal Ca2+-activated tension, resting
tension, and the kinetics of active tension generation and
relaxation.9,29 Myofibril suspensions were also used to
measure resting ATPase by assessing phosphate production
with a rapid colorimetric method. by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from Hallmarks of Human HCM Phenotype are Present
in Both R92Q and E163R Mouse Hearts Echocardiographic measurements were performed in anes-
thetized male mice from the 3 study groups (WT, R92Q, and
E163R), using a standardized protocol.31 Average heart rate
during echocardiographic measurement was similar in the 3
groups: 58542 beats per minute in WT, 57938 beats per
minute in E163R, and 57345 beats per minute in R92Q
mice. Representative images of parasternal long-axis views of
the left ventricle are reported in Figure 1A. In both R92Q and
E163R mice, septal thickness was significantly increased
compared with WT mice (Figure 1B), highlighting the presence
of asymmetric LVH in both HCM models. Short-axis views at
different levels of the left ventricle were used to estimate LV
volumes (Figure 1C). In R92Q compared with WT and E163R
mice, we found a modest reduction of end-diastolic and end-
systolic LV volumes. This is consistent with more severe LVH
in R92Q mice, as the increase of wall thickness diminishes LV
cavity volume. Results Single cardiomyocytes were isolated from excised hearts via
enzymatic dissociation and used for intracellular Ca2+mea-
surements using Ca-sensitive fluorescent dyes18,27 to evalu-
ate the amplitude and kinetics of Ca2+ transients, diastolic
[Ca2+], and the rate of spontaneous Ca2+ release, during
stimulation with field electrodes.30 To assess T-tubule density,
myocytes were stained with membrane-selective dyes and
observed with a confocal microscope.18,30 Fast-frozen LV
myocardial samples were processed to obtain total proteins,
which were used for Western blot studies to assess expres-
sion and phosphorylation
of CaMKII,22 phospholamban,
troponin I, and nuclear factor of activated T cell, as well as
the expression of SERCA. Formalin-fixed LV slices were
stained with Picrosirius red and used to assess intramyocar-
dial fibrosis.30 Myocardial Mechanics From the Whole Heart to
the Single Myofibril Echocardiography was performed on isoflurane-anesthetized
mice as previously described25 to characterize LV morphology DOI: 10.1161/JAHA.116.005407 DOI: 10.1161/JAHA.116.005407 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al Journal of the American Heart Association
3 Statistical Analysis Data from myofibrils, cells, and muscles are expressed as
meanSEM (number of samples and animals are indicated in
the respective figure legends). Statistical analysis was per-
formed as previously described,22,30 using SPSS version 23.0
(IBM) and STATA version 12.0 (StataCorp). In brief, all sets of
variables were checked for normality (Shapiro-Wilk test) and
for homogeneity of variances among groups (Levene’s test). The statistical tests used to calculate P values for each data
set are indicated in the respective figure legends. For variables
where a single measurement for each mouse is included (eg,
echocardiography, Western blot), the 3 different groups were
compared using 1-way ANOVA with Tukey correction (for
normally distributed homoscedastic data sets), Kruskal-Wallis In both R92Q and E163R mice compared with WT mice, we
observed increased ejection fraction and septal fractional
thickening during contraction (Figure 1D). In E163R mice,
stroke volume and cardiac output were increased compared
with that in WT mice (Figure 1E). In R92Q mice, instead, no
differences were noted compared with that in WT mice,
indicating that enhanced LV ejection fraction compensates for
the reduced LV volumes. Doppler studies of transmitral blood flow velocity (Fig-
ure 1F) were performed in 4-chamber views to assess LV DOI: 10.1161/JAHA.116.005407 DOI: 10.1161/JAHA.116.005407 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al ORIGINAL RESEARCH F
G
R92Q
WT
E163R
E
A
E A
A
E
R92Q
WT
E163R
A
Diastolic
Systolic
0.8
1.0
1.2
1.4
Septal Thickness, mm
IVS Thickening
LV-EF%
40
50
60
70
%
Diastolic
Systolic
15
30
45
60
75
WT
R92Q
E163R
Volume, μl
S.V. C.O. 10
20
30
40
S.V., μl - C.O., ml/min
B
0.6
0.8
1.0
1.2
1.4
1.6
E/A ratio
0
5
10
15
20
25
30
IVRT, ms
C
D
E
*
**
*
*
*
*
*
*
*
**
**
*
*
*
*
*
*
*
*
*
*
Figure 1. Echocardiographic measurements. A, Representative parasternal long-axis views of the left ventricle at end diastole from wild-type
(WT), R92Q, and E163R mice. The horizontal scale bars equal 1 mm. B, Thickness of the interventricular septum (IVS) measured at end diastole
(left) and at end systole (right) in WT, R92Q, and E163R mice. C, Left ventricular (LV) volumes are calculated using the Simpson technique at end
diastole (left) and end systole (right) in mice from the 3 study groups. Statistical Analysis D, Systolic thickening of the IVS (left) and LV ejection fraction (LV-EF%,
right), are expressed as percentage of the diastolic values, measured in mice from the 3 cohorts. E, Stroke volume (SV) calculated from Simpson
volumes and cardiac output (CO). F, Representative transmitral blood flow velocity curves recorded using pulsed-wave Doppler
echocardiography in apical views from WT, R92Q, and E163R mice. The vertical scale bar equals 100 mm/s. Early (passive, E) and late
(atrial contraction, A) LV filling waves are indicated on the traces. G, Ratio between E and A waves (E/A ratio, left) and isovolumic LV relaxation
time (IVRT, right) in mice from the 3 study cohorts. (B, E) Statistical tests: 1-way ANOVA with Tukey correction. G, Statistical test: Welch’s
ANOVA with Games-Howell test. (B through E and G) MeansSE from 10 mice per group. *=0.05>P>0.01; **=0.01>P>0.001. R92Q
WT
E163R
A
Diastolic
Systolic
0.8
1.0
1.2
1.4
Septal Thickness, mm
IVS Thickening
LV-EF%
40
50
60
70
%
Diastolic
Systolic
15
30
45
60
75
WT
R92Q
E163R
Volume, μl
S.V. C.O. 10
20
30
40
S.V., μl - C.O., ml/min
B
C
D
E
*
*
*
*
*
*
**
**
*
*
*
*
*
*
* Diastolic
Systolic
0.8
1.0
1.2
1.4
Septal Thickness, mm
Diastolic
Systolic
15
30
45
60
75
WT
R92Q
E163R
Volume, μl
B
C
*
*
*
*
*
*
* R92Q
WT
Diastolic
Systolic
0.8
1.0
1.2
1.4
Septal Thickness, mm
Diastolic
Systolic
15
30
45
60
75
WT
R92Q
E163R
Volume, μl
B
C
*
*
*
*
*
*
* Diastolic
Systolic
15
30
45
60
75
WT
R92Q
E163R
Volume, μl
C
*
*
* A IVS Thickening
LV-EF%
40
50
60
70
%
D
**
**
*
* S.V. C.O. 10
20
30
40
S.V., μl - C.O., ml/min
E
*
*
*
* E D D G
0.6
0.8
1.0
1.2
1.4
1.6
E/A ratio
0
5
10
15
20
25
30
IVRT, ms
*
**
*
*
*
* F
G
R92Q
WT
E163R
E
A
E A
A
E
0.6
0.8
1.0
1.2
1.4
1.6
E/A ratio
0
5
10
15
20
25
30
IVRT, ms
*
**
*
*
*
* F
WT
E
A F Figure 1. Echocardiographic measurements. A, Representative parasternal long-axis views of the left ventricle at end diastole from wild-type
(WT), R92Q, and E163R mice. Journal of the American Heart Association Twitch Relaxation is Prolonged and Spontaneous
Activity Increased in Both R92Q and E163R
Trabeculae Twitch Relaxation is Prolonged and Spontaneous
Activity Increased in Both R92Q and E163R
Trabeculae arrhythmogenic propensity associated with each mutation. In
all mouse lines, the rate of spontaneous contractions was
extremely low at baseline conditions (at 1-Hz stimulation, <1%
of preparations showed spontaneous contractions). A specific
protocol to increase the probability of spontaneous events
(Figure 2D) was applied, featuring the combination of isopro-
terenol 107 mol/L and a burst of high-rate (3 Hz) stimuli
followed by a 30-second simulation pause. With this protocol,
spontaneous activity occurred in more than 40% of E163R and
R92Q trabeculae, while arrhythmic beats arose in only 10% of
WT trabeculae (Figure 2E). Isometric force was measured from intact left and right
ventricular trabeculae or thin papillary muscles during field
stimulation at 1 Hz, 30°C, and 2 mmol/L extracellular [Ca2+]
(Figure 2). The amplitude of twitch contraction was similar in
WT, R92Q, and E163R mice (Figure 2A and 2B). The baseline
diastolic tension tended to be higher in E163R and R92Q
compared with WT mice, although the difference was not
significant (Figure 2B). Both E163R and R92Q trabeculae
showed prolonged twitch duration compared with WT, caused
by a significant prolongation of relaxation times (Figure 2A
and 2C). This result is in line with the prolonged isovolumic LV
relaxation time observed in vivo. The prolonged relaxation
cannot be explained by changes in the myosin heavy chain
isoform as myosin heavy chain-a is the only significantly
expressed isoform (>99%) in all 3 mouse lines. ORIGINAL RESEARCH ORIGINAL RESEARCH Statistical Analysis The horizontal scale bars equal 1 mm. B, Thickness of the interventricular septum (IVS) measured at end diastole
(left) and at end systole (right) in WT, R92Q, and E163R mice. C, Left ventricular (LV) volumes are calculated using the Simpson technique at end
diastole (left) and end systole (right) in mice from the 3 study groups. D, Systolic thickening of the IVS (left) and LV ejection fraction (LV-EF%,
right), are expressed as percentage of the diastolic values, measured in mice from the 3 cohorts. E, Stroke volume (SV) calculated from Simpson
volumes and cardiac output (CO). F, Representative transmitral blood flow velocity curves recorded using pulsed-wave Doppler
echocardiography in apical views from WT, R92Q, and E163R mice. The vertical scale bar equals 100 mm/s. Early (passive, E) and late
(atrial contraction, A) LV filling waves are indicated on the traces. G, Ratio between E and A waves (E/A ratio, left) and isovolumic LV relaxation
time (IVRT, right) in mice from the 3 study cohorts. (B, E) Statistical tests: 1-way ANOVA with Tukey correction. G, Statistical test: Welch’s
ANOVA with Games-Howell test. (B through E and G) MeansSE from 10 mice per group. *=0.05>P>0.01; **=0.01>P>0.001. in both R92Q and E163R compared with WT mice, we found
prolonged isovolumic LV relaxation time, an additional index
of diastolic dysfunction (Figure 1G). Taken together, these
echocardiographic measurements indicate that 3 hallmarks
of human HCM phenotype, ie, LVH, LV hypercontractility,
and diastolic dysfunction are reproduced in both HCM
models. diastolic function. Reduced transmitral blood flow velocity
during early diastole (ie, reduced E-wave amplitude), paral-
leled by an increased proportion of LV diastolic filling during
atrial contraction (ie, increased A-wave amplitude), is widely
recognized as an index of diastolic dysfunction. In both
mutants, and more severely in R92Q, early LV filling was
reduced and E/A ratio decreased (Figure 1G). Furthermore, Journal of the American Heart Association DOI: 10.1161/JAHA.116.005407 DOI: 10.1161/JAHA.116.005407 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al Responses to b-Adrenergic Stimulation and
Changes in Stimulus Interval Suggest Impaired
SR Function in R92Q but not E163R Myocardium Next, we studied the response of muscle contraction to
steady-state stimulation at different frequencies (0.2–6 Hz) at
baseline (Figure 3) and in the presence of isoproterenol
(107 mol/L) (Figure S1). While twitch amplitude at 1 Hz was by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from We then analyzed the occurrence of spontaneous contrac-
tions in R92Q, E163R, and WT trabeculae to estimate the 120
160
200
n (ms)
C
30
40
WT
R92Q
E163R
N/mm2)
B
A
WT
E163R
*
*
**
*
100 ms
TTP
R50%
R90%
0
40
80
120
Duration
Active
Passive
0
10
20
Tension (mN
10 mN/mm2
R92Q
*
D
40
60
80
eculae With
Activity (% of total)
E
10 mN/mm2
5 s
WT
R92Q
*
**
0
20
Trabe
Spontaneous
E163R
Figure 2. Steady-state isometric twitches and occurrence of spontaneous contractions during stimulation
pauses in intact trabeculae. A, Representative twitches elicited at 1 Hz in left ventricular trabeculae of wild-
type (WT), R92Q, and E163R mice. B, Passive and active twitch tension measured during 1-Hz stimulation. C, Time from stimulus to peak contraction (TTP) and time from peak to 50% and 90% relaxation (R50% and
R90%, respectively), measured in twitches at 1 Hz. D, Representative traces of the occurrence of
spontaneous activity following a stimulation pause after 3-Hz stimulation burst in the presence of
isoproterenol 107 mol/L in intact trabeculae from the 3 study groups. At variance with WT, R92Q and
E163R muscles display frequent spontaneous beats during pauses. The grey lines indicate electrical stimuli. E, Percentage of trabeculae that show spontaneous contractions following the protocol described in D. (B,
C, E) MeansSE from 9 WT (6 mice), 11 R92Q (8 mice), and 9 E163R (6 mice) trabeculae. Statistical tests:
linear mixed models with Tukey-Kramer correction (corrected for heteroschedasticity in E). *=0.05>P>0.01;
**=0.01>P>0.001. Responses to b-Adrenergic Stimulation and
Changes in Stimulus Interval Suggest Impaired
SR Function in R92Q but not E163R Myocardium 120
160
200
n (ms)
C
30
40
WT
R92Q
E163R
N/mm2)
B
A
WT
E163R
*
*
**
*
100 ms
TTP
R50%
R90%
0
40
80
120
Duration
Active
Passive
0
10
20
Tension (mN
10 mN/mm2
R92Q
*
D
40
60
80
eculae With
Activity (% of total)
E
10 mN/mm2
5 s
WT
R92Q
*
**
0
20
Trabe
Spontaneous
E163R 120
160
200
n (ms)
C
30
40
WT
R92Q
E163R
N/mm2)
B
A
WT
E163R
*
*
**
*
100 ms
TTP
R50%
R90%
0
40
80
120
Duration
Active
Passive
0
10
20
Tension (mN
10 mN/mm2
R92Q
* 40
60
80
eculae With
Activity (% of total)
E
*
**
0
20
Trabe
Spontaneous E Figure 2. Steady-state isometric twitches and occurrence of spontaneous contractions during stimulation
pauses in intact trabeculae. A, Representative twitches elicited at 1 Hz in left ventricular trabeculae of wild-
type (WT), R92Q, and E163R mice. B, Passive and active twitch tension measured during 1-Hz stimulation. C, Time from stimulus to peak contraction (TTP) and time from peak to 50% and 90% relaxation (R50% and
R90%, respectively), measured in twitches at 1 Hz. D, Representative traces of the occurrence of
spontaneous activity following a stimulation pause after 3-Hz stimulation burst in the presence of
isoproterenol 107 mol/L in intact trabeculae from the 3 study groups. At variance with WT, R92Q and
E163R muscles display frequent spontaneous beats during pauses. The grey lines indicate electrical stimuli. E, Percentage of trabeculae that show spontaneous contractions following the protocol described in D. (B,
C, E) MeansSE from 9 WT (6 mice), 11 R92Q (8 mice), and 9 E163R (6 mice) trabeculae. Statistical tests:
linear mixed models with Tukey-Kramer correction (corrected for heteroschedasticity in E). *=0.05>P>0.01;
**=0.01>P>0.001. Journal of the American Heart Association DOI: 10.1161/JAHA.116.005407 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al erimental groups, the rate-dependent
increase of contractile force in respons
2
4
6
20
40
60
80
100
Frequency (Hz)
WT
E163R
R92Q
0
2
4
6
20
40
60
80
Twitch Tension
(mN/mm2)
Frequency (Hz)
A
B
C
D
0.2 Hz
2 Hz
5 Hz
0.2 Hz
2 Hz
5 Hz
0.2 Hz
2 Hz
5 Hz
WT
E163R
R92Q
200 ms
10 mN/mm2
3 s
10 mN/mm2
§
## ##
Figure 3. Steady-state and short-term interval force relationships. ORIGINAL RESEARCH 2
4
6
20
40
60
80
100
Frequency (Hz)
WT
E163R
R92Q
0
2
4
6
20
40
60
80
Twitch Tension
(mN/mm2)
Frequency (Hz)
A
B
C
D
0.2 Hz
2 Hz
5 Hz
0.2 Hz
2 Hz
5 Hz
0.2 Hz
2 Hz
5 Hz
WT
E163R
R92Q
200 ms
10 mN/mm2
3 s
10 mN/mm2
§
## ##
Figure 3 St
d
t t
d
h
t t
i t
l f
l ti
hi
A R
t ti 2
4
6
20
40
60
80
100
Frequency (Hz)
WT
E163R
R92Q
0
2
4
6
20
40
60
80
Twitch Tension
(mN/mm2)
Frequency (Hz)
A
B
C
0.2 Hz
2 Hz
5 Hz
0.2 Hz
2 Hz
5 Hz
0.2 Hz
2 Hz
5 Hz
200 ms
10 mN/mm2
§
## ## 2
4
6
20
40
60
80
100
Frequency (Hz)
0
2
4
6
20
40
60
80
Twitch Tension
(mN/mm2)
Frequency (Hz)
B
C
§
## ## 0
2
4
6
20
40
60
80
Twitch Tension
(mN/mm2)
Frequency (Hz)
B
§ B Frequen
D
WT
3 s
10 mN/mm2 D q
y (
)
q
y (
)
D
WT
E163R
R92Q
3 s
10 mN/mm2 Figure 3. Steady-state and short-term interval force relationships. A, Representative
superimposed force traces from trabeculae of wild-type (WT), R92Q, and E163R mice,
stimulated at 0.2 Hz, 2 Hz, and 5 Hz. B, Relationship between active twitch force and
stimulation frequency (0.1–6 Hz) in trabeculae from the 3 groups of mice. §=0.05>P>0.01
(R92Q vsWT) for frequencies 0.1 to 0.5 Hz and 4 Hz and 0.01>P>0.001 for 5 to 6 Hz. C,
Time from peak to 50% relaxation (R50%, right), measured in steady-state twitches at
different stimulation frequencies. ##=0.01>P>0.001 at all frequencies. D, Representative
force traces of post-rest potentiation in trabeculae from WT, R92Q, and E163R mice;
steady-state stimulation 3 Hz, pause duration 10 seconds. E, Average percentage increase
of twitch force after different pause intervals from a basal frequency of 3 Hz in the 3 study
groups. ‡=0.05>P>0.01 at resting intervals of 20 to 60 seconds. F, Ca2+ recirculation
fraction estimated by the decline of potentiated beats following a period of high stimulation
rate in trabeculae from WT, R92Q, and E163R mice at baseline and in the presence of
isoproterenol 107 mol/L (Iso). Journal of the American Heart Association Responses to b-Adrenergic Stimulation and
Changes in Stimulus Interval Suggest Impaired
SR Function in R92Q but not E163R Myocardium A, Representative
superimposed force traces from trabeculae of wild-type (WT), R92Q, and E163R mice,
stimulated at 0.2 Hz, 2 Hz, and 5 Hz. B, Relationship between active twitch force and
stimulation frequency (0.1–6 Hz) in trabeculae from the 3 groups of mice. §=0.05>P>0.01
(R92Q vsWT) for frequencies 0.1 to 0.5 Hz and 4 Hz and 0.01>P>0.001 for 5 to 6 Hz. C,
Time from peak to 50% relaxation (R50%, right), measured in steady-state twitches at
different stimulation frequencies. ##=0.01>P>0.001 at all frequencies. D, Representative
force traces of post-rest potentiation in trabeculae from WT, R92Q, and E163R mice;
steady-state stimulation 3 Hz, pause duration 10 seconds. E, Average percentage increase
of twitch force after different pause intervals from a basal frequency of 3 Hz in the 3 study
groups. ‡=0.05>P>0.01 at resting intervals of 20 to 60 seconds. F, Ca2+ recirculation
fraction estimated by the decline of potentiated beats following a period of high stimulation
rate in trabeculae from WT, R92Q, and E163R mice at baseline and in the presence of
isoproterenol 107 mol/L (Iso). G, Mechanical restitution protocol: representative traces
from WT, R92Q, and E163R trabeculae: steady-state stimulation 1 Hz, premature interval
250 ms. H, Restitution curves show the fractional recovery of force (percent of 1 Hz
steady-state peak force) in response to the premature stimulus plotted against the
premature interval. Each curve was fitted by a single exponential to obtain the time
constant of mechanical restitution (s): mean s values are reported in the inset. (B, C through
F, H) MeanSE from 9 WT (6 mice), 11 R92Q (8 mice), and 9 E193R (6 mice) trabeculae,
Statistical tests: linear mixed models with Tukey-Kramer correction (corrected for
heteroschedasticity in C and F). *=0.05>P>0.01; **=0.01>P>0.001. echanisms in Hypertrophic Cardiomyopathy
Ferrantini et al ORIGINAL RESEARCH ORIGINAL RESEARCH ORIGINAL RESEARCH 0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
Premature Interval (s)
Relative Tension (%)
Baseline
Iso
70
80
90
100
Recirculation Fraction (%)
0.0
0.2
0.4
0.6
τ (ms)
E
F
G
H
500 ms
20 mN/mm2
WT
E163R
R92Q
*
*
**
**
***
0
10
20
30
40
50
60
200
400
600
800
Relative Tension (%)
Resting Interval (s)
‡
Figure 3. Continued. Baseline
Iso
70
80
90
100
Recirculation Fraction (%)
*
*
**
** E
F
0
10
20
30
40
50
60
200
400
600
800
Relative Tension (%)
Resting Interval (s)
‡ E 0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
Premature Interval (s)
Relative Tension (%)
0.0
0.2
0.4
0.6
τ (ms)
H
*** G
500 ms
20 mN/mm2
WT
E163R
R92Q G H by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from Figure 3. Continued. presence of isoproterenol (Figure 3F), suggesting impaired
SERCA function in the R92Q model. adrenergic stimulation mainly rely on SR function.32 The
results suggest that SR function is impaired in R92Q but not
E163R trabeculae. Relaxation times were prolonged in the 2
mutants at all investigated pacing rates, but rate adaptation of
twitch duration (ie, acceleration of contraction kinetics with
increase in stimulation frequency) was preserved in both
R92Q and E163R models (Figure 3A through 3C). The positive
lusitropic effect of b-adrenergic stimulation was also pre-
served in both mutants (Figure S1). Last, in intact trabeculae of the 3 mouse lines, we analyzed
mechanical restitution by introducing a premature stimulus
into a regular stimulus sequence at 1 Hz. The premature
“extrasystolic” contraction was reduced in amplitude com-
pared with steady-state twitches (Figure 3G). The amplitude
of the extrasystolic beat was plotted against the premature
stimulus interval to obtain mechanical restitution curves
(Figure 3F). In both R92Q and E163R versus WT trabeculae,
the rate of mechanical restitution was faster (Figure 3F
through G), suggestive of a shorter refractoriness of SR Ca2+
release27 and/or a lower Ca2+ threshold for myofilament
activation.35 Stimulation pauses of variable duration were inserted into
a steady-state 1-Hz series (Figure 3D) and the relative
increase in amplitude of the first twitch after the pause was
estimated (post-rest potentiation) and plotted against the rest
interval (Figure 3F). Maximum post-rest potentiation was
markedly lower in R92Q compared with WT and E163R
trabeculae and was reached at shorter rest intervals. ORIGINAL RESEARCH Pauses
enhance twitch force mainly by allowing more Ca2+ to fill the
SR during the prolonged diastolic period, thus favoring a
larger Ca2+ release after the pause.33 Therefore, diminished
post-rest
potentiation
of
R92Q
myocardium
suggests
impaired SR Ca2+ reuptake. ORIGINAL RESEARCH G, Mechanical restitution protocol: representative traces
from WT, R92Q, and E163R trabeculae: steady-state stimulation 1 Hz, premature interval
250 ms. H, Restitution curves show the fractional recovery of force (percent of 1 Hz
steady-state peak force) in response to the premature stimulus plotted against the
premature interval. Each curve was fitted by a single exponential to obtain the time
constant of mechanical restitution (s): mean s values are reported in the inset. (B, C through
F, H) MeanSE from 9 WT (6 mice), 11 R92Q (8 mice), and 9 E193R (6 mice) trabeculae,
Statistical tests: linear mixed models with Tukey-Kramer correction (corrected for
heteroschedasticity in C and F). *=0.05>P>0.01; **=0.01>P>0.001. the same in all experimental groups, the rate-dependent
increase of contractile force observed in WT and E163R
trabeculae at frequencies above 2 Hz was significantly
blunted in the R92Q trabeculae (Figure 3A and 2B). The increase of contractile force in response to isoproterenol was
also significantly reduced in R92Q versus WT and E163R
trabeculae (Figure S1). Mechanisms responsible of potentiat-
ing contraction in response to a high pacing rate or b- DOI: 10.1161/JAHA.116.005407 DOI: 10.1161/JAHA.116.005407 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al DOI: 10.1161/JAHA.116.005407 Sarcomere Mechanical and Kinetic Properties are
Altered in E163R but Preserved in R92Q
Myofibrils Sarcomere Mechanical and Kinetic Properties are
Altered in E163R but Preserved in R92Q
Myofibrils Next, we studied the occurrence of spontaneous Ca2+
waves and spontaneous Ca2+ transients after applying a
specific stimulation protocol, ie, a burst of high-rate stimuli
followed by a stimulation pause (Figure 4E). The occurrence
of spontaneous Ca2+ waves and spontaneous Ca2+ transients
was significantly higher in R92Q and E163R versus WT models
at baseline, and the difference increased after the adminis-
tration of isoproterenol (107 mol/L). Ventricular myofibrils from R92Q, E163R, and WT hearts,
mounted for force recording at 15°C and optimum myofila-
ment overlap, were maximally calcium-activated (pCa 4.5) and
fully relaxed (pCa 8.0) by fast solution switching, as previously
described.9 Figure 6A through 6C shows representative
traces of activation-relaxation protocols, while average data
for the 3 myofibril groups are reported in Figure 6D through
6F. R92Q myofibrils showed no changes compared with WT. Specifically, maximal isometric tension, resting tension, the
rates of tension generation and tension redevelopment, and
all kinetic parameters of tension relaxation were the same in
R92Q and WT myofibrils (Figure 6). E163R myofibrils, instead,
showed a number of significant changes compared with both
WT and R92Q. While maximal isometric tension was pre-
served (Figure 6A), resting tension was significantly increased
and tension activation and tension redevelopment were
faster, indicative of faster crossbridge turnover in the E163R
sarcomeres (Figure 6B, 6D, and 6E). Significant changes were
also found in the kinetics of tension relaxation of E163R
myofibrils (Figure 6C and 6F). As previously described,34,35
the time course of tension relaxation upon Ca2+ removal in all
myofibril groups was biphasic with an early slow relaxation
phase, which occurs while sarcomeres are isometric, followed
by a fast exponential phase, starting with the “give” of few
sarcomeres and dominated by intersarcomere dynamics. The
rate of the early phase (slow kREL), predominantly reflecting
the apparent rate with which attached crossbridges leave
force-generating states,34,35 was significantly faster in the
E163R myofibrils, indicating that this mutation may increase
the energy cost of active tension generation10 (Figure 6F). In
spite of the acceleration in kinetics of the slow isometric
relaxation phase, the overall duration of the force relaxation
transient was prolonged in E163R myofibrils compared with
both WT and R92Q myofibrils. Journal of the American Heart Association Ca2+ Handling, CaMKII Signaling, and Fibrosis in
R92Q, WT, and E163R Myocardium To further investigate the E-C coupling process, intracellular
Ca2+ measurements were performed in isolated, Fluoforte-
loaded cardiomyocytes during electrical field stimulation at
35°C. Representative traces are reported in Figure 4A and
show considerable differences between the 2 mutants. At all
stimulation frequencies tested, intracellular Ca2+ transient
decay was markedly prolonged and its peak amplitude
decreased in R92Q cardiomyocytes compared with both WT
and
E163R
cardiomyocytes. Notably,
WT
and
E163R The decay of post-rest potentiation was studied to
estimate SR Ca2+ recirculating fraction, a parameter that is
generally assumed to reflect the relative contributions of
SERCA and sarcolemma to cytosolic Ca2+ removal.34 In R92Q
compared with WT or E163R trabeculae, SR Ca2+ recirculating
fraction was significantly reduced both at baseline and in the DOI: 10.1161/JAHA.116.005407 DOI: 10.1161/JAHA.116.005407 Journal of the American Heart Association
7 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al ORIGINAL RESEARCH E163R hearts. However, the increase was more pronounced in
R92Q versus E163R myocardium (Figure 5D). cardiomyocytes exhibited the same Ca2+ transient amplitude
and kinetics (Figure 4A through 4C). Diastolic [Ca2+]i was
increased in both mutants but the largest change occurred in
the R92Q model (Figure 4D). We conclude that Ca2+ handling alterations in R92Q mice
(driven, at least in part, by altered CaMKII signaling) may
account for prolonged relaxation in trabeculae and diastolic
dysfunction observed in vivo. In E163R mice, the extent of
secondary myocardial remodeling, although qualitatively sim-
ilar to that of R92Q mice, is much more limited. Indeed, the
phenotype of E163R hearts is determined by different patho-
mechanisms. To detect potential structural changes of T-tubules, which
may affect E-C coupling,18 we labelled intact cardiomyocytes
with a membrane-selective dye and imaged them with a
confocal myocroscope. T-tubule profile was quantified with
dedicated software36 based on fast Fourier transform. The
loss of T-tubules, which was modest but statistically signif-
icant, was more pronounced in R92Q cardiomyocytes com-
pared with E163R cardiomyocytes (Figure S2). Sarcomere Mechanical and Kinetic Properties are
Altered in E163R but Preserved in R92Q
Myofibrils D, Diastolic Ca2+ levels expressed as arbitrary units of fluorescence
intensity during steady-state stimulation at different frequencies in cells from mice of the 3 cohorts. (B
through D) MeanSE from 79 WT (6 mice), 91 R92Q (7 mice), and 109 E163R (8 mice) cardiomyocytes. *P<0.05. (E) Representative traces showing the stimulation pause protocol used to elicit spontaneous Ca2+ **
F
0.1
0.2
0.3
taneous Ca2+ Waves * s-1
*
*
*
**
Basal
Iso
0.0
Spon **
E
F
2 s
2 s
Basal Conditions
Iso
0.2 F/F0
0.1
0.2
0.3
taneous Ca2+ Waves * s-1
*
*
*
**
2 s
2 s
0.2 F/F0
Basal
Iso
0.0
Spon
0.15
0.20
a2+ Transients * s-1
*
*
**
*
2 s
2 s
0.2 F/F0
Basal
Iso
0.00
0.05
0.10
Spontaneous Ca E
F
2 s
2 s
Basal Conditions
Iso
0.2 F/F0
2 s
2 s
0.2 F/F0
2 s
2 s
0.2 F/F0 E F 0.15
0.20
a2+ Transients * s-1
*
*
**
*
Basal
Iso
0.00
0.05
0.10
Spontaneous Ca 0.2 F/F0 0.2 F/F0 Figure 4. Intracellular Ca2+ measurements in intact ventricular cardiomyocytes. A, Representative
superimposed Ca2+ transients elicited at 1 Hz in wild-type (WT) and R92Q cardiomyocytes (left) and WT and
E163R cells (right). B, Time from stimulus to peak (Peak Time), time from peak to 50% and 90% decay of
Ca2+ transients (D50% and D90%, respectively) elicited at 1 Hz in cardiomyocytes from WT, R92Q, and
E163R mouse hearts. C, Amplitude of Ca2+ transients in cardiomyocytes from the 3 study groups at
different stimulation frequencies. D, Diastolic Ca2+ levels expressed as arbitrary units of fluorescence
intensity during steady-state stimulation at different frequencies in cells from mice of the 3 cohorts. (B
through D) MeanSE from 79 WT (6 mice), 91 R92Q (7 mice), and 109 E163R (8 mice) cardiomyocytes. *P<0.05. (E) Representative traces showing the stimulation pause protocol used to elicit spontaneous Ca2+
events in WT, R92Q, and E163R cardiomyocytes at basal conditions and in the presence of isoproterenol
107 mol/L (Iso). Notably, R92Q and E163R cardiomyocytes showed frequent Ca2+ waves (denoted by gray
arrows) and premature Ca2+ transients (black arrows). The black lines below the traces indicate the times of
stimulation. F, Frequency of spontaneous Ca2+ waves and spontaneous Ca2+ transients during stimulation
pauses in WT, R92Q, and E163R cardiomyocytes at basal conditions and in the presence of Iso. Sarcomere Mechanical and Kinetic Properties are
Altered in E163R but Preserved in R92Q
Myofibrils Indeed, the duration of the slow
tension decay was significantly longer while the rate of the
fast exponential phase (fast kREL) was significantly slower in
E163R myofibrils (Figure 6F). by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from To identify some of the potential mechanisms underlying the
observed Ca2+ handling differences, fast-frozen LV myocardial
samples from WT, R92Q, and E163R animals were processed to
obtain total proteins, which were used for Western blot studies
aimed at identifying changes in CaMKII signaling and other
molecular markers of cardiac remodelling, as previously
described in human myocardium (Figure 5A).22,37 Notably,
CaMKII autophosphorylation, a marker of CaMKII activation,
was increased in hearts from both R92Q and E163R compared
with WT mice. Interestingly, the increase of CaMKII autophos-
phorylation was significantly more pronounced in R92Q com-
pared with E163R hearts (Figure 5B). The total amount of
SERCA protein was reduced in both R92Q and E163R hearts
(Figure S3). Contrarily, phospholamban expression was signif-
icantly increased only in R92Q mice. Interestingly, the ratio
between SERCA and phospholamban expression levels was
markedly decreased in R92Q hearts. In E163R hearts, SERCA/
phospholamban ratio was also reduced compared with WT, but
was 3 times higher than that in R92Q. Finally, phospholamban
phosphorylation at the protein kinase A site was decreased in
both mutant lines, but the decrease was larger in R92Q hearts
(Figure S3). The observed changes of the expression and
phosphorylation of Ca2+-handling proteins are in line with the
marked reduction of the kinetics of Ca2+ transient decay
observed in R92Q cardiomyocytes. Intramyocardial fibrosis is a pathological hallmark of HCM,
both in human and in mouse models.30 We therefore explored
disease-related remodeling of extracellular matrix in the 2
transgenic lines by using Picrosirius red staining on LV tissue
sections (Figure 5C). Sarcomere Mechanical and Kinetic Properties are
Altered in E163R but Preserved in R92Q
Myofibrils The amount of intramyocardial fibrosis,
as identified by the extent of Picrosirius red–positive area in
stained tissue sections, was increased in both R92Q and DOI: 10.1161/JAHA.116.005407 DOI: 10.1161/JAHA.116.005407 8 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al ORIGINAL RESEARCH E163R
R92Q
0.5 F/F0
0.5 F/F0
A
1 5
2.0
[Ca
2+]
ence)
*
**
*
**
***
150 ms
150 ms
0
240
300
360
420
480
mes (ms)
**
**
2.5
3.0
ence)
*
*
*
B
C
D
**
**
**
*
*
*
*
1
3
5
0.5
1.0
1.5
Diastolic [
(Fluoesce
Frequency (Hz)
**
Peak Time
D50%
D90%
0
20
40
120
180
Tim
1
3
5
1.0
1.5
2.0
Ca
2+-Transient
(Δ Fluoresce
Frequency (Hz)
E
F
2 s
2 s
Basal Conditions
Iso
0.2 F/F0
0.1
0.2
0.3
taneous Ca2+ Waves * s-1
*
*
*
**
2 s
2 s
0.2 F/F0
Basal
Iso
0.0
Spon
0.15
0.20
a2+ Transients * s-1
*
*
**
*
2 s
2 s
0.2 F/F0
Basal
Iso
0.00
0.05
0.10
Spontaneous Ca
Amplitude
Figure 4. Intracellular Ca2+ measurements in intact ventricular cardiomyocytes. A, Representative
superimposed Ca2+ transients elicited at 1 Hz in wild-type (WT) and R92Q cardiomyocytes (left) and WT and
E163R cells (right). B, Time from stimulus to peak (Peak Time), time from peak to 50% and 90% decay of
Ca2+ transients (D50% and D90%, respectively) elicited at 1 Hz in cardiomyocytes from WT, R92Q, and
E163R mouse hearts. C, Amplitude of Ca2+ transients in cardiomyocytes from the 3 study groups at
different stimulation frequencies. D, Diastolic Ca2+ levels expressed as arbitrary units of fluorescence
intensity during steady-state stimulation at different frequencies in cells from mice of the 3 cohorts. (B
through D) MeanSE from 79 WT (6 mice), 91 R92Q (7 mice), and 109 E163R (8 mice) cardiomyocytes. Sarcomere Mechanical and Kinetic Properties are
Altered in E163R but Preserved in R92Q
Myofibrils *P<0 05 (E) R
t ti
t
h
i
th
ti
l ti
t
l
d t
li it
t
C 2+ E163R
R92Q
0.5 F/F0
0.5 F/F0
A
150 ms
150 ms
0 A A 1 5
2.0
[Ca
2+]
ence)
*
**
*
**
***
2.5
3.0
ence)
*
*
*
C
D
**
*
*
*
*
1
3
5
0.5
1.0
1.5
Diastolic [
(Fluoesce
Frequency (Hz)
1
3
5
1.0
1.5
2.0
Ca
2+-Transient
(Δ Fluoresce
Frequency (Hz)
Amplitude 240
300
360
420
480
mes (ms)
**
**
B
C
**
**
Peak Time
D50%
D90%
0
20
40
120
180
Tim 2.5
3.0
ence)
*
*
*
C
D
*
*
*
1
3
5
1.0
1.5
2.0
Ca
2+-Transient
(Δ Fluoresce
Frequency (Hz)
Amplitude 1 5
2.0
[Ca
2+]
ence)
*
**
*
**
***
240
300
360
420
480
mes (ms)
**
**
2.5
3.0
ence)
*
*
*
B
C
D
**
**
**
*
*
*
*
1
3
5
0.5
1.0
1.5
Diastolic [
(Fluoesce
Frequency (Hz)
Peak Time
D50%
D90%
0
20
40
120
180
Tim
1
3
5
1.0
1.5
2.0
Ca
2+-Transient
(Δ Fluoresce
Frequency (Hz)
Amplitude 1 5
2.0
[Ca
2+]
ence)
*
**
*
**
***
*
D
**
*
*
1
3
5
0.5
1.0
1.5
Diastolic [
(Fluoesce
Frequency (Hz)
5 D C B q
y (
)
**
Pe
q
y (
)
E
F
2 s
2 s
Basal Conditions
Iso
0.2 F/F0
0.1
0.2
0.3
taneous Ca2+ Waves * s-1
*
*
*
**
2 s
2 s
0.2 F/F0
Basal
Iso
0.0
Spon
0.15
0.20
a2+ Transients * s-1
*
*
**
*
2 s
2 s
0.2 F/F0
Basal
Iso
0.00
0.05
0.10
Spontaneous Ca
Figure 4. Intracellular Ca2+ measurements in intact ventricular cardiomyocytes. A, Representative
superimposed Ca2+ transients elicited at 1 Hz in wild-type (WT) and R92Q cardiomyocytes (left) and WT and
E163R cells (right). B, Time from stimulus to peak (Peak Time), time from peak to 50% and 90% decay of
Ca2+ transients (D50% and D90%, respectively) elicited at 1 Hz in cardiomyocytes from WT, R92Q, and
E163R mouse hearts. C, Amplitude of Ca2+ transients in cardiomyocytes from the 3 study groups at
different stimulation frequencies. ORIGINAL RESEARCH ORIGINAL RESEARCH B
1.6
ensity
*
*
*
*
AMK/GAPDH
286/GAPDH
K-286/CAMK
0.8
1.2
Relative Inte
CAM
p-CAMK-28
p-CAMK-2 B Figure 5. Molecular
investigation
of
calcium/calmodulin-dependent
protein
kinase II (CaMKII) signaling and myocardial fibrosis. A, Representative Western
blots for total CaMKII, phospho-CaMKII at tyrosine 287, and GAPDH. B, Average
values in wild-type (WT), R92Q, and E163R hearts (7 samples for each groups). The
intensity of individual bands was quantified and normalized to that of GAPDH. Relative intensities of WT were set at 1. Statistical test: 1-way ANOVA with Tukey
correction. *=0.05>P>0.01. C, Representative histological sections of mouse left
ventricular myocardium from WT, R92Q, and E163R mice. The left and central
images show sections stained with hematoxylin (purple, marking cell nuclei) and
Picrosirius red (pink) at different magnification; the bright pink strands denote
collagen fibers. Sections shown on the right are stained with hematoxylin and eosin. D, Fraction of the myocardium occupied by collagen fibers, expressed as
percentage of the whole section’s area, measured in samples from mice of the 3
groups. MeanSE from 9 mice per group. Five histological sections of the same
size from each mouse were analyzed to calculate the extent of Picrosirius red–
positive areas. Statistics: Kruskal–Wallis test with Dunn’s multiple comparisons
test. *=0.05>P>0.01; **=0.01>P>0.001. Figure 5. Molecular
investigation
of
calcium/calmodulin-dependent
protein
kinase II (CaMKII) signaling and myocardial fibrosis. A, Representative Western
blots for total CaMKII, phospho-CaMKII at tyrosine 287, and GAPDH. B, Average
values in wild-type (WT), R92Q, and E163R hearts (7 samples for each groups). The
intensity of individual bands was quantified and normalized to that of GAPDH. Relative intensities of WT were set at 1. Statistical test: 1-way ANOVA with Tukey
correction. *=0.05>P>0.01. C, Representative histological sections of mouse left
ventricular myocardium from WT, R92Q, and E163R mice. The left and central
images show sections stained with hematoxylin (purple, marking cell nuclei) and
Picrosirius red (pink) at different magnification; the bright pink strands denote
collagen fibers. Sections shown on the right are stained with hematoxylin and eosin. D, Fraction of the myocardium occupied by collagen fibers, expressed as
percentage of the whole section’s area, measured in samples from mice of the 3
groups. MeanSE from 9 mice per group. Five histological sections of the same
size from each mouse were analyzed to calculate the extent of Picrosirius red–
positive areas. Statistics: Kruskal–Wallis test with Dunn’s multiple comparisons
test. *=0.05>P>0.01; **=0.01>P>0.001. ORIGINAL RESEARCH Prolonged tension
relaxation
and
increased
diastolic
tension suggest that diastolic dysfunction of E163R hearts
is primarily related to sarcomere dysfunction. Mutation-driven
impairment of the blocked state of thin filaments, by allowing
recruitment of force-generating crossbridges in the absence
of Ca2+, may be responsible for both the relaxation and
diastolic tension changes found in the E163R myofibrils. This
mechanism can also lead to increased sarcomere energy
consumption at rest. obtained by simultaneously measuring isometric tension and
ATPase activity in demembranated ventricular trabeculae at
20°C. Representative recordings are reported in Figure 7A. Maximal Ca2+-activated tension was the same in all groups of
preparations (Figure 7B), in line with what observed in
myofibrils. Resting tension tended to be higher in E163R
trabeculae compared with WT and R92Q, although, at variance
with myofibrils, this difference did not reach statistical
significance (Figure 7B). Importantly, in E163R skinned tra-
beculae, both resting and maximal Ca2+-activated ATPase were
increased compared with WT and R92Q (Figure 7C). An
increase in resting ATPase was confirmed in E163R trabeculae
by measuring steady-state ATP hydrolysis from myofibril
suspensions in relaxing solution (Figure S4). The ratio between
maximal Ca2+-activated ATPase activity and active tension,
representing the energetic cost of tension generation, was Sarcomere Mechanical and Kinetic Properties are
Altered in E163R but Preserved in R92Q
Myofibrils MeanSE
from 88 WT (5 mice), 102 R92Q-KET (7 mice), and 131 R92Q-RAN (7 mice) cardiomyocytes. (B, D, F)
Statistical tests: linear mixed models with Tukey-Kramer correction (correction for heteroschedasticity was
applied in F). *=0.05>P>0.01; **=0.01>P>0.001; ***=P<0.001. Journal of the American Heart Association DOI: 10.1161/JAHA.116.005407 DOI: 10.1161/JAHA.116.005407 A
CAMKIIδ
WT
R92Q
E163R
GAPDH
B
p-CAMKIIδ
GAPDH
1.6
ensity
*
*
*
*
AMK/GAPDH
286/GAPDH
K-286/CAMK
0.8
1.2
Relative Inte
CAM
p-CAMK-28
p-CAMK-2
Figure 5. Molecular
investigation
of
calcium/calmodulin-dependent
protein
kinase II (CaMKII) signaling and myocardial fibrosis. A, Representative Western
blots for total CaMKII, phospho-CaMKII at tyrosine 287, and GAPDH. B, Average
values in wild-type (WT), R92Q, and E163R hearts (7 samples for each groups). The
intensity of individual bands was quantified and normalized to that of GAPDH. Relative intensities of WT were set at 1. Statistical test: 1-way ANOVA with Tukey
correction. *=0.05>P>0.01. C, Representative histological sections of mouse left
ventricular myocardium from WT, R92Q, and E163R mice. The left and central
images show sections stained with hematoxylin (purple, marking cell nuclei) and
Picrosirius red (pink) at different magnification; the bright pink strands denote
collagen fibers. Sections shown on the right are stained with hematoxylin and eosin. D, Fraction of the myocardium occupied by collagen fibers, expressed as
percentage of the whole section’s area, measured in samples from mice of the 3
groups. MeanSE from 9 mice per group. Five histological sections of the same
size from each mouse were analyzed to calculate the extent of Picrosirius red–
positive areas. Statistics: Kruskal–Wallis test with Dunn’s multiple comparisons
test. *=0.05>P>0.01; **=0.01>P>0.001. echanisms in Hypertrophic Cardiomyopathy
Ferrantini et al A
CAMKIIδ
WT
R92Q
E163R
GAPDH
B
p-CAMKIIδ
GAPDH
1.6
ensity
*
*
*
*
AMK/GAPDH
286/GAPDH
K-286/CAMK
0.8
1.2
Relative Inte
CAM
p-CAMK-28
p-CAMK-2
nisms in Hypertrophic Cardiomyopathy
Ferrantini et al Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al A
CAMKIIδ
WT
R92Q
E163R
GAPDH
p-CAMKIIδ
GAPDH
nisms in Hypertrophic Cardiomyopathy
Ferrantini et al ORIGINAL RESEARCH ORIGINAL RESEARCH C
WT
R92Q
200μm
40μm
30μm
E163R
0
2
4
6
8
WT
R92Q
E163R
Fibrosis (% Picrosirius Red Area)
**
*
D
Figure 5. Continued. C
WT
R92Q
200μm
40μm
30μm
E163R
0
2
4
6
8
WT
R92Q
E163R
Fibrosis (% Picrosirius Red Area)
**
*
D
Figure 5. Continued. C by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from D Figure 5. Continued. markedly higher in E163R compared with both WT and R92Q
trabeculae (Figure 7D). The result was confirmed by measuring
tension cost from the slope of the ATPase/tension plot
obtained in skinned trabeculae by simultaneously measuring
isometric ATPase and tension at different levels of Ca2+
activation (Figure S5). contribution of primary, mutation-driven, sarcomeric changes
versus adverse cardiomyocyte remodelingtothe developmentof
HCM phenotype is unclear, and, in the present work, we suggest
that it varies depending on the underlying mutation. Previous
work on human samples17,22 described a number of “disease-
specific” aspects of HCM-related myocardial remodeling, differ-
ent from those observed in heart failure or secondary hypertro-
phy and without a major impact of the specific patient genotype. Here, we characterized in vivo heart phenotype and in vitro
biophysical changes in sarcomere function and E-C coupling in 2
HCM mouse models carrying different mutations in cardiac TnT
and found that HCM pathophysiology is “mutation specific”
rather than “disease specific” (Figure 8). Mean pCa-active tension curves obtained from skinned
trabeculae showed that myofilament Ca2+ sensitivity was
significantly increased in both R92Q and E163R compared
with WT preparations (Figure 7E). The change was milder in
the E163R trabeculae and much more marked in the R92Q
preparations that showed the highest pCa50 value. Journal of the American Heart Association
1 Energetic Changes are Only Observed in E163R
Skinned Trabeculae, While Myofilament Ca2+
Sensitivity is Enhanced in Both Mutants Direct demonstration of the specific impact of E163R on
sarcomere energetics at rest and during contraction was 10 Journal of the American Heart Association DOI: 10.1161/JAHA.116.005407 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al Similar Phenotypes Through Different Pathogenic
Pathways Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al WT
E163R
R92Q
A
50 mN mm
-2
0.2 l0
1 s
1 s
1 s
B
C
kACT
Fast kREL
Slow kREL , Dslow
0.2 s
0.1 s
Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al ORIGINAL RESEARCH WT
E163R
R92Q
A
50 mN mm
-2
0.2 l0
1 s
1 s
1 s A B
C
kACT
Fast kREL
Slow kREL , Dslow
0.2 s
0.1 s C
Fast kREL
Slow kREL , Dslow
0 1 s B
C
kACT
0.2 s C B by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from 2.4
40
*
*
*
**
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
Slow kREL
10
15
20
25
30
35
40
Fast kREL (s
-1)
*
* F
110
2.4
40
**
*
*
*
*
**
60
70
80
90
100
110
Dslow (ms)
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
Slow kREL
10
15
20
25
30
35
40
Fast kREL (s
-1)
**
*
* D
E
F
14
2)
8.5
7.5
110
2.4
40
**
**
**
*
*
*
**
*
*
*
*
**
4
6
8
10
12
14
Resting Tension (mN/mm
2
6.0
6.5
7.0
7.5
8.0
8.5
kACT (s
-1)
5.0
5.5
6.0
6.5
7.0
7.5
kTR (s
-1)
60
70
80
90
100
110
Dslow (ms)
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
Slow kREL
10
15
20
25
30
35
40
Fast kREL (s
-1)
**
**
*
**
*
* E
8.5
7.5
**
*
*
*
6.0
6.5
7.0
7.5
8.0
8.5
kACT (s
-1)
5.0
5.5
6.0
6.5
7.0
7.5
kTR (s
-1)
**
* E D
2)
2 E D Figure 6. Mechanical measurements on single myofibrils. A, Tension generation and relaxation: representative tension responses of wild-type
(WT), E163R, and R92Q myofibrils, maximally activated (pCa 4.5) and fully relaxed (pCa 8) by fast solution switching technique at 15°C, [MgATP]
5 mmol/L; [Pi] <5 lmol/L. The timing of solution switch is indicated by black arrows. (B through C) Superimposed representative tension
curves from single myofibrils isolated from WT, R92Q, and E163R hearts, highlighting the kinetics of force generation (B) and relaxation (C). Journal of the American Heart Association Similar Phenotypes Through Different Pathogenic
Pathways In HCM, functional alterations at the sarcomere level9,14 are
associated with secondary modifications in E-C coupling that are
responsiblefortheproarrhythmogenicphenotypeandcontribute
to alterations of cardiac mechanical function.22 The relative Findings from echocardiographic and Doppler measurements
showed that several hallmarks of human HCM phenotype 11 Journal of the American Heart Association
11 DOI: 10.1161/JAHA.116.005407 WT
E163R
R92Q
A
50 mN mm
-2
0.2 l0
1 s
1 s
1 s
B
C
kACT
Fast kREL
Slow kREL , Dslow
0.2 s
0.1 s
D
E
F
14
2)
8.5
7.5
110
2.4
40
**
**
**
*
*
*
**
*
*
*
*
**
4
6
8
10
12
14
Resting Tension (mN/mm
2
6.0
6.5
7.0
7.5
8.0
8.5
kACT (s
-1)
5.0
5.5
6.0
6.5
7.0
7.5
kTR (s
-1)
60
70
80
90
100
110
Dslow (ms)
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
Slow kREL
10
15
20
25
30
35
40
Fast kREL (s
-1)
**
**
*
**
*
*
Figure 6. Mechanical measurements on single myofibrils. A, Tension generation and relaxation: representative tension responses of wild-type
(WT), E163R, and R92Q myofibrils, maximally activated (pCa 4.5) and fully relaxed (pCa 8) by fast solution switching technique at 15°C, [MgATP]
5 mmol/L; [Pi] <5 lmol/L. The timing of solution switch is indicated by black arrows. (B through C) Superimposed representative tension
curves from single myofibrils isolated from WT, R92Q, and E163R hearts, highlighting the kinetics of force generation (B) and relaxation (C). Of
note, relaxation is biphasic, with an initial linear phase (slow) followed by an exponential decay of force (fast phase). D, Resting tension in WT,
R92Q, and E163R myofibrils. E, Kinetic parameters of tension generation in WT, R92Q, and E163R myofibrils: time constant of tension activation
(kACT, left) and of tension redevelopment upon myofibril relengthening after forced slacking (kTR, right). F, Kinetic parameters of myofibri
relaxation: duration of the slow linear phase (Dslow, left) and time constant of the slow phase of relaxation (Slow kREL, center) and of the fast
exponential phase (Fast kREL, right). (D through F) MeanSE from 32 WT, 34 R92Q, and 36 E163R myofibrils, 3 mice per group. Statistical tests:
linear mixed models with Tukey-Kramer correction. *=0.05>P>0.01; **=0.01>P>0.001. ORIGINAL RESEARCH ORIGINAL RESEARCH A
WT
E163R
R92Q
mm2
30 mN/
300 pmol ATP
B
60
80
mm
2)
50 ms
E
14
16
st
N/mm
2)
*
**
1000
1200
μl
-1 s
-1)
C
*
*
Maximal Tension
Resting Tension
0
20
40
Tension (mN/m
6
8
10
12
Tension cos
(pmol μl
-1 s
-1/ mN
imal ATPase
ting ATPase
0
200
400
600
800
ATPase (pmol μ
*
**
Maxima
Resting
D
0 6
0.8
1.0
(%)
6 0
6.2
6.4
*
**
*
Maximal A
Resting A
9
8
7
6
5
4
3
0.0
0.2
0.4
0.6
Tension
pCa
5.4
5.6
5.8
6.0
PCa 50
p
Figure 7. Isometric tension and ATPase measurements in skinned ventricular trabeculae. A, Representative
tension (top) and ATPase activity (bottom) traces from wild-type (WT), R92Q, and E163R trabeculae that were
maximally Ca2+-activated (pCa 4.5) at 20°C. (B through D) MeanSE of maximal active tension (P0) and resting
tension (B), maximal Ca2+-activated ATPase and resting ATPase activities (C), tension cost estimated as the
ratio between maximal Ca2+-activated ATPase and maximal active tension (D) in WT (n=16), R92Q (n=12), and
E163R (n=15) trabeculae. (E) Tension-pCa curves (left) and mean values of pCa at half-maximal activation
(pCa50, right) from WT (n=10), R92Q (n=7), and E163R (n=9) trabeculae. Statistical tests: linear mixed models
with Tukey-Kramer correction. *=0.05>P>0.01; **=0.01>P>0.001. Similar Phenotypes Through Different Pathogenic
Pathways Of
note, relaxation is biphasic, with an initial linear phase (slow) followed by an exponential decay of force (fast phase). D, Resting tension in WT,
R92Q, and E163R myofibrils. E, Kinetic parameters of tension generation in WT, R92Q, and E163R myofibrils: time constant of tension activation
(kACT, left) and of tension redevelopment upon myofibril relengthening after forced slacking (kTR, right). F, Kinetic parameters of myofibril
relaxation: duration of the slow linear phase (Dslow, left) and time constant of the slow phase of relaxation (Slow kREL, center) and of the fast
exponential phase (Fast kREL, right). (D through F) MeanSE from 32 WT, 34 R92Q, and 36 E163R myofibrils, 3 mice per group. Statistical tests:
linear mixed models with Tukey-Kramer correction. *=0.05>P>0.01; **=0.01>P>0.001. and resting tension tended to increase in both HCM mice
compared with WT mice (Figure 2A through C), in line with the
diastolic dysfunction observed in vivo (Figure 1F and 1G). Compared with WT, both HCM models exhibited a significantly
higher frequency of spontaneous activity during pauses,
following a period of high-frequency stimulation in the (increased septal thickness, increased ejection fraction,
diastolic dysfunction) are present in both HCM mouse models,
with the phenotype more pronounced in the R92Q mice
(Figure 1). Mechanical experiments in intact trabeculae
demonstrated that (in the absence of a switch in myosin
isoforms) isometric twitch relaxation was markedly prolonged Journal of the American Heart Association DOI: 10.1161/JAHA.116.005407 12 DOI: 10.1161/JAHA.116.005407 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al ORIGINAL RESEARCH 13 DOI: 10.1161/JAHA.116.005407 Journal of the American Heart Association DOI: 10.1161/JAHA.116.005407 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al ORIGINAL RESEARCH Impaired Thin-filament switch-off
↑↑ Ca2+ sensivity
↑CaMKII acvity
↑ ↑ Diastolic Ca2+ levels
↓SERCA acvity
↑ Diastolic Ca2+ levels
↑ Ca2+ sensivity
↑Tension Cost
Increased myofilament resng tension
Major role of
Sarcomeric changes
HCM
phenotype
R92Q
E163R
Major role of
EC coupling alteraons
RyR2
Na+
channel
Ca2+
channel
NCX
SERCAPLB
↑ CaMKII acvity
LV hypertrophy, diastolic dysfuncon, arrhythmias
Figure 8. Different mutation-specific mechanisms lead to similar phenotypes. Scheme depicting the
proposed mutation-specific molecular and cellular mechanisms leading to hypertrophic cardiomyopathy
(HCM)–related cardiac structural and functional abnormalities in mice carrying the R92Q or E163
mutations. The figure includes 2 short-axis magnetic resonance scans (mid-left ventricular [LV] level) at end
diastole from mice of the 2 strains, obtained with a 7T magnetic resonance setup. CaMKII indictes calcium/
calmodulin-dependent protein kinase II; E-C, excitation-contraction. E163R ↑Tension Cost by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from Major role of
Sarcomeric changes
HCM
phenotype
Major role of
EC coupling alteraons
LV hypertrophy, diastolic dysfuncon, arrhythmias Major role of
Sarcomeric changes LV hypertrophy, diastolic dysfuncon, arrhythmias Figure 8. Different mutation-specific mechanisms lead to similar phenotypes. Scheme depicting the
proposed mutation-specific molecular and cellular mechanisms leading to hypertrophic cardiomyopathy
(HCM)–related cardiac structural and functional abnormalities in mice carrying the R92Q or E163
mutations. The figure includes 2 short-axis magnetic resonance scans (mid-left ventricular [LV] level) at end
diastole from mice of the 2 strains, obtained with a 7T magnetic resonance setup. CaMKII indictes calcium/
calmodulin-dependent protein kinase II; E-C, excitation-contraction. presence of isoproterenol (Figure 2D and 2E). Previous
telemetry
evaluations
in
the
R92Q
models
highlighted
increased occurrence of both atrial and ventricular arrhyth-
mias,38 while no evaluation of the arrhythmic burden was
performed on living E163R mice. R92Q cardiomyocytes exhibited smaller and slower Ca2+
transients compared with WT, providing a reasonable expla-
nation for diastolic dysfunction in vivo. Differently, in E163R
cardiomyocytes, the kinetics of Ca2+ transients were normal
(Figure 4) and no major E-C coupling changes were found that
could explain prolonged twitch duration in trabeculae. ORIGINAL RESEARCH A
WT
E163R
R92Q
mm2
30 mN/
300 pmol ATP
B
60
80
mm
2)
50 ms
E
14
16
st
N/mm
2)
*
**
1000
1200
μl
-1 s
-1)
C
*
*
Maximal Tension
Resting Tension
0
20
40
Tension (mN/m
6
8
10
12
Tension cos
(pmol μl
-1 s
-1/ mN
imal ATPase
ting ATPase
0
200
400
600
800
ATPase (pmol μ
*
**
Maxima
Resting
D
0 6
0.8
1.0
(%)
6 0
6.2
6.4
*
**
*
Maximal A
Resting A
0.4
0.6
nsion
5 8
6.0
Ca 50 A
WT
E163R
R92Q
mm2
30 mN/
300 pmol ATP
50 ms A B
60
80
mm
2)
E
14
16
st
N/mm
2)
*
**
1000
1200
μl
-1 s
-1)
C
*
*
Maximal Tension
Resting Tension
0
20
40
Tension (mN/m
6
8
10
12
Tension cos
(pmol μl
-1 s
-1/ mN
imal ATPase
ting ATPase
0
200
400
600
800
ATPase (pmol μ
*
**
Maxima
Resting
Maximal A
Resting A E
14
16
st
N/mm
2)
*
**
6
8
10
12
Tension cos
(pmol μl
-1 s
-1/ mN
e C E B D
0 6
0.8
1.0
(%)
6 0
6.2
6.4
*
**
*
9
8
7
6
5
4
3
0.0
0.2
0.4
0.6
Tension
pCa
5.4
5.6
5.8
6.0
PCa 50
p D D Figure 7. Isometric tension and ATPase measurements in skinned ventricular trabeculae. A, Representative
tension (top) and ATPase activity (bottom) traces from wild-type (WT), R92Q, and E163R trabeculae that were
maximally Ca2+-activated (pCa 4.5) at 20°C. (B through D) MeanSE of maximal active tension (P0) and resting
tension (B), maximal Ca2+-activated ATPase and resting ATPase activities (C), tension cost estimated as the
ratio between maximal Ca2+-activated ATPase and maximal active tension (D) in WT (n=16), R92Q (n=12), and
E163R (n=15) trabeculae. (E) Tension-pCa curves (left) and mean values of pCa at half-maximal activation
(pCa50, right) from WT (n=10), R92Q (n=7), and E163R (n=9) trabeculae. Statistical tests: linear mixed models
with Tukey-Kramer correction. *=0.05>P>0.01; **=0.01>P>0.001. ORIGINAL RESEARCH ORIGINAL RESEARCH similarities were observed. Myofilament Ca2+ sensitivity in
skinned trabeculae was increased in both HCM mutant lines
compared with WT, although the change was much greater in
the R92Q model (Figure 7). Intramyocardial fibrosis was
increased in both mouse lines as compared with WT, but the
extent was much larger in R92Q hearts (Figure 4). In
cardiomyocytes, diastolic Ca2+ levels and CaMKII autophos-
phorylation were abnormally elevated in both mutant lines,
with greater changes occurring in the R92Q model (Figure 4). Increased myofilament Ca2+ sensitivity and cytosolic Ca2+
overload may be plausibly linked to the increased rate of
arrhythmias observed with both mutations. Overall, these
results indicate that similar HCM phenotypes can be gener-
ated through different pathogenic pathways (see scheme in
Figure 8). In the E163R model, mutation-driven abnormalities
in myofilament function largely account for cardiac dysfunc-
tion. In the R92Q model, instead, profound E-C coupling
changes related to aggressive cardiomyocyte remodeling
seem to drive the cardiac HCM phenotype (Figure 8). detachment rate in addition to its ability to control the
attachment of crossbridges to the thin filament. However, the
potential molecular mechanism of this effect of the E163R
mutation remains unclear. In spite of the faster rate of the slow isometric phase of
relaxation following Ca2+ removal, overall relaxation from
maximal activation was prolonged in E163R myofibrils
(slower fast kREL and prolonged duration of the early
isometric phase), while resting tension was significantly
higher compared with WT and R92Q myofibrils (Figure 6). This behavior, together with the increase in resting tension,
suggests that in E163R sarcomeres, the thin filament may
not turn off properly upon Ca2+ removal and may be unable
to fully maintain its switched-off state during diastole, in
analogy with previously described HCM-related tropomyosin
mutations.41–43 The modest increase in myofilament Ca2+
sensitivity in E163R myocardium is in line with previous
findings44 and may be a consequence of the impaired thin
filament blocked state that allows recruitment of some
force-generating
crossbridges
at
low
Ca2+.24,45,46
We
excluded that the increase of myofilament Ca2+ sensitivity
depended on a reduction of troponin I phosphorylation at the
protein kinase A site (Figure S3), although we have not
explored other possible posttranslational modifications of
sarcomere proteins (eg, MyPBC3). The observed increase of
resting ATPase in E163R skinned trabeculae and myofibrils
is likely a direct consequence of residual thin filament
activation at low [Ca2+]. ORIGINAL RESEARCH Residual thin filament activation
following Ca2+ removal may promote formation of new
crossbridges,41–43 contributing to slow down relaxation, and
increase diastolic tension in E163R myocardium. These
sarcomeric abnormalities are likely to be the basis of the
prolonged twitch contraction and the impaired diastolic
function observed in vivo (Figure 8). by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from In addition to the comprehensive biophysical character-
ization in 6-month-old mice, data obtained at different ages,
including 3 months and 12 months, are reported in Fig-
ure S6. Studies of 3-month-old mutant mice demonstrated
that both sarcomere and E-C coupling changes are already
present, although in R92Q mice, some of the intracellular
calcium alterations (eg, increased diastolic Ca2+ levels,
prolonged Ca2+ transient duration) are more severe in older
animals, while no progression was observed in the E163R
model. ORIGINAL RESEARCH Exper-
iments performed in single myofibrils and skinned ventricular
trabeculae showed that E163R uniquely alters the switched-
off state of the thin filament in the absence of Ca2+
(Figures 6, 7 and Figure S4), thus directly impairing the
mechanics and energetics of cardiac relaxation and diastole. In addition, simultaneous measurements of isometric tension
and ATPase demonstrated that the energy cost of tension
generation was increased in E163R but unchanged in R92Q
trabeculae (Figure 7 and Figure S5), the latter in line with
previous observations.28 Despite these differences, some In spite of the similarities between the 2 mouse models,
further experiments investigating the inotropic reserve of
intact trabeculae indicated that the function of the Ca2+-
handling machinery was substantially different (Figure 3 and
Figure S1). While E163R trabeculae behaved like WT, R92Q
trabeculae exhibited a blunted response to changes in
stimulation frequency and to other inotropic interventions,
as
previously
shown
in
Langendorff-perfused
working
hearts.39 Experiments in isolated cardiomyocytes confirmed
that E-C coupling abnormalities were mostly present in the
R92Q mouse model (Figure 4). As previously described,38 14 Journal of the American Heart Association DOI: 10.1161/JAHA.116.005407 DOI: 10.1161/JAHA.116.005407 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al Journal of the American Heart Association
1 Secondary E-C Coupling Changes are the Major
Cause for Contractile Dysfunction in R92Q transient decay, resulting in increased end-diastolic [Ca2+].35
Increased myofilament Ca2+ sensitivity in combination with
reduced SERCA activity can explain the elevated diastolic
[Ca2+]i and excessive increase in baseline [Ca2+]i at high
stimulation frequencies, which were particularly prominent in
R92Q myocytes. In E163R myocardium, we found a slight
reduction of SERCA expression and a minimal increase of
phospholamban expression and phosphorylation compared
with WT myocardium (Figure S3). Such changes were much
less pronounced with respect to R92Q hearts and are unlikely
to determine a significant slowing of SERCA Ca2+-reuptake
rate, as indicated by the preserved Ca2+ transient decay rate
in E163R myocytes. Therefore, increased myofilament Ca2+
sensitivity could be the sole cause of the increase of diastolic
[Ca2+]i observed in E163R myocardium.35 Diastolic dysfunction evaluated in vivo through echocardiog-
raphy was more severe in R92Q compared with E163R mice. Mechanical measurements in intact trabeculae confirmed that
twitch
contraction
was
prolonged. In
addition,
despite
preserved baseline systolic contraction, inotropic reserve
was reduced in R92Q, at variance with E163R preparations
(Figure 2 and 3). Reduced contractile reserve is a feature of
adverse remodeling and disease progression in human HCM,
previously observed in transgenic mice carrying different TnT
mutations in the same locus.47 Approximately 20% of HCM
patients showed an insufficient blood pressure response to
exercise and 10% displayed overt systolic dysfunction (ejec-
tion fraction <50%).3 Both features are validated predictors of
adverse outcome and sudden cardiac death in patients with
HCM.48 As previously shown in human HCM myocardium,22 a
sustained increase in diastolic [Ca2+]i is associated with
enhanced activation of CaMKII, as confirmed by the higher
degree of CaMKII autophosphorylation observed in R92Q
myocardium. Increased CaMKII activity plays a central role in
driving cellular and extracellular remodeling in a number of
genetic and acquired cardiac diseases,52 including HCM,22
and may play a leading role in R92Q mice. In line with the
lesser degree of cellular remodeling and the less prominent
increase of diastolic [Ca2+]i in E163R versus R92Q myocytes
(Figure 4), the increase of CaMKII autophosphorylation was
less pronounced in E163R myocardium as compared with
R92Q. CaMKII activates transcription factors that drive
increased myocardial fibrosis in cardiac diseases.53 In line
with that, fibrosis was more pronounced in hearts from R92Q
mice compared with E163R mice. Secondary E-C Coupling Changes are the Major
Cause for Contractile Dysfunction in R92Q Of note, both mouse lines
well
reproduce
the
increased
CaMKII
activity,
recently
observed in myectomy samples from HCM patients22,37 and
described in humans as a hallmark of myofilament-positive
HCM. by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from At variance with E163R cells, R92Q cardiomyocytes
display severe prolongation of Ca2+ transient decay, which
can account for prolonged twitch duration and contribute to
the diastolic dysfunction observed in vivo (Figure 8). A
number of mechanisms may contribute to this aspect. First,
we confirmed47 that the reduced SERCA activity underlies the
slower kinetics of Ca2+ transients in R92Q cells and is caused
by a lower SERCA expression paralleled by increased
phospholamban expression and phosphorylation levels (Fig-
ure S3). The reduction of SR Ca2+ recirculation fraction
observed in R92Q trabeculae further confirms the reduction of
SERCA activity.26 Interestingly, cardiomyocytes isolated from
septal samples of patients with obstructive HCM undergoing
myectomy consistently displayed slower Ca2+ transients,
associated with reduced SERCA expression.22 Furthermore,
slower Ca2+ transient kinetics and reduction in SERCA
expression have been observed in the majority of animal
models with experimental LVH and LV failure and in
cardiomyocytes from patients with terminal heart failure.49
Thus, it does not represent a disease-specific feature. Similarly, as in human HCM,50 we describe a certain degree
of T-tubule disruption in R92Q cardiomyocytes, in line with our
previous data on D160 cardiac TnT mutant mouse model.18
The extent of T-tubule structural remodeling in R92Q
cardiomyocytes, however, is much less pronounced compared
with models of secondary hypertrophy and failure51 and is
likely to play only a minor role in the slowing of Ca2+ transient
and contraction kinetics. In parallel with a large number of
HCM mutations in thin and thick filament proteins, we
observed a large increase in myofilament Ca2+ sensitivity in
R92Q skinned trabeculae, likely related to higher Ca2+ binding
affinity of troponin complexes incorporating the mutant
protein.46 As a consequence, myofilament Ca2+ dissociation
is slower and may contribute to prolong the last phase of Ca2+ Other pathways of hypertrophic remodeling, including the
nuclear factor of activated T cell pathway, do not appear to be
changed in either of the 2 mutant lines (Figure S3), consistent
with studies on human samples.37 ORIGINAL RESEARCH ORIGINAL RESEARCH Primary Sarcomeric Changes Directly Impair
Energetics and Diastolic Function in E163R
Hearts While all mechanical and kinetic parameters of maximally
activated R92Q myofibrils were the same as those of WT
myofibrils, in E163R myofibrils, kinetics of force generation
(tension activation and tension redevelopment) and isometric
relaxation (slow kREL) were significantly faster compared with
both WT and R92Q, preparations suggesting that the mutation
primarily affects crossbridge kinetics and energetics. Accord-
ing to a simple crossbridge model, these results indicate that
a modest (but significant) increase in gapp (the apparent rate
with which crossbridges leave their force-generating state
under isometric conditions) in E163R mice accounts for both
faster crossbridge turnover and increased tension cost,
without significantly affecting maximal tension generation.9
As to the impact of E163R on gapp, pioneer studies reviewed
by Gordon et al40 reported that HCM-associated cardiac TnT
mutations may cause changes in regulated acto-S1 ATPase
and unloaded thin filament sliding speed. This implies that
cardiac
TnT
can
modulate
strongly
bound
crossbridge While these considerations suggest that altered sarcomere
function is the main pathogenic element underlying impaired
relaxation in E163R myocardium, a contribution of other
mechanisms cannot be excluded. For instance, previous
work24 described considerable degrees of myofibril disarray
and Z-line misalignment in E163R cardiomyocytes. This is
consistent with the modest structural disorganization of T-
tubules observed in E163R myocytes and may contribute to
impair cardiomyocyte relaxation properties.26 Moreover, the expression of mutant TnT protein in the
R92Q line is 67%,13,23 while it is 50% in the E163R line.24
Such a difference is unlikely to have accounted for the
observed divergences in myofilament energetics and myofibril
relaxation between the 2 lines; however, we cannot exclude
that the higher expression of mutant protein in R92Q mice
affects some features of myocardial dysfunction in mutant
mice (eg, myofilament Ca2+ sensitivity) and acknowledge this
issue as a limitation of our work. 15 Journal of the American Heart Association DOI: 10.1161/JAHA.116.005407 DOI: 10.1161/JAHA.116.005407 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al ORIGINAL RESEARCH ORIGINAL RESEARCH ORIGINAL RESEARCH trabeculae was likely an increased rate of spontaneous Ca2+
waves and Ca2+ transients in cardiomyocytes, providing a
trigger for propagated contractions. In recent work on a
mouse model of catecholaminergic polymorphic ventricular
tachycardia,27
we
showed
that
Ca2+-dependent
cellular
arrhythmias depend on the increased open probability of
the ryanodine receptor (RyR), and that increased RyR open
probability alone causes faster mechanical restitution. Indeed,
in both R92Q and E163R mutants, and previously in the D160
TnT mouse line,18 we observed a faster rate of mechanical
restitution. This suggests that in TnT mutants, Ca2+ waves are
mediated by an increased RyR open probability during
diastole. The increase in diastolic cytosolic [Ca2+]i, observed
in cells from both TnT mouse lines, is the most direct
explanation for the observed instability of RyR channels, as
the main trigger for RyR opening is [Ca2+] at the dyad level. However, in our experimental conditions, maximal diastolic
[Ca2+]i is observed at high pacing rates, while Ca2+ waves are
more frequent during prolonged pauses. An alternative
explanation has been provided by Knollman et al,35 who
found that in HCM-linked TnT mutations, high myofilament
Ca2+ sensitivity leads to increased cytosolic Ca2+-buffering
capacity. The excessive amount of Ca2+ that binds to the
myofilaments is slowly released during pauses and accumu-
lates in the SR, determining SR Ca2+ overload, thus enhancing
RyR open probability through luminal control.54 This may
represent the main trigger for Ca2+ waves and delayed
afterdepolarizations during pauses. The proposed mechanism
occurs only in the presence of a mutation that increases the
binding rate of Ca2+ to the troponin complex. This is likely the
case for R92Q mutation. Whether the E163R mutation affects
Ca2+ binding to myofilaments and intracellular Ca2+ buffering
remains to be determined. cardiomyocytes, the late Na+-current blocker ranolazine
(10 lmol/L) significantly hastened Ca2+ transient kinetics
and reduced diastolic Ca2+ and the rate of spontaneous beats
and Ca2+ waves.30 Additionally, lifelong treatment with
ranolazine prevented most of the aspects of cardiac remodeling
and dysfunction in R92Q mice.30 Contrarily, recent unpublished
experiments show that the acute administration of ranolazine is
unable to exert similar effects on E163R cardiomyocytes. Sources of Funding This work was supported by Telethon Italy (GGP13162 and
GGP16191),
the
European
Commission
(STREP
Project
241577 “BIG HEART,” 7th European Framework Program),
the Italian Ministry of Health (GR-2011-02350583, RF-2013-
02356787
and
NET-2011-02347173),
Regione
Toscana
(FAS-Salute 2014, ToRSADE project), and the National
Institues of Health (projects HL075619 and HL62426). ORIGINAL RESEARCH The same can be said for drugs that act on extracellular
remodeling, such as sartans or statins, which have shown
promising results in transgenic HCM animal models with
selected “severe” mutations59,60 but failed to show convinc-
ing results in patients who were not selected based on
genotype.61,62 Conversely, in the presence of sarcomeric
mutations leading to cardiac dysfunction through direct
mechanisms, sarcomere-targeting drugs may be preferred,
such as novel myosin inhibitors, currently in the pipeline for
the treatment of HCM.63 by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from Implications for Human Disease Our data support the view that different mutations, even within
the same gene, can generate cardiac dysfunction with
markedly different mechanisms. While diastolic dysfunction
and arrhythmogenesis are common consequences of all HCM
mutations, a complex structural and functional remodeling of
cardiomyocytes may occur only with certain variants, such as
R92Q, and may not be present in others, such as E163R. Interestingly, mutations at the R92 site of TNNT2 are regarded
as severe, with a high risk of lethal arrhythmias during
adolescence and early adulthood,55,56 while mutations at the
E163 site lead to milder forms of HCM.57 The notion that
cellular remodeling is mutation specific may be relevant for
treatment. Drugs targeting membrane ion channels, such as
late Na-current blockers22 or diltiazem,58 may be more
effective in the presence of significant cardiomyocyte E-C
coupling remodeling. We previously showed that in R92Q Conclusions Based on the results of the present investigation, we envision
a future where the treatment of every patient will be decided
based on his/her specific mutation, driven by studies of the
specific mutation-linked mechanisms of disease. Although we
understand that producing animal models as a predictive tool
is not practical, cardiomyocytes differentiated from patient-
specific induced pluripotent stem cells may be a feasible tool
to assess mutation-specific pathomechanisms and predict the
efficacy of drugs in individual patients.19,64 1. Semsarian C, Ingles J, Maron MS, Maron BJ. New perspectives on the
prevalence of hypertrophic cardiomyopathy. J Am Coll Cardiol. 2015;65:1249–
1254. Cellular Mechanisms of Arrhythmias in R92Q and
E163R Myocardium: Open Questions Increased arrhythmogenesis in hearts from patient and animal
models with HCM is the result of pathological changes that
occur both at the cell and tissue level. The latter include
intramyocardial
fibrosis, myocyte disarray, and
regional
hypertrophy, which all promote the establishment of reentrant
arrhythmias. In the present work, we found that the mech-
anism underlying the increased rate of premature beats
during prolonged pauses and b-adrenergic stimulation in Journal of the American Heart Association 16 DOI: 10.1161/JAHA.116.005407 DOI: 10.1161/JAHA.116.005407 Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al 2. Ho CY, Charron P, Richard P, Girolami F, Van Spaendonck-Zwarts KY, Pinto Y.
Genetic advances in sarcomeric cardiomyopathies: state of the art. Cardiovasc
Res. 2015;105:397–408. References Decreased energetics in murine
hearts bearing the R92Q mutation in cardiac troponin T. J Clin Investig. 2003;112:768–775. 32. Briston SJ, Dibb KM, Solaro RJ, Eisner DA, Trafford AW. Balanced changes in Ca
buffering by SERCA and troponin contribute to Ca handling during beta-
adrenergic stimulation in cardiac myocytes. Cardiovasc Res. 2014;104:347–
354. 14. Witjas-Paalberends ER, Guclu A, Germans T, Knaapen P, Harms HJ, Vermeer
AM, Christiaans I, Wilde AA, Dos Remedios C, Lammertsma AA, van Rossum
AC, Stienen GJ, van Slegtenhorst M, Schinkel AF, Michels M, Ho CY, Poggesi C,
van der Velden J. Gene-specific increase in the energetic cost of contraction in
hypertrophic cardiomyopathy caused by thick filament mutations. Cardiovasc
Res. 2014;103:248–257. 33. Maier LS, Bers DM, Pieske B. Differences in Ca(2+)-handling and sarcoplasmic
reticulum Ca(2+)-content in isolated rat and rabbit myocardium. J Mol Cell
Cardiol. 2000;32:2249–2258. 34. Roe AT, Frisk M, Louch WE. Targeting cardiomyocyte Ca2+ homeostasis in
heart failure. Curr Pharm Des. 2015;21:431–448. 15. Hoskins AC, Jacques A, Bardswell SC, McKenna WJ, Tsang V, dos Remedios
CG, Ehler E, Adams K, Jalilzadeh S, Avkiran M, Watkins H, Redwood C, Marston
SB, Kentish JC. Normal passive viscoelasticity but abnormal myofibrillar force
generation in human hypertrophic cardiomyopathy. J Mol Cell Cardiol. 2010;49:737–745. 35. Schober T, Huke S, Venkataraman R, Gryshchenko O, Kryshtal D, Hwang HS,
Baudenbacher FJ, Knollmann BC. Myofilament Ca sensitization increases
cytosolic Ca binding affinity, alters intracellular Ca homeostasis, and causes
pause-dependent Ca-triggered arrhythmia. Circ Res. 2012;111:170–179. 16. Willott RH, Gomes AV, Chang AN, Parvatiyar MS, Pinto JR, Potter JD. Mutations
in troponin that cause HCM, DCM and RCM: what can we learn about thin
filament function? J Mol Cell Cardiol. 2010;48:882–892. 36. Guo A, Song LS. AutoTT: automated detection and analysis of T-tubule
architecture in cardiomyocytes. Biophys J. 2014;106:2729–2736. 17. Sequeira V, Wijnker PJ, Nijenkamp LL, Kuster DW, NajafiA, Witjas-Paalberends
ER, Regan JA, Boontje N, Ten Cate FJ, Germans T, Carrier L, Sadayappan S, van
Slegtenhorst MA, Zaremba R, Foster DB, Murphy AM, Poggesi C, Dos
Remedios C, Stienen GJ, Ho CY, Michels M, van der Velden J. Perturbed length-
dependent activation in human hypertrophic cardiomyopathy with missense
sarcomeric gene mutations. Circ Res. 2013;112:1491–1505. 37. Helms AS, Alvarado FJ, Yob J, Tang VT, Pagani F, Russell MW, Valdivia HH, Day
SM. Genotype-dependent and -independent calcium signaling dysregulation in
human hypertrophic cardiomyopathy. Circulation. 2016;134:1738–1748. 38. Jimenez J, Tardiff JC. References Ferrantini C, Coppini R, Scellini B, Ferrara C, Pioner JM, Mazzoni L, Priori S,
Cerbai E, Tesi C, Poggesi C. R4496c RYR2 mutation impairs atrial and
ventricular contractility. J General Physiol. 2016;147:39–52. 8. Poggesi C, Ho CY. Muscle dysfunction in hypertrophic cardiomyopathy: what is
needed to move to translation? J Muscle Res Cell Motil. 2014;35:37–45. 9. Belus A, Piroddi N, Scellini B, Tesi C, Amati GD, Girolami F, Yacoub M, Cecchi
F, Olivotto I, Poggesi C. The familial hypertrophic cardiomyopathy-associated
myosin mutation R403Q accelerates tension generation and relaxation of
human cardiac myofibrils. J Physiol. 2008;586:3639–3644. 28. Chandra M, Rundell VL, Tardiff JC, Leinwand LA, De TOMBE PP, Solaro RJ. Ca(2+)
activation of myofilaments from transgenic mouse hearts expressing R92Q
mutant cardiac troponin T. Am J Physiol Heart Circ Physiol. 2001;280:H705–713. 10. Ferrantini C, Belus A, Piroddi N, Scellini B, Tesi C, Poggesi C. Mechanical and
energetic consequences of HCM-causing mutations. Cardiovasc Transl Res. 2009;2:441–451. 29. Belus A, Piroddi N, Ferrantini C, Tesi C, Cazorla O, Toniolo L, Drost M, Mearini
G, Carrier L, Rossi A, Mugelli A, Cerbai E, van der Velden J, Poggesi C. Effects
of chronic atrial fibrillation on active and passive force generation in human
atrial myofibrils. Circ Res. 2010;107:144–152. 11. Tardiff JC, Carrier L, Bers DM, Poggesi C, Ferrantini C, Coppini R, Maier LS,
Ashrafian H, Huke S, van der Velden J. Targets for therapy in sarcomeric
cardiomyopathies. Cardiovasc Res. 2015;105:457–470. 30. Coppini R, Mazzoni L, Ferrantini C, Gentile F, Pioner JM, Laurino T, Santini L,
Bargelli V, Rotellini M, Bartolucci G, Crocini C, Sacconi L, Tesi C, Belardinelli L,
Tardiff J, Mugelli A, Olivotto I, Cerbai E, Poggesi C. Ranolazine prevents
phenotype development in a mouse model of hypertrophic cardiomyopathy. Circ Heart Fail. 2017;10(3). pii: e003565. 12. Witjas-Paalberends ER, Ferrara C, Scellini B, Piroddi N, Montag J, Tesi C,
Stienen GJ, Michels M, Ho CY, Kraft T, Poggesi C, van der Velden J. Faster
cross-bridge detachment and increased tension cost in human hypertrophic
cardiomyopathy with the R403Q MYH7 mutation. J Physiol. 2014;592:3257–
3272. 31. Merx MW, Gorressen S, van de Sandt AM, Cortese-Krott MM, Ohlig J, Stern M,
Rassaf T, Godecke A, Gladwin MT, Kelm M. Depletion of circulating blood nos3
increases severity of myocardial infarction and left ventricular dysfunction. Basic Res Cardiol. 2014;109:398. 13. Javadpour MM, Tardiff JC, Pinz I, Ingwall JS. References 1. Semsarian C, Ingles J, Maron MS, Maron BJ. New perspectives on the
prevalence of hypertrophic cardiomyopathy. J Am Coll Cardiol. 2015;65:1249–
1254. 2. Ho CY, Charron P, Richard P, Girolami F, Van Spaendonck-Zwarts KY, Pinto Y. Genetic advances in sarcomeric cardiomyopathies: state of the art. Cardiovasc
Res. 2015;105:397–408. 2. Ho CY, Charron P, Richard P, Girolami F, Van Spaendonck-Zwarts KY, Pinto Y. Genetic advances in sarcomeric cardiomyopathies: state of the art. Cardiovasc
Res. 2015;105:397–408. 17 DOI: 10.1161/JAHA.116.005407 DOI: 10.1161/JAHA.116.005407 Journal of the American Heart Association Ferrantini et al Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al ORIGINAL RESEARCH Mugelli
A. Late
sodium
current
inhibition
reverses
electromechanical
dysfunction
in
human
hypertrophic
cardiomyopathy. Circulation. 2013;127:575–584. 3. Olivotto I, Cecchi F, Poggesi C, Yacoub MH. Patterns of disease progression in
hypertrophic cardiomyopathy: an individualized approach to clinical staging. Circ Heart Fail. 2012;5:535–546. 4. Olivotto I, Girolami F, Nistri S, Rossi A, Rega L, Garbini F, Grifoni C, Cecchi F,
Yacoub MH. The many faces of hypertrophic cardiomyopathy: from develop-
mental biology to clinical practice. Cardiovasc Transl Res. 2009;2:349–367. 23. Tardiff JC, Hewett TE, Palmer BM, Olsson C, Factor SM, Moore RL, Robbins J,
Leinwand LA. Cardiac troponin T mutations result in allele-specific phenotypes
in a mouse model for hypertrophic cardiomyopathy. J Clin Investig. 1999;104:469–481. 5. Olivotto I, d’Amati G, Basso C, Van Rossum A, Patten M, Emdin M, Pinto Y,
Tomberli B, Camici PG, Michels M. Defining phenotypes and disease
progression in sarcomeric cardiomyopathies: contemporary role of clinical
investigations. Cardiovasc Res 2015;105:409–423. 24. Moore RK, Abdullah S, Tardiff JC. Allosteric effects of cardiac troponin TNT1
mutations on actomyosin binding: a novel pathogenic mechanism for
hypertrophic cardiomyopathy. Arch Biochem Biophys. 2014;552–553:21–28. 6. Maron MS, Rowin EJ, Olivotto I, Casey SA, Arretini A, Tomberli B, Garberich RF,
Link MS, Chan RH, Lesser JR, Maron BJ. Contemporary natural history and
management of nonobstructive hypertrophic cardiomyopathy. J Am Coll
Cardiol. 2016;67:1399–1409. 25. Pistner A, Belmonte S, Coulthard T, Blaxall B. Murine echocardiography and
ultrasound imaging. J Vis Exp. 2010;42:pii:2100. 26. Ferrantini C, Coppini R, Sacconi L, Tosi B, Zhang ML, Wang GL, de Vries E,
Hoppenbrouwers E, Pavone F, Cerbai E, Tesi C, Poggesi C, ter Keurs HE. Impact of detubulation on force and kinetics of cardiac muscle contraction. J
General Physiol. 2014;143:783–797. 7. Maron BJ, Maron MS. Hypertrophic cardiomyopathy. Lancet. 2013;381:242–
255. 27. References Abnormal heart rate regulation in murine hearts with
familial hypertrophic cardiomyopathy-related cardiac troponin T mutations. Am
J Physiol Heart Circ Physiol. 2011;300:H627–635. 18. Crocini C, Ferrantini C, Scardigli M, Coppini R, Mazzoni L, Lazzeri E, Pioner JM,
Scellini B, Guo A, Song LS, Yan P, Loew LM, Tardiff J, Tesi C, Vanzi F, Cerbai E,
Pavone FS, Sacconi L, Poggesi C. Novel insights on the relationship between T-
tubular defects and contractile dysfunction in a mouse model of hypertrophic
cardiomyopathy. J Mol Cell Cardiol. 2016;91:42–51. 39. He H, Hoyer K, Tao H, Rice R, Jimenez J, Tardiff JC, Ingwall JS. Myosin-driven
rescue of contractile reserve and energetics in mouse hearts bearing familial
hypertrophic cardiomyopathy-associated mutant troponin T is mutation-
specific. J Physiol. 2012;590:5371–5388. 40. Gordon AM, Homsher E, Regnier M. Regulation of contraction in striated
muscle. Physiol Rev. 2000;80:853–924. 19. Lan F, Lee AS, Liang P, Sanchez-Freire V, Nguyen PK, Wang L, Han L, Yen M,
Wang Y, Sun N, Abilez OJ, Hu S, Ebert AD, Navarrete EG, Simmons CS, Wheeler
M, Pruitt B, Lewis R, Yamaguchi Y, Ashley EA, Bers DM, Robbins RC, Longaker
MT, Wu JC. Abnormal calcium handling properties underlie familial hyper-
trophic cardiomyopathy pathology in patient-specific induced pluripotent stem
cells. Cell Stem Cell. 2013;12:101–113. 41. Bai F, Weis A, Takeda AK, Chase PB, Kawai M. Enhanced active cross-bridges
during diastole: molecular pathogenesis of tropomyosin’s HCM mutations. Biophys J. 2011;100:1014–1023. 42. Janco M, Kalyva A, Scellini B, Piroddi N, Tesi C, Poggesi C, Geeves MA. Alpha-
tropomyosin with a D175N or E180G mutation in only one chain differs from
tropomyosin with mutations in both chains. Biochemistry. 2012;51:9880–
9890. 20. Haim TE, Dowell C, Diamanti T, Scheuer J, Tardiff JC. Independent fhc-related
cardiac troponin T mutations exhibit specific alterations in myocellular
contractility and calcium kinetics. J Mol Cell Cardiol. 2007;42:1098–1110. 21. Predmore JM, Wang P, Davis F, Bartolone S, Westfall MV, Dyke DB, Pagani F,
Powell SR, Day SM. Ubiquitin proteasome dysfunction in human hypertrophic
and dilated cardiomyopathies. Circulation. 2010;121:997–1004. 43. Ly S, Lehrer SS. Long-range effects of familial hypertrophic cardiomyopathy
mutations E180G and D175N on the properties of tropomyosin. Biochemistry. 2012;51:6413–6420. 22. Coppini R, Ferrantini C, Yao L, Fan P, Del Lungo M, Stillitano F, Sartiani L, Tosi
B, Suffredini S, Tesi C, Yacoub M, Olivotto I, Belardinelli L, Poggesi C, Cerbai E, 44. Harada K, Potter JD. References Familial hypertrophic cardiomyopathy mutations from
different functional regions of troponin T result in different effects on the pH 44. Harada K, Potter JD. Familial hypertrophic cardiomyopathy mutations from
different functional regions of troponin T result in different effects on the pH 18 Journal of the American Heart Association DOI: 10.1161/JAHA.116.005407 Ferrantini et al Mutation-Specific Pathomechanisms in Hypertrophic Cardiomyopathy
Ferrantini et al ORIGINAL RESEARCH for cardiac troponin T and alpha-tropomyosin in hypertrophic cardiomyopathy. New Engl J Med. 1995;332:1058–1064. and
Ca2+
sensitivity
of
cardiac
muscle
contraction. J
Biol
Chem. 2004;279:14488–14495. for cardiac troponin T and alpha-tropomyosin in hypertrophic cardiomyopathy. New Engl J Med. 1995;332:1058–1064. 56. Moolman JC, Corfield VA, Posen B, Ngumbela K, Seidman C, Brink PA, Watkins
H. Sudden death due to troponin T mutations. J Am Coll Cardiol. 1997;29:549–
555. 45. Manning EP, Guinto PJ, Tardiff JC. Correlation of molecular and functional
effects of mutations in cardiac troponin T linked to familial hypertrophic
cardiomyopathy: an integrative in silico/in vitro approach. J Biol Chem. 2012;287:14515–14523. 57. Koga Y, Toshima H, Kimura A, Harada H, Koyanagi T, Nishi H, Nakata M,
Imaizumi T. Clinical manifestations of hypertrophic cardiomyopathy with
mutations in the cardiac beta-myosin heavy chain gene or cardiac troponin T
gene. J Cardiac Fail. 1996;2:S97–103. 46. Manning EP, Tardiff JC, Schwartz SD. Molecular effects of familial hypertrophic
cardiomyopathy-related mutations in the TNT1 domain of cTnT. J Mol Biol. 2012;421:54–66. 47. Ertz-Berger BR, He H, Dowell C, Factor SM, Haim TE, Nunez S, Schwartz SD,
Ingwall JS, Tardiff JC. Changes in the chemical and dynamic properties of
cardiac troponin T cause discrete cardiomyopathies in transgenic mice. Proc
Natl Acad Sci USA. 2005;102:18219–18224. 58. Ho CY, Lakdawala NK, Cirino AL, Lipshultz SE, Sparks E, Abbasi SA, Kwong RY,
Antman EM, Semsarian C, Gonzalez A, Lopez B, Diez J, Orav EJ, Colan SD,
Seidman CE. Diltiazem treatment for pre-clinical hypertrophic cardiomyopathy
sarcomere mutation carriers: a pilot randomized trial to modify disease
expression. JACC Heart Fail. 2015;3:180–188. 48. Authors/Task Force Members; Elliott PM, Anastasakis A, Borger MA,
Borggrefe M, Cecchi F, Charron P, Hagege AA, Lafont A, Limongelli G,
Mahrholdt H, McKenna WJ, Mogensen J, Nihoyannopoulos P, Nistri S, Pieper
PG, Pieske B, Rapezzi C, Rutten FH, Tillmanns C, Watkins H. 2014 ESC
guidelines on diagnosis and management of hypertrophic cardiomyopathy: the
task force for the diagnosis and management of hypertrophic cardiomyopathy
of the European Society of Cardiology (ESC). Eur Heart J. 2014;35:2733–2779. 59. References Patel R, Nagueh SF, Tsybouleva N, Abdellatif M, Lutucuta S, Kopelen HA,
Quinones MA, Zoghbi WA, Entman ML, Roberts R, Marian AJ. Simvastatin
induces regression of cardiac hypertrophy and fibrosis and improves cardiac
function in a transgenic rabbit model of human hypertrophic cardiomyopathy. Circulation. 2001;104:317–324. 60. Lim DS, Lutucuta S, Bachireddy P, Youker K, Evans A, Entman M, Roberts R,
Marian AJ. Angiotensin II blockade reverses myocardial fibrosis in a transgenic
mouse
model
of
human
hypertrophic
cardiomyopathy. Circulation. 2001;103:789–791. 49. Kranias EG, Hajjar RJ. Modulation of cardiac contractility by the phospholam-
ban/SERCA2a regulatome. Circ Res. 2012;110:1646–1660. 50. Lyon AR, MacLeod KT, Zhang Y, Garcia E, Kanda GK, Lab MJ, Korchev YE,
Harding SE, Gorelik J. Loss of T-tubules and other changes to surface
topography in ventricular myocytes from failing human and rat heart. Proc Natl
Acad Sci USA. 2009;106:6854–6859. 61. Nagueh SF, Lombardi R, Tan Y, Wang J, Willerson JT, Marian AJ. Atorvastatin
and cardiac hypertrophy and function in hypertrophic cardiomyopathy: a pilot
study. Eur J Clin Invest. 2010;40:976–983. by guest on September 11, 2017
http://jaha.ahajournals.org/
Downloaded from 51. Wei S, Guo A, Chen B, Kutschke W, Xie YP, Zimmerman K, Weiss RM, Anderson
ME, Cheng H, Song LS. T-tubule remodeling during transition from hypertrophy
to heart failure. Circ Res. 2010;107:520–531. 62. Axelsson A, Iversen K, Vejlstrup N, Ho C, Norsk J, Langhoff L, Ahtarovski K,
Corell P, Havndrup O, Jensen M, Bundgaard H. Efficacy and safety of the
angiotensin ii receptor blocker losartan for hypertrophic cardiomyopathy: the
inherit randomised, double-blind, placebo-controlled trial. Lancet Diabetes
Endocrinol. 2015;3:123–131. 52. Swaminathan PD, Purohit A, Hund TJ, Anderson ME. Calmodulin-dependent
protein
kinase
II:
linking
heart
failure
and
arrhythmias. Circ
Res. 2012;110:1661–1677. 63. Green EM, Wakimoto H, Anderson RL, Evanchik MJ, Gorham JM, Harrison BC,
Henze M, Kawas R, Oslob JD, Rodriguez HM, Song Y, Wan W, Leinwand LA,
Spudich JA, McDowell RS, Seidman JG, Seidman CE. A small-molecule inhibitor
of sarcomere contractility suppresses hypertrophic cardiomyopathy in mice. Science. 2016;351:617–621. 53. Kreusser MM, Backs J. Integrated mechanisms of camkii-dependent ventric-
ular remodeling. Front Pharmacol. 2014;5:36. 54. Chen W, Wang R, Chen B, Zhong X, Kong H, Bai Y, Zhou Q, Xie C, Zhang J, Guo
A, Tian X, Jones PP, O’Mara ML, Liu Y, Mi T, Zhang L, Bolstad J, Semeniuk L,
Cheng H, Zhang J, Chen J, Tieleman DP, Gillis AM, Duff HJ, Fill M, Song LS,
Chen SR. References The ryanodine receptor store-sensing gate controls Ca2+ waves and
Ca2+-triggered arrhythmias. Nat Med. 2014;20:184–192. 64. Pioner JM, Racca AW, Klaiman JM, Yang KC, Guan X, Pabon L, Muskheli
V, Zaunbrecher R, Macadangdang J, Jeong MY, Mack DL, Childers MK,
Kim DH, Tesi C, Poggesi C, Murry CE, Regnier M. Isolation and
mechanical
measurements
of
myofibrils
from
human
induced
pluripotent
stem
cell-derived
cardiomyocytes. Stem
Cell
Reports. 2016;6:885–896. 55. Watkins H, McKenna WJ, Thierfelder L, Suk HJ, Anan R, O’Donoghue A, Spirito
P, Matsumori A, Moravec CS, Seidman JG, Seidman CE. Mutations in the genes 19 DOI: 10.1161/JAHA.116.005407 Journal of the American Heart Association for more information.
http://jaha.ahajournals.org
Access publication. Visit the Journal at
is an online only Open
Journal of the American Heart Association
Subscriptions, Permissions, and Reprints: The Olivotto, Elisabetta Cerbai, Chiara Tesi and Corrado Poggesi
Spinelli, Leonardo Sacconi, Pieter De Tombe, Rachel Moore, Jil Tardiff, Alessandro Mugelli, Iacopo
Mazzoni, Beatrice Scellini, Nicoletta Piroddi, Annunziatina Laurino, Lorenzo Santini, Valentina
Cecilia Ferrantini, Raffaele Coppini, Josè Manuel Pioner, Francesca Gentile, Benedetta Tosi, Luca
Comparison of the Cardiac Troponin T E163R and R92Q Mouse Models
Pathogenesis of Hypertrophic Cardiomyopathy is Mutation Rather Than Disease Specific: A Olivotto, Elisabetta Cerbai, Chiara Tesi and Corrado Poggesi
Spinelli, Leonardo Sacconi, Pieter De Tombe, Rachel Moore, Jil Tardiff, Alessandro Mugelli, Iacopo
Mazzoni, Beatrice Scellini, Nicoletta Piroddi, Annunziatina Laurino, Lorenzo Santini, Valentina
Cecilia Ferrantini, Raffaele Coppini, Josè Manuel Pioner, Francesca Gentile, Benedetta Tosi, Luca
Comparison of the Cardiac Troponin T E163R and R92Q Mouse Models
Pathogenesis of Hypertrophic Cardiomyopathy is Mutation Rather Than Disease Specific: A http://jaha.ahajournals.org/content/6/7/e005407
World Wide Web at:
The online version of this article, along with updated information and services, is located on the http://jaha.ahajournals.org/content/6/7/e005407
World Wide Web at:
The online version of this article, along with updated information and services, is located on the for more information. http://jaha.ahajournals.org
Access publication. Visit the Journal at
is an online only Open
Journal of the American Heart Association
Subscriptions, Permissions, and Reprints: The
|
https://openalex.org/W3205037902
|
https://www.nature.com/articles/s41534-021-00487-y.pdf
|
English
| null |
Symmetry-resolved entanglement detection using partial transpose moments
|
npj quantum information
| 2,021
|
cc-by
| 16,253
|
INTRODUCTION This powerful entanglement condition, which found many
applications in theoretical works15–22, is difficult to apply in
experimental conditions as the PT spectrum is difficult to access. To overcome this challenge, it was shown in ref. 23 that valuable
information about the PT spectrum can be obtained from a few PT
moments trðρΓÞk only. Using the first three PT moments, an
entanglement condition, called p3-PPT, was proposed and shown
to be useful for detecting entangled states in several different
contexts. Moments trðρΓÞk have the advantage that they can be
estimated using shadow tomography23 in a more efficient way
than if one had to reconstruct ρ via full quantum state
tomography. In the past years, a considerable effort led to the building of larger
and larger Noisy Intermediate-Scale Quantum (NISQ) devices1–3. For the benchmarking of such devices comes the need for more
scalable tools in order to characterize the underlying many-body
quantum state (see e.g.4 and references therein). For instance,
characterizing the entanglement properties of these quantum
states is, besides the intrinsic theoretical interest, essential to
gauge the performance and verify the proper working of the NISQ
devices. As a first prominent example among the tools to characterize
entanglement, there is the concept of entanglement witness5,6. An
entanglement witness is a functional of the quantum density
matrix that separates a specific entangled state from the set of all
separable states (for examples of entanglement witnesses in
various types of systems, see e.g. ref. 7 and references therein). When this function is linear, it can be identified with an observable
whose expectation value can be used to decide whether the
target state is entangled or not. By contrast, in this work, we shall
focus on a superset of the set of separable states: the set of states
with positive partial transpose. In other words, we will focus on
sufficient conditions for entanglement (equivalently, necessary
conditions for separability). g
p y
Indeed, the k-th order PT moment of a state ρ can be expressed
as the expectation value of some k-copy permutation operator O23,
i.e. trðρΓÞ
k ¼ tr Oρk
. As shown in ref. 24, the classical shadow
formalism allows to estimate such an expectation value from
independent classical snapshots ^ρ1; ¼ ; ^ρk (which can each be
obtained from post-processing single-qubit measurement out-
comes) through the U-statistics estimator ^ok ¼ tr O^ρ1 ^ρk
ð
Þ. 1Institute for Theoretical Physics, University of Innsbruck, Innsbruck, Austria. 2The Abdus Salam International Center for Theoretical Physics, Trieste, Italy. 3SISSA, Trieste, Italy.
4Center for Quantum Physics, University of Innsbruck, Innsbruck, Austria. 5Institute for Quantum Optics and Quantum Information of the Austrian Academy of Sciences,
Innsbruck, Austria. 6INFN, Trieste, Italy. 7Univ. Grenoble Alpes, CNRS, LPMMC, Grenoble, France. 8Institute for Integrated Circuits, Johannes Kepler University Linz, Linz, Austria.
✉email: antoine.neven@uibk.ac.at; jose.carrasco@uibk.ac.at ARTICLE
OPEN
Symmetry-resolved entanglement detection using partial
transpose moments Antoine Neven
1✉, Jose Carrasco
1✉, Vittorio Vitale
2,3, Christian Kokail4,5, Andreas Elben
4,5, Marcello Dalmonte2,3,
Pasquale Calabrese2,3,6, Peter Zoller
4,5, Benoȋt Vermersch
4,5,7, Richard Kueng
8 and Barbara Kraus
1 Antoine Neven
1✉, Jose Carrasco
1✉, Vittorio Vitale
2,3, Christian Kokail4,5, Andreas Elben
4,5, Marcello Dalmonte2,3,
Pasquale Calabrese2,3,6, Peter Zoller
4,5, Benoȋt Vermersch
4,5,7, Richard Kueng
8 and Barbara Kraus
1 We propose an ordered set of experimentally accessible conditions for detecting entanglement in mixed states. The k-th condition
involves comparing moments of the partially transposed density operator up to order k. Remarkably, the union of all moment
inequalities reproduces the Peres-Horodecki criterion for detecting entanglement. Our empirical studies highlight that the first four
conditions already detect mixed state entanglement reliably in a variety of quantum architectures. Exploiting symmetries can help
to further improve their detection capabilities. We also show how to estimate moment inequalities based on local random
measurements of single state copies (classical shadows) and derive statistically sound confidence intervals as a function of the
number of performed measurements. Our analysis includes the experimentally relevant situation of drifting sources, i.e. non-
identical, but independent, state copies. npj Quantum Information (2021) 7:152 ; https://doi.org/10.1038/s41534-021-00487-y www.nature.com/npjqi Published in partnership with The University of New South Wales INTRODUCTION Therefore, as in other randomized measurements protocols
probing entanglement23–33, the classical shadows formalism only
requires (randomized) single-qubit measurements in experiments
realizing the single-copy state ρ. In this paper, we follow this idea
of using PT moments to build experimentally computable
entanglement conditions, and extend the p3-PPT condition in
two directions. From
the
numerous
theoretical
sufficient
conditions
for
entanglement that have been developed in the literature, many
cannot be straightforwardly implemented experimentally, mainly
because they require the (exponentially expensive) knowledge of
the full density matrix8–10. This is for instance the case of the
celebrated PPT condition5,11, which states that a separable state ρ
always has a positive semi-definite (PSD) partial transpose (PT) ρΓ
for any bipartite splitting of its subsystems. Thus, if ρΓ has (at least)
a single negative eigenvalue, then ρ is entangled. The negativity,
which resulted from this condition, is a highly used entanglement
measure for mixed states12–14. On the one hand, we propose different entanglement detection
strategies depending on how many PT moments can be
estimated. Starting from the third-order moment, we show that
the estimation of each higher-order moment gives access to an
independent entanglement condition. Interestingly, if all the PT
moments can be estimated, this set of conditions is then
necessary and sufficient for the state to be PPT (i.e. to have a
positive semi-definite partial transpose). Of course, the higher the Published in partnership with The University of New South Wales A. Neven et al. 2 The p3-PPT condition will serve as a reference point below in
accessing the predictive power of the proposed relations. moment, the larger the number of experimental runs needed. In
case of higher-order moments cannot be accessed, we show how
to obtain an optimal entanglement condition using PT moments
of order up to three. Throughout this paper, we will establish and analyse different
sets of necessary (and sometimes also sufficient) PPT conditions
based on PT moments. An exhaustive list summarizing these
conditions is given below, together with two important results
regarding the experimental estimation of the PPT conditions using
classical shadows. p
On the other hand, we investigate the effect of symmetries on
this entanglement detection method. As shown in ref. Definitions and summary of results and similar conditions are obtained including higher moments; In this section, we introduce the basic definitions needed for the
entanglement detection criteria below, and summarize in a
succinct manner our main results. Given a bipartite state ρ =
ρAB, we denote by ρΓ its partial transpose with respect to
subsystem B. We say that ρ is PPT if ρΓ is positive semi-definite,
and NPT otherwise. All NPT states are entangled, however there
are entangled states, known as bound-entangled states, that are
not NPT. We focus here on the detection of NPT entangled states. iii) in case high-order moments are difficult or too expensive to
access, we also show how to obtain an optimized, necessary
condition for PPT using only PT moments of order up to three. We
call this condition Dopt
3 ; 3
iv) all of the above conditions can be cast in terms of ρΓ
ðqÞ, in
which case we add the prefix SR (for symmetry-resolved). For
instance, the SR-p3-PPT condition for sector q reads g
We denote the k-th order moment of a matrix M by p3ðρΓ
ðqÞÞp1ðρΓ
ðqÞÞ ⩾ðp2ðρΓ
ðqÞÞÞ
2:
(5) (5) pkðMÞ tr Mk:
(1) (1) pkðMÞ tr Mk: Since these conditions are sensitive to the presence of negative
eigenvalues in a specific symmetry sector, they are typically much
more sensitive than their non-SR counterparts, as illustrated in Fig. 1. We will mostly consider moments of ρΓ, and sometimes use the
short-hand notation pk ≡pk(ρΓ). In the presence of symmetries, the
partial transpose can be cast in block diagonal form: we denote as
ρΓ
ðqÞ the resulting blocks, where q indicates a quantum number,
and define the corresponding moments pkðρΓ
ðqÞÞ as from Eq. (1). p
g
In the SR case, it is worth mentioning that also the SR-D2
condition, p2ðρΓ
ðqÞÞ ⩽ðp1ðρΓ
ðqÞÞÞ
2 ;
(6) (6) p
g
pkðρðqÞÞ
q ( )
We start by recalling the p3-PPT condition of ref. 23, i.e. that any
PPT state satisfies is non-trivial; is non-trivial; is non-trivial; p3ðρΓÞp1ðρΓÞ ⩾ðp2ðρΓÞÞ
2 :
(2) (2) v) we show how SR conditions can, in fact, be applied to
arbitrary states, via the application of a proper transformation on
the density matrix of interest. In practice, this transformation is v) we show how SR conditions can, in fact, be applied to
arbitrary states, via the application of a proper transformation on
the density matrix of interest. INTRODUCTION 34 for the
case of dynamical purification, taking symmetries into account to
define symmetry-resolved (SR) versions of the tools usually used
to characterize quantum states can enable a finer characterization
of some quantum features and even reveal phenomena that
cannot be observed without symmetry-resolution. For states
preserving an extensive quantity, we define SR versions of the
PT-moment inequalities mentioned previously and show that
these are indeed better suited to detect the entanglement of such
states. Furthermore, we also show that these SR inequalities
provide a sufficient entanglement condition for states that do not
possess any symmetry. i) the first set of conditions, which we dub Dn conditions, also
contains polynomial inequalities in the moments pk of order up to
k ⩽n. The first non-trivial such a condition is D3, and reads: p3ðρΓÞ ⩾ 1
2 ðp1ðρΓÞÞ
3 þ 3
2 p1ðρΓÞp2ðρΓÞ:
(3) (3) Knowing only the first three moments p1(ρΓ), p2(ρΓ) and p3(ρΓ), this
condition is optimal for detecting entanglement if 1/2 ⩽p2(ρΓ) ⩽1. Knowing moments of order up to the dimension of ρΓ, the set of
Dn conditions provides a necessary and sufficient condition for
NPT entanglement; The conditions derived here are particularly interesting from an
experimental (and numerical) point of view, as low moments of
(partially transposed) density operators are accessible. We show
how source drifts in an experiment can be taken into account and
how the quantities which are of interest here can be accurately
estimated via local measurements on single copies of the state. g
ii) the second set of conditions, dubbed Stieltjesn, involves
inequalities among the moments pk of order up to n. The
condition Stieltjes3 is equivalent to p3-PPT, while Stieltjes5 reads: det
p1
p2
p3
p2
p3
p4
p3
p4
p5
0
B
@
1
C
A ⩾0
(4) (4) npj Quantum Information (2021) 152 Entanglement detection from partial transpose moments for i = 1, …, d, and e0(x1, …, xd) = 1, we derive in the Methods
section the following lemma. Lemma 1. A Hermitian matrix A of dimension d is PSD if and only if
ei(λ1, …, λd) ⩾0 for all i = 1, …, d, where λ1, …, λd are the eigenvalues
of A, and ei denote the elementary symmetric polynomials (Eq. (7)). Using Newton’s identities, which relate the elementary sym-
metric polynomials, ek, in the eigenvalues of A to the moments of
A through the recursive formula Fig. 2
Entanglement conditions of order three. The plot of the
value of the third moment p3 saturating the p3-PPT (dashed orange
curve), the D3 (dashed green line), and the optimal Dopt
3
(thick black
curve) conditions as a function of the second moment p2 for a
normalized Hermitian matrix. According to the p3-PPT condition, any
state ρ with a value of p3(ρΓ) below the dashed orange curve is
entangled. Similarly, the condition D3 shows that any state ρ with a
value of p3(ρΓ) below the dashed green line is entangled. From this
plot, it is clear that, for p2(ρΓ) > 1/2, all entangled states detected by
the p3-PPT condition are also detected by D3, which coincides with
Dopt
3
in this case. When p2(ρΓ) < 1/2, the p3-PPT condition is then
stronger than D3, and Dopt
3
represents a slight improvement over the
p3-PPT condition. As illustrated in the applications, this slight
improvement can nevertheless be important for the detection of
physically relevant states. k ek ¼
X
k
i¼1
ð1Þi1eki piðAÞ;
(8) (8) each inequality ei ⩾0 can be transformed into an inequality
involving moments of A of order up to i. We denote by Di these
moments inequalities. As an illustration, the conditions D1 to D4
read p1ðAÞ ⩾0;
(9) p1ðAÞ ⩾0;
(9)
p2ðAÞ ⩽ðp1ðAÞÞ2;
(10)
p3ðAÞ ⩾ 1
2 ðp1ðAÞÞ3 þ 3
2 p1ðAÞp2ðAÞ;
(11)
p4ðAÞ ⩽1
2 ðp1ðAÞÞ2 p2ðAÞ
2
1
3 ðp1ðAÞÞ4 þ 4
3 p1ðAÞp3ðAÞ;
(12) p1ðAÞ ⩾0; (9) p3ðAÞ ⩾ 1
2 ðp1ðAÞÞ3 þ 3
2 p1ðAÞp2ðAÞ;
(11)
p4ðAÞ ⩽1
2 ðp1ðAÞÞ2 p2ðAÞ
2
1
3 ðp1ðAÞÞ4 þ 4
3 p1ðAÞp3ðAÞ; (11) Fig. 2
Entanglement conditions of order three. Entanglement detection from partial transpose moments The plot of the
value of the third moment p3 saturating the p3-PPT (dashed orange
curve), the D3 (dashed green line), and the optimal Dopt
3
(thick black
curve) conditions as a function of the second moment p2 for a
normalized Hermitian matrix. According to the p3-PPT condition, any
state ρ with a value of p3(ρΓ) below the dashed orange curve is
entangled. Similarly, the condition D3 shows that any state ρ with a
value of p3(ρΓ) below the dashed green line is entangled. From this
plot, it is clear that, for p2(ρΓ) > 1/2, all entangled states detected by
the p3-PPT condition are also detected by D3, which coincides with
Dopt
3
in this case. When p2(ρΓ) < 1/2, the p3-PPT condition is then
stronger than D3, and Dopt
3
represents a slight improvement over the
p3-PPT condition. As illustrated in the applications, this slight
improvement can nevertheless be important for the detection of
physically relevant states. Fig. 2
Entanglement conditions of order three. The plot of the
value of the third moment p3 saturating the p3-PPT (dashed orange g
a g
o d
o
o o d
. e p ot o t e
value of the third moment p3 saturating the p3-PPT (dashed orange
curve), the D3 (dashed green line), and the optimal Dopt
3
(thick black
curve) conditions as a function of the second moment p2 for a
normalized Hermitian matrix. According to the p3-PPT condition, any
state ρ with a value of p3(ρΓ) below the dashed orange curve is
entangled. Similarly, the condition D3 shows that any state ρ with a
value of p3(ρΓ) below the dashed green line is entangled. From this
plot, it is clear that, for p2(ρΓ) > 1/2, all entangled states detected by
the p3-PPT condition are also detected by D3, which coincides with
Dopt
3
in this case. When p2(ρΓ) < 1/2, the p3-PPT condition is then
stronger than D3, and Dopt
3
represents a slight improvement over the
p3-PPT condition. As illustrated in the applications, this slight
improvement can nevertheless be important for the detection of
physically relevant states. (12) respectively. One has p1(ρΓ) = 1 for any quantum state ρ, implying
that D1 is trivially satisfied. Similarly, since p2(ρΓ) is equal to p2(ρ)
(i.e., to the purity of ρ) for any quantum state ρ, the inequality D2 is
also trivially satisfied. A. Neven et al. 3 effectively carried out in the post-processing step of the classical
shadows; so that the inequality D2 is not trivially satisfied and can already
reveal the presence of entanglement. p
g
When applied to ρΓ, Lemma 1 and Newton’s identities (8) can
thus be used to detect NPT entangled states from PT moments
only. From an experimental point of view, this is an important
aspect of this entanglement detection scheme, as PT moments are
experimentally more affordable to estimate than, for instance, the
whole spectrum of ρΓ. As PT moments are more expensive to be
estimated the higher the order, these inequalities should be
considered starting from those involving the lowest moment
orders. Even though showing that a state is NPT with this method
can in principle require the knowledge of all the PT moments, we
will provide many experimentally relevant instances where
entanglement
can
be
effectively
detected
from
low-order
moments even in the presence of errors. To this end, we provide
confidence intervals for the quantities of interest granting that a
certain inequality is violated with high probability (see Theorem 1
and the Supplementary Information). vi) as illustrated in Fig. 1, the uncertainty in estimating the
moments can be bounded, in principle, using the classical
shadows formalism. Here, we show how to combine those
bounds to provide rigorous confidence intervals for SR-D2, which
considerably strengthens the impact of our results in real
experiments. Entanglement detection from partial transpose moments In this section, we present entanglement conditions based on PT
moments. We extend the idea behind the p3-PPT condition (c.f. Eq. (2)) of ref. 23 in two directions. On the one hand, we present a set
of inequalities involving all the PT moments which provides a
necessary and sufficient condition for the underlying state to be
PPT. In addition, each condition of this set is itself a necessary PPT
condition. On the other hand, we show how to optimize such
entanglement conditions when only a few low-order PT moments
are accessible. Similarly, let us mention here that necessary and sufficient
conditions for a matrix to be PSD can be expressed as different
sets of polynomial inequalities in its moments. One of such sets
can be deduced from the well-known (truncated) Stieltjes
moment problem (see Methods). In the part dedicated to
applications, we illustrate the usefulness of these inequalities by
applying them to the entanglement detection of the ground state
of the XXZ model (c.f. Fig. 5). Let us finally also mention that, from
a few moments of a Hermitian matrix, one can also obtain bounds
on the distance between this matrix and the PSD cone35. The idea behind this set of conditions is to use Descartes’ rule of
signs on the characteristic polynomial of a Hermitian matrix to
obtain a set of moment inequalities that has to be satisfied by any
PSD matrix. Applied to the partially transposed matrix ρΓ, such
conditions can then be used to detect the entanglement of NPT
quantum states. More precisely, using the definition of the
elementary symmetric polynomials on d variables, eiðx1; ¼ ; xdÞ ¼
X
1 ⩽j1<<ji ⩽d
xj1 xji;
(7) (7) Optimized condition for low-order moments. Due to (experimen-
tal) constraints, it might not be possible to determine all, but only
a few PT moments. This is why, we show here how to optimize
necessary PPT conditions using only PT moments of order up to
three. From the previous sections, we already have two examples
of such conditions, namely the p3-PPT and D3 conditions. As
illustrated in Fig. 2, the p3-PPT (D3) condition is tighter than D3 (p3- for i = 1, …, d, and e0(x1, …, xd) = 1, we derive in the Methods
section the following lemma. Published in partnership with The University of New South Wales Published in partnership with The University of New South Wales Definitions and summary of results In practice, this transformation is Any state violating this condition is NPT and therefore entangled. Any state violating this condition is NPT and therefore entangled. Fig. 1
An illustration of the proposed method for entanglement detection. We assume the experimentally relevant situation of a source
producing non-identical but independent copies {ρ1, …, ρN} ("drift''). Randomly chosen unitaries Ui are applied to the qubits of each copy and
then measured in the standard basis. Using classical shadows24, these measurement outcomes are post-processed to obtain the moments
pj ¼ trðρΓ
avgÞj. As explained in the main text, we combine those moments to derive inequalities whose violation implies that the state ρavg is
NPT, thus showing that at least one of the states ρk produced by the source is entangled. We also show how symmetry-resolution techniques
can be used to enhance entanglement detection capabilities. npj Quantum Information (2021) 152
Published in partnership with The University of New South Wales Fig. 1
An illustration of the proposed method for entanglement detection. We assume the experimentally relevant situation of a source
producing non-identical but independent copies {ρ1, …, ρN} ("drift''). Randomly chosen unitaries Ui are applied to the qubits of each copy and
then measured in the standard basis. Using classical shadows24, these measurement outcomes are post-processed to obtain the moments
pj ¼ trðρΓ
avgÞj. As explained in the main text, we combine those moments to derive inequalities whose violation implies that the state ρavg is
NPT, thus showing that at least one of the states ρk produced by the source is entangled. We also show how symmetry-resolution techniques
can be used to enhance entanglement detection capabilities. Published in partnership with The University of New South Wales Entanglement detection from partial transpose moments One can see that
ðN A N BÞ ab0
j
i ¼ ðna nb0Þ ab0
j
i
and
ðN A N BÞ a0b
j
i ¼
ðna0 nbÞ a0b
j
i with na nb0 ¼ na0 nb. This shows that matrix
elements within a block of ρ are mapped, via partial transposition,
to matrix elements within a block of ρΓ. Given a d × d PSD matrix A, with non-zero eigenvalues λ1, …, λr,
for some r ∈[1, d], this optimization can be performed with the
help of Lagrange multipliers. As shown in the Methods section,
this leads to solutions with only two distinct eigenvalues λa, λb
with multiplicity ra, r −ra, respectively, for ra ∈[1, r]. Assuming,
without loss of generality, that λa ⩾λb, the optimization of p3 leads
then to ra = r −1. For each value of r, the optimal value of p3 can
be easily determined in the interval [1/r, 1/(r −1)]. For r = 2 this
leads to D3 whereas for r > 2 one obtains an optimal value of p3
which is slightly better than p3-PPT. Observe that pmin
3
ðp2Þ is a
piece-wise function and the derivative ∂pmin
3
=∂p2 is discontinuous
at points p2 = 1/r (see Fig. 2). ρ
The size of the sector corresponding to the eigenvalue q in the
block-decomposition of ρΓ is given by tr Pq ¼
X
ab¼q
n
a
m
b
¼
n þ m
q þ m
:
(17) In case higher-order PT moments can be accessed experimen-
tally, one may consider optimizing higher-order PPT conditions. In
principle, the Lagrange multipliers method we used to optimize
the entanglement condition using PT moments of order at most
three can be straightforwardly extended to optimize higher-order
PPT conditions. In practice, however, the number of variables
would grow with the moment order, making the analytical
resolution of the optimization problem more involved. In the
simultaneous but independent work36, the authors also consider
this optimization problem and explicitly carry out the optimization
of the PPT condition using PT moments of order up to four
(though with a slightly different method). In addition, they provide
a numerical code to perform the optimization with PT moments of
order up to five. Published in partnership with The University of New South Wales Entanglement detection from partial transpose moments (17) When the partial transpose of a density matrix has a block
structure, it is naturally PSD iff each block is itself a PSD matrix. Therefore, one can apply the conditions of the previous section
directly to the blocks ρΓ
ðqÞ of the partial transpose. For the p3-PPT
condition, the corresponding symmetry-resolved (SR) inequalities
are simply p3ðρΓ
ðqÞÞp1ðρΓ
ðqÞÞ ⩾ðp2ðρΓ
ðqÞÞÞ
2
(18) (18) for all q = −m, −m + 1, …, n. Any violation of a PSD condition in
one of the blocks is then sufficient to show that ρΓ has at least one
negative eigenvalue and that ρ is therefore entangled. When using the Di conditions, symmetry-resolution can be a
significant advantage (see e.g. the applications section). First, the
necessary and sufficient PSD conditions involve moments of order
at most equal to the dimension of the largest block, that is
Entanglement detection from partial transpose moments Therefore, when ρ is a quantum state, the
first non-trivial inequality for ρΓ is D3. As will be shown in the next
section, it is sometimes more efficient (in order to detect
entanglement) to apply these inequalities to projections of ρΓ
onto specific subspaces, rather than to ρΓ itself. We would like to
stress here that, in that case, the argument above does not hold, npj Quantum Information (2021) 152 A. Neven et al. 4 …, n + m} of the operator N and has support in the correspond-
ing eigenspace. Here, …, n + m} of the operator N and has support in the correspond-
ing eigenspace. Here, PPT) for states with purity larger (smaller) than 1/2, respectively. As
the low-order moments are easier to access experimentally, we
now address the question about the optimal inequality involving
PT moments of order up to three. PPT) for states with purity larger (smaller) than 1/2, respectively. As
the low-order moments are easier to access experimentally, we
now address the question about the optimal inequality involving
PT moments of order up to three. Qq ¼
X
aþb¼q
ΠaðAÞ ΠbðBÞ;
(14) (14) p
To answer this question, we use the following approach. For
fixed values p1 and p2 of the first two moments, we determine the
minimal value pmin
3
that the third moment can reach for any PSD
matrix. From this bound, we know that any Hermitian matrix with
a smaller third moment is necessarily not PSD. Note that we want
here to minimize p3 because it is an odd moment (for which
negative eigenvalues would have the tendency to decrease the
value of the moment). For an even moment, we would instead
maximize the value of this moment over PSD matrices. This is also
reflected in the Dn conditions (9)-(12), where the inequality sign
alternates between even and odd values of n. Naturally, we restrict
ourselves to values of p1 and p2 which are compatible with a PSD
matrix, and therefore satisfy Eqs. (9) and (10). Recall that for the
partial transpose of a density operator, this is always fulfilled. with ΠkðAÞ ¼
X
i1þþin¼k
i1 in
j
i i1 in
h
j ;
(15) (15) and similarly for B. It has been shown48 that, for this type of
symmetry, the partial transpose ρΓ is also block diagonal, but in a
different basis. Entanglement detection from partial transpose moments In fact, ρΓ ¼ n
q¼mρΓ
ðqÞ ¼ P
q PqρΓPq, where Pq is
the projector onto the eigenspace of N A N B with eigenvalue
q ∈{ −m, −m + 1, …, n}, i.e. and similarly for B. It has been shown48 that, for this type of
symmetry, the partial transpose ρΓ is also block diagonal, but in a
different basis. In fact, ρΓ ¼ n
q¼mρΓ
ðqÞ ¼ P
q PqρΓPq, where Pq is
the projector onto the eigenspace of N A N B with eigenvalue
q ∈{ −m, −m + 1, …, n}, i.e. Pq ¼
X
ab¼q
ΠaðAÞ ΠbðBÞ:
(16) (16) This can be easily seen as follows. Consider a matrix element
ρab;a0b0 ab
j
i a0b0
h
j of ρ with eigenvalue i of N A þ N B. Precisely, let
us write N A aj i ¼ na aj i, N A a0
j i ¼ na0 a0
j i, N B bj i ¼ nb bj i, and
N B b0
j i ¼ nb0 b0
j i with na þ nb ¼ na0 þ nb0 ¼ i. After partial trans-
position, ρab;a0b0 ab
j
i a0b0
h
j7!ρab;a0b0 ab0
j
i a0b
h
j. One can see that
ðN A N BÞ ab0
j
i ¼ ðna nb0Þ ab0
j
i
and
ðN A N BÞ a0b
j
i ¼
ðna0 nbÞ a0b
j
i with na nb0 ¼ na0 nb. This shows that matrix
elements within a block of ρ are mapped, via partial transposition,
to matrix elements within a block of ρΓ. This can be easily seen as follows. Consider a matrix element
ρab;a0b0 ab
j
i a0b0
h
j of ρ with eigenvalue i of N A þ N B. Precisely, let
us write N A aj i ¼ na aj i, N A a0
j i ¼ na0 a0
j i, N B bj i ¼ nb bj i, and
N B b0
j i ¼ nb0 b0
j i with na þ nb ¼ na0 þ nb0 ¼ i. After partial trans-
position, ρab;a0b0 ab
j
i a0b0
h
j7!ρab;a0b0 ab0
j
i a0b
h
j. Symmetry-resolved entanglement detection Symmetries, as they often occur in physical situations, can be
exploited to observe relevant phenomena (see e.g. refs. 34,37–47). H
i
h
d
i
f
l Symmetries, as they often occur in physical situations, can be
exploited to observe relevant phenomena (see e.g. refs. 34,37–47). Here, we use symmetries to ease the detection of entanglement. More precisely, we apply the previously developed tools to
symmetric states, which will lead to conditions of entanglement
involving much lower moments of the partial transpose projected
onto certain subspaces. Despite the fact that these quantities
differ significantly from the moments of ρΓ, we will show later on
that they can nevertheless be estimated using the framework of
classical shadows. n þ m
bðn þ mÞ=2c
, which is necessarily smaller than the dimension
Γ
Γ n þ m
bðn þ mÞ=2c
, which is necessarily smaller than the dimension
Γ
Γ of the density matrix itself. Second, since a block ρΓ
ðqÞ of ρΓ is (in
general) not the partial transpose of any density matrix (i.e. there
could be no σ > 0 such that ρΓ
ðqÞ ¼ σΓ), the inequality: p2ðρΓ
ðqÞÞ ⩽ðp1ðρΓ
ðqÞÞÞ
2
(19) (19) We consider a bipartite state ρ = ρAB, with subsystems A and B
containing n and m qubits, respectively. We assume that this state
commutes with Pnþm
i¼1 Zi, or similarly with the total number
operator N ¼ N A þ N B. Here and in the following, we denote by
X, Y, Z the Pauli operators. Obviously, such a state has a block
diagonal form, i.e., is not necessarily satisfied. This implies that moments of order two
can already be sufficient to detect entanglement. y
g
As stressed in the introduction, using PT-moment inequalities to
detect entanglement is particularly interesting from an experi-
mental point of view, because such PT moments can be
estimated, for instance using shadow tomography23. As we show
in the following lemma, the shadow tomography protocol used in
ref. 23 can also be used to estimate moments of blocks of the
partial transpose (which differ significantly from the PT moments)
by slightly modifying the non-linear observable that has to be
measured. npj Quantum Information (2021) 152 Symmetry-resolved entanglement detection We
refer
the
reader
to24
or
to
the
Supplementary
Information
for
an
introduction
to
classical
shadows. The original classical shadow formalism is contingent
on noiseless measurements and sources that produce i.i.d. states. Subsequently, it was shown that classical shadows can also handle
noisy measurements49,50. As we show in detail in the Supplemen-
tary Information, the formalism can also be extended to take non-
identical, but independent, state preparations ("drifts”) into
account, i.e. the source produces the states {ρ1, ρ2, …, ρN}. In this
case, each snapshot ^ρi will have a different expectation value and p
p
Finally, let us stress that this error bound addresses the
estimation of DðiÞ
2 ðρavgÞ in terms of a single U-statistics estimator. The poor scaling in 1/δ can be exponentially improved by dividing
the classical shadow into equally-sized batches and performing
a median-of-U-statistics estimation instead24: 1=δ ! const ´
log ð1=δÞ. However, numerical experiments conducted in ref. 23
suggest that this trade-off is only worthwhile if one attempts to
predict many properties with the same data set. 1
N
X
N
i¼1
^ρi ! 1
N
X
N
i¼1
E^ρi ¼ 1
N
X
N
i¼1
ρi ¼: ρavg ;
(25) (25) SR inequalities applied to states without symmetries. In the last
part of this section, we show that the SR inequalities can also be
used to detect the entanglement of arbitrary states, including
those that do not have any symmetry. i.e., the average of the snapshots converges to the average state. Since different snapshots are statistically independent, it turns out
that one can estimate linear functions, say tr ðOρavgÞ. y y
y
The reason for that is that there exists a local channel C that
transforms any state ρ into a state σ CðρÞ that has the desired
block structure. The channel can be realized with local operations
assisted by classical communication, and can thus not generate
entanglement. Therefore, the initial state ρ must be at least as
entangled as the final block diagonal state σ. This statement holds
for any entanglement measure. As a consequence, if entangle-
ment is detected in σ (which can be investigated using the
symmetry-resolved tools), then ρ is necessarily also entangled. In
other words, looking at the entanglement of σ, the "block-
diagonalized” version of ρ, provides a sufficient condition of
entanglement for ρ. Symmetry-resolved entanglement detection ρ ¼
M
nþm
q¼0
ρðqÞ ¼
X
q
QqρQq;
(13) (13) where each block (or sector) is labeled by an eigenvalue q ∈{0, 1, where each block (or sector) is labeled by an eigenvalue q ∈{0, 1, where each block (or sector) is labeled by an eigenvalue q ∈{0, 1, Published in partnership with The University of New South Wales A. Neven et al. 5 Lemma 2. Given a symmetric state ρ = ∑iQiρQi, for each eigenvalue i
of N A N B, it holds that Lemma 2. Given a symmetric state ρ = ∑iQiρQi, for each eigenvalue i
of N A N B, it holds that where, using Lemma 2, where, using Lemma 2, where, using Lemma 2, Q ¼ Lð1Þ
i
Lð1Þ
i
Lð2Þ
i
:
(27)
ð Þ (27) tr ðPiρΓPiÞ
k ¼ tr ðLðkÞ
i ρkÞ
(20) tr ðPiρΓPiÞ
k ¼ tr ðLðkÞ
i ρkÞ (20) Recall that DðiÞ
2 ðρÞ < 0 implies that ρ is entangled. Here, we will
provide confidence intervals in the estimation of DðiÞ
2 ðρavgÞ given a
fixed number of measurements. Observe that DðiÞ
2 ðρavgÞ < 0 implies
that there is at least a value k ∈{1, 2, …, N} such that DðiÞ
2 ðρkÞ < 0. Thus, DðiÞ
2 ðρavgÞ < 0 implies that the source is able to produce
entangled states. Recall that DðiÞ
2 ðρÞ < 0 implies that ρ is entangled. Here, we will
provide confidence intervals in the estimation of DðiÞ
2 ðρavgÞ given a
fixed number of measurements. Observe that DðiÞ
2 ðρavgÞ < 0 implies
that there is at least a value k ∈{1, 2, …, N} such that DðiÞ
2 ðρkÞ < 0. Thus, DðiÞ
2 ðρavgÞ < 0 implies that the source is able to produce
entangled states. Symmetry-resolved entanglement detection st, t s st a g t o
a d to see t at
trnA1Bk ~SðA1; ¼ ; AkÞ SðB1; ¼ ; BkÞρk
¼ ððρΓÞ
kÞ
Γ
:
(22)
Then, we have that
tr ðLðkÞ
i ρkÞ ¼ tr
Pi ððρΓÞkÞ
Γ
¼ tr
PΓ
i ðρΓÞk
¼ tr
Pi ðρΓÞk
:
(23) trnA1Bk ~SðA1; ¼ ; AkÞ SðB1; ¼ ; BkÞρk
¼ ððρΓÞ
kÞ
Γ
:
(22) trnA1Bk ~SðA1; ¼ ; AkÞ SðB1; ¼ ; BkÞρk
¼ ððρΓÞ
kÞ
Γ
:
(22)
Then, we have that (22) (22) tr ðLðkÞ
i ρkÞ ¼ tr
Pi ððρΓÞkÞ
Γ
¼ tr
PΓ
i ðρΓÞk
¼ tr
Pi ðρΓÞk
:
(23) Theorem 1. (Error bound for DðiÞ
2 ) Fix ϵ ∈(0, 1) (accuracy of the
approximation), δ ∈(0, 1) (probability-of-error threshold), a biparti-
tion AB, as well as a symmetry sector i. Suppose that we perform (23) N ⩾2nþmtr ðPiÞ
ϵ2δ
1
2
4 þ
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
42 þ 2 ϵ2δ2nþm
tr ðPiÞ
s
! þ 1
(29) Here, the first equality follows from the definition of LðkÞ
i
and Eq. (22); the second, from tr ðRSΓÞ ¼ tr ðRΓSÞ for any two matrices R, S;
and the third, from PΓ
i ¼ Pi. Finally, using that Pi ¼ P2
i
are
orthogonal projectors, the cyclic property of the trace, and the
block structure of ρΓ, we have (29) randomized, single-qubit measurements on independent states. Then, bD
ðiÞ
2 DðiÞ
2 ðρavgÞ
⩽ϵ
with prob: ðat leastÞ1 δ: tr
Pi ðρΓÞ
k
¼ tr
Pi ðρΓÞ
kPi
¼ tr PiρΓPi
k;
(24) (24) which completes the proof. which completes the proof. It should be noted that this bound corresponds to randomized
measurements performed using local Clifford unitaries (see
Supplementary Information). Using a different measurement
strategy, for instance with global Clifford unitaries, we could
expect a much better scaling of the measurement budget24, but
such global unitary gates are more difficult to implement, and are
still out of reach for most of the current experimental platforms. Classical shadows. Classical shadows are a convenient formalism
to reason about predicting properties of a quantum system based
on randomized single-qubit measurements performed in single-
copy
experiments24. npj Quantum Information (2021) 152 Published in partnership with The University of New South Wales Symmetry-resolved entanglement detection where the operators LðkÞ
i
are given by where the operators LðkÞ
i
are given by LðkÞ
i
¼
P
ab¼i
ΠaðA1Þ 1 1 ΠbðBkÞ
~SðA1; ¼ ; AkÞ SðB1; ¼ ; BkÞ g
Let us finally introduce (see the Supplementary Information for
details) the empirical average, over N(N −1) pairs of independent
snapshots, with ~SðA1; ¼ ; AkÞ
¼
P
a1
P
ak
aka1 ak1
j
i a1 ak
h
j ;
SðB1; ¼ ; BkÞ
¼
P
b1
P
bk
b2 bkb1
j
i b1 bk
h
j:
(21) (21) bD
ðiÞ
2 ¼
1
NðN 1Þ
X
i≠j
tr Q^ρi ^ρj
:
(28) (28) Here, the sum over each ai (bi) runs from 1 to 2n (2m) respectively. Here, the sum over each ai (bi) runs from 1 to 2n (2m) respectively. We can fix the desired approximation accuracy ϵ and a probability-
of-error threshold δ to obtain a lower bound on the measurement
budget N. For simplicity, let us consider a non-trivial sector (i.e. q ≠
−m, n) and assume n + m ⩾4. Then one has the following
theorem (see the Supplementary Information for a slightly better
bound). We can fix the desired approximation accuracy ϵ and a probability-
of-error threshold δ to obtain a lower bound on the measurement
budget N. For simplicity, let us consider a non-trivial sector (i.e. q ≠
−m, n) and assume n + m ⩾4. Then one has the following
theorem (see the Supplementary Information for a slightly better
bound). Proof. First, it is straightforward to see that Proof. First, it is straightforward to see that Proof. First, it is straightforward to see that budget N. For simplicity, let us consider a non-trivial sector (i.e. q ≠
−m, n) and assume n + m ⩾4. Then one has the following
theorem (see the Supplementary Information for a slightly better
bound). oo . Applications can be used to reveal the presence of entanglement at short times. We show in Fig. 3 a numerical confirmation of these results for
various values of γ/J and N = 8, which was obtained by simulating
Eq. (32). We note that, in the present context, utilizing symmetry-
resolution is fundamental to detect entanglement: this is due to
the fact that the negative eigenvalues in ρΓ appear in sectors that
are not macroscopically populated34, so that moments without
symmetry resolution would not be able to detect them. In this section, we apply and compare the entanglement
conditions presented in the previous sections on various physical
systems. For the systems possessing a symmetry as discussed
above, we highlight some of the advantages that can result from
considering symmetry-resolved entanglement detection tools. Entanglement detection in quench dynamics. We begin
by
considering the situation of quench dynamics, where entangle-
ment emerges from the dynamics of a many-body Hamiltonian. We consider the model presented in ref. 34, where the interplay
between coherent dynamics with U(1) symmetry and dissipation
leads to a dynamics of "purification”. Here, we will use the same
formalism to show how entanglement is generated at short
times, and can be detected via the symmetry-resolved versions
of the D2 and p3-PPT conditions. In the next subsection, we will
consider an analogous experimental situation obtained with
trapped ions28. Experimental demonstration in a trapped-ion quantum simulator. In the previous section, we showed in an idealized theoretical
setting that entanglement is generated –and can be revealed via
SR-entanglement conditions– at early times after a quantum
quench. Here, we demonstrate this effect experimentally via the
measurement of the SR-D2 and SR-p3-ppT condition using
randomized measurement data taken at early times after quantum Fig. 3
Symmetry resolved entanglement detection in quench
dynamics with spin excitation loss. We study SR-entanglement in
quench dynamics in a system consisting of N = 8 spins initialized in
a Néel state #"
j
iN=2 and evolved with HXX subject to spin excitation
loss with various rates γ (γ/J increases with the darkness of the color,
see insets). We take A = [1, 2, 3, 4] and B = [5, 6, 7, 8]. In panels (a) and
(b), the D2 ratio and p3-PPT ratio of sector q = −1 are shown,
respectively. Entanglement is detected for values below unity in the
shaded gray areas. Applications The insets in (a) and (b) show the early time value
at t = 0+ of the D2-ratio a) and p3-PPT ratio (b), respectively, as a
function of the decoherence rate γ/J. Black lines are the perturbation
theory results displayed in Eqs. (40) and (39). pp
Our model is described by a master equation Our model is described by a master equation
∂tρ ¼ i
_ ½HXY; ρ þ
X
j
γ σ
j ρσþ
j 1
2 fσþ
j σ
j ; ρg
;
(32)
with the lowering and raising operators σ
j ¼ ðXj iYjÞ=2,
σþ
j ¼ ðXj þ iYjÞ=2, and the Hamiltonian ∂tρ ¼ i
_ ½HXY; ρ þ
X
j
γ σ
j ρσþ
j 1
2 fσþ
j σ
j ; ρg
;
(32) ∂tρ
_ ½HXY; ρ þ
X
j
γ σj ρσj
2 fσj σj ; ρg
;
(32)
with the lowering and raising operators σ
j ¼ ðXj iYjÞ=2,
σþ
j ¼ ðXj þ iYjÞ=2, and the Hamiltonian with the lowering and raising operators σ
j ¼ ðXj iYjÞ=2,
σþ
j ¼ ðXj þ iYjÞ=2, and the Hamiltonian HXY ¼ _
2
X
i<j
JijðXiXj þ YiYjÞ
(33) (33) Fig. 3
Symmetry resolved entanglement detection in quench
dynamics with spin excitation loss. We study SR-entanglement in
quench dynamics in a system consisting of N = 8 spins initialized in
a Néel state #"
j
iN=2 and evolved with HXX subject to spin excitation
loss with various rates γ (γ/J increases with the darkness of the color,
see insets). We take A = [1, 2, 3, 4] and B = [5, 6, 7, 8]. In panels (a) and
(b), the D2 ratio and p3-PPT ratio of sector q = −1 are shown,
respectively. Entanglement is detected for values below unity in the
shaded gray areas. The insets in (a) and (b) show the early time value
at t = 0+ of the D2-ratio a) and p3-PPT ratio (b), respectively, as a
function of the decoherence rate γ/J. Black lines are the perturbation
theory results displayed in Eqs. (40) and (39). and where γ is the spontaneous emission rate. Here, we consider
open boundary conditions. The hopping between spins i and j is
described by the coefficient Jij and we consider in this subsection
nearest-neighbor hopping Jij = Jδi+1,j with strength J. A. Neven et al. The fact that this
channel maps ρ to a state σ that is block-diagonal in the number-
of-excitations basis can easily be seen as follows. First, observe
that for any j ∈{0, …, 2(n+m) −1}, the computational basis state jj i
is an eigenvector of Ui, associated to an eigenvalue, ð1Þjjji=2k, that
essentially depends on ∣j∣, the number of excitations of
jj i. Therefore, we have þJt iσþ
NAþ1ρð0Þσþ
NA þ h:c
:
(35) (35) The presence of a negative eigenvalue in this sector can be
detected from the value of the moments is an eigenvector of Ui, associated to an eigenvalue, ð1Þjjji=2k, that
essentially depends on ∣j∣, the number of excitations of
jj i. Therefore, we have p1ðρΓ
ð1ÞðtÞÞ ¼ γNAt
2
;
(36)
p2ðρΓ
ð1ÞðtÞÞ ¼ 2J2t2;
(37)
p3ðρΓ
ð1ÞðtÞÞ ¼ 3γJ2t3;
(38)
l
d
d
l
h (36) σ ¼ 1
2k
X
2ðmþnÞ1
j;j0¼0
X
2k1
i¼0
ð1Þ
iðjjjjj0jÞ
2k ρj;j0 jj i j0
h j:
(31) (31) p3ðρΓ
ð1ÞðtÞÞ ¼ 3γJ2t3;
(38) For any j and j0 having different number of excitations, i.e. such
that jjj ≠jj0j, the sum over i in Eq. (31) vanishes, explaining why σ is
diagonal in the number-of-excitations basis. p3ðρΓ
ð1ÞðtÞÞ ¼ 3γJ2t3; (38) in leading order in J ≫γNA. In particular, the p3-PPT ratio in leading order in J ≫γNA. In particular, the p3-PPT ratio As can be seen from the argument above, the non-zero
elements of σ are all equal to the corresponding elements of ρ. This implies that the channel can effectively be replaced by a sum
of projectors onto all the number-of-excitations sectors. From an
experimental point of view, the practical implementation of this
channel can thus be circumvented by using the observables of
Lemma 2 in the post-processing of the classical shadows. p3ðρΓ
ð1ÞðtÞÞp1ðρΓ
ð1ÞðtÞÞ
p2ðρΓ
ð1ÞðtÞÞ2
¼ 3γ2NA
8J2
1;
(39) (39) and the D2 condition and the D2 condition and the D2 condition p1ðρΓ
ð1ÞðtÞÞ
2
p2ðρΓ
ð1ÞðtÞÞ ¼ γ2N2
A
8J2 1;
(40) (40) npj Quantum Information (2021) 152 Published in partnership with The University of New South Wales A. Neven et al. A. Neven et al. A. Neven et al. 6 that the channel C destroys all the entanglement of ρ. that the channel C destroys all the entanglement of ρ. The local channel that can be used for this approach is the
f ll that the channel C destroys all the entanglement of ρ. The local channel that can be used for this approach is the
following: As we are interested in short-time dynamics, we can solve Eq. (32) in first-order in perturbation theory, which is valid for t ≪1/
J, 1/γ. Considering for concreteness a half-partition A, made of
the first NA sites, we obtain a block with a negative eigenvalue34. Assuming for simplicity NA = N/2, NA even, we obtain C : ρ ! CðρÞ ¼ 1
2k
X
2k1
i¼0
UðnþmÞ
i
ρ ðUy
i Þ
ðnþmÞ
(30) (30) ρΓ
ð1ÞðtÞ ¼ γt
X
NA=2
m¼1
σ
2mρð0Þσþ
2m
(34) ρΓ
ð1ÞðtÞ ¼ γt
X
NA=2
m¼1
σ
2mρð0Þσþ
2m
(34)
þJt iσþ
NAþ1ρð0Þσþ
NA þ h:c
:
(35) (34) where k ¼ blog ðn þ mÞc þ 1 and Ui ¼ Zi=2k. The fact that this
channel maps ρ to a state σ that is block-diagonal in the number-
of-excitations basis can easily be seen as follows. First, observe
that for any j ∈{0, …, 2(n+m) −1}, the computational basis state jj i
k where k ¼ blog ðn þ mÞc þ 1 and Ui ¼ Zi=2k. The fact that this
channel maps ρ to a state σ that is block-diagonal in the number-
of-excitations basis can easily be seen as follows. First, observe
that for any j ∈{0, …, 2(n+m) −1}, the computational basis state jj i
k where k ¼ blog ðn þ mÞc þ 1 and Ui ¼ Zi=2k. The fact that this
channel maps ρ to a state σ that is block-diagonal in the number-
of-excitations basis can easily be seen as follows. First, observe
that for any j ∈{0, …, 2(n+m) −1}, the computational basis state jj i
is an eigenvector of Ui, associated to an eigenvalue, ð1Þjjji=2k, that
essentially depends on ∣j∣, the number of excitations of
jj i. Therefore, we have where k ¼ blog ðn þ mÞc þ 1 and Ui ¼ Zi=2k. Symmetry-resolved entanglement detection This condition is not necessary as it could be ð
a gÞ
These ideas regarding the prediction of linear observables do
extend to higher-order polynomials. Here, we restrict ourselves to
the quadratic case24, involving up to second-order moments of a
block of the partial transpose. An extension to higher-order
polynomials is conceptually straightforward, but can become
somewhat tedious to analyze23. Let us fix a block label i and
consider the second-order moment inequality restricted to this
block: DðiÞ
2 ðρÞ ¼ tr PiρΓPi
ð
Þ
ð
Þ
2 tr PiρΓPi
ð
Þ
2
¼ p1ðPiρΓPiÞ
ð
Þ2 p2ðPiρΓPiÞ ¼ trðQρ ρÞ ;
(26) (26) npj Quantum Information (2021) 152 Published in partnership with The University of New South Wales Applications In
particular, we show that the SR-D2 condition and SR-p3-PPT
condition allow for a fine-grained detection of bipartite entangle-
ment, in regimes where the corresponding global conditions23
and conditions relying on the purities of different subsystems28
are not conclusive. irrespective of the quantum state in question, for the specific
states in the experiment only 8 × 105 have been performed. The
errorbars of the experimental are then drawn at 1.96σ where the
standard error of the mean σ has been estimated for each data
point using Jackknife resampling. For normally distributed data
with empirical mean μ, μ ± 1.96σ defines a 95% confidence
interval. While normal distribution is here not guaranteed a priori,
we checked through additional numerical simulations of many
experiments (with fixed number of runs per experiment) that
errorbars of length 1.96σ indeed approximate a confidence
interval with confidence level 95%. In the experiment reported in ref. 28, a one-dimensional spin-1/
2-chain, consisting of N = 10 spins, was initialized in the Néel state
"#
j
i5 and time-evolved with the Hamiltonian HXY [Eq. (33)] where
the coupling parameter Jij follows the approximate power-law
decay Jij ≈J0/∣i −j∣α, with α ≈1.24, J0 = 420s−1. The Hamiltonian
evolution exhibits a global U(1)-symmetry conserving the total
magnetization of the system (i.e., [H, ∑iZi] = 0). Symmetry-breaking
terms (such as σþ
i σþ
j þ h.c. ) are strongly suppressed due to a
large effective magnetic field28. As detailed in refs. 28,34 weak
decoherence effects are present in the experiment, including
imperfect initial state preparation, local spin-flips and spontaneous
emission during the dynamics, and measurement errors model as
local depolarization. Note that coherent spin-flips do not preserve
the
global
magnetization
and
block-diagonal
form
of
the
(reduced) density matrix. On the timescales accessed in the
experiment, these effects are however very weak (causing in
numerical simulations including the above decoherence model a
purity mismatch of the order of 10−5 of the full 10-spin density
matrix ρ vs. the projected one ρQ = ∑qQqρQq at t = 5 ms). While the SR-D2 condition requires only the estimation of first
and second PT-moment, the third order SR-p3-PPT condition
[panel b)], allows to detect entanglement in an even wider time
window. In comparison to the global p3-PPT condition [red curve
in panel b)], this clearly demonstrates the benefit of taking
symmetry-resolution into account. Entanglement detection in the ground state of the XXZ model. Applications Since the XXZ
spin chain exhibits a U(1) symmetry related to magnetization
conservation, we can exploit the symmetry-resolved counterpart
of the Dk conditions, the p3-PPT and their optimized version Dopt
3 . 3
The simulation results are shown in Fig. 5. We consider the
ground state of an open chain of L = 14 sites. In a) and b), we
select ℓ= 10 sites in the middle as subsystem A and divide it in
two parts A = A1 ∪A2. We use the negativity as a reference to
benchmark the efficiency of some entanglement conditions to
detect entanglement between A1 and A2. The fact that the SR-D2 condition can reveal the presence of
entanglement is particularly interesting from an experimental
point of view as it implies that entanglement can be detected
from the estimation of only two moments of the partial transpose
(in a sector). For the shadow estimation of D2( −1), our rigorous
bound from Theorem 1 ensures that ~ 1.3 × 106 measurements
would be sufficient to guarantee entanglement detection with a
probability of 95%. While this represents an upper bound, valid In Fig. 5a), we calculate the p3-PPT, the D3, the optimal Dopt
3
condition, and the Stieltjes condition using moments up to order
five (see Methods). The convention we choose in the plot is that
entanglement is detected whenever the value is positive. All the
conditions work in most of the interval Jz ∈[ −4, 1], where we
expect entanglement to be sizeable, except for D3 failing in the
vicinity of Jz = 1. The presence of entanglement is confirmed by
the calculation of the negativity (red line). Fig. 4
Experimental SR-entanglement detection in a trapped ion
quantum simulator using data obtained in ref. 28. For a total
system of N = 10 spins and subsystem A, B = [4, 5], [6, 7], we present
in a) the SR-D2 ratio and b) SR-p3-PPT ratio as a function of time for
various symmetry sectors . In both panels, entanglement is detected
in regimes where the corresponding global conditions do not reveal
entanglement, as indicated in the shaded grey areas (values below
unity). The points with error bars correspond to the values and
uncertainties extracted from the experimental data from ref. 28,
whereas the dashed (solid) lines are theoretical simulations of
unitary dynamics (taking decoherence effects into account), as
detailed in refs. 28,34. In Fig. Applications The initial
state is the Néel state ρð0Þ ¼ ψð0Þ
j
i ψð0Þ
h
j, with ψð0Þ
j
i ¼ #"
j
iN=2. As shown in ref. 34, the time evolved state ρ(t) of the N spin
system has the block diagonal form of Eq. (13). Moreover, the
partially transposed matrix w.r.t a partition A, ρΓ is also block
diagonal with blocks ρΓ
ðqÞ. Here, the index q represents the
difference between the number of spin excitations in A and the
one in the complement partition B. Fig. 3
Symmetry resolved entanglement detection in quench
dynamics with spin excitation loss. We study SR-entanglement in
quench dynamics in a system consisting of N = 8 spins initialized in
N=2 and where γ is the spontaneous emission rate. Here, we consider
open boundary conditions. The hopping between spins i and j is
described by the coefficient Jij and we consider in this subsection
nearest-neighbor hopping Jij = Jδi+1,j with strength J. The initial
state is the Néel state ρð0Þ ¼ ψð0Þ
j
i ψð0Þ
h
j, with ψð0Þ
j
i ¼ #"
j
iN=2. As shown in ref. 34, the time evolved state ρ(t) of the N spin
system has the block diagonal form of Eq. (13). Moreover, the
partially transposed matrix w.r.t a partition A, ρΓ is also block
diagonal with blocks ρΓ
ðqÞ. Here, the index q represents the
difference between the number of spin excitations in A and the
one in the complement partition B. a Néel state #"
j
iN=2 and evolved with HXX subject to spin excitation
loss with various rates γ (γ/J increases with the darkness of the color,
see insets). We take A = [1, 2, 3, 4] and B = [5, 6, 7, 8]. In panels (a) and
(b), the D2 ratio and p3-PPT ratio of sector q = −1 are shown,
respectively. Entanglement is detected for values below unity in the
shaded gray areas. The insets in (a) and (b) show the early time value
at t = 0+ of the D2-ratio a) and p3-PPT ratio (b), respectively, as a
function of the decoherence rate γ/J. Black lines are the perturbation
theory results displayed in Eqs. (40) and (39). Published in partnership with The University of New South Wales A. Neven et al. 7 quench in a trapped-ion quantum simulator (c.f. ref. 28). Applications The
XXZ spin chain is a generalization of the Heisenberg chain
including an anisotropy in the interaction along the z direction,
whose Hamiltonian reads: H ¼ J
X
i
XiXiþ1 þ
X
i
YiYiþ1 þ Jz
X
i
ZiZiþ1
! :
(41) (41) ρ
p
j
ρQ
q
qρ
q
In ref. 28 randomized measurements were performed at various
times (t = 0ms, …, 5ms) after the quantum quench. As described
in detail in ref. 34 (see also the Supplementary Information), we can
use this data to estimate SR-PT moments and the SR entangle-
ment conditions via classical shadow formalism24. In Fig. 4, we
present the SR D2 and p3-PPT conditions in the different sectors,
for a subsystem consisting of the neighbouring spins A, B = [4, 5],
[6, 7] and where the partial transpose is taken in the subsystem
A = [4, 5]. Similar to the results of the previous subsection, both
conditions detect entanglement at short times in the sector q =
−1. The corresponding global conditions, in particular the global
p3-PPT condition, do not reveal the presence of entanglement in
this regime [see Fig. 4 b)]. We will fix J = 1 as energy unit: Jz sets the strength of the
anisotropy
along
the
z-axis. The
phase
diagram
at
zero
temperature is known51: the system hosts an antiferromagnetic
phase when Jz < −1, a Luttinger liquid for Jz ∈[ −1, 1], and a
ferromagnetic one for Jz > 1. We might expect that the entangle-
ment conditions we described in the previous sections will detect
that the state is not PPT in the range Jz 2 1; 1. Since the XXZ
spin chain exhibits a U(1) symmetry related to magnetization
conservation, we can exploit the symmetry-resolved counterpart
of the Dk conditions, the p3-PPT and their optimized version Dopt
3 . We will fix J = 1 as energy unit: Jz sets the strength of the
anisotropy
along
the
z-axis. The
phase
diagram
at
zero
temperature is known51: the system hosts an antiferromagnetic
phase when Jz < −1, a Luttinger liquid for Jz ∈[ −1, 1], and a
ferromagnetic one for Jz > 1. We might expect that the entangle-
ment conditions we described in the previous sections will detect
that the state is not PPT in the range Jz 2 1; 1. npj Quantum Information (2021) 152 Published in partnership with The University of New South Wales Applications 5b) we focus on the q = 1 sector. In this case, we observe
that all conditions indicate the presence of at least a negative
eigenvalue in the sector ρΓ
Aðq ¼ 1Þ - that is, they are informative
about which sector for the reduced density matrix contributes to
violating PPT. In this specific instance, SR is however not
fundamental in detecting entanglement beyond what non-SR
conditions can. In Fig. 5c) and d), we carry out the same analysis for
disconnected partitions. We consider A = A1 ∪A2, where A1
consists of the first l/2 sites and A2 of the last l/2, and L = 14, l
= 10. In Fig. 5c) for Jz ~ −1.9 all the quantities except Stieltjes5 are
below zero, thus not revealing entanglement even though the
negativity is positive. In this plot, one can also see that, for Jz < −2,
the optimized condition Dopt
3
detects entanglement whereas both
p3-PPT and D3 fail. This illustrates that the slight improvement
obtained from the optimization (see Fig. 2) can be decisive to
detect the entanglement of physically relevant states from the first
three moments only. Fig. 4
Experimental SR-entanglement detection in a trapped ion
quantum simulator using data obtained in ref. 28. For a total
system of N = 10 spins and subsystem A, B = [4, 5], [6, 7], we present
in a) the SR-D2 ratio and b) SR-p3-PPT ratio as a function of time for
various symmetry sectors . In both panels, entanglement is detected
in regimes where the corresponding global conditions do not reveal
entanglement, as indicated in the shaded grey areas (values below
unity). The points with error bars correspond to the values and
uncertainties extracted from the experimental data from ref. 28,
whereas the dashed (solid) lines are theoretical simulations of
unitary dynamics (taking decoherence effects into account), as
detailed in refs. 28,34. Published in partnership with The University of New South Wales A. Neven et al. 8 4
2
0
2
Jz
0.00
0.25
0.50
0.75
1.00
X
a)
N
D3
Stieltjes5
p2
2
p3p1
Dopt
4
2
0
2
Jz
0.00
0.05
0.10
0.15
X(1)
b)
4
2
0
2
Jz
0.0
0.1
0.2
X
c)
4
2
0
2
Jz
0.00
0.05
0.10
0.15
X(1)
d)
Fig. 5
Entanglement conditions on the ground state of XXZ model. (42) where Pi ¼ 0j ii 0
h j are local projectors. As can be seen from Eq. (42), each spin undergoes independent Rabi-oscillations as long as
the neighbouring spins are in their ground state
0j i. This
constraint breaks the tensor product structure of the Hilbert
space (as it is the case in a lattice gauge theory54). The model
effectively resembles the experimental situation in52 if the
Rydberg atoms are driven close to resonance and neighbouring
atoms cannot be simultaneously in the state
1j i due to the
Rydberg blockade mechanism. The Hamiltonian (42) has recently
attracted great interest in the context of quantum many-body
scarring55,56. In particular, performing a quantum quench on
special unentangled product states results in long-lived periodic
revivals which have been attributed to the existence of quantum
scarred eigenstates in the many-body spectrum55. where Pi ¼ 0j ii 0
h j are local projectors. As can be seen from Eq. (42), each spin undergoes independent Rabi-oscillations as long as
the neighbouring spins are in their ground state
0j i. This
constraint breaks the tensor product structure of the Hilbert
space (as it is the case in a lattice gauge theory54). The model
effectively resembles the experimental situation in52 if the
Rydberg atoms are driven close to resonance and neighbouring
atoms cannot be simultaneously in the state
1j i due to the
Rydberg blockade mechanism. The Hamiltonian (42) has recently
attracted great interest in the context of quantum many-body
scarring55,56. In particular, performing a quantum quench on
special unentangled product states results in long-lived periodic
revivals which have been attributed to the existence of quantum
scarred eigenstates in the many-body spectrum55. Finally, we investigate the required number of experimental
runs in order to measure the conditions up to the given error
bar. The classical shadow is constructed by sampling bit strings
from the quantum state after applying a global random unitary
on subsystem A. At each point in Fig. 6c), we collect 5000-bit
strings in different random basis. Such global random unitaries
in Rydberg systems can be implemented via random quenches
with local disorder potentials57. The entire estimation of the
conditions is repeated 20 times in order to obtain statistical
uncertainties. Note that statistical covariances among the
measured moments trðρΓÞ
n can give rise to nonuniform sizes
for the error bars. In Fig. Applications With X 2 ½D3; Dopt
3 ; Stieltjes5; p2
2 p3p1; N we denote the conditions
computed on ρA and with X(1) the ones on ρA(q = 1). Chain length L = 14, subsystem length ℓ= 10. We consider in a)-b) a connected
subsystem A of length ℓat the center of the chain; in c)-d) a disjoint interval A consisting of ℓ/2 sites at the beginning and ℓ/2 sites at the end
of the chain. We set the sign of the inequalities such that a positive value indicates the violation of a PPT condition, and thus the presence of
entanglement. To compare data of different magnitude we multiply the Stieltjes5 condition by 102 in a), 105 in b), 104 in c), 105 in d). 4
2
0
2
Jz
0.00
0.05
0.10
0.15
X(1)
b) b) Jz
4
2
0
2
Jz
0.00
0.05
0.10
0.15
X(1)
d) d) Jz Jz Fig. 5
Entanglement conditions on the ground state of XXZ model. With X 2 ½D3; Dopt
3 ; Stieltjes5; p2
2 p3p1; N we denote the conditions
computed on ρA and with X(1) the ones on ρA(q = 1). Chain length L = 14, subsystem length ℓ= 10. We consider in a)-b) a connected
subsystem A of length ℓat the center of the chain; in c)-d) a disjoint interval A consisting of ℓ/2 sites at the beginning and ℓ/2 sites at the end
of the chain. We set the sign of the inequalities such that a positive value indicates the violation of a PPT condition, and thus the presence of
entanglement. To compare data of different magnitude we multiply the Stieltjes5 condition by 102 in a), 105 in b), 104 in c), 105 in d). Entanglement detection under constrained dynamics. As a third
example, we study the detection of mixed-state entanglement in
subsystems of constrained spin models after a global quantum
quench. Such models have been realized experimentally with
neutral atoms in optical tweezer arrays coupled to Rydberg
states52,53. Below we simulate an experiment, in which moments
of the partially transposed density matrix are obtained from a
classical shadow involving global random unitaries available in
current experimental setups. In particular, we demonstrate that
periodic revivals of mixed state entanglement can be detected
from the conditions D3 and D4 (Eqs. (11) and (12)) requiring only a
small number of experimental runs. Published in partnership with The University of New South Wales npj Quantum Information (2021) 152 Applications In the following we study the conditions given in Eqs. (11)
and (12), when a quench is performed from a product state
that leads to kinetically constrained dynamics. To this end, the
initial state Ψ0
j
i ¼ 10
j
iN=2 is time evolved with the Hamilto-
nian (42) up to t = 50/Ω. Figure 6a) shows the local Zi-
expectation values exhibiting long-lived persistent oscillations. This striking departure from a thermalizing behaviour is also
reflected in the slow growth of entanglement entropy (Panel
b). We now analyse the time-resolved behaviour of mixed state
entanglement for a subsystem depicted in the inset of Fig. 6c). The revivals in the negativity indicate that spins in the
subsystem get periodically entangled and disentangled with
each other. Interestingly, the p3-PPT condition is unable to
detect the revivals, while D3 yields positive values at the first 3
peaks in the negativity. At later times, the D3 fails to detect the
entanglement present in the system, but this entanglement is
still captured by D4. Entanglement detection under constrained dynamics. As a third
example, we study the detection of mixed-state entanglement in
subsystems of constrained spin models after a global quantum
quench. Such models have been realized experimentally with
neutral atoms in optical tweezer arrays coupled to Rydberg
states52,53. Below we simulate an experiment, in which moments
of the partially transposed density matrix are obtained from a
classical shadow involving global random unitaries available in
current experimental setups. In particular, we demonstrate that
periodic revivals of mixed state entanglement can be detected
from the conditions D3 and D4 (Eqs. (11) and (12)) requiring only a
small number of experimental runs. We consider the Fibonacci chain with open boundary condi-
tions described by the Hamiltonian H ¼ Ω
X
i
Pi1XiPiþ1 ; H ¼ Ω
X
i
Pi1XiPiþ1 ;
(42) (42) A. Neven et al. 9 Fig. 6
Entanglement detection in quench dynamics of a kineti-
cally constrained Rydberg chain (42). a Coherent oscillations of the
Zi-expectation values in a quench with an 18-site Fibonacci chain
from a staggered initial state: Ψ0
j
i ¼ 1010 ¼
j
i. b Von Neumann
entropy of half partition of the chain as a function of time if a global
quench is performed on 2 different initial states. c Entanglement
detection in a subsystem of 5 spins (corresponding to the red dots
in the inset), with partial transpose taken with respect to the first
three. All plotted curves are tuned so that positive values indicate
entanglement. The revivals in the negativity indicate periodic
entangling and disentangling of spins within the subsystem. The
goblal p3 −PPT condition is unable to detect entanglement in the
entire time window in contrast to the D3 and D4 conditions. The
points are obtained from a classical shadow consisting of 5000
global random unitaries. Error bars are obtained by repeating the
procedure 20 times and estimating the standard error. derived error bounds and confidence intervals for D2, with a
natural extension to quantities involving higher-order moments. Empirical
evaluations
complement
our
theoretical
findings. Applications to several theoretical models, as well as experimental
data, demonstrate both tractability and viability of our approach. ,
y
y
pp
We are confident that this work opens up several interesting
future research directions. Firstly, the sequence of Dk’s is designed
to detect bipartite entanglement in a reliable and experimentally
accessible fashion. A natural next step is to try to extend similar
ideas
to
multipartite
entanglement
detection,
e.g. using
non–linear entanglement witnesses 58,59. Secondly, the complete
sequence of Dk’s is used to answer a binary question: is the partial
transpose positive or not? Entanglement measures, like the
negativity, address entanglement in a quantitative fashion, but
are also harder to estimate. Is it possible to use moments (or other
density matrix functionals) to define entanglement measures that
are experimentally tractable? The statistical analysis of the
estimation procedure is also far from complete. We have shown
that independence between the states that are produced in each
iteration of an experiment is enough to derive statistically sound
confidence intervals for estimating matrix moments with classical
shadows. This addresses the practically relevant case of drifting
sources, but further extensions to correlated states would also be
appealing. Note While completing the writing of the present work, we became
aware of a work by Yu et al.36, in which similar questions have
been addressed. DISCUSSION The study of entanglement has a long and prominent history in a
variety of disciplines. And with the advent of serious quantum
technologies, reliable entanglement generation is more important
than ever. This work provides a principled approach to reliably
detect bipartite entanglement between subsystem A and sub-
system B. We have presented a set of inequality conditions Dk (1 ⩽
k ⩽2∣AB∣). Each Dk is an inequality that involves the first k moments
of the partially transposed density operator. Violation of a single
inequality implies that the underlying density operator cannot
have a PSD partial transpose. This in turn implies that the state
must be entangled. Conversely, if the underlying state is not PPT,
then, there must exist at least one Dk that is violated. This
motivates a sequence of one-sided entanglement tests. Start with
D3 – the lowest non-trivial condition – and check whether it is
violated. If this is the case, we are done. If not, we move on to the
next higher condition (D4) and repeat until we find a violation. For
states having an extensive conserved quantity (such as total
magnetization, in the case of spin systems), both the density
matrix and its partial transpose have a block-diagonal structure48. In this case, it is advisable to apply these conditions directly to
individual symmetry sectors of the partial transpose. The resulting
sequence of symmetry-resolved conditions is stronger in the
sense that lower-order moments (of blocks of the partial
transpose) suffice to detect entanglement. Importantly, this
approach is not only conceptually sound, but also tractable from
an experimental perspective. The classical shadows formalism24
allows for reliably estimating moments of the partial transpose
from randomized single-qubit measurements. We demonstrated
how to include the experimentally relevant situation of non-
identical (however independent) copies in the analysis and In this context, the quantum de Finetti theorem seems
highly relevant. In future work, we will also investigate how
importance sampling60,61 and/or derandomization24,62 can further
improve moment estimation based on classical shadows. Finally,
another promising direction of research would be to try to detect
and
characterize
phase
transitions
in
quantum
mechanical
Hamiltonians at finite (non-zero) temperatures. Quantum phase
transitions at zero temperature originate from quantum fluctua-
tions, whereas quantum phase transitions at finite temperature
are due to thermal fluctuations. Following ref. 63, quantum phase
transitions at finite temperature can be studied using negativity. It
would be interesting to investigate whether low-order PT
moments, intimately related to the negativity, can also be used
to this end. Fig. 6
Entanglement detection in quench dynamics of a kineti-
cally constrained Rydberg chain (42). a Coherent oscillations of the
Zi-expectation values in a quench with an 18-site Fibonacci chain
from a staggered initial state: Ψ0
j
i ¼ 1010 ¼
j
i. b Von Neumann
entropy of half partition of the chain as a function of time if a global
quench is performed on 2 different initial states. c Entanglement
detection in a subsystem of 5 spins (corresponding to the red dots
in the inset), with partial transpose taken with respect to the first
three. All plotted curves are tuned so that positive values indicate
entanglement. The revivals in the negativity indicate periodic
entangling and disentangling of spins within the subsystem. The
goblal p3 −PPT condition is unable to detect entanglement in the
entire time window in contrast to the D3 and D4 conditions. The
points are obtained from a classical shadow consisting of 5000
global random unitaries. Error bars are obtained by repeating the
procedure 20 times and estimating the standard error. Published in partnership with The University of New South Wales npj Quantum Information (2021) 152 Published in partnership with The University of New South Wales (42) 6c) we depict the 2σ error bars,
showing that entanglement can be detected with a moderate
experimental effort. DATA AVAILABILITY The data that support the findings of this study are available from the corresponding
author upon reasonable request. Received: 31 March 2021; Accepted: 8 September 2021; Received: 31 March 2021; Accepted: 8 September 2021; Stieltjes moment problem Note that, because we consider here an atomic density function, we
have m0 = r. We could naturally renormalize the density function so that
m0 = 1, but it would imply a re-scaling of the first moment, i.e., the trace of
ρΓ would be 1/r. Since the partial transpose and the density function
cannot be normalized at the same time, we chose to keep normalized
partial transposes. Given a sequence of moments, ðmnÞd
n¼0, the (truncated) Stieltjes moment
problem consists in finding necessary and sufficient conditions for the
existence of a measure μ on the half-line [0, ∞) such that mn ¼
Z 1
0
xndμðxÞ; 8n 2 f0; ¼ ; dg:
(46) (46) If such a measure exists, one may wonder whether it is unique or not. For
our purposes, it will be enough to discuss only its existence. Defining the
matrices If such a measure exists, one may wonder whether it is unique or not. For
our purposes, it will be enough to discuss only its existence. Defining the
matrices Descartes’ rule of signs Let A be a Hermitian matrix of dimension d. Its eigenvalues λ1, …, λd are the
roots of the characteristic polynomial PðtÞ ¼ det A t 1
ð
Þ ¼
Y
d
i¼1
ðλi tÞ:
(43) (43) For convenience, let us now consider the polynomial P( −t), which
effectively replaces the positive eigenvalues of A by negative ones and vice
versa. The coefficients of this polynomial can be expressed using the
elementary symmetric polynomials in its roots, ei(λ1, …, λd), defined as For convenience, let us now consider the polynomial P( −t), which
effectively replaces the positive eigenvalues of A by negative ones and vice
versa. The coefficients of this polynomial can be expressed using the
elementary symmetric polynomials in its roots, ei(λ1, …, λd), defined as eiðλ1; ¼ ; λdÞ ¼
X
1 ⩽j1<<ji ⩽d
λj1 ¼ λji;
(44) (44) for i = 1, …, d and with e0(λ1, …, λd) = 1. This yields for i = 1, …, d and with e0(λ1, …, λd) = 1. This yields PðtÞ ¼
X
d
i¼0
eiðλ1; ¼ ; λdÞ tdi:
(45) (45) For a polynomial with real roots (as it is the case here), Descartes’ rule of
sign states that the number of positive roots is given by the number of
sign changes between consecutive elements in the ordered list of its non-
zero coefficients (see ref. 64 and references therein). The matrix A is PSD iff
P( −t) has only negative roots, which by Descartes’ rule is the case iff there For a polynomial with real roots (as it is the case here), Descartes’ rule of
sign states that the number of positive roots is given by the number of
sign changes between consecutive elements in the ordered list of its non-
zero coefficients (see ref. 64 and references therein). The matrix A is PSD iff
P( −t) has only negative roots, which by Descartes’ rule is the case iff there A. Neven et al. 10 is no sign change in the ordered list of its non-zero coefficients, i.e., iff ei(λ1,
…, λd) ⩾0 for all i = 1, …, d, since e0(λ1, …, λd) = 1. detects more random entangled states than either popt
3
or D5. Not all
Stieltjes moment conditions are this powerful, though. For instance, the
principal minor condition for the first two rows and columns of A(k) is
trivial. Optimizing conditions involving moments up to degree three
Given a PSD matrix A, with non-zero eigenvalues λ1, …, λr, for some
r 2 ½1; dim A, consider the Lagrangian function In the first case, in
which all the eigenvalues are equal, one obtains the isolated points (p2, p3) Since this argument holds for any k ≠1, it must hold that the eigenvalues λi
are either all equal or can only take two distinct values. In the first case, in
which all the eigenvalues are equal, one obtains the isolated points (p2, p3)
= (1/r, 1/r2) in Fig. 2 in the main text. In the second case, the rank r PSD
matrices corresponding to the stationary points of the Lagrangian (54)
have a spectrum with ra degenerate eigenvalues λa and r −ra eigenvalues
λb. Assuming, without loss of generality, λa > λb, one can show that the
minimal value of the third moment is obtained when ra = r −1. where δ is the Dirac delta distribution. If ρ is PPT, this density function has
support on [0, ∞) and reproduces the moments of ρΓ, as mn ¼
Z 1
0
xndμðxÞ ¼ pnðρΓÞ:
(52) (52) Therefore, according to the solution of the Stieltjes moment problem
mentioned above, the moments of any PPT state necessary satisfy either
condition (49) or (50), depending on the value of r. The violation of any of
these conditions for a set of PT moments thus reveals that the
corresponding state must be entangled. The range condition may require
the knowledge of all the moments to be checked, but the PSD conditions
can be broken into sets of simpler conditions. And some of them only
involve low order moments. Indeed, it is well known (see e.g.66) that a
matrix is PSD if and only if all its principal minors are non-negative. For
instance, looking at the principal minor at the intersection of the first two
rows and columns of B(k), one obtains the condition m1m3 ðm2Þ2 ⩾0. This condition is nothing but the p3-PPT condition (which we know is
useful to detect entanglement23). Extending this principal minor to the
third row and column, one gets another PPT condition: npj Quantum Information (2021) 152 Optimizing conditions involving moments up to degree three
Given a PSD matrix A, with non-zero eigenvalues λ1, …, λr, for some
r 2 ½1; dim A, consider the Lagrangian function Optimizing conditions involving moments up to degree three
Given a PSD matrix A, with non-zero eigenvalues λ1, …, λr, for some
r 2 ½1; dim A, consider the Lagrangian function AðnÞ ¼
m0
m1
m2
mn
m1
m2
m3
mnþ1
m2
m3
m4
mnþ2
... ... ... .. . ... mn
mnþ1
mnþ2
m2n
0
B
B
B
B
B
B
B
@
1
C
C
C
C
C
C
C
A
(47)
and
BðnÞ ¼
m1
m2
m3
mnþ1
m2
m3
m4
mnþ2
m3
m4
m5
mnþ3
.. . .. . .. . .. . .. . mnþ1
mnþ2
mnþ3
m2nþ1
0
B
B
B
B
B
B
B
@
1
C
C
C
C
C
C
C
A
;
(48) AðnÞ ¼
m0
m1
m2
mn
m1
m2
m3
mnþ1
m2
m3
m4
mnþ2
... ... ... .. . ... mn
mnþ1
mnþ2
m2n
0
B
B
B
B
B
B
B
@
1
C
C
C
C
C
C
C
A
(47)
and
BðnÞ ¼
m1
m2
m3
mnþ1
m2
m3
m4
mnþ2
m3
m4
m5
mnþ3
.. . .. . .. . .. . .. . mnþ1
mnþ2
mnþ3
m2nþ1
0
B
B
B
B
B
B
B
@
1
C
C
C
C
C
C
C
A
;
(48) Lðλ1; ¼ ; λr; C1; C2Þ ¼
X
r
i¼1
λ3
i þ C1
X
r
i¼1
λ2
i p2
! þ C2
X
r
i¼1
λi p1
! ; (47) (54) where C1 and C2 are Lagrange multipliers. g
g
p
Here we show that, for all 1⩽r⩽dim A, the stationary points (λ1, …, λr) of
the Lagrangian function (54) are such that the variables λi can take at most
two distinct values. REFERENCES det
m1
m2
m3
m2
m3
m4
m3
m4
m5
0
B
@
1
C
A ⩾0:
(53) 1. Deutsch, I. H. Harnessing the power of the second quantum revolution. PRX
Quantum 1, 020101 (2020). (53) Optimizing conditions involving moments up to degree three
Given a PSD matrix A, with non-zero eigenvalues λ1, …, λr, for some
r 2 ½1; dim A, consider the Lagrangian function These stationary points, for which the derivatives of the
Lagrangian (54) with respect to each variable vanish, satisfy the set of
equations (48) 3λ2
i þ 2C1λi þ C2 ¼ 0 ; i ¼ 1; ¼ ; r
(55) (55) 3λ2
i þ 2C1λi þ C2 ¼ 0 ; i ¼ 1; ¼ ; r
(55)
X
r
i¼1
λ2
i ¼ p2 ;
(56)
X
r
i¼1
λi ¼ p1 :
(57) X
r
i¼1
λ2
i ¼ p2 ;
X
r
i¼1
λi ¼ p1 : X
r
i¼1
λ2
i ¼ p2 ; (56) mnþ1
mnþ2
mnþ3
m2nþ1
@
A a solution to this problem can be stated as follows65. If d is odd – such that
d = 2k + 1 for some integer k – there exists such a measure μ if and only if X
r
i¼1
λi ¼ p1 : (57) AðkÞ ⩾0; BðkÞ ⩾0 and ðmk; ¼ ; m2kþ1ÞT 2 R½AðkÞ;
(49) (49) We first sum up Eq. (55) for all values of i and then insert Eqs.(56) and (57)
into it. This yields where, given a matrix M, the notation M ⩾0 indicates that M is PSD and
RðMÞ denotes the range of M. If d is even – such that d = 2k for some
integer k – there exists such a measure μ if and only if C2 ¼ 2C1p1 3p2
r
:
(58) (58) AðkÞ ⩾0; Bðk 1Þ ⩾0 and
ðmkþ1; ¼ ; m2kÞT 2 R½Bðk 1Þ:
(50) (50) Inserting this relation into Eq. (55) and considering this equation for two
distinct values of i, say 1 and k ≠1, one can eliminate the variable C1 to get
a relation between λ1 and λk. After some algebra, one finds These solutions to the Stiltjes moment problem can be used to obtain
entanglement conditions. Given λ1, …, λr the eigenvalues of ρΓ for some
density matrix ρ, let us define the (atomic) eigenvalue distribution function λk ¼ λ1 or λk ¼ λ1p1 p2
λ1r p1
:
(59) (59) dμðxÞ ¼
X
r
i¼1
δðx λiÞ ;
(51) (51) Since this argument holds for any k ≠1, it must hold that the eigenvalues λi
are either all equal or can only take two distinct values. (53) 2. Preskill, J. Quantum computing in the nisq era and beyond. Quantum 2, 79
(2018). 3. Manipulating Quantum Systems: An Assessment of Atomic, Molecular, and Optical
Physics in the United States (The National Academies Press, 2020). We call this condition Stieltjes5. We illustrate in the Applications section
(c.f. Fig. 5) that this condition is also useful for entanglement detection. Numerical computations suggest that this condition is a powerful tool to
detect the entanglement of random mixed states, in the sense that it 4. Eisert, J. et al. Quantum certification and benchmarking. Nat. Rev. Phys. 2,
382–390 (2020). Rev. Lett. 109, 130502 (2012). 49. Chen, S., Yu, W., Zeng, P. & Flammia, S. T. Robust shadow estimation. PRX
Quantum 2, 030348 (2021). 16. Calabrese, P., Cardy, J. & Tonni, E. Entanglement negativity in extended systems: a
field theoretical approach. J. Stat. Mech. 2013, P02008 (2013). 50. Koh, D. E. & Grewal, S. Classical shadows with noise. https://arxiv.org/abs/
2011.11580 (2020). field theoretical approach. J. Stat. Mech. 2013, P02008 (2013) 17. Castelnovo, C. Negativity and topological order in the toric code. Phys. Rev. A 88,
042319 (2013). 51. Gogolin, A. O., Nersesyan, A. A. & Tsvelik, A. M. Bosonization and Strongly Corre-
lated Systems (Cambridge Univ. Press, Cambridge, 2004). 18. Eisler, V. & Zimborás, Z. Entanglement negativity in the harmonic chain out of
equilibrium. New J. Phys. 16, 123020 (2014). 52. Bernien, H. et al. Probing many-body dynamics on a 51-atom quantum simulator. Nature 551, 579–584 (2017). 19. Wen, X., Chang, P.-Y. & Ryu, S. Topological entanglement negativity in chern-
simons theories. J. High Energ. Phys. 2016, 12 (2016). 53. Ebadi, S. et al. Quantum phases of matter on a 256-atom programmable quan-
tum simulator. Nature 595, 227 (2021). simons theories. J. High Energ. Phys. 2016, 12 (2016). 54. Surace, F. M. et al. Lattice gauge theories and string dynamics in rydberg atom
quantum simulators. Phys. Rev. X 10, 021041 (2020). 20. Ruggiero, P., Alba, V. & Calabrese, P. Entanglement negativity in random spin
chains. Phys. Rev. B 94, 035152 (2016). 21. Blondeau-Fournier, O., Castro-Alvaredo, O. A. & Doyon, B. Universal scaling of the
logarithmic negativity in massive quantum field theory. J. Phys. A: Math. Theor. 49, 125401 (2016). 55. Turner, C. J., Michailidis, A. A., Abanin, D. A., Serbyn, M. & Papić, Z. Weak ergodicity
breaking from quantum many-body scars. Nat. Phys. 14, 745–749 (2018). 56. Ho, W. W., Choi, S., Pichler, H. & Lukin, M. D. Periodic orbits, entanglement, and
quantum many-body scars in constrained models: Matrix product state approach. Phys. Rev. Lett. 122, 040603 (2019). 22. Ruggiero, P., Alba, V. & Calabrese, P. Negativity spectrum of one-dimensional
conformal field theories. Phys. Rev. B 94, 195121 (2016). 23. Elben, A. et al. Mixed-state entanglement from local randomized measurements. Phys. Rev. Lett. 125, 200501 (2020). 57. Vermersch, B., Elben, A., Dalmonte, M., Cirac, J. I. & Zoller, P. Unitary n-designs via
random quenches in atomic hubbard and spin models: Application to the
measurement of rényi entropies. Phys. Rev. A. Neven et al. 11 5. Horodecki, M., Horodecki, P. & Horodecki, R. Separability of mixed states:
necessary and sufficient conditions. Phys. Lett. A 223, 1–8 (1996). 41. Tan, M. T. & Ryu, S. Particle number fluctuations, rényi entropy, and symmetry-
resolved entanglement entropy in a two-dimensional fermi gas from multi-
dimensional bosonization. Phys. Rev. B 101, 235169 (2020). 6. Terhal, B. M. Bell inequalities and the separability criterion. Phys. Lett. A 271, 319
(2000). 42. Fraenkel, S. & Goldstein, M. Symmetry resolved entanglement: exact results in 1d
and beyond. J. Stat. Mech. 2020, 033106 (2020). 7. Friis, N., Vitagliano, G., Malik, M. & Huber, M. Entanglement certification from
theory to experiment. Nat. Rev. Phys. 1, 72 (2019). 43. Murciano, S., Giulio, G. D. & Calabrese, P. Symmetry resolved entanglement in
gapped integrable systems: a corner transfer matrix approach. SciPost Phys. 8, 46
(2020). y
y
8. Horodecki, R., Horodecki, P., Horodecki, M. & Horodecki, K. Quantum entangle-
ment. Rev. Mod. Phys. 81, 865–942 (2009). 44. Azses, D. & Sela, E. Symmetry-resolved entanglement in symmetry-protected
topological phases. Phys. Rev. B 102, 235157 (2020). 9. Gühne, O. & Tòth, G. Entanglement detection. Phys. Rep. 474, 1 (2009). 10. Amico, L., Fazio, R., Osterloh, A. & Vedral, V. Entanglement in many-body systems. Rev. Mod. Phys. 80, 517–576 (2008). 45. Turkeshi, X., Ruggiero, P., Alba, V. & Calabrese, P. Entanglement equipartition in
critical random spin chains. Phys. Rev. B 102, 014455 (2020). 11. Peres, A. Separability criterion for density matrices. Phys. Rev. Lett. 77, 1413
(1996). 46. Murciano, S., Bonsignori, R. & Calabrese, P. Symmetry decomposition of negativity
of massless free fermions. SciPost Phys. 10, 111 (2021). 12. Życzkowski, K., Horodecki, P., Sanpera, A. & Lewenstein, M. Volume of the set of
separable states. Phys. Rev. A 58, 883–892 (1998). 47. Parez, G., Bonsignori, R. & Calabrese, P. Quasiparticle dynamics of symmetry-
resolved entanglement after a quench: Examples of conformal field theories and
free fermions. Phys. Rev. B 103, L041104 (2021). 13. Vidal, G. & Werner, R. F. Computable measure of entanglement. Phys. Rev. A 65,
032314 (2002). 48. Cornfeld, E., Goldstein, M. & Sela, E. Imbalance entanglement: Symmetry
decomposition of negativity. Phys. Rev. A 98, 032302 (2018). 14. Plenio, M. B. Logarithmic negativity: A full entanglement monotone that is not
convex. Phys. Rev. Lett. 95, 090503 (2005). 15. Calabrese, P., Cardy, J. & Tonni, E. Entanglement negativity in quantum field
theory. Phys. A 97, 023604 (2018). 24. Huang, H.-Y., Kueng, R. & Preskill, J. Predicting many properties of a quantum
system from very few measurements. Nat. Phys. 16, 1050 (2020). 58. Gühne, O. & Lütkenhaus, N. Nonlinear entanglement witnesses. Phys. Rev. Lett. 96,
170502 (2006). 25. Van Enk, S. J. & Beenakker, C. W. Measuring Trρn on single copies of ρ using
random measurements. Phys. Rev. Lett. 108, 110503 (2012). 59. Jungnitsch, B., Moroder, T. & Gühne, O. Taming multiparticle entanglement. Phys. Rev. Lett. 106, 190502 (2011). 26. Elben, A., Vermersch, B., Dalmonte, M., Cirac, J. I. & Zoller, P. Rényi entropies from
random quenches in atomic hubbard and spin models. Phys. Rev. Lett. 120,
050406 (2018). 60. Hadfield, C., Bravyi, S., Raymond, R. & Mezzacapo, A. Measurements of quantum
hamiltonians
with
locally-biased
classical
shadows. https://arxiv.org/abs/
2006.15788 (2020). 27. Elben, A., Vermersch, B., Roos, C. F. & Zoller, P. Statistical correlations between
locally randomized measurements: A toolbox for probing entanglement in many-
body quantum states. Phys. Rev. A 99, 052323 (2019). 61. Rath, A., van Bijnen, R., Elben, A., Zoller, P. & Vermersch, B. Importance sampling
of randomized measurements for probing entanglement. http://arxiv.org/abs/
2102.13524 (2021). 28. Brydges, T. et al. Probing rényi entanglement entropy via randomized mea-
surements. Science 364, 260–263 (2019). 62. Huang, H.-Y., Kueng, R. & Preskill, J. Efficient estimation of Pauli observables by
derandomization. Phys. Rev. Lett. 127, 030503 (2021). 29. Knips, L. et al. Multipartite entanglement analysis from random correlations. npj
Quantum Inf 6, 51 (2020). 63. Lu, T.-C. & Grover, T. Structure of quantum entanglement at a finite temperature
critical point. Phys. Rev. Research 2, 043345 (2020). 30. Ketterer, A., Wyderka, N. & Gühne, O. Characterizing multipartite entangle-
ment with moments of random correlations. Phys. Rev. Lett. 122, 120505
(2019). 64. Bensimhoun, M. Historical account and ultra-simple proofs of descartes’s rule of
signs, de gua, fourier, and budan’s rule. https://arxiv.org/abs/1309.6664 (2013). 31. Zhou, Y., Zeng, P. & Liu, Z. Single-copies estimation of entanglement negativity. Phys. Rev. Lett. 125, 200502 (2020). 65. Curto, R. E. & Fialkow, L. A. Recursiveness, positivity, and truncated moment
problems. Houston J. Math. 17, 603–635 (1991). 66. Meyer, C. D. Matrix Analysis and Applied Linear Algebra (Society for Industrial and
Applied Mathematics, 2000). 32. Ketterer, A., Wyderka, N. & Gühne, O. Entanglement characterization using
quantum designs. Quantum 4, 325 (2020). 33. Ketterer, A., Imai, S., Wyderka, N. & Gühne, O. Certifying multiparticle entangle-
ment with randomized measurements. https://arxiv.org/abs/2012.12176 (2020). npj Quantum Information (2021) 152 Published in partnership with The University of New South Wales Published in partnership with The University of New South Wales Published in partnership with The University of New South Wales A. Neven et al. A. Neven et al. 12 Reprints and permission information is available at http://www.nature.com/
reprints Reprints and permission information is available at http://www.nature.com/
reprints B.K., P.C., and P.Z. conceived the ideas behind this work. J.C., A.N., and B.V. developed
the theory for the partial transpose moment inequalities presented in this work. R.K. developed the statistical analysis related to the classical shadow formalism in the
case of a drifting source. A.E., B.V., and M.D. developed the theory, performed the
numerical simulations and the analysis of the experimental data for the application to
quench dynamics. V.V. and M.D. performed the numerical simulations for the XXZ
spin chain. C.K. performed the numerical simulations for the Fibonacci chain. All
authors discussed the results and contributed to the writing of the final manuscript. J. C. and A.N. contributed equally to this work. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims
in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made. The images or other third party
material in this article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not included in the
article’s Creative Commons license and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly
from the copyright holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/. ACKNOWLEDGEMENTS 34. Vitale, V. et al. Symmetry-resolved dynamical purification in synthetic quantum
matter. https://arxiv.org/abs/2101.07814 (2021). We would like to thank Ion Nechita for pointing out the Stieltjes moment problem to
us. J.C., B.K., and A.N. acknowledge financial support from the Austrian Science Fund
(FWF) stand-alone project: P32273-N27, the FWF: FG-5, and the SFB BeyondC. BV
acknowledges funding from the Austrian Science Foundation (FWF, P 32597 N), and
the French National Research Agency (ANR-20-CE47-0005, JCJC project QRand). The
work of MD and VV is partly supported by the ERC under grant number 758329
(AGEnTh), by the MIUR Programme FARE (MEPH), and has received funding from the
European Union’s Horizon 2020 research and innovation programme under grant
agreement No. 817482 (Pasquans). P.C. acknowledges support from ERC under
Consolidator grant number 771536 (NEMO). C.K., A.E., and P.Z. acknowledge support
by European Union’s Horizon 2020 research and innovation programme under Grant
Agreement No. 817482 (Pasquans) and Simons Collaboration on Ultra-Quantum
Matter, which is a grant from the Simons Foundation (651440, P.Z.). 35. De Las Cuevas, G., Fritz, T. & Netzer, T. Optimal bounds on the positivity of a
matrix from a few moments. Commun. Math. Phys. 375, 105–126 (2020). 36. Yu, X.-D., Imai, S. & Gühne, O. Optimal entanglement certification from moments
of the partial transpose. Phys. Rev. Lett. 127, 060504 (2021). of the partial transpose. Phys. Rev. Lett. 127, 060504 (2021). 37. Goldstein, M. & Sela, E. Symmetry-resolved entanglement in many-body systems. Phys. Rev. Lett. 120, 200602 (2018). 38. Xavier, J. C., Alcaraz, F. C. & Sierra, G. Equipartition of the entanglement entropy. Phys. Rev. B 98, 041106 (2018). 39. Feldman, N. & Goldstein, M. Dynamics of charge-resolved entanglement after a
local quench. Phys. Rev. B 100, 235146 (2019). 40. Bonsignori, R., Ruggiero, P. & Calabrese, P. Symmetry resolved entanglement in
free fermionic systems. J. Phys. A: Math. Theor. 52, 475302(23) (2019). Published in partnership with The University of New South Wales Published in partnership with The University of New South Wales npj Quantum Information (2021) 152 AUTHOR CONTRIBUTIONS Reprints and permission information is available at http://www.nature.com/
reprints COMPETING INTERESTS The authors declare no competing interests. npj Quantum Information (2021) 152 Published in partnership with The University of New South Wales Published in partnership with The University of New South Wales Correspondence and requests for materials should be addressed to Antoine Neven
or Jose Carrasco. ADDITIONAL INFORMATION Supplementary information The online version contains supplementary material
available at https://doi.org/10.1038/s41534-021-00487-y. Correspondence and requests for materials should be addressed to Antoine Neven
or Jose Carrasco. © The Author(s) 2021 npj Quantum Information (2021) 152
|
https://openalex.org/W3116319841
|
https://digital.library.adelaide.edu.au/dspace/bitstream/2440/134683/2/hdl_134683.pdf
|
English
| null |
Periscope Proteins are variable length regulators of bacterial cell surface interactions
|
bioRxiv (Cold Spring Harbor Laboratory)
| 2,020
|
cc-by
| 11,890
|
protein structure | bacterial surface proteins | multidomain proteins protein structure | bacterial surface proteins | multidomain proteins B B
acteria encounter complex and dynamic environments, in-
cluding within human hosts, and have thus evolved various
mechanisms that enable a rapid response for survival within, and
exploitation of, new conditions. In addition to classical control by
regulation of gene expression, bacteria exploit mechanisms that
give rise to random variation to facilitate adaptation [e.g., phase
and antigenic variation (1)]. In Gram-positive and Gram-negative
human pathogens, DNA inversions (2, 3), homologous recombi-
nation (4), DNA methylation (1), and promoter sequence poly-
morphisms (5) govern changes in bacterial surface components,
including capsular polysaccharide and protein adhesins, which can
impact bacterial survival and virulence in the host (1, 6). Many of
these mechanisms are very well studied and widespread across
bacteria. PNAS 2021 Vol. 118 No. 23 e2101349118 Periscope Proteins are variable-length regulators of
bacterial cell surface interactions Fiona Whelana,1,2, Aleix Lafitab,1, James Gilburta,1, Clément Déguta,1, Samuel C. Griffithsa,1, Huw T. Jenkinsc,
Alexander N. St Johnd, Emanuele Pacid, James W. B. Moira, Michael J. Plevina, Christoph G. Baumanna,
Alex Batemanb,3, and Jennifer R. Pottsa,3,4 aDepartment of Biology, The University of York, YO10 5DD York, United Kingdom; bEuropean Molecular Biology Laboratory, European Bioinformatics
Institute, Wellcome Genome Campus, CB10 1SD, United Kingdom; cDepartment of Chemistry, The University of York, YO10 5DD York, United Kingdom;
and dAstbury Centre for Structural Molecular Biology, The University of Leeds, LS2 9JT Leeds, United Kingdom Aap, and Rib contain N-terminal host ligand binding domains
and C-terminal wall attachment motifs; thus our recent demon-
stration that the repetitive regions of both SasG (16) and Rib
(17) form unusual highly elongated rods suggests that host-
colonization domains will be projected differing distances from
the bacterial surface. Changes at the cell surface enable bacteria to survive in dynamic
environments, such as diverse niches of the human host. Here, we
reveal “Periscope Proteins” as a widespread mechanism of bacte-
rial surface alteration mediated through protein length variation. Tandem arrays of highly similar folded domains can form an elon-
gated rod-like structure; thus, variation in the number of domains
determines how far an N-terminal host ligand binding domain
projects from the cell surface. Supported by newly available long-
read genome sequencing data, we propose that this class could
contain over 50 distinct proteins, including those implicated in host
colonization and biofilm formation by human pathogens. In large
multidomain proteins, sequence divergence between adjacent do-
mains appears to reduce interdomain misfolding. Periscope Proteins
break this “rule,” suggesting that their length variability plays an
important role in regulating bacterial interactions with host sur-
faces, other bacteria, and the immune system. Here, we show that repeat number variation in predicted
bacterial surface proteins is more widespread and we characterize
a third rod-like repetitive region in the Streptococcus gordonii
protein (Sgo_0707) formed by tandem array of Streptococcal High
Identity Repeats in Tandem (SHIRT) domains. Thus, we propose
a growing class of “Periscope Proteins,” in which long, highly
similar DNA repeats facilitate expression of surface protein stalks
of variable length. This mechanism could enable changes in re-
sponse to selection pressures and confer key advantages to the
organism that include evasion of the host immune system (8) and
regulation of surface interactions (11) involved in biofilm forma-
tion and host colonization. Significance The structure of single and tandem SHIRT domains from the
streptococcal surface protein Sgo_0707 were determined. In
conjunction with biophysics and molecular dynamics simula-
tions, the results show that the observed gene length variation
would result in differential projection of the host ligand bind-
ing domain on the bacterial cell surface. An analysis of long-
read DNA sequence data reveals many other repetitive bacte-
rial surface proteins that appear to undergo gene length vari-
ation. We propose that these variable-length “Periscope
Proteins” represent an important mechanism of bacterial cell
surface modification with potential roles in infection and
immune evasion. A less well-studied mechanism is length variation in bacterial
surface proteins. Variability in the number of sequence repeats
in the Rib domain (7)–containing proteins on the surface of
Group B streptococci has been linked to pathogenicity and im-
mune evasion (8). The repetitive regions of the Staphylococcus
aureus surface protein G (SasG) (9) and Staphylococcus epi-
dermidis SasG homolog, Aap (10), also demonstrate sequence
repeat number variability. In SasG, this variability regulates li-
gand binding by other bacterial proteins in vitro (11) in a process
that has been proposed to enable bacterial dissemination in the
host. Variations in repeat number have also been noted in the
biofilm forming proteins Esp from Enterococcus faecalis (12)
and, more recently, CdrA from Pseudomonas aeruiginosa (13). High DNA sequence identity in the genes that encode these
proteins is likely to facilitate intragenic recombination events
that would lead to repeat number variation (14) and, in turn, to
protein sequence repetition. However, such sequence repetition
is usually highly disfavored in large multidomain proteins (15), so
its existence in these bacterial surface proteins suggests that
protein length variation provides an evolutionary benefit. SasG, Author contributions: F.W., A.L., J.G., C.D., S.C.G., A.N.S.J., E.P., M.J.P., C.G.B., A.B., and
J.R.P. designed research; F.W., A.L., J.G., C.D., S.C.G., H.T.J., A.N.S.J., E.P., C.G.B., A.B., and
J.R.P. performed research; F.W., A.L., J.G., C.D., S.C.G., H.T.J., A.N.S.J., E.P., M.J.P., C.G.B.,
A.B., and J.R.P. analyzed data; and F.W., A.L., J.G., C.D., S.C.G., H.T.J., A.N.S.J., E.P.,
J.W.B.M., M.J.P., C.G.B., A.B., and J.R.P. wrote the paper. The authors declare no competing interest. This article is a PNAS Direct Submission. This article is a PNAS Direct Submission. This open access article is distributed under Creative Commons Attribution License 4.0
(CC BY). 1F.W., A.L., J.G., C.D., and S.C.G. contributed equally to this work. 1F.W., A.L., J.G., C.D., and S.C.G. Results Defining the Structural Repeats of Sgo_0707 from S. gordonii. Having
revealed the unusual repetitive, rod-like characteristics of both
SasG (16) and Rib (17) in our previous studies, we used bio-
informatic approaches to search for other cell-wall attached bacte-
rial proteins with similar domain architectures. A8AW49 (herein
Sgo_0707) encoded by the gene Sgo_0707 from S. gordonii (Fig. 1A)
has a C-terminal wall attachment motif, homologs with repeat
number variation, and a structurally defined two-domain N termi-
nus (N1-N2; residues 36 through 458, Protein Data Bank [PDB]:
4igb) that is proposed to be involved in collagen binding (18). As the
repeats had no Pfam definition, we called the putative domain
“SHIRT” (Fig. 1B). S. gordonii is a member of the S. sanguinis
group of viridans streptococci (19) and is a common colonizer of the
oral cavity. It is a pioneer organism in the establishment of dental
plaque (20) and also implicated in infective endocarditis (21). Tandemly Arrayed Sgo_0707 SHIRT Domains Form an Extended
Rod-like Structure. A tandem domain construct (Sgo_R3-4; resi-
dues 621 through 789) was crystallized, and the structure was
solved via MR using the ΔN-Sgo_R2 model; data collection and
refinement statistics are summarized in Table 1. The structure
(Fig. 2A) reveals two complete domains with a very short (Pro-
Ala-Pro) linker (SI Appendix, Fig. S1 C and D); the structure of
Sgo_R3-4 is ordered throughout (residues 623 through 789). Each domain adopts the SHIRT fold, and the interdomain in-
terface is limited; this was confirmed by comparing the Tm of
Sgo_R3 (75.7 °C) and Sgo_R3-4 (76.6 °C; SI Appendix, Fig. S3). The similarity of unfolding curves for single and double SHIRT
constructs suggests that the two domains in the tandem construct
unfold independently. Small angle X-ray scattering (SAXS)
analysis substantiates the anisotropic head-to-tail domain ar-
rangement in solution (SI Appendix, Fig. S4A). Notably, there is
a significant twist between domains when viewed along the long
axis of the molecule. p q
(
)
p
(
)
Defining the structural, rather than sequence, repeat bound-
aries in repetitive bacterial proteins is challenging. For Sgo_0707
the T-REKS server (22) predicts a repeat frame of 460 through
543 and 13 repeats of 84 through 90 residues. A construct based
on the second repeat (residues 544 through 627; ΔN-Sgo_R2) was
folded (SI Appendix, Fig. S1A) and solved to 0.95 Å resolution using
X-ray crystallography (SI Appendix, Fig. Significance contributed equally to this work. 2Present address: Department of Molecular and Biomedical Science, The University of
Adelaide, Adelaide, South Australia, 5005, Australia. 3To whom correspondence may be addressed. Email: agb@ebi.ac.uk or jennifer.potts@
sydney.edu.au. 3To whom correspondence may be addressed. Email: agb@ebi.ac.uk or jennifer.potts@
sydney.edu.au. 4Present address: School of Life and Environmental Sciences, University of Sydney,
Camperdown, NSW 2006, Australia. 4Present address: School of Life and Environmental Sciences, University of Sydney,
Camperdown, NSW 2006, Australia. This article contains supporting information online at https://www.pnas.org/lookup/suppl/
doi:10.1073/pnas.2101349118/-/DCSupplemental. This article contains supporting information online at https://www.pnas.org/lookup/suppl/
doi:10.1073/pnas.2101349118/-/DCSupplemental. Published May 31, 2021. https://doi.org/10.1073/pnas.2101349118
|
1 of 10 the data collection and refinement statistics are summarized in
Table 1. The model confirms that SHIRT has an α/β fold orga-
nized around a single α-helix and two distinct β-sheets (Fig. 1B). Fig. 1A shows a schematic of Sgo_0707 based on the structural
boundaries of the repeats; SI Appendix, Fig. S1C shows the high
level of protein sequence identity (82 to 100%) between adjacent
SHIRT domains. Comparison of Sgo_0707 genes from different
bacterial strains, including the homologous protein fibA from S. sanguinis, shows a high variability in the number of SHIRT do-
main repeats forming the stalk, ranging from 3 to 14 copies
(Fig. 1C). SHIRT domains are found in many other proteins,
often in tandem array (SI Appendix, Fig. S2). Results A
B
SGO_0707
HMPREF0847_01079
TZ86_00090
RN86_10690
WH25_04385
UA00_01863
AXF18_01495
SORDD15_01302
HMPREF1047_1371
TZ95_01757
V470_10525
SOR_0366
HMPREF8578_0049
SSV_0571
HMPREF9382_1128
HMPREF9381_0540
HMPREF9383_1788
HMPREF9395_1672
HMPREF9398_1728
SSA_0684
HMPREF9394_1795
HMPREF9384_1170
HMPREF9397_1121
HMPREF9386_1124
HMPREF9388_1149
S.gordonii str. Challis substr. CH1
S. sp. 2_1_36FAA
S.gordoniiG9B
S.gordoniiKCOM 1506
S.gordoniiIE35
S.gordoniiI141
S. sp. oral taxon 064
S.oralis DD15
S.oralis SK1074
S.oralis subsp.tigurinus
S. sp.VT162
S.oralis Uo5
S.oralisATCC 49296
S.sanguinis2908
S.sanguinisSK115
S.sanguinisSK72
S.sanguinisSK150
S.sanguinisSK1058
S.sanguinisVMC66
S.sanguinisSK36
S.sanguinisSK1057
S.sanguinisSK160
S.sanguinisSK1087
S.sanguinisSK330
S.sanguinisSK353
0
7
14
fibA
C
N
C
N1
N2
L
1
1643
tandem repeat region SGO_0707
HMPREF0847_01079
TZ86_00090
RN86_10690
WH25_04385
UA00_01863
AXF18_01495
SORDD15_01302
HMPREF1047_1371
TZ95_01757
V470_10525
SOR_0366
HMPREF8578_0049
SSV_0571
HMPREF9382_1128
HMPREF9381_0540
HMPREF9383_1788
HMPREF9395_1672
HMPREF9398_1728
SSA_0684
HMPREF9394_1795
HMPREF9384_1170
HMPREF9397_1121
HMPREF9386_1124
HMPREF9388_1149
S.gordonii str. Challis substr. CH1
S. sp. 2_1_36FAA
S.gordoniiG9B
S.gordoniiKCOM 1506
S.gordoniiIE35
S.gordoniiI141
S. sp. oral taxon 064
S.oralis DD15
S.oralis SK1074
S.oralis subsp.tigurinus
S. sp.VT162
S.oralis Uo5
S.oralisATCC 49296
S.sanguinis2908
S.sanguinisSK115
S.sanguinisSK72
S.sanguinisSK150
S.sanguinisSK1058
S.sanguinisVMC66
S.sanguinisSK36
S.sanguinisSK1057
S.sanguinisSK160
S.sanguinisSK1087
S.sanguinisSK330
S.sanguinisSK353
0
7
14
fibA
C
L
1643 A
B
f
N
C
N1
N2
L
1
tandem repeat region A Downloaded from https://www.pnas.org by UNIVERSITY IF ADELAIDE LIBRARIES on M B Fig. 1. Close homologs of A8AW49 contain variable numbers of repeats that each form the “SHIRT” fold. (A) Schematic of Sgo_0707 showing the N-terminal
adhesin domains N1 and N2, tandem repeat (SHIRT) domains (red) and C-terminal LPXTG cell-wall crosslinking motif (black box “L”). (B) Structure of Sgo_R10
and topology; sheets S1 (red) and S2 (blue) are highlighted in boxes. (C) A phylogenetic tree of Sgo_0707 homologs (>70% identity) identified using PHMMER
(61) from ENSEMBL bacterial genomes using the N-terminal domain sequence (residue range 1 through 450 for Sgo_0707) and containing the LPXTG cell wall
anchor motif at their C termini. The number of SHIRT domain repeats in the stalk of each protein is shown as a bar plot with scale bar below. Results S1A, Inset and Table 1)
utilizing ab initio molecular replacement (MR) with ideal fragments
(23). Based on the significant truncation of the N-terminal β-strand
(SI Appendix, Fig. S1A, Inset), we hypothesized that shifting the
frame of the repeat by seven residues toward the N terminus of
Sgo_0707 would complete the fold. Sgo_R3 (residues 621 through
705) and Sgo_R10 (residues 1211 through 1299) based on this new
definition were thus expressed and purified. They were found to
have significantly higher melting temperatures (Tm) than the
N-terminally truncated Sgo_R2 (ΔN-Sgo_R2, 55.9 °C; Sgo_R3,
75.7 °C; Sgo_R10, 75.9 °C; SI Appendix, Fig. S1B). Å Downloaded from https://www.pnas.org by UNIVERSITY IF ADELAIDE LIBRARIES on March 27, 2022 from IP address 129.127.145.230. Molecular dynamics (MD) simulations of the Sgo_R3-4 con-
struct show that individual domains are particularly stable (rmsd
<1.5 Å for Cα atoms) over the length of the trajectory (0.8 μs);
their individual length is conserved during the simulation and the ;
g _
,
;
pp
,
g
)
The structure of Sgo_R10 (Fig. 1B) was solved at 0.82 Å res-
olution using MR with the structure of ΔN-Sgo_R2 for phasing; A
B
SGO_0707
HMPREF0847_01079
TZ86_00090
RN86_10690
WH25_04385
UA00_01863
AXF18_01495
SORDD15_01302
HMPREF1047_1371
TZ95_01757
V470_10525
SOR_0366
HMPREF8578_0049
SSV_0571
HMPREF9382_1128
HMPREF9381_0540
HMPREF9383_1788
HMPREF9395_1672
HMPREF9398_1728
SSA_0684
HMPREF9394_1795
HMPREF9384_1170
HMPREF9397_1121
HMPREF9386_1124
HMPREF9388_1149
S.gordonii str. Challis substr. CH1
S. sp. 2_1_36FAA
S.gordoniiG9B
S.gordoniiKCOM 1506
S.gordoniiIE35
S.gordoniiI141
S. sp. oral taxon 064
S.oralis DD15
S.oralis SK1074
S.oralis subsp.tigurinus
S. sp.VT162
S.oralis Uo5
S.oralisATCC 49296
S.sanguinis2908
S.sanguinisSK115
S.sanguinisSK72
S.sanguinisSK150
S.sanguinisSK1058
S.sanguinisVMC66
S.sanguinisSK36
S.sanguinisSK1057
S.sanguinisSK160
S.sanguinisSK1087
S.sanguinisSK330
S.sanguinisSK353
0
7
14
fibA
C
N
C
N1
N2
L
1
1643
tandem repeat region
Fig. 1. Close homologs of A8AW49 contain variable numbers of repeats that each form the “SHIRT” fold. (A) Schematic of Sgo_0707 showing the N-terminal
adhesin domains N1 and N2, tandem repeat (SHIRT) domains (red) and C-terminal LPXTG cell-wall crosslinking motif (black box “L”). (B) Structure of Sgo_R10
and topology; sheets S1 (red) and S2 (blue) are highlighted in boxes. (C) A phylogenetic tree of Sgo_0707 homologs (>70% identity) identified using PHMMER
(61) from ENSEMBL bacterial genomes using the N-terminal domain sequence (residue range 1 through 450 for Sgo_0707) and containing the LPXTG cell wall
anchor motif at their C termini. The number of SHIRT domain repeats in the stalk of each protein is shown as a bar plot with scale bar below. PNAS
|
3 of 10
https://doi.org/10.1073/pnas.2101349118 2 of 10
|
PNAS Taken together, these data are consistent fits of χ2 = 1.1 and χ2 = 1.2, respectively; SI Appendix, Fig. S4 A and
B and Materials and Methods). Analysis of the data for Sgo_R3-4
and Sgo_R2-8 results in equal Porod exponents (1.2, Fig. 3A), as
well as similar radii of gyration of a cross-section (Rg
c) values (R3
through 4 = 6.4 ± 0.0 Å; R2 through 8 = 7.0 ± 0.0 Å; Fig. 3 B
and C). Consistent with the models (Fig. 2D), the Dmax that was
determined using SAXS scales with the number of domains;
Sgo_R3-4 exhibits a Dmax of 107 Å and Sgo_R2-8 has a Dmax of
371 Å (SI Appendix, Fig. S4 A and B). Therefore, while dis-
playing a much larger intraparticle maximum dimension, the
Sgo_R2-8 construct has a comparable shape and Rg
c to Sgo_R3-4. length of Sgo_R3-4 fluctuates only moderately around 97 Å
(Fig. 2B). The distributions of α, β, γ interdomain (17) angles (SI
Appendix, Fig. S5A) observed in the simulations of the Sgo_R3-4
construct (Fig. 2C) were used to generate models of longer
constructs (Fig. 2D). The radius of gyration (Rg) of the simulated
constructs increases following the relation Rg ∝Nν, where ν is the
Flory exponent and describes the increase in size of a polymer
(protein) made of N monomers (amino acids). Such an exponent
is ∼0.6 for denatured proteins and ∼0.4 for folded ones (24). Polymers formed of sequential SHIRT domains are highly ex-
tended; Rg scales with the number of domains (or equivalently,
amino acids) with an exponent ν ∼0.8 (Fig. 2E), which is re-
markable given the width of the distribution of the angles de-
scribing the mutual orientation of adjacent domains (Fig. 2C). h
l
l
ll
d length of Sgo_R3-4 fluctuates only moderately around 97 Å
(Fig. 2B). The distributions of α, β, γ interdomain (17) angles (SI
Appendix, Fig. S5A) observed in the simulations of the Sgo_R3-4
construct (Fig. 2C) were used to generate models of longer
constructs (Fig. 2D). The radius of gyration (Rg) of the simulated
constructs increases following the relation Rg ∝Nν, where ν is the
Flory exponent and describes the increase in size of a polymer
(protein) made of N monomers (amino acids). Such an exponent
is ∼0.6 for denatured proteins and ∼0.4 for folded ones (24). Whelan et al.
Periscope Proteins are variable-length regulators of bacterial cell surface interactions 2 of 10
|
PNAS Therefore, while dis-
playing a much larger intraparticle maximum dimension, the
Sgo_R2-8 construct has a comparable shape and Rg
c to Sgo_R3-4. In our previous study (17), we observed elongation in two-
domain Rib constructs with rotation of angle α, while angles β
and γ were smaller. We conducted the same analysis for Sgo_0707
constructs, based on fitting to our experimental SAXS data
(SI Appendix, Fig. S5). As with Rib, MD simulations of the two-
domain Sgo_R3-4 construct show that a range of α angles give
good fits to the observed SAXS data, while β and γ are restricted
to a narrow range, consistent with an elongated conformation
(SI Appendix, Fig. S5B). For fitting of MD simulations of the
7-domain Sgo_R2-8 construct to SAXS data, we observed that
longer end-to-end distances improved the quality of the fit
(SI Appendix, Fig. S5C). Taken together, these data are consistent fits of χ2 = 1.1 and χ2 = 1.2, respectively; SI Appendix, Fig. S4 A and
B and Materials and Methods). Analysis of the data for Sgo_R3-4
and Sgo_R2-8 results in equal Porod exponents (1.2, Fig. 3A), as
well as similar radii of gyration of a cross-section (Rg
c) values (R3
through 4 = 6.4 ± 0.0 Å; R2 through 8 = 7.0 ± 0.0 Å; Fig. 3 B
and C). Consistent with the models (Fig. 2D), the Dmax that was
determined using SAXS scales with the number of domains;
Sgo_R3-4 exhibits a Dmax of 107 Å and Sgo_R2-8 has a Dmax of
371 Å (SI Appendix, Fig. S4 A and B). Therefore, while dis-
playing a much larger intraparticle maximum dimension, the
Sgo_R2-8 construct has a comparable shape and Rg
c to Sgo_R3-4. In our previous study (17), we observed elongation in two-
domain Rib constructs with rotation of angle α, while angles β
and γ were smaller. We conducted the same analysis for Sgo_0707
constructs, based on fitting to our experimental SAXS data
(SI Appendix, Fig. S5). As with Rib, MD simulations of the two-
domain Sgo_R3-4 construct show that a range of α angles give
good fits to the observed SAXS data, while β and γ are restricted
to a narrow range, consistent with an elongated conformation
(SI Appendix, Fig. S5B). For fitting of MD simulations of the
7-domain Sgo_R2-8 construct to SAXS data, we observed that
longer end-to-end distances improved the quality of the fit
(SI Appendix, Fig. S5C). 2 of 10
|
PNAS Whelan et al. Periscope Proteins are variable-length regulators of bacterial cell surface interactions 2 of 10
|
PNAS https://doi.org/10.1073/pnas.2101349118 Table 1. Crystallographic data collection and refinement statistics
ΔN-Sgo_R2
Sgo_R10
Sgo_R3-4
Data collection statistics
Wavelength (Å)
0.85
0.78
0.976
Space group
P212121
P21212
P21
Cell dimensions
a, b, c (Å)
21.4, 40.8, 82.5
65.4, 48.0, 48.4
24.1, 37.2, 101.1
β (°)
90.0
90.0
95.8
Resolution limits (Å)*
82.5–0.95 (0.96–0.95)
38.9–0.82 (0.85–0.82)
37.2–1.35 (1.37–1.35)
No. reflections
Total*
281,952 (12,914)
292,386 (24,481)
157,956 (7,439)
Unique*
44,546 (2,604)
147,184 (11,281)
39,416 (1,911)
Rmerge*, †
0.034 (0.681)
0.045 (2.057)
0.066 (0.960)
Mean I/σI*
24.8 (2.3)
28.3 (0.3)
8.9 (1.4)
Half-set correlation CC (1/2)*
0.999 (0.839)
1.000 (0.239)
0.998 (0.562)
Wilson B-factor (Å2)
8.9
10.8
10.8
Completeness (%)*
95.2 (75.8)
96.8(76.5)
99.9 (99.6)
Redundancy*
6.3 (5.0)
2.0(1.9)
4.0 (3.9)
Refinement statistics
Resolution (Å)*
43.2 (0.95)
38.9 (0.82)
34.9 (1.35)
No. of reflections
Working
42,240
144,297
37,511
Free
2,193
1,454
1,892
Rwork/Rfree (%)
12.5/13.3
14.7/16.2
13.4/17.0
Rms from ideality
Bond length (Å)
0.017
0.017
0.023
Angles (°)
1.96
1.48
2.15
No. of atoms
Protein
661
1,442
1,404
Ligand/ion
—
15
33
Water
104
247
200
Average B (Å2)
Protein
14.9
18.1
18
Ligand/ion
—
19.7
22.4
Water
28.3
27.8
34.0
Ramachandran angles (%)
Favored
98.0
98.15
97.0
Allowed
2.0
1.85
3.0
Outliers
0.0
0.0
0.0
*Values in parentheses are for the highest resolution shell. †Rmerge = ΣjI - <I>j/ΣI. Table 1. Crystallographic data collection and refinement statistics Downloaded from https://www.pnas.org by UNIVERSITY IF ADELAIDE LIBRARIES on March 27, 2022 from IP address 129.127.145.230. *Values in parentheses are for the highest resolution shell. †Rmerge = ΣjI - <I>j/ΣI. fits of χ2 = 1.1 and χ2 = 1.2, respectively; SI Appendix, Fig. S4 A and
B and Materials and Methods). Analysis of the data for Sgo_R3-4
and Sgo_R2-8 results in equal Porod exponents (1.2, Fig. 3A), as
well as similar radii of gyration of a cross-section (Rg
c) values (R3
through 4 = 6.4 ± 0.0 Å; R2 through 8 = 7.0 ± 0.0 Å; Fig. 3 B
and C). Consistent with the models (Fig. 2D), the Dmax that was
determined using SAXS scales with the number of domains;
Sgo_R3-4 exhibits a Dmax of 107 Å and Sgo_R2-8 has a Dmax of
371 Å (SI Appendix, Fig. S4 A and B). 2 of 10
|
PNAS Polymers formed of sequential SHIRT domains are highly ex-
tended; Rg scales with the number of domains (or equivalently,
amino acids) with an exponent ν ∼0.8 (Fig. 2E), which is re-
markable given the width of the distribution of the angles de-
scribing the mutual orientation of adjacent domains (Fig. 2C). To assess the elongation of Sgo_0707 in solution, we collected
SAXS data for constructs comprising two (Sgo_R3-4) and seven
(Sgo_R2-8) tandemly arrayed SHIRT domains (Fig. 3 and SI
Appendix, Fig. S4). Both Sgo_R3-4 and Sgo_R2-8 are mono-
meric in solution, eluting as monodisperse peaks from size ex-
clusion chromatography (SEC) columns (SI Appendix, Fig. S4 A
and B, Inset). Both the crystal structure of Sgo_R3-4 (Fig. 2A)
and an elongated model for Sgo_R2-8 (SI Appendix, Fig. S4B)
are consistent with the SAXS data measured in solution (model:data g _
p
p
g
g _
In our previous study (17), we observed elongation in two-
domain Rib constructs with rotation of angle α, while angles β
and γ were smaller. We conducted the same analysis for Sgo_0707
constructs, based on fitting to our experimental SAXS data
(SI Appendix, Fig. S5). As with Rib, MD simulations of the two-
domain Sgo_R3-4 construct show that a range of α angles give
good fits to the observed SAXS data, while β and γ are restricted
to a narrow range, consistent with an elongated conformation
(SI Appendix, Fig. S5B). For fitting of MD simulations of the
7-domain Sgo_R2-8 construct to SAXS data, we observed that
longer end-to-end distances improved the quality of the fit
(SI Appendix, Fig. S5C). Taken together, these data are consistent To assess the elongation of Sgo_0707 in solution, we collected
SAXS data for constructs comprising two (Sgo_R3-4) and seven
(Sgo_R2-8) tandemly arrayed SHIRT domains (Fig. 3 and SI
Appendix, Fig. S4). Both Sgo_R3-4 and Sgo_R2-8 are mono-
meric in solution, eluting as monodisperse peaks from size ex-
clusion chromatography (SEC) columns (SI Appendix, Fig. S4 A
and B, Inset). Both the crystal structure of Sgo_R3-4 (Fig. 2A)
and an elongated model for Sgo_R2-8 (SI Appendix, Fig. S4B)
are consistent with the SAXS data measured in solution (model:data PNAS
|
3 of 10
https://doi.org/10.1073/pnas.2101349118 PNAS
|
3 of 10
https://doi.org/10.1073/pnas.2101349118 A A B
C
E
D
Fig. 2. 2 of 10
|
PNAS The tandem SHIRT repeat Sgo_R3-4 adopts an anisotropic (head-to-tail) structure with limited interdomain “bend” but significant “twist.” (A) The
structure of Sgo_R3-4 (Left) and connecting short, well-ordered Pro-Ala-Pro interdomain linker; alternate conformers of P704 are included (electron density
2mFo-DFc map contoured at 0.1182 electrons/Å3 blue chickenwire, Right). (B) Frequency of domain lengths and (C) interdomain angles (SI Appendix, Fig. S5A)
over a 0.8 μs all-atom, fully solvated molecular dynamics simulation of Sgo_R3-4 at 303 K. The ends of the domains Sgo_R3, Sgo_R4, and Sgo_R3-4 and linker
are identified by the Cα atoms of residues T623–P704, T707–A789, and T627–A789, respectively; arrows show the value of the distances in the crystal structure. (D) Models of seven tandemly arrayed domains based on the α, β, γ angles in C. (E) Scaling of Rg of simulated model SHIRT constructs with increasing number
of domains (or, equivalently, amino acids) compared with denatured and native proteins (approximated by blue and red lines, respectively). Error bars are SDs
over 100 models generated. by UNIVERSITY IF ADELAIDE LIBRARIES on March 27, 2022 from IP address 129.127.145.230. B
C
E
D B C NIVERSITY IF ADELAIDE LIBRARIES on March 27, 2022 from IP address 129.127.145.230. Downloaded from https://www.pnas.org by UNIVERSITY IF ADELAIDE LIBRARIES on March 27, 2022 from IP address 129.127.145.230. D E Fig. 2. The tandem SHIRT repeat Sgo_R3-4 adopts an anisotropic (head-to-tail) structure with limited interdomain “bend” but significant “twist.” (A) The
structure of Sgo_R3-4 (Left) and connecting short, well-ordered Pro-Ala-Pro interdomain linker; alternate conformers of P704 are included (electron density
2mFo-DFc map contoured at 0.1182 electrons/Å3 blue chickenwire, Right). (B) Frequency of domain lengths and (C) interdomain angles (SI Appendix, Fig. S5A)
over a 0.8 μs all-atom, fully solvated molecular dynamics simulation of Sgo_R3-4 at 303 K. The ends of the domains Sgo_R3, Sgo_R4, and Sgo_R3-4 and linker
are identified by the Cα atoms of residues T623–P704, T707–A789, and T627–A789, respectively; arrows show the value of the distances in the crystal structure. (D) Models of seven tandemly arrayed domains based on the α, β, γ angles in C. (E) Scaling of Rg of simulated model SHIRT constructs with increasing number
of domains (or, equivalently, amino acids) compared with denatured and native proteins (approximated by blue and red lines, respectively). Error bars are SDs
over 100 models generated. with multiple tandemly arrayed SHIRT domains from Sgo_0707
behaving as an elongated, rod-like particle. Downloaded from https://www.pnas.o (Fig. 4 of 10
|
PNAS
https://doi.org/10.1073/pnas.2101349118 Whelan et al.
Periscope Proteins are variable-length regulators of bacterial cell surface interactions 2 of 10
|
PNAS As expected, the distance between the N and C termini of domain
repeats is large, and their termini are oriented close to 180° (3.14
radians; SI Appendix, Table S3), enabling linear arrangements of
tandem domain repeats. In addition, most interdomain linkers are
short (smaller than 5 residues), and some of them are rich in
prolines, conferring additional rigidity to the linear domain ar-
rangements. Some exceptions with short intertermini distance and
small angles, such as LysM and SH3_3 domains, contain longer
and more flexible linkers. Large-Scale Identification of Potential Periscope Proteins in Bacterial
Genomes. We call SasG, Rib, and Sgo_0707 “Periscope Proteins”
due to their having variable-length stalk-like regions that define
the distance that the N-terminal functional domain projects from
the bacterial surface. We conducted an unbiased identification of
other potential Periscope Proteins from the NCTC3000 bacterial
genomes: an ongoing project lead by the Wellcome Sanger Insti-
tute (UK) that provides high quality annotated genome assemblies
for 3,000 bacterial strains from Public Health England’s National
Collection of Type Cultures (NCTC) using the long-read Pac-
Bio technology (https://www.sanger.ac.uk/resources/downloads/
bacteria/nctc/). The use of long-reads spanning whole genes
overcomes the challenges in the assembly of repetitive genes in
bacterial genomes (26), providing reliable numbers of repeats in
the assembled genes and associated proteins and allowing the
identification of length variability in Periscope Proteins. We
identified 1,576 proteins containing long and highly identical re-
peats (i.e., similar to those forming stalks in Periscope Proteins),
out of a total of 2.5 million proteins in the NCTC3000 data set,
from different species including both Gram-positive and Gram-
negative bacteria. Clustering of the repeating sequences reveals
most of them are classified into existing Pfam families (Fig. 4 and
Dataset S1) that correspond to globular domains, as in known We further clustered full-length repetitive proteins and identified
a total of 180 unique groups, 84 of which exhibit repeat number
variation (Dataset S1). Manual inspection confirmed that 56 of
these groups are examples of Periscope Proteins, 30 of which could
be assigned to recognizable gene names. We retrieved known
Periscope Proteins, such as SasG, Rib, and Sgo_0707 (clusters 4, 25,
and 26, respectively), as well as novel ones, for example, CdrA from
Pseudomonas aeruginosa (27), which has been reported recently to
exhibit variability in the size of the repeat region (13), CshA from S. gordonii (28), and SAP077A_019 (UniProt D2JAN8) from S. au-
reus (Fig. 5). 2 of 10
|
PNAS (D) SHRImP-TIRFM determination of the
interdye distances for AF488-labeled Sgo_R2-8S666C/S1086C (Schematic Inset) immobilized on 2 μg/mL (blue) or 20 μg/mL (red) poly-D-lysine-treated quartz
surfaces. Color-coded, dashed lines indicate Gaussian fits to the histograms for the 2 μg/mL (n = 90, R2 = 0.86) and 20 μg/mL (n = 85, R2 = 0.90) poly-D-lysine-
treated quartz surfaces (mean ± SE = 25.5 ± 2.39 nm and 27.8 ± 2.21 nm, respectively). B
C
D A C A B C D D Fig. 3. Tandemly arrayed SHIRT domains form a highly elongated rod. (A) Porod plots from SAXS analysis of Sgo_R3-4 (Top) and Sgo_R2-8 (Bottom). Porod
exponents were calculated from the negative slope of the linear region of the data (R3-4, blue, R2 = 1.0; R2-8, red, R2 = 1.0). (B) Modified Guinier region for
Sgo_R3-4 with cross-sectional radius of gyration (Rg
c, annotated) calculated from fitted region (blue line, R2 = 0.98, q*Rg limits = 0.3 to 1.1). (C) Modified
Guinier region for Sgo_R2-8 with Rg
c calculated from fitted region (red line, R2 = 0.99, q*Rg limits = 0.3 to 1.0. (D) SHRImP-TIRFM determination of the
interdye distances for AF488-labeled Sgo_R2-8S666C/S1086C (Schematic Inset) immobilized on 2 μg/mL (blue) or 20 μg/mL (red) poly-D-lysine-treated quartz
surfaces. Color-coded, dashed lines indicate Gaussian fits to the histograms for the 2 μg/mL (n = 90, R2 = 0.86) and 20 μg/mL (n = 85, R2 = 0.90) poly-D-lysine-
treated quartz surfaces (mean ± SE = 25.5 ± 2.39 nm and 27.8 ± 2.21 nm, respectively). Periscope Proteins, and from a wide range of secondary structure
and fold types. measurements were performed in the presence of 100-fold molar
excess of unlabeled Sgo_R2-8 (termed “blocking” protein). This
implies the rod-like conformation of Sgo_R2-8 protein is mal-
leable, adopting, compared with solution, a slightly more elon-
gated state when on a surface which is accentuated (Δx = 4 to 5
nm) at high-protein concentrations that presumably favor lateral
protein–protein interactions. Most of the domain families we find correspond to
Immunoglobulin-like beta sandwich folds in the E-set clan (Pfam:
CL0159), but we also find domain families with beta grasp folds in
the Ubiquitin clan (Pfam: CL0072) and three-helix bundles in the
bacterial immunoglobulin/albumin-binding clan (Pfam:CL0598). 2 of 10
|
PNAS 3 D, Inset) and simulating the increased volume accessible
to each dye due to the chemical linker (25). AF488-labeled
Sgo_R2-8S666C/S1086C was imaged using total internal reflection
fluorescence microscopy (TIRFM) using two different poly-D-
lysine concentrations to coat the slides. Both surface treatments
effectively
immobilized
AF488-labeled
Sgo_R2-8S666C/S1086C,
and each SHRImP-TIRFM histogram had a single peak consis-
tent with monodispersity. The measured interdye distances were
25.5 nm and 27.8 nm (Fig. 3D) with 2 μg/mL and 20 μg/mL
poly-D-lysine
concentration, respectively, and increased to
29.8 nm and 33.0 nm, (SI Appendix, Fig. S6B), respectively, when (Fig. 3 D, Inset) and simulating the increased volume accessible
to each dye due to the chemical linker (25). AF488-labeled
Sgo_R2-8S666C/S1086C was imaged using total internal reflection p
To further assess the elongation of multidomain SHIRT
constructs, we used a high-resolution single-molecule technique
[SHRImP (16)] to measure the intramolecular distance between
two Alexa Fluor 488 (AF488) dyes covalently attached to cys-
teine residues engineered at specific sites in Sgo_R2-8 (S666C
and S1086C). If the extended interdomain topology observed in
the two-domain construct is maintained, a distance of 24.1 nm
between the mean dye positions (SI Appendix, Fig. S6A) is pre-
dicted by using the distance between the two cysteine residues fluorescence microscopy (TIRFM) using two different poly-D-
lysine concentrations to coat the slides. Both surface treatments
effectively
immobilized
AF488-labeled
Sgo_R2-8S666C/S1086C,
and each SHRImP-TIRFM histogram had a single peak consis-
tent with monodispersity. The measured interdye distances were
25.5 nm and 27.8 nm (Fig. 3D) with 2 μg/mL and 20 μg/mL
poly-D-lysine
concentration, respectively, and increased to
29.8 nm and 33.0 nm, (SI Appendix, Fig. S6B), respectively, when Whelan et al. Whelan et al. Periscope Proteins are variable-length regulators of bacterial cell surface interactions A
B
C
D
Fig. 3. Tandemly arrayed SHIRT domains form a highly elongated rod. (A) Porod plots from SAXS analysis of Sgo_R3-4 (Top) and Sgo_R2-8 (Bottom). Porod
exponents were calculated from the negative slope of the linear region of the data (R3-4, blue, R2 = 1.0; R2-8, red, R2 = 1.0). (B) Modified Guinier region for
Sgo_R3-4 with cross-sectional radius of gyration (Rg
c, annotated) calculated from fitted region (blue line, R2 = 0.98, q*Rg limits = 0.3 to 1.1). (C) Modified
Guinier region for Sgo_R2-8 with Rg
c calculated from fitted region (red line, R2 = 0.99, q*Rg limits = 0.3 to 1.0. PNAS
|
5 of 10
https://doi.org/10.1073/pnas.2101349118 Whelan et al.
Periscope Proteins are variable-length regulators of bacterial cell surface interactions Discussion The work presented in this study identifies a class of bacterial
surface proteins; through length variation, these proteins are
likely to modulate surface interactions. Sgo0707 is the third ex-
ample of this class for which we have characterized both isolated
domains and tandem arrays. SasG, the first example (16), is
composed of arrays of E and G5 domains. E domains are unstable
in the absence of G5 domains. The interdomain interfaces (E-G5
and G5-E) make very significant contributions to the overall sta-
bility of the tandem array of E-G5 domains. G5 domains also have
an unusual flat structure rather than a “typical” hydrophobic core. The second example, Rib, is different. We reported the 3D
structure of a Rib domain (17), which appears to be an interesting
example of domain atrophy from the common Ig-fold. In Rib,
there is no evidence for a significant interdomain interface. SHIRT
(presented here) is different again. We report the 3D structure of a
SHIRT domain which, while also containing β-sheets, has a to-
pology with no obvious relationship to the Ig fold. Like Rib, there
is no evidence for significant interdomain interfaces in tandem
arrays of SHIRT domains. Rib and Sgo_0707 have relatively short
interdomain linkers between Rib and SHIRT domains, respec-
tively, and/or linkers that contain proline residues; both features To classify this diverse family of proteins with a memorable
analogy, we call them “Periscope Proteins.” We propose that high-
sequence similarity at the DNA level enables intragenic recom-
bination events (14) (and thus loss or gain of repeats) with se-
lection pressures resulting in enrichment of bacteria with shorter
or longer proteins. Periscope Proteins include proteins implicated
in biofilm formation in both Gram-positive and Gram-negative
bacteria, for example, SAP077A_019 from S. aureus and related
proteins in S. epidermidis, Staphylococcus xylosus, Staphylococcus
capitis, Staphylococcus simiae, and Staphylococcus warneri, the
Enterococcus faecalis protein Esp (12) and related proteins, and
CdrA from Pseudomonas aeruginosa (13). We expect the total
number of Periscope Proteins across bacterial species to be much
larger and more widespread than reported in this study. Our
knowledge of Periscope Proteins will increase as more long-read
sequence data for a diverse set of strains not included in the
NCTC3000 collection become available. 2 of 10
|
PNAS The majority of confirmed Periscope Proteins (40/56)
have at least one Gene Ontology (GO) location term assigned, all
of them associating to the cell-wall or bacterial membrane. g
We further find that the number of repeats per gene can be
large [>40 in one case (29); Fig. 6], spanning thousands of amino PNAS
|
5 of 10
https://doi.org/10.1073/pnas.2101349118 Fig. 4. Sequence clustering of long tandem highly identical repeats identified across proteins of the NCTC3000 genomes. The largest clusters are annotated
using Pfam (Pfam IDs are found in Dataset S1). Clusters with unknown Pfam classification are marked with “?”. New protein domain families built from
sequence clusters into Pfam are highlighted in blue. Downloaded from https://www.pnas.org by UNIVERSITY IF ADELAIDE LIBRARIES on March 27, 2022 from IP address 129.127.145.230. Fig. 4. Sequence clustering of long tandem highly identical repeats identified across proteins of the NCTC3000 genomes. The largest clusters are annotated
using Pfam (Pfam IDs are found in Dataset S1). Clusters with unknown Pfam classification are marked with “?”. New protein domain families built from
sequence clusters into Pfam are highlighted in blue. are likely to contribute to the observed elongation of the tandemly
arrayed domains. In addition, due to requirements of forming an
elongated array, we predict that, as observed for SasG [PDB: 3TIP;
(30)], Rib [PDB: 6S5X; (17)], and Sgo_0707 (Fig. 1), Periscope
Proteins will often contain domain folds that locate the N and C
termini of the protein sequence at either end of the fold (SI Ap-
pendix, Table S3). Thus, in Periscope Proteins, we find that dif-
ferent types of domains, when arrayed in tandem, can form an
elongated rod that typically serves to project a functional domain
distal to the cell surface. acids, and that variation in the number of stalk repeats can be
extreme, increasing the length of the protein by an order of
magnitude in some cases. Importantly, we found the most ex-
treme repeat numbers and length variation in proteins with the
highest repeat identity (Fig. 6). Furthermore, as shown in Fig. 5,
the number of repeats changes drastically even between similar
bacterial strains, suggesting a rapid evolutionary rate of repeat
number change in Periscope Proteins. Whelan et al.
Periscope Proteins are variable-length regulators of bacterial cell surface interactions 6 of 10
|
PNAS
https://doi.org/10.1073/pnas.2101349118 Downloaded from https://www.pnas.org by UNIVERSITY IF ADELAIDE LIBRARIES on March 27, 2022 from IP address 129.127.145.230. Downloaded from https://www.pnas.org by UNIVERSITY IF ADELAIDE LIBRARIES on March 27, 2022 from IP address 129.127.145.230. D 12
13
13
10
12
9
12
23
21
19
20
18
18
SGO_0707
CshA Fig. 5. Variation of stalk region repeat numbers in Periscope Proteins. Phylogenetic trees of (A) Staphylococcus aureus, (B) Pseudomonas aeruginosa, (C)
various streptococcal species including S. agalactiae and S. pyogenes, and (D) S. gordonii genomes in the NCTC3000 collection, mapped to the number of
repeats in stalk regions in Periscope genes; respectively, SasG with G5/E repeats, SAP077A_019 with Big_6 domain repeats, CdrA with MBG_2 domain repeats,
surface protein Rib with Rib domain repeats, Sgo_0707 homolog containing SHIRT domain repeats, and surface adhesin CshA with ∼100 residue globular
domain repeats of a new Pfam family (CshA_repeat). Large multidomain proteins usually have <50% sequence iden-
tity between adjacent domains (15), which has been proposed as an
evolutionary strategy to minimize interdomain misfolding (31–33). The existence of Periscope Proteins, which contain tandemly
arrayed structured domains with high sequence similarity, con-
founds this observation. Considering this potential “misfolding
problem,” the existence of Periscope Proteins suggests that length
variation (driven by the requirement for high identity at the DNA
level and thus high protein similarity) is functionally important and
confers a significant advantage to the organism. Simply expressing a
Periscope Protein (an extreme form of length variation) results in
changes in ligand binding by other bacterial surface proteins
through spatial competition. For example, expression of Pls (34)
and SasG significantly inhibits binding of S. aureus to the human
plasma proteins fibronectin and fibrinogen, respectively; the latter
resulting in altered bacterial colony morphology (35). In both cases,
it is suggested that the elongated Periscope Protein is blocking
interactions of the host protein with bacterial proteins closer to the
cell surface. There is also more direct evidence for the role of
length variation in Periscope Proteins under selection pressure. For
example, in mice immunized with anti-serum raised against the nine-repeat alpha C, GBS expressing alpha C with one Rib repeat
were 100-fold more pathogenic than GBS expressing alpha C with
nine Rib repeats (36). It was proposed that the shorter protein is
not recognized because it is less exposed. When SasG was
expressed on S. aureus with differing numbers of repeats, only the
longer variants blocked binding of other bacterial surface proteins
to their target (11). Downloaded from https://www.pnas.org by UNIVERSITY IF ADELAIDE LIBRARIES on March 27, 2022 from IP address 129.127.145.230. These authors noted that the ability of a bac-
terium to detach from a ligand could be important for dissemina-
tion in the host. Finally, deletion of CdrA had the most deleterious
effect on biofilm formation for P. aeruginosa strains expressing the
longest CdrA variants (13). Rib repeat number variation within a
single strain can be observed within 24 h following inoculation of a
mouse with group B streptococci (37), suggesting the potential for
dynamic regulation by Periscope Proteins on a physiologically rel-
evant timescale. In summary, surface variation through the Peri-
scope Protein class appears to be a highly distinctive way to alter a
variety of interactions linked to colonization and infection. Whelan et al.
Periscope Proteins are variable-length regulators of bacterial cell surface interactions Discussion We suggest that length
variation is the main function of the tandemly arrayed regions of
Periscope Proteins; however, as some of the domains found in such
arrays (such as LysM domains) have had a ligand function assigned,
they could play an additional functional role in some contexts. ERS825155
ERS807423
ERS825170
ERS636088
ERS659566
ERS798847
ERS798855
ERS836414
ERS846851
ERS950459
ERS825165
ERS811740
ERS846853
ERS798850
ERS836417
ERS811727
ERS825154
ERS806200
ERS798851
ERS811741
ERS811724
ERS1066616
ERS798864
ERS980038
ERS825159
ERS811728
ERS654926
ERS819824
ERS636089
ERS636087
ERS811722
ERS846858
ERS798844
ERS825153
ERS654930
ERS846852
ERS846849
8
7
8
7
7
5
8
4
8
7
7
9
8
8
10
6
6
7
9
9
8
4
9
8
8
11
8
8
8
8
7
4
8
8
SasG
ERS1022024
ERS554120
ERS1106926
ERS1117683
ERS647587
ERS1043811
ERS1018586
ERS1031246
ERS666353
ERS666352
ERS1018592
ERS1043809
ERS1018584
ERS1022026
ERS1022025
ERS1018585
ERS1022027
ERS554121
ERS1058909
ERS666355
ERS666351
ERS1018587
ERS1043810
ERS1110714
12
15
17
13
18
16
16
15
15
16
15
15
17
11
15
15
12
15
15
14
16
17
18
15
CdrA
16
ERS970082
ERS1037087
ERS956180
ERS1058928
ERS956184
ERS1037084
ERS1043815
ERS950458
ERS956188
12
6
2
10
8
14
2
6
Rib
P. aeruginosa
3
13
7
21
3
13
SAP077A_019
ERS950456
ERS970085
ERS956189
ERS1022031
ERS901511
ERS956190
ERS1022030
12
13
13
10
12
9
12
23
21
19
20
18
18
S. gordonii
SGO_0707
CshA
B
Streptococcus sp. A
C
D
S. aureus
Fig. 5. Variation of stalk region repeat numbers in Periscope Proteins. Phylogenetic trees of (A) Staphylococcus aureus, (B) Pseudomonas aeruginosa, (C)
various streptococcal species including S. agalactiae and S. pyogenes, and (D) S. gordonii genomes in the NCTC3000 collection, mapped to the number of
repeats in stalk regions in Periscope genes; respectively, SasG with G5/E repeats, SAP077A_019 with Big_6 domain repeats, CdrA with MBG_2 domain repeats,
surface protein Rib with Rib domain repeats, Sgo_0707 homolog containing SHIRT domain repeats, and surface adhesin CshA with ∼100 residue globular
domain repeats of a new Pfam family (CshA_repeat). BIOPHYSICS AND ERS1022024
ERS554120
ERS1106926
ERS1117683
ERS647587
ERS1043811
ERS1018586
ERS1031246
ERS666353
ERS666352
ERS1018592
ERS1043809
ERS1018584
ERS1022026
ERS1022025
ERS1018585
ERS1022027
ERS554121
ERS1058909
ERS666355
ERS666351
ERS1018587
ERS1043810
ERS1110714
12
15
17
13
18
16
16
15
15
16
15
15
17
11
15
15
12
15
15
14
16
17
18
15
CdrA
16
082
7087
180
8928
184
7084
3815
458
188
12
6
2
10
8
14
2
6
Rib
P. PNAS
|
7 of 10
https://doi.org/10.1073/pnas.2101349118 Discussion aeruginosa
B B B A 11 16
ERS970082
ERS1037087
ERS956180
ERS1058928
ERS956184
ERS1037084
ERS1043815
ERS950458
ERS956188
12
6
2
10
8
14
2
6
Rib
Streptococcus sp. C C Materials and Methods ΔN-Sgo_R2
crystallized by sitting drop vapor diffusion within 9 mo in conditions com-
prising 0.1 M Hepes (pH 7), 2.4 M ammonium sulfate at 291 K. This crystal
was passed through 4 M sodium malonate prior to flash cooling. Sgo_R10
crystallized in 5 wk at 277 K in conditions comprising 2.2 M ammonium
sulfate and 150 mM potassium thiocyanate. The crystal was harvested under
mineral oil and flash cooled in liquid nitrogen prior to data collection. Sgo_R3-4
crystallized
in
4
d
in
conditions
comprising
65%
(vol/vol)
2-methyl-2,4-pentanediol and 0.1 M Tris HCl (pH 8) at 277 K and flash cooled
in liquid nitrogen prior to data collection. A8AW49) was synthesized (Genewiz; SI Appendix, Table S2) and the truncated
single repeat ΔN-Sgo_R2 (amino acids 544 through 627), single repeats Sgo_R3
(aa 621 through 705) and Sgo_R10 (aa 1211 through 1299) (synthesized by
Eurofins Genomics), and tandem repeat Sgo_R3-4 (aa 621 through 789) were
cloned downstream of a hexahistidine tag and 3C protease specific linker by
the In-fusion method (Clontech; primer sequences listed in SI Appendix, Table
S1). DNA coding for the 7-repeat construct (Sgo_R2-8) comprising the second
to eighth SHIRT repeats (aa 537 through 1125) was synthesized (Eurofins
Genomics) and inserted by homologous recombination into a modified pBAD
vector (pBADcLIC2005), generating a coding sequence comprising GGGFA-
Sgo_R2-8-His10. Site-directed mutagenesis was used to introduce cysteine
mutations in the Sgo_R2-8 construct for fluorescent dye modification at po-
sitions S666 and S1086 (Sgo_R2-8S666C/S1086C; mutagenesis primer sequences
listed in SI Appendix, Table S1). Structure Determination. X-ray data were collected at 100 K on the I03
beamline at Diamond Light Source (Didcot, UK) using a Pilatus 3 6M de-
tector. Data were indexed and integrated by XDS (38) and scaled and
merged by Aimless (39). The structure of ΔN-Sgo_R2 was solved by MR with
2 antiparallel 5 residue ideal β-strands using Fragon (23), and the model was
built automatically with ARP/wARP (40). Sgo_R3-4 and Sgo_R10 were phased
by molecular replacement with Phaser (41) using search model ΔN-Sgo_R2. Models were manually built using Coot (42) and refined to completion with
REFMAC5 (43) for Sgo_R3-4 and PHENIX (44) for Sgo_R10 (Table 1). The
coordinates and structure factors have been deposited in the Protein Data
Bank with accession codes (ΔN-Sgo_R2, 7AVJ; Sgo_R10, 7AVK; Sgo_R3-4,
7AVH). Materials and Methods Cloning. The Escherichia coli codon-optimized coding sequence for S. gordonii
strain Challis (substrain DL1) Sgo_0707 residues 544 through 795 (UniProt: PNAS
|
7 of 10
https://doi.org/10.1073/pnas.2101349118 PNAS
|
7 of 10
https://doi.org/10.1073/pnas.2101349118 CshA
YeeJ
Rib
Fap
PavB_3
SAP077A_019
SraP
HxuA
SraP_2
PavB_2
Sgo0707
ShdA
ShdA_2
Esp
CdrA
Bca_2
SAP077A_019_2
Bag
Rib_2
Spg
BapA
PavB
ZmpC
Spa
SasG
R28
Epf
Vwbl
SasG_2
0
10
20
30
75
80
85
90
95
100
Repeat DNA identity (%)
Cluster size
2
5
10
50
0
10
20
30
40
50
# repeats CshA
YeeJ
Rib
Fap
PavB_3
SAP077A_019
SraP
HxuA
SraP_2
PavB_2
Sgo0707
ShdA
ShdA_2
Esp
CdrA
Bca_2
SAP077A_019_2
Bag
Rib_2
Spg
BapA
PavB
ZmpC
Spa
SasG
R28
Epf
Vwbl
SasG_2
0
10
20
30
75
80
85
90
95
100
Repeat DNA identity (%)
Cluster size
2
5
10
50
0
10
20
30
40
50
# repeats
Fig. 6. Repeat number variation in Periscope Proteins as a function of repeat sequence identity. The sequence identity of tandem repeats in the genome
plotted against the variation in repeat number observed for each Periscope Protein cluster. The repeat number variation is calculated as the difference
between the maximum and minimum observed repeat numbers in proteins within each cluster. The maximum number of repeats is shown as a viridis color
scale. The repeat DNA identity is calculated as the maximum repeat sequence identity across proteins in each cluster. The number of proteins in each cluster is
shown as point size. Names for the most relevant Periscope Protein clusters are shown as labels. Fig. 6. Repeat number variation in Periscope Proteins as a function of repeat sequence identity. The sequence identity of tandem repeats in the genome
plotted against the variation in repeat number observed for each Periscope Protein cluster. The repeat number variation is calculated as the difference
between the maximum and minimum observed repeat numbers in proteins within each cluster. The maximum number of repeats is shown as a viridis color
scale. The repeat DNA identity is calculated as the maximum repeat sequence identity across proteins in each cluster. The number of proteins in each cluster is
shown as point size. Names for the most relevant Periscope Protein clusters are shown as labels. with 3 kDa molecular weight cutoff (MWCO) PES (Vivaspin). Materials and Methods The structures were aligned by secondary structure matching with Su-
perpose (45) and cartoons rendered with CCP4mg (46) with secondary structure
defined by the database of protein secondary structure assignment (DSSP) (47). Protein Expression and Purification. For ΔN-Sgo_R2, Sgo_R3, and Sgo_R3-4,
expression was induced in E. coli BL21 (DE3) cells in log phase growth with
the addition of 0.1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) and
subsequently incubated with shaking at 20 °C for 20 h. Cells were harvested
by centrifugation, resuspended in 20 mM Tris HCl, 150 mM NaCl, 20 mM
imidazole (pH 7.5), and lysed by sonication. Soluble protein was purified by
standard nickel affinity chromatography methods, 3C protease was added at
a ratio of 1:100 (wt/wt) to remove the hexahistidine tag, and proteolysed
material isolated by a second nickel affinity chromatography purification. Sgo_R10 and Sgo_R3-4 were further purified by preparative SEC using a
Superdex 75 16/600 column (GE Healthcare) equilibrated in 20 mM Tris HCl,
150 mM NaCl (pH 7.5) (Sgo_R10) and (pH 8) (Sgo_R3-4). Sgo_R2-8 constructs
were transformed into Rosetta-2 E. coli (DE3) cells and grown in LB-Miller
media containing 100 μg/mL ampicillin at 37 °C to an optical density (OD)600
of 0.5 to 0.6. Gene expression was induced by addition of 0.1% (wt/vol)
L-arabinose, followed by incubation at 20 °C. Soluble protein was purified as
described
for
Sgo_R10. Protein-containing
cysteine
mutations
were
purified by affinity chromatography and SEC with the addition of 5 mM
β-mercaptoethanol at all steps. Protein samples prepared for SEC-SAXS were
dialyzed into 20 mM Tris HCl, 150 mM NaCl, 1 mM EDTA (pH 7.5). Determination of Tm. Differential scanning fluorimetry (DSF) was peformed
using a Nanotemper Prometheus NT.48 instrument. Protein concentrations were
1 mM, and solution conditions were 20 mM Tris HCl (pH 7.5), 150 mM NaCl. SAXS. SAXS experiments were performed at beamline B21, Diamond Light
Source (Didcot, UK) over a momentum transfer range (q) of 0.01 Å−1 < q <
0.4 Å−1. Scattering intensity (I versus q, where q = 4πsinθ/λ and 2θ is the
scattering angle) was collected using a Pilatus 2M detector, with a beam-to-
detector distance of 4,014 mm and an incident beam energy of 12.4 keV. Sgo_R3-4 and Sgo_R2-8S666C/S1086C were injected on an inline Shodex KW-203.5
column equilibrated in 20 mM Tris·HCl, 150 mM NaCl, 3 mM KNO3, 5mM
β-mercaptoethanol (pH 7.5), each at 7.5 mg/mL, and data processing and Protein Crystallization. Whelan et al.
Periscope Proteins are variable-length regulators of bacterial cell surface interactions Materials and Methods Proteins were concentrated (ΔN-Sgo_R2 to 48 mg/mL;
Sgo_R10 to 30 mg/mL and Sgo_R3-4 to 35.2 mg/mL) by centrifugal filtration 8 of 10
|
PNAS
https://doi.org/10.1073/pnas.2101349118 8 of 10
|
PNAS Whelan et al. Periscope Proteins are variable-length regulators of bacterial cell surface interactions reduction was performed with Chromixs (48). Rg
c (Ln[I(q) versus q2] and Dis-
tance Distribution (P(r)) plots were calculated with Primus (49). The range of
useful scattering angles was assessed using Shanum (50). SWISS-MODEL (51)
was used for the generation of a structural model of Sgo_R2-8, using the
Sgo_R3-4 crystal structure to generate iterative tandem domains for R3-5, R5-
6, and R7-8. Sgo_R2 of the Sgo_R2-8 model was generated based on Sgo_R4
from the Sgo_R3-4 structure. For validation of the Sgo_R3-4 crystal structure,
tag residues unresolved in the crystal structure were added using Modeler and
all-atom ensembles generated using Allosmod (51). In each case, 50 indepen-
dent pools of 100 models were created, and calculation and fitting of theo-
retical scattering curves to experimental data were performed using FoXS (52). This process was automated using Allosmod-FoXS (53). Plots were generated
using OriginPro version 9.5.5.409 (OriginLab), as were the gradients of the
linear regions of double logarithmic plots for the calculation of Porod
exponents (54). in two steps were performed as previously described (16). In addition, an
eccentricity ratio was calculated for each spot (using the intensity profile in
the x and y directions) to remove events that included weakly surface-
adsorbed AF488-labeled protein or partially photobleached clusters of
AF488-labeled protein. The intensity profile was calculated for a central re-
gion (2 × 10 pixels2) along the x and y axes of each spot image. This image was
the sum of the first 5 video frames for each 10 × 10 pixel2 image stack (100
frames). The x and y intensity profiles were fit with a one-dimensional
Gaussian function in MATLAB (MathWorks), and a ratio of the widths for
the Gaussian fits (= σ y-axis/σ x-axis) was used to obtain the spot eccentricity. Fluorescent spots with an eccentricity ratio between 0.9 and 1.1 were retained
(this filter removed 15 to 30% of the spots in an experiment). Bin size for the
inter-AF488 dye distance histograms was calculated using the Freedman-
Diaconis rule (55), and a single Gaussian distribution was fit to each histo-
gram in KaleidaGraph (Synergy Software). Production of Fluorescently Labeled Protein. Materials and Methods As described above, cysteine
residues were engineered into the third and eighth repeats of the Sgo_R2-8
construct at positions (S666, S1086), where fluorescence quenching by
nearby amino acid residues is minimized. Sgo_R2-8S666C/S1086C (27 μM) was
dialyzed into 150 mM NaCl, 20 mM Tris HCl, 1 mM EDTA (pH 7.5), followed
by dialysis into 150 mM NaCl, 20 mM Tris HCl, 1.35 mM Tris(2-carboxyethyl)
phosphine (TCEP) (pH 7.5). The protein was then reacted at 20 °C in low-
light conditions with a 20× molar excess of Alexa Fluor 488 C5 maleimide
(ThermoFisher, 540 μM), which was added step-wise over a period of 1.5 h
using a 10 mM stock in anhydrous DMSO. The labeling reaction was quenched
by adding dithiothreitol (DTT) at 10× molar excess to the maleimide. The
protein solution was dialyzed into 150 mM NaCl, 20 mM Tris HCl, 1 mM DTT
(pH 7.5) prior to purification by SEC on a S200 30/10 column (Amersham)
equilibrated in the same buffer to remove remaining free dye. Purified pro-
teins were stored at −80 °C. The labeling efficiency (∼2 fluorophores/protein)
was estimated from the spectrophotometrically determined concentrations of
fluorophore (e495 nm = 72,000 M−1 · cm−1) and protein (e280 nm = 99,475 M−1
· cm−1) after correction for absorption at 280 nm by the fluorophore. Molecular Dynamics Simulations. Molecular dynamics simulations were per-
formed starting from the X-ray crystal structure model of Sgo_R3-4. Simu-
lations have been performed using the CHARMM36m (56) force field and
NAMD (57). The protein was energy minimized and solvated in a periodic
rectangular box 123 Å × 46 Å × 40 Å needed to guarantee a layer of at least
12 Å solvent around the elongated protein. After a 1-ns equilibration, the
systems were simulated at 303 K for 0.8 μs. Simulations were performed in the
isothermal-isobaric ensemble, where the temperature was kept constant on
average through a Langevin thermostat and the pressure was set to 1 atm
through an isotropic Langevin piston manostat. For each saved frame of the
simulation (one every 2 ns), the positions of the Ca atoms of Sgo_R4 were least
square superposed to those of Sgo_R3, which implies a translation and a ro-
tation around the each of the three principal axes of Sgo_R3 (hence applying
the reverse transformation to the coordinates of Sgo_R4 the original confor-
mation of that frame is recovered). Materials and Methods Constructs with N domains were obtained
by taking the original X-ray crystal structure, duplicating the coordinates of
Sgo_R4 and applying to them the reverse transformation (using translation
and rotations from random frames of the trajectory) N times. Sample Preparation for SHRImP-TIRF Microscopy. A 100-mM Trolox solution
was freshly prepared by solubilizing 25 mg of Trolox powder (Fluka) in 50 μL
methanol, followed by dilution with 850 μL of 0.31 M NaOH solution. All
stock buffer solutions were passed through a 0.22-μm pore filter. Adsorption
buffer contained 10 mM Hepes, 10 mM NaCl (pH 7.0), 1 mM Trolox, 0.02%
(wt/vol) 5-μm diameter silica beads (Bangs Labs), 8 pM Alexa Fluor 488
(AF488)-labeled Sgo_R2-8S666C/S1086C protein, and, in the “blocked” samples
only, 800 pM unlabeled Sgo_R2-8 protein. Imaging buffer contained 10 mM
Hepes, 10 mM NaCl (pH 7.0), and 1 mM Trolox. Poly-D-lysine-coated quartz
slides were prepared as described previously (16). A total of 1 μM AF488-
Sgo_R2-8S666C/S1086C and 200 nM Sgo_R2-8 stock solutions were thawed
from −80 °C storage and diluted with 20 mM Tris HCl, 150 mM NaCl (pH 7.5)
buffer to the desired concentration before addition to the adsorption buffer
at a 25× or 50× dilution. In “blocked” samples, the molar concentration of
unlabeled Sgo_R2-8 in the adsorption buffer was maintained at 100× the
molar concentration of AF488-Sgo_R2-8S666C/S1086C. Periscope Protein Identification in NCTC3000 Collection. A total of 2,579,577
proteins were extracted from 734 annotated bacterial genomes downloaded
from the NCTC3000 project website (October 2019). Tandem repeats longer
than 50 residues and with at least 80% sequence identity were detected with
the T-REKS tool. Repeat sequences were clustered using BLASTp with a bit
score threshold of 30 and extracting connected components of the resulting
sequence similarity network (SSN). Similarly, proteins containing the long re-
peats were clustered using BLASTp with a bit score threshold of 100 and ad-
ditional minimum sequence identity of 90% and coverage of 50% thresholds,
by extracting the connected components of the SSN. Proteins were mapped to
UniProt (58) identifiers (version 2020_04), with their associated Gene Ontology
(GO) terms, and to Pfam (59) families (version 32.0) using PHMMER. Several
new Pfam families were created from repeat sequences during the course of
this study, including SHIRT (PF18655), YDG (PF18657), MBG_2 (PF18676), SSSPR-
51 (PF18877), CshA_repeat (PF19076), and Big_13 (PF19077), among others. Materials and Methods Phylogenetic trees of bacterial genomes were generated by first selecting all
pairs of homologous genes for each genome pair using BLASTn, then com-
puting a genomic sequence identity matrix, and finally creating a dendrogram
of strains by hierarchical clustering. In low-light conditions, 50 μL of adsorption buffer was distributed along
the center line of a 2 μg/mL or 20 μg/mL poly-D-lysine-coated quartz slide
and then covered with a clean coverslip (No. 1, 22 mm × 64 mm, Menzel-
Gläser). A flow chamber was created by sealing the two opposite, short sides
of the coverslip with nail varnish. A small amount of imaging buffer was
added along the unsealed sides of the flow chamber to prevent it from
drying out. After 10 min of incubation at room temperature (20 to 22 °C),
∼500 μL of imaging buffer was flowed through the chamber created by the
slide-silica bead-coverslip sandwich to wash away unbound protein, and the
chamber was sealed with nail varnish. Analysis of Interdomain Linkers and Domain Termini Orientation. A subset of 11
proteins with diverse domain repeats were selected from the Periscope
Proteins table. Interdomain linkers were extracted from the protein se-
quences according to the Pfam domain boundaries for each family, with
minor adjustments to cover the structural boundaries for incomplete Pfam
models (missing terminal residues). Structures of the domain repeats, or close
homologs from the same family or families within the same Pfam clan, were
selected for each protein. The distance between the terminal residues and
their relative orientation were calculated using the TADOSS software (60). The distance is measured as the Euclidean distance between the C-alpha
atoms of the N-terminal and C-terminal residues, while the orientation is
measured as the angle between the vectors formed by the four N- and four
C-terminal residues (C-alpha atoms). Downloaded from https://www.pnas.org SHRImP-TIRF Microscopy. Fluorescence excitation and detection of AF488 dye
emission was achieved using a custom, prism-coupled TIRF microscope as
described previously (16) with the following modification and increased
optical magnification. Quantum dots were not routinely added to the ad-
sorption buffer as an image-focusing aid. Video data (100 frames) were
collected using an Evolve 512 (Photometrics) electron-multiplying CCD
camera (500 ms exposure) and the pixel size was equivalent to 96 nm in the
magnified image. Data Availability. Atomic models of ΔN-Sgo_R2, Sgo_R10, and Sgo_R3-4 have
been deposited in the PDB with IDs 7AVJ, 7AVK, and 7AVH, respectively. Materials and Methods All
other study data are included in the article text and supporting information. Detection and Localization of Single Fluorophores. Fluorescent spot detection
and calculation of inter-AF488 dye distances for each spot that photobleached PNAS
|
9 of 10
https://doi.org/10.1073/pnas.2101349118 PNAS
|
9 of 10
https://doi.org/10.1073/pnas.2101349118 PNAS
|
9 of 10
https://doi.org/10.1073/pnas.2101349118 Whelan et al. prosthetic hip and knee joint infections. Biomaterials 28, 1711–1720 (2 42. P. Emsley, K. Cowtan, Coot: Model-building tools for molecular graphics. Acta Crys-
tallogr. D Biol. Crystallogr. 60, 2126–2132 (2004). 11. R. M. Corrigan, D. Rigby, P. Handley, T. J. Foster, The role of Staphylococcus aureus
surface protein SasG in adherence and biofilm formation. Microbiology (Reading)
153, 2435–2446 (2007). 43. M. D. Winn et al., Overview of the CCP4 suite and current developments. Acta Crys-
tallogr. D Biol. Crystallogr. 67, 235–242 (2011). 12. A. Toledo-Arana et al., The enterococcal surface protein, Esp, is involved in Entero-
coccus faecalis biofilm formation. Appl. Environ. Microbiol. 67, 4538–4545 (2001). 44. D. Liebschner et al., Macromolecular structure determination using X-rays, neutrons
and electrons: Recent developments in phenix. Acta Crystallogr. D Struct. Biol. 75,
861–877 (2019). 13. C. Reichhardt et al., The versatile Pseudomonas aeruginosa biofilm matrix protein
CdrA promotes aggregation through different extracellular exopolysaccharide in-
teractions. J. Bacteriol. 202, e00216–e00220 (2020). 45. R. Maiti, G. H. Van Domselaar, H. Zhang, D. S. Wishart, SuperPose: A simple server for
sophisticated structural superposition. Nucleic Acids Res. 32, W590–W594 (2004). 14. C. S. Lachenauer, R. Creti, J. L. Michel, L. C. Madoff, Mosaicism in the alpha-like
protein genes of group B streptococci. Proc. Natl. Acad. Sci. U.S.A. 97, 9630–9635 (2000). 46. S. McNicholas, E. Potterton, K. S. Wilson, M. E. M. Noble, Presenting your structures:
The CCP4mg molecular-graphics software. Acta Crystallogr. D Biol. Crystallogr. 67,
386–394 (2011). 15. C. F. Wright, S. A. Teichmann, J. Clarke, C. M. Dobson, The importance of sequence
diversity in the aggregation and evolution of proteins. Nature 438, 878–881 (2005). 47. W. Kabsch, C. Sander, Dictionary of protein secondary structure: Pattern recognition
of hydrogen-bonded and geometrical features. Biopolymers 22, 2577–2637 (1983). 16. D. T. Gruszka et al., Cooperative folding of intrinsically disordered domains drives
assembly of a strong elongated protein. Nat. Commun. 6, 7271 (2015). 48. A. Panjkovich, D. I. Svergun, CHROMIXS: Automatic and interactive analysis of
chromatography-coupled
small-angle X-ray
scattering data. Bioinformatics 34,
1944–1946 (2018). 17. F. Whelan et al., Defining the remarkable structural malleability of a bacterial surface
protein Rib domain implicated in infection. Proc. Natl. Acad. Sci. U.S.A. 116,
26540–26548 (2019). g
y
protein Rib domain implicated in infection. Proc. Natl. Acad. Sci. U.S.A. 116,
26540–26548 (2019). 49. P. V. Konarev, V. V. Volkov, A. V. Sokolova, M. H. J. Koch, D. I. Whelan et al. Svergun, PRIMUS: A
windows PC-based system for small-angle scattering data analysis. J. Appl. Cryst. 36,
1277–1282 (2003). 18. Å. Nylander et al., Structural and functional analysis of the N-terminal domain of the
Streptococcus gordonii adhesin Sgo0707. PLoS One 8, e63768 (2013). 19. C. D. Doern, C.-A. D. Burnham, It’s not easy being green: The viridans group streptococci, 19. C. D. Doern, C.-A. D. Burnham, It’s not easy being green: The viridans group streptococci,
with a focus on pediatric clinical manifestations. J. Clin. Microbiol. 48, 3829–3835 (2010). 50. P. V. Konarev, D. I. Svergun, A posteriori determination of the useful data range for
small-angle scattering experiments on dilute monodisperse systems. IUCrJ 2, 352–360
(2015). with a focus on pediatric clinical manifestations. J. Clin. Microbiol. 48, 38 20. N. S. Jakubovics, S. A. Yassin, A. H. Rickard, Community interactions of oral strepto-
cocci. Adv. Appl. Microbiol. 87, 43–110 (2014). 51. A. Waterhouse et al., SWISS-MODEL: Homology modelling of protein structures and
complexes. Nucleic Acids Res. 46, W296–W303 (2018). 21. I. M. Tleyjeh et al., Temporal trends in infective endocarditis: A population-based
study in Olmsted County, Minnesota. JAMA 293, 3022–3028 (2005). study in Olmsted County, Minnesota. JAMA 293, 3022–3028 (2005 52. D. Schneidman-Duhovny, M. Hammel, J. A. Tainer, A. Sali, Accurate SAXS profile
computation and its assessment by contrast variation experiments. Biophys. J. 105,
962–974 (2013). 22. J. Jorda, A. V. Kajava, T-REKS: Identification of tandem REpeats in sequences with a
K-means based algorithm. Bioinformatics 25, 2632–2638 (2009). 22. J. Jorda, A. V. Kajava, T-REKS: Identification of tandem REpeats in sequences with a
K means based algorithm Bioinformatics 25 2632 2638 (2009) K-means based algorithm. Bioinformatics 25, 2632–2638 (2009). 23. H. T. Jenkins, Fragon: Rapid high-resolution structure determination from ideal pro-
tein fragments. Acta Crystallogr. D Struct. Biol. 74, 205–214 (2018). 23. H. T. Jenkins, Fragon: Rapid high-resolution structure determina 53. M. Guttman, P. Weinkam, A. Sali, K. K. Lee, All-atom ensemble modeling to analyze
small-angle x-ray scattering of glycosylated proteins. Structure 21, 321–331 (2013). tein fragments. Acta Crystallogr. D Struct. Biol. 74, 205–214 (201 24. H. Hofmann et al., Polymer scaling laws of unfolded and intrinsically disordered
proteins quantified with single-molecule spectroscopy. Proc. Natl. Acad. Sci. U.S.A. 109, 16155–16160 (2012). Hammouda, A new Guinier-Porod model. J. Appl. Cryst. 43, 716–719 (20 55. D. Freedman, P. Diaconis, On the histogram as a density estimator: L 2 theory. Z. Wahrscheinlichkeitstheorie 57, 453–476 (1981). 25. Whelan et al. Whelan et al. Periscope Proteins are variable-length regulators of bacterial cell surface interactions that contributed to the results presented here. This work used the Bioscience
Technology Facility at the University of York. J.G. was funded by a Medical Re-
search Council Discovery Award (MC_PC_15073). A.B. and A.L. are funded by the
European Molecular Biology Laboratory. J.R.P., F.W., and S.C.G. were funded by
the British Heart Foundation (Grants FS/12/36/29588 and PG/17/19/32862). ACKNOWLEDGMENTS. We acknowledge Johan Turkenburg and Sam Hart
for assistance with crystal testing and data collection, Judith Hawkhead for
protein expression and purification, and Andrew Leech, Rachael Cooper, and
Michael Hodgkinson for Tm measurements and figure preparation. We also thank
the Diamond Light Source for access to beamline I03 (proposal number mx-9948) 30. D. T. Gruszka et al., Staphylococcal biofilm-forming protein has a contiguous rod-like
structure. Proc. Natl. Acad. Sci. U.S.A. 109, E1011–E1018 (2012). 1. M. van der Woude, B. Braaten, D. Low, Epigenetic phase variation of the pap operon
in Escherichia coli. Trends Microbiol. 4, 5–9 (1996). 2. J. Li et al., Epigenetic switch driven by DNA inversions dictates phase variation in
Streptococcus pneumoniae. PLoS Pathog. 12, e1005762 (2016). 31. J. H. Han, S. Batey, A. A. Nickson, S. A. Teichmann, J. Clarke, The folding and evolution
of multidomain proteins. Nat. Rev. Mol. Cell Biol. 8, 319–330 (2007). Streptococcus pneumoniae. PLoS Pathog. 12, e1005762 (2016). p
p
g
3. J. M. Abraham, C. S. Freitag, J. R. Clements, B. I. Eisenstein, An invertible element of 32. A. Borgia et al., Transient misfolding dominates multidomain protein folding. Nat. Commun. 6, 8861 (2015). 3. J. M. Abraham, C. S. Freitag, J. R. Clements, B. I. Eisenstein, An invertible element of
DNA controls phase variation of type 1 fimbriae of Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 82, 5724–5727 (1985). DNA controls phase variation of type 1 fimbriae of Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 82, 5724–5727 (1985). 33. M. B. Borgia et al., Single-molecule fluorescence reveals sequence-specific misfolding
in multidomain proteins. Nature 474, 662–665 (2011). 4. C. Vink, G. Rudenko, H. S. Seifert, Microbial antigenic variation mediated by homol-
ogous DNA recombination. FEMS Microbiol. Rev. 36, 917–948 (2012). 34. K. Savolainen et al., Expression of pls, a gene closely associated with the mecA gene
of methicillin-resistant Staphylococcus aureus, prevents bacterial adhesion in vitro. Infect. Immun. 69, 3013–3020 (2001). 5. Z. Wen, Y. Liu, F. Qu, J.-R. Whelan et al. Zhang, Allelic variation of the capsule promoter diversifies
encapsulation and virulence in Streptococcus pneumoniae. Sci. Rep. 6, 30176 (2016). 5. Z. Wen, Y. Liu, F. Qu, J.-R. Zhang, Allelic variation of the capsule promoter diversifies
l ti
d
i
l
i
St
t
i
S i R
6 30176 (2016) 35. H. A. Crosby et al., The Staphylococcus aureus global regulator MgrA modulates clumping
and virulence by controlling surface protein expression. PLoS Pathog. 12, e1005604 (2016). 6. J. O. Kim, J. N. Weiser, Association of intrastrain phase variation in quantity of cap-
sular polysaccharide and teichoic acid with the virulence of Streptococcus pneumo-
niae. J. Infect. Dis. 177, 368–377 (1998). 36. C. Gravekamp, B. Rosner, L. C. Madoff, Deletion of repeats in the alpha C protein
enhances the pathogenicity of group B streptococci in immune mice. Infect. Immun. 66, 4347–4354 (1998). 7. M. Stålhammar-Carlemalm, L. Stenberg, G. Lindahl, Protein rib: A novel group B
streptococcal cell surface protein that confers protective immunity and is expressed by
most strains causing invasive infections. J. Exp. Med. 177, 1593–1603 (1993). 37. L. C. Madoff, J. L. Michel, E. W. Gong, D. E. Kling, D. L. Kasper, Group B streptococci
escape host immunity by deletion of tandem repeat elements of the alpha C protein. Proc. Natl. Acad. Sci. U.S.A. 93, 4131–4136 (1996). most strains causing invasive infections. J. Exp. Med. 177, 1593–1603 8. C. Gravekamp, D. S. Horensky, J. L. Michel, L. C. Madoff, Variation in repeat number
within the alpha C protein of group B streptococci alters antigenicity and protective
epitopes. Infect. Immun. 64, 3576–3583 (1996). 38. W. Kabsch, XDS. Acta Crystallogr. D Biol. Crystallogr. 66, 125–132 (2010). 39. P. R. Evans, G. N. Murshudov, How good are my data and what is the resolution? Acta
Crystallogr. D Biol. Crystallogr. 69, 1204–1214 (2013). 9. F. M. Roche et al., Characterization of novel LPXTG-containing proteins of Staphy-
lococcus aureus identified from genome sequences. Microbiology (Reading) 149,
643–654 (2003). 40. G. Langer, S. X. Cohen, V. S. Lamzin, A. Perrakis, Automated macromolecular model
building for X-ray crystallography using ARP/wARP version 7. Nat. Protoc. 3,
1171–1179 (2008). 10. H. Rohde et al., Polysaccharide intercellular adhesin or protein factors in biofilm ac-
cumulation of Staphylococcus epidermidis and Staphylococcus aureus isolated from
prosthetic hip and knee joint infections. Biomaterials 28, 1711–1720 (2007). 41. A. J. McCoy et al., Phaser crystallographic software. J. Appl. Cryst. 40, 658–674 (2007). Whelan et al. S. Kalinin et al., A toolkit and benchmark study for FRET-restrained high-precision
structural modeling. Nat. Methods 9, 1218–1225 (2012). 56. J. Huang et al., CHARMM36m: An improved force field for folded and intrinsically
disordered proteins. Nat. Methods 14, 71–73 (2017). 26. O. K. Tørresen et al., Tandem repeats lead to sequence assembly errors and impose
multi-level challenges for genome and protein databases. Nucleic Acids Res. 47,
10994–11006 (2019). 57. J. C. Phillips et al., Scalable molecular dynamics with NAMD. J. Comput. Chem. 26,
1781–1802 (2005). 58. T. U. Consortium; UniProt Consortium, UniProt: A worldwide hub of protein knowl-
edge. Nucleic Acids Res. 47, D506–D515 (2019). 27. B. R. Borlee et al., Pseudomonas aeruginosa uses a cyclic-di-GMP-regulated adhesin to
reinforce the biofilm extracellular matrix. Mol. Microbiol. 75, 827–842 (2010). orce the biofilm extracellular matrix. Mol. Microbiol. 75, 827–842 (2010 59. S. El-Gebali et al., The Pfam protein families database in 2019. Nucleic Acids Res. 47,
D427–D432 (2019). 28. C. R. Back et al., The streptococcal multidomain fibrillar adhesin CshA has an elon-
gated polymeric architecture. J. Biol. Chem. 295, 6689–6699 (2020). 60. A. Lafita, P. Tian, R. B. Best, A. Bateman, TADOSS: Computational estimation of
tandem domain swap stability. Bioinformatics 35, 2507–2508 (2019). gated polymeric architecture. J. Biol. Chem. 295, 6689–6699 (2020 29. J. Lebeurre et al., Comparative genome analysis of Staphylococcus lugdunensis shows
clonal complex-dependent diversity of the putative virulence factor, ess/type VII locus. Front. Microbiol. 10, 2479 (2019). 61. S. R. Eddy, Accelerated profile HMM searches. PLOS Comput. Biol. 7, e1002195
(2011). 10 of 10
|
PNAS
https://doi.org/10.1073/pnas.2101349118 10 of 10
|
PNA Whelan et al. Whelan et al. Periscope Proteins are variable-length regulators of bacterial cell surface interactions
|
https://openalex.org/W2871774252
|
https://europepmc.org/articles/pmc6706143?pdf=render
|
English
| null |
RNA isolation from Peyer
|
Journal of biological methods
| 2,018
|
cc-by
| 10,998
|
RNA isolation from Peyer’s patch lymphocytes and
mononuclear phagocytes to determine gene expression
profiles using NanoString technology 1Division of Infectious Diseases, Wadsworth Center, New York State Department of Health, Albany NY, USA
2Division of Molecular Genetics Wadsworth Center, New York State Department of Health, Albany NY, USA
3Biochemistry and Immunology Core, Wadsworth Center, New York State Department of Health, Albany NY, USA
4Department of Biomedical Sciences, University at Albany, School of Public Health Albany, Albany NY, USA
5Program in Molecular Medicine, University of Massachusetts Medical School, Worcester, MA, USA *Corresponding author: Magdia De Jesus, Email: magdia.dejesus@health.ny.gov Competing interests: The authors have declared that no competing interests exist. Competing interests: The authors have declared that no competing interests exist. Abbreviations used: DCs: dendritic cells; DN, double negative; Dmbt1, deleted in malignant brain tumors 1; FAE: follicle associated epithelium; Fn1, fibronectin-1; GALT, gut-as
sociated lymphoid tissues; GPs, glucan particles; HBSS, Hanks’ balanced salt solution; IFR, interfollicular regions; IL-1r1, interleukin 1 receptor type 1; IL-22, interleukin-22; Lyso
DCs, lysozyme-expressing DCs; M-cell: Microfold Cell; MTE, multiplexed target enrichment; PCA, principal component analysis; PCR, polymerase chain reaction; PP: Peyer’s
patch; RT-PCR, reverse transcription PCR; RT-qPCR, quantitative reverse transcription PCR; SED, sub-epithelial dome Received January 9, 2018; Revision received February 28, 2018; Accepted March 8, 2018; Published July 2, 2018 protocol protocol Journal of Biological Methods | 2018 | Vol. 5(3) | e95
DOI: 10.14440/jbm.2018.246 ABSTRACT Sampling and immune surveillance within gut-associated lymphoid tissues such as the intestinal Peyer’s patch (PP)
occurs by an elegantly orchestrated effort that involves the epithelial barrier, B and T lymphocytes, and an extensive
network of mononuclear phagocytes. Although we now understand more about the dynamics of antigen and microbial
sampling within PPs, the gene expression changes that occur in individual cell subsets during sampling are not well
characterized. This protocol describes the isolation of high-quality RNA from sorted PP, B and T-lymphocytes, and CD11c+
phagocytes for use with nCounter-NanoString technology. This method allows investigators to study gene expression
changes within PPs in response to antigens, microbes, and oral vaccine delivery vehicles of interest that are sampled. Keywords: B-lymphocytes, mononuclear phagocytes, Peyer’s patches, NanoString technology, T-lym How to cite this article: Singh N , Gallagher HC, Song R, Dhinsa JK, Ostroff GR, De Jesus M. RNA isolation from Peyer’s patch lympho
cytes and mononuclear phagocytes to determine gene expression profiles using NanoString technology. J Biol Methods 2018;5(3):e95. DOI:
10.14440/jbm.2018.246 Animals Deoxynucleotide (dNTP) Solution Set (New England BioLabs,
cat. # N0446S) Swiss Webster mice (8–12 weeks old) were obtained from Taconic
Farms (Hudson, NY). Animals were housed under conventional, spe
cific pathogen-free conditions and were treated in compliance with the
Wadsworth Center’s Institutional Animal Care and Use Committee
(IACUC) guidelines. We recommend usage of age-matched mice for
consistency to reduce outlier variability. 100 bp DNA Ladder (Thermo Fisher, cat. # 10488058)
Equipment Kleen Guard Gloves (Kimberly Clark, cat. # 57370) Straight Surgical Scissor (FST, cat. # 14060-09) Curved Surgical Scissor (FST, cat. # 14061-09) 70 µm nylon cell strainer (CELLTREAT, cat. # 229483) MATERIALS 9
DNA Oglios (IDT, # See Fig. S1 for sequences) 9
9
Gold Taq Flexi DNA polymerase (Promega, cat. # M8295) New Hyde Park, NY) 9
9
Fetal Bovine Serum (Thermo Fisher, cat. # A3160401) Spin-X Centrifuge Tube 0.22 µm Filter (Costar, cat. # 8161) 2% E-gel (Thermo Fisher Scientific, cat. # G501802) Ethics statement 12 × 75 mm Polystyrene Tubes with Cell Strainer Cap (BD,
cat. # 352235) Experiments described in this study that involve mice were reviewed
and approved by the Wadsworth Center’s Institutional Animal Care and
Use Committee (IACUC) under protocol # 15-450. The Wadsworth
Center complies with the Public Health Service Policy on Humane
Care and Use of Laboratory Animals and was issued assurance number
A3183-01. Moreover, the Wadsworth Center is fully accredited by the
Association for Assessment and Accreditation of Laboratory Animal
Care (AAALAC). Obtaining this voluntary accreditation status reflects
that Wadsworth Center’s Animal Care and Use Program meets all of the
standards required by law, and goes beyond the standards as it strives
to achieve excellence in animal care and use. 9
9
Cell Strainer Pestle (CELLTREAT, cat. # 229480) 9
9
Countess Cell Counter (Thermo Fisher Scientific, cat. #
C10281) 9
9
Countess Cell Counting Chamber Slides (Thermo Fisher
Scientific, cat. # C10312) 9
9
Squared Petri Dishes (Fisher Scientific, cat. # FB0875711A
9
9
22 G × 1.5 in blunt-end feeding needle (Popper Scientific
New Hyde Park, NY) 9
9
Squared Petri Dishes (Fisher Scientific, cat. # FB0875711A) 9
9
22 G × 1.5 in blunt-end feeding needle (Popper Scientific, 9
9
22 G × 1.5 in blunt-end feeding needle (Popper Scientific,
New Hyde Park, NY) BACKGROUND that lack both CD11b- and CD8α- surface expression and localize to
the SED and IFR [7]. More recently, lysozyme-expressing DCs (Lyso
DCs) have also been identified in the SED, as well as DCs that express
the C-type lectin Langerin [7-10]. These DC subsets in the SED cap
ture and present antigens to resident T- and B-lymphocytes, and are
also known to influence specific T-helper (Th) responses [7,11,12]. Although we understand more about the mechanisms of how antigen
and microbial sampling occur, the gene expression changes that take
place within specific PP immune cells during sampling are not well
characterized [10,13]. Recently, Bonnardel and colleagues developed
a method to isolate and perform transcriptional analysis on PP mono
nuclear phagocyte subsets such as the CD11b+ conventional DCs, the
lysozyme-expressing monocyte-derived DC termed LysoDC, and the
CD11c(hi) lysozyme-expressing macrophages [14-16]. Bonnardel’s
method demonstrates that there is much value in studying different PP
cell subsets using transcriptomics [14,16]. In addition to its absorptive functions, the small intestine is faced
with a two-fold challenge; it must tolerate incoming antigens and
microbes while serving as a site of immune surveillance [1]. This
carefully orchestrated immune surveillance takes place within aggre
gate lymphoid structures called Peyer’s patches (PPs) [1]. Selective
transcytosis of luminal antigens, microbes, and oral vaccine delivery
vehicles into PPs occurs through specialized enterocytes called micro
fold cells (M-cells) that are interspersed within the epithelial barrier
called the follicle-associated epithelium (FAE) [2-6]. Below the FAE,
in the sub-epithelial dome (SED), there is an extensive and dynamic
network of dendritic cell (DCs) and macrophage subsets. A decade
ago, Iwasaki and Kelsall identified three distinct (CD11c+) DC subsets
within PPs, including the lymphoid DCs (CD11b-CD8α+) that reside
in the interfollicular regions (IFR), the myeloid DCs (CD11b+ CD8α)
localized in the SED, and the “so-called” double negative (DN) DCs To gain insight into the gene expression changes of specific PP cells at www.jbmethods.org
1
POL Scientific 1 1 www.jbmethods.org POL Scientific protocol protocol baseline and in response to vaccine delivery vehicles, we have developed
a protocol that isolates high-quality RNA from specific PP cells such
as B- and T-lymphocytes and CD11c+ phagocytes. Total RNA isolated
from sorted cells is then subjected to NanoString nCounter® technology
that uses a novel digital color-coded molecular barcode technology to
measure gene expression based on the counts of the target RNA [17]. In
these experiments, we specifically used the nCounter® PanCancer immune
profiling array panel to gain insight into the gene expression levels of
over 770 genes within PPs by interrogating B- and T-lymphocytes and
CD11c+ phagocytes. To compare gene expression changes at baseline
levels, we used Saccharomyces cerevisiae-derived β-glucan particles
(GPs). Earlier studies using GPs demonstrated that these particles are
taken up by both PP M-cells and CD11c+ phagocytes and are a good
positive control as they are known to stimulate an immune response
[10,18,19]. It is important to note that PP cells are extremely delicate and
the RNA is prone to rapid degradation; therefore, stabilization of RNA
early in this procedure is critically important. This method allows for
the investigator to reproducibly isolate stable RNA from sorted B- and
T lymphocytes and CD11c+ phagocytes to determine gene expression
changes using any antigen, microbe or delivery vehicle of interest in
a murine model. 9
9
SUPERase·In RNase Inhibitor (20 U/µl) (Thermo Fisher
Scientific, cat. # AM 2694) 9
9
Agilent RNA 600 Pico Kit (Agilent Technologies, cat. #
5067-1513) 9
9
Agilent RNA 600 Nano Kit (Agilent Technologies, cat. #
5067-1511) 9
9
nCounter Low RNA input Amplification kit (Nanostring, cat. # LOW-RNA-48) 9
9
Target enrichment Primer pool (NanoString, cat. # PP-MIP1-12) 9
9
nCounter® PanCancer Immune Profiling Kit (NanoString, cat. ounter® PanCancer Immune Profiling Kit (NanoString, # GXA-PATH1-12) Rat anti-mouse CD45R/B220 PerCp Clone RA3-6B2 (BD
Pharmigen, cat. # 553093) Pharmigen, cat. # 553093) Rat anti-mouse CD3 FITC-Clone 17A2 (BD Pharmigen, cat. # 555274) Rat anti-mouse Fc-Block CD16/CD32-Clone 2.4G2 (BD Rat anti-mouse Fc-Block CD16/CD32-Clone 2.4G2 (BD
Pharmigen, cat. # 553142) (
Pharmigen, cat. # 553142) Pharmigen, cat. # 553142) Armenian Hamster anti-mouse CD11c-Clone N418 (eBiosci
ence, cat. # 17-0114-82) 9
9
Luna Universal qPCR Master Mix (New England Biolabs,
cat. # M3003X) 9
9
ProtoScript II First Strand cDNA Synthesis Kit (New England
BioLabs, cat. # E6560S) Recipes 1× phosphate buffered saline pH 7.4, 2% fetal bovine serum Oral gavage 1. Gavage mice with phosphate buffered saline (PBS) or with antigen of interest. In these experiments, we used
1 × 108 highly purified, Saccharomyces cerevisiae derived β-glucan particles (GPs) as a positive control. As
a second positive control, 1 × 108 purified β-glucan particles with exposed mannans (GMPs) were used in
200 µl of 1× PBS. Delivery of particles was achieved using a 22 G × 1.5 in blunt-end feeding needle. Both
GPs and GMPs are known to stimulate the immune response [19,21]. 1. Gavage mice with phosphate buffered saline (PBS) or with antigen of interest. In these experiments, we used
1 × 108 highly purified, Saccharomyces cerevisiae derived β-glucan particles (GPs) as a positive control. As
a second positive control, 1 × 108 purified β-glucan particles with exposed mannans (GMPs) were used in
200 µl of 1× PBS. Delivery of particles was achieved using a 22 G × 1.5 in blunt-end feeding needle. Both
GPs and GMPs are known to stimulate the immune response [19,21]. NOTE: Delivery volumes should not exceed 400 µl per mouse. 9
9
nCounter digital analyzer Version 4.0.0.3
9
9
nSolver analysis software 3.0
9
9
FACSAria IIu Flow Cytometry Sorter (BD, cat. # 643895)
9
9
BD FACSDiva Flow Cytometry Sort/Analysis Softer Ware
Version 6.1.3 (BD, cat. # 643629)
9
9
BD Falcon 5 ml Polystryrene Round-Bottom Tube with
Cell-Strainer Cap (BD, cat. # 32235) Isolation of total Peyer’s patch cells from mice 2. Before euthanizing mice, prepare one tube containing 2 ml of Hanks’ balanced salt solution (HBSS) with 50 µl
of SUPERase·In RNase inhibitor (20 U/µl), label as tube number 1, and place it on ice. Label one additional
tube as number 2, add 2 ml of HBSS with 30 µl of SUPERase·In RNase Inhibitor (20 U/µl) and place on ice. 2. Before euthanizing mice, prepare one tube containing 2 ml of Hanks’ balanced salt solution (HBSS) with 50 µl
of SUPERase·In RNase inhibitor (20 U/µl), label as tube number 1, and place it on ice. Label one additional
tube as number 2, add 2 ml of HBSS with 30 µl of SUPERase·In RNase Inhibitor (20 U/µl) and place on ice. NOTE: This step is critical, as addition of the SUPERase·In RNase inhibitor immediately stabilizes RNA when
PPs are harvested. Without the SUPERase·In RNase inhibitor, RNA recovery will be poor. Keeping everything on
ice is extremely important. NOTE: This step is critical, as addition of the SUPERase·In RNase inhibitor immediately stabilizes RNA when
PPs are harvested. Without the SUPERase·In RNase inhibitor, RNA recovery will be poor. Keeping everything on
ice is extremely important. CAUTION: DO NOT USE RNAlater RNA Stabilizer Reagent (Qiagen cat. #76104) specifically for PPs, as this
reagent will significantly reduce the viability of PP cells to 40%. The RNA quality was also found to be very poor. We found that RNAlater made PPs float in solution during the harvesting step, suggesting that these tissues were
damaged. PPs naturally sink to the bottom of the tube in solution. 2.1. Euthanize mice by CO2 asphyxiation, according to institutional guidelines. 2.1. Euthanize mice by CO2 asphyxiation, according to institutional guidelines. 2.1. Euthanize mice by CO2 asphyxiation, according to institutional guidelines. 2.2. Cleanse the mouse abdomen with 70% ethanol prior to necropsy. Perform a standard necropsy that in
volves a 1 cm incision using straight surgical grade scissors along the midline beginning about 1.5 cm
from the base of the rib cage. Expose the peritoneal cavity and identify the cecum. Snip the terminal
small intestine at the ileal-cecal junction, gently remove the small intestine in its entirety, and snip
the stomach-ileal junction to release intact intestine. 2.3. Identify individual PPs located on the anti-mesenteric side of the intestine. Mice typically contain 5
to 10 PPs spaced more or less evenly from the duodenum (proximal) to ileum (distal). 2.3. Reagents iBase (Thermo Fisher Scientific, cat. # G6400) g
9
9
RNase Zap Decontamination Solution (Thermo Fisher Sci
entific, cat. # AM 9780)
9
9
70% Ethanol (Pharmco, cat. # 04343-19)
9
9
β-glucan particles (GPs) [20]
9
9
β-glucan particles with residual mannan content (GMPs) [20]
9
9
β-Mercaptoethanol (Sigma, cat. # M6250)
9
9
0.4% Trypan Blue (Amresco, cat. # K940)
9
9
RNeasy Mini Kit (Qiagen, cat. # 74106)
9
9
DNase RNase-Free Set (1500 Kunitz units) (Qiagen, cat. #
79254) 9
9
Freeze Zone 4.5 Benchtop Lyophilizer (Labconco, cat. #7750020) 9
9
96 Well Thermal iCycler (Biorad, cat. # 170-8740) 9
9
Chip Priming Station (Agilent, cat. # 5065-4401) Vortex Mixer (IKA, cat. # 0025001607) 16 pin bayonet electrode cartridges (Agilent, cat. #5065-4413) PCR Machine DNA Engine Dyad (MJ Research, cat. # PTC-
220) 9
9
2100 Expert Software (Agilent, Version B.02.08.SI648 (SR2))
9
9
nCounter Prep Station Version 4.1.0.1 2100 Expert Software (Agilent, Version B.02.08.SI648 (SR2))
nCounter Prep Station Version 4.1.0.1 POL Scientific J Biol Methods | 2018 | Vol. 5(3) | e95 J Biol Methods | 2018 | Vol. 5(3) | e95 2 POL Scientific protocol 9
9
nCounter digital analyzer Version 4.0.0.3
9
9
nSolver analysis software 3.0
9
9
FACSAria IIu Flow Cytometry Sorter (BD, cat. # 643895)
9
9
BD FACSDiva Flow Cytometry Sort/Analysis Softer Ware
Version 6.1.3 (BD, cat. # 643629)
9
9
BD Falcon 5 ml Polystryrene Round-Bottom Tube with
Cell-Strainer Cap (BD, cat. # 32235) phenol red pH 7.2 (1 L, Sigma cat. # H1387)
8 g NaCl, 0.4 g KCl, 0.06 g KH2PO4, 0.048 g NaHPO4,1 g D-Glu
cose, 0.098 g MgSO4 (anhydrous), 0.14 g CaCl2 (anhydrous), Adjust
the pH to 7.2 with HCl
1× phosphate buffered saline pH 7.4 (1 L)
8 g of NaCl, 0.2 g of KCl, 1.44 g of Na2HPO4, 0.24 g of KH2PO4,
Adjust the pH to 7.4 with HCl
Flow buffer
1× phosphate buffered saline pH 7.4, 2% fetal bovine serum phenol red pH 7.2 (1 L, Sigma cat. # H1387)
8 g NaCl, 0.4 g KCl, 0.06 g KH2PO4, 0.048 g NaHPO4,1 g D-Glu
cose, 0.098 g MgSO4 (anhydrous), 0.14 g CaCl2 (anhydrous), Adjust
the pH to 7.2 with HCl 1× phosphate buffered saline pH 7.4 (1 L)
8 g of NaCl, 0.2 g of KCl, 1.44 g of Na2HPO4, 0.24 g of KH2PO4,
Adjust the pH to 7.4 with HCl p
Hanks’ balanced salt solution without sodium bicarbonate and PROCEDURE Oral gavage Isolation of total Peyer’s patch cells from mice Identify individual PPs located on the anti-mesenteric side of the intestine. Mice typically contain 5
to 10 PPs spaced more or less evenly from the duodenum (proximal) to ileum (distal). 2.4. Using curved surgical scissors, gently excise individual PPs and place them immediately in cold HBSS
that contains 50 µl of SUPERase·In RNase inhibitor (tube 1). The scissors should be placed curve-side
up just above the PP and then gently applied to the tissue. Add RNAase Zap to surgical instruments
and then rinse with ethanol between every mouse. 2.4. Using curved surgical scissors, gently excise individual PPs and place them immediately in cold HBSS
that contains 50 µl of SUPERase·In RNase inhibitor (tube 1). The scissors should be placed curve-side
up just above the PP and then gently applied to the tissue. Add RNAase Zap to surgical instruments
and then rinse with ethanol between every mouse. NOTE: Two to three mice yield plenty of cells to be able to isolate RNA from all three PP cell types. To avoid cell
clumping, isolate PPs from less than 5 mice per group. NOTE: Two to three mice yield plenty of cells to be able to isolate RNA from all three PP cell types. To avoid cell
clumping, isolate PPs from less than 5 mice per group. 2.5. Decant HBSS in tube 1 and transfer PPs into tube 2 that contains 2 ml of HBSS with 30 µl of SU
PERase·In RNase inhibitor. 2.5. Decant HBSS in tube 1 and transfer PPs into tube 2 that contains 2 ml of HBSS with 30 µl of SU
PERase·In RNase inhibitor. 2.5. Decant HBSS in tube 1 and transfer PPs into tube 2 that contains 2 ml of HBSS with 30 µl of SU
PERase·In RNase inhibitor. J Biol Methods | 2018 | Vol. 5(3) | e95 3 protocol Staining cells for cell sorting Prepare a (1:100) dilution of Rat anti-mouse Fc-block in flow buffer and resuspend cells in 100 µl of
this buffer and incubate on ice for 15 min. Place the ice bucket on a rocker during incubation. 3.4. Prepare a (1:100) dilution of Rat anti-mouse Fc-block in flow buffer and resuspend cells in 100 µl of
this buffer and incubate on ice for 15 min. Place the ice bucket on a rocker during incubation. 3.5. During the incubation step with the Fc block, count viable cells in the suspension by preparing a 1:20
dilution of cells with trypan blue. Add 10 μl of the cell suspension to the countess chamber slides and
count cells. 20 PPs should yield approximately 1.5 × 106 cells per ml with a viability range of 90%–98%. 3.5. During the incubation step with the Fc block, count viable cells in the suspension by preparing a 1:20
dilution of cells with trypan blue. Add 10 μl of the cell suspension to the countess chamber slides and
count cells. 20 PPs should yield approximately 1.5 × 106 cells per ml with a viability range of 90%–98%. 3.6. After the 15 min incubation, spin the plate for 2 min at 2100 rpm (887× g), and decant supernatant
by inverting the plate. 3.7. Resuspend the cells in 100 µl of fluorophore-conjugated antibodies added directly to cell suspensions
(1:200 for B and T-cells and 1:40 for CD11c+ phagocytes) and incubate on ice for 30 min with con
tinuous rocking. 3.7. Resuspend the cells in 100 µl of fluorophore-conjugated antibodies added directly to cell suspensions
(1:200 for B and T-cells and 1:40 for CD11c+ phagocytes) and incubate on ice for 30 min with con
tinuous rocking. 3.8. Spin the plate for 2 min at 2100 rpm (887× g), and decant supernatant by inverting the plate. 3.10. Spin the plate for 2 min at 2100 rpm (887× g), and remove the supernatant by inverting the plate. 3.11. Resuspend the cells from the plate into 3 ml of flow buffer; some of cells will clump. Pass the 3 ml
cell suspension through a filtered cap flow tube that contains 1ml of flow buffer resulting in a total
of 4 ml of cell suspension for cell sorting. Staining cells for cell sorting Staining cells for cell sorting NOTE: Before moving on to this step, wipe petri dish, pestle and cell strainer with RNase Zap decontamination
solution. NOTE: Before moving on to this step, wipe petri dish, pestle and cell strainer with RNase Zap decontamination
solution. 3. To generate a cell suspension, decant tube 2 into a 70 μm nylon mesh cell strainer that is resting on a petri
dish. This setup should be on ice. Grind PPs with pestle or with the back part of a syringe plunger. The 2 ml
should yield approximately 1200 µl after grinding the PPs. Transfer the cell suspension from the petri dish
into a microfuge tube [21]. If you use RNAlater RNA Stabilizer Reagent instead of SUPERase·In RNase
inhibitor, PPs will feel like rubber at this step and will not yield high quality RNA. 3.1. Save 10–15 µl of cells to count them in step 3.5. CRITICAL STEP: Prepare flow buffer that contains SUPERase· In RNase inhibitor, as this buffer will be used
for all subsequent steps to stain and wash cells for cell sorting. It is recommended to prepare 20 ml of flow buffer
at a 1:250 dilution. CRITICAL STEP: Prepare flow buffer that contains SUPERase· In RNase inhibitor, as this buffer will be used
for all subsequent steps to stain and wash cells for cell sorting. It is recommended to prepare 20 ml of flow buffer
at a 1:250 dilution. 3.2. Using a 96-round bottom plate, add 50 µl of cells per well for single color controls. For B-cells, T-cells
and DCs that will be sorted, fill ten wells per cell type, with 100 µl of cells per well. Spreading cells
into multiple wells prevents cell clumping in subsequent steps. 3.2. Using a 96-round bottom plate, add 50 µl of cells per well for single color controls. For B-cells, T-cells
and DCs that will be sorted, fill ten wells per cell type, with 100 µl of cells per well. Spreading cells
into multiple wells prevents cell clumping in subsequent steps. 3.3. Spin the plate for 2 min at 2100 rpm (887× g), to pellet the cells and decant supernatant by inverting
the plate. 3.3. Spin the plate for 2 min at 2100 rpm (887× g), to pellet the cells and decant supernatant by inverting
the plate. 3.4. Cell sorting of B- and T-lymphocytes and CD11c+ phagocytes and RNA isolation Cells were analyzed and sorted on a FACS Aria IIu cell sorter. Prior to the collection of sorted cells, a cell
purity test sort should be performed to check that the cell types of interest are free of debris and contamina
tion by other cell types in the sample. 4. Cells were analyzed and sorted on a FACS Aria IIu cell sorter. Prior to the collection of sorted cells, a cell
purity test sort should be performed to check that the cell types of interest are free of debris and contamina
tion by other cell types in the sample. NOTE: Magnetic bead isolation yields mixed populations of cells, making it difficult to interpret NanoString data. A combination of magnetic bead isolation with cell sorting also decreases RNA yield significantly. NOTE: Magnetic bead isolation yields mixed populations of cells, making it difficult to interpret NanoString data. A combination of magnetic bead isolation with cell sorting also decreases RNA yield significantly. NOTE: RNA samples derived from PPs are sensitive to degradation, even with the addition of RNase inhibitors. The RNeasy Mini Kit calls for cell wall lysis by using a syringe, which leads to RNA degradation. We therefore
developed an alternative method for cell lysis using the Spin-X filter column (step 4.1). NOTE: RNA samples derived from PPs are sensitive to degradation, even with the addition of RNase inhibitors. The RNeasy Mini Kit calls for cell wall lysis by using a syringe, which leads to RNA degradation. We therefore
developed an alternative method for cell lysis using the Spin-X filter column (step 4.1). 4.1. Sort CD45R/B220+(CD3-/CD11c-) B-cells, CD3+(CD45R/B220-/CD11c-) T-cells and CD11c+(CD45R/
B220-/CD3-) CD11c+ phagocytes onto a filtered Spin-X column that is moistened with 100 µl of flow
buffer that containing SUPERase·In RNase inhibitor. For PP B and T-cells we sorted 300000 cells. 4.1. Sort CD45R/B220+(CD3-/CD11c-) B-cells, CD3+(CD45R/B220-/CD11c-) T-cells and CD11c+(CD45R/
B220-/CD3-) CD11c+ phagocytes onto a filtered Spin-X column that is moistened with 100 µl of flow
buffer that containing SUPERase·In RNase inhibitor. For PP B and T-cells we sorted 300000 cells. L Scientif J Biol Methods | 2018 | Vol. 5(3) | e95 POL Scientific protocol Since CD11c+ phagocytes are the least abundant of the three cell types in PPs, we sorted a maximum
of 50000–80000 cells. Since CD11c+ phagocytes are the least abundant of the three cell types in PPs, we sorted a maximum
of 50000–80000 cells. 4.2. During sorting, centrifuge the Spin-X column every 100000 cells at 2000 rpm for 1 min at room
temperature (20°C to 30°C) as this is the maximum volume that the tubes hold during cell sorting. Discard flow buffer through and continue sorting. 4.3. When sorting is finished, centrifuge the Spin-X column at 2000 rpm for 2 min. Place the filter in a
new 1.5 ml collection tube. Discard flow buffer through an old collection tube. 4.4. Add 350 µl of Qiagen’s RLT buffer that contains β-Mercaptoethanol (BME) (10 µl of BME in 1 ml
of RLT) to the filter, pipette 20–25 times to mix. Place tube on dry ice for 5 min. 4.5. Incubate on rotary spinner for 10 min and vortex at the end of the incubation. The RNA extraction is
performed at room temperature (20°C to 30°C). 4.6. Centrifuge the Spin-X column for 2 min at 13400 rpm. Keep flow-through. To obtain a maximum
yield of RNA, the flow through is passed a second time unto the filter. er the flow-through back onto the filter, and centrifuge for 2 min at 13400 rpm. Keep flow-through. 4.7. Transfer the flow-through back onto the filter, and centrifuge for 2 min at 13400 rpm. Kee 4.8. Add 100 µl of RNase free H2O to the filter and pipette 20–25 times, and centrifuge for 1 min at 13400
rpm. Keep flow through and discard the filter. 4.9. Add 250 µl 100% ethanol to the tube. 4.10. Transfer the liquid to the RNeasy spin column and pipette 20–25 times, incubate for 5 min on bench
top. Centrifuge for 1 min at 13400 rpm. Discard flow-through. Column DNase digestion Column DNase digestion NOTE: The DNase RNase-Free Set was used according to manufacturer’s instructions. 4.11. Add 350 µl buffer RW1 to the RNeasy spin column. Close the lid gently, and centrifuge for 15 s at
13400 rpm to wash the spin column membrane. Discard the flow-through. 4.12. Add 10 µl DNase I stock solution to 70 µl buffer RDD. Vortex to mix and add 80 µl directly to each
RNeasy spin column membrane, then incubate on the bench top for 15 min. POL Scientific NOTE: Be sure to add the DNase I incubation directly to the RNeasy spin column membrane. DNase digestion
will be incomplete if part of the mix sticks to the walls of the spin column. 4.13. Add 350 µl buffer RW1 directly onto the RNeasy spin column that contains the 80 µl of DNase I and
spin for 15 s at ≥ 10000 rpm. Discard the flow-through. 4.14. Add 500 µl buffer RPE to the RNeasy spin column. Close the lid gently, and centrifuge for 15 s 13400
rpm to wash the spin column membrane. Discard the flow-through. 4.15. Add 500 µl buffer RPE to the RNeasy spin column. Close the lid gently, and centrifuge for 2 min at
13400 rpm to wash the spin column membrane. Discard the flow-through. 4.16. Add another 500 µl buffer RPE to the RNeasy spin column. Close the lid gently, and centrifuge for 2
min at 13400 rpm to wash the spin column membrane. Discard the flow-through. 4.17. Place the RNeasy spin column in a new 2 ml collection tube and discard the old collection tube contain
ing the flow-through. Spin at 13400 rpm for 1 min to eliminate any possible carryover of buffer RPE. 4.18. Place the RNeasy spin column in a new 1.5 ml collection tube. Add 30 µl of RNase-free water directly
to the spin column membrane, incubate for 2 min, then spin at 13400 rpm for 2 min to elute the RNA. Keep flow-through. 4.18. Place the RNeasy spin column in a new 1.5 ml collection tube. Add 30 µl of RNase-free water directly
to the spin column membrane, incubate for 2 min, then spin at 13400 rpm for 2 min to elute the RNA. Keep flow-through. 4.19. Add 10 µl of RNase-free water directly to the spin column membrane, incubate for 2 min, then spin
at 13400 rpm for 2 min to elute the RNA. Keep flow-through. Vortex the 40 µl of collected RNA. 4.19. Add 10 µl of RNase-free water directly to the spin column membrane, incubate for 2 min, then spin
at 13400 rpm for 2 min to elute the RNA. Keep flow-through. Vortex the 40 µl of collected RNA. 4.20. Sample can be stored short-term at −20°C; for longer storage, store at −80°C. 4.20. Sample can be stored short-term at −20°C; for longer storage, store at −80°C. Measurement of RNA using bioanalyzer For T-cells, the yield was in the range of 1.5–14 ng,
and for CD11c+ phagocytes, the yield was 0.06–0.6 ng, in a total volume of 5 µl each. Prior to hybridization with
the reporter and capture probe set, we used the nCounter Low RNA Input Amplification kit to generate cDNA
followed by the multiplexed target enrichment primer pool to amplify target genes in the codeset. Since the yield
of RNA is low, it is important to check the desired amplification of targets in all cell types by reverse transcription
polymerase chain reaction (PCR) and by reverse transcription quantitative PCR before proceeding with nCounter
Low RNA Input amplification. nCounter low RNA input amplification assay 8. Add RNA (2 ng or less) with the provided reagents in the kit; 0.5 µl of 10× RT enzyme mix, 0.5 µl of 10×
Primer Mix and bring up to 5 µl with RNase-free water. 8.1. Synthesize cDNA in thermocycler using the following conditions: anneal primer at 25°C for 10 min,
first strand cDNA synthesis at 42°C for 60 min, enzyme inactivation for 85°C for 5 min and hold at 4°C. 8.2. Add the following to the cDNA: 1.5 µl 5× dT Amp Master Mix, 1 µl gene specific primers (500 nM
per primer) corresponding to the codeset for multiplexed target enrichment. Gently, flick to mix and
centrifuge. NOTE: Estimate the number of cycles to generate dsDNA for each cell type based on manufacturer’s instructions
and the data collected during the RT-PCR/RT-qPCR in step 7.2. NOTE: Estimate the number of cycles to generate dsDNA for each cell type based on manufacturer’s instructions
and the data collected during the RT-PCR/RT-qPCR in step 7.2. 8.3. Set up the amplification reaction in thermocycler. Initial denaturation 95°C for 10 min, 95°C for 15 s
and 60°C for 4 min, followed by hold at 4°C. Number of cycles needs to be estimated on cell type,
as stated in the note. 8.3. Set up the amplification reaction in thermocycler. Initial denaturation 95°C for 10 min, 95°C for 15 s
and 60°C for 4 min, followed by hold at 4°C. Number of cycles needs to be estimated on cell type,
as stated in the note. 8.4. Proceed with NanoString hybridization or store the dsDNA at −80°C. RT-PCR and RT-qPCR assays for detection of gene expression . Synthesize the cDNA using ProtoScript II first strand cDNA Synthesis Kit from NEB using 1 ng of RNA for
each cell type with random hexamer primers. 7.1. Use gene-specific primers to perform either reverse transcription-PCR (RT-PCR) or quantitative RT-PCR
(RT-qPCR) on the cDNA. RT-PCR was performed on cDNA to ensure the amplification efficiency and
specificity of the primers. Using both methods, we chose three housekeeping genes from the immune
profiling codeset Hprt, Tubb5 and Sf3a3 (Fig. S1). 7.2. RT-PCR conditions: Initial denaturation 94°C for 3 min, 94°C for 20 s, 55°C for 20 s, 72°C for 20 s
(30×) and 72°C for 5 min, followed by hold at 4°C. RT-qPCR conditions: Initial denaturation 95°C
for 60 s, 95°C for 15 s and 60°C for 30 s (40×), followed by hold at 4°C. 7.3. To estimate the minimum number of cycles needed for cDNA amplification, we used the RT-qPCR assay,
using Luna Universal qPCR Master Mix and the primers for the housekeeping genes mentioned in 7.1. 7.4. The estimated number of cycles from the RT-qPCR was used in the nCounter Low RNA Input Am
plification assay. protocol protocol for 3 h. To ensure that the flask remains cold and the RNA remains frozen, place the flask on dry ice while
it is attached to the lyophilizer. Resuspend lyophilized RNA in 6 µl of RNase-free water. for 3 h. To ensure that the flask remains cold and the RNA remains frozen, place the flask on dry ice while
it is attached to the lyophilizer. Resuspend lyophilized RNA in 6 µl of RNase-free water. Lyophilize RNA to concentrate Lyophilize RNA to concentrate 5. RNA stored at −80°C must be placed on dry ice and the lid of the tube covered with parafilm. Make small
holes on the top of the tube using a flame-heated needle and place inside a lyophilization flask. Lyophilize 5. RNA stored at −80°C must be placed on dry ice and the lid of the tube covered with parafilm. Make small
holes on the top of the tube using a flame-heated needle and place inside a lyophilization flask. Lyophilize J Biol Methods | 2018 | Vol. 5(3) | e95 5 5 Measurement of RNA using bioanalyzer ement of RNA using bioanalyzer
6. Before use, thaw sample and ladder on ice until fully resuspended. Take 1 µl of RNA sample and ladder and
heat-denature at 70°C for 2 min. 6.1. Agilent RNA 6000 Pico kit protocol was followed as per manufacturer’s instructions. NOTE: The RNA yield for B-cells was in the range of 3–20 ng. For T-cells, the yield was in the range of 1.5–14 ng,
and for CD11c+ phagocytes, the yield was 0.06–0.6 ng, in a total volume of 5 µl each. Prior to hybridization with
the reporter and capture probe set, we used the nCounter Low RNA Input Amplification kit to generate cDNA
followed by the multiplexed target enrichment primer pool to amplify target genes in the codeset. Since the yield
of RNA is low, it is important to check the desired amplification of targets in all cell types by reverse transcription
polymerase chain reaction (PCR) and by reverse transcription quantitative PCR before proceeding with nCounter
Low RNA Input amplification. . Before use, thaw sample and ladder on ice until fully resuspended. Take 1 µl of RNA sample and ladder and
heat-denature at 70°C for 2 min. 6. Before use, thaw sample and ladder on ice until fully resuspended. Take 1 µl of RNA sample and ladder and
heat-denature at 70°C for 2 min. 6.1. Agilent RNA 6000 Pico kit protocol was followed as per manufacturer’s instructions. 6.1. Agilent RNA 6000 Pico kit protocol was followed as per manufacturer’s instructions. NOTE: The RNA yield for B-cells was in the range of 3–20 ng. For T-cells, the yield was in the range of 1.5–14 ng,
and for CD11c+ phagocytes, the yield was 0.06–0.6 ng, in a total volume of 5 µl each. Prior to hybridization with
the reporter and capture probe set, we used the nCounter Low RNA Input Amplification kit to generate cDNA
followed by the multiplexed target enrichment primer pool to amplify target genes in the codeset. Since the yield
of RNA is low, it is important to check the desired amplification of targets in all cell types by reverse transcription
polymerase chain reaction (PCR) and by reverse transcription quantitative PCR before proceeding with nCounter
Low RNA Input amplification. NOTE: The RNA yield for B-cells was in the range of 3–20 ng. J Biol Methods | 2018 | Vol. 5(3) | e95 protocol Close the tubes and immediately transfer to the 65°C thermocycler set up in step 9.2. 9.10. Incubate in 65°C for 12–30 h. .10. Incubate in 65°C for 12–30 h. PREP station PREP station 9.11. Warm sealed cartridge from −20°C to room temperature and sealed reagent plates from 4°C to room
temperature (20°C–30°C). 9.11. Warm sealed cartridge from −20°C to room temperature and sealed reagent plates from 4°C to room
temperature (20°C–30°C). 9.12. Centrifuge reagent plates 2000 g for 2 min. 9.12. Centrifuge reagent plates 2000 g for 2 min. 9.13. Follow prep station automated instructions until it requests sample loading. Immediately remove
sample from 65°C, spin in Picofuge, carefully remove caps, and load in prep station. Immediately
initiate protocol by pressing start. 9.13. Follow prep station automated instructions until it requests sample loading. Immediately remove
sample from 65°C, spin in Picofuge, carefully remove caps, and load in prep station. Immediately
initiate protocol by pressing start. 9.13. Follow prep station automated instructions until it requests sample loading. Immediately remove
sample from 65°C, spin in Picofuge, carefully remove caps, and load in prep station. Immediately
initiate protocol by pressing start. 9.14. Remove cartridge and seal with adhesive tape to prevent evaporation. Tape is provided in kit. 9.14. Remove cartridge and seal with adhesive tape to prevent evaporation. Tape is provided in kit. 9.14. Remove cartridge and seal with adhesive tape to prevent evaporation. Tape is provided in kit. .15. Take cartridge and place in digital analyzer for RNA copy counts. Digital analyzer 9.16. Upload the Reporter Library file in the provided flash drive. 9.17. Create a CDF file with sample name and description and upload into the digital analyzer. 9.18. Insert cartridge with the seal (step 9.14) into the digital analyzer. 9.19. Initiate counts by pressing start. 9.20. When the digital analyzer is complete download your data for analysis. 9.21. Follow instructions on nSolver or self-analyze. .16. Upload the Reporter Library file in the provided flash drive. 9.16. Upload the Reporter Library file in the provided flash drive. .17. Create a CDF file with sample name and description and upload into the digital analyzer 9.18. Insert cartridge with the seal (step 9.14) into the digital analyzer. .18. Insert cartridge with the seal (step 9.14) into the digital analyzer. 9.20. When the digital analyzer is complete download your data for analysis. 9.21. Follow instructions on nSolver or self-analyze. protocol .3. Thaw reporter code set and capture code set on ice. Mix by flicking or inverting tubes. NOTE: DO NOT VORTEX. Mix by flicking or inverting tubes, followed by a quick spin in picofuge. Make sure
to change pipette tip every time. NOTE: DO NOT VORTEX. Mix by flicking or inverting tubes, followed by a quick spin in picofuge. Make sure
to change pipette tip every time. 9.4. Prepare master mix for the 12 samples. The master mix includes 70 µl of sample hybridization buffer
to the provided reporter code set. NOTE: DO NOT VORTEX. Mix by flicking or inverting tubes, followed by a quick spin in picofuge. Make sure
to change pipette tip every time. NOTE: DO NOT VORTEX. Mix by flicking or inverting tubes, followed by a quick spin in picofuge. Make sure
to change pipette tip every time. 9.5. Label strip tubes and cut the strip in the middle to yield 6 tubes each. Aliquot 8 µl of master mix into
strip tubes. 9.5. Label strip tubes and cut the strip in the middle to yield 6 tubes each. Aliquot 8 µl of master mix into
strip tubes. 9.6. Adjust dsDNA input to 5 µl in dH2O. NOTE: It is important to empirically adjust the input amount so that the concentration of dsDNA is not too high
or too low. NOTE: It is important to empirically adjust the input amount so that the concentration of dsDNA is not too high
or too low. 9.7. Add dsDNA to the 8 µl of master mix into labeled strip tubes. 9.7. Add dsDNA to the 8 µl of master mix into labeled strip tubes. NOTE: DO NOT VORTEX. Mix by flicking or inverting tubes, followed by a quick spin in picofuge. Make sure
to change pipette tip every time. NOTE: DO NOT VORTEX. Mix by flicking or inverting tubes, followed by a quick spin in picofuge. Make sure
to change pipette tip every time. 9.8. Add 2 µl of capture probe set to every tube. NOTE: DO NOT VORTEX, Mix by flicking or inverting tubes, followed by a quick spin in picofuge. Make sure
to change pipette tip every time. NOTE: DO NOT VORTEX, Mix by flicking or inverting tubes, followed by a quick spin in picofuge. Make sure
to change pipette tip every time. 9.9. Reading RNA counts with NanoString 9. Thaw the dsDNA if frozen or proceed with hybridization. 9. Thaw the dsDNA if frozen or proceed with hybridization. 9.1. Heat denatures the dsDNA at 95°C for 2 min. Immediately cool on ice. 9.2. Pre-heat PCR thermocycler to 65°C with the lid set to 70°C. J Biol Methods | 2018 | Vol. 5(3) | e95 6 6 protocol ANTICIPATED RESULTS quality from PPs is extremely poor with less than 10% recovery. To
overcome these challenges, we treated PP single cell suspensions with
two different commercially available RNase inhibitors, RNAlater and
SUPERase·In RNase inhibitor. We found that treatment with RNAlater
caused PPs to float in solution, yielding a hard and tacky texture that
was not ideal for making single cell suspensions. Cell viability using
the RNAlater reagent was determined to be 46% (Fig. S2). We noted
that PPs treated with SUPERase·In RNase inhibitor remained in the
bottom of the tube, as expected, and single cell suspensions yielded
98% viability (Fig. S2). Most importantly, we found that viability of
PP single cell suspensions remained above 90% after 2–4 h on ice with
SUPERase·In RNase inhibitor, a significant improvement from our
previous protocol, which was done without inhibitors [24]. quality from PPs is extremely poor with less than 10% recovery. To
overcome these challenges, we treated PP single cell suspensions with
two different commercially available RNase inhibitors, RNAlater and
SUPERase·In RNase inhibitor. We found that treatment with RNAlater
caused PPs to float in solution, yielding a hard and tacky texture that
was not ideal for making single cell suspensions. Cell viability using
the RNAlater reagent was determined to be 46% (Fig. S2). We noted
that PPs treated with SUPERase·In RNase inhibitor remained in the
bottom of the tube, as expected, and single cell suspensions yielded
98% viability (Fig. S2). Most importantly, we found that viability of
PP single cell suspensions remained above 90% after 2–4 h on ice with
SUPERase·In RNase inhibitor, a significant improvement from our
previous protocol, which was done without inhibitors [24]. PPs are aggregate lymphoid structures that can be easily seen with
the naked eye and are located on the anti-mesenteric side of the small
intestine (Fig. 1A). Each PP contain multiple follicles that are indi
vidually equipped with an epithelial barrier, an extensive network
of mononuclear phagocytes, and T- and B-lymphocytes that play an
important role in sampling luminal antigens, microbes, and oral vac
cine delivery vehicles (Fig. 1B) [10]. To assess the transcriptomics of
individual cell types in response to β-glucan delivery vehicles, PPs were
harvested and RNA was isolated. It is important to note that isolation
of PPs cells can be challenging because PP cells are extremely delicate
and difficult to maintain as viability decreases by roughly 15% every
30 min [21]. ANTICIPATED RESULTS Additionally, in the absence of RNA inhibitors, the RNA Figure 1. Peyer’s patches are aggregate lymphoid structures. A. Top view image of PP show that these are aggregate lymphoid structures that
are on the anti-mesenteric side of the small intestine. B. Profile image of PP reveal multiple follicles, all of which participate in sampling and immune
surveillance. PPs are surrounded by an extensive network of blood and lymphatic vessels that allow the dynamic movement of immune cells in and out
of the gut. Scale bar is 1 mm. Figure 1. Peyer’s patches are aggregate lymphoid structures. A. Top view image of PP show that these are aggregate lymphoid structures that
are on the anti-mesenteric side of the small intestine. B. Profile image of PP reveal multiple follicles, all of which participate in sampling and immune
surveillance. PPs are surrounded by an extensive network of blood and lymphatic vessels that allow the dynamic movement of immune cells in and out
of the gut. Scale bar is 1 mm. After determining which RNase inhibitor was optimal in prolong
ing cell viability and providing RNA stability, we used cell sorting to
separate PP B220+ B-cells, CD3+ T-cells and CD11c+ phagocytes (Fig. 2A). Using this method, we sorted a single PP suspension of cells into
three separate tubes that contained CD45R/B220+(CD3-/CD11c-) B-cells,
CD3+(CD45R/B220- /CD11c) T-cells and CD11c+(CD45R/B220-/CD3-)
phagocytes. During sorting, we carefully selected the gates denoted in
green (Fig. 2A-2E) such that there was no cross-contamination of each
of the cell types of interest. Reanalysis of the sorted cells revealed that
we accomplished 99% purity for B220+ B-cells (Fig. 2G), 99% purity
for CD3+ T-cells (Fig. 2J), and 73% purity for CD11c+ phagocytes (Fig. 2M). Although we could not achieve a higher percentage of purity for
CD11c+ phagocytes due to their complicated shape, we determined that
there was no contamination by B220+ or CD3+ cells (Fig. 2N). We also
found that using a combination of both magnetic bead cell isolation
and cell sorting did not improve cell purity but, rather, it decreased
RNA yield to 10%–15%, due to the increased experimental time and
cell losses in each step (data not shown). Using both bioanalyzer nano and pico chips, we determined that RNA isolated using RNA-later was
highly degraded (Fig. S3A, lanes 2–4). SUPERase·In RNase inhibitor
alone yielded multiple bands. However, many of these bands were
not clearly defined due to DNA contamination (Fig. J Biol Methods | 2018 | Vol. 5(3) | e95 protocol ANTICIPATED RESULTS STATISTICAL ANALYSIS mune profiling kit. The results were analyzed using the raw count with
DEseq2 [22,23]. A transcript was considered differentially expressed
when up or downregulated at least 2-fold and the P-adjusted value ≤ 0.05. mune profiling kit. The results were analyzed using the raw count with
DEseq2 [22,23]. A transcript was considered differentially expressed
when up or downregulated at least 2-fold and the P-adjusted value ≤ 0.05. NanoString analysis was performed in duplicate per experimental
condition with the nCounter Analysis System using the PanCancer Im POL Scientific J Biol Methods | 2018 | Vol. 5(3) | e95
POL 7 J Biol Methods | 2018 | Vol. 5(3) | e95 POL Scientific protocol ANTICIPATED RESULTS S3A, lanes 6–8). RNA purified by SUPERase·In RNase inhibitor was of higher quality,
although the preparation was contaminated with DNA. We therefore
subjected the RNA preparation to DNase treatment, which resulted in
DNA-free RNA. The bioanalyzer picochip results yielded reproducible
bands that represent 28S (4.7 kb) and 18S (1.9 kb) rRNA (Fig. S3B, lanes
2–4). We also compared two RNA extraction methods by sorting cells
directly unto Trizol reagent or Spin-X columns that contained a 0.22
µm membrane. Spin-X columns were subjected to dry ice and RNeasy
reagents to extract RNA from cells. We determined that the Spin-X
column RNA extraction method consistently yielded DNA-free RNA
of high quality in the bioanalyzer data (Fig. S3B and data not shown). We also determined the limits of detection of the RNA isolated per cell
type and per single mouse. We found that PP RNA isolation for B- and
T -lymphocytes can be achieved at an individual mouse level as both POL Scientific 8 POL Scientific protocol used for this cell type (data not shown). used for this cell type (data not shown). are abundant within PPs. For CD11c+ phagocytes, RNA amounts fall
below the limits of detection; therefore, more than one mouse must be below the limits of detection; therefore, more than one mouse must be
Figure 2. Cell sorting and reanalysis of isolated PP B-cells, T-cells and CD11c+ phagocytes. Single cell suspension of PPs were stained for the
cell surface markers: B220 for B-cells, CD3 for T-cells and CD11c for CD11c+ phagocytes. FACS sorting panels: A. Forward (FSC) and side scatter
(SSC) plot represents the entire single cell suspension. Cells were gated on (B) B220-/CD3- cells, which represented 31% of the population. C. CD11c+
phagocytes which represented 8% of the population. D. CD11c- cells (representing 85% of the population). E. B220+ CD3- cells (representing 48% of
the population) and B220- CD3+ cells (representing 16% of the population). Reanalysis panels: F-H. The purity for each cell type. For B220+ B-cells, a
99% purity was obtained. I-K. For CD3+ T-cells, a 99% purity was obtained. Neither B or T-cells were contaminated with CD11c+ phagocytes. L-N. For
CD11c+ phagocytes, 73% purity was obtained; these also were not contaminated by B- or T-cells. Figure 2. Cell sorting and reanalysis of isolated PP B-cells, T-cells and CD11c+ phagocytes. ANTICIPATED RESULTS Single cell suspension of PPs were stained for the
cell surface markers: B220 for B-cells, CD3 for T-cells and CD11c for CD11c+ phagocytes. FACS sorting panels: A. Forward (FSC) and side scatter
(SSC) plot represents the entire single cell suspension. Cells were gated on (B) B220-/CD3- cells, which represented 31% of the population. C. CD11c+
phagocytes which represented 8% of the population. D. CD11c- cells (representing 85% of the population). E. B220+ CD3- cells (representing 48% of
the population) and B220- CD3+ cells (representing 16% of the population). Reanalysis panels: F-H. The purity for each cell type. For B220+ B-cells, a
99% purity was obtained. I-K. For CD3+ T-cells, a 99% purity was obtained. Neither B or T-cells were contaminated with CD11c+ phagocytes. L-N. For
CD11c+ phagocytes, 73% purity was obtained; these also were not contaminated by B- or T-cells. Figure 3. Comparison of GPs and GMPs 24 h post gavage. PPs from mice gavaged with FITC-labeledGPs (A) and GMPs (green) (B) were harvested
after 24 h. Both GPs and GMPs are localized within CD11c+ phagocytes (magenta) in the SED of the patch. Scale Bar is 100 µm. Figure 3. Comparison of GPs and GMPs 24 h post gavage. PPs from mice gavaged with FITC-labeledGPs (A) and GMPs (green) (B) were harvested
after 24 h. Both GPs and GMPs are localized within CD11c+ phagocytes (magenta) in the SED of the patch. Scale Bar is 100 µm. Biol Methods | 2018 | Vol. 5(3) | e95
9
POL Scientific J Biol Methods | 2018 | Vol. 5(3) | e95 9 protocol To assess the gene expression profiles of each cell type, we used the
nCounter® mouse Pan Cancer Immune profiling kit that profiles over
770 murine immune system-related genes. To obtain both reproducible
and biologically relevant results from PP RNA, it was critical to start
with high quality RNA. To compare the differential expression of genes
per cell type in comparison to the untreated control, we gavaged mice
with β-glucan particles (GPs) derived from Saccharomyces cerevisiae
cell walls. GPs are highly purified, spherical hollow shells that mainly
consist of β-1, 3-D-glucans. Because GPs are recognized by Dectin-1
and complement 3 receptors, they are considered useful antigen present
ing cell-targeted vaccine adjuvant/delivery vehicles. The inner hollow
cavity of GPs can be loaded to deliver antigens to macrophages and
dendritic cells [20]. ANTICIPATED RESULTS In addition to GPs, we also used GMPs, which are
considered a lower purity β-glucan particle that contain residual mannan
content in the particle cell wall. GMPs have also been shown to have
immunostimulatory properties [19]. Confocal microscopy of PPs from
mice gavage with GPs and GMPs show that both types of particles
are sampled by PPs CD11c+ phagocytes (Fig. 3A and 3B). Since GPs
and GMPs can serve as both antigen-presenting cell-targeted delivery
systems and as adjuvants, we were particularly interested in differential
gene expression of CD11c+ phagocytes [20]. downregulated at least 2-fold and the P-adjusted value to be ≤ 0.05. Mice were gavaged with GPs and GMPs, and PPs were harvested after
24 h. When comparing the differential expression within each cell type
at baseline and with stimulation by GPs and GMPs, the results reveal
that stimulation with GPs did not yield any significant differential gene
expression in B-cells and T-cells. Only one gene, interleukin 1 receptor
type 1, (IL-1r1), showed differential expression in CD11c+ phagocytes. In contrast, GMPs stimulated gene expression in CD11c+ phagocytes
but not in B- and T-cells. The four genes that were most significant
ly upregulated in CD11c+ phagocytes by GMPs were interleukin-22
(IL-22), (21-fold), deleted in malignant brain tumors 1 (Dmbt1) 11-
fold, IL-1r1 (6-fold), and Fibronectin-1 (Fn1) (6-fold). IL-22, is a key
cytokine that has been demonstrated to induce antimicrobial activity,
stimulating tissue-damage protection and tissue repair and remodeling
[25]. The Dmbt1 gene encodes a glycoprotein with multiple scavenger
receptor domains that may be involved in the activation of the com
plement mannose lectin pathway, and influences IL-22 [26]. IL-1r1 is
the receptor for IL-1, a proinflammatory cytokine that influences IL-22
[27]. Fibronectin-1 (Fn1) is involved in cell adhesion and motility [28]
(Table 1). In addition, principal component analysis of the data show
that gene expression profiles are clustered based on cell types regardless
of treatment, indicating that there are larger changes in gene expression
between cell types (Fig. 4). To determine highly significant differential gene expression, we
analyzed the RNA raw counts with DEseq2. Differential expression
of transcripts was stringently defined, requiring transcripts to be up or Figure 4. Principal component analysis of PP cell types. Principal component analysis demonstrates gene expression profiles are clustered based
on cell types B-cells, T-cells and CD11c+ phagocytes, regardless of treatment with GP, GMPs or control. Figure 4. Principal component analysis of PP cell types. ANTICIPATED RESULTS Principal component analysis demonstrates gene expression profiles are clustered bas
on cell types B-cells, T-cells and CD11c+ phagocytes, regardless of treatment with GP, GMPs or control. Figure 4. Principal component analysis of PP cell types. Principal component analysis demonstrates gene expression profiles are clustered based
on cell types B-cells, T-cells and CD11c+ phagocytes, regardless of treatment with GP, GMPs or control. 10
J Biol Methods | 2018 | Vol. 5(3) | e95
POL Scientific J Biol Methods | 2018 | Vol. 5(3) | e95 10 mvv121. PMID: 26634447 6. Mabbott NA, Donaldson DS, Ohno H, Williams IR, Mahajan A (2013) Microfold
(M) cells: important immunosurveillance posts in the intestinal epithelium. Mucosal Immunol 6: 666-677. doi: 10.1038/mi.2013.30. PMID: 23695511 We thank the Wadsworth Center Applied Genomic Technologies
Core and the Wadsworth Center Biochemistry/Immunology Core. We
thank Kelly Miller of NanoString technologies for the technical support
and data analysis. We also thank the Richard Cole of the Wadsworth
Center Advanced Light Microscopy and Image Analysis Core. We thank
Kimberly McClive-Reed, Kathleen McDonough and Keith Derbyshire
for the critical reading of this manuscript. Ostroff GR was supported by
HDTRA-12-D-0003 0008. De Jesus M was supported by the University
at Albany and Wadsworth Center start-up funds. 7. Iwasaki A, Kelsall BL (2001) Unique functions of CD11b+, CD8 alpha+, and
double-negative Peyer's patch dendritic cells. J Immunol 166: 4884-4890. doi:
10.4049/jimmunol.166.8.4884. PMID: 11290765 8. Rochereau N, Verrier B, Pin J, Genin C, Paul S (2011) Phenotypic localization
of distinct DC subsets in mouse Peyer Patch. Vaccine 29: 3655-3661. doi:
10.1016/j.vaccine.2011.03.012. PMID: 21439318 9. Lelouard H, Henri S, De Bovis B, Mugnier B, Chollat-Namy A, et al. (2009)
Pathogenic bacteria and dead cells are internalized by a unique subset of Peyer's
patch dendritic cells that express lysozyme. Gastroenterology 138: 173-184. doi: 10.1053/j.gastro.2009.09.051. PMID: 19800337 10. De Jesus M, Ostroff GR, Levitz SM, Bartling TR, Mantis NJ (2014) A
population of Langerin-positive dendritic cells in murine Peyer's patches
involved in sampling β-glucan microparticles. PLoS One 9: doi: 10.1371/
journal.pone.0091002. PMID: 24632738 TROUBLESHOOTING The reported method here provides a reproducible approach to isolat
ing high-quality RNA from PP, B-cells, T-cells and CD11c+ phagocytes
which allowed the application of NanoString technology to determine
changes in the transcriptional profiles of these cells. To achieve optimal Possible problems and their troubleshooting solutions are listed in
Table 2. Table 2. Possible problems and their troubleshooting solutions. Table 2. Possible problems and their troubleshooting solutions. Step #
Problems
Causes
Suggestions
2
Rapid RNA degradation
Adequate amounts of SUPERase·In RNase
inhibitor were not added
Add suggested amounts of SUPERase·In RNase inhibitor
3
Rapid RNA degradation
PPs were not ground on ice
PPs must always be dissociated on ice
3.11
Cells clump before sorting
Small volumes of flow buffer
Start with a minimum of 1–2 ml of flow buffer as cells are
filtered through
4
Rapid RNA degradation
Harsh cell lysis
Use a Spin-X column not a syringe
7.1
NanoString low input kit
does not seem to be efficient
Very low levels of RNA
Estimate the minimum number of cycles needed for cDNA
amplification, we used the RT-qPCR assay protocol results with this protocol, it is important to use an RNase inhibitor that is
suited for PPs immediately after harvesting tissues. Careful cell sorting
is also an important part of this method, as cells must be properly gated
to ensure the highest cell purity during cell sorting. Lastly, the stringent
data analysis that we describe will identify differentially expressed genes
with high confidence. As an example of the utility of this method, we
were able identify changes in genetic expression of GPs and GMPs
β-glucan delivery vehicles that can target PP CD11c+ phagocytes. A
major advantage of the presented method is that the investigator can
perform intestinal immune profiling per cell type with any antigen or
microbe of interest that is introduced into the intestinal mucosa. results with this protocol, it is important to use an RNase inhibitor that is
suited for PPs immediately after harvesting tissues. Careful cell sorting
is also an important part of this method, as cells must be properly gated
to ensure the highest cell purity during cell sorting. Lastly, the stringent
data analysis that we describe will identify differentially expressed genes
with high confidence. As an example of the utility of this method, we
were able identify changes in genetic expression of GPs and GMPs
β-glucan delivery vehicles that can target PP CD11c+ phagocytes. A
major advantage of the presented method is that the investigator can
perform intestinal immune profiling per cell type with any antigen or
microbe of interest that is introduced into the intestinal mucosa. *NS denotes no significant differential gene expression. protocol protocol To determine whether we could find additional cell specific tran
scripts, we performed DEseq2 analysis between control B-cells, T-cells,
and CD11c+ phagocytes. Our stringent cut-off criteria were the same as
described earlier. The results reveal that there were a total of 396 differ
entially expressed genes. Among those transcripts, 29 genes were only
upregulated in B-cells, 103 genes were only upregulated in T-cells, and
97 genes were only upregulated in CD11c+ phagocytes (Fig. 5). We also found that 63 differentially expressed genes were only downregulated
in B-cells, 70 genes that were only downregulated in T-cells, and 34
genes that were only downregulated in DCs (Fig. S4). The data used
in our representative heatmaps demonstrates the reproducibility of this
method across all cell types under experimental conditions involving
stimulation with GPs and GMPs. found that 63 differentially expressed genes were only downregulated
in B-cells, 70 genes that were only downregulated in T-cells, and 34
genes that were only downregulated in DCs (Fig. S4). The data used
in our representative heatmaps demonstrates the reproducibility of this
method across all cell types under experimental conditions involving
stimulation with GPs and GMPs. 18 | Vol. 5(3) | e95
POL Scientific
-cells, 103 genes were only upregulated in T-cells, and
nly upregulated in CD11c+ phagocytes (Fig. 5). We also
stimulation with GPs and GMPs. map compares gene expression profiles of upregulated genes in B-cells, T-cells and DCs. Z-scores in this heatmap sho
regulated specifically per cell type. Red indicates downregulation, green indicates upregulation, C denotes control, and GP
nt with particles. Data is presented in duplicate. Figure 5. Heatmap compares gene expression profiles of upregulated genes in B-cells, T-cells and DCs. Z-scores in this heatmap show that a set
of genes are upregulated specifically per cell type. Red indicates downregulation, green indicates upregulation, C denotes control, and GP and GMPs
indicate treatment with particles. Data is presented in duplicate. 11 J Biol Methods | 2018 | Vol. 5(3) | e95 protocol Table 1. Differential gene expression of GPs and GMPs in DCs. GPs
GMPs
Fold change
Adjusted P
Fold change
Adjusted P
IL-22
NS*
21.8
1.31E-06
Dmbt1
NS*
10.9
0.000183
IL-1r1
2.5
0.16481
5.6
0.03035
Fn1
NS*
5.9
0.040187
*NS denotes no significant differential gene expression. Table 1. Differential gene expression of GPs and GMPs in DCs. protocol 24. De Jesus M, Ahlawat S, Mantis NJ (2013) Isolating and immunostaining
lymphocytes and dendritic cells from murine Peyer's patches. J Vis Exp: e50167. doi: 10.3791/50167. PMID: 23525039 14. Bonnardel J, Da Silva C, Wagner C, Bonifay R, Chasson L, et al. (2017)
Distribution, location, and transcriptional profile of Peyer's patch conventional
DC subsets at steady state and under TLR7 ligand stimulation. Mucosal Immunol
10: 1412-1430. doi: 10.1038/mi.2017.30. PMID: 28378808 25. Dudakov JA, Hanash AM, van den Brink RM (2015) Interleukin-22:
immunobiology and pathology. Annu Rev Immunol 33: 747-785. doi: 10.1146/
annurev-immunol-032414-112123. PMID: 25706098 15. Da Silva C, Wagner C, Bonnardel J, Gorvel J, Lelouard H (2017) The Peyer's
Patch Mononuclear Phagocyte System at Steady State and during Infection. Front Immunol 8: 1254. doi: 10.3389/fimmu.2017.01254. PMID: 29038658 26. Renner M, Bergmann G, Krebs I, End C, Lyer S, et al. (2007) DMBT1 confers
mucosal protection in vivo and a deletion variant is associated with Crohn's
disease. Gastroenterology 133: 1499-1509. doi: 10.1053/j.gastro.2007.08.007. PMID: 17983803 16. Bonnardel J, Da Silva C, Masse M, Montañana-Sanchis F, Gorvel J, et al. (2015)
Gene expression profiling of the Peyer's patch mononuclear phagocyte system. Genom Data 5: 21-24. doi: 10.1016/j.gdata.2015.05.002. PMID: 26484215 17. Geiss GK, Bumgarner RE, Birditt B, Dahl T, Dowidar N, et al. (2008) Direct
multiplexed measurement of gene expression with color-coded probe pairs. Nat
Biotechnol 26: 317-325. doi: 10.1038/nbt1385. PMID: 18278033 27. Chen VL, Surana NK, Duan J, Kasper DL (2013) Role of murine intestinal
interleukin-1 receptor 1-expressing lymphoid tissue inducer-like cells in
Salmonella infection. PLoS One 8: doi: 10.1371/journal.pone.0065405. PMID:
23750260 18. Huang H, Ostroff GR, Lee CK, Specht CA, Levitz SM (2013) Characterization
and optimization of the glucan particle-based vaccine platform. Clin Vaccine
Immunol 20: 1585-1591. doi: 10.1128/CVI.00463-13. PMID: 23945157 28. Kay RA, Ellis IR, Jones SJ, Perrier S, Florence MM, et al. (2005) The expression
of migration stimulating factor, a potent oncofetal cytokine, is uniquely controlled
by 3'-untranslated region-dependent nuclear sequestration of its precursor
messenger RNA. Cancer Res 65: 10742-10749. doi: 10.1158/0008-5472. CAN-05-2038. PMID: 16322219 19. Young S, Ostroff GR, Zeidler-Erdely PC, Roberts JR, Antonini JM, et al. (2007) A comparison of the pulmonary inflammatory potential of different
components of yeast cell wall. J Toxicol Environ Health A 70: 1116-1124. doi:
10.1080/15287390701212224. PMID: 17558806 20. Mirza Z, Soto ER, Dikengil F, Levitz SM, Ostroff GR (2017) Beta-Glucan
Particles as Vaccine Adjuvant Carriers. Methods Mol Biol 1625: 143-157. doi:
10.1007/978-1-4939-7104-6_11. protocol PMID: 28584989 References 1. Jung C, Hugot J, Barreau F (2010) Peyer's Patches: The Immune Sensors
of the Intestine. Int J Inflam 2010: 823710-12. doi: 10.4061/2010/823710. PMID: 21188221 11. Reboldi A, Arnon TI, Rodda LB, Atakilit A, Sheppard D, et al. (2016) IgA
production requires B cell interaction with subepithelial dendritic cells in
Peyer's patches. Science 352: doi: 10.1126/science.aaf4822. PMID: 27174992 2. Didierlaurent A, Sirard J, Kraehenbuhl J, Neutra MR (2002) How the gut senses
its content. Cell Microbiol 4: 61-72. doi: 10.1046/j.1462-5822.2002.00177.x. PMID: 11896763 12. Salazar-Gonzalez RM, Niess JH, Zammit DJ, Ravindran R, Srinivasan A, et al. (2006) CCR6-mediated dendritic cell activation of pathogen-specific T cells in
Peyer's patches. Immunity 24: 623-632. doi: 10.1016/j.immuni.2006.02.015. PMID: 16713979 3. Kanaya T, Ohno H (2014) The Mechanisms of M-cell Differentiation. Biosci
Microbiota Food Health 33: 91-97. doi: 10.12938/bmfh.33.91. PMID: 25032083 3. Kanaya T, Ohno H (2014) The Mechanisms of M-cell Differentiation. Biosci
Microbiota Food Health 33: 91-97. doi: 10.12938/bmfh.33.91. PMID: 25032083 4. Kraehenbuhl JP, Neutra MR (2000) Epithelial M cells: differentiation and
function. Annu Rev Cell Dev Biol 16: 301-332. doi: 10.1146/annurev. cellbio.16.1.301. PMID: 11031239 4. Kraehenbuhl JP, Neutra MR (2000) Epithelial M cells: differentiation and
function. Annu Rev Cell Dev Biol 16: 301-332. doi: 10.1146/annurev. cellbio.16.1.301. PMID: 11031239 13. De Jesus M, Rodriguez AE, Yagita H, Ostroff GR, Mantis NJ (2015) Sampling
of Candida albicans and Candida tropicalis by Langerin-positive dendritic
cells in mouse Peyer's patches. Immunol Lett 168: 64-72. doi: 10.1016/j. imlet.2015.09.008. PMID: 26386376 5. Ohno H (2015) Intestinal M cells. J Biochem 159: 151-160. doi: 10.1093/jb/ 5. Ohno H (2015) Intestinal M cells. J Biochem 159: 151-160. doi: 10.1093/jb/ 5. Ohno H (2015) Intestinal M cells. J Biochem 159: 151-160. doi: 10.1093/jb/ POL Scientific J Biol Methods | 2018 | Vol. 5(3) | e95 J Biol Methods | 2018 | Vol. 5(3) | e95 12 POL Scientific J Biol Methods | 2018 | Vol. 5(3) | e95 J Biol Methods | 2018 | Vol. 5(3) | e95 Supplementary information Figure S1. Specific gene targets in all cell types are tested by RT-PCR. Figure S1. Specific gene targets in all cell types are tested by RT-PCR. 21. De_Jesus M, Ahlawat S, Mantis NJ (2013) Isolating and immunostaining
lymphocytes and dendritic cells from murine Peyer's patches. J Vis Exp: e50167. doi: 10.3791/50167. PMID: 23525039 Figure S2. Percent viability of PP single cell suspensions. Figure S2. Percent viability of PP single cell suspensions. Figure S2. Percent viability of PP single cell suspensions. Figure S3. Stabilization of PP RNA is important for nanostring-based
methods. Figure S3. Stabilization of PP RNA is important for nanostring-based
methods. Figure S3. Stabilization of PP RNA is important for nanostring-based
methods. 22. Love MI, Huber W, Anders S (2014) Moderated estimation of fold change
and dispersion for RNA-seq data with DESeq2. Genome Biol 15: 550. doi:
10.1186/s13059-014-0550-8. PMID: 25516281 Figure S4. Heatmap compares gene expression profiles of down
regulated genes in B-cells, T-cells and DCs. Figure S4. Heatmap compares gene expression profiles of down
regulated genes in B-cells, T-cells and DCs. Figure S4. Heatmap compares gene expression profiles of down
regulated genes in B-cells, T-cells and DCs. 23. Chen J, Tambalo M, Barembaum M, Ranganathan R, Simões-Costa M, et
al. (2017) A systems-level approach reveals new gene regulatory modules in
the developing ear. Development 144: 1531-1543. doi: 10.1242/dev.148494. PMID: 28264836 upplementary information of this article can be found online a http://www.jbmethods.org/jbm/rt/suppFiles/246. http://www.jbmethods.org/jbm/rt/suppFiles/246. POL Scientific 13 POL Scientific
|
https://openalex.org/W2987668296
|
https://www.duo.uio.no/bitstream/10852/70778/1/12964_2019_Article_467.pdf
|
English
| null |
GREM1 is associated with metastasis and predicts poor prognosis in ER-negative breast cancer patients
|
Cell communication and signaling
| 2,019
|
cc-by
| 13,187
|
RESEARCH Open Access GREM1 is associated with metastasis and
predicts poor prognosis in ER-negative
breast cancer patients Ulrike Neckmann1,2,3†, Camilla Wolowczyk1,2,3†, Martina Hall4,5, Eivind Almaas4,5, Jiang Ren6, Sen Zhao7,
Bjarne Johannessen7, Rolf I. Skotheim7,8, Geir Bjørkøy1,2, Peter ten Dijke6 and Toril Holien9,10* Neckmann et al. Cell Communication and Signaling (2019) 17:140
https://doi.org/10.1186/s12964-019-0467-7 Neckmann et al. Cell Communication and Signaling (2019) 17:140
https://doi.org/10.1186/s12964-019-0467-7 Abstract Background: In breast cancer, activation of bone morphogenetic protein (BMP) signaling and elevated levels of
BMP-antagonists have been linked to tumor progression and metastasis. However, the simultaneous upregulation
of BMPs and their antagonist, and the fact that both promote tumor aggressiveness seems contradictory and is not
fully understood. Methods: We analyzed the transcriptomes of the metastatic 66cl4 and the non-metastatic 67NR cell lines of the
4T1 mouse mammary tumor model to search for factors that promote metastasis. CRISPR/Cas9 gene editing was
used for mechanistic studies in the same cell lines. Furthermore, we analyzed gene expression patterns in human
breast cancer biopsies obtained from public datasets to evaluate co-expression and possible relations to clinical
outcome. Results: We found that mRNA levels of the BMP-antagonist Grem1, encoding gremlin1, and the ligand Bmp4 were
both significantly upregulated in cells and primary tumors of 66cl4 compared to 67NR. Depletion of gremlin1 in
66cl4 could impair metastasis to the lungs in this model. Furthermore, we found that expression of Grem1
correlated with upregulation of several stem cell markers in 66cl4 cells compared to 67NR cells. Both in the mouse
model and in patients, expression of GREM1 associated with extracellular matrix organization, and formation,
biosynthesis and modification of collagen. Importantly, high expression of GREM1 predicted poor prognosis in
estrogen receptor negative breast cancer patients. Analyses of large patient cohorts revealed that amplification of
genes encoding BMP-antagonists and elevation of the corresponding transcripts is evident in biopsies from more
than half of the patients and much more frequent for the secreted BMP-antagonists than the intracellular inhibitors
of SMAD signaling. Conclusion: In conclusion, our results show that GREM1 is associated with metastasis and predicts poor prognosis
in ER-negative breast cancer patients. Gremlin1 could represent a novel target for therapy. Keywords: 4T1 model, 66cl4, 67NR, BMP4, GREM1, Metastasis, Gremlin 1 * Correspondence: toril.holien@ntnu.no
Ulrike Neckmann and Camilla Wolowczyk are Co-first author
9Department of Clinical and Molecular Medicine (IKOM), NTNU, Gastro
Center, Prinsesse Kristinas gt 1, 7030 Trondheim, Norway
10Department of Hematology, St. Olavs Hospital, Trondheim, Norway
Full list of author information is available at the end of the article * Correspondence: toril.holien@ntnu.no
Ulrike Neckmann and Camilla Wolowczyk are Co-first author
9Department of Clinical and Molecular Medicine (IKOM), NTNU, Gastro
Center, Prinsesse Kristinas gt 1, 7030 Trondheim, Norway
10Department of Hematology, St. Olavs Hospital, Trondheim, Norway
Full list of author information is available at the end of the article Background g
Members of the transforming growth factor β (TGF-β)
family, including the TGF-βs, activins, nodal, bone
morphogenetic proteins (BMPs), and growth and differ-
entiation factors (GDFs), play important roles in tumor
progression and formation of metastases [1]. However,
these signaling molecules have pleiotropic effects in
cancer, as seen with TGF-β signaling, which has been
linked to both tumor suppression and tumor promotion
depending on the tumor stage [2, 3]. After binding of
the TGF-βs to transmembrane receptors, SMAD2 and
SMAD3 become phosphorylated and bind to SMAD4;
subsequent translocation of this complex to the nucleus
result in the expression of TGF-β controlled genes. In
contrast, BMP signaling primarily occurs via SMAD1/
5/8 instead of SMAD2/3. Receptor activation by TGF-β
family ligands can also result in induction of non-
SMAD pathways, including mitogen-activated protein
kinases (MAPK), and PI3K-Akt-mTOR pathways [4]. Like TGF-βs, BMPs have been implicated in both
tumor suppression and tumor progression [5]. Under-
standing the role of TGF-β family members in cancer
and the development of TGF-β targeted therapies have
been hampered by the complex interplay of these sig-
naling molecules with intracellular SMAD-inhibitors
and extracellular BMP-antagonists. The expression of
intracellular
inhibitory
SMADs,
i.e. SMAD6
and
SMAD7, is induced in response to several TGF-β li-
gands and form a negative feedback loop. Extracellular
BMP-antagonists, like noggin, gremlin1, Dand5, and
follistatin, bind to BMPs and prevent binding of the
BMPs to their respective receptor [6, 7]. Like BMPs and
TGF-βs, also BMP-antagonists can have a tumor-
promoting role. For instance, Dand5, also known as
Coco, facilitated formation of lung metastases in the
4T1 mammary tumor model by blocking lung-derived
BMPs [8], whereas gremlin1 promoted stem cell main-
tenance in both glioma [9] and colorectal cancer [10]. Here, we focused on 67NR, which does not metastasize,
and 66cl4, which metastasizes to the lungs. We hypoth-
esized that their different metastatic propensity relates
to differences in the secretome of the two cell lines,
since cancer cell-derived factors influence the pheno-
type of the cancer cells and/or the stroma cells. We
therefore performed transcriptome analysis of 66cl4
and 67NR cells from in vitro culture and grown as tu-
mors in BALB/c nude mice and focused on differen-
tially expressed mRNAs encoding secreted proteins. Expression of the BMP-antagonist Grem1 was signifi-
cantly upregulated in 66cl4 compared to 67NR both in
cell culture and primary tumors. Gremlin1 binds and
inhibits BMP2, BMP4, and BMP7 [15], which play
important roles during cellular differentiation [1]. In
line with the elevated expression of Grem1 in 66cl4,
several stem cell markers were upregulated in the 66cl4
cells. Interestingly, compared to the non-metastatic
67NR cells, the metastatic 66cl4 cells also expressed
significantly elevated levels of Bmp4. Consistent with a
role of these secreted proteins in metastasis, we found
that high GREM1 expression correlated with reduced
relapse-free survival (RFS) in estrogen receptor (ER)-
negative breast cancer patients and that the association
with poor prognosis remains, even with elevated co-
expression of a BMP in the tumor. Methods For information about Transcriptome analysis; Quantita-
tive real-time PCR; Immunoblotting; Harvest of condi-
tioned medium; ELISA; In vitro extravasation assay; Cell
proliferation assay; Soft-agar assay; Flow cytometry; see
Additional file 1: Supplementary Methods. Metastasis is a complex multi-step process, includ-
ing local invasion, intravasation of tumor cells into
the circulation, tumor cell dissemination, extravasa-
tion at metastatic site, survival at secondary site, and
finally proliferation and formation of metastases [11]. To metastasize, cancer cells of epithelial origin have
to acquire new properties such as increased motility,
anchorage-independent
growth
and
the
ability
to
adapt to a foreign microenvironment. Only cancer
cells that display a high degree of phenotypic plasti-
city and thus are able to adapt to stressful conditions
and changing environments, will survive. For example,
migratory tumor cells require stem cell-like properties
including a low differentiation status [12]. © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Neckmann et al. Cell Communication and Signaling (2019) 17:140 Neckmann et al. Cell Communication and Signaling (2019) 17:140 Page 2 of 17 Page 2 of 17 lines isolated from the same spontaneously formed
tumor [13, 14]. All the cell lines form primary tumors
when injected into the fat pad of BALB/c mice, but they
differ widely in their ability to form distant metastasis. Here, we focused on 67NR, which does not metastasize,
and 66cl4, which metastasizes to the lungs. We hypoth-
esized that their different metastatic propensity relates
to differences in the secretome of the two cell lines,
since cancer cell-derived factors influence the pheno-
type of the cancer cells and/or the stroma cells. We
therefore performed transcriptome analysis of 66cl4
and 67NR cells from in vitro culture and grown as tu-
mors in BALB/c nude mice and focused on differen-
tially expressed mRNAs encoding secreted proteins. Expression of the BMP-antagonist Grem1 was signifi-
cantly upregulated in 66cl4 compared to 67NR both in
cell culture and primary tumors. Gremlin1 binds and
inhibits BMP2, BMP4, and BMP7 [15], which play
important roles during cellular differentiation [1]. In
line with the elevated expression of Grem1 in 66cl4,
several stem cell markers were upregulated in the 66cl4
cells. Interestingly, compared to the non-metastatic
67NR cells, the metastatic 66cl4 cells also expressed
significantly elevated levels of Bmp4. Consistent with a
role of these secreted proteins in metastasis, we found
that high GREM1 expression correlated with reduced
relapse-free survival (RFS) in estrogen receptor (ER)-
negative breast cancer patients and that the association
with poor prognosis remains, even with elevated co-
expression of a BMP in the tumor. lines isolated from the same spontaneously formed
tumor [13, 14]. All the cell lines form primary tumors
when injected into the fat pad of BALB/c mice, but they
differ widely in their ability to form distant metastasis. KM plotter KM plotter (http://kmplot.com/analysis/) is an online
tool for examining prognostic markers in breast cancer
subtypes, which utilizes
data from multiple
cDNA
microarray experiments [20, 21]. Survival analysis (RFS
or OS) was done using the KM plotter database 2017
version. High and low expression were defined as above
(hazard ratio (HR) > 1.2, p-value < 0.05) and below
(HR < 0.83, p-value < 0.05) median, except for the OS
analysis where best cutoff was used. Gene co-expression network analysis p
y
RNA-seq data were downloaded from the study by Cir-
iello et al. in Cell, 2015 [19] from The Cancer Genome
Atlas (TCGA; http://cancergenome.nih.gov). The co-
hort consists of 817 cases of breast cancer with 17,214
genes, normalized within samples to the median gene
expression. Due to missing values, only 421 cases and
16,749 genes were available for download and analysis. To assess the association between GREM1 and other
genes, an ego-centric network with GREM1 in the cen-
ter was created using sample Pearson correlation. The
sample Pearson correlation represents the similarity be-
tween GREM1 and the corresponding gene, in terms of
linear correlation. The correlations range from −1 to 1,
where a value of 1 represents a total positive linear cor-
relation, e.g. similar expression values of GREM1 and
the corresponding gene, −1 represents a total negative
linear correlation, e.g. opposite expression values of
GREM1 and the corresponding gene. A value near zero
represents no linear correlation, e.g. no linear associ-
ation between the expression values of GREM1 and the
corresponding gene. Orthotopic mouse tumors and in vivo lung colonization
assay y
For all experiments, female mice (8–11 weeks old, Jan-
vier Labs, France) were used. For orthotopic tumors,
mice were anaesthetized and injected with 5 × 105 viable
cells resuspended in PBS into the fourth mammary fat
pad. When palpable, tumor size was measured twice
weekly using electronic calipers. Tumor volume was cal-
culated: VT = (length x width2) / 2. NT controls and
gremlin1 depleted cell lines were injected into nude mice
(BALB/cAnNRj-Foxn1nu/nu), n = 10 per cell line. Mice
were sacrificed after 21 days, unless otherwise stated. Weight of primary tumors and lungs was recorded. For
the in vivo lung colonization assay female mice were
anaesthetized and injected with 5 × 105 cells/100 μl PBS
in the lateral tail vein. Mice were monitored daily and
sacrificed 15 days after injection. Entire lungs were
weighed. Cell culture and generation of stable cell lines Cell culture and generation of stable cell lines
The mouse mammary tumor cell lines 67NR and 66cl4
cell lines were obtained from Barbara Ann Karmanos
Cancer Institute, Detroit, MI, USA. The cells were cul-
tured in DMEM (Lonza, BioWhittaker, BE12- 604F)
supplemented with 10% fetal calf serum (FCS, Thermo
Fischer
Scientific,
#10270–106),
2 mM L-Glutamine
(Lonza Group, De-17-605E) and 50 U/ml penicillin-
streptomycin
(ThermoFischer
Scientific,
Gibco,
#15070–063). Primary umbilical vein endothelial cells,
normal, human (HUVEC) were obtained from ATCC
(ATCC-PCS-100-010). The
cells
were
cultured
in
Medium 200 (Gibco, M200500) supplemented with low
serum growth supplement (LSGS) (Gibco, S00310). All
cells were incubated at 37 °C with 5% CO2. The immunocompetent mouse model for metastatic
mammary carcinoma called 4T1 consists of several cell Page 3 of 17 Neckmann et al. Cell Communication and Signaling (2019) 17:140 Page 3 of 17 Neckmann et al. Cell Communication and Signaling CRISPR/Cas9-GREM1
knockouts
and
CRISPR/Cas9
vector control cell lines were generated by viral transduc-
tion. The mouse lentiviral particles were purchased from
Sigma Aldrich (CRISPR/Cas9-GREM1 Mm0000325632
and
Mm0000325634;
CRISPR/Cas9-NT,
CRISPR12V-
1EA). 6 h after seeding, 66cl4 cells were infected with
lentiviral particles (MOI 0.1) in medium containing hexa-
dimethrine bromide (8 μg/ml). After 24 h cells were split
1:17. Starting 48 h after infection, cells were selected with
puromycin (3.25 μg/ml) for 1 week. The selection medium
was replaced every 2–3 days. Single cell colonies were
picked using cloning cylinders and tested for gremlin1
expression levels by western blot. The CRISPR/Cas9 NT_
mix was generated by trypsinizing a plate containing thou-
sands of puromycin resistant colonies. STED confocal microscope. Confocal stacks were proc-
essed for maximum intensity projections with Image J. Brightness and contrast of images were adjusted with
Adobe Photoshop CS6. For each cell line, at least fifty
embryos were analyzed. Zebrafish xenograft model g
The transgenic fish line Tg(fli1:GFP) was used [16, 17]. The tumor cell lines were fluorescently labelled with
mCherry using plv-mCherry lentiviruses. G418 was used
as selection marker. Embryo preparation and tumor cell
implantation was done as previously described [18]. Briefly, Tg(fli1:GFP) zebrafish embryos were dechorio-
nated 2 days post fertilization. Single cell suspensions of
66cl4 wildtype cells and variants were re-suspended in
PBS and kept at 4 °C until transplantation. Cell suspen-
sions were loaded into borosilicate glass capillary needles
(1 mm O.D. × 0.78 mm I.D.; Harvard Apparatus) and
injected into the duct of Cuvier (DoC) using a Pneu-
matic Picopump and a manipulator (WPI, Stevenage,
UK). After injection of approximately 400 cells, zebrafish
embryos were maintained at 33 °C. Invasive clusters were
quantified 6 days post-implantation (dpi). Zebrafish em-
bryos were fixated with 4% paraformaldehyde at 4 °C
overnight and were imaged in PBS using a Leica SP5 Genetic alterations in BMP-antagonists are common in
invasive breast cancer Cancer cells’ ability to metastasize may be affected by
factors that are secreted by the cancer cells (autocrine)
or by factors secreted by cells in the tumor microenvir-
onment (TME) (paracrine). In search of factors that pro-
mote metastasis, we therefore compared the expression
of genes affecting the TME in the non-metastatic 67NR
and the metastatic 66cl4 cells from the 4T1 mouse
model. Using RNA-sequencing (RNA-seq), we compared
the transcriptomes of cells grown in culture and isolated
from primary tumors in mice and we focused on tran-
scripts
encoding
matrisome-associated
extracellular
matrix (ECM) affiliated proteins (164 genes) and se-
creted factors (363 genes) as defined by Naba et al. [30]. We detected 220 of these transcripts in either cell cul-
tures or primary tumors of 66cl4 or 67NR (cut-off: 1
fragment per kilobase of mRNA per million mapped
reads [FPKM]). Of these, 28 genes were significantly
overexpressed in both cells and primary tumors in 66cl4
compared to 67NR (cut-off: fold-change ≥1.5, p-value
≤0.05) (Fig. 1a). We then analyzed if increased mRNA
expression levels of any of these 28 secreted factors cor-
related with reduced RFS in breast cancer patients. Using KM plotter [20, 21], we found that high expres-
sion of GREM1 correlated with RFS in all breast cancer
cases (HR = 1.32 (1.18–1.47), p-value = 6.9e−07) and ER- Increased gremlin1 activity has been shown to promote
epithelial-to-mesenchymal transition (EMT) and main-
tenance of stem cell-like properties in glioma and
colorectal cancer cells [9, 10, 31]. In addition, other
BMP-antagonists, including Dand5 [8] and noggin [32],
have been linked to tumor progression and metastasis. The RNA-seq analysis revealed that also other BMP-
antagonists were expressed (> 1 FPKM) and significantly
upregulated either in cell culture or primary tumors in
66cl4 compared to 67NR (Fig. 2a and Additional file 4:
Table S2). Furthermore, the RNA-seq analysis showed a
significant upregulation of the two inhibitory SMADs,
Smad6 and Smad7, in 66cl4 (Fig. 2a). Thus, BMP activ-
ity may be more repressed in the metastatic 66cl4 cells
than the non-metastatic 67NR cells, both by extracellu-
lar antagonists and intracellular inhibitors. Consistently,
we found increased SMAD4-related gene expression in
67NR by gene enrichment analysis (Enrichr [24, 25]) of
the 1252 transcripts that were significantly upregulated
in the 67NR cells and primary tumors compared with
66cl4 (Fig. 2b). Statistics Statistical analyses were performed in GraphPad Prism
7. Values are expressed as mean ± standard deviation
(SD), or mean ± standard error of the mean (SEM). Stat-
istical analyses were performed by paired 2-tailed Stu-
dent’s t-test after log-transformation (*0.01 < P < 0.05,
**0.001 < P < 0.01, *** P < 0.001). Gene enrichment analysis Gene enrichment analysis negative
subtype
(HR = 1.51
(1.2–1.9),
p-value =
0.00037), whereas in the ER-positive subtype there was a
significant correlation with RFS, but the HR was just
below
cut-off
(HR = 1.19
(1.01–1.4),
p-value = 0.035)
(Fig. 1b). Of note, a poorer overall survival (OS) was also
seen in the patients with the highest levels of GREM1
mRNA in tumor biopsies (Additional file 2: Figure S1). We confirmed overexpression of Grem1 mRNA levels in
66cl4 by quantitative PCR (Q-PCR) (Fig. 1c). Moreover,
western blot and ELISA analysis showed that gremlin1
was also highly expressed on protein level and was se-
creted by 66cl4, but not 67NR (Fig. 1d and e). Surpris-
ingly, Bmp4, which is one of the three preferred ligands
that gremlin1 binds and inhibits, was also among the 28
upregulated secreted factors in the metastatic 66cl4 cells
compared to the non-metastatic 67NR cells (Fig. 1a). Overexpression of Bmp4 in 66cl4 was validated by Q-
PCR (Fig. 1f) and western blot (Fig. 1g). However, be-
sides GREM1, high mRNA levels of BMP4 or any of the
other 28 secreted factors that were upregulated in the
metastatic 66cl4 cells did not correlate with reduced RFS
(Additional file 3: Table S1). On the contrary, a reduced
RFS correlated with low mRNA levels for some of these
genes. We therefore wanted to study the role of grem-
lin1 in aggressive tumor development and metastasis
more carefully. y
Gene enrichment analyses were done with the online tools
Enrichr (http://amp.pharm.mssm.edu/Enrichr/) [24, 25] for
the 66cl4/67NR transcriptome data (Fig. 2b) and Reactome
(https://reactome.org/) [26] for the TCGA cohort (Fig. 3b). Co-expression analysis Gene co-expression analysis was performed using the
online gene co-expression search engine SEEK (Search-
Based Exploration of Expression Compendium [Hu-
man]) (http://seek.princeton.edu/) [27] searching 331
breast cancer patient data sets. In addition, both Expres-
sion Atlas (https://www.ebi.ac.uk/gxa) [28] and CCLE
(Broad Institute Cancer Cell Line Encyclopedia, https://
portals.broadinstitute.org/ccle) [29] was used for gene
expression analysis in human breast cancer cell lines. Results Grem1 is highly expressed in the metastatic 66cl4 cells
and elevated GREM1 in tumor biopsies correlates with
reduced relapse-free survival in patients cBioPortal cBioPortal
cBioPortal
(http://www.cbioportal.org/)
is
an
open-
access database that allows visualization and analysis of
large-scale cancer genomics data sets [22, 23]. Our ana-
lyses utilize the OncoPrints visualization to identify ge-
nomic alterations, including somatic mutations, mRNA
expression and amplifications across a set of cases. This
visualization shows the genes as rows, while individual
cases are shown as columns. For this analysis we used
the TCGA Provisional data set for invasive breast car-
cinoma, and selected mutations, putative copy-number
alterations and mRNA expression as genomic profiles. Page 4 of 17 Neckmann et al. Cell Communication and Signaling (2019) 17:140 Neckmann et al. Cell Communication and Signaling Genetic alterations in BMP-antagonists are common in
invasive breast cancer To assess if our finding in the mouse
model could have clinical relevance, we searched the
TCGA provisional data set for invasive breast carcinoma
in the online tool cBioPortal [22, 23] for gene alterations
in BMP-antagonists and SMADs in invasive breast Neckmann et al. Cell Communication and Signaling (2019) 17:140 Page 5 of 17 Fig. 1 (See legend on next page.) Fig. 1 (See legend on next page. Fig. 1 (See legend on next page.) Neckmann et al. Cell Communication and Signaling (2019) 17:140 Page 6 of 17 (See figure on previous page.)
Fig. 1 Grem1 is highly expressed in 66cl4 and GREM1 correlates with RFS in breast cancer patients. (a) Scatter plot of in vivo versus in vitro
differential expression of mRNAs encoding secreted proteins in 66cl4 and 67NR. A positive number indicates higher expression in 66cl4, whereas
a negative number indicates higher expression in 67NR. The 28 factors were significantly overexpressed in both 66cl4 cells and 66cl4 primary
tumors (cut-off: fold-change ≥1.5, p-value ≤0.05) are indicated by name. (b) Relationship between GREM1 gene expression and RFS in breast
cancer patients using KM plotter. High and low expression were defined as above and below median. (c) Q-PCR analysis of Grem1 mRNA
expression in 67NR and 66cl4 cell lines in vitro (n = 3). Fold change is relative to Actb and Tbp. Results are shown as mean ± SEM. Student’s t-test,
**0.001 < P < 0.01. (d) Representative gremlin1 immunoblot (n = 4) of whole lysates of the 67NR and 66cl4 cell lines without or with PTI for 6 h. (e)
Levels of gremlin1 in conditioned medium from 67NR and 66cl4 (n = 3) measured by ELISA. Results are shown as mean ± SEM. (f) Q-PCR analysis
of Bmp4 mRNA expression in 67NR and 66cl4 cell lines in vitro (n = 3). Fold change is relative to Actb and Tbp. Results are shown as mean ± SEM. Student’s t-test, **0.001 < P < 0.01. (g) Representative BMP4 immunoblot (n = 4) of whole lysates of the 67NR and 66cl4 cell lines without or with
PTI for 6 h phenotype. This phenotype is associated with expres-
sion of genes that modulate ECM and allows for
tumor cell motility and invasion [34]. We hypothe-
sized that gremlin1 is involved in modulation of ECM
and that GREM1 expression correlates with expres-
sion of genes known to be involved in this process in
patients. Genetic alterations in BMP-antagonists are common in
invasive breast cancer To test this, we performed a co-expression
network
analysis
of
RNA-seq
data
of
421
breast
cancer cases from TCGA (study by Ciriello et al.,
Cell, 2015) [19]. The correlation values between the
16,749 genes tested and GREM1 were in the range
−0.32 - 0.72. An ego-centric network was made by the
50 top-scoring transcripts with positive correlations [range
0.59–0.72], meaning that these genes and GREM1 have
similar expression patterns across samples (Fig. 3a). The
location of the genes in the network represents the
strength of association, where the ones closest to GREM1
correlate best. Gene enrichment analysis of the same 50
top-scoring genes using the Reactome pathway database
[26]
demonstrated
that
genes
involved
in
ECM
organization, collagen formation, and collagen biosyn-
thesis and modification were enriched (Fig. 3b). These
results were confirmed by the gene co-expression search
engine SEEK [27], searching 331 breast cancer data sets
(Additional file 6: Table S4 and Additional file 7: Table
S5). Combined, the data suggest that gremlin1 may be
involved in ECM modification, which favor tumor cell
migration and intravasation, leading to increased metasta-
sis and consequently poor prognosis. cancer patient samples. Since there is no complete list of
existing BMP-antagonists, we chose to include the most
recognized extracellular antagonists, as well as the two
membrane-bound antagonists, BAMBI and CRIM1, in
this comparison [7, 33]. Interestingly, we found genetic
changes (mutations, amplifications, deletions or altered
mRNA levels) in the selected genes in as much as 62%
(53% if the SMADs are omitted) of the 960 samples
analyzed (Fig. 2c). The data indicate that deregulation of
BMP-antagonists in patients can happen via different
mechanisms. For instance, GREM2 is commonly ampli-
fied in breast cancer patients, indicating that elevated
levels of GREM2 origins from the cancer cells and not
the surrounding tissue. In comparison, increased levels
of GREM1 mRNA (found in 6% of the tumor biopsies)
are rarely caused by amplifications. Thus, elevated
GREM1 in breast cancer tumors may either originate
from normal cells in the tumor microenvironment or
from the tumor cells themselves. We then asked if any of the extracellular antagonists be-
sides GREM1, or inhibitory SMADs, could predict prog-
nosis of breast cancer patients. In all breast cancer cases,
elevated levels of GREM1 and SMAD6 correlated with
poor prognosis (Fig. 1b and Additional file 5: Table S3). Genetic alterations in BMP-antagonists are common in
invasive breast cancer In
contrast, low mRNA levels of six BMP-antagonists (CER1,
CHRD, CHRDL1, CRIM1, DAND5, and FST) correlated
with reduced RFS. Moreover, low mRNA levels of
BMPER, CHRDL1, CRIM1 and SOSTDC1 correlated with
reduced RFS in ER-positive breast cancer patients and
high mRNA levels of CRIM1, GREM1 and SMAD6 corre-
lated with reduced RFS in ER-negative breast cancer pa-
tients. In summary, these data indicate that in contrast to
other BMP-antagonists, high GREM1 expression is par-
ticularly relevant for aggressive tumor development in
breast cancer patients. Grem1 depletion might impair 66cl4’s ability to form
metastases in the lungs To determine if gremlin1 expression is important for
tumor formation and progression, we used CRISPR/
Cas9 to generate gremlin1 depleted 66cl4 cells. Pro-
tein expression levels were assessed by immunoblot
analyses (Fig. 4a). Five clones were selected for fur-
ther studies that all displayed an 80–100% reduction
in gremlin1 protein levels compared to cells harboring
a non-targeting (NT) guide sequence. The gremlin1
depleted
clones
showed
no
consistent
change
in
growth rate or ability to form colonies in soft agar Gene co-expression network analysis identifies genes
involved in collagen formation and ECM organization
associated with GREM1 A feature of metastatic cells is the ability to undergo
EMT, a process where the tumor cells lose their epi-
thelial
characteristics
and
gain
a
mesenchymal Neckmann et al. Cell Communication and Signaling (2019) 17:140 Page 7 of 17 Fig. 2 BMP-antagonists are genetically altered in a high frequency of breast cancer biopsies. (a) Heat-map showing the RNA-seq expression levels
of 11 BMP-antagonists (cut-off: expression level ≥1 in either in vitro cultivated cells or tumors of 67NR and 66cl4). Values are given in fragments
per kilobase of transcripts per million fragments mapped (FPKM). (b) Gene enrichment analysis of the 1252 genes significantly upregulated in
67NR cells and primary tumors using the Enrichr online tool (TRANSFAC and JASPAR position weight matrices). (c) Genetic alterations in
extracellular, intracellular and transmembrane BMP-antagonists, as well as BMP2, BMP4 and BMP7 were analyzed using cBioPortal in 960 patient
samples from The Cancer Genome Atlas (TCGA) provisional data set for invasive breast carcinoma Fig. 2 BMP-antagonists are genetically altered in a high frequency of breast cancer biopsies. (a) Heat-map showing the RNA-seq expression levels
of 11 BMP-antagonists (cut-off: expression level ≥1 in either in vitro cultivated cells or tumors of 67NR and 66cl4). Values are given in fragments
per kilobase of transcripts per million fragments mapped (FPKM). (b) Gene enrichment analysis of the 1252 genes significantly upregulated in
67NR cells and primary tumors using the Enrichr online tool (TRANSFAC and JASPAR position weight matrices). (c) Genetic alterations in
extracellular, intracellular and transmembrane BMP-antagonists, as well as BMP2, BMP4 and BMP7 were analyzed using cBioPortal in 960 patient
samples from The Cancer Genome Atlas (TCGA) provisional data set for invasive breast carcinoma However, all cell lines formed tumors in immuno-
compromised, nude mice. Compared to control cells,
the Grem1 depleted clones Grem1_1.3 and Grem1_2.1
formed smaller primary tumors and lost their ability
to metastasize to the lungs (Fig. 4b-d). In addition to
the non-target cell line NT_mix, two control single cell
clones (NT_1 and NT_2) were analyzed in vivo. Surpris-
ingly, these two control sub-clones did not behave like compared
to
the
pooled
66cl4
NT
control
cells
(Additional file 8: Figure S2A and B). Next, we ana-
lyzed the ability of two randomly selected clones to
form primary tumors and lung metastases after injec-
tion into the fat pad and tail vein of nude mice,
respectively. Gene co-expression network analysis identifies genes
involved in collagen formation and ECM organization
associated with GREM1 Unfortunately, none of the Cas9 express-
ing clones could form primary tumors in immuno-
competent BALB/c mice, irrespective of gene editing. Neckmann et al. Cell Communication and Signaling (2019) 17:140 Page 8 of 17 Fig. 3 (See legend on next page.) Fig. 3 (See legend on next page.) Fig. 3 (See legend on next page.) (See figure on previous page.)
Fig. 3 GREM1 expression is associated with genes involved in collagen formation and extracellular matrix (ECM) organization. Co-expression
analysis was done on RNA-seq data from 421 breast cancer cases downloaded from the TCGA (study by Ciriello et al., Cell, 2015) using Pearson
correlation coefficient. (a) Depicts the 50-top scoring genes in an ego-centric network and (b) shows the results of the gene enrichment analysis
of these genes using the Reactome pathway database
Neckmann et al. Cell Communication and Signaling (2019) 17:140
Page 9 of 17 Neckmann et al. Cell Communication and Signaling (2019) 17:140 Page 9 of 17 (See figure on previous page.)
Fig. 3 GREM1 expression is associated with genes involved in collagen formation and extracellular matrix (ECM) organization. Co-expression
analysis was done on RNA-seq data from 421 breast cancer cases downloaded from the TCGA (study by Ciriello et al., Cell, 2015) using Pearson
correlation coefficient. (a) Depicts the 50-top scoring genes in an ego-centric network and (b) shows the results of the gene enrichment analysis
of these genes using the Reactome pathway database (
g
p
p g )
Fig. 3 GREM1 expression is associated with genes involved in collagen formation and extracellular matrix (ECM) organization. Co-expression
analysis was done on RNA-seq data from 421 breast cancer cases downloaded from the TCGA (study by Ciriello et al., Cell, 2015) using Pearson
correlation coefficient. (a) Depicts the 50-top scoring genes in an ego-centric network and (b) shows the results of the gene enrichment analysis
of these genes using the Reactome pathway database 66cl4 wildtype or the NT_mix control cells, but rather dis-
played impaired primary tumor growth and metastases
formation in the lungs similar to the gremlin1 depleted
clones. From this, it is apparent that sub-clones origi-
nating from single 66cl4 cells have different ability to form
primary and secondary tumors without any further manipulation. Nonetheless, gremlin1 is overexpressed in
the metastatic 66cl4 versus the non-metastatic 67NR cells. Gene co-expression network analysis identifies genes
involved in collagen formation and ECM organization
associated with GREM1 This, and the finding that GREM1 is frequently over-
expressed in breast cancers and correlates with poor
prognosis, merge on the notion that gremlin1 is important
for aggressive breast cancer development. Fig. 4 Gremlin1 depletion impairs metastases formation in the lungs. (a) Analysis of CRISPR/Cas9-mediated knockout efficiency in 66cl4. Representative immunoblot of gremlin1 of whole lysates of 66cl4 non-target controls and gremlin1 knockout clones. (b-c) In vivo analysis of
66cl4 non-target control and two gremlin1 knockout clones (10 mice per group). Tumor cells were injected into the fat-pad of nude mice. (b)
The size of the primary tumors was measured regularly during the 21 days of experiment and (c) tumor weight was recorded after the mice had
been sacrificed. The line indicates mean tumor weight. Student’s t-test, *0.01 < P < 0.05. (d) Tumor cells were injected into the tail vein of nude
mice. After 15 days the mice were sacrificed and the lung weight was recorded. The line indicates mean lung weight. Student’s t-test,
**0.001 < P < 0.01, *** P < 0.001 Fig. 4 Gremlin1 depletion impairs metastases formation in the lungs. (a) Analysis of CRISPR/Cas9-mediated knockout efficiency in 66cl4. Representative immunoblot of gremlin1 of whole lysates of 66cl4 non-target controls and gremlin1 knockout clones. (b-c) In vivo analysis of
66cl4 non-target control and two gremlin1 knockout clones (10 mice per group). Tumor cells were injected into the fat-pad of nude mice. (b)
The size of the primary tumors was measured regularly during the 21 days of experiment and (c) tumor weight was recorded after the mice had
been sacrificed. The line indicates mean tumor weight. Student’s t-test, *0.01 < P < 0.05. (d) Tumor cells were injected into the tail vein of nude
mice. After 15 days the mice were sacrificed and the lung weight was recorded. The line indicates mean lung weight. Student’s t-test,
**0.001 < P < 0.01, *** P < 0.001 Fig. 4 Gremlin1 depletion impairs metastases formation in the lungs. (a) Analysis of CRISPR/Cas9-mediated knockout efficiency in 66cl4. Representative immunoblot of gremlin1 of whole lysates of 66cl4 non-target controls and gremlin1 knockout clones. (b-c) In vivo analysis of
66cl4 non-target control and two gremlin1 knockout clones (10 mice per group). Tumor cells were injected into the fat-pad of nude mice. Gene co-expression network analysis identifies genes
involved in collagen formation and ECM organization
associated with GREM1 (b)
The size of the primary tumors was measured regularly during the 21 days of experiment and (c) tumor weight was recorded after the mice had
been sacrificed. The line indicates mean tumor weight. Student’s t-test, *0.01 < P < 0.05. (d) Tumor cells were injected into the tail vein of nude
mice. After 15 days the mice were sacrificed and the lung weight was recorded. The line indicates mean lung weight. Student’s t-test,
**0.001 < P < 0.01, *** P < 0.001 Page 10 of 17 Page 10 of 17 Neckmann et al. Cell Communication and Signaling (2019) 17:140 breast cancer, a mesenchymal CD44hiCD24low phenotype
is suggested to characterize the cancer stem cells [34]. However, of the stem cell markers analyzed in 66cl4,
Cd24a was the most clearly elevated. Flow cytometry ana-
lysis of CD24 expression on 66cl4 and 67NR in vitro cul-
tured cells confirmed the results of the RNA-seq analysis
(Fig. 5b). Cd44, on the other hand, showed significantly
lower expression in 66cl4 compared to 67NR (Fig. 5a and
Additional File 9: Table S6). Thus, the expression of these
two markers indicate the opposite of expected for a classic
stem cell-like metastatic phenotype (CD44hiCD24low) in
66cl4. This is consistent with previous studies showing
that the metastatic potential of the 4T1 cell line is much
stronger than the non-metastatic 67NR, even though 4T1
is characterized by an epithelial phenotype and the 67NR
cells are more differentiated towards a mesenchymal 66cl4 displayed increased expression of stem cell markers
The first step in metastasis is invasion of tumor cells into
the surrounding stroma. Cancer cells of epithelial origin
acquire the ability to invade as a part of the EMT, during
which they reverse to an undifferentiated, stem cell-like
state [12]. Since BMP-antagonists, such as gremlin1, can
interfere with differentiation, we were interested in the dif-
ferentiation status of the metastatic 66cl4 and the non-
metastatic 67NR cells in the initial RNA-seq experiment. Analyzing the transcripts of 13 expressed stem cell
markers, we found that four markers were significantly
upregulated in 66cl4 cells and primary tumors (Cd24a,
Krt8, Itgb4 and Esr1), compared to 67NR cells and 67NR
tumors (Fig. 5a and Additional file 9: Table S6). On the
other side, none of the stem cell markers analyzed were
significantly upregulated in 67NR compared to 66cl4. In Fig. 5 High Grem1 levels correlates with high expression of stem cell markers in 66cl4. Gene co-expression network analysis identifies genes
involved in collagen formation and ECM organization
associated with GREM1 (a) Heat-map showing the RNA-Seq expression levels of
13 known stem cell markers (cut-off: expression level ≥1 in either cells or tumors of 67NR and 66cl4). Values are given in fragments per kilobase
of transcripts per million fragments mapped (FPKM). (b) Representative histogram of 67NR and 66cl4 in vitro cultured cells stained with CD24
anti-mouse antibody. (c) Relationship between CD24 gene expression and RFS in breast cancer patients using KM plotter. (d) Relationship
between combined CD24 and GREM1 expression and RFS in breast cancer patients using KM plotter Fig. 5 High Grem1 levels correlates with high expression of stem cell markers in 66cl4. (a) Heat-map showing the RNA-Seq expression levels of
13 known stem cell markers (cut-off: expression level ≥1 in either cells or tumors of 67NR and 66cl4). Values are given in fragments per kilobase
of transcripts per million fragments mapped (FPKM). (b) Representative histogram of 67NR and 66cl4 in vitro cultured cells stained with CD24
anti-mouse antibody. (c) Relationship between CD24 gene expression and RFS in breast cancer patients using KM plotter. (d) Relationship
between combined CD24 and GREM1 expression and RFS in breast cancer patients using KM plotter Fig. 5 High Grem1 levels correlates with high expression of stem cell markers in 66cl4. (a) Heat-map showing the RNA-Seq expression levels of
13 known stem cell markers (cut-off: expression level ≥1 in either cells or tumors of 67NR and 66cl4). Values are given in fragments per kilobase
of transcripts per million fragments mapped (FPKM). (b) Representative histogram of 67NR and 66cl4 in vitro cultured cells stained with CD24
anti-mouse antibody. (c) Relationship between CD24 gene expression and RFS in breast cancer patients using KM plotter. (d) Relationship
between combined CD24 and GREM1 expression and RFS in breast cancer patients using KM plotter Neckmann et al. Cell Communication and Signaling (2019) 17:140 Neckmann et al. Cell Communication and Signaling (2019) 17:140 Neckmann et al. Cell Communication and Signaling (2019) 17:140 Page 11 of 17 Page 11 of 17 To test if the impaired metastatic potential of 66cl4
GREM1_1.3 and GREM1_2.1 in vivo was due to a re-
duced capability to extravasate, we repeated the in vitro
extravasation assay with these two gremlin1 depleted cell
lines. We found that both clones had a slightly reduced
ability to penetrate the HUVEC monolayer compared to
the NT control cells (Fig. 6c). Extravasation capacity is only marginally affected by
gremlin1 depletion p
The finding that Bmp4 and its antagonist Grem1 were co-
expressed in the 66cl4 metastatic cell line puzzled us. We
therefore looked for co-expression of GREM1 and its pre-
ferred BMPs, BMP2, BMP4 or BMP7, in human breast
cancer cell lines using Expression Atlas [28]. Interestingly,
all the 13 breast cancer cell lines that expressed GREM1
co-expressed at least one of the three BMPs, with BMP4
being most commonly co-expressed (Additional file 11:
Figure S4). To explore the clinical significance of these
findings, we used the TCGA provisional data set for in-
vasive breast carcinoma in cBioPortal to analyze if BMP-
antagonists were co-expressed with either BMP2, BMP4,
or BMP7. We found that many of the biopsies with a
BMP amplification, also had an amplification of a gene en-
coding one of the BMP-antagonists (Fig. 2c). For BMP7,
several amplifications were also seen without a concomi-
tant amplification of any known antagonist. Since gremlin1 depletion might impair the metastatic
potential of 66cl4, we next asked if 66cl4 and 67NR also
display a different ability to extravasate and if gremlin1
has a role in controlling such differences. We first took
advantage of an in vitro approach, where the cancer cells
were scored for their ability to interfere with the integ-
rity of a confluent layer of human umbilical vein endo-
thelial
cells
(HUVECs)
measured
real
time
using
impedance [37]. After seeding the HUVECs, the imped-
ance increased until it reached a plateau. Visual inspec-
tion of the wells verified that the HUVECs had adhered
and that the cultures were confluent. A small volume of
culture medium containing the same number of 67NR
or 66cl4 cells was added to the respective wells and im-
pedance measured. The same volume of conditioned
medium (CM) from the two cell lines was used as a
negative control. Presence of 66cl4 cells caused a rapid
loss in impedance (Fig. 6a). In contrast, adding the same
number of 67NR cells resulted in a less prominent effect. These
data
suggest
that
the
metastatic,
gremlin1-
expressing 66cl4 cells are better in penetrating an endo-
thelial layer than 67NR. Interestingly, we did not see any
change in impedance when adding conditioned medium
of 66cl4 and 67NR to the HUVEC monolayer (Fig. 6b),
indicating that changes in the endothelial cell monolayer
are caused by direct interactions of the tumor cells with
the HUVECs. Gene co-expression network analysis identifies genes
involved in collagen formation and ECM organization
associated with GREM1 In summary, these data
suggest that the metastatic 66cl4 cells are more able to
penetrate an endothelial layer than the non-metastatic
67NR cells. However, depletion of gremlin1 in 66cl4
only marginally affected the extravasation ability. As an
alternative and more invasion relevant approach, we
employed a xenograft model using the transgenic zebra-
fish line Tg(fli1:GFP) [16–18]. After injecting fluores-
cently labelled cancer cells in the duct of Cuvier (DoC)
of zebrafish embryos at 48 h post fertilization, the move-
ment of cancer cells into the avascular tail fin area,
which is a measure for invasion, was documented using
a fluorescent microscope. Consistent with the findings of
the in vitro extravasation assay, 66cl4 wildtype cells and
NT_mix cells displayed similar numbers of invasion
clusters, whereas 66cl4 Grem1_1.3 formed fewer clusters
in the avascular tail fin area (Fig. 6d and e). phenotype [35]. It was also proposed that the 66cl4 cells
display a poorer ability to invade and migrate in vitro
compared to the 67NR cells, although 66cl4 metastasizes
to lungs in vivo [36]. To get a better overview of the
phenotypic profile, we used ChIP-X enrichment analysis
(ChEA) in the Enrichr database to analyze the 1270 genes
significantly upregulated in both 66cl4 cells and 66cl4 tu-
mors. The analysis showed activation of several signaling
pathways that are essential for stem cell maintenance, in-
cluding Oct4, Sox2, Sox9, and Nanog (Additional file 10:
Figure S3). Taken together, although the expression of
commonly used breast cancer stem cell markers showed
opposite expression than what would be expected, we
found increased expression of several stem cell markers in
66cl4 compared to 67NR, supporting a metastatic stem
cell-like
phenotype. Since
increased GREM1
mRNA
expression levels in tumors correlated with reduced RFS,
we asked if high CD24 mRNA levels also could predict
poor prognosis. Similar to GREM1, CD24 correlated with
reduced RFS in all breast cancer cases when highly
expressed (Fig. 5c). Of note, combining the expression
levels for GREM1 and CD24 did not give any additive
effect in prediction value, suggesting that their functions
are closely related (Fig. 5d). Extravasation capacity is only marginally affected by
gremlin1 depletion Since deregulated BMP activity has been linked to
invasion and metastasis [38], we wanted to explore the
clinical significance of elevated BMP levels for survival
in breast cancer patients. However, we found no corre-
lation between high expression level of BMP2, BMP4 or
BMP7 and prognosis, irrespective of sub-grouping (all
breast cancer cases, ER-positive subtype, or ER-negative
subtype (Table 1). Based on our findings that elevated
BMP expression does not correlate with reduced RFS,
whereas high GREM1 expression is associated with poor Neckmann et al. Cell Communication and Signaling (2019) 17:140 Neckmann et al. Cell Communication and Signaling Page 12 of 17 Fig. 6 Gremlin1 depletion may affect 66cl4’s ability to extravasate. (a-c) Analysis of the tumor cells ability to penetrate a confluent monolayer of
human umbilical vein-endothelial cells (HUVECs) using the xCELLigence Real-Time Cell Analysis (RTCA) Systems. Representative graphs depicting
the change in normalized cell index after addition of tumor cells or conditioned medium. Pure growth medium was used as a control. (a) 67NR
and 66cl4. (b) Conditioned medium of 67NR and 66cl4. (c) 66cl4 non-target control and two Grem1 knockout clones. Results are shown as
mean ± SD (A-C). (d) Representative image of 6 dpi zebrafish larvae showing the invasion of 66cl4 wildtype cells, non-target control (NT_mix) and
Grem1_1.3 (red). The vasculature is shown in green. (e) Quantification of invasive cluster numbers in 66cl4 injected zebrafish larvae. Probability
represents the percentage of zebrafish larvae in which tumor cell extravasation was observed. Results are shown as mean ± SEM. The line
indicates mean cluster number. Student’s t-test, *0.01 < P < 0.05 Fig. 6 Gremlin1 depletion may affect 66cl4’s ability to extravasate. (a-c) Analysis of the tumor cells ability to penetrate a confluent monolayer of
human umbilical vein-endothelial cells (HUVECs) using the xCELLigence Real-Time Cell Analysis (RTCA) Systems. Representative graphs depicting
the change in normalized cell index after addition of tumor cells or conditioned medium. Pure growth medium was used as a control. (a) 67NR
and 66cl4. (b) Conditioned medium of 67NR and 66cl4. (c) 66cl4 non-target control and two Grem1 knockout clones. Results are shown as
mean ± SD (A-C). (d) Representative image of 6 dpi zebrafish larvae showing the invasion of 66cl4 wildtype cells, non-target control (NT_mix) and
Grem1 1 3 (red) The vasculature is shown in green (e) Quantification of invasive cluster numbers in 66cl4 injected zebrafish larvae Probability Fig. Extravasation capacity is only marginally affected by
gremlin1 depletion 6 Gremlin1 depletion may affect 66cl4’s ability to extravasate. (a-c) Analysis of the tumor cells ability to penetrate a confluent monolayer of
human umbilical vein-endothelial cells (HUVECs) using the xCELLigence Real-Time Cell Analysis (RTCA) Systems. Representative graphs depicting
the change in normalized cell index after addition of tumor cells or conditioned medium. Pure growth medium was used as a control. (a) 67NR
and 66cl4. (b) Conditioned medium of 67NR and 66cl4. (c) 66cl4 non-target control and two Grem1 knockout clones. Results are shown as
mean ± SD (A-C). (d) Representative image of 6 dpi zebrafish larvae showing the invasion of 66cl4 wildtype cells, non-target control (NT_mix) and
Grem1_1.3 (red). The vasculature is shown in green. (e) Quantification of invasive cluster numbers in 66cl4 injected zebrafish larvae. Probability
represents the percentage of zebrafish larvae in which tumor cell extravasation was observed. Results are shown as mean ± SEM. The line
indicates mean cluster number. Student’s t-test, *0.01 < P < 0.05 Increased levels of GREM1 mRNA are associated with lack
of estrogen receptor expression High and low expression were defined as above and below median
Gene(s)
Gene ID
all BC patients
ER+ BC patients
ER- BC patients
HR
p-value
HR
p-value
HR
p-value
GREM1
218469_at
1.32 (1.18–1.47)
6.90E-07
1.19 (1.01–1.4)
0.035
1.51 (1.2–1.9)
0.00037
BMP2
205289_at
0.88 (0.79–0.98)
0.02
0.88 (0.75–1.04)
0.13
1.01 (0.81–1.26)
0.94
BMP4
211518_s_at
0.9 (0.81–1)
0.058
0.98 (0.83–1.16)
0.83
1.16 (0.93–1.46)
0.19
BMP7
209590_at
0.88 (0.79–0.98)
0.023
1.08 (0.92–1.27)
0.36
0.92 (0.73–1.15)
0.45
BMP2/GREM1
205289_at / 218469_at
1.33 (1.19–1.48)
3.70E-07
1.21 (1.03–1.42)
0.022
1.64 (1.31–2.07)
1.80E-05
BMP4/GREM1
211518_s_at / 218469_at
1.28 (1.15–1.43)
7.30E-06
1.2 (1.02–1.42)
0.026
1.63 (1.3–2.06)
2.30E-05
BMP7/GREM1
209590_at / 218469_at
1.29 (1.16–1.44)
4.10E-06
1.19 (1.01–1.4)
0.04
1.63 (1.3–2.06)
2.30E-05 Analyzing the metastatic 66cl4 and non-metastatic
67NR cell lines of the 4T1 mouse mammary tumor
model, we found that Grem1 was upregulated in 66cl4
cells and primary tumors. Consistent with the role of
gremlin1 as an antagonist of BMP-induced differen-
tiation, RNA-seq revealed that 66cl4 cells have elevated
expression of several stem cell markers compared to
67NR. In line with Goa et al. [8], who found that over-
expression of Dand5 in the 4TO7 mouse mammary
tumor cell line inhibits lung-derived BMPs and thereby
promotes metastases in the lungs, we observed a reduc-
tion in lung metastases after injection of 66cl4 Grem1
depleted cells into the tail vein of nude mice. However,
high levels of DAND5 in patients did not predict survival
and may be less relevant than GREM1 in breast cancer
development. Interestingly,
of
the
BMP-antagonists
tested here, only elevated expression of GREM1 and
CRIM1 correlated with poor prognosis in patients with
ER-negative tumors. cell lines lacked expression of estrogen receptor, ESR1,
mRNA (Fig. 7b). Moreover, in the TCGA provisional
dataset (cBioportal) there was no overlap between breast
cancer biopsies overexpressing GREM1 (n = 61) or ESR1
(n = 44) mRNA (data not shown), further supporting a
functional link between GREM1 expression and lack of
ER-signaling. Increased levels of GREM1 mRNA are associated with lack
of estrogen receptor expression Increased levels of GREM1 mRNA are associated with lack
of estrogen receptor expression Increased levels of GREM1 mRNA are associated with lack
of estrogen receptor expression prognosis in breast cancer, we hypothesized that com-
bining the expression levels of BMPs and GREM1 annuls
the observed correlation of GREM1 with RFS. Surpris-
ingly, this was not the case. We found that in ER-
negative breast cancer cases combining high expression
level of GREM1 with high expression of BMP2, BMP4 or
BMP7 showed a slightly poorer prognosis compared to
GREM1 alone (Table 1). Combined, these data demon-
strate that breast cancer cells may co-express BMPs and
GREM1, and that co-expression of BMP and GREM1 in
breast cancer biopsies does not reduce GREM1’s value
as a predictor of poor outcome. The prognostic value of GREM1 in ER-negative breast
cancers made us ask if there could be a link between
GREM1 and ER-signaling. We therefore analyzed cell
line
data
available
from
the
Cancer
Cell
Line
Encyclopedia (Broad Institute). We found that 10% (6
out of 60) breast cancer cell lines expressed elevated
GREM1 mRNA levels compared to the average of > 1000
different cell lines (Fig. 7a). The expression of GREM1
mRNA was not caused by copy number alterations in
the gene, and interestingly, all the GREM1 expressing Page 13 of 17 Neckmann et al. Cell Communication and Signaling (2019) 17:140 Page 13 of 17 Table 1 Combined elevated GREM1 and BMP expression level improves prognostic value in ER-negative breast cancer patients. Analysis of RFS in breast cancer patients using KM plotter. High and low expression were defined as above and below median Table 1 Combined elevated GREM1 and BMP expression level improves prognostic value in ER-negative breast cancer patients. Analysis of RFS in breast cancer patients using KM plotter. High and low expression were defined as above and below median Analysis of RFS in breast cancer patients using KM plotter. Discussion Transcript levels were determined using RNA-sequencing and compared to the average expression in more than 1000
human cancer cell lines for each transcript Page 14 of 17 Neckmann et al. Cell Communication and Signaling (2019) 17:140 Page 14 of 17 correlated with reduced relapse-free survival in ER-
negative breast cancer patients. Using publicly available
databases, we found that GREM1 is also expressed on
mRNA level by various human breast cancer cell lines
and that GREM1 is significantly upregulated in primary
tumor biopsies of breast cancer patients compared to
normal tissue samples. However, it remains unclear
which cells are the source of GREM1 in the tumor bi-
opsies. The gene alteration data (Fig. 2c) shows that
GREM1 mRNA is elevated in approximately 6% of the
breast cancer biopsies, and that amplifications are not
common for this gene. It was recently shown that
GREM1 can be produced by cancer-associated fibro-
blasts (CAFs) in breast cancer patients and that CAFs
are the main source of GREM1 in colorectal cancer tis-
sue [46, 47]. Of note, GREM1 was one of eight elevated
genes shown in cancer cells laser-dissected from inva-
sive breast carcinoma patients compared with cancer
cells from ductal carcinoma in situ patients [48], sup-
porting that GREM1 could be expressed by the trans-
formed cancer cells themselves and associated with an
invasive
phenotype. Taken
together,
the
elevated
GREM1 mRNA found in breast cancer biopsies could
either come from the cancer cells themselves, or from
cells infiltrating the tumor. combining high expression levels of GREM1 and BMP2,
BMP4 or BMP7. These data indicate that gremlin1 and
BMP may not counteract each other with respect to can-
cer progression. Since gremlin1 has been shown to bind
to cell surfaces [54], one could speculate that gremlin1,
secreted by tumor cells or surrounding cells, binds to
the surface of the transformed cancer cells and protects
them
from
BMP-induced
differentiation. On
the
other hand, BMPs could be further distributed in the
TME and thereby induce differentiation of immune
cells and stroma cells to promote tumor growth and
metastasis. In this study, we show that GREM1 mRNA in tumor
biopsies correlates with poor survival in ER-negative
breast cancer patients. We propose that there may be
an association between expression of GREM1 and lack
of expression of the estrogen receptor, ESR1, as we
only found GREM1 expression among the ESR1 nega-
tive tumor cells. Discussion Our findings are consistent with a
previous study showing that BMPs can counteract
estrogen-induced cell division and that estrogen treat-
ment led to downregulation of several BMP-receptors,
including BMPR1A [55]. BMPR1A is needed for proper
BMP2 and BMP4 activity and downregulating this re-
ceptor could be one way that ER-positive breast cancer
cells avoid BMP activity. We
speculate
that
ER-
negative cells may develop other mechanisms to avoid
BMP activity, for instance by producing their own
BMP antagonists or by modulating the microenviron-
ment to secrete factors that inhibit BMPs. We also ob-
served that 6% of the tumor biopsies had elevated
GREM1 mRNA levels (Fig. 2c) as opposed to 10% of
the breast cancer cell lines (Fig. 7). Since the source of
GREM1 in some of the tumor biopsies likely are CAFs,
the proportion of cell lines that make their own
GREM1 is relatively large and this may indicate that
GREM1 gives them a growth advantage that enrich for
such cells among the breast cancer cell lines. g
Among the 28 mRNAs encoding secreted proteins
that were significantly upregulated in 66cl4 cells and
tumors, we found not only Grem1, but also its ligand,
Bmp4. Like other members of the TGF-β superfamily,
BMP4
has
both
tumor-suppressing
and
tumor-
promoting roles in tumor cells [40]. On one side, it has
been demonstrated that BMP4 inhibits tumor cell
growth. On the other side, BMP4 has been shown to in-
crease cell migration and invasion of tumor cells. Re-
cently, it has been described that BMP4 might regulate
autophagy [49] and polarize macrophages towards an
anti-inflammatory or M2-like phenotype [50]. Using
web-based databases we analyzed mRNA expression
levels of GREM1 and BMP4 in breast cancer cell lines
and a breast cancer patient cohort (TCGA) and found
that these two genes are also co-expressed in some
tumor cell lines and primary tumors. In this way, we
confirmed a possible relevance of this unexpected co-
expression of both BMP4 and its antagonist gremlin1
specifically in the aggressive 66cl4 cells. Discussion Both activated BMP signaling [39–42] and elevated
levels of SMAD-inhibitors [43, 44] and BMP-antagonists
[8–10, 31, 32] have been linked to tumor progression
and metastasis. Moreover, amplification and/or upregula-
tion of BMPs as well as BMP-antagonists, including NOG,
GREM1, GREM2, and CHRD have been reported in breast
cancer [39, 45]. Yet, the simultaneous upregulation of
BMPs and their antagonists, and the fact that both pro-
mote tumor aggressiveness appears contradictory. In the
present study, we found that high expression of the BMP-
antagonist GREM1 correlates strongly with reduced RFS
in ER-negative breast cancer. g
The clinical relevance of our findings in the 4T1
mouse mammary tumor model is supported by several
observations. Most importantly, high GREM1 expression Fig. 7 Increased levels of GREM1 mRNA are associated with lack of estrogen receptor expression. Level of GREM1 transcript in 60 different human
breast cancer cell lines analyzed using the Cancer Cell Line Encyclopedia made available by Broad Institute. (a) GREM1 transcript levels were
elevated in some cell lines, but this did not correlate with GREM1 copy number variations. (b) Correlation of the GREM1 and estrogen receptor
(ERS1) transcript levels. Transcript levels were determined using RNA-sequencing and compared to the average expression in more than 1000
human cancer cell lines for each transcript Fig. 7 Increased levels of GREM1 mRNA are associated with lack of estrogen receptor expression. Level of GREM1 transcript in 60 different human
breast cancer cell lines analyzed using the Cancer Cell Line Encyclopedia made available by Broad Institute. (a) GREM1 transcript levels were
elevated in some cell lines, but this did not correlate with GREM1 copy number variations. (b) Correlation of the GREM1 and estrogen receptor
(ERS1) transcript levels. Transcript levels were determined using RNA-sequencing and compared to the average expression in more than 1000
human cancer cell lines for each transcript Fig. 7 Increased levels of GREM1 mRNA are associated with lack of estrogen receptor expression. Level of GREM1 transcript in 60 different human
breast cancer cell lines analyzed using the Cancer Cell Line Encyclopedia made available by Broad Institute. (a) GREM1 transcript levels were
elevated in some cell lines, but this did not correlate with GREM1 copy number variations. (b) Correlation of the GREM1 and estrogen receptor
(ERS1) transcript levels. Funding
Thi
k Funding
This work was supported by grants from the Norwegian Cancer Society
(project numbers: 63843 and 63825), and from the Research Council of
Norway through its Centres of Excellence funding program (project number:
223255/F50), by a grant from the Norwegian Women’s Public Health
Association (CW) and Cancer Genomics Center Netherlands (PtD). Additional file 5: Table S3. Relationship between gene expression of
BMP-antagonists and RFS in breast cancer patients. High and low
expression were defined as above (HR > 1.2, p-value < 0.05) and below
(HR < 0.83, p-value < 0.05) median. Additional file 6: Table S4. The 50 top-scoring genes that are co-
expressed with GREM1 in breast cancer. Co-expression analysis of the 50
top-scoring hits that are found co-expressed with GREM1 in a search of
331 breast cancer data sets in the SEEK database. Supplementary information Supplementary information
Supplementary information accompanies this paper at https://doi.org/10. 1186/s12964-019-0467-7. Acknowledgements y
Supplementary information accompanies this paper at https://doi.org/10. 1186/s12964-019-0467-7. We thank Unni Nonstad for excellent technical assistance. The RNA-
sequencing was provided by the Genomics Core Facility (GCF), Norwegian
University of Science and Technology (NTNU). GCF is funded by the Faculty
of Medicine and Health Sciences at NTNU and Central Norway Regional
Health Authority. We also thank the Comparative medicine Core Facility
(CoMed), NTNU, for the help regarding the mice experiments. CoMed is
funded by the Faculty of Medicine and Health Sciences at NTNU and Central
Norway Regional Health Authority. Parts of the results shown here are based
upon data generated by the TCGA Research Network: https://www.cancer. gov/tcga. Additional file 1: Supplementary methods: Transcriptome analysis,
Quantitative PCR, Immunoblotting, Conditioned medium, ELISA, Cell
proliferation assay, Soft-agar assay, Flow cytometry, and In vitro extravasa-
tion assay using xCELLigence Real-Time Cell Analysis (RTCA) Systems. Additional file 2: Figure S1. High GREM1 expression correlates with OS
in breast cancer patients. Relationship between GREM1 gene expression
and OS in breast cancer patients using KM plotter. High and low
expression was determined using best cut off. Additional file 2: Figure S1. High GREM1 expression correlates with OS
in breast cancer patients. Relationship between GREM1 gene expression
and OS in breast cancer patients using KM plotter. High and low
expression was determined using best cut off. Availability of data and materials
h
d
b
d b The transcriptome data obtained by sequencing mRNA isolated from cells
and primary breast tumors of 67NR and 66cl4 is accessible from NCBI
(https://www.ncbi.nlm.nih.gov/biosample, SRA accession PRJNA577616). Additional file 7: Table S5. GREM1 expression is associated with genes
involved in extracellular matrix (ECM) and collagen fibril organization. Gene enrichment analysis (GO Biological Process (BP) terms) of 50 top-
scoring hits that co-expressed with GREM1 using the SEEK database. T,
term size; A, Number of genes in the co-expressed gene set with annota-
tions in the functional database; A&T, size of overlap between the term’s
gene-set and the co-expressed gene set. Ethics approval The mice studies were approved by the National Animal Research
Authorities and carried out according to the European Convention for the
Protection of Vertebrates used for Scientific Purposes (FOTS ID 10049). The
zebrafish experiment was approved by The Institutional Committee for
Animal Welfare of the Leiden University Medical Center (LUMC). All
experiments were performed according to approved guidelines and
regulations. Additional file 8: Figure S2. In vitro analysis of CRISPR/Cas9-mediated
Grem1 knockouts in 66cl4. (A) Measurement of proliferation in culture
(n = 4). Results are shown as mean ± SEM. Student's t-test,
*0.01 < P < 0.05, *** P < 0.001. (B) Soft-agar assay. Colony area was
measured in pixels (n = 3). Results are shown as mean ± SEM. Additional file 9: Table S6. RNA-Seq expression levels of 13 known
stem cell markers. Expression level ≥1 in either cells or tumors of 67NR
and 66cl4. Values are given in fragments per kilobase of transcripts per
million fragments mapped (FPKM), as well as Log2 and p-values. Additional file 9: Table S6. RNA-Seq expression levels of 13 known
stem cell markers. Expression level ≥1 in either cells or tumors of 67NR
and 66cl4. Values are given in fragments per kilobase of transcripts per
million fragments mapped (FPKM), as well as Log2 and p-values. Authors’ contributions
d Additional file 3: Table S1. Genes encoding secreted proteins that are
significantly upregulated in 66cl4 and their prognostic value in breast
cancer patients. High and low expression were defined as above
(HR > 1.2, p-value < 0.05) and below (HR < 0.83, p-value < 0.05) median. Additional file 3: Table S1. Genes encoding secreted proteins that are
significantly upregulated in 66cl4 and their prognostic value in breast
cancer patients. High and low expression were defined as above
(HR > 1.2, p-value < 0.05) and below (HR < 0.83, p-value < 0.05) median. UN, CW, GB and TH prepared the manuscript, UN, CW, MH, EA, JR, GB, PtD
and TH collected the data, UN, CW, MH, EA, JR, SZ, BJ, RIS, GB, PtD and TH
analyzed the data, GB, PtD and TH did the conceptualization, UN, CW, MH,
EA, JR, SZ, BJ, RIS, GB, PtD and TH edited the manuscript. All authors read
and approved the final manuscript. Additional file 4: Table S2. RNA-Seq expression levels of BMP-
antagonists and SMADs. Expression level ≥1 in either cells or tumors of
67NR and 66cl4. Values are given in fragments per kilobase of transcripts
per million fragments mapped (FPKM), as well as Log2 and p-values. Additional file 4: Table S2. RNA-Seq expression levels of BMP-
antagonists and SMADs. Expression level ≥1 in either cells or tumors of
67NR and 66cl4. Values are given in fragments per kilobase of transcripts
per million fragments mapped (FPKM), as well as Log2 and p-values. Consent for publication
Not applicable. Additional file 10: Figure S3. Signaling pathways maintaining
stemness are activated in 66cl4. Using CHiP-X enrichment analysis (ChEA)
of the 1,270 genes significantly upregulated in both 66cl4 cells and 66cl4
tumors, we found activation of several signaling pathways that are
essential for stem cell maintenance. Additional file 10: Figure S3. Signaling pathways maintaining
stemness are activated in 66cl4. Using CHiP-X enrichment analysis (ChEA)
of the 1,270 genes significantly upregulated in both 66cl4 cells and 66cl4
tumors, we found activation of several signaling pathways that are
essential for stem cell maintenance. Competing interests
The authors declare that they have no competing interests. Author details
1 1Centre of Molecular Inflammation Research, Department of Clinical and
Molecular Medicine, Faculty of Medicine and Health Sciences, NTNU -
Norwegian University of Science and Technology, Trondheim, Norway. 2Department of Biomedical Laboratory Science, Faculty of Natural Sciences,
NTNU - Norwegian University of Science and Technology, Trondheim,
Norway. 3Clinic of Laboratory Medicine, St Olavs Hospital, Trondheim,
Norway. 4Department of Biotechnology and Food Science, Faculty of Natural
Sciences, NTNU - Norwegian University of Science and Technology,
Trondheim, Norway. 5K.G. Jebsen Center for Genetic Epidemiology,
Department of Public Health and General Practice, Faculty of Medicine and
Health Sciences, NTNU - Norwegian University of Science and Technology,
Trondheim, Norway. 6Department of Cell and Chemical Biology, Oncode
Institute, Leiden University Medical Center, Leiden, The Netherlands. 7Department of Molecular Oncology, Institute for Cancer Research, Oslo
University Hospital-Radiumhospitalet, Oslo, Norway. 8Department of
Informatics, University of Oslo, Oslo, Norway. 9Department of Clinical and
Molecular Medicine (IKOM), NTNU, Gastro Center, Prinsesse Kristinas gt 1,
7030 Trondheim, Norway. 10Department of Hematology, St. Olavs Hospital,
Trondheim, Norway. Additional file 11: Figure S4. GREM1 is co-expressed with BMPs in
several human breast cancer cell lines. Co-expression analysis of GREM1
and selected BMPs (BMP2, BMP4, and BMP7) in human breast cancer cell
lines using Expression atlas. Conclusions We find that gene expression of BMP-antagonists is
more frequently elevated than expression of inhibitory
SMADs in breast cancer biopsies. Elevated expression
level of GREM1 in tumor biopsies correlate with adverse
outcome irrespective of the expression level of the BMPs
in the biopsy. This suggests a dominant role for
gremlin1 secretion with respect to tumor progression. The predictive value of the gene expression level for
the different BMP-antagonists varies and suggests that
the antagonists are functionally different in ways that
must
be
better
understood
to
explore
these
as
possible drug targets against metastatic breast cancer
development. In addition to BMP4, also elevated levels of other
members of the BMP-family, including BMP6 [51, 52]
and BMP7 [53], have been linked to increased tumor
aggressiveness. Interestingly, in contrast to GREM1, we
found no correlation between high mRNA expression
level and poor prognosis for breast cancer patients for
any of the three BMPs tested. However, the correlation
between GREM1 expression levels and RFS in ER-
negative breast cancer cases was not annulled when Page 15 of 17 Page 15 of 17 Neckmann et al. Cell Communication and Signaling (2019) 17:140 Neckmann et al. Cell Communication and Signaling References Lamouille S, Xu J, Derynck R. Molecular mechanisms of epithelial-
mesenchymal transition. Nat Rev Mol Cell Biol. 2014;15(3):178–96. 14. Miller FR, Miller BE, Heppner GH. Characterization of metastatic
heterogeneity among subpopulations of a single mouse mammary tumor:
heterogeneity in phenotypic stability. Invasion & metastasis. 1983;3(1):22–31. 35. Marsan M, Van den Eynden G, Limame R, Neven P, Hauspy J, Van Dam
PA, Vergote I, Dirix LY, Vermeulen PB, Van Laere SJ. A core invasiveness
gene signature reflects epithelial-to-mesenchymal transition but not
metastatic potential in breast cancer cell lines and tissue samples. PLoS
One. 2014;9(2):e89262. 15. Church RH, Krishnakumar A, Urbanek A, Geschwindner S, Meneely J, Bianchi
A, Basta B, Monaghan S, Elliot C, Stromstedt M, et al. Gremlin1 preferentially
binds to bone morphogenetic protein-2 (BMP-2) and BMP-4 over BMP-7. The Biochemical journal. 2015;466(1):55–68. 36. Lou Y, Preobrazhenska O, Auf dem Keller U, Sutcliffe M, Barclay L, PC MD,
Roskelley C, Overall CM, Dedhar S. Epithelial-mesenchymal transition (EMT)
is not sufficient for spontaneous murine breast cancer metastasis. Developmental dynamics : an official publication of the American
Association of Anatomists. 2008;237(10):2755–68. 16. Lawson ND, Weinstein BM. In vivo imaging of embryonic vascular
development using transgenic zebrafish. Dev Biol. 2002;248(2):307–18. 17. He S, Lamers GE, Beenakker JW, Cui C, Ghotra VP, Danen EH, Meijer AH,
Spaink HP, Snaar-Jagalska BE. Neutrophil-mediated experimental metastasis
is enhanced by VEGFR inhibition in a zebrafish xenograft model. J Pathol. 2012;227(4):431–45. 37. Rahim S, Uren A. A real-time electrical impedance based technique to
measure invasion of endothelial cell monolayer by cancer cells. Journal of
visualized experiments : JoVE. 2011;50. 18. Ren J, Liu S, Cui C, Ten Dijke P. Invasive behavior of human breast
Cancer cells in embryonic Zebrafish. Journal of visualized experiments :
JoVE. 2017;122. 38. Ren J, Ten Dijke P. Bone Morphogenetic Proteins in the Initiation and
Progression of Breast Cancer. In: Vukicevic S, Sampath KT, editors. Bone
Morphogenetic Proteins: Systems Biology Regulators: Cham: Springer
International Publishing; 2017. p. 409–33. 19. Ciriello G, Gatza ML, Beck AH, Wilkerson MD, Rhie SK, Pastore A, Zhang H,
McLellan M, Yau C, Kandoth C, et al. Comprehensive molecular portraits of
invasive lobular breast Cancer. Cell. 2015;163(2):506–19. 39. Owens P, Pickup MW, Novitskiy SV, Giltnane JM, Gorska AE, Hopkins CR,
Hong CC, Moses HL. Inhibition of BMP signaling suppresses metastasis in
mammary cancer. Oncogene. 2015;34(19):2437–49. 20. Gyorffy B, Lanczky A, Eklund AC, Denkert C, Budczies J, Li Q, Szallasi Z. Abbreviations BMP: Bone morphogenetic protein; ChEA: ChIP-X enrichment analysis;
CM: Conditioned medium; DoC: Duct of Cuvier; Dpi: Days post implantation;
ECM: Extracellular matrix; EMT: Epithelial-to-mesenchymal transition;
ER: Estrogen receptor; FPKM: Fragment per kilobase of mRNA per million
mapped reads; GDF: Growth and differentiation factor; HR: Hazard ratio;
HUVECs: human umbilical vein endothelial cells; NT: Non-targeting;
OS: Overall survival; Q-PCR: Quantitative-PCR; RFS: Relapse free survival; RNA-
seq: RNA-sequencing; SD: Standard deviation; SEEK: Search-Based Exploration
of Expression Compendium [Human]; SEM: Standard error of the mean;
TCGA: The Cancer Genome Atlas; TGF-β: transforming growth factor-β;
TME: Tumor microenvironment Page 16 of 17 Page 16 of 17 Page 16 of 17 Neckmann et al. Cell Communication and Signaling (2019) 17:140 Neckmann et al. Cell Communication and Signaling (2019) 17:140 Received: 16 July 2019 Accepted: 22 October 2019 References Walsh DW, Godson C, Brazil DP, Martin F. Extracellular BMP-antagonist
regulation in development and disease: tied up in knots. Trends Cell Biol. 2010;20(5):244–56. 29. Ghandi M, Huang FW, Jane-Valbuena J, Kryukov GV, Lo CC, McDonald ER 3rd,
Barretina J, Gelfand ET, Bielski CM, Li H, et al. Next-generation characterization
of the Cancer cell line encyclopedia. Nature. 2019;569(7757):503–8. 8. Gao H, Chakraborty G, Lee-Lim AP, Mo Q, Decker M, Vonica A, Shen R, Brogi
E, Brivanlou AH, Giancotti FG. The BMP inhibitor coco reactivates breast
cancer cells at lung metastatic sites. Cell. 2012;150(4):764–79. 8. Gao H, Chakraborty G, Lee-Lim AP, Mo Q, Decker M, Vonica A, Shen R, Brogi
E, Brivanlou AH, Giancotti FG. The BMP inhibitor coco reactivates breast
cancer cells at lung metastatic sites. Cell. 2012;150(4):764–79. 30. Naba A, Clauser KR, Hoersch S, Liu H, Carr SA, Hynes RO. The matrisome: in
silico definition and in vivo characterization by proteomics of normal and
tumor extracellular matrices. Molecular & cellular proteomics : MCP. 2012;
11(4):M111.014647. 9. Yan K, Wu Q, Yan DH, Lee CH, Rahim N, Tritschler I, DeVecchio J, Kalady MF,
Hjelmeland AB, Rich JN. Glioma cancer stem cells secrete Gremlin1 to
promote their maintenance within the tumor hierarchy. Genes Dev. 2014;
28(10):1085–100. 31. Karagiannis GS, Musrap N, Saraon P, Treacy A, Schaeffer DF, Kirsch R, Riddell
RH, Diamandis EP. Bone morphogenetic protein antagonist gremlin-1
regulates colon cancer progression. Biol Chem. 2015;396(2):163–83. 10. Davis H, Irshad S, Bansal M, Rafferty H, Boitsova T, Bardella C, Jaeger E, Lewis
A, Freeman-Mills L, Giner FC, et al. Aberrant epithelial GREM1 expression
initiates colonic tumorigenesis from cells outside the stem cell niche. Nat
Med. 2015;21(1):62–70. 32. Hsu MY, Rovinsky SA, Lai CY, Qasem S, Liu X, How J, Engelhardt JF,
Murphy GF. Aggressive melanoma cells escape from BMP7-mediated
autocrine growth inhibition through coordinated Noggin upregulation. Laboratory investigation; a journal of technical methods and pathology. 2008;88(8):842–55. 11. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144(5):646–74. 33. Wilkinson L, Kolle G, Wen D, Piper M, Scott J, Little M. CRIM1 regulates the
rate of processing and delivery of bone morphogenetic proteins to the cell
surface. J Biol Chem. 2003;278(36):34181–8. 12. Batlle E, Clevers H. Cancer stem cells revisited. Nat Med. 2017;23(10):1124–34. 13. Dexter DL, Kowalski HM, Blazar BA, Fligiel Z, Vogel R, Heppner GH. Heterogeneity of tumor cells from a single mouse mammary tumor. Cancer
Res. 1978;38(10):3174–81. 34. References 1. Wakefield LM, Hill CS. Beyond TGFβ: roles of other TGFβ superfamily 25. Kuleshov MV, Jones MR, Rouillard AD, Fernandez NF, Duan Q, Wang Z,
Koplev S, Jenkins SL, Jagodnik KM, Lachmann A, et al. Enrichr: a
comprehensive gene set enrichment analysis web server 2016 update. Nucleic Acids Res. 2016;44(W1):W90–7. y
β
β
p
y
members in cancer. Nat Rev Cancer. 2013;13(5):328–41. members in cancer. Nat Rev Cancer. 2013;13(5):328–41. members in cancer. Nat Rev Cancer. 2013;13(5):328–41. 2. Colak S, Ten Dijke P. Targeting TGF-β signaling in Cancer. Trends in cancer. 2017;3(1):56–71. 2. Colak S, Ten Dijke P. Targeting TGF-β signaling in Cancer. Trends in cancer. 2017;3(1):56–71. 3. Tang B, Vu M, Booker T, Santner SJ, Miller FR, Anver MR, Wakefield LM. TGF-
β switches from tumor suppressor to prometastatic factor in a model of
breast cancer progression. J Clin Invest. 2003;112(7):1116–24. 26. Fabregat A, Jupe S, Matthews L, Sidiropoulos K, Gillespie M, Garapati P, Haw
R, Jassal B, Korninger F, May B, et al. The Reactome pathway
knowledgebase. Nucleic Acids Res. 2018;46(D1):D649–d655. 4. Mu Y, Gudey SK, Landstrom M. Non-Smad signaling pathways. Cell Tissue
Res. 2012;347(1):11–20. 27. Zhu Q, Wong AK, Krishnan A, Aure MR, Tadych A, Zhang R, Corney DC,
Greene CS, Bongo LA, Kristensen VN et al: Targeted exploration and analysis
of large cross-platform human transcriptomic compendia. Nature methods
2015, 12(3):211–214, 213 p following 214. 5. Zhang L, Ye Y, Long X, Xiao P, Ren X, Yu J. BMP signaling and its
paradoxical effects in tumorigenesis and dissemination. Oncotarget. 2016;
7(47):78206–18. 5. Zhang L, Ye Y, Long X, Xiao P, Ren X, Yu J. BMP signaling and its
paradoxical effects in tumorigenesis and dissemination. Oncotarget. 2016;
7(47):78206–18. 28. Petryszak R, Keays M, Tang YA, Fonseca NA, Barrera E, Burdett T, Fullgrabe A,
Fuentes AM, Jupp S, Koskinen S, et al. Expression atlas update--an
integrated database of gene and protein expression in humans, animals
and plants. Nucleic Acids Res. 2016;44(D1):D746–52. 6. Avsian-Kretchmer O, Hsueh AJ. Comparative genomic analysis of the eight-
membered ring cystine knot-containing bone morphogenetic protein
antagonists. Molecular endocrinology (Baltimore, Md). 2004;18(1):1–12. 6. Avsian-Kretchmer O, Hsueh AJ. Comparative genomic analysis of the eight-
membered ring cystine knot-containing bone morphogenetic protein
antagonists. Molecular endocrinology (Baltimore, Md). 2004;18(1):1–12. 7. Walsh DW, Godson C, Brazil DP, Martin F. Extracellular BMP-antagonist
regulation in development and disease: tied up in knots. Trends Cell Biol. 2010;20(5):244–56. 7. Received: 16 July 2019 Accepted: 22 October 2019 open platform for exploring multidimensional cancer genomics data. Cancer discovery. 2012;2(5):401–4. open platform for exploring multidimensional cancer genomics data. Cancer discovery. 2012;2(5):401–4. open platform for exploring multidimensional cancer genomics data. Cancer discovery. 2012;2(5):401–4. 24. Chen EY, Tan CM, Kou Y, Duan Q, Wang Z, Meirelles GV, Clark NR, Ma'ayan
A. Enrichr: interactive and collaborative HTML5 gene list enrichment analysis
tool. BMC bioinformatics. 2013;14:128. References BMP-6 over-expression
in prostate cancer is associated with increased id-1 protein and a more
invasive phenotype. J Pathol. 2008;214(3):394–404. 53. Alarmo EL, Korhonen T, Kuukasjärvi T, Huhtala H, Holli K, Kallioniemi A. Bone
morphogenetic protein 7 expression associates with bone metastasis in
breast carcinomas. Ann Oncol. 2008;19(2):308–14. 53. Alarmo EL, Korhonen T, Kuukasjärvi T, Huhtala H, Holli K, Kallioniemi A. Bone
morphogenetic protein 7 expression associates with bone metastasis in
breast carcinomas. Ann Oncol. 2008;19(2):308–14. 54. Costello CM, Cahill E, Martin F, Gaine S, McLoughlin P. Role of gremlin
in the lung: development and disease. Am J Respir Cell Mol Biol. 2010;
42(5):517–23. 55. Takahashi M, Otsuka F, Miyoshi T, Otani H, Goto J, Yamashita M, Ogura
T, Makino H, Doihara H. Bone morphogenetic protein 6 (BMP6) and
BMP7 inhibit estrogen-induced proliferation of breast cancer cells by
suppressing p38 mitogen-activated protein kinase activation. J
Endocrinol. 2008;199(3):445–55. 55. Takahashi M, Otsuka F, Miyoshi T, Otani H, Goto J, Yamashita M, Ogura
T, Makino H, Doihara H. Bone morphogenetic protein 6 (BMP6) and
BMP7 inhibit estrogen-induced proliferation of breast cancer cells by
suppressing p38 mitogen-activated protein kinase activation. J
Endocrinol. 2008;199(3):445–55. References Fibroblast-derived
Gremlin1 localises to epithelial cells at the base of the intestinal crypt. Oncotarget. 2019;10(45):4630–9. 47. Ren J, Smid M, Iaria J, Salvatori DCF, van Dam H, Zhu HJ, Martens JWM,
Ten Dijke P. Cancer-associated fibroblast-derived gremlin 1 promotes
breast cancer progression. Breast cancer research :
BCR. 2019;21(1):109. 47. Ren J, Smid M, Iaria J, Salvatori DCF, van Dam H, Zhu HJ, Martens JWM,
Ten Dijke P. Cancer-associated fibroblast-derived gremlin 1 promotes
breast cancer progression. Breast cancer research :
BCR. 2019;21(1):109. 48. Schultz S, Bartsch H, Sotlar K, Petat-Dutter K, Bonin M, Kahlert S, Harbeck N,
Vogel U, Seeger H, Fehm T, et al. Progression-specific genes identified in
microdissected formalin-fixed and paraffin-embedded tissue containing
matched ductal carcinoma in situ and invasive ductal breast cancers. BMC
Med Genet. 2018;11(1):80. 48. Schultz S, Bartsch H, Sotlar K, Petat-Dutter K, Bonin M, Kahlert S, Harbeck N,
Vogel U, Seeger H, Fehm T, et al. Progression-specific genes identified in
microdissected formalin-fixed and paraffin-embedded tissue containing
matched ductal carcinoma in situ and invasive ductal breast cancers. BMC
Med Genet. 2018;11(1):80. 49. Deng G, Zeng S, Qu Y, Luo Q, Guo C, Yin L, Han Y, Li Y, Cai C, Fu Y, et al. BMP4 promotes hepatocellular carcinoma proliferation by autophagy
activation through JNK1-mediated Bcl-2 phosphorylation. Journal of
experimental & clinical cancer research : CR. 2018;37(1):156. 49. Deng G, Zeng S, Qu Y, Luo Q, Guo C, Yin L, Han Y, Li Y, Cai C, Fu Y, et al. BMP4 promotes hepatocellular carcinoma proliferation by autophagy
activation through JNK1-mediated Bcl-2 phosphorylation. Journal of
experimental & clinical cancer research : CR. 2018;37(1):156. 50. Martinez VG, Rubio C, Martinez-Fernandez M, Segovia C, Lopez-Calderon F,
Garin MI, Teijeira A, Munera-Maravilla E, Varas A, Sacedon R, et al. BMP4
induces M2 macrophage polarization and favors tumor progression in
bladder Cancer. Clinical cancer research : an official journal of the American
Association for Cancer Research. 2017;23(23):7388–99. 51. Stieglitz D, Lamm S, Braig S, Feuerer L, Kuphal S, Dietrich P, Arndt S,
Echtenacher B, Hellerbrand C, Karrer S, et al. BMP6-induced modulation of
the tumor micro-milieu. Oncogene. 2018. 51. Stieglitz D, Lamm S, Braig S, Feuerer L, Kuphal S, Dietrich P, Arndt S,
Echtenacher B, Hellerbrand C, Karrer S, et al. BMP6-induced modulation of
the tumor micro-milieu. Oncogene. 2018. 52. Darby S, Cross SS, Brown NJ, Hamdy FC, Robson CN. References An
online survival analysis tool to rapidly assess the effect of 22,277 genes on
breast cancer prognosis using microarray data of 1,809 patients. Breast
Cancer Res Treat. 2010;123(3):725–31. 40. Ketolainen JM, Alarmo EL, Tuominen VJ, Kallioniemi A. Parallel inhibition of
cell growth and induction of cell migration and invasion in breast cancer
cells by bone morphogenetic protein 4. Breast Cancer Res Treat. 2010;
124(2):377–86. 21. Nagy A, Lanczky A, Menyhart O, Gyorffy B. Validation of miRNA prognostic
power in hepatocellular carcinoma using expression data of independent
datasets. Sci Rep. 2018;8(1):9227. 41. Alarmo EL, Parssinen J, Ketolainen JM, Savinainen K, Karhu R, Kallioniemi A. BMP7 influences proliferation, migration, and invasion of breast cancer cells. Cancer Lett. 2009;275(1):35–43. 22. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, Sun Y,
Jacobsen A, Sinha R, Larsson E et al: Integrative analysis of complex
cancer genomics and clinical profiles using the cBioPortal. Sci Signal
2013, 6(269):pl1. 42. Owens P, Polikowsky H, Pickup MW, Gorska AE, Jovanovic B, Shaw AK,
Novitskiy SV, Hong CC, Moses HL. Bone morphogenetic proteins stimulate
mammary fibroblasts to promote mammary carcinoma cell invasion. PLoS
One. 2013;8(6):e67533. 23. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, Jacobsen A,
Byrne CJ, Heuer ML, Larsson E, et al. The cBio cancer genomics portal: an Page 17 of 17 Neckmann et al. Cell Communication and Signaling (2019) 17:140 43. de Boeck M, Cui C, Mulder AA, Jost CR, Ikeno S, Ten Dijke P. Smad6
determines BMP-regulated invasive behaviour of breast cancer cells in a
zebrafish xenograft model. Sci Rep. 2016;6:24968. 44. Stolfi C, Marafini I, De Simone V, Pallone F, Monteleone G. The dual role of
Smad7 in the control of cancer growth and metastasis. Int J Mol Sci. 2013;
14(12):23774–90. 45. Tarragona M, Pavlovic M, Arnal-Estape A, Urosevic J, Morales M, Guiu M,
Planet E, Gonzalez-Suarez E, Gomis RR. Identification of NOG as a
specific breast cancer bone metastasis-supporting gene. J Biol Chem. 2012;287(25):21346–55. 46. Dutton LR, Hoare OP, McCorry AMB, Redmond KL, Adam NE, Canamara S,
Bingham V, Mullan PB, Lawler M, Dunne PD, et al. Fibroblast-derived
Gremlin1 localises to epithelial cells at the base of the intestinal crypt. Oncotarget. 2019;10(45):4630–9. 46. Dutton LR, Hoare OP, McCorry AMB, Redmond KL, Adam NE, Canamara S,
Bingham V, Mullan PB, Lawler M, Dunne PD, et al. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
|
https://openalex.org/W4225101418
|
https://www.frontiersin.org/articles/10.3389/fimmu.2022.873701/pdf
|
English
| null |
Evolving Strategies to Eliminate the CD4 T Cells HIV Viral Reservoir via CAR T Cell Immunotherapy
|
Frontiers in immunology
| 2,022
|
cc-by
| 16,343
|
REVIEW REVIEW REVIEW
published: 29 April 2022
doi: 10.3389/fimmu.2022.873701 Jarrod York 1,2, Kavitha Gowrishankar 2,3,4, Kenneth Micklethwaite 2,4,5,6, Sarah Palmer 1,4,
Anthony L. Cunningham 1,4*† and Najla Nasr 1,4*† Jarrod York 1,2, Kavitha Gowrishankar 2,3,4, Kenneth Micklethwaite 2,4,5,6, Sarah Palmer 1,4,
Anthony L. Cunningham 1,4*† and Najla Nasr 1,4*† 1 Centre for Virus Research, The Westmead Institute for Medical Research, Westmead, NSW, Australia, 2 Centre for Cancer
Research, The Westmead Institute for Medical Research, Westmead, NSW, Australia, 3 Children’s Cancer Research Unit,
Kids Research, The Children’s Hospital at Westmead, Sydney Children’s Hospitals Network, Westmead, NSW, Australia,
4 Faculty of Medicine and Health, Sydney Institute for Infectious Diseases, School of Medical Sciences, The University of
Sydney, Sydney, NSW, Australia, 5 Blood Transplant and Cell Therapies Program, Department of Haematology, Westmead
Hospital, Sydney, NSW, Australia, 6 NSW Health Pathology Blood Transplant and Cell Therapies Laboratory – Institute of
Clinical Pathology and Medical Research (ICPMR) Westmead, Sydney, NSW, Australia Edited by: Edited by:
William Anderson Paxton,
University of Liverpool,
United Kingdom Reviewed by:
Jean-Pierre Routy,
McGill University, Canada
Masakazu Kamata,
University of Alabama at Birmingham,
United States *Correspondence:
Anthony L. Cunningham
tony.cunningham@sydney.edu.au
Najla Nasr
najla nasr@sydney edu au *Correspondence:
Anthony L. Cunningham
tony.cunningham@sydney.edu.au
Najla Nasr
najla.nasr@sydney.edu.au
†These authors share last authorship †These authors share last authorship Specialty section:
This article was submitted to
Viral Immunology,
a section of the journal
Frontiers in Immunology Keywords: HIV-1, CD4, CD8, chimeric antigen receptor, latency, reactivation INTRODUCTION Received: 11 February 2022
Accepted: 04 April 2022
Published: 29 April 2022 Over forty years of HIV research has resulted in the management of this infection as a chronic
disease, primarily due to the advent of effective antiretroviral therapy (ART) in addition to
improved community, sexual, drug and occupational health practices (1). However, despite these
successes, there is currently no effective vaccine for the virus, and HIV cannot be fully eradicated
from patients due to the establishment of the HIV latent reservoir, defined here as cells containing
replication-competent HIV. In 2018, 1.7 million people were newly infected with HIV worldwide,
bringing the global disease burden to over 37 million. Currently anti-retroviral therapy (ART)
inhibits HIV replication at several stages of the virus lifecycle (2). If ART is received early during
acute infection, long-term control without viral rebound following withdrawal of treatment has Evolving Strategies to Eliminate
the CD4 T Cells HIV Viral Reservoir
via CAR T Cell Immunotherapy Jarrod York 1,2, Kavitha Gowrishankar 2,3,4, Kenneth Micklethwaite 2,4,5,6, Sarah Palmer 1,4,
Anthony L. Cunningham 1,4*† and Najla Nasr 1,4*† Although the advent of ART has significantly reduced the morbidity and mortality
associated with HIV infection, the stable pool of HIV in latently infected cells requires
lifelong treatment adherence, with the cessation of ART resulting in rapid reactivation of
the virus and productive HIV infection. Therefore, these few cells containing replication-
competent HIV, known as the latent HIV reservoir, act as the main barrier to immune
clearance and HIV cure. While several strategies involving HIV silencing or its reactivation
in latently infected cells for elimination by immune responses have been explored, exciting
cell based immune therapies involving genetically engineered T cells expressing synthetic
chimeric receptors (CAR T cells) are highly appealing and promising. CAR T cells, in
contrast to endogenous cytotoxic T cells, can function independently of MHC to target
HIV-infected cells, are efficacious and have demonstrated acceptable safety profiles and
long-term persistence in peripheral blood. In this review, we present a comprehensive
picture of the current efforts to target the HIV latent reservoir, with a focus on CAR T cell
therapies. We highlight the current challenges and advances in this field, while discussing
the importance of novel CAR designs in the efforts to find a HIV cure. Citation: York J, Gowrishankar K,
Micklethwaite K, Palmer S,
Cunningham AL and Nasr N (2022)
Evolving Strategies to Eliminate the
CD4 T Cells HIV Viral Reservoir via
CAR T Cell Immunotherapy. Front. Immunol. 13:873701. doi: 10.3389/fimmu.2022.873701 York J, Gowrishankar K,
Micklethwaite K, Palmer S,
Cunningham AL and Nasr N (2022)
Evolving Strategies to Eliminate the
CD4 T Cells HIV Viral Reservoir via
CAR T Cell Immunotherapy. Front. Immunol. 13:873701. doi: 10.3389/fimmu.2022.873701 April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org Reducing the HIV Reservoir York et al. been observed in patients, referred to as post-treatment
controllers (PTCs) (3). The immune system in PTCs is able to
control viremia and limit the HIV latent reservoir size even after
the cessation of ART (3). Levels of CD4 T cell activation have
been found to be lower in PTCs than in non-controllers. Furthermore, CD8 T and natural killer (NK) cell responses are
more efficient, and levels of inflammation are low (4). Although
ART has significantly reduced the morbidity and mortality
associated with HIV infection (2), the enduring pool of
replication-competent proviruses in latently infected cells
requires lifelong treatment adherence which is associated with
long-term toxicities, issues of compliance, expense, and
inconvenience. Furthermore, it has been shown that HIV-
infected individuals on ART have an increased risk of
malignancies, cardiovascular and neurologic disease (5). The
latter is attributed to HIV infection of macrophages, astrocytes
and microglia in the central nervous system (6). However, the
largest HIV reservoir reside in resting CD4 T cells. We and
others have previously reviewed the CD4 T cell subsets in which
HIV persists (7, 8). In addition, our recent study has revealed the
HIV proviral landscape is different across naïve and memory
CD4 T cell subsets and that the viral protein Nef plays a role in
the persistence of genetically intact HIV within the effector
memory CD4 T cells (9). Reactivation of replication-competent
virus (intact) in CD4 T cells upon cessation of treatment remains
a major drawback of ART (10). Therefore, the latent HIV
reservoir acts as the main barrier to immune clearance and
HIV cure. production in DCs and macrophages (15) via the HIV-
accessory protein Vpr which blocks the phosphorylation of
TBK1, preventing the phosphorylation of IRF3 and its
subsequent translocation to the nucleus to induce IFNb (16). In CD4 T cells, Vpu, Vpr or Vif directly degrade IRF3 (17). Natural Killer Cells and the Innate
Immune Response To control initial HIV infection, NK cells are activated and
proliferate prior to peak viremia (22). NK cell immunoglobulin-
like receptors can interact with cells expressing human leukocyte
antigen (HLA) molecules to limit HIV viral replication during
acute infection. However, similar to the inhibition of type I IFN
responses in early HIV infection, HIV has evolved mechanisms
to decrease expression of ligands important in triggering NK cell
cytotoxic responses and therefore interferes with lysis of infected
cells (23). Citation: Despite HIV mediated blocking of IFNb production in its key
target cells, IFN has been detected in the circulation of HIV
patients within 1-2 weeks of infection (18) with pDCs being the
main source of IFN production. Therefore, pDCs can
compensate for the loss of IFN inhibition in HIV key target
cells. pDCs and IFN and have been associated with antiviral
responses that limit early SIV/HIV replication and dissemination
in macaque and humanized mouse models respectively (19, 20). Moreover, Type I IFNs enhance immune cell activation and
effector functions, particularly the proliferation and survival of
NK cells, DC maturation, priming of T cell responses, promotion
of Th cell survival, activation and expansion of CD8 T cell, B cell
class switching and affinity maturation (21). Adaptive Immune Responses Adaptive immune responses are initiated as innate immunity
involving the NK cytotoxic response and IFN induction is
inhibited. However, there is a dampened CD4 T cell response
in acute HIV infection due to their significant early depletion
upon exposure to HIV (24). Early ART administration to control
viremia prevents the killing of CD4 T cells and rescues potent
CD4 T cell responses (25). The initial adaptive immune response
depends mainly on cytotoxic CD8 T cells to limit or inhibit viral
spread. However, a rapid decline of CD8 T cell responses occurs
limiting its effectiveness. The recapitulation of this phenomenon
in a CD4 depleted mouse model suggests that CD4 T cells are
required for the maintenance of long-lived memory CD8 T cells
in the context of acute HIV infection (26). Innate immunity is mediated upon pathogen encounter by cells
to directly destroy invading pathogens or control them via
production of inflammatory cytokines such as interferon (IFN)
and interleukin (IL)-15. Adaptive immunity can be humoral
acting in extracellular spaces via antibody production by B cells,
or T cell-mediated effectors which act once pathogens invade
cells leading to the induction of memory CD4 and CD8 T cells
for faster responses during pathogen re-exposure. Interferon and the Innate Immune
Response p
Type I IFNs include IFNa, IFNb, IFNϵ, IFNw, and IFNk (11). Thirteen subtypes of human IFNa and IFNb signal through the
IFNa receptor (IFNAR) (12). Type II and III IFNs include IFNg
and IFNl1-4, which signal through the IFNg and IFNl
receptors, respectively (11). Type I and type II IFNs are key
mediators of antiviral immunity, whilst type III IFN activity is
limited to epithelial cell surfaces due to restricted expression of
the IFN-l receptors (13). We have previously reviewed the
induction of IFN in myeloid DCs, macrophages, CD4 T cells
and pDCs (14) and outlined that many viruses interfere with IFN
induction to evade innate immune recognition. HIV is no
exception as we have shown that HIV inhibits IFNb Although the early CD8 T cell response peaks later than viral
load, there is much evidence for the importance of CD8 T cells in
controlling HIV replication via: the association of HIV-specific
CD8 T cells with a decrease in HIV-infected cells in acutely
infected patients (27); restriction of viral replication and disease
progression in chronically infected patients, especially elite
controllers (28); prevention of viral escape mutations (29);
maintaining viral suppression in SIV-infected macaques (30). Various studies in macaques have shown that stimulation of SIV-
specific CD8 T cell responses by vaccines can attenuate SIV
infection (31) and in vivo exhaustion of CD8 T cells in this model April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 2 Reducing the HIV Reservoir York et al. region of HIV (45) were effective in preventing viral RNA
transcription. Such strategies involving recombinant proteins or
RNA interference require specific targeting and expression cassettes
for delivery to latently infected cells. Heat shock protein 90 (HSP90)
inhibitors have also been explored as a block and lock strategy. Since
heat shock proteins are required for the production of viral proteins
(46), HSP90 inhibitors have been shown to suppress HIV
transcription and replication (47). Perhaps the most advanced
block and lock strategy employs didehydro-cortistatin A (dCA), a
Tat inhibitor, to silence HIV transcription. It is known that viral Tat
recruits and activates RNAPII for stimulation of HIV
transcriptional elongation (48). Using dCA, Kessing et al. showed
that prior treatment with dCA delayed and reduced viral rebound
both in vitro and in vivo (49). Kick and Kill Latently infected resting memory T cells do not express HIV
peptides on their cell surface and consequently evade immune
detection and cell lysis. Therefore, HIV activation and infected cell
elimination strategies, termed “kick and kill”, are based on
reactivation of HIV provirus from latently infected cells for
expression of viral peptides followed by elimination via viral
cytopathic effects or CD8 T cell mediated immune responses (51). The first generation of LRAs, including histone deacetylase
inhibitors (HDACi) and protein kinase C (PKC) activators were
disappointing due mainly to a lack of potency and/or unacceptable
toxicity (52). However, other studies have generated positive results
both in vitro and using animal models for HIV latency,
demonstrating induction of robust HIV expression using
bryostatin, a PKC activator (53), and second mitochondria-
derived activator of caspase (SMAC) mimetics to activate NF-kB
signalling and reverse HIV-1 latency (54). Clinical trials have
highlighted that a high concentration of LRAs is required for
potent reactivation of latently infected cells, but this led to off-
target effects and cytotoxicity (55). Lowering the toxicity of LRAs
combined with multiple dosing were effective in improving the
‘kick’ strategies (55). One negative feature of LRAs is their
suppressive effects on the cytolytic function of CTLs (56). Therefore, alternative strategies have been proposed, including use
of either the SMAC mimetics that activate the non-canonical
nuclear factor kB (NFkB) pathway (54) or the toll-like receptor 7
(TLR-7) agonists that mediate activation of HIV-specific CTLs (57). These strategies have shown increased reactivation and cell killing of
latently infected cells, indicating that a robust immune response
competent enough to remove the reactivated viral reservoir is
needed (58). STRATEGIES TO ELIMINATE THE HIV
LATENT RESERVOIR Current ART therapies targeting HIV suppress viral replication
to undetectable levels allowing infected individuals to live
without clinical symptoms, provided ART is maintained (43). Therefore, development of a cure to HIV patients relies on
approaches aimed at eliminating cells harbouring latent and
replication competent virions or removing the HIV provirus
from the genome by gene editing. Some of these approaches have
been addressed recently by Deeks et al. and are summarised
below (8). Interferon and the Innate Immune
Response In addition to silencing strategies,
genome editing with CRISPR/Cas9, ZFNs and TALENs have been
applied to highly conserved sequences within the LTR to eliminate
the HIV provirus genome in vitro (50). These approaches would
allow for highly specific targeting of latently infected cells, however
off-target endonuclease activity may occur, and development of
effective gene delivery vehicles remain important challenges. limited viral control in acute infection (32). CD8 T cell responses
during initial HIV infection target mutable regions of the virus to
decrease replication efficiency and contribute to the control of
HIV infection (33). This early CD8 T cell response is observed
days prior to the peak of viremia and targets epitopes found in
HIV Env and Nef (34). However, it has been demonstrated that
escape mutants, which cannot be recognized by initially
generated cytotoxic CD8 T cells, are generated within 10 to 21
days post infection in response to potent T cell responses (35). Therefore, although the initial CD8 T cell response restricts
viremia in acute HIV infection, their target epitopes are
subsequently mutated resulting in immune escape. In addition, HIV-specific CD8 T cell cytotoxic responses
decrease with disease progression, due to a reduction in their
frequency, increased activation, and immune exhaustion (36). CD8 T cells with higher expression of activation and immune
exhaustion markers such as programmed death (PD)-1 and T-
cell immunoglobulin and mucin-domain containing-3 (TIM-3)
have been detected in untreated HIV-infected patients in
contrast to seronegative patients (37). Expression of PD-1 is
associated with a lower degranulation capacity and a diminished
population of IL-17-producing cells following polyclonal
stimulation (38). Exhausted CD8 T cells are also more
susceptible to apoptosis (39). The loss of CD4 T cell assistance,
persistent viral antigen load, chronic inflammation (40) and lack
of co-stimulation (41) have been proposed to contribute to CD8
T cell dysfunction during chronic HIV infection. Furthermore,
increased expression of inhibitory molecules, competitive
signalling between TCRs, upregulated inhibitory receptors on
the responding CD8 T cells and the modulation of intracellular
signalling pathways can all contribute to the lack of co-
stimulation and dampening of the CD8 T cell response to
infection (42). April 2022 | Volume 13 | Article 873701 Block and Lock Several compounds have been developed to inhibit both the basal
and signal induced activity of the nuclear transcription factor NFkB,
a key mediator of HIV gene expression (44). However, the HIV LTR
is sensitive to a variety of activating factors and inhibiting a single
mediator such as NFkB is unlikely to silence all HIV viral RNA
expression. To this end, constitutive suppression of the HIV LTR
transcription using short hairpin RNAs targeting the LTR enhancer Recently, a novel finding of HIV reactivation via IFN was
reported (59) and showed that IFNa but no other IFN subtypes
was able to efficiently reverse latency in both an in vitro model
and in CD4 T cells collected from HIV patients on suppressive
ART (59). This is similar to our recent findings of latency April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 3 Reducing the HIV Reservoir York et al. envelope transmembrane gp41, potent broadly neutralising
antibodies (bnAbs) specific for conserved regions of HIV gp120
are generated in 1-2% of patients many years after initial infection
(69). The mechanism by which bnAbs are generated in chronic
patients is not well understood. However, recent structural,
virologic and immunological studies have provided strong
evidence of how virus–antibody co-evolution throughout the
natural course of the disease are synergistically associated with
sequential development of potent bnAbs via somatic
hypermutations and maturation of antibody genes (70). Early
studies employing passive immunotherapies of such anti-HIV
antibodies have demonstrated improved clinical outcomes
associated with reduction in plasma viral RNA load and delayed
appearance of AIDS-defining illnesses (71). A limited number of
studies has demonstrated that bnAbs may act on latently infected
cells after their reactivation. For example, 3BNC117 induced
cytolysis of latently infected cells and PGT121 significantly
reduced HIV DNA after passive administration to non-human
primates (72). bnAbs can bind and neutralise viral envelopes
expressed on infected cells, following viral reactivation. They can
enhance lysis of infected cells by antigen-dependent cellular
cytotoxicity (ADCC) with the recruitment of NK cells and can
also prevent seeding of additional latently infected cells (73). However, although bnAbs exhibit remarkable breadth and
potency in their ability to neutralize HIV and prevent spread,
they rely on functional NK cells to directly recognise and lyse
infected cells during chronic HIV infection (74). Block and Lock In addition,
latently infected cells express only integrated HIV DNA and
thus remain invisible to the NK and CD8 T cells without HIV
reactivation by latency reversal agents (LRAs). Currently, bnAbs
are under investigation for treatment and prevention of HIV
infection (75) and have demonstrated pre-clinical success in use
as novel immunotherapies targeting the virus (76). reversal in in vitro infected resting memory CD4 T cells (60)
treated with IFNa8 or co-cultured with pDCs that secreted IFNa
upon HIV exposure. Data from in vitro and humanized mice
studies showed that IFNa8 and IFNa14 are more efficient at
inhibiting HIV viral replication than IFNa2 (61) due to their
higher affinities for the IFN receptor and the increased induction
of antiviral proteins. Therefore, HIV reactivation by a specific
type of IFNa should provide another promising strategy to
reactivate and purge latently infected cells with the added
benefit of inhibiting viral spread to adjacent cells (60) while
avoiding the suppressive effects of LRAs on the cytolytic function
of CD8 T cells. Altogether, this suggests that a combination of
the above strategies may be required for an optimal “kick and
kill” strategy. Activation and elimination components are influenced by
several factors. As HIV integrates within transcribed regions of
the chromosome, it has been demonstrated that responsiveness
of HIV provirus to T cell signalling and chromatin modifying
agonists dependent on the site of integration (62). Therefore, a
major challenge in developing robust LRAs is the broad range of
responses of the integrated proviruses at distinct chromosomal
locations (62). For example, although most provirus is
responsive to T cell signalling induced by a combination of
phorbol 12-myristate 13-acetate (PMA) and ionomycin, it has
been shown that only small proportion of provirus is induced by
treating patients with chromatin modifying agents (63). Due to
regulation of viral transcription by cell signalling pathways
linked to TCR engagement and T cell activation responses,
forced induction of viral expression without promotion of
mass T cell activation, leading to cytokine release syndrome,
remains a major challenge (64). j
g
As described earlier increased PD-1 and TIM-3 expression on
CD8 T cells in HIV patients results in lower degranulation capacity,
indicating that the immune system after ART is incapable of a
sufficient anti-HIV cytotoxic T lymphocyte response to eliminate
reactivated cells (65). To combat this, various strategies to augment
anti-HIV cellular immune responses are being explored. Block and Lock IL-15
treatment has been shown ex vivo to enhance elimination of
resting CD4 T cells isolated from HIV-patients treated with LRAs
by inducing NK activity (66). B cell lymphoma 2 (Bcl-2) regulates
apoptosis, inhibits T-cell mediated cytotoxicity pathways and has
been identified in primary CD4 T cells that survived co-culture with
HIV-specific CTLs (67). Addition of Bcl-2 antagonists in co-
cultures of latently infected CD4 and CTLs reduced latently
infected target cells when compared with CTL and LRAs
alone (68). Adoptive cell therapies have revolutionised the way we
can empower the immune response and direct it against specific
targets in a precise and controlled manner. In recent years major
advances have been made in engineering human T cells by
introducing chimeric antigen receptors (CARs) to enable specific
lysis. This approach has been approved as treatment for
haematological malignancies and it is being extended to other
diseases, including autoimmunity, fungal and viral infections. CHIMERIC ANTIGEN RECEPTOR T CELLS Chimeric antigen receptors (CARs) are synthetic receptors
expressed on T cells that confer novel specificity by combining
chosen antigen binding moieties and T cell activating functional
domains. CAR expressing T cells (CAR T cells) recognise antigens
independent of major histocompatibility class (MHC) presentation
and can therefore be used in immunotherapy to recognise, target,
and destroy antigen expressing cells in a specific manner. This
MHC independent mechanism of action can be particularly
attractive in targeting HIV infected cells as the virus is well
known to down regulate MHC-I molecules (77). Chimeric
antigen receptors have been well described by Rafiq et al. and are
structured as discrete modules (Figure 1A) with an antigen
recognition domain typically derived from single chain fragments
of the variable regions (scFv) of the light and heavy chains of a
monoclonal antibody (mAb) (78). In addition, novel non-antibody-
based approaches have been employed which can include natural
receptors, ligands, nanobodies, cytokines and peptides (78). A hinge
or spacer region links the antigen recognition domain to the
transmembrane domain and a T cell isignalling domain (CD3z Broadly Neutralising Antibodies Broadly Neutralising Antibodies Although the initial production of antibodies in response to HIV is
non-neutralizing, in part due to targeting antigens on the HIV April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 4 York et al. Reducing the HIV Reservoir A
B
FIGURE 1 | (A) Structure of chimeric antigen receptor. Antigen receptor domains are linked to transmembrane domains and intracellular signalling domains consisting of co-
stimulatory domains and CD3z by hinge regions. scFvs are common as antigen recognition domains, although ligands can be used to take advantage of ligand-receptor
interactions. (B) Evolution of chimeric antigen receptors. Antigen recognition domains are linked to CD3z in first generation CARs and activity was improved by addition of co-
stimulatory domains in second and third generations. Fourth generation TRUCKs include inducible transcription of transgenes (such as IL-12) to second generation CARs. scFv, single chain variable fragment; CAR, chimeric antigen receptor TRUCK, T cell redirected for universal cytokine killing; bnAb, broadly neutralising antibody; CRD,
carbohydrate recognition domain. Created with BioRender. Adapted from Rafiq et al. (78). A
B FIGURE 1 | (A) Structure of chimeric antigen receptor. Antigen receptor domains are linked to transmembrane domains and intracellular signalling domains consisting of co-
stimulatory domains and CD3z by hinge regions. scFvs are common as antigen recognition domains, although ligands can be used to take advantage of ligand-receptor
interactions. (B) Evolution of chimeric antigen receptors. Antigen recognition domains are linked to CD3z in first generation CARs and activity was improved by addition of co-
stimulatory domains in second and third generations. Fourth generation TRUCKs include inducible transcription of transgenes (such as IL-12) to second generation CARs. scFv, single chain variable fragment; CAR, chimeric antigen receptor TRUCK, T cell redirected for universal cytokine killing; bnAb, broadly neutralising antibody; CRD,
carbohydrate recognition domain. Created with BioRender. Adapted from Rafiq et al. (78). domain from the TCR complex) (78). Incorporation of
costimulatory domains provide the required co-stimulation for a
robust response. Domains from proteins of the CD28 family (CD28
and inducible T cell co-stimulator (ICOS)), or the tumour necrosis
factor receptor (TNFR) family (41BB (CD137)) and TNFR
superfamily (member 4 (OX40) or CD27) have been used for this
co-stimulation (78). Although several costimulatory domains have
been trialled in vitro, CD28 and 4-1BB have been the most
successful and are used in clinical products (78). For a
comprehensive review of CAR engineering strategies refer to
Rafiq et al. Broadly Neutralising Antibodies (78). Various vectors and trategies have been utilized
in the production of CAR-T cells and are summarised in Table 1. storm (CRS) and neurotoxicity (86). CRS involves the release of
effector cytokines like IL-6, IFNg and TNFa and can be
accompanied by fever, endothelial activation and vascular
instability. Neurotoxicity can range from mild headache to more
severe delirium, cerebral oedema or intracranial haemmorhage. Inclusion of safety switches can potentially prevent organ damage
and fatality and have become an essential component of current
CAR designs (78). Engineering CAR – T Cells to Target the
HIV Reservoir CAR T cells for HIV have been trialled since the 90s. Some have
demonstrated long term persistence in peripheral blood (87–89)
and have been found to traffic to areas of the HIV reservoir
including the central nervous system and peripheral lymphoid
tissue (90). As such, CAR T cells have the potential to
revolutionise immunotherapy for HIV, when appropriate
design and safety strategies are incorporated. CAR T cells have shown phenomenal success against
hematologic malignancies with close to 90% response in B-cell
acute lymphoblastic leukaemia and lymphoma targeting CD19 (84). A recent reportshowed persistence of CD19 CAR T cells 10 years
after infusion and leading to complete remission (85), further
validating CAR T cells as paradigm shifting therapies. Four CAR
T cell products (targeting CD19) for leukaemia and lymphomas and
one for myeloma (targeting BCMA) have been approved by the
FDA as therapy. The key challenges of CAR therapy include
managing adverse events like life threatening cytokine release Frontiers in Immunology | www.frontiersin.org First Generation The first-generation anti-HIV CAR-T cells linked singular CD3z
intracellular signalling domains to antigen recognition domains April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 5 Reducing the HIV Reservoir York et al. TABLE 1 | Current ex vivo genetic engineering strategies for CAR expression. Genetic engineering
Transgene
delivery/
Promoter
Transgene insertion/
gene expression
Strengths
Limitations
Retrovirus: Lentivirus, gamma
retrovirus
(79)
Transduction/
Exogenous
Non-targeted
integration/Stable
high transduction efficiency
size restriction, expensive production, induction of an
immune response, insertional mutagenesis
Transposase enzyme: PiggyBac,
and sleeping Beauty (80)
Electroporation/
Exogenous
Non-targeted
integration/Stable
integrating larger transgenes,
inexpensive production
Less developed technology than retroviral vectors,
off-target cleavage and insertional mutagenesis
mRNA (81)
Electroporation/
NA
Transient
Limits off-target effects
expression is rapidly diluted during T cell expansion
Non-integrative lentivirus: NILV-S/
MAR
(82)
Transduction/
Exogenous
Episomal/Transient
Episomal maintenance prevents
insertional mutagenesis
Expensive production, constant redosing required
Endonuclease enzymes: ZFNs,
TALENs, CRISPR/Cas9
(83)
Electroporation/
Endogenous
Targeted integration/
Stable
Site-directed insertion
Not yet fully optimized
non-integrative lentivirus containing scaffold/matrix attachment region (NILV-S/MAR); zinc-finger nucleases (ZFNs); transcription activator-like effector nucleases (TALENS); clustered
regularly interspaced short palindromic repeat (CRISPR). TABLE 1 | Current ex vivo genetic engineering strategies for CAR expression. non-integrative lentivirus containing scaffold/matrix attachment region (NILV-S/MAR); zinc-finger nucleases (ZFNs); transcription activator-like effector nucleases (TALENS); clustered
regularly interspaced short palindromic repeat (CRISPR). non-integrative lentivirus containing scaffold/matrix attachment region (NILV-S/MAR); zinc-finger nucleases (ZFNs); transcription activator-like effector nucleases (TALENS); clustered
regularly interspaced short palindromic repeat (CRISPR). cells exhibited extremely high cytolytic activities by JAM assay, a
cell lysis assay which measures the DNA retained by living cells
rather than the cellular components lost by dying cells, against
the CEM cell line infected with the HIV-1 IIIB strain and
HEK293 expressing HIV env. Furthermore, using Jurkat cells
expressing CD4 or scFv derived from anti-HIV mAb 98.6 -CD3z
based CARs, Roberts et al. demonstrated that HEK293 env target
cells stimulated Jurkat CAR-T cells to secrete IL-2 at high levels
(92). Subsequently, Yang et al. confirmed that primary CD8 T
cells expressing CD4 ligand or scFv derived from anti-HIV mAb directed against HIV Env using either CD4 or scFvs derived from
anti-HIV Env bNAbs (Figure 1B). These CAR-T cells were very
specific in recognizing and lysing cells expressing HIV Env in
vitro (Table 2A). Outcome Outcome - CD8 exhibited extremely high cytolytic activities against gp120 - CD8 exhibited extremely high cytolytic activities against gp120
expressing HeLa cells by chromium-release assay - CD8 exhibited extremely high cytolytic activities against gp120
expressing HeLa cells by chromium-release assay
- CD8 exhibited extremely high cytolytic activities toward CEM and
HEK293 cells expressing HIV env
by JAM assay. - HEK293 expressing HIV env activated the CAR expressing Jurkat
cells to secrete IL-2 at high levels. - Exhibited lysis of HIV-1 IIIB infected T1 and H9-B14 cells by
chromium-release assay
- CAR expression detected in 1% - 3% of circulating T cells at 8
weeks and 0.1% at 1 year
- Survival of CAR was not enhanced by IL-2
- CAR DNA was observed in biopsies of bulk rectal tissue and
lamina propria in 2 of 3 patients at 1 year
- No significant reduction in HIV RNA or DNA
- Greater persistence for at least one year compared to 8 weeks
following multiple infusions of CD8 CAR-T cells co-administered
with CD4 CAR-T cells. - CAR T cells detected in lymphoid tissue biopsies at 1 year
- Multiple infusions were well tolerated without substantive
immunologic or virologic changes
- Reduction in mean HIV RNA levels observed in patients who
received repeated infusions of CD8 CAR T cells
- Decline from baseline in HIV after CAR T cell treatment, but no
meaningful difference in HIV burden was observed between
treatment and placebo groups
- Increase of CD4 T cell count following therapy expressing HeLa cells by chromium-release assay - Both Jurkat cells and primary CD8 T cells expressed CD4 binding site or scFv derived from anti-
HIV mAb 98.6 -CD3z (92). - CD8 exhibited extremely high cytolytic activities toward CEM and
HEK293 cells expressing HIV env
by JAM assay. - HEK293 expressing HIV env activated the CAR expressing Jurkat
cells to secrete IL-2 at high levels. - Primary CD8 T cells expressing CD4 ligand or scFv derived from anti-HIV mAb 98.6 -CD3z (93). - Exhibited lysis of HIV-1 IIIB infected T1 and H9-B14 cells by
chromium-release assay B. Clinical studies
- Phase II randomized trial: Single-dose administration of autologous CD4 and CD8 CAR-T cells
both expressing CD4-CD3z with or without IL-2 to HIV- infected patients (n=24) with a detectable
level of virus (1000 copies/mL) (88). First Generation Using a CD8 cell line (WH3) expressing CD4-
CD3z CARs, Romeo and Seed demonstrated that after binding to
gp120 expressing HeLa cells, CD4-CD3z CAR-T cells exhibited
extremely high cytolytic activities by chromium-release assay
(91). These data were replicated in human PBMC derived CD8 T
cells expressing CD4 or scFv derived from anti-HIV mAb 98.6
-CD3z based CARs (92). After binding to HIV Env, these CAR-T TABLE 2 | Preclinical and clinical testing of first-generation anti-HIV CAR-T cells. Design of CAR
A. Preclinical studies
- CD8 cell line (WH3) expressing CD4 binding site-zCAR (91). - CD8
express
- Both Jurkat cells and primary CD8 T cells expressed CD4 binding site or scFv derived from anti-
HIV mAb 98.6 -CD3z (92). - CD8
HEK29
by JAM
- HEK2
cells to
- Primary CD8 T cells expressing CD4 ligand or scFv derived from anti-HIV mAb 98.6 -CD3z (93). - Exhib
chromi
B. Clinical studies
- Phase II randomized trial: Single-dose administration of autologous CD4 and CD8 CAR-T cells
both expressing CD4-CD3z with or without IL-2 to HIV- infected patients (n=24) with a detectable
level of virus (1000 copies/mL) (88). - CAR
weeks
- Surviv
- CAR
lamina
- No sig
- Single and multiple infusions of autologous CD4-CD3z CD8 CAR-T cells with or without CD4-
CD3z CD4 CAR-T cells between identical twins (n=33) serodifferent for HIV infection (87). - Great
followin
with CD
- CAR
- Multip
immun
- Redu
receive
- Phase II randomized trial: Three infusions of autologous CD4-CD3z CD4 and CD8 CAR-T cells
compared to unmodified T cells in HIV-infected individuals (n=40) on ART with low plasma viral
loads (<50 copies/mL) (89). - Declin
meanin
treatme
- Increa
Frontiers in Immunology | www.frontiersin.org
6 Second Generation Furthermore, HIV-resistant CAR expressing
CD4 T cells containing costimulatory domains from the CD28
receptor family exhibited the greatest production of effector
cytokines in vitro. whilst 41BB co-stimulated CAR expressing
CD4 T cells showed superior expansion and reduced HIV
pathogenesis in vivo using a humanized mouse model of HIV
infection (100). g
y
First generation CD4 ligand-based CAR-T cells also
demonstrated stable engraftment and long-term safety in the
clinical setting (Table 2B). However, despite these observations
and the previously demonstrated anti-HIV activity in vitro, the
CD4-CD3z CAR-T cells were not capable of reducing viral
burden permanently in most clinical studies (87, 88). Specifically, Mitsuyasu et al. noted CAR expression in 1% - 3%
of circulating T cells at 8 weeks and 0.1% at 1 year following
administration of autologous CD4 and CD8 T cells expressing
CD4-CD3z CARs with or without IL-2 to HIV-infected patients
(n=24) with a detectable level of virus (1000 copies/mL). However, no significant reduction in HIV RNA or DNA was
observed (88). Moreover, administration of exogenous IL-2 with
CAR T cells was not shown to enhance survival. Another study
showed greater persistence of CD8 T cells expressing CD4-CD3z
CARs at 1 year after multiple infusions when co-administered
with CD4 T cells expressing CD4-CD3z CARs compared to CD8
CAR-T cells infused without CD4 CAR-T cells in identical twins
(n=33) sero-different for HIV infection (87). Furthermore, a
reduction in mean HIV RNA levels was observed in patients who
received repeated infusions of CD8 CAR T cells (87). Subsequently, it has been demonstrated that infusion of CD4-
CD3z expressing CD4 and CD8 T cells led to a decline from
baseline in HIV after CAR T cell treatment, but no meaningful
difference in HIV burden was observed between treatment and
placebo groups (89). Notably, both groups experienced a
treatment-related increase in CD4 T-cell counts (89). Second Generation Based on the success of second generation anti-CD19 CAR-T
cells for hematologic malignancies, various laboratories have
developed second generation anti-HIV CAR T cells and
analysed them in preclinical studies (Figure 1B). Landmark
studies have been summarised in Table 3. Second-generation
CD8 CAR T cells incorporating the CD4 ligand or scFv derived
from anti-HIV bnAbs as antigen recognition domains co-
stimulated by 41BB were directly compared with first
generation CARs and shown to be at least 50-fold more potent
at suppressing HIV replication in vitro (94). Furthermore,
second generation CARs demonstrated superior antigen
dependent expansion, CD4 T cell protection via reduced cell
loss in vivo when compared to HIV-specific CARs without co-
stimulatory molecules (101). In the same study, a direct
comparison of 41BB and CD28 co-stimulatory domains
showed that 41BB costimulatory domains promote antigen
independent T cell persistence in vivo (101) and that they are
superior to the CD28 domain in reducing viral rebound
following cessation of ART treatment. This is in accordance
with other 4-1BB co-stimulatory domain containing CARs
showing long term persistence (96). Primary CD4 and CD8 T
cells expressing scFv derived from anti-HIV bnAbs including
PGT128, PGT145, VRC07-523, or 10E8 in second generation
CARs with 41BB co-stimulatory domains were developed and
compared (99). All CARs incorporating scFv from each bnAb
were shown to be efficacious as anti-HIV designs using HIV
infected cell lines and Env transfected cells (99). However,
PGT145- and VRC07-523-CARs showed the most consistent
potency in cytotoxic activity (99). Perhaps most importantly, this
study demonstrated the success of homology-directed
recombination of CAR transgene into the CCR5 (HIV co-
receptor) locus to suppress entry of HIV into these cells (99). By showing the feasibility of CAR integration to disrupt CCR5 in
T cells, Hale et al. demonstrated an important consideration
when developing a CAR T cell immunotherapy for HIV. The
efficacy of CAR expressing CD4 T cells in the context of HIV has
also been demonstrated in other studies. Focussing on HIV-
resistant CD4 CAR-T cells modified by introduction of a single
Asp mutation (D97N) and expressing co-stimulatory domains
derived from 41BB, CD28, CD27, OX40 or ICOS, Maldini et al. showed direct elimination of env expressing cell lines in vitro. They also demonstrated improved proliferation and persistence
of CD4 ligand based CAR expressing CD8 T cells, especially
when co-injected with CD4 ligand based CAR expressing CD4 T
cells in vivo (101). Outcome - CAR expression detected in 1% - 3% of circulating T cells at 8
weeks and 0.1% at 1 year - Survival of CAR was not enhanced by IL-2
- CAR DNA was observed in biopsies of bulk rectal tissue and
lamina propria in 2 of 3 patients at 1 year - No significant reduction in HIV RNA or DNA - Single and multiple infusions of autologous CD4-CD3z CD8 CAR-T cells with or without CD4-
CD3z CD4 CAR-T cells between identical twins (n=33) serodifferent for HIV infection (87). - Greater persistence for at least one year compared to 8 weeks
following multiple infusions of CD8 CAR-T cells co-administered
with CD4 CAR-T cells. - Reduction in mean HIV RNA levels observed in patients who
received repeated infusions of CD8 CAR T cells - Phase II randomized trial: Three infusions of autologous CD4-CD3z CD4 and CD8 CAR-T cells
compared to unmodified T cells in HIV-infected individuals (n=40) on ART with low plasma viral
loads (<50 copies/mL) (89). - Decline from baseline in HIV after CAR T cell treatment, but no
meaningful difference in HIV burden was observed between
treatment and placebo groups - Increase of CD4 T cell count following therapy April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 6 York et al. Reducing the HIV Reservoir 98.6 -CD3z based CARs lysed HIV-1 IIIB infected T1 and H9-
B14 cells and inhibited HIV-1 replication in vitro (93). The
kinetics of lysis and efficiency of inhibition were comparable to
that of naturally occurring HIV-1-specific CTL clones isolated
from infected individuals (93). Such preclinical studies provided
clear evidence that first generation CD4 or scFv derived from
anti-HIV mAb 98.6 -CD3z based CAR T cells exhibited
significant anti-HIV activity in vitro. in vitro led to the development of second and third generation
CARs incorporating the signals require for full T-cell activation
within the CAR structure to enhance their function. Frontiers in Immunology | www.frontiersin.org Design - Primary CD8 T cells expressing CD4 or scFv derived from anti-HIV bnAbs (VRC01,
3BNC60, PG9, PGT128, PGDM1400) based second generation CARs with 41BB or
CD28 co-stimulatory domains (94). - Had 50-fold increase in potency at suppressing HIV replication in vitro than
first generation CAR T cells by chromium-release assay. - Superior in vivo expansion in response to antigen, - Superior in vivo expansion in response to antigen,
Reduced CD4 T cell loss compared to first generation CARs. - Superior in vivo expansion in response to antigen,
Reduced CD4 T cell loss compared to first generation CARs. Reduced CD4 T cell loss compared to first generation CARs. - Incorporation of 41BB costimulatory domains was superior to CD28
domains in reducing viral rebound after ART treatment and promoting T cell
persistence in vivo - Incorporation of 41BB costimulatory domains was superior to CD28
domains in reducing viral rebound after ART treatment and promoting T cell
persistence in vivo - Primary CD4 and CD8 T cells expressing scFv derived from anti-HIV bnAbs (PGT128,
PGT145, VRC07-523, 10E8) second generation CARs with 41BB as a co-stimulatory
domain (99). - PGT145- and VRC07-523-CARs had consistent potency in killing of HIV
infected cell lines and Env transfected cells - CCR5-edited CAR T cells more effectively suppressed viral replication - AAV6 mediated homology-directed recombination of the CAR gene into the CCR5
locus (99). - Primary CD4 and CD8 T cells expressing CD4 ligand based second generation CARs
and incorporating co-stimulatory domains derived from 41BB, CD28, CD27, OX40 and
ICOS (100). - CAR T cells with CD28 costimulatory domains exhibited the greatest
production of IL-2, TNF, IFNg, (MIP)-1b and GMCSF when co-cultured with
env expressing cell lines in vitro. - CAR modified CD4 T cells rendered HIV-resistant by introduction of a single Asp
mutation (D97N) of CXCR4 (100). - CAR T cells with 4-1BB-costimulatory domains directly eliminated env
expressing cell lines in vitro and exhibited profound expansion. - CD4 CAR T Cells had improved proliferation and persistence of CD8 CAR T
cells in vivo. target effects, and navigate immunoregulatory factors in the
tissue microenvironment for a sustained and effective response. The versatile and modular nature of CARs and advances in
molecular and synthetic biology thus allows for the development
of a myriad of designs for these purposes to increase the efficacy
and precision of anti-HIV cell therapies. Overcoming Immune Escape g
Antigen escape, including antigen loss or downregulation, is a
major limitation of CARs designed with single antigen
recognition domains, particularly for scFv-based CARs. Immune escape was demonstrated in recent clinical trials
where a decrease in plasma viremia followed by viral rebound
using monotherapy of bNAb VRC01 (103) was observed. In
contrast, treatment with combinations of two or more bNAbs
significantly reduced the viral reservoir and demonstrated long-
term viral suppression (104). Therefore, single bnAb scFv-based
CARs may be insufficient for long term HIV suppression due to
the emergence of escape mutants. Thus, combination of multiple
antigen recognition domain by CARs, using duo or Tandem
designs may provide long-term suppression and are currently
under investigation. Duo CARs expressing complimentary
antigen recognition domains directed against different antigens
via transfection as a single bicistronic vector (Figure 2A) have
been used successfully in cancer to overcome immune escape
(106). Tandem CAR-T cells consists of a CAR where distinct
antigen recognition domains are fused to generate smaller
transgenes compared to dual CARs whilst retaining capacity to
target multiple antigens can prevent immune escape (Figure 2B)
(107). More recently, Hajduczki et al. further improved a
bispecific CAR design based on the CRD of human mannose- Design Prompted by novel
innovations in cancer immunotherapy, various research groups
have begun designing and testing CARs to overcome shared
challenges between cancer and chronic viral infections as well as
HIV specific hurdles including CAR – T cell expansion and
persistence, and their susceptibility to HIV infection. These data demonstrate that co-stimulatory domains along with
helper functions supplied from CAR expressing CD4 T cells
differentially enhance HIV-specific CAR expressing CD8 T cell
function in vivo. Future studies may benefit from adopting strategies
to develop HIV-resistant CAR expressing CD4 T cells and
preferentially incorporating 41BB co-stimulatory domains. Third and Fourth Generation G
Third-generation HIV-specific CARs (Figure 1B) have not been
extensively studied in HIV. A single study combining multiple
intracellular signalling domains from 41BB, CD28 and CD3z linked
to the scFv of anti-HIV bnAb VRC01 displayed increased potency
in lysing env expressing cell lines in vitro compared to HIV
recognition via CD4-based third generation CAR (102). Unfortunately, this study failed to directly compare the potency of
their third generation CAR to an appropriately designed second
generation CAR counterpart. However, an important finding of the
study was that the VRC01-based third generation CAR-T cell
effectively eliminated LRA-reactivated HIV-1-infected CD4 T cells
isolated from infected individuals receiving ART, thus validating the
development of CAR-T cell immunotherapies against latent HIV
(102). The fourth generation of CAR-T cells are known as T cells
redirected for universal cytokine-mediated killing (TRUCKs) and
have yet to be employed in the context of HIV infection
(Figure 1B). TRUCKs can utilise the additional secretion of IL-12
or other inflammatory mediators to attract innate immune cells and
eliminate antigen negative cancer cells. CAR designs are constantly
evolving with novel features being added to provide multiple
functions, especially to overcome the challenges of the tissue
microenvironment and future anti-HIV CARs will no doubt
incorporate such features. Outcome Outcome Design Second Generation The failure of these clinical trials to generate robust cell lysis
of their targets may be attributed to several factors: (1) low
transduction efficiency and/or dose of CD4-CD3z CAR-T cells
leading to lower efficacy (94); (2) T cell exhaustion during ex vivo
CD4-CD3z CAR-T cell expansion by excessive IL-2 stimulation
(95); (3) CAR design leading to low expansion rate and effector
functions (96); (4) Initial ART administration prior to CD4-
CD3z CAR T cell infusion may block the efficacy of CAR-T cells
by reducing HIV Env antigen required for optimal expansion of
a CD4-CD3z CAR-T cells in vivo and the production of a robust
anti-HIV response (97); (5) CD4-CD3z CAR T cells may induce
selective pressure on HIV Env to select for escape mutants (98);
however, (6) the most common explanation for the failure of the
CD4-CD3z CAR T cell clinical trials is the inclusion of CD4 as
the antigen recognition domain, rendering the CD8 T cells
expressing CD4 susceptible to HIV infection (97). Overall, the
application of first-generation anti-HIV CAR T cell therapy in
HIV patients has demonstrated a lack of adverse effects, and the
generation of compartmental immunity to HIV (87–89). However, the lack of efficacy of these CAR T cells in vivo
despite previously demonstrated significant anti-HIV activity April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 7 Reducing the HIV Reservoir York et al. TABLE 3 | Recent preclinical testing of second-generation anti -HIV CAR T cells. Advances in Developing CAR-T Cell
Therapy for HIV Cure Upon binding of a secondary DARPin linked to a key protein to its antigen, the latch is released exposing Bim to bind to a Bcl-2 CAR. CAR, chimeric
antigen receptor; NKG2D, MICA receptor; MICA, NKG2D ligand; DARPin, designer ankyrin repeat proteins. Created with BioRender. Adapted from Guedan et al. (105). binding lectin (MBL) recognizing the highly conserved
oligomannose patch on gp120 and CD4 ligand by addition of a
third antigen recognition domain against a distinctly conserved
region on Env. The group employed a polypeptide sequence
derived from the N-terminus of the HIV coreceptor CCR5 to
develop the trispecific CAR to demonstrate enhanced in
vitro anti-HIV potency compared to the bispecific CAR (108). Binding of either CAR in dual or tandem systems to its
associated antigen is sufficient to stimulate full T cell
activation. Moreover, cancer studies have demonstrated that
enhanced T-cell function is observed when both targets are
present (106). A preliminary study of anti-HIV dual CARs
designed with CD4 antigen recognition domains linked to
bnAb 17b scFvs or carbohydrate recognition domains of a
human C-type lectin (109) transduced into CD8 T cells
showed enhanced potency against genetically diverse strains of
HIV when compared to single CD4 antigen recognition domain-
based CARs in vitro. Furthermore, Liu et al. demonstrated that
bispecific CAR function is influenced by the length between
various antigen-binding domains to inform future studies
incorporating dual antigen recognition domain designs (110). Subsequent studies have included CD4-based duo CARs linked
to various bnAb scFvs or the carbohydrate recognition domain of a human C-type lectin receptor and have all shown long -term
suppression of HIV in vitro (109, 111). Most recently, a study
describes duo CARs targeting two or three antigens
simultaneously via HIV neutralising antibody fragments in
transduced primary human T cells in vitro (111). Multi-
specific CAR - T cells were able to eliminate PBMCs infected
with single bnAb-resistant HIV strains and displayed long-term
control of HIV infection in vivo and the prevention of CD4 T cell
loss (111). Further, the group demonstrated long-term control of
HIV infection in vivo and prevented the loss of CD4+ T cells
during HIV infection using a humanized NSG mouse model
(111). Advances in Developing CAR-T Cell
Therapy for HIV Cure py
Successful immunotherapies for chronic viral infections, such as
HIV, would need to overcome challenges of immune escape, off- April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 8 York et al. Reducing the HIV Reservoir A
B
D
C
FIGURE 2 | Engineering CAR-T cells for improved function. (A) DuoCARs target independent antigens by CAR co-expression. (B) TanCARs adopt a tandem antigen
recognition domain to target multiple antigens. (C) CD3z and costimulatory domains are split between independent antigen recognition domains in SplitCARs for T cell
activation upon recognition of both antigens. (D) Upon recognition of antigen, synNotch receptors undergo transmembrane cleavage and release of their intracellular
transcriptional domain which in turn induces the transcription of a secondary CAR for recognition of a second antigen and T cell activation. (E) convertibleCAR, a universal
system where the antigen-targeting domain and the T cell signalling unit are split. Effector cells express CARs incorporating NKG2D, the natural receptor of the MIC/ULBP
ligand family. MicAbody, comprised of the MICA ligand bound to the scFv of an antibody of interest, is administered separately. (F) DARPin linked to a Cage protein bind
to surface antigens. Upon binding of a secondary DARPin linked to a key protein to its antigen, the latch is released exposing Bim to bind to a Bcl-2 CAR. CAR, chimeric
antigen receptor; NKG2D, MICA receptor; MICA, NKG2D ligand; DARPin, designer ankyrin repeat proteins. Created with BioRender. Adapted from Guedan et al. (105). FIGURE 2 | Engineering CAR-T cells for improved function. (A) DuoCARs target independent antigens by CAR co-expression. (B) TanCARs adopt a tandem antigen
recognition domain to target multiple antigens. (C) CD3z and costimulatory domains are split between independent antigen recognition domains in SplitCARs for T cell
activation upon recognition of both antigens. (D) Upon recognition of antigen, synNotch receptors undergo transmembrane cleavage and release of their intracellular
transcriptional domain which in turn induces the transcription of a secondary CAR for recognition of a second antigen and T cell activation. (E) convertibleCAR, a universal
system where the antigen-targeting domain and the T cell signalling unit are split. Effector cells express CARs incorporating NKG2D, the natural receptor of the MIC/ULBP
ligand family. MicAbody, comprised of the MICA ligand bound to the scFv of an antibody of interest, is administered separately. (F) DARPin linked to a Cage protein bind
to surface antigens. Frontiers in Immunology | www.frontiersin.org Advances in Developing CAR-T Cell
Therapy for HIV Cure Using a bicistronic lentiviral vector to allow for
simultaneous expression of both CARs, the group targeted dual
antigens on HIV env, first the mD1.22 domain to its CD4
binding site, and subsequently the m36.4 domain to its
conformationally exposed binding site (111). Collectively, these
data demonstrate that dual-targeting antigen recognition
domains are important in overcoming immune escape in the
context of an anti-HIV immunotherapy. Overcoming Off-Target Toxicitites
HIV specific CAR-T cells have demonstrated long-term safety in
the clinical setting (87–89). However, the primary concern with Tissue Microenvironment A major shared challenge to target cell clearance in chronic HIV
infection and cancer is the need for T cell localisation and
exertion of effector functions within the immunosuppressive
tissue microenvironment. Anti-HIV CAR T cells that over-
express the chemokine receptor CXCR5, to promote trafficking
into lymphoid tissues have been developed (119). This was based
on the identification of follicular CD8 T cells expressing CXCR5
and their control of HIV replication within lymphoid tissue
(120), in addition to a proof-of-concept study demonstrating
CXCR5 overexpression in CD8 T cells led to localisation within
the germinal centres of lymphoid tissue in macaques (121). These CD4-based CXCR5 expressing CAR T cells
demonstrated suppression of SIV infection in vitro and
trafficking in response to CXCL13 in migration assays using
transwells and lymph node organoid cultures (122). Most
recently, Barber-Anxthelm et al. employed hematopoietic stem
cells modified with CD4 ligand based CARs to demonstrate
trafficking and multilineage engraftment at lymphoid germinal
centres, gut associated lymphoid tissue and the central nervous
system (123). Furthermore, CAR expressing cells persisted for
nearly two years at these sites of tissue-associated viral reservoirs
(123). Since lymph nodes are important sites of HIV reservoirs,
these data highlight important innovations that can be included
in the development of a dual-targeting HIV-specific CAR-T cell
with increased capacities to traffic to anatomical sites that
contain large fractions of the latent HIV reservoir. Both the
solid tumour microenvironment and lymph nodes can be
characterised by the secretion of various cytokines and growth
factors by stromal and immune cells to remodel the extracellular
matrix of these tissues and contribute to the suppression of T cell
responses. To overcome such barriers, CAR T cells have been
engineered to secrete extracellular matrix modifying enzymes Another strategy to reduce toxicity is by allowing for
controlled expression of CAR T cells. For example, CARs are
directed against an inert molecule which is conjugated to an
antigen binding antibody. By infusing various antibodies, the
universal CAR can be redirected against various antigens and
CAR activity can be controlled by regulating the times and
amount of antibody infusions. Most recently Herzig et al. engineered a novel version of universal CAR known as
convertibleCAR (Figure 2E) where they combine the cytotoxic
T cell with many antibodies (117). Overcoming Off-Target Toxicitites Overcoming Off-Target Toxicitites g
g
HIV specific CAR-T cells have demonstrated long-term safety in
the clinical setting (87–89). However, the primary concern with April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 9 Reducing the HIV Reservoir York et al. CD4-ligand based anti-HIV CAR-T cell therapies is the capacity
for CD4-MHC II interactions resulting in non-specific T cell
activation. Fortunately, cytolysis of MHC II-expressing cells has
not been demonstrated in CD4-based anti-HIV CAR-T cell
studies (91, 110). Development of anti-HIV bnAb-based CAR
T cell therapies specifically targeting the HIV envelope
glycoproteins is safer than targeting the CD4 binding partners. It is well established that the non-covalent interaction of the
gp120 and gp41 envelope glycoproteins of HIV is disrupted by
soluble CD4 binding, resulting in irreversible gp120 release
(112). Since CAR-T cell recognition of its specific antigen is an
MHC I independent process, another important consideration is
the recognition of cell-free or virion-associated HIV env by HIV
specific CAR-T cells leading to activation and potentially lethal
cytokine release. Although this area has not been extensively
explored in HIV CAR-T cell therapies, dual targeting of HIV env
with cell surface markers may provide a novel strategy to
mitigate off-target cytotoxicity resulting from non-specific T
cell responses to cell-free or virion-associated HIV env. protein called MIC-A with A being any antibody of interest. Consequently, neither the CAR-T cells nor the MicAbody can
kill target cells until they have bound to each other and the
MicAbody has specifically engaged its epitope on an HIV-
infected cell to create an immunologic synapse. This strategy
compensated for the cleavage of soluble ligands (MICA and
ULBP) for the natural NKG2D receptor from the surface of
infected cells, thereby compromising the ability of CTL and NK
cells to kill infected cells (118). The possibility of multiplexing a single convertibleCAR cell
with several MicAbodies makes this platform quite promising for
tackling multiple diseases or pathogen variants and for avoiding
the universal problem of drug resistance. Another advantage of
this type of CAR T cells is the exclusive ligand-receptor
interaction where CAR can be introduced into patients in an
inert state. Only after the administration of the specific
MicAbody or MicAbody mix will the CAR-T get activated. This increases safety and controls dependent on both the dose
and timing of administered HIV-specific MicAbody. By contrast,
classical CAR T cells are ‘‘on’’ all the time. Overcoming Off-Target Toxicitites Using a platform that
is inert unless both parts are present can potentially
enhance safety. p
However, targeting two antigens may lead to increased risk
for off-target toxicity following CAR-T cell treatment. Strategies
to mitigate off-target cytotoxicity include synNotch, LOCKR and
convertible systems where CARs are activated upon recognition
of a specific combination of antigens. The primary signal can be
provided following stimulation of the CAR linked to the CD3z
chain whilst co-stimulation is split and provided by another CAR
to limit non-specific activation (Figure 2C) (113, 114). A more
recent strategy employs a two-step process where activation of
the first receptor known as the synthetic Notch (synNotch)
induces the expression of a secondary antigen specific receptor
(Figure 2D) (115). This AND-gate recognition system ensures T
cell activation only occurs when both antigens are expressed on
target cells thus enabling precision killing and sparing cells
expressing either of the single antigens in isolation. Other
AND-gating strategies involve dimerising protein switches,
which can serve as lock and keys that are activated only upon
the two independent antigens (lock-bound and key-bound)
present together. An interesting example of this is described in
Lajoie et al. (116) (Figure 2F). Frontiers in Immunology | www.frontiersin.org Tissue Microenvironment This is crucial for controlling
HIV as different variants exist and there would be no long-term
success with a CAR-T cell carrying a single antibody to fight
HIV. In this scenario, Herzig et al. used a ligand from the MIC/
ULBP family, expressed on stressed cells, and its receptor
NKG2D, expressed on CTL and NK cells. They engineered the
convertibleCAR cell T cell to express NKG2D thus converting
them into a potent killer, but only when bound to its partner, a April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 10 Reducing the HIV Reservoir York et al. phosphorylated upon IL-2 engagement of its receptor (136). After translocation to the nucleus STAT5 promotes T cell
proliferation and cell cycle progression. By engineering
orthogonal IL-2 cytokine-receptors pairs, Sockolosky et al. demonstrated that orthogonal IL-2 potently activated STAT5
on orthogonal IL-2 receptor transduced primary mouse CAR – T
cells when compared to wild-type resulting in specific expansion
of primary mouse CAR – T cells with negligible toxicity whilst
retaining anti-tumour effector functions (135). As mentioned
previously, 41BB co-stimulated CARs induce a delayed T cell
effector response and increased persistence whilst CD28 co-
stimulated CARs lead to quicker T cell activation, proliferation,
cytolysis. Dual-CARs incorporating distinct CD28 and 4-1BB
costimulatory domains can combine both antigen-driven
expansion and long-term persistence with potent effector
functions (101). Using LRAs to stimulate robust reactivation of
latent reservoirs, may help in antigen dependent expansion and
persistence of HIV-specific CAR T cells during clinical
translation of HIV specific T cell therapies. Thus, a
combinatorial approach of LRA and CAR T cells can be
expected to aid in increased in vivo expansion and persistence. such as heparinase to degrade heparin sulphate proteoglycans in
the extracellular matrix (124). These CAR-T cells demonstrated
an improved capacity to infiltrate xenograft tumours in mice and
prolonged survival compared with CAR T cells lacking
heparinase expression (124). Such strategies could be included
in future HIV CAR designs. After overcoming the physical barriers and localising to the
tumour microenvironment, upregulation of inhibitory ligands
such as PDL-1 that bind to inhibitory receptor PD1 to suppress
effector T cell responses have been observed (125). Several
approaches to overcome the effect of PD-1 using CAR T cells
have been reported and are succinctly described in Yoon et al. (126). Expansion and Persistence p
The low number of HIV-infected target cells comprising the
latent HIV reservoir and their low antigen density upon
reactivation leads to challenges surrounding the expansion and
persistence of HIV specific CAR – T cells. One strategy involving
orthogonal IL-2 cytokine-receptor pairs may overcome poor
engineered T cell expansion and persistence upon patient
infusion (135). It is known that transcription factor STAT5 is Susceptibility to HIV Infection Suscept b ty to
ect o
CD4 ligand-based HIV specific CARs have illustrated significant
barriers to HIV cure due to their susceptibility to HIV infection
(137). To overcome this limitation, one strategy describes co-
expressing short hairpin RNA (shRNA) sequences targeting
CCR5 and HIV LTR to silence target genes (138). It was
shown that expressing CCR5 and HIV LTR shRNA decreased
the risk of CAR T cell infection leading to overall increased levels
and persistence whilst maintaining HIV specific effector
functions in vitro (138). Targeted CAR integration at the
CCR5 locus in primary human T cells has been also
demonstrated by RNA-based nuclease expression coupled with
adeno-associated virus-mediated delivery of a CCR5-targeting
donor template and was shown to effectively supress viral
replication in primary HIV infected human T cells (99). Thus,
targeted integration of the CAR into the CCR5 locus is an
important milestone in developing robust T cell therapies
against HIV. More recently, Anthony-Gonda et al. showed
infection resistance when the C46 viral fusion inhibitor peptide
was co-expressed in anti- HIV CAR T cells without interfering
with their effector function (111). Various studies have shown
that anti-HIV bnAbs may be used as an alternative to CD4
antigen recognition domains in HIV specific CARs to stimulate
specific T cell activation and killing of HIV infected cells without
HIV infection of CAR T cells (99, 102, 111). These studies
demonstrate the feasibility of utilising bnAb scFv in CAR
constructs as immunotherapies for HIV infection. Cytokine Release Syndrome The risk of cytokine release syndrome (CRS) is most acute during
the first infusion when the cancer tumour burden is at its peak
(131). However, the number of HIV-infected target cells in
patients on ART is dramatically less than target cell numbers
in individuals with leukaemia and HIV reactivation of latently
infected cells is characterised by low surface antigen density
(132). Tumour specific antigens provoke a potent immune
response relative to reactivated latently infected resting
memory T cells. For this reason, CRS has not been extensively
studied in the context of HIV specific CAR – T cells in vivo. However, should life threatening CRS ensue, it would be
beneficial to incorporate safety switches to “turn off” the CAR
T cells. The popular switches currently being trialled are based on
inducible caspase-9 (133) and truncated epidermal growth factor
receptor (tEGFR) (elimination induced by the infusion of
clinically approved antibody Cetuximab) (134). Tissue Microenvironment Engineered CAR-T cells co-transduced with truncated
PD-1 receptors lacking intracellular signalling functions which
still bind PDL-1 but are unable to propagate inhibitory signals
have been developed and demonstrated resistance to T cell
exhaustion mediated by PDL-1 and prolonged survival in mice
bearing xenograft tumours, compared with CAR T cells lacking
these truncated PD-1 receptors (127). More recently, Jiang et al. demonstrated that expression of PD-1 dominant negative
receptor (DNR) in a 3BNC117 bnAb scFv based CAR T cell
resulted in superior lytic and functional responses in vitro and in
vivo when compared to 3BNC117 bnAb scFv CAR expression
alone (128). PD-1 receptor expression has also been disrupted
completely via CRISPR/Cas9 resulting in augmentation of CAR
T cell mediated killing of tumor cells in vitro and enhanced
clearance of PDL-1 expressing tumor xenografts in vivo (129). The complexity of the tissue microenvironment continues to
drive novel CAR design as previously outlined by our
group (130). REFERENCES Persistence in Effector Memory CD4+ T-Cells. J Clin Invest (2022) 132:1–17. doi: 10.1172/JCI154422 Persistence in Effector Memory CD4+ T-Cells. J Clin Invest (2022) 132:1–17. doi: 10.1172/JCI154422 1. Barré-Sinoussi F, Ross AL, Delfraissy J-F. Past, Present and Future: 30 Years
of HIV Research. Nat Rev Microbiol (2013) 11:877–83. doi: 10.1038/
nrmicro3132 1. Barré-Sinoussi F, Ross AL, Delfraissy J-F. Past, Present and Future: 30 Years
of HIV Research. Nat Rev Microbiol (2013) 11:877–83. doi: 10.1038/
nrmicro3132 10. Murray AJ, Kwon KJ, Farber DL, Siliciano RF. The Latent Reservoir for
HIV-1: How Immunologic Memory and Clonal Expansion Contribute to
HIV-1 Persistence. J Immunol (2016) 197:407–17. doi: 10.4049/
jimmunol.1600343 2. Arts EJ, Hazuda DJ. HIV-1 Antiretroviral Drug Therapy. Cold Spring Harb
Perspect Med (2012) 2:1–23. doi: 10.1101/cshperspect.a007161 2. Arts EJ, Hazuda DJ. HIV-1 Antiretroviral Drug Therapy. Cold Spring Harb
Perspect Med (2012) 2:1–23. doi: 10.1101/cshperspect.a007161 11. McNab F, Mayer-Barber K, Sher A, Wack A, O’Garra A. Type I Interferons
in Infectious Disease. Nat Rev Immunol (2015) 15:87–103. doi: 10.1038/
nri3787 3. Sáez-Cirión A, Bacchus C, Hocqueloux L, Avettand-Fenoel V, Girault I,
Lecuroux C, et al. Post-Treatment HIV-1 Controllers With a Long-Term
Virological Remission After the Interruption of Early Initiated
Antiretroviral Therapy ANRS VISCONTI Study. PloS Pathog (2013) 9:
e1003211. doi: 10.1371/journal.ppat.1003211 12. Pestka S, Krause CD, Walter MR. Interferons, Interferon-Like Cytokines,
and Their Receptors. Immunol Rev (2004) 202:8–32. doi: 10.1111/j.0105-
2896.2004.00204.x 13. Witte K, Witte E, Sabat R, Wolk K. IL-28a, IL-28B, and IL-29: Promising
Cytokines With Type I Interferon-Like Properties. Cytokine Growth Factor
Rev (2010) 21:237–51. doi: 10.1016/j.cytogfr.2010.04.002 4. Samri A, Bacchus-Souffan C, Hocqueloux L, Avettand-Fenoel V, Descours B,
Theodorou I, et al. Polyfunctional HIV-Specific T Cells in Post-Treatment
Controllers. AIDS (2016) 30:2299–302. doi: 10.1097/QAD.0000000000001195 4. Samri A, Bacchus-Souffan C, Hocqueloux L, Avettand-Fenoel V, Descours B,
Theodorou I, et al. Polyfunctional HIV-Specific T Cells in Post-Treatment
Controllers. AIDS (2016) 30:2299–302. doi: 10.1097/QAD.0000000000001195 14. Bertram KM, Botting RA, Baharlou H, Rhodes JW, Rana H, Graham JD, et al. Identification of HIV Transmitting CD11c + Human Epidermal Dendritic
Cells. Nat Commun (2019) 10:2759. doi: 10.1038/s41467-019-10697-w 5. Deeken JF, Tjen-A-Looi A, Rudek MA, Okuliar C, Young M, Little RF, et al. The Rising Challenge of Non-AIDS-Defining Cancers in HIV-Infected
Patients. Clin Infect Dis (2012) 55:1228–35. doi: 10.1093/cid/cis613 15. Nasr N, Maddocks S, Turville SG, Harman AN, Woolger N, Helbig KJ, et al. HIV-1 Infection of Human Macrophages Directly Induces Viperin Which
Inhibits Viral Production. CONCLUSIONS Many LRAs are toxic, can
reactivate HIV but fail to kill the target CD4 T cells, and some
inhibit endogenous CD8 T cell function. Currently, there is no
licensed treatment for the reduction of the latent HIV reservoir
and novel approaches are required to tackle latency. Individual
strategies are unikely to eliminate the entire HIV reservoir. However, reactivation via multiple approaches such as LRAs or
IFN combined with immunotherapy using CD8 CAR T cells to
kill reactivated cells need to be tested further. CAR T cells are one
such promising immunotherapeutic approach for killing HIV
infected cells. An important feature of CD8 CAR T cell
immunotherapy is that HIV infected cells that evade CD8 T
recognition via mutations in MHC-I restricted epitopes or
downregulation of MHC-I will be targeted by CAR T cells
which recognise surface antigens independent of MHC-I
presentation. IFN has also many advantages over classical
LRAs: it does not impair the CD8 antiviral function, prevents
HIV spread, upregulate MHC-I and is not toxic. In addition to
the inclusion of scFvs from bnAbs to induce CAR T cells
cytotoxic anti-HIV effects, one could also consider arming the
CAR T cells to secrete cytokines that activate latently infected
cells and to increase their ability to migrate to key sites of HIV
infection such as lymph nodes or mucosal lymphoid tissue or
even to immune privileged sites such as the brain via the expression of specific homing markers. First generation CAR-T
cells were not capable of reducing viral burden permanently in
most clinical studies while third and fourth generation led to
activation-induced cell death (AICD). However, second
generation CAR T cells are currently the most promising
candidates as they show superior expansion and reduced HIV
infected cells in many in vitro and humanised mouse studies. As
CAR designs are constantly evolving, their modifications may
provide more effective cell therapies to cure HIV infected
patients in the future. FUNDING The Westmead Institute for Medical Research paid for open
access publication. JY was supported by a grant from the
Australian Centre for HIV and Hepatitis Virology Research
(ACH2) granted to AC, KM, NN and KG. SP was supported
by the Delaney AIDS Research Enterprise (DARE) to Find a
Cure funded by the US National Institutes of Health
(1UM1AI164560-01) and the Australian National Health and
Medical Research Council (APP1149990). CONCLUSIONS Immunotherapy for the treatment of infection and malignancy
has revolutionised therapeutic options. While one major aspect
of HIV research focusses on developing alternative or April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org Reducing the HIV Reservoir York et al. complementary methods to LRAs to reactivate HIV, the other
exciting aspect focusses on harnessing the immune system more
effectively to eliminate HIV reactivated cells. However, none of
the attempts at reducing latently infected cells have yet succeeded
as CD8 T cells are dysfunctional in chronic HIV infection and
LRAs have been suboptimal so far. Many LRAs are toxic, can
reactivate HIV but fail to kill the target CD4 T cells, and some
inhibit endogenous CD8 T cell function. Currently, there is no
licensed treatment for the reduction of the latent HIV reservoir
and novel approaches are required to tackle latency. Individual
strategies are unikely to eliminate the entire HIV reservoir. However, reactivation via multiple approaches such as LRAs or
IFN combined with immunotherapy using CD8 CAR T cells to
kill reactivated cells need to be tested further. CAR T cells are one
such promising immunotherapeutic approach for killing HIV
infected cells. An important feature of CD8 CAR T cell
immunotherapy is that HIV infected cells that evade CD8 T
recognition via mutations in MHC-I restricted epitopes or
downregulation of MHC-I will be targeted by CAR T cells
which recognise surface antigens independent of MHC-I
presentation. IFN has also many advantages over classical
LRAs: it does not impair the CD8 antiviral function, prevents
HIV spread, upregulate MHC-I and is not toxic. In addition to
the inclusion of scFvs from bnAbs to induce CAR T cells
cytotoxic anti-HIV effects, one could also consider arming the
CAR T cells to secrete cytokines that activate latently infected
cells and to increase their ability to migrate to key sites of HIV
infection such as lymph nodes or mucosal lymphoid tissue or
even to immune privileged sites such as the brain via the complementary methods to LRAs to reactivate HIV, the other
exciting aspect focusses on harnessing the immune system more
effectively to eliminate HIV reactivated cells. However, none of
the attempts at reducing latently infected cells have yet succeeded
as CD8 T cells are dysfunctional in chronic HIV infection and
LRAs have been suboptimal so far. AUTHOR CONTRIBUTIONS JY primarily wrote the review under the guidance of KG, AC and
NN. SP and KM provided intellectual input. All authors
contributed to the article and approved the submitted version. REFERENCES Mortality in the Highly Active
Antiretroviral Therapy Era: Changing Causes of Death and Disease in the
HIV Outpatient Study. J Acquir Immune Defic Syndr (2006) 43:27–34. doi: 10.1097/01.qai.0000233310.90484.16 25. Oxenius A, Price DA, Easterbrook PJ, O’Callaghan CA, Kelleher AD,
Whelan JA, et al. Early Highly Active Antiretroviral Therapy for Acute
HIV-1 Infection Preserves Immune Function of CD8+ and CD4+ T
Lymphocytes. Proc Natl Acad Sci USA (2000) 97:3382–7. doi: 10.1073/
pnas.97.7.3382 44. Vansant G, Bruggemans A, Janssens J, Debyser Z. Block-And-Lock
Strategies to Cure HIV Infection. Viruses (2020) 12:E84. doi: 10.3390/
v12010084 26. Sun JC, Williams MA, Bevan MJ. CD4+ T Cells are Required for the
Maintenance, Not Programming, of Memory CD8+ T Cells After Acute
Infection. Nat Immunol (2004) 5:927–33. doi: 10.1038/ni1105 45. Ahlenstiel CL, Suzuki K, Marks K, Symonds GP, Kelleher AD. Controlling
HIV-1: Non-Coding RNA Gene Therapy Approaches to a Functional Cure. Front Immunol (2015) 6:474. doi: 10.3389/fimmu.2015.00474 27. Takata H, Buranapraditkun S, Kessing C, Fletcher JLK, Muir R, Tardif V,
et al. Delayed Differentiation of Potent Effector CD8+ T Cells Reducing
Viremia and Reservoir Seeding in Acute HIV Infection. Sci Transl Med
(2017) 9:1–22. doi: 10.1126/scitranslmed.aag1809 46. Boukli NM, Shetty V, Cubano L, Ricaurte M, Coelho-Dos-Reis J, Nickens Z,
et al. Unique and Differential Protein Signatures Within the Mononuclear
Cells of HIV-1 and HCV Mono-Infected and Co-Infected Patients. Clin
Proteomics (2012) 9:11. doi: 10.1186/1559-0275-9-11 28. Sáez-Cirión A, Lacabaratz C, Lambotte O, Versmisse P, Urrutia A, Boufassa
F, et al. HIV Controllers Exhibit Potent CD8 T Cell Capacity to Suppress
HIV Infection Ex Vivo and Peculiar Cytotoxic T Lymphocyte Activation
Phenotype. PNAS (2007) 104:6776–81. doi: 10.1073/pnas.0611244104 47. Vozzolo L, Loh B, Gane PJ, Tribak M, Zhou L, Anderson I, et al. Gyrase B
Inhibitor Impairs HIV-1 Replication by Targeting Hsp90 and the Capsid
Protein. J Biol Chem (2010) 285:39314–28. doi: 10.1074/jbc.M110.155275 29. Borrow P, Lewicki H, Wei X, Horwitz MS, Peffer N, Meyers H, et al. Antiviral Pressure Exerted by HIV-L-Specific Cytotoxic T Lymphocytes
(CTLs) During Primary Infection Demonstrated by Rapid Selection of CTL
Escape Virus. Nat Med (1997) 3:205–11. doi: 10.1038/nm0297-205 48. Rice AP. The HIV-1 Tat Protein: Mechanism of Action and Target for HIV-
1 Cure Strategies. Curr Pharm Des (2017) 23:4098–102. doi: 10.2174/
1381612823666170704130635 30. Cartwright EK, Spicer L, Smith SA, Lee D, Fast R, Paganini S, et al. REFERENCES Course
and Clinical Significance of CD8+ T-Cell Counts in a Large Cohort of HIV-
Infected Individuals. J Infect Dis (2015) 211:1726–34. doi: 10.1093/infdis/
jiu669 19. Sandler NG, Bosinger SE, Estes JD, Zhu RTR, Tharp GK, Boritz E, et al. Type
I Interferon Responses in Rhesus Macaques Prevent SIV Infection and Slow
Disease Progression. Nature (2014) 511:601–5. doi: 10.1038/nature13554 38. Perdomo-Celis F, Taborda NA, Rugeles MT. CD8+ T-Cell Response to HIV
Infection in the Era of Antiretroviral Therapy. Front Immunol (2019)
10:1896. doi: 10.3389/fimmu.2019.01896 20. Li Q, Estes JD, Schlievert PM, Duan L, Brosnahan AJ, Southern PJ, et al. Glycerol Monolaurate Prevents Mucosal SIV Transmission. Nature (2009)
458:1034–8. doi: 10.1038/nature07831 39. Petrovas C, Casazza JP, Brenchley JM, Price DA, Gostick E, Adams WC,
et al. PD-1 is a Regulator of Virus-Specific CD8+ T Cell Survival in HIV
Infection. J Exp Med (2006) 203:2281–92. doi: 10.1084/jem.20061496 21. Kolumam GA, Thomas S, Thompson LJ, Sprent J, Murali-Krishna K. Type I
Interferons Act Directly on CD8 T Cells to Allow Clonal Expansion and
Memory Formation in Response to Viral Infection. J Exp Med (2005)
202:637–50. doi: 10.1084/jem.20050821 40. Dornadula G, Zhang H, VanUitert B, Stern J, Livornese L, Ingerman MJ,
et al. Residual HIV-1 RNA in Blood Plasma of Patients Taking Suppressive
Highly Active Antiretroviral Therapy. JAMA (1999) 282:1627–32. doi: 10.1001/jama.282.17.1627 22. Mikulak J, Oriolo F, Zaghi E, Di Vito C, Mavilio D. Natural Killer Cells in
HIV-1 Infection and Therapy. AIDS (2017) 31:2317–30. doi: 10.1097/
QAD.0000000000001645 41. Schweneker M, Favre D, Martin JN, Deeks SG, McCune JM. HIV-Induced
Changes in T Cell Signaling Pathways. J Immunol (2008) 180:6490–500. doi: 10.4049/jimmunol.180.10.6490 23. Ward J, Bonaparte M, Sacks J, Guterman J, Fogli M, Mavilio D, et al. HIV
Modulates the Expression of Ligands Important in Triggering Natural Killer
Cell Cytotoxic Responses on Infected Primary T-Cell Blasts. Blood (2007)
110:1207–14. doi: 10.1182/blood-2006-06-028175 42. Wherry EJ, Ha S-J, Kaech SM, Haining WN, Sarkar S, Kalia V,
et al. Molecular Signature of CD8+ T Cell Exhaustion During Chronic
Viral Infection. Immunity (2007) 27:670–84. doi: 10.1016/j.immuni. 2007.09.006 24. Douek DC, Brenchley JM, Betts MR, Ambrozak DR, Hill BJ, Okamoto Y,
et al. HIV Preferentially Infects HIV-Specific CD4+ T Cells. Nature (2002)
417:95–8. doi: 10.1038/417095a 43. Palella FJ, Baker RK, Moorman AC, Chmiel JS, Wood KC, Brooks JT, et al. HIV Outpatient Study Investigators. REFERENCES Blood (2012) 120:778–88. doi: 10.1182/blood-
2012-01-407395 6. Sonti S, Sharma AL, Tyagi M. HIV-1 Persistence in the CNS: Mechanisms of
Latency, Pathogenesis and an Update on Eradication Strategies. Virus Res
(2021) 303:198523. doi: 10.1016/j.virusres.2021.198523 7. O’Neil TR, Hu K, Truong NR, Arshad S, Shacklett BL, Cunningham AL,
et al. The Role of Tissue Resident Memory CD4 T Cells in Herpes Simplex
Viral and HIV Infection. Viruses (2021) 13:359. doi: 10.3390/v13030359 16. Harman AN, Nasr N, Feetham A, Galoyan A, Alshehri AA, Rambukwelle D,
et al. HIV Blocks Interferon Induction in Human Dendritic Cells and
Macrophages by Dysregulation of TBK1. J Virol (2015) 89:6575–84. doi: 10.1128/JVI.00889-15 8. Deeks SG, Archin N, Cannon P, Collins S, Jones RB, de Jong MAWP, et al. Research Priorities for an HIV Cure: International AIDS Society Global
Scientific Strategy 2021. Nat Med (2021) 27:2085–98. doi: 10.1038/s41591-
021-01590-5 8. Deeks SG, Archin N, Cannon P, Collins S, Jones RB, de Jong MAWP, et al. Research Priorities for an HIV Cure: International AIDS Society Global
Scientific Strategy 2021. Nat Med (2021) 27:2085–98. doi: 10.1038/s41591-
021-01590-5 17. Doehle BP, Chang K, Rustagi A, McNevin J, McElrath MJ, Gale M. Vpu
Mediates Depletion of Interferon Regulatory Factor 3 During HIV Infection
by a Lysosome-Dependent Mechanism. J Virol (2012) 86:8367–74. doi: 10.1128/JVI.00423-12 9. Duette G, Hiener B, Morgan H, Mazur FG, Mathivanan V, Horsburgh BA,
et al. The HIV-1 Proviral Landscape Reveals Nef Contributes to HIV-1 9. Duette G, Hiener B, Morgan H, Mazur FG, Mathivanan V, Horsburgh BA,
et al. The HIV-1 Proviral Landscape Reveals Nef Contributes to HIV-1 April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 12 Reducing the HIV Reservoir York et al. 18. Stacey AR, Norris PJ, Qin L, Haygreen EA, Taylor E, Heitman J, et al. Induction of a Striking Systemic Cytokine Cascade Prior to Peak Viremia in
Acute Human Immunodeficiency Virus Type 1 Infection, in Contrast to
More Modest and Delayed Responses in Acute Hepatitis B and C Virus
Infections. J Virol (2009) 83:3719–33. doi: 10.1128/JVI.01844-08 36. Roberts ER, Carnathan DG, Li H, Shaw GM, Silvestri G, Betts MR. Collapse
of Cytolytic Potential in SIV-Specific CD8+ T Cells Following Acute SIV
Infection in Rhesus Macaques. PloS Pathog (2016) 12:1–21. doi: 10.1371/
journal.ppat.1006135 37. Helleberg M, Kronborg G, Ullum H, Ryder LP, Obel N, Gerstoft J. REFERENCES CD8(+)
Lymphocytes Are Required for Maintaining Viral Suppression in SIV-
Infected Macaques Treated With Short-Term Antiretroviral Therapy. Immunity (2016) 45:656–68. doi: 10.1016/j.immuni.2016.08.018 49. Kessing CF, Nixon CC, Li C, Tsai P, Takata H, Mousseau G, et al. In Vivo
Suppression of HIV Rebound by Didehydro-Cortistatin A, a “Block-And-
Lock” Strategy for HIV-1 Treatment. Cell Rep (2017) 21:600–11. doi: 10.1016/j.celrep.2017.09.080 31. Liu H, Liu L, Liu K, Bizargity P, Hancock WW, Visner GA. Reduced
Cytotoxic Function of Effector CD8+ T Cells Is Responsible for Indoleamine
2,3-Dioxygenase-Dependent Immune Suppression. J Immunol (2009)
183:1022–31. doi: 10.4049/jimmunol.0900408 50. Xiao Q, Guo D, Chen S. Application of CRISPR/Cas9-Based Gene Editing in
HIV-1/AIDS Therapy. Front Cell Infect Microbiol (2019) 9:69. doi: 10.3389/
fcimb.2019.00069 32. Jin X, Bauer DE, Tuttleton SE, Lewin S, Gettie A, Blanchard J, et al. Dramatic
Rise in Plasma Viremia After CD8(+) T Cell Depletion in Simian
Immunodeficiency Virus-Infected Macaques. J Exp Med (1999) 189:991–8. doi: 10.1084/jem.189.6.991 51. Thorlund K, Horwitz MS, Fife BT, Lester R, Cameron DW. Landscape
Review of Current HIV ‘Kick and Kill’ Cure Research - Some Kicking, Not
Enough Killing. BMC Infect Dis (2017) 17:595. doi: 10.1186/s12879-017-
2683-3 52. Rasmussen TA, Tolstrup M, Søgaard OS. Reversal of Latency as Part of a
Cure for HIV-1. Trends Microbiol (2016) 24:90–7. doi: 10.1016/
j.tim.2015.11.003 33. Leslie AJ, Pfafferott KJ, Chetty P, Draenert R, Addo MM, Feeney M, et al. HIV Evolution: CTL Escape Mutation and Reversion After Transmission. Nat Med (2004) 10:282–9. doi: 10.1038/nm992 34. Turnbull EL, Wong M, Wang S, Wei X, Jones NA, Conrod KE, et al. Kinetics
of Expansion of Epitope-Specific T Cell Responses During Primary HIV-1
Infection. J Immunol (2009) 182:7131–45. doi: 10.4049/jimmunol.0803658 53. Marsden MD, Wu X, Navab SM, Loy BA, Schrier AJ, DeChristopher BA,
et al. Characterization of Designed, Synthetically Accessible Bryostatin
Analog HIV Latency Reversing Agents. Virology (2018) 520:83–93. doi: 10.1016/j.virol.2018.05.006 35. Ferrari G, Korber B, Goonetilleke N, Liu MKP, Turnbull EL, Salazar-
Gonzalez JF, et al. Relationship Between Functional Profile of HIV-1
Specific CD8 T Cells and Epitope Variability With the Selection of Escape
Mutants in Acute HIV-1 Infection. PloS Pathog (2011) 7:e1001273. doi: 10.1371/journal.ppat.1001273 54. Pache L, Dutra MS, Spivak AM, Marlett JM, Murry JP, Hwang Y, et al. BIRC2/cIAP1 Is a Negative Regulator of HIV-1 Transcription and Can Be
Targeted by Smac Mimetics to Promote Reversal of Viral Latency. Cell Host
Microbe (2015) 18:345–53. REFERENCES Garrido C, Abad-Fernandez M, Tuyishime M, Pollara JJ, Ferrari G, Soriano-
Sarabia N, et al. Interleukin-15-Stimulated Natural Killer Cells Clear HIV-1-
Infected Cells Following Latency Reversal Ex Vivo. J Virol (2018) 92:1–13. doi: 10.1128/JVI.00235-18 85. Melenhorst JJ, Chen GM, Wang M, Porter DL, Chen C, Collins MA, et al. Decade-Long Leukaemia Remissions With Persistence of CD4+ CAR T
Cells. Nature (2022) 602:503–9. doi: 10.1038/s41586-021-04390-6 67. Chandrasekar AP, Cummins NW, Badley AD. The Role of the BCL-2 Family
of Proteins in HIV-1 Pathogenesis and Persistence. Clin Microbiol Rev
(2019) 33:1–25. doi: 10.1128/CMR.00107-19 86. Neelapu SS. Managing the Toxicities of CAR T-Cell Therapy. Hematol Oncol
(2019) 37(Suppl 1):48–52. doi: 10.1002/hon.2595 68. Ren Y, Huang SH, Patel S, Alberto WDC, Magat D, Ahimovic D, et al. BCL-2
Antagonism Sensitizes Cytotoxic T Cell–Resistant HIV Reservoirs to
Elimination Ex Vivo. J Clin Invest (2020) 130:2542–59. doi: 10.1172/
JCI132374 87. Walker RE, Bechtel CM, Natarajan V, Baseler M, Hege KM, Metcalf JA, et al. Long-Term In Vivo Survival of Receptor-Modified Syngeneic T Cells in
Patients With Human Immunodeficiency Virus Infection. Blood (2000)
96:467–74. doi: 10.1182/blood.V96.2.467 69. Tomaras GD, Yates NL, Liu P, Qin L, Fouda GG, Chavez LL, et al. Initial B-
Cell Responses to Transmitted Human Immunodeficiency Virus Type 1:
Virion-Binding Immunoglobulin M (IgM) and IgG Antibodies Followed by
Plasma Anti-Gp41 Antibodies With Ineffective Control of Initial Viremia. J Virol (2008) 82:12449–63. doi: 10.1128/JVI.01708-08 88. Mitsuyasu RT, Anton PA, Deeks SG, Scadden DT, Connick E, Downs MT,
et al. Prolonged Survival and Tissue Trafficking Following Adoptive Transfer
of CD4zeta Gene-Modified Autologous CD4(+) and CD8(+) T Cells in
Human Immunodeficiency Virus-Infected Subjects. Blood (2000) 96:785–93. doi: 10.1182/blood.V96.3.785.015k10_785_793 70. Bonsignori M, Liao H-X, Gao F, Williams WB, Alam SM, Montefiori DC,
et al. Antibody-Virus Co-Evolution in HIV Infection: Paths for HIV Vaccine
Development. Immunol Rev (2017) 275:145–60. doi: 10.1111/imr.12509 89. Deeks SG, Wagner B, Anton PA, Mitsuyasu RT, Scadden DT, Huang C, et al. A Phase II Randomized Study of HIV-Specific T-Cell Gene Therapy in
Subjects With Undetectable Plasma Viremia on Combination Antiretroviral
Therapy. Mol Ther (2002) 5:788–97. doi: 10.1006/mthe.2002.0611 71. Jacobson JM, Colman N, Ostrow NA, Simson RW, Tomesch D, Marlin L,
et al. Passive Immunotherapy in the Treatment of Advanced Human
Immunodeficiency Virus Infection. J Infect Dis (1993) 168:298–305. doi: 10.1093/infdis/168.2.298 90. Iwamoto N, Patel B, Song K, Mason R, Bolivar-Wagers S, Bergamaschi C,
et al. REFERENCES doi: 10.1016/j.chom.2015.08.009 April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 13 York et al. Reducing the HIV Reservoir 55. Kim Y, Anderson JL, Lewin SR. Getting the “Kill” Into “Shock and Kill”:
Strategies to Eliminate Latent HIV. Cell Host Microbe (2018) 23:14–26. doi: 10.1016/j.chom.2017.12.004 73. Halper-Stromberg A, Nussenzweig MC. Towards HIV-1 Remission:
Potential Roles for Broadly Neutralizing Antibodies. J Clin Invest (2016)
126:415–23. doi: 10.1172/JCI80561 74. Lal KG, Naluyima P, Costanzo MC, Kijak GH, Eller LA, Creegan M, et al. Terminally Differentiated CD8 Effector T Cells Have NK-Like Features and
are Potent Mediators of HIV-Specific ADCC. J Immunol (2017) 198:125.4–4. 56. Walker-Sperling VE, Pohlmeyer CW, Tarwater PM, Blankson JN. The Effect
of Latency Reversal Agents on Primary CD8+ T Cells: Implications for
Shock and Kill Strategies for Human Immunodeficiency Virus Eradication. EBioMedicine (2016) 8:217–29. doi: 10.1016/j.ebiom.2016.04.019 75. Caskey M. Broadly Neutralizing Antibodies for the Treatment and
Prevention of HIV Infection. Curr Opin HIV AIDS (2020) 15:49–55. doi: 10.1097/COH.0000000000000600 57. Tsai A, Irrinki A, Kaur J, Cihlar T, Kukolj G, Sloan DD, et al. Toll-Like
Receptor 7 Agonist GS-9620 Induces HIV Expression and HIV-Specific
Immunity in Cells From HIV-Infected Individuals on Suppressive
Antiretroviral Therapy. J Virol (2017) 91:e02166-16. doi: 10.1128/
JVI.02166-16 76. Gaudinski MR, Houser KV, Doria-Rose NA, Chen GL, Rothwell RSS,
Berkowitz N, et al. Safety and Pharmacokinetics of Broadly Neutralising
Human Monoclonal Antibody VRC07-523LS in Healthy Adults: A Phase 1
Dose-Escalation Clinical Trial. Lancet HIV (2019) 6:e667–e679. doi: 10.1016/S2352-3018(19)30181-X 58. Petravic J, Rasmussen TA, Lewin SR, Kent SJ, Davenport MP. Relationship
Between Measures of HIV Reactivation and Decline of the Latent Reservoir
Under Latency-Reversing Agents. J Virol (2017) 91:1–14. doi: 10.1128/
JVI.02092-16 77. Dirk BS, Pawlak EN, Johnson AL, Van Nynatten LR, Jacob RA, Heit B, et al. HIV-1 Nef Sequesters MHC-I Intracellularly by Targeting Early Stages of
Endocytosis and Recycling. Sci Rep (2016) 6:37021. doi: 10.1038/srep37021 59. Sluis RMV d, Kumar NA, Pascoe RD, Zerbato JM, Evans VA,
Dantanarayana AI, et al. Combination Immune Checkpoint Blockade to
Reverse HIV Latency. J Immunol (2020) 204(5):1242–54. doi: 10.4049/
jimmunol.1901191 78. Rafiq S, Hackett CS, Brentjens RJ. Engineering Strategies to Overcome the
Current Roadblocks in CAR T Cell Therapy. Nat Rev Clin Oncol (2020)
17:147–67. doi: 10.1038/s41571-019-0297-y 60. Tong O, Duette G, O’Neil TR, Royle CM, Rana H, Johnson B, et al. REFERENCES Plasmacytoid Dendritic Cells Have Divergent Effects on HIV Infection of
Initial Target Cells and Induce a Pro-Retention Phenotype. PloS Pathog
(2021) 17:e1009522. doi: 10.1371/journal.ppat.1009522 79. Hamilton JR, Tsuchida CA, Nguyen DN, Shy BR, McGarrigle ER, Sandoval
Espinoza CR, et al. Targeted Delivery of CRISPR-Cas9 and Transgenes
Enables Complex Immune Cell Engineering. Cell Rep (2021) 35:109207. doi: 10.1016/j.celrep.2021.109207 61. Lavender KJ, Gibbert K, Peterson KE, Dis EV, Francois S, Woods T, et al. Interferon Alpha Subtype-Specific Suppression of HIV-1 Infection In Vivo. J Virol (2016) 90:6001–13. doi: 10.1128/JVI.00451-16 80. Bishop DC, Xu N, Tse B, O’Brien TA, Gottlieb DJ, Dolnikov A, et al. PiggyBac-Engineered T Cells Expressing CD19-Specific CARs That Lack
IgG1 Fc Spacers Have Potent Activity Against B-ALL Xenografts. Mol Ther
(2018) 26:1883–95. doi: 10.1016/j.ymthe.2018.05.007 62. Schröder ARW, Shinn P, Chen H, Berry C, Ecker JR, Bushman F. HIV-1
Integration in the Human Genome Favors Active Genes and Local Hotspots. Cell (2002) 110:521–9. doi: 10.1016/s0092-8674(02)00864-4 81. Beatty GL, O’Hara MH, Lacey SF, Torigian DA, Nazimuddin F, Chen F, et al. Activity of Mesothelin-Specific Chimeric Antigen Receptor T Cells Against
Pancreatic Carcinoma Metastases in a Phase 1 Trial. Gastroenterology (2018)
155:29–32. doi: 10.1053/j.gastro.2018.03.029 63. Jones RB, Mueller S, O’Connor R, Rimpel K, Sloan DD, Karel D, et al. A
Subset of Latency-Reversing Agents Expose HIV-Infected Resting CD4+ T-
Cells to Recognition by Cytotoxic T-Lymphocytes. PloS Pathog (2016) 12:
e1005545. doi: 10.1371/journal.ppat.1005545 82. Jin C, Fotaki G, Ramachandran M, Nilsson B, Essand M, Yu D. Safe
Engineering of CAR T Cells for Adoptive Cell Therapy of Cancer Using
Long-Term Episomal Gene Transfer. EMBO Mol Med (2016) 8:702–11. doi: 10.15252/emmm.201505869 64. Sadowski I, Hashemi FB. Strategies to Eradicate HIV From Infected Patients:
Elimination of Latent Provirus Reservoirs. Cell Mol Life Sci (2019) 76:3583–
600. doi: 10.1007/s00018-019-03156-8 83. Roth TL, Puig-Saus C, Yu R, Shifrut E, Carnevale J, Li PJ, et al. Reprogramming
Human T Cell Function and Specificity With non-Viral Genome Targeting. Nature (2018) 559:405–9. doi: 10.1038/s41586-018-0326-5 65. Huang S-H, Ren Y, Thomas AS, Chan D, Mueller S, Ward AR, et al. Latent
HIV Reservoirs Exhibit Inherent Resistance to Elimination by CD8+ T Cells. J Clin Invest (2018) 128:876–89. doi: 10.1172/JCI97555 84. Maude SL, Frey N, Shaw PA, Aplenc R, Barrett DM, Bunin NJ, et al. Chimeric Antigen Receptor T Cells for Sustained Remissions in Leukemia. N
Engl J Med (2014) 371:1507–17. doi: 10.1056/NEJMoa1407222 66. REFERENCES Dissociation of Gp120
From HIV-1 Virions Induced by Soluble CD4. Science (1990) 250:1139–42. doi: 10.1126/science.2251501 96. Long AH, Haso WM, Shern JF, Wanhainen KM, Murgai M, Ingaramo M,
et al. 4-1BB Costimulation Ameliorates T Cell Exhaustion Induced by Tonic
Signaling of Chimeric Antigen Receptors. Nat Med (2015) 21:581–90. doi: 10.1038/nm.3838 113. Kloss CC, Condomines M, Cartellieri M, Bachmann M, Sadelain M. Combinatorial Antigen Recognition With Balanced Signaling Promotes
Selective Tumor Eradication by Engineered T Cells. Nat Biotechnol (2013)
31:71–5. doi: 10.1038/nbt.2459 97. Zhen A, Peterson CW, Carrillo MA, Reddy SS, Youn CS, Lam BB, et al. Long-
Term Persistence and Function of Hematopoietic Stem Cell-Derived Chimeric
Antigen Receptor T Cells in a Nonhuman Primate Model of HIV/AIDS. PloS
Pathog (2017) 13:e1006753. doi: 10.1371/journal.ppat.1006753 114. Lanitis E, Poussin M, Klattenhoff AW, Song D, Sandaltzopoulos R, June CH,
et al. Chimeric Antigen Receptor T Cells With Dissociated Signaling
Domains Exhibit Focused Antitumor Activity With Reduced Potential for
Toxicity In Vivo. Cancer Immunol Res (2013) 1:43–53. doi: 10.1158/2326-
6066.CIR-13-0008 98. Seif M, Einsele H, Löffler J. CAR T Cells Beyond Cancer: Hope for
Immunomodulatory Therapy of Infectious Diseases. Front Immunol
(2019) 10:2711. doi: 10.3389/fimmu.2019.02711 115. Roybal KT, Williams JZ, Morsut L, Rupp LJ, Kolinko I, Choe JH, et al. Engineering T Cells With Customized Therapeutic Response Programs
Using Synthetic Notch Receptors. Cell (2016) 167:419–432.e16. doi: 10.1016/j.cell.2016.09.011 99. Hale M, Mesojednik T, Ibarra GSR, Sahni J, Bernard A, Sommer K, et al. Engineering HIV-Resistant, Anti-HIV Chimeric Antigen Receptor T Cells. Mol Ther (2017) 25:570–9. doi: 10.1016/j.ymthe.2016.12.023 116. Lajoie MJ, Boyken SE, Salter AI, Bruffey J, Rajan A, Langan RA, et al. Designed Protein Logic to Target Cells With Precise Combinations of
Surface Antigens. Science (2020) 369:1637–43. doi: 10.1126/science.aba6527 100. Maldini CR, Gayout K, Leibman RS, Dopkin DL, Mills JP, Shan X, et al. HIV-
Resistant and HIV-Specific CAR-Modified CD4+ T Cells Mitigate HIV
Disease Progression and Confer CD4+ T Cell Help In Vivo. Mol Ther
(2020) 28:1585–99. doi: 10.1016/j.ymthe.2020.05.012 117. Herzig E, Kim KC, Packard TA, Vardi N, Schwarzer R, Gramatica A, et al. Attacking Latent HIV With convertibleCAR-T Cells, a Highly Adaptable
Killing Platform. Cell (2019) 179:880–894.e10. doi: 10.1016/j.cell.2019.10.002 101. Maldini CR, Claiborne DT, Okawa K, Chen T, Dopkin DL, Shan X, et al. Dual CD4-Based CAR T Cells With Distinct Costimulatory Domains
Mitigate HIV Pathogenesis In Vivo. Nat Med (2020) 26:1776–87. doi: 10.1038/s41591-020-1039-5 118. REFERENCES Evaluation of Chimeric Antigen Receptor T Cell Therapy in non-
Human Primates Infected With SHIV or SIV. PloS One (2021) 16:e0248973. doi: 10.1371/journal.pone.0248973 72. Barouch DH, Whitney JB, Moldt B, Klein F, Oliveira TY, Liu J, et al. Therapeutic Efficacy of Potent Neutralizing HIV-1-Specific Monoclonal
Antibodies in SHIV-Infected Rhesus Monkeys. Nature (2013) 503:224–8. doi: 10.1038/nature12744 91. Romeo C, Seed B. Cellular Immunity to HIV Activated by CD4 Fused to T
Cell or Fc Receptor Polypeptides. Cell (1991) 64:1037–46. doi: 10.1016/0092-
8674(91)90327-u April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 14 Reducing the HIV Reservoir York et al. the CD4 Binding Site and High-Mannose Glycans of Gp120 Optimized for
Anti-Human Immunodeficiency Virus Potency and Breadth With Minimal
Immunogenicity. Cytotherapy (2018) 20:407–19. doi: 10.1016/
j.jcyt.2017.11.001 the CD4 Binding Site and High-Mannose Glycans of Gp120 Optimized for
Anti-Human Immunodeficiency Virus Potency and Breadth With Minimal
Immunogenicity. Cytotherapy (2018) 20:407–19. doi: 10.1016/
j.jcyt.2017.11.001 92. Roberts M, Qin L, Zhang D, Smith D, Tran A, Dull T, et al. Targeting of
Human Immunodeficiency Virus-Infected Cells by CD8+ T Lymphocytes
Armed With Universal T-Cell Receptors. Blood (1994) 84:2878–89. doi: 10.1182/blood.V84.9.2878.2878 93. Yang OO, Tran A-C, Kalams SA, Johnson RP, Roberts MR, Walker BD. Lysis
of HIV-1-Infected Cells and Inhibition of Viral Replication by Universal
Receptor T Cells. PNAS (1997) 94:11478–83. doi: 10.1073/pnas.94.21.11478 110. Liu L, Patel B, Ghanem MH, Bundoc V, Zheng Z, Morgan RA, et al. Novel
CD4-Based Bispecific Chimeric Antigen Receptor Designed for Enhanced
Anti-HIV Potency and Absence of HIV Entry Receptor Activity. J Virol
(2015) 89:6685–94. doi: 10.1128/JVI.00474-15 94. Leibman RS, Richardson MW, Ellebrecht CT, Maldini CR, Glover JA,
Secreto AJ, et al. Supraphysiologic Control Over HIV-1 Replication
Mediated by CD8 T Cells Expressing a Re-Engineered CD4-Based
Chimeric Antigen Receptor. PloS Pathog (2017) 13:e1006613. doi: 10.1371/journal.ppat.1006613 111. Anthony-Gonda K, Bardhi A, Ray A, Flerin N, Li M, Chen W, et al. Multispecific Anti-HIV duoCAR-T Cells Display Broad In Vitro Antiviral
Activity and Potent In Vivo Elimination of HIV-Infected Cells in a
Humanized Mouse Model. Sci Transl Med (2019) 11:1–40. doi: 10.1126/
scitranslmed.aav5685 95. Kaartinen T, Luostarinen A, Maliniemi P, Keto J, Arvas M, Belt H, et al. Low
Interleukin-2 Concentration Favors Generation of Early Memory T Cells
Over Effector Phenotypes During Chimeric Antigen Receptor T-Cell
Expansion. Cytotherapy (2017) 19:689–702. doi: 10.1016/j.jcyt.2017.03.067 112. Moore JP, McKeating JA, Weiss RA, Sattentau QJ. REFERENCES Matusali G, Tchidjou HK, Pontrelli G, Bernardi S, D’Ettorre G, Vullo V, et al. Soluble Ligands for the NKG2D Receptor are Released During HIV-1
Infection and Impair NKG2D Expression and Cytotoxicity of NK Cells. FASEB J (2013) 27:2440–50. doi: 10.1096/fj.12-223057 102. Liu B, Zou F, Lu L, Chen C, He D, Zhang X, et al. Chimeric Antigen Receptor
T Cells Guided by the Single-Chain Fv of a Broadly Neutralizing Antibody
Specifically and Effectively Eradicate Virus Reactivated From Latency in CD4
+ T Lymphocytes Isolated From HIV-1-Infected Individuals Receiving
Suppressive Combined Antiretroviral Therapy. J Virol (2016) 90:9712–24. doi: 10.1128/JVI.00852-16 119. Haran KP, Hajduczki A, Pampusch MS, Mwakalundwa G, Vargas-
Inchaustegui DA, Rakasz EG, et al. Simian Immunodeficiency Virus
(SIV)-Specific Chimeric Antigen Receptor-T Cells Engineered to Target B
Cell Follicles and Suppress SIV Replication. Front Immunol (2018) 9:492. doi: 10.3389/fimmu.2018.00492 120. Leong YA, Chen Y, Ong HS, Wu D, Man K, Deleage C, et al. CXCR5(+)
Follicular Cytotoxic T Cells Control Viral Infection in B Cell Follicles. Nat
Immunol (2016) 17:1187–96. doi: 10.1038/ni.3543 103. Bar KJ, Sneller MC, Harrison LJ, Justement JS, Overton ET, Petrone ME,
et al. Effect of HIV Antibody VRC01 on Viral Rebound After Treatment
Interruption. N Engl J Med (2016) 375:2037–50. doi: 10.1056/
NEJMoa1608243 121. Ayala VI, Deleage C, Trivett MT, Jain S, Coren LV, Breed MW, et al. CXCR5-
Dependent Entry of CD8 T Cells Into Rhesus Macaque B-Cell Follicles
Achieved Through T-Cell Engineering. J Virol (2017) 91:1–17. doi: 10.1128/
JVI.02507-16 104. Mendoza P, Gruell H, Nogueira L, Pai JA, Butler AL, Millard K, et al. Combination Therapy With Anti-HIV-1 Antibodies Maintains Viral
Suppression. Nature (2018) 561:479–84. doi: 10.1038/s41586-018-0531-2 105. Guedan S, Calderon H, Posey AD, Maus MV. Engineering and Design of
Chimeric Antigen Receptors. Mol Ther Methods Clin Dev (2019) 12:145–56. doi: 10.1016/j.omtm.2018.12.009 122. Pampusch MS, Hajduczki A, Mwakalundwa G, Connick E, Berger EA,
Skinner PJ. Production and Characterization of SIV-Specific CAR/CXCR5
T Cells. In: Rast J, Buckley K, editors. Immune Receptors: Methods and
Protocols. Methods in Molecular Biology. New York, NY: Springer US (2022). p. 171–85. doi: 10.1007/978-1-0716-1944-5_12 106. Ruella M, Barrett DM, Kenderian SS, Shestova O, Hofmann TJ, Perazzelli J,
et al. Dual CD19 and CD123 Targeting Prevents Antigen-Loss Relapses After
CD19-Directed Immunotherapies. J Clin Invest (2016) 126:3814–26. doi: 10.1172/JCI87366 123. Barber-Axthelm IM, Barber-Axthelm V, Sze KY, Zhen A, Suryawanshi GW,
Chen ISY, et al. REFERENCES J Exp Med (2004) 200:749–
59. doi: 10.1084/jem.20040874 129. Rupp LJ, Schumann K, Roybal KT, Gate RE, Ye CJ, Lim WA, et al. CRISPR/
Cas9-Mediated PD-1 Disruption Enhances Anti-Tumor Efficacy of Human
Chimeric Antigen Receptor T Cells. Sci Rep (2017) 7:737. doi: 10.1038/
s41598-017-00462-8 138. Ringpis G-EE, Shimizu S, Arokium H, Camba-Colón J, Carroll MV, Cortado
R, et al. Engineering HIV-1-Resistant T-Cells From Short-Hairpin RNA-
Expressing Hematopoietic Stem/Progenitor Cells in Humanized BLT Mice. PloS One (2012) 7:e53492. doi: 10.1371/journal.pone.0053492 PloS One (2012) 7:e53492. doi: 10.1371/journal.pone.0053492 130. Habib R, Nagrial A, Micklethwaite K, Gowrishankar K. Chimeric Antigen
Receptors for the Tumour Microenvironment. In: A Birbrair, editor. Tumor
Microenvironment: State of the Science. Advances in Experimental Medicine
and Biology. Cham: Springer International Publishing (2020). p. 117–43. doi: 10.1007/978-3-030-44518-8_8 Conflict of Interest: The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could be construed as a
potential conflict of interest. 131. Shimabukuro-Vornhagen A, Gödel P, Subklewe M, Stemmler HJ, Schlößer
HA, Schlaak M, et al. Cytokine Release Syndrome. J ImmunoTher Cancer
(2018) 6:56. doi: 10.1186/s40425-018-0343-9 Publisher’s Note: All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their affiliated organizations, or those of
the publisher, the editors and the reviewers. Any product that may be evaluated in
this article, or claim that may be made by its manufacturer, is not guaranteed or
endorsed by the publisher. 132. Massanella M, Richman DD. Measuring the Latent Reservoir In Vivo. J Clin
Invest (2016) 126:464–72. doi: 10.1172/JCI80567 133. Zhang P, Raju J, Ullah MA, Au R, Varelias A, Gartlan KH, et al. Phase I Trial
of Inducible Caspase 9 T Cells in Adult Stem Cell Transplant Demonstrates
Massive Clonotypic Proliferative Potential and Long-Term Persistence of
Transgenic T Cells. Clin Cancer Res (2019) 25:1749–55. doi: 10.1158/1078-
0432.CCR-18-3069 Copyright © 2022 York, Gowrishankar, Micklethwaite, Palmer, Cunningham and
Nasr. This is an open-access article distributed under the terms of the Creative
Commons Attribution License (CC BY). The use, distribution or reproduction in other
forums is permitted, provided the original author(s) and the copyright owner(s) are
credited and that the original publication in this journal is cited, in accordance with
accepted academic practice. No use, distribution or reproduction is permitted which
does not comply with these terms. 134. Paszkiewicz PJ, Fräßle SP, Srivastava S, Sommermeyer D, Hudecek M,
Drexler I, et al. REFERENCES Stem Cell–Derived CAR T Cells Traffic to HIV Reservoirs in
Macaques. JCI Insight (2021) 6:e141502. doi: 10.1172/jci.insight.141502 107. Fry TJ, Shah NN, Orentas RJ, Stetler-Stevenson M, Yuan CM, Ramakrishna
S, et al. CD22-Targeted CAR T Cells Induce Remission in B-ALL That is
Naive or Resistant to CD19-Targeted CAR Immunotherapy. Nat Med (2018)
24:20–8. doi: 10.1038/nm.4441 124. Caruana I, Savoldo B, Hoyos V, Weber G, Liu H, Kim ES, et al. Heparanase
Promotes Tumor Infiltration and Antitumor Activity of CAR-Redirected T
Lymphocytes. Nat Med (2015) 21:524–9. doi: 10.1038/nm.3833 108. Hajduczki A, Danielson DT, Elias DS, Bundoc V, Scanlan AW, Berger EA. A
Trispecific Anti-HIV Chimeric Antigen Receptor Containing the CCR5 N-
Terminal Region. Front Cell Infect Microbiol (2020) 10:242. doi: 10.3389/
fcimb.2020.00242 125. Spranger S, Spaapen RM, Zha Y, Williams J, Meng Y, Ha TT, et al. Up-
Regulation of PD-L1, IDO, and Tregs in the Melanoma Tumor
Microenvironment Is Driven by CD8+ T Cells. Sci Trans Med (2013)
5:200ra116–200ra116. doi: 10.1126/scitranslmed.3006504 109. Ghanem MH, Bolivar-Wagers S, Dey B, Hajduczki A, Vargas-Inchaustegui
DA, Danielson DT, et al. Bispecific Chimeric Antigen Receptors Targeting 126. Yoon DH, Osborn MJ, Tolar J, Kim CJ. Incorporation of Immune
Checkpoint Blockade Into Chimeric Antigen Receptor T Cells (CAR-Ts): April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 15 York et al. Reducing the HIV Reservoir Combination or Built-In CAR-T. Int J Mol Sci (2018) 19:E340. doi: 10.3390/
ijms19020340 135. Sockolosky JT, Trotta E, Parisi G, Picton L, Su LL, Le AC, et al. Selective
Targeting of Engineered T Cells Using Orthogonal IL-2 Cytokine-Receptor
Complexes. Science (2018) 359:1037–42. doi: 10.1126/science.aar3246 127. Cherkassky L, Morello A, Villena-Vargas J, Feng Y, Dimitrov DS, Jones DR,
et al. Human CAR T Cells With Cell-Intrinsic PD-1 Checkpoint Blockade
Resist Tumor-Mediated Inhibition. J Clin Invest (2016) 126:3130–44. doi: 10.1172/JCI83092 136. Moriggl R, Sexl V, Piekorz R, Topham D, Ihle JN. Stat5 Activation is
Uniquely Associated With Cytokine Signaling in Peripheral T Cells. Immunity (1999) 11:225–30. doi: 10.1016/s1074-7613(00)80097-7 128. Jiang Z, Liang H, Pan H, Liang Y, Wang H, Yang X, et al. HIV-1-Specific
CAR-T Cells With Cell-Intrinsic PD-1 Checkpoint Blockade Enhance Anti-
HIV Efficacy In Vivo. Front Microbiol (2021) 12:684016. doi: 10.3389/
fmicb.2021.684016 137. Brenchley JM, Schacker TW, Ruff LE, Price DA, Taylor JH, Beilman GJ, et al. CD4+ T Cell Depletion During All Stages of HIV Disease Occurs
Predominantly in the Gastrointestinal Tract. April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org REFERENCES Targeted Antibody-Mediated Depletion of Murine CD19
CAR T Cells Permanently Reverses B Cell Aplasia. J Clin Invest (2016)
126:4262–72. doi: 10.1172/JCI84813 April 2022 | Volume 13 | Article 873701 Frontiers in Immunology | www.frontiersin.org 16
|
https://openalex.org/W2808453157
|
https://link.springer.com/content/pdf/10.1007/s40860-018-0064-3.pdf
|
English
| null |
Is the internet spatial?
|
Journal of reliable intelligent environments
| 2,018
|
cc-by
| 5,416
|
Abstract One of the internet’s design principles is to make physical distance transparent, which has made it a motor of globalization
and economic development. However, the physical world within which it operates is spatial, and reduction of actual spatial
distance is a factor of high impact leveraged, e.g., in today’s commercially successful content distribution networks as well
as new technologies, such as edge computing. Predictable network behavior is key for reliable intelligent environments. It is
therefore worthwhile to study whether and how spatial the internet actually is. We studied a small set of globally dispersed
public universities on six continents, recording delays reported by the ping tool to a server in the Pacific Northwest area
of the U.S. at four different times of the day (morning, noon, evening, midnight) and over both weekends and weekdays,
thus retrieving more than 17,210 data points. We then determined the Pearson’s correlations between distance and delay
and between hop count and delay. The experiment yielded that there is an overall correlation of 0.57 (a border case strong
correlation) between distance and delay and a weaker correlation of 0.4 (medium strength correlation) between hop count
and delay. The results suggest that there is a physical distance effect and that this effect is stronger than the hop count effect. Keywords Spatial networks · Internet · Round trip time Keywords Spatial networks · Internet · Round trip time Is the internet spatial? Hedda R. Schmidtke1 Received: 29 December 2017 / Accepted: 31 May 2018 / Published online: 15 June 2018
© The Author(s) 2018 Journal of Reliable Intelligent Environments (2018) 4:123–129
https://doi.org/10.1007/s40860-018-0064-3 Journal of Reliable Intelligent Environments (2018) 4:123–129
https://doi.org/10.1007/s40860-018-0064-3 Journal of Reliable Intelligent Environments (2018) 4:123–129
https://doi.org/10.1007/s40860-018-0064-3 ORIGINAL ARTICLE B Hedda R. Schmidtke
schmidtke@acm.org 1
University of Oregon, Eugene, OR 97403-1251, USA 1 Introduction If a packet is lost, it
simply disappears, and the recipient may hear a noise or brief
distortion in a VoIP connection. If packets arrive in a different
ordering than the one they were sent in, the recipient’s UDP
layer cannot detect this, but they can be discarded without
much loss. TCP, on the other hand, establishes a connection
before starting to send data as a sequence of packets, each
identified with a sequence number, which allows the recip-
ient’s TCP layer to reassemble the original data stream. In
order to do this, it uses an acknowledgement and retransmis-
sion mechanism to make sure all packets have been received. We need to look at this process in slightly more detail to
understand the impact of RTT. A detailed introduction and
comparison of techniques is beyond the scope of this paper
and can be found in Dordal [6]; an evaluation of the different
newer proposals with respect to satellite links, in Shin et al,
[16]. The purpose of this paper was to investigate the question
of whether there is any spatial factor to internet delays, mea-
sured with RTT, besides hop count. We found this question
of interest as everyday experience and the success of CDNs
may suggest that there is a spatial factor but the current inter-
net specification and network theory suggest that this should
be an at most spurious effect. Our study set out to investigate
this by measuring RTT from a terminal in the US to several
public universities in the world and checking for the cor-
relation between RTT and spatial distance. While we look
at the literature with respect to potential explanations and
assessment of the generality of the problem with respect to
today’s and tomorrow’s networks, the focus of this article is
on reporting the experimental findings. Structure of the article We start with a short introduction
of TCP (Transmission Control Protocol) and a discussion of
literature mentioning spatial properties in TCP (Sect. 2). We
then describe the experiment and its results and discuss the
outcome (Sect. 3). We outline conclusions and potential for
future work in Sect. 4. In establishing a TCP connection, the client and server,
inter alia, exchange information about their respective capa-
bilities. 1 Introduction interconnect with each other, creating a virtual network
layered on top of an existing packet network infrastruc-
ture. [p. 20] When someone chooses a mirror from which to download a
software they usually pick the ones nearest to them, a strat-
egy derived from human experience in a physical world. In
the physical world, longer distances mean more effort and
we drive to the shop nearby rather than the one further away
if both offer the same interface and products. From a network
perspective, we may suggest to pick one such that traffic is
with a high probability smaller, however, far away. In com-
puter networks distance seems to disappear shrinking the
world into a “global village.” In today’s internet, however,
mirrors are increasingly important as the growing sector for
content delivery networks (CDNs) suggests. Content own-
ers, such as streaming services, send their content to CDN
operators to host and distribute it to a worldwide network
of servers close to the locations of customers and clients. Hofmann and Beaumont [9] define: Search engines, map providers, streaming services, and
social networks, for instance, benefit from employing content
networks. One of the advantages of CDNs, is that they provide dis-
tribution in proximity to the end-users. CDNs thus reduce
global traffic, but also, one might think, that they provide an
advantage by delivering information over shorter distances. This is worthwhile noticing because we think of computer
networks rather as non-spatial graphs, potentially depend-
ing on hop count, but not on spatial distance. Moreover, hop
count may be greatly divergent within areas, i.e., is gener-
ally not spatial in the sense of physical geographic space, but
only in an abstract topological sense. Internet protocols, for
instance, do not make use of a notion of physical distance
and leverage a temporal rather than spatial measure: delays
or roundtrip time (RTT). Electric signals travel at the speed
of light, i.e., 300 km/ms. Given the Earth’s circumference of
about 40,000 km, a signal traveling at this speed could circle
the whole planet within 134 ms. Distance thus should not
play any significant role. The term content network refers to a communication
network that deploys infrastructure components [that] B Hedda R. Schmidtke
schmidtke@acm.org 1
University of Oregon, Eugene, OR 97403-1251, USA 3 3 124 Journal of Reliable Intelligent Environments (2018) 4:123–129 messages and provides checksums for error detection and
port numbers for identifying services. 2 Related works To our knowledge, there are no studies on the spatial proper-
ties of the internet per se. However, roundtrip time (RTT) has
been studied in the context of transport protocol issues and
spatial properties are mentioned in some of these studies. We briefly introduce how TCP operates in Sect. 2.1. Then
we look at the packet reordering problem, a problem that
has been reported to particularly affect long-distance con-
nections, and solution proposals (Sect. 2.2). Finally, we move
beyond the TCP/IP suite to proposals for more atypical dis-
tances (Sect. 2.3). 1 Introduction A main concern at this stage is to avoid congestion
on the client’s side by establishing the maximum window
rwnd of packets the receiver can process, as congestion leads
to packets being dropped by the underlying IP layer. After
the connection has been started, data packets are sent from
the server to the client, which responds by acknowledg-
ing receipt. The receiver sends an acknowledgement (ACK)
with a packet number when a new packet arrives and the
receiver has received each packet up to that number. If the
packet numbered n is lost or delayed and the packet with
number n + 1 arrives, the receiver thus, sends an acknowl-
edgement for sequence number n −1 again. If packet n now
arrives, n + 1 can be acknowledged (cumulative acknowl-
edgements). If the same packet, e.g., n −1, is acknowledged
three times, the sender assumes the packet is lost. The orig-
inal TCP specification [14] of 1981 lets the sender restart
sending all packets from n again, even if other packets have
arrived. In contrast, the 2009 TCP [2] then resends n, and
continues from where it was (fast retransmit). The sender
also keeps a timer for each packet. If the sender does not
receive an acknowledgement before a timeout is reached,
the not acknowledged packet is assumed lost and retransmit-
ted. 123 2.1 TCP/IP over long distances The transport layer of the TCP/IP model is built upon the
internet layer (IP). While IP connects individual hosts and
routers physically linked to each other and is responsible
for routing packets, TCP is responsible for providing reli-
able end-to-end connections upon which internet application
services such as HTTP but also, for instance, transaction-
based services are built, such as a travel booking tool, an
educational tool, a business process management tool, or a
banking application. A connectionless alternative to TCP is
UDP (User Datagram Protocol), a protocol without relia-
bility guarantees often found in real-time applications, such
as VoIP or video streaming applications, where missing a
packet is less critical than the delay that would be caused by
retransmission. During the transmission phase, the sender is responsible
for avoiding congestion on the network. It, therefore, tries to
adjust its sending rate to the conditions of the network. The
only information available to TCP is roundtrip time (RTT). Packet loss is inferred using this mechanism. The timeout
window is a conservative estimate of RTT. The basic assump-
tion is that delays and packet loss are caused by congestion in
the network and can, therefore, be addressed by reducing the
packet rate. Several updates to TCP [2] focus on congestion
recognition and avoidance. The specific difficulties regarding RTT in TCP can best be
illustrated by comparison with UDP. UDP sends individual 123 Journal of Reliable Intelligent Environments (2018) 4:123–129 125 Fig. 1 Current and planned submarine cables [18] The congestion window (cwnd) controls the packet rate. The “slow start” algorithm was introduced as an algorithm
for beginning “transmission into a network with unknown
conditions.” It starts from a very small cwnd and succes-
sively increases the rate by one with each ACK. This doubles
the cwnd after each RTT. This doubling of packet rate stops
when the initially set maximum window size of the receiver
is reached. When the slow start threshold ssthresh is reached
the increase is slowed to a linear increase of packets per
RTT. When a packet is lost, congestion avoidance is initiated. Different TCP implementations use different congestion
avoidance algorithms. In the simplest version (TCP Tahoe)
the lost packet is sent again (fast retransmit), the cwnd is set
to its very small initial size and ssthresh is halved. 2.1 TCP/IP over long distances Another
version (TCP Reno and NewReno) differs in that it sets the
cwnd not to the minimal size but to the halved ssthresh, essen-
tially stopping the slow start phase and moving directly to
the linear growth phase. Other proposals (e.g., Vegas, FAST,
Westwood, Veno, and Illinois) improve on Reno’s greedy
loss-based strategy using a delay-based strategy and are able
to speed up a high-speed connection faster, but also assume
that RTT depends solely on congestion. TCP Veno targets
noisy wireless connections. High-speed connections are bet-
ter supported, in particular, by Highspeed TCP and TCP
CUBIC, which aim for a more aggressive cwnd increase. Fig. 1 Current and planned submarine cables [18] as long-distance networks. The problem in both cases is the
number of packets in transit in relation to reordering capabil-
ity. With a high-speed connection this number is high even
over short distances; likewise, with lower speeds and longer
distances. The number of packets in transit for a connection
can be summarized as: τ = ρ ∗RTT, where ρ is the packet
send rate in packets per second. In both cases, τ is high. In
the case of high speed networks, because ρ is large. In case
of long distance networks, because RTT is large. An important aspect is the availability of alternative paths. Today’s internet makes considerable use of path diversity to
increase throughput, and transmission speeds are steadily
increasing letting the problem surface over increasingly
shorter distances [3]. From a spatiotemporal point of view,
the measure τ reflects that high-speed connections generate
a similar effect as spatial distance. With a short RTT, how-
ever, control messages can be exchanged relatively quickly
between sender and receiver to restart a connection. A long-
distance connection does not have this option. Additionally,
restarting a connection after a timeout, happens faster with a
smaller RTT. That distance could play a significant role in wired net-
works seems counterintuitive, given that electric signals
travel at the speed of light. However, “TCP extensions for
long-delay paths” have been a topic since the 1980s [11]
when addressing satellite links. TCP Hybla [5] is a TCP implementation that targets
specifically satellite connections. Interestingly, it achieves
its improvements by replacing RTT measurements for esti-
mates of minimum RTT with a reference value chosen to be
suitable for satellite links, i.e., a fixed value depending on
the satellites’ orbit, that is, a distance. 2.2 Packet reordering A generally high variability of RTT has been experimentally
confirmed by Aikat et al [1]. This has a large impact on
TCP as it bases its strategy only on measurements of RTT. High variability leads to reordering of packets. As would
be expected, the variability for small RTT is smaller, while
intercontinental travel employs submarine cables and is thus
more prone to meet bottlenecks (Fig. 1). Parallel paths are the main reason for reordering [4], but
also a way to improve packet rate. Eliminating parallel paths
is a theoretical solution [8], but comes at the expense of
decreasing maximum utilization of resources and thus effi-
ciency [4], which is not desirable especially in the case of
long-distance connections. The fact that reordering is not pathological behavior but
an issue of TCP in particular over long-haul connections is
well known [4] and was studied experimentally, in particular
with respect to high-speed networks [7] with renewed inter-
est more recently. High speed networks face the same issues 2.1 TCP/IP over long distances The geostationary net-
working satellite orbits are in 35,786 km above the equator,
and satellite links suffer from about 600 ms delays [16]. The consequences of packet reordering for TCP are that
it “makes inefficient use of available link bandwidth,” and it
“loses self-clocking and becomes highly bursty” [4]. Increas-
ing reordering capability by selective acknowledgements
(TCP SACK) has been shown to be an effective measure
[7] and is supported by some of the newer protocols. How-
ever, relaxing the fixed three packet boundary comes at the
price of increased complexity. 3 We used the prefix search at: https://en.wikipedia.org/.
4 https://www.iplocation.net 3 Experiment The goal of our experiment was to determine the correlation
between RTT and distance on its own and as a secondary
concern compare it to the correlation between hop count and
RTT. We, therefore, did a detailed study of a widely dispersed
network of servers at public universities. We chose public
universities, because private companies including universi-
ties often already employ CDNs, which would disable our
ability to study distance. Our rationale in selecting specific
universities was to have at least two from each continent. However, in both Africa and South America we could only
find one university each that fulfilled the necessary criteria
and participated over the whole duration. We tried to equally
select servers in both developed and developing regions on
each continent.1 However, we could not find a server among
the low income countries in Africa that would have fulfilled
our criteria. All in all, 16 experiments were performed. Eight experi-
ments were executed on a weekend day (at the origin), the
remaining eight during weekdays. Experiments also were
varied by the times of the day (at the origin): four were exe-
cuted at midnight, four at 6am, four at noon, and four at 6pm. In each experiment, two separate bursts of 50 ping probes
were sent to each server. The returned RTT values were
recorded and stored. We thus all in all sent 16 × 11 × 100 =
17,600 pings, and received 17,210 responses. There were
only two anomalies. The server in Rio de Janeiro temporar-
ily dropped out of the study and was unreachable. The second
anomaly was that the university in Cape Town, maybe inten-
tionally, as it serves a continent where the internet is still
under development, replied to nearly all pings with duplicate
responses. As it was the only university in Africa that passed
our criteria, removing it would have removed Africa from
the study. We thus only removed the duplicates. 2.3 Approaches for atypical distances Several approaches for providing reliable connection based
on TCP have been cited in the above discussion. The current 12 3 Journal of Reliable Intelligent Environments (2018) 4:123–129 126 geodistances. Through a web search of lists of universities
represented on Wikipedia3, we selected 50 public univer-
sities worldwide. We checked all servers for the following
minimum criteria: method to address the problem of long-haul connections is
to establish mirrors close to where the clients are. A protocol
that specifically supports this approach is the Stream Control
Transmission Protocol [SCTP, 17]. SCTP provides multi-
homing support and establishes parallel channels. It uses path
selection and monitoring to improve on the reordering issue. It also leverages using larger packet sizes (Ethernet jumbo
frames). – The servers had to be enabled to respond to queries. We
tested this using ping. – They should not be using a CDN so as to not distort local-
ization. We determined this together with the hop count
by using traceroute and checking the final responding
server’s geolocation with the IPLocation service.4 Using spatial properties for network operation has been
proposed in other domains of networking. Georouting appro-
aches [13] are employed for ad hoc networks, such as
geosensor networks. – They should have a less than 1% overall packet drop rate
in an initial test. A high packet drop rate could indicate a
distinct problem or misconfiguration on the path or at the
server. After initial ping probes, we removed all servers
that did not meet this criterion. The problems of long-haul transmissions have been tra-
ditionally studied in the area of delay-tolerant (DTN) or
intermittently connected networks (ICN) [12], including
interplanetary communications, MANETs, or communica-
tion to underdeveloped regions. ICNs are currently operating
on distinct protocols employing so-called bundles [15]. All servers passing the initial vetting process were kept in
the pool over the whole duration. No servers were removed
after this. Overall, only 11 of the selected university servers
fulfilled all requirements. Table 1 shows the locations of the
servers together with their distances to our server. 1 For this classification, we relied mainly on the IMF classification
[10], into advanced economies (A) and emerging market and develop-
ing economies (E): South Korea (A), Vietnam (E), New Zealand (A),
Australia (A), South Africa (E, here A), Germany (A), Romania (E),
USA (A), Mexico (E), Brazil (E). Because of its special status as the
internet hub for a large part of Africa, we classified South Africa within
the advanced category for our purposes. 2 https://nathanrooy.github.io/posts/2016-09-07/haversine-with-
python/ 3.3 Discussion The data support the intuition that physical distance is a pre-
dictor for network behavior. The two servers in Mexico and
Florida, for instance, have similar values in Table 1, although
one is in a developing country and one is in a developed
country. Likewise, Hamburg and Daejeon/Seoul have simi-
lar distance and similar values in Table 1. Table 2 shows the results of the analysis. Over all coun-
tries, we can see that the correlation between distance and
delay (column: CD) > 0.5 for every entry, except during
the evening on weekdays at the destination (0.39) and dur-
ing weekend nights at the origin (0.43). When we look at
developed countries, it also drops below 0.5 over weekday
evenings at the destination (0.35) and over nights during the
weekend at the origin (0.39). With the exception of these
four cases all other eight analyzed cases have a value larger
than 0.5, the maximum being during weekday afternoons at
the destinations in developed countries (0.74), with the same
time slot at destinations over all countries being the second
highest value (0.72). A value of 0.7 is reached also in devel-
oping countries when looking at time at the origin in the
weekend morning and weekday night slots. In the develop-
ing countries, CD varies between 0.59 and 0.7 for time at the
origin and between 0.57 and 0.69 for time at the destination,
that is, the correlation CD is strong at all times for developing
countries. Due to the larger number of experiments, however, the
more detailed analysis of Table 2 allows us to further study
how the different parameters, location, time, and hop count,
influence RTT at a given time. When looking at the correla-
tions for the smaller time slices in Table 2, the influence of
how busy the network is at a given time influences the cor-
relation considerably with correlation values increasing as
measured RTT would be expected to be more homogenous
within a given time slice. The observation that the difference between the influence
of hop count between developing and developed regions is
large suggests a crucial difference between the different net-
works. A detailed interpretation, however, is not possible
from the conducted experiments. However, this difference
is worth noting as such differences may considerably impact
the ability of developing countries to participate in the global
economy. 3.2 Results We designed an experiment using the widely available tools
ping and traceroute to retrieve RTT and hop count and
used a Haversine distance implementation2 to approximate We determined the Pearson correlation coefficient between
distance and delay, and also between hop count and delay. The result over the whole point set was a correlation of
0.57 > 0.5, that is, a boundary case strong correlation,
for distance. We obtained 0.4 < 0.5, a weaker correlation
of medium strength, for hop count. To understand better
whether time influences the result we also analyzed the dif-
ferent categories of time slices and locations. Since an earlier 3 Journal of Reliable Intelligent Environments (2018) 4:123–129 127 Table 1 Experimental Setup:
cities of servers queried with
number of pings returned not
counting duplicates (n), GMT
timezone (TZ), distance from
the client in Oregon (timezone
GMT -8) in km, hop count
(HC), minimum delay (Min),
average delay (Avg), maximum
delay (Max), and standard
deviation of delays (StD) over
all n pings sent to that location City
n
TZ
Dist
HC
Min
Avg
Max
StD
Daejeon/Seoul
1591
9.0
8531
22
176
198
586
61
Hanoi
1594
7.0
11234
21
284
400
915
106
Christchurch
1598
13.0
11636
13
189
198
514
31
Adelaide
1588
10.5
13221
23
192
255
4937
199
Cape Town
1599
2.0
16539
32
348
465
937
107
Hamburg
1600
1.0
8280
12
199
221
559
59
Bucharest
1588
2.0
9715
14
208
314
766
104
Gainesville FL
1600
−5.0
3913
19
106
112
410
22
Mexico City
1600
−6.0
3527
19
94
100
404
19
Rio de Janeiro
1292
−2.0
10994
27
220
337
798
106
Manaus
1560
−4.0
8150
24
220
340
812
110 (0.36 ≤Dif ≤0.59) and small for developed countries
(0.07 ≤Dif ≤0.14). (0.36 ≤Dif ≤0.59) and small for developed countries
(0.07 ≤Dif ≤0.14). version of this research was rejected with the comment that
time had not been accounted for, we studied the detailed
behavior for different times, taking into account both time at
the fixed origin in Oregon and time at the various destina-
tions. To do this, we partitioned the day into four slices: night
= [0 : 00, 6 : 00), morning = [6 : 00, 12 : 00), afternoon
= [12 : 00, 18 : 00), evening = [18 : 00, 24 : 00) to classify
the responses by time at the destination. Table 1 Experimental Setup:
cities of servers queried with
number of pings returned not
counting duplicates (n), GMT
timezone (TZ), distance from
the client in Oregon (timezone
GMT -8) in km, hop count
(HC), minimum delay (Min),
average delay (Avg), maximum
delay (Max), and standard
deviation of delays (StD) over
all n pings sent to that location 3.3 Discussion The correlation between hop count and delay (CH) is in
every case a smaller correlation as visualized by the last col-
umn in each country group: Dif = CD −CH. It shows
by how much the correlation between distance and delay
exceeds the correlation between hop count and delay. It is
positive for each location type and time, relative to origin
as well as relative to destination. The difference between
CD and CH is particularly large for developing countries One may object that the comparison between distance
and hop count would be unfair as, depending on the cho-
sen route, hop count could considerably vary. What this only
emphasizes is, however, that hop count is an unreliable pre-
dictor, while distance is not. 3.3 Discussion would not only be considerably more costly to create, but
may even be not as useful as a simple distance calculation. While we did not make recommendations or try to give
a deeper explanation beyond what has been stated in the lit-
erature, we can conclude that, given that distance can be
calculated easily, a more flexible, spatiotemporally distance-
adaptive protocol could be developed with relatively small
overhead. Such a spatiotemporally adaptive protocol would
not only serve currently underserved regions and satellite
connections at the long range, but equally also high-speed
connections at the short range. 3.3 Discussion A statistical database of average
hop count distance between any two network locations, thus, 12 123 3 3 Journal of Reliable Intelligent Environments (2018) 4:123–129 128 Table 2 Experimental Results: number of pings (n), correlation
between distance and delay (CD), correlation between hop count and
delay (CH), and difference between the two correlations (Dif = CD −
CH), distinguished by type of time of day, type day of of week, and for
both time classified by the origin or by the destination of pings
Loc
Day
Time
All countries
Developing countries
Developed countries
n
CD
CH
Dif
n
CD
CH
Dif
n
CD
CH
Dif
Origin
Weekend
Night
2200
0.43
0.30
0.13
1000
0.61
0.16
0.45
1200
0.39
0.32
0.07
Morning
2180
0.59
0.44
0.15
980
0.70
0.18
0.52
1200
0.67
0.59
0.08
Afternoon
2190
0.60
0.43
0.17
991
0.59
0.20
0.39
1199
0.69
0.55
0.14
Evening
2083
0.64
0.39
0.25
890
0.61
0.09
0.52
1193
0.67
0.52
0.15
Weekday
Night
2086
0.63
0.46
0.17
893
0.70
0.19
0.51
1193
0.68
0.60
0.08
Morning
2089
0.60
0.46
0.14
892
0.68
0.20
0.48
1197
0.68
0.61
0.07
Afternoon
2196
0.57
0.41
0.16
996
0.58
0.22
0.36
1200
0.65
0.51
0.14
Evening
2186
0.63
0.45
0.18
992
0.67
0.19
0.48
1194
0.68
0.56
0.12
Destination
Weekend
Night
2096
0.58
0.38
0.20
896
0.57
0.10
0.47
1200
0.65
0.53
0.12
Morning
2185
0.59
0.44
0.15
986
0.60
0.20
0.40
1199
0.68
0.56
0.12
Afternoon
2190
0.62
0.44
0.18
991
0.69
0.17
0.52
1199
0.68
0.59
0.09
Evening
2191
0.60
0.40
0.20
991
0.61
0.19
0.42
1200
0.67
0.53
0.14
Weekday
Night
2189
0.59
0.45
0.14
993
0.63
0.20
0.43
1196
0.67
0.57
0.10
Morning
2091
0.58
0.45
0.13
893
0.69
0.20
0.49
1198
0.66
0.59
0.07
Afternoon
2078
0.72
0.47
0.25
894
0.69
0.10
0.59
1184
0.74
0.62
0.12
Evening
2190
0.39
0.29
0.10
990
0.60
0.24
0.36
1200
0.35
0.28
0.07 CH), distinguished by type of time of day, type day of of week, and for
both time classified by the origin or by the destination of pings Table 2 Experimental Results: number of pings (n), correlation
between distance and delay (CD), correlation between hop count and
delay (CH), and difference between the two correlations (Dif = CD − Table 2 Experimental Results: number of pings (n), correlation
between distance and delay (CD), correlation between hop count and
delay (CH), and difference between the two correlations (Dif = CD − would not only be considerably more costly to create, but
may even be not as useful as a simple distance calculation. 4 Conclusions We reported a global experiment in which we sent 17,600
pings to 11 servers on six continents and collected 17,210
responses. Our study is unique in its reproducibility as it
only uses publicly available internet tools. Large-scale stud-
ies reporting on variability in RTT as depending on distance,
such as [4], were performed at internet service provider (ISP)
facilities, and are thus not easily reproducible. The study is
also unique in that it probes the internet with a geographic
rather than technical perspective as a complex ecosystem. Future work should also study the surprising difference
between the influence of hop count for servers in developing
and developed countries. The internet plays an increasingly
important role in developing countries as everywhere. For
developingcountries,however,itconnectslocalbusinessesto
potentialcustomersandpresentsthenecessaryaccessrouteto
creating local employment and revenues on a global market,
a vital factor in economic development and thus regional
stability. We studied the question whether the internet is a spa-
tial network. To our knowledge, this question has not been
asked before, although there are a number of studies report-
ing a distance effect and a number of technologies dedicated
to handling specific atypical constellations. We tested how
strongly correlated distance and RTT are and found a strong
correlation for seven of the eight time slices by type of day
in the week studied, and for both developed and developing
regions. Moreover, the correlation was in all cases stronger
for distance than for hop count. In developing countries the
correlation was at least 0.57 for all days at all times. Given
this result, we conclude that our research question has been
affirmed. Acknowledgements The author is grateful to the anonymous reviewers
of this article for helpful comments. Open Access This article is distributed under the terms of the Creative
Commons Attribution 4.0 International License (http://creativecomm
ons.org/licenses/by/4.0/), which permits unrestricted use, distribution,
and reproduction in any medium, provided you give appropriate credit
to the original author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made. 3 123 123 Journal of Reliable Intelligent Environments (2018) 4:123–129 129 References 11. Jacobson V, Braden R (1988) RFC 1072: TCP extensions for long-
delay paths. Technical report. IETF. https://tools.ietf.org/html/
rfc1072. Accessed 26 Oct 2017 1. Aikat J, Kaur J, Smith FD, Jeffay K (2003) Variability in TCP
round-trip times. In: Proceedings of the 3rd ACM SIGCOMM con-
ference on Internet measurement. ACM, New York, pp 279–284 12. Khabbaz MJ, Assi CM, Fawaz WF (2012) Disruption-tolerant net-
working: a comprehensive survey on recent developments and
persisting challenges. IEEE Commun Surv Tutor 14(2):607–640 2. Allman M, Paxson V (2009) RFC 5681: TCP congestion con-
trol. Technical report. IETF. https://tools.ietf.org/html/rfc5681. Accessed 26 Oct 2017 13. Maihofer C (2004) A survey of geocast routing protocols. IEEE
Commun Surv Tutor 6(2):32–42 14. Postel J (1981) RFC 793: transmission control protocol. Technical
report. IETF. https://tools.ietf.org/html/rfc793. Accessed 26 Oct
2017 3. Bellardo J, Savage S (2002) Measuring packet reordering. In:
Proceedings of the 2nd ACM SIGCOMM workshop on internet
measurement. ACM, New York, pp 97–105 15. Scott K, Burleigh S (2007) RFC 5050: bundle protocol specifi-
cation. Technical report. IETF. https://tools.ietf.org/html/rfc5050. Accessed 26 Oct 2017 4. Bennett JC, Partridge C, Shectman N (1999) Packet reordering is
not pathological network behavior. IEEE/ACM Trans Netw (TON)
7(6):789–798 16. Shin M, Park M, Oh D, Kim B, Lee J (2014) Adaptive logarithmic
increase congestion control algorithm for satellite networks. TIIS
8(8):2796–2813 5. Caini C, Firrincieli R (2004) TCP Hybla: a TCP enhancement for
heterogeneous networks. Int J Sat Commun Netw 22(5):547–566 6. Dordal PL (2014) An introduction to computer networks. Loyola
University Chicago, Chicago 17. Stewart R (2007) RFC 4960: Stream control transmission pro-
tocol. Technical report. IETF. https://tools.ietf.org/html/rfc4960. Accessed 26 Oct 2017 7. Feng J, Ouyang Z, Xu L, Ramamurthy B (2009) Packet reordering
in high-speed networks and its impact on high-speed TCP variants. Comput Commun 32(1):62–68 18. TeleGeography
(2017)
Submarine
cable
map. http://www. submarinecablemap.com/. Accessed 14 June 2017 8. Gao C, Ling Z, Yuan Y (2014) Packet reordering analysis for con-
current multipath transfer. Int J Commun Syst 27(12):4510–4526 9. Hofmann M, Beaumont LR (2005) Content networking: architec-
ture, protocols, and practice. Elsevier, Amsterdam 10. International Monetary Fund (2018) World economic outlook:
cyclical upswing, structural change. Technical report. International
Monetary Fund, Washington, DC 12 123
|
https://openalex.org/W2923480858
|
https://ejournal.undip.ac.id/index.php/jitaa/article/download/20772/pdf
|
English
| null |
Cross-sectional survey on environmental pollution surrounding poultry production cluster area
|
Journal of The Indonesian Tropical Animal Agriculture/Journal of the Indonesian tropical animal agriculture
| 2,019
|
cc-by-sa
| 5,264
|
ABSTRAK Penelitian ini dilakukan untuk mengetahui pencemaran lingkungan sekitar kluster produksi unggas/
poultry production cluster (PPC). Survei cross-sectional terkait pencemaran lingkungan di sekitar PPC
telah dilakukan di Kabupaten Subang dan Ciamis, Provinsi Jawa Barat, Indonesia. Aspirasi peternak
tentang pencemaran lingkungan dan masalah sanitasi terkait dengan keberadaan PPC diperoleh melalui
kuesioner semi-terstruktur. Gas amonia (NH3) dalam feses diuji secara kualitatif, dan air di sekitar
kandang dianalisis guna melihat kemungkinan adanya pencemaran bakteri Coliform dan Salmonella sp. Hasil studi menunjukkan bahwa debu dan gas amonia yang berasal dari feses (Subang 300-450 ppm,
Ciamis 25-525 ppm) telah mencemari lingkungan dan menimbulkan bau tidak sedap di sekitar PPC. Selain itu, populasi lalat juga meningkat ketika panen ayam, sehingga kehidupan masyarakat tidak
nyaman. Kualitas air di sekitar PPC menunjukkan bahwa kontaminasi Salmonella sp. dapat diabaikan,
akan tetapi sebagian besar sampel dari Subang dan Ciamis terkontaminasi dengan bakteri coliform pada
kisaran <3 MPN/ml-27 MPN/ml. Kebijakan perbaikan manajemen untuk mengurangi pencemaran
lingkungan masih diperlukan dalam mengembangkan daerah PPC. Kata kunci: ammonia, Coliform, lingkungan, pencemaran, kluster produksi unggas E. Martindah1 and N. Ilham2
1Indonesian Research Centre for Veterinary Science (IRCVS),
Jl. RE. Martdinata No. 30, Bogor - Indonesia.
2Indonesian Center for Agriculture Socio Economic and Policy Studies (ICASEPS),
Jl. Tentara Pelajar Cimanggu Blok Kompleks Pertanian No. 3. Bogor - Indonesia
Corresponding E-mail: emartindah@gmail.com E. Martindah1 and N. Ilham2
1Indonesian Research Centre for Veterinary Science (IRCVS),
Jl. RE. Martdinata No. 30, Bogor - Indonesia. 2Indonesian Center for Agriculture Socio Economic and Policy Studies (ICASEPS),
Jl. Tentara Pelajar Cimanggu Blok Kompleks Pertanian No. 3. Bogor - Indonesia
Corresponding E-mail: emartindah@gmail.com Received October 24, 2018; Accepted November 29, 2018 Received October 24, 2018; Accepted November 29, 2018 J. Indonesian Trop. Anim. Agric.
pISSN 2087-8273 eISSN 2460-6278
http://ejournal.undip.ac.id/index.php/jitaa
44(1):56-64, March 2019
DOI: 10.14710/jitaa.44.1.56-64 J. Indonesian Trop. Anim. Agric. pISSN 2087-8273 eISSN 2460-6278
http://ejournal.undip.ac.id/index.php/jitaa
44(1):56-64, March 2019
DOI: 10.14710/jitaa.44.1.56-64 J. Indonesian Trop. Anim. Agric. pISSN 2087-8273 eISSN 2460-6278
http://ejournal.undip.ac.id/index.php/jitaa
44(1):56-64, March 2019
DOI: 10.14710/jitaa.44.1.56-64 J I T A A
Journal of the Indonesian Tropical Animal Agriculture
Accredited by Ditjen Penguatan Risbang No. 60/E/KPT/2016 Keywords: ammonia, Coliform, environment, pollution, poultry production cluster Environmental Problem Surrounding PPC Environmental Problem Surrounding PPC
Concentrations of poultry operations under
PPC scheme increased in operation size. Consequently, environmental pollution problems Examination of Water A total of 18 samples of ground water was
collected from various sources and tested for
microbes of Salmonella sp. and Colliform. Samples of water in PPC Subang were collected
from dug wells (6 samples), artesian wells (2
samples), ditches (1 sample) and pond (1 sample)
with distance from the pen were 200 m, 200-300
m and 1-2 m, respectively. Samples of water in
PPC Ciamis was collected from dug well (8
samples), with distance from the pen was about
100-150 m. Water from dug wells and artesian
wells are used for human consumption and
chickens drinks. Laboratory test for water
pollution, was done at Diagnostic Laboratory
Unit, in Indonesian Research Centre for
Veterinary Science (IRCVS/BBLitvet), Bogor. ABSTRACT This study was carried out to determine the environmental pollution surrounding poultry
production cluster (PPC). A cross-sectional survey on environmental pollution surrounding PPC was
conducted in the districts of Subang and Ciamis, West Java Province, Indonesia. Information of farmers
aspiration on environmental pollution and sanitation issues related to the existence of PPC was collected
by semi-structured questionnaires. Ammonia gas (NH3) in feces was tested qualitatively, and ground
water was analyzed for Coliform and Salmonella sp. contamination. The result showed that dust and
ammonia gases from feces (Subang 300-450 ppm, Ciamis 25-525 ppm) pollute the environment and
caused an unpleasant odor surrounding the pens. Fly population was increasing during the harvest time
of chickens, causing community daily lives were not comfortable. Water quality surrounding PPC
indicated that Salmonella sp. contamination was negligible however most samples from Subang and
Ciamis were contaminated with coliform bacteria (<3 MPN/ml–27 MPN/ml). Improvement on
management policy to reduce the environmental pollution is thus still needed to develop surrounding the
PPC areas. Keywords: ammonia, Coliform, environment, pollution, poultry production cluster 56 J.Indonesian Trop.Anim.Agric. 44(1):56-64, March 2019 56 Ammonia Level Measurement Sample feces were taken from the inside of
pen in PPC Subang (n=8 farmers) and PPC
Ciamis (n=17 farmers) when the the poultry were
between 3-4 weeks of age. Each sample site was
collected from 12 pick-up points,
mixed
homogeneously, so that the collected samples
were representing fresh and dried manure. Ammonia level in manure is determined indirectly
using Nessler’s method (Bartik and Piskac, 1991)
and compared against the standard colour
indicator, from yellow, orange, redness to yellow
rusty (Stair and Whaley 1990). Studies related to environmental pollution by
livestock waste have been reported (Charles and
Hariono 1991; Harper et al. (2010) and reviewed
by Rahmawati (2000). Animal production tends to
be concentrated in relatively small geographical
areas and may increase localized ammonia (HN3)
(Harper et al., 2010). The environmental impacts
are highly dependent on poultry production,
especially on manure management practices. Poultry waste may lead to reduced air quality with
high concentrations of organic and inorganic dust,
microorganism as well as harmful gases such as
ammonia, hydrogen sulfide and methane (Gates et
al., 2008). However, the consequence of the
change from a single small-farm to poultry
production clusters on environmental pollution is
not well understood. The objective of this study
was to determine the environmental pollution
surrounding poultry production cluster (PPC). INTRODUCTION Indonesia. The study was conducted coincide with
dry season. Two PPC in each district were
selected as a study site, they were 53 broiler
farmers in PPC Subang and 63 male-layers
farmers (PPC Ciamis). Both locations were using
open litter pen type farming system. Poultry pens
size in PPC Subang is greater than in PPC Ciamis,
which are correlated linearly with poultry
ownership. Average poultry ownership in Subang
was 4,600 to 5,000 birds per production cycle,
while in PPC Ciamis was 2,200 to 2,900 birds. Information of farmers aspiration on
environmental pollution and sanitation issues
related to the existence of PPC was collected by
semi-structured questionnaires. Poultry production clusters (PPC) compose
of small-scale farmers and it is defined as “areas
of concentrated poultry production in rural areas
usually separated from residential areas” where
farmers operate certain economies of scale
(Aengwanich et al., 2012). The government of
Indonesia has established various policies to
encourage the growth of poultry production
cluster (PPC) in rural areas (Ilham, 2015). This
PPC has been developed through a partnership
with large-scale business, which is accompanied
by the intensification and concentration of poultry
operations. Wang et al. (2015) noted that in many Asian
countries PPC is key strategies to engage small
commercial poultry producers in high-value
production chains and to control infectious
poultry diseases. Consequently, the existence of
PPC has a great impact on the welfare of farmers
(Ilham et al., 2013; Ilham, 2015). In contrast, the
use of large facilities associated with PPC, has
given rise to not only limited to the local
production settings, but extend to environmental
concerns. Poultry farm often considered as a
business which contributes to polluting the
environment, so that chicken density in PPC can
cause many pollution problems. Study Design A cross-sectional study was designed in the
district of Subang and Ciamis West Java, Environmental Pollution Surrounding Poultry Production Cluster (E. Martindah and N. Ilham) 57 have occurred. Access to the farm disposal
facilities to accommodate manure is limited. As a
result, the poultry farm is a source of the smell
and attracts flies, rats and other pests that create
local disturbances. However, Rodić et al. (2011)
pointed out that poultry industry will not threaten
the environment only if both economically and
environmentally acceptable management practices
are applied. Gerber et al. (2007) stated that the
development of PPC would increase waste,
especially from poultry pen, which cannot be
managed by land disposal and cause
environmental problems parasite. The shift from small farm flocks to
poultry production cluster (PPC) operations had
greatly increased people concerns of the fly
population. Flies could breed prolifically in
moist, litters, spoiled feed and plant material as
well as all kinds of manure including poultry
manure (Kalu, 2015). Manure moisture of 70 to
80% is most suitable for fly breeding; and fly
breeding usually less occur in manure with
moisture below 60% (Robertson, et al. 2015). Farmers in PPC Subang (34%) and PPC
Ciamis (17%) had applied flies control in a
combination of mechanical (by using nets and
insect glue) and chemical (spraying and put
larvadex in the feed) (Table 1). This showed that
farmers had a positive effort to control flies not
depends only on chemical control. J.Indonesian Trop.Anim.Agric. 44(1):56-64, March 2019 Ammonia and Odour About 81% of farmers in Subang and 43% in
Ciamis recognized that chicken manure caused a
very disturbing stinging smell (Tabel 2). Odour
issues are serious in the residential area that is
close to PPC facility especially in Subang. Some
people complained about the hostile smell of
ammonia from the poultry pens, especially in
Subang as the poultry density per production
cycle is higher than in Ciamis. This was also
reflected from the aspiration of 87% farmers in
PPC Ciamis that said easy to overcome odour. Similar situation was also occurred in Nigeria,
about 63 % of the residents near by the poultry
farms either resolved to permanently shut their
doors and windows or make verbal complaints
against the foul odours to the headship of the
farms (Akanni and Benson, 2014). Odour may
arise from improper disposal of poultry waste. These odours are from gases that arise as a result
of uncontrolled decomposition of manure (Kalu,
2015). In all excreted animal manures, nitrogen in
the form of ammonia (NH3) is potential to create Table 2 shows that 53% of farmers in
Subang and 87% in Ciamis did not have any
problem in reducing the odour in their farm,
however, in general, did not provide any
particular space to collect the feces (>80%). This
result was in line with studied carried out by Kalu
et al. (2016), that 73.1% of the farmers were not
aware that improper disposal of their waste affects
the environment and human health. Moreover,
most of the farmers do not know how to handle
their waste efficiently. Farmers in Ciamis fed their layers with
broilers feed. Their effort to reduce bad odour was
by adding rice-bran (48%) for covering the
surface of manure and replace the litter more
often, or combination of both. Methods of odour
handling using a compound containing a
microorganism (probiotic) into the feed was
practiced by 10% of farmers in Ciamis. By adding
probiotic into feed would improve feed efficiency,
reduces protein which is not digested, and
expected lessen the formation of gas that causes Table 2. Farmer Perception and Practice to Reduce the Odour
Perception and Farmers’ Effort in Overcoming Odour
PPC Subang
(n= 53)
PPC Ciamis
(n= 63)
Farmers’ Perception of odour
1. Feces generates strong odour (%)
- Yes
81
43
- No
19
57
2. Control of Fly Population The results of the survey based on farmers'
perceptions (Table 1) showed that most farmers in
PPC Subang (70%) admit their difficulties in
eradicating flies populations. While in PPC
Ciamis the proportion between farmers
experiencing difficulties and farmers have no
problems to control flies were almost similar,
being 49 and 51%, respectively. People in the
study sites complaint frequently of discomfort due
to the increased of flies population, especially at
harvest time of poultry. Flies are considered
environmental pollutants just by their presence. The population of flies may cause a public health
problem. It is known that flies could act as a
disease vector and can spread various pathogen
agents of the diseases such as typhoid fever,
salmonellosis, diarrhea, cholera, and another According to Roberstson et al. (2015), four
basic management strategies such as barn
management, biological control, mechanical
control and chemical control make up a successful
integrated fly control program. Some control
methods can be applied simultaneously. In the
study sites, 60-65% of farmers used insecticide
(chemical) by spraying to control the adult flies
(Table 1). They found that spraying was the most
effective and economical method to control heavy
populations of adult flies. Even though chemical
control methods have shown a reduction in fly
density, its control for routine long-term use can
lead to the development of insecticide resistance
(Dogra and Aggarwal, 2010). Robertson et al. 58
J.Indonesian Trop.Anim.Agric. 44(1):56-64, March 2019
Table 1. Farmers’ Perception and Effort to Control Flies Population
Perception and effort to control flies
PPC Subang
(n= 53)
PPC Ciamis
(n=63)
Difficulty of controlling flies in the poultry sheds (%):
a. Yes
70
49
b. No
30
51
Effort to control flies (%):
a. Spray
60
65
b. Insect Net
0
2
c. Mechanical control by Insect glue
2
2
d. Larvadex in the feed
4
0
e. No action
0
14
f. Others / Combination as mentions above
34
17 Table 1. Farmers’ Perception and Effort to Control Flies Population J.Indonesian Trop.Anim.Agric. 44(1):56-64, March 2019 58 (2015) noted that insecticides only control the
adult flies, as any pupae in manure will still
emerge as adults, moreover, it harm the
environment and affect birds, if
applied
improperly. odours and negatively impact on air quality as
well as animal and human health. Kalu et al. Control of Fly Population (2016) stated that odour from animal feeding
operations is caused of a large number of
contributing compounds including ammonia
(NH3), volatile organic compounds (VOCs), and
hydrogen sulfide (H2S). Environmental Pollution Surrounding Poultry Production Cluster (E. Martindah and N. Ilham)
59 Ammonia and Odour Coping with odour (%):
- Difficult
47
13
- Easy to overcome odour
53
87
3. A place to collect feces (%)
- Available
15
14
- Not available
85
86
Effort to reduce odour (%)
1. Adding rice bran in litter
40
48
2. Providing starbio/probiotic into the feed
0
10
3. Replacing bran more often/once a week
15
21
4. Combination
45
21 Table 2. Farmer Perception and Practice to Reduce the Odour Table 2. Farmer Perception and Practice to Reduce the Odour Table 2. Farmer Perception and Practice to Reduce the Odour 1. Adding rice bran in litter Environmental Pollution Surrounding Poultry Production Cluster (E. Martindah and N. Ilham) Environmental Pollution Surrounding Poultry Production Cluster (E. Martindah and N. Ilham) 59 smells in
manure
(Rahmawati, 2000). Furthermore, the used of 1-3% lime and 0.025-
0.05% starbio-probiotics appears to be a good
choice compared to zeolites and EM4®. poultry farms. Ammonia emission from poultry
manure can cause several problems such as poor
poultry performance, reduce the poultry’s
immunity, and damage the bird’s respiratory
systems (Aziz and Barnes, 2009). In addition,
they stated that at high concentrations, ammonia
is irritating the conjunctivae of the eyes and
damage the mucous membranes of the respiratory
system which increases the susceptibility of birds
to bacterial infection, especially E. coli infection. Broilers reared in an environment with ammonia
concentration over 25 ppm showed a reduction in
antioxidant capacity (Wei et al., 2012), and reduce
carcass traits and immune organ indices and
increased the kidney and hepatic indices (Xing et
al., 2016). Hutabarat (2007) summarized that in a
range of 200-400 ppm, ammonia could cause
nasopharyngeal irritation, while the level of >400
– 500 ppm causes direct hazardous impact to the
human health. The frequency relative of ammonia level
more than 300 ppm in PPC Subang and Ciamis
was 75% and 47.6% respectively (Table 3). This
indicted that in those two sites the level of
ammonia is relatively high, even though in a
cluster of Broiler (Subang) was slightly higher
than male-layers cluster (Ciamis). This result was
in line with the farmer effort in reducing odour as
reflected in Table 2, in which farmers in Ciamis
(48%) adding rice bran and 10% of farmers added
probiotic in the chicken feed. Ammonia and Odour The emission rates
of the pollutants depend on many factors
including temperature, humidity, wind speed and
weather conditions, ventilation, housing type, and
manure properties and characteristics – for
example, dryer manure and litters result in more
particulate emission, while moist conditions are
likely to result in increased emission of ammonia
(Williams, 2013). Aerosol contamination from
poultry production can generally be characterized
as pollutants, including gases (such as ammonia),
particulates (dust) and microbial pathogens. Kalu
(2015) noted that environmental problems such as
odor nuisance and land pollution resulting from
improperly discharged manure. The ammonia level needs to be controlled,
not only for the animal health but also for the
public health. According to Harper et al. (2010),
the ammonia emission will increase steadily after
the third week of poultry growth. The emission
rate of ammonia increase in a linear relationship
with age from chick placement to the end of the
flock (Gates et al., 2008). Choi et al. (2011)
indicated that adding liquid aluminum chloride to
rice husk would be a useful in reducing the
negative environmental impact of litter. The
decreased volatile fatty acids (VFA) production
and (NH3) volatilization was associated with
reduction in litter pH. To achieve efficient farming
and to maintain good environmental quality,
attention has to be given to farm management,
housing and waste handling. The heat produced The workers in PPC and people who live
nearby the PPC are potentially exposed to
ammonia if feces (litter) are not handled properly. Maguire et al. (2006) stated that if manure left
unattended to for more than 72 hours, the rate of
ammonia volatilization would be higher, thereby
creating environmental pollution for the birds,
worker in the farm and people living close to the Table 3. The Ammonia Level in PPC Subang and Ciamis 60
J.Indonesian Trop.Anim.Agric. 44(1):56-64, March 2019
Table 3. The Ammonia Level in PPC Subang and Ciamis
Location and Ammonia Level
Frequency
Frequency relative (%)
PCC Subang (n = 8)
• 250 - 300 ppm
2
25.00
• > 300 ppm
6
75.00
PCC Ciamis (n = 17)
• 250 - 300 ppm
9
52. 94
• > 300 ppm
8
47. Ammonia and Odour 06
TOTAL (n = 25)
• 250 - 300 ppm
11
44.00
• > 300 pp m
14
56.00 PCC Subang (n = 8)
• 250 - 300 ppm
2
25.00
• > 300 ppm
6
75.00
PCC Ciamis (n = 17)
• 250 - 300 ppm
9
52. 94
• > 300 ppm
8
47. 06
TOTAL (n = 25)
• 250 - 300 ppm
11
44.00
• > 300 pp m
14
56.00 PCC Subang (n = 8)
• 250 - 300 ppm
• > 300 ppm
PCC Ciamis (n = 17)
• 250 - 300 ppm
• > 300 ppm
TOTAL (n = 25)
• 250 - 300 ppm
• > 300 pp m 60
J.Indonesian Trop.Anim.Agric. 44(1):56-64, March 2019 J.Indonesian Trop.Anim.Agric. 44(1):56-64, March 2019 J.Indonesian Trop.Anim.Agric. 44(1):56-64, March 2019 60 during composting completely reduces the
pathogenic organisms in the waste (Adeoye et al.,
2014). In this regards, the government of
Indonesia
has
issued
decree
No. 31/Permentan/OT.140/2/2014 through Minister of
Agriculture. The decree state about guidelines of
good management for broilers and layers in which
a farm of broilers and layers needs to be equipped
with good management and environmentally
friendly (Kementerian Pertanian, 2014). Sartika et
al. (2015) stated that the development of PPC
triggers another serious problem, not only to
animal diseases that threaten the production, but
also impact on public health. Aerosol emissions
from poultry production can transmit
communicable diseases to nearby poultry flocks;
scientific evidence shows that some pathogenic
microorganisms can remain viable and able to be
transported for considerable distances (from 50 to
more than 500 m) in ambient air (Williams,
2013). collected and tested for microbes of Salmonella
sp. and Coliform. The laboratory test result
(Tables 4 and 5) shows that all samples from
water sources are negative of Salmonella sp. However, for Coliform mostly present at <3
MPN/ml – 27 MPN/ml. There were two samples
with a high level of coliform contamination, i.e a
sample from PPC Subang with 290 MPN/ml of
the coliform level, and a sample from Ciamis with
160 MPN/ml of coliform. Water with this level of
coliform must be boiled for human consumption,
considers the coliform level is already on the
verge (PP No. 20/1990). Those two samples were
collected from the buckets where the source is
artesian well and dug well respectively. Ammonia and Odour Although
the water appearance looks clear, there is a
possibility the water is contaminated from septic
tank seep or is contaminated when in the water
container. The water source in the study sites might be
contaminated by absorption or seepage of pen
waste-water (pen washed activity). Coliform is a
family of bacteria made up of several groups, one
of which is the fecal coliform group, which is
found in the intestinal tracts of warm-blooded
animals including humans. The presence of
coliform bacteria is typically an indication of
fecal contamination. When water has a high
bacterial count, the best option is to eliminate the
source of the contamination or to locate an
alternative water source. Study conducted by Onu
et al. (2015) indicated that the effects of poultry
production activities include the degradation of Water Quality Surrounding PPC In the poultry farm, the source and quality of
water are important. The source of drinking water
must be free from contamination of microbial
pathogens. A study in Thailand found where egg-
laying chickens are raised over fish ponds,
resulted in water both in the fish ponds and the
public water sources (Aengwanich et al., 2014). Mostly the water source in the study sites comes
from dug wells and is distributed through pipes
with the electric water pump. A sample of water is Table 4. Water Analysis Result (Salmonella sp. and Coliform) in PPC Subang
Water
Source
Distance from
Poultry Pens (m)
Usage
Physic, pH
Salmonella
sp. Coliform
MPN/mL
Artesian l
5
poultry
clear, 5.5
negative
290
Dug Well
1
poultry
clear, 6.5
negative
< 3
Dug Well
400
human
clear, 5.5
negative
14
Ditch
400
-
turbid, 7.5
negative
< 3
Artesian
1
poultry, human
clear, 6
negative
< 3
Pond
0.5
pen wash, fish pond
turbid, 7.5
negative
4
Dug Well
400
human
clear, 5.5
negative
< 3
Dug Well
50
human
clear, 6
negative
27
Spring
200
poultry, human
turbid, 6.5
negative
< 3
Dug Well
200
poultry, human
clear, 6.5
negative
< 3 Table 4. Water Analysis Result (Salmonella sp. and Coliform) in PPC Subang Table 4. Water Analysis Result (Salmonella sp. and Coliform) in PPC Subang Environmental Pollution Surrounding Poultry Production Cluster (E. Martindah and N. Ilham) Environmental Pollution Surrounding Poultry Production Cluster (E. Martindah and N. Ilham) 61 Table 5. Water Analysis Result (Salmonella sp. and Coliform) in PPC Ciamis
Source
Distance from
Poultry pens (m)
Usage
Physic, pH
Salmonella
sp. Coliform
MPN/mL
Dug Well
1
poultry, human
clear, 5.5
negative
4
Dug Well
50
poultry, human
clear, 7
negative
15
Dug Well
200
poultry, human
clear, 5.5
negative
7
Dug Well
1
poultry
clear, 5.5
negative
< 3
Dug Well
150
poultry, human
clear, 6
negative
4
Dug Well
70
poultry, human
clear, 5.5
negative
39
Dug Well
10
poultry, human
clear, 5.5
negative
160
Dug Well
10
poultry, human
clear, 6
negative
< 3 Table 5. Water Analysis Result (Salmonella sp. and Coliform) in PPC Ciamis nearby surface and /or underground water, as well
as pollution of the environment through the
emission of foul odour; thus causing discomfort to
both the human and animal lives. Environmental
hygiene control is not prioritized by the local
government in study the sites. CONCLUSION The direct consequences of PPC
development are the increase of waste from the
poultry pens which cause environment problem,
such as odour issues in residential surrounding
PPC, increasing fly population, as well as water
and air contamination. The shift from small farm
flocks to PPC had greatly increased farmers’
concerns of a fly population that was difficult to
be eradicated especially during the harvest time of
chickens. A combination of mechanical and
chemical for fly control is a positive effort in
managing pesticide resistance. Acidity and alkalinity of water are one that
indicates the quality of drinking water. The pH,
hardness, and total dissolved solids (TDS) can all
affect consumption patterns. Water with a pH of 7
is neutral; a pH greater than 7 indicates alkalinity,
while a pH less than 7 indicates acidity. Good
water/normal water has a pH around 6.5 – 7.2. The resulting test of water sample in study sites
show the water pH used for poultry drink is about
5.5 – 7, and physically the water is clear. When
the water pH is lower or higher in the normal
range, it can affect the medicine solubility,
especially for medicine that difficult to dissolve
homogenously in water. Water with a low pH can
be unpalatable, while high pH water can clog
watering systems because of excessive mineral
levels, especially calcium and magnesium. Water
outside maximum acceptable levels for both high
and low pH can negatively impact performance
(Tabler et al., 2013). It can affect chickens Water Quality Surrounding PPC Such a situation is
common in developing countries however, there is
a need for collateral personal hygiene and sanitary
education to achieve improved outcome (Mara et
al., 2010; Jenkins et al., 2014). However, Elsaidy
et al. (2015) stated that different water sources is
safe as drinking water for poultry; as long as it is
acceptable within the range of drinking water
quality for chickens, he suggested of maintaining
the hygienic quality of stored water. drinking consumption which also affect the
growth and productivity of the
chickens. Moreover, he suggested that regular water
sanitation program on the farm will assist farmers
in preventing unhealthy environments. Providing
a clean and safe water supply is critical to
ensuring that poultry performs at their best. REFERENCES 864. Gates, R.S., K.D. Casey, E.F. Wheeler, H. Xin and
A.J. Pescatore. 2008. U.S. broiler ammonia
emissions inventory model. Atmos. Environ. 42:3342-3350. Adeoye, P.A., C.M. Hasfalina, M.S.M. Amin,
A.M. Thamer and C.O. Akinbile. 2014. Environmental implication of poultry waste
generation and management techniques in
Minna, Semi-arid Region of Nigeria. Annual Res. Rev. Biol. (ARRB). 4(10):1669-1681 Gerber, P., C. Opio and H. Steinfeld. 2007. Poultry production and the environment – a
review. Poultry in the 21st Century. Animal
Production and Health Division, FAO,
Rome. Aengwanich, W., M. Intarakhamhaeng, J. Wandee,
T. Nongbua, S. Chaiyasak, P. Srikot, K. Thammasar, N. Junsanitsri, K.Sritongtuam
and
T. Tawinwaang. 2012. Poultry
Production Clusters (PPCs)
after AI
outbreaks in Thailand: Past, present and
future direction. Int. J. Poultry Sci.11: 541–
550 Harper, L.A., T.K. Flesch, and J.D. Wilson. 2010. Ammonia emmisions from broiler
production in the San Joaquin Valley. Poult. Sci. 89:1802-1814. Hutabarat, I.O. 2007. Analisis dampak gas
ammonia dan khlorin pada faal paru pekerja
pabrik sarung tangan karet “X” Medan. Thesis. Universitas Sumatra Utara, Medan. Aengwanich, W., T. Boonsorn and K. Thammasar. 2014. Mathematical equations for reducing
water pollution problems among Poultry
Production Clusters in Nong Khai Province,
Thailand. J. Geosci. Environ. Protection
(GEP). 2:48-51 Ilham, N., Y. Yusdja, E. Basuno, E. Martindah,
and R.A.D. Sartika. 2013. Ecohealth
assesment on poultry production clusters for
the livelihood of improvement small
producers. Final Report (Indonesia). Indonesia Center for Sosio-Economic and
Agriculture Policies. Indonesian Agency for
Agricultural Reserch and Development. Ministry of Agriculture – Republic of
Indonesia collaboration with International
Development Research Centre (IDRC) –
Canada. Akanni, K.A and O.B. Benson. 2014. Poultry
wastes management strategies and
environmental implications on human
health in Ogun State of Nigeria. Advances
in Econ. Business 2(4):164-171. Aziz, T. and H.J. Barnes. 2009. Harmful effect of
ammonia on birds. World Poult. 26(3):28-
30. Ilham, N. 2015. Kebijakan pemerintah terhadap
usaha unggas skala kecil dan kesehatan
lingkungan di Indonesia. Wartazoa. 25(2):
95-105. Bartik, M. and A. Piskac. 1991. The Main
Chemical and Toxicological Method of
Examination in Farm Animal Poisoning
Veterinary Toxicology. Eksebisi Suci. Pub. Co. New York. 298-319. Jenkins, M.W, M.C. Freeman and P. Routry. 2014. Measuring the safety of excreta disposal
behaviour in India with the new Safe San
Index: Reliability, validity and utility. Int. J. Environ. Res. Public Health. 11:8319-46. Charles, R.T. and B. Hariono. 1991. Pencemaran
lingkungan oleh limbah peternakan dan
pengelolaannya. Bull. FKH-UGM. X(2):
71-75. Kalu, E. 2015. ACKNOWLEDGMENTS This study is funded through a capacity-
building initiative for Ecohealth Research on
Emerging Infectious Disease in Southeast Asia
supported by the International Development
Research Centre (IDRC), the Canadian
International Development ment Agency (CIDA)
and the Australian Agency for International
Development (AusAID) in partnership with the
Global Health Research Initiative. J.Indonesian Trop.Anim.Agric. 44(1):56-64, March 2019 62 REFERENCES Poultry litter/manure management
practices in intensively managed poultry
farms in Portharcourt. IOSR J. Agric. Vet. Sci. 8(3):53-58. Choi, I.H., J.H. Choi, S.H. Ko and P.A. Jr. Moore. 2011. Reducing ammonia emissions and
volatile fatty acids in poultry litter with
liquid aluminum chloride. J. Environ. Sci. Health. B. 46(5):432-435. Kalu E., A.N. Ajaruonye, and N. Okwara. 2016. Waste management practices in selected
poultry farms in Umuahia, Abia State. J. Vet
Adv. 6(9):1310-1316. Dogra, V. and A.K. Aggarwal. 2010. Association
of poultry farms with housefly and
morbidity: A comparative study from
Raipur Rani, Haryana. IJCM. 35(4):473-
477. Kementerian Pertanian. 2014. Peraturan Menteri
Pertanian Republik Indonesia Nomor
31/Permentan/OT.140/2/2014,
tentang
Pedoman budi daya ayam pedaging dan
ayam petelur yang baik. Kementerian
Pertanian, Jakarta Indonesia. Elsaidy, N., Mohamed R.A. And F. Abouelenien. 2015. Assessment of variable drinking
water sources used in Egypt on broiler
health and welfare. Vet. World 8(7):855- Mara D., J. Lane, B. Scott and D. Trouba. 2010. Environmental Pollution Surrounding Poultry Production Cluster (E. Martindah and N. Ilham) 63 Sanitation and health. PloS Med
7:e1000363 Sci. 2(4):15-23 Sci. 2(4):15-23 Sci. 2(4):15-23 Stair, E.L. and M. Whaley. 1990. Rapid screening
and spot tests for the presence of common
poisons. Vet. Hum. Toxicol. 32(6): 564-566 Maguire R.O., D. Hesterberg, A. Gernat, K. Anderson, M. Wineland, and J. Grimes. 2006. Liming poultry manures to decrease
soluble phosphorus and suppress the
bacterial population. J. Environm Qual. 35:
849-857. Tabler T., J. Wells and W. Zhai. 2013.Water
Quality Critical to Broiler Performance. Extension Service of Mississippi State
University, cooperating with U.S. Department of Agriculture. Publication
2754. Onu, D.O., E.I. Offor and B.O. Okpara. 2015. Poultry wastes management strategies and
environmental implications in Abia State. Int. Res. J. Agric. Sci. Soil Sci. 5(6):159-
164. Wang, L, E. Basuno, T. Nguyen, W. Aengwanich,
N. Ilham and X. Li. 2015. An ecohealth
assessment of poultry production clusters
(PPCs) for the livelihood and biosecurity
improvement of small poultry producers in
Asia. Infectious Diseases of Poverty. 4:6. Peraturan Pemerintah. 1990. Peraturan Pemerintah
No. 20 Tahun 1990 Tentang: Pengendalian
Pencemaran Air. Jakarta. Williams, C.M. 2013. Poultry waste management
in developing countries: Aerosol
contamination. Poultry Development
Review. Food and Agriculture Organization
(FAO) of the United Nations. Rahmawati, S. 2000. Upaya pengelolaan
lingkungan
usaha
peternakan ayam. Wartazoa. 9(2):73-80. Robertson, A., D. Ward and S. Lachance. 2015. House Fly Control in Poultry Barns. Publication 849. Queen’s Printer for
Ontario. Toronto. Canada. Wei F.X., B. Xu, R.N. J.Indonesian Trop.Anim.Agric. 44(1):56-64, March 2019 REFERENCES Sa, S.Y. Li , F.Z. Liu and
Q.Y. Sun. 2012. The effect of ambient
relative humidity
and ammonia on
antioxidant capacity and meat quality of
broiler chickens. Acta Vet Zootech Sin. 43(10):1573-1581 Rodić, V., L. Perić, M. Đukić-Stojčić and N. Vukelić. 2011. The environmental impact of
poultry
Production. Biotechnol. Anim. Husbandry. 27(4):1673-1679. Xing, H., S. Luan, Y. Sun, R. Sa and H. Zhang. 2016. Effects of ammonia exposure on
carcass traits and fatty acid composition of
broiler meat. Animal Nutrition. 2(2016):
282-287. Sartika, R.A.D., N. Ilham and R.A. Wulandari. 2015. The impact of poultry production
cluster (PPC) and Non PPC on child health
in Indonesia (A qualitative research). J. Advanced Studies in Agric. Biol. Environ. 64 J.Indonesian Trop.Anim.Agric. 44(1):56-64, March 2019
|
https://openalex.org/W3106479234
|
https://link.springer.com/content/pdf/10.1007%2Fs00340-016-6402-3.pdf
|
English
| null |
An efficient source of frequency anti-correlated entanglement at telecom wavelength
|
Applied physics. B, Lasers and optics
| 2,016
|
cc-by
| 5,648
|
1 Introduction Abstract We demonstrate an efficient generation of fre-
quency anti-correlated entangled photon pairs at telecom
wavelength. The fundamental laser is a continuous-wave
high-power fiber laser at 1560 nm, through an extracavity
frequency doubling system, a 780-nm pump with a power
as high as 742 mW is realized. After single-passing through
a periodically poled KTiOPO4 (PPKTP) crystal, degenerate
down-converted photon pairs are generated. With an overall
detection efficiency of 14.8 %, the count rates of the sin-
gle photons and coincidence of the photon pairs are meas-
ured to be 370 kHz and 22 kHz, respectively. The spectra
of the signal and idler photons are centered at 1560.23 and
1560.04 nm, while their 3-dB bandwidths being 3.22 nm
both. The joint spectrum of the photon pair is observed
to be frequency anti-correlated and have a spectral band-
width of 0.52 nm. According to the ratio of the single-pho-
ton spectral bandwidth to the joint spectral bandwidth of
the photon pairs, the degree of frequency entanglement is
quantified to be 6.19. Based on a Hong–Ou–Mandel inter-
ferometric coincidence measurement, a frequency indistin-
guishability of 95 % is demonstrated. The good agreements
with the theoretical estimations show that the inherent extra
intensity noise in fiber lasers has little influence on fre-
quency entanglement of the generated photon pairs. Entangled photon pairs based on spontaneous parametric
down-conversion (SPDC) process [1] are not only at the
heart of the most fundamental tests of quantum mechanics
[2, 3] but also essential resources for quantum information
processing (QIP) [4]. The entanglement between the gener-
ated signal and idler photons can be in polarization, photon
quadrature, angular momentum, frequency, wave vector,
etc. Recently, frequency entanglement have attracted many
interests due to its wide applications in quantum enhanced
positioning and clock synchronization [5–14], quantum
spectroscopy [15], quantum optical coherence tomography
[16–20], quantum imaging [21, 22], nonlinear microscopy
[23–25], and quantum key distribution [26, 27]. Because of low-loss, high-stability features of optical
fibers, they become a widely used transmission medium. In a fiber-based quantum information network, frequency-
entangled photon pair source at telecom wavelength is
required for minimum absorption. Several experiments
on generation of frequency-correlated photon pairs at tel-
ecom wavelength have been reported [28–31] since pulse
pumped SPDC can be generated and detected much more
efficiently. However, in application to the entanglement-
based quantum communication, their broad spectral width
of the photon pairs limits the fiber transmission distance. An efficient source of frequency anti‑correlated entanglement
at telecom wavelength Feiyan Hou1,2 · Xiao Xiang1,2 · Runai Quan1,2 · Mengmeng Wang1,2 · Yiwei Zhai1,2 ·
Shaofeng Wang1,2 · Tao Liu1 · Shougang Zhang1 · Ruifang Dong1 Received: 9 December 2015 / Accepted: 24 March 2016 / Published online: 27 April 2016
© The Author(s) 2016. This article is published with open access at Springerlink.com DOI 10.1007/s00340-016-6402-3
Appl. Phys. B (2016) 122:128 * Ruifang Dong
dongruifang@ntsc.ac.cn 1
Key Laboratory of Time and Frequency Primary Standards,
National Time Service Center, Chinese Academy of Science,
Xi’an 710600, China 2
University of Chinese Academy of Sciences, Beijing 100049,
China 1 Introduction By using of frequency anti-correlated photons, the del-
eterious effects of chromatic dispersion can be mitigated. Particularly, frequency anti-correlated photon pairs have
unprecedented advantage of dispersion cancelation in
quantum synchronization [6, 11, 14]. Therefore, the gen-
eration of frequency anti-correlated entangled photon
pairs in the telecom wavelength regime and with a high
brightness is required for the long-distance quantum
communication. 1 3 3 F. Hou et al. 128 Page 2 of 8 Fig. 1 The experimental setup
for generation and characteriza-
tion of the frequency anti-corre-
lated entangled biphoton source
at 1560 nm. The inset labeled
by A denotes the spectral
measurement setup, while B is
for the HOM interferometric
measurement
EOM
1560nm
Fiber laser
Isolator
SHG
PDH cavity
locking circuit
780nm
Block
HWP-1
PPKTP
crystal
Filters
Dichroic
mirrors
FPBS
Detector
Equilateral
Dispersive
Prism
Infrared
monochromator1
D1
(SPD4)
Coincidence
measurement
device
Lens
Aspheric
lens
HWP-2
Infrared
monochromator2
MDL
ODL
FPC
50/50FC
A
B
D2
(SPD4) Fig. 1 The experimental setup
for generation and characteriza-
tion of the frequency anti-corre-
lated entangled biphoton source
at 1560 nm. The inset labeled
by A denotes the spectral
measurement setup, while B is
for the HOM interferometric
measurement PDH cavity
locking circuit 1560nm
Fiber laser SHG Isolator 780nm 780nm Equilateral
Dispersive
Prism Detector D1
(SPD4) Infrared
monochromator1 Fil
HWP-2 Lens Coincidence
measurement
device Block Infrared
monochromator2 Aspheric
lens frequency anti-correlated and have a spectral bandwidth
of 0.52 nm. Thus the degree of frequency entanglement is
quantified to be 6.19. Based on a Hong–Ou–Mandel
(HOM) interferometric coincidence measurement, a fre-
quency indistinguishability of 95 % is demonstrated. The
good agreements with the theoretical estimations show
that the inherent extra intensity noise in fiber lasers has
little influence on frequency entanglement of the gener-
ated photon pairs. For the generation of frequency anti-correlated entan-
gled photon pairs at 1560 nm, a continuous-wave (cw) 780-
nm laser source is required as the pump of a type-II SPDC. However, most of such 780-nm laser sources with a power
higher than 200 mW are acquired by a Ti:Sapphire laser
pump, the application of which is limited by their high
cost. On the other hand, with the rapid development of fiber
amplifiers, a low-cost, high-power, single-frequency fiber
laser at 1560 nm can be obtained based on a Er3+/Yb3+
-doped single-mode phosphate glass fiber [32, 33]. 1 The laser is provided by Prof. Z. Yangs group of South China Uni-
versity of Technology, which is a collaborative product of the Cross
and Cooperative Science and Technology Innovation Team Project of
the CAS, China. 1 Introduction For
acquiring a high-power cw 780-nm laser source, frequency
doubling of a high-power 1560-nm fiber laser becomes a
preferable method [32, 34]. The rest of the paper is organized as follows. The experi-
mental setup for the generation of the frequency anti-cor-
related entangled state at 1560 nm is described in Sect. 2,
and a theoretical model of the setup is presented in Sect. 3. Experimental results are presented in Sect. 4. Finally,
Sect. 5 contains a brief conclusion. In this paper, we demonstrate an efficient generation of
frequency anti-correlated entangled photon pairs at
1560 nm. First the generation of a high-power cw laser
source at 780 nm with an extracavity frequency doubling
of a custom-made 1560-nm single-frequency fiber laser is
presented,1 the power of which can achieve 742 mW. Based on this source and a type-II SPDC crystal of
PPKTP, frequency anti-correlated entangled photon pairs
at 1560 nm are then generated. With an overall detection
efficiency of 14.8 %, the count rates of the single photons
and coincidence of the photon pairs are measured to be
370 kHz and 22 kHz respectively. Compared with the first
report on such photon pairs generation [41], the single-
count rate has been improved for sevenfold while the
coincidence-count rate for fourfold. According to the
spectroscopic measurements, the spectra of the signal and
idler photons are centered at 1560.23 and 1560.04 nm
while their 3-dB bandwidths being 3.22 nm both, while
the joint spectrum of the photon pair is observed to be Fig. 1 The experimental setup
for generation and characteriza-
tion of the frequency anti-corre-
lated entangled biphoton source
at 1560 nm. The inset labeled
by A denotes the spectral
measurement setup, while B is
for the HOM interferometric
measurement 2 Experimental setup The setup used for experimental generation of frequency
anti-correlated entangled biphoton source at 1560 nm
can be divided into two parts. As shown in Fig. 1, the
first part is for generation of the frequency-doubling-
based quasi-monochromatic 780-nm pump source. The
fundamental laser source we use is a single-frequency
1560-nm fiber laser with a linewidth of 2 kHz and a
maximum output power of 2.5 W (see footnote 1). The
frequency doubling external cavity is composed of two
concave mirrors with an identical radius of 50 mm, and
the cavity length is set to be 103.5 mm. The incoupler
M1 has a high reflectivity of R1 > 99 % for the sec-
ond harmonic 780 nm beam, while the transmittance to
1560 nm is 5 %. The outcoupler M2 is 99.9 % reflective
to 1560 nm and its transmittance to 780 nm is 95.2 %. A
piezo-actuator (PZT) is attached to M2 for scanning and
locking the cavity length. To realize the second harmonic
generation, a type-I PPKTP crystal (Raicol Ltd.) with a
size of 1 mm × 2 mm × 10 mm and a poling period 1 3 An efficient source of frequency anti-correlated entanglement at telecom wavelength Page 3 of 8 128 PPKTP
P2
P1
Pr
Pc
M1,r1,t1
M2,r2,t2
Fig. 2 Theoretical model of a singly-resonant two-mirror SHG cav-
ity of 24.925 µm is inserted into the center of the cavity. The working temperature is optimized and fixed at 70◦C
for maximum output harmonic power. Before coupling
into the second harmonic cavity, the fundamental 1560-
nm fiber laser passes first through an EO modulator and
an optical isolator and then is collimated to focus onto
the center of PPKTP with a beam waist of 66 µm. The
total transmittance out of these optics is measured to be
64.2 %. Applying PDH locking technique [35], a stable
second harmonic generation at 780 nm can be output and
used as the pump of the subsequent SPDC. Fig. 2 Theoretical model of a singly-resonant two-mirror SHG cav-
ity 3 Theoretical model After filtering out the residual fundamental 1560-nm
laser by an equilateral dispersive prism and a series of
short-pass filters (FESH1000, Thorlabs), the generated
780-nm laser subsequently single passes through a type-
II PPKTP crystal for generating the down-converted
signal and idler photons with orthogonal polarizations. The PPKTP has a length of 10 mm and a poling period
of 46.146 µm, while the 780-nm beam is focused to a
waist of 48 µm in the center of PPKTP. Through care-
ful evaluation, the temperature of PPKTP is set at 64◦C
to ensure frequency degeneracy of the generated pho-
ton pairs. Via a series of dichroic mirrors and long-pass
filters, the 780-nm laser is filtered out from the gener-
ated photon pairs. Afterward, the photon pairs are cou-
pled into a fiber polarization beam splitter (FPBS). With
the help of a half-wave-plate (HWP) at the entrance, the
signal and idler are separated into two output ports of
FPBS. Linking the two ports to two infrared monochro-
mators (MONO 1&2, Jobin Yvon MicroHR, see inset
A of Fig. 1), the joint spectral properties of the photon
pairs can be measured [36] by scanning the MONOs and
monitoring the outputs of them with the help of two sin-
gle photon detectors (SPD4, D1 & D2) and subsequent
coincidence measurement device (Ortec 453). When we
measure the individual spectrum of the photon pairs,
only one of the MONOs is connected, while the other
output of the FPBS is sent directly to the single pho-
ton detector. With the information of the individual and
joint spectral properties of the down-converted photon
pairs, the degree of frequency entanglement can then
be evaluated. Furthermore, replacing the MONOs setup
with a fiber-based HOM interferometer [37] as shown in
inset B, the frequency indistinguishability can be meas-
ured. In the HOM interferometer shown in inset B, the
ODL is a manually adjusted optical delay line (ODL,
General Photonics Inc.), while the MDL is a motorized
optical delay line (MDL-002, General Photonics Inc.)
which is adjustable with a resolution of 1 fs within the
range of 0–560 ps. A fiber polarization controller (FPC)
is inserted in one arm to match the polarization in the
other arm 3.1 Second harmonic generation of 780‑nm source In the experiment, we use a singly-resonant two-mirror
standing wave cavity for second harmonic generation
(SHG). Its theoretical model is shown in Fig. 2. In the experiment, we use a singly-resonant two-mirror
standing wave cavity for second harmonic generation
(SHG). Its theoretical model is shown in Fig. 2. Assume that the reflectivity and transmittance of incou-
pler (outcoupler) M1(M2) to the fundamental 1560-nm
source are r1(r2) and t1(t2) respectively. The transmittance
of M2 to 780 nm is t2SH. t denotes the single-pass efficiency
of the fundamental source through the cavity, t = 1 −δ ,
where δ represents the total intra-cavity loss. γSH is the
nonlinear conversion coefficient, which can be deducted
according to experimental parameters [38]. The output
power of the second harmonic generation is then given by (1)
P2 = 2γSHP2
ct2SH, (1) P2 = 2γSHP2
ct2SH, where Pc denotes the intra-cavity cycling power of the fun-
damental source. It is related to the input power by the fol-
lowing formula: (2)
Pc =
t1
(1 −√r1rm)2 P1, (2) where rm describes the total transmission efficiency of the
fundamental source after a single round-trip in the SHG
cavity, it is given by (3)
rm = t2 · (1 −γSH · Pc)2 · r2. (3) Combining the above formulae, the second harmonic
power out of the SHG cavity as a function of the input fun-
damental power can be analyzed. 3.2 Characterization of the generated photon pairs With the second harmonic generated 780-nm source as
the pump, a type-II PPKTP crystal is used to generate fre-
quency anti-correlated entangled photon pairs. The two-
photon state can be expressed as [39, 40, 42, 43] |Ψ =
dωsdωiA(ωs, ωi; T)a†
s(ωs)a†
i (ωi)|0 , (4) 1 3 3 128 Page 4 of 8 F. Hou et al. 128 Page 4 128 Alternatively, it can be measured by the overlap between
the individual spectra of signal and idler photons [29]. where ωs, ωi denote the frequencies of the signal and idler pho-
ton, respectively; A(ωs, ωi; T) denotes the joint spectral ampli-
tude function of the generated photon pairs, which is given by
the product of the pump spectrum α(ωs, ωi) and the phase-
matching function ΦL(ωs, ωi; T) of the nonlinear crystal. The degree of frequency entanglement denotes the non-
classical correlation between the spectral distributions of
the signal and idler, which is fundamentally characterized
by the Schmidt number K [47]. The larger K is, the higher
the entanglement. Since the Schmidt number K cannot be
measured directly, an operational entanglement parameter
R is introduced [48, 49], which is given by (5)
A(ωs, ωi; T) = α(ωs, ωi)ΦL(ωs, ωi; T). A(ωs, ωi; T) = α(ωs, ωi)ΦL(ωs, ωi; T). (5) The spectrum of the pump is written as The spectrum of the pump is written as (6)
α(ωs, ωi) ∝exp
−
(ωs + ωi −ω0
p)2
4Bp2
, (10)
R = ωs/δωc, R = ωs/δωc, (6) (10) where ωs and δωc represent the single-particle and coin-
cidence spectral bandwidths, respectively. The coinci-
dence photon spectrum can be given by the joint spectral
density function |A(ωs, ωi)|2 at a given value of ωi, e.g.,
Sc(ωs) = |A(ωs, ωi = ω0
p/2)|2. The single-particle photon
spectrum can be determined by |A(ωs, ωi)|2 integrated over
ωi, i.e., Ss(ωs) =
dωi|A(ωs, ωi)|2. where ω0
p and Bp denote the center frequency and spectral
bandwidth of the pump. The phase-matching function satis-
fied by the SPDC photon pairs is given by ΦL(ωs, ωi; T) ≡sin[k(ωs, ωi; T)L/2]
k(ωs, ωi; T)/2
,
k(ωs, ωi; T) ≡kp(ωp; T) −ks(ωs; T) −ki(ωi; T) ± 2π/Λ. In the experiment, the pump radiation is centered at
wavelength of 780 nm and the 3-dB bandwidth is measured
to be around 0.05 nm. 3.2 Characterization of the generated photon pairs In the phase-matching condition, the
pump spectrum function and the phase-matching function
based on the wavelength of signal and idler photons are
shown in Fig. 3a, b. The joint spectrum of the photon pairs
determined by the pump spectrum function and the phase-
matching function is shown in Fig. 3c. The spectra of both
signal and idler are shown in Fig. 3d. (7) (7) where L is the crystal length, T and Λ represent the
working temperature and poled period of PPKTP,
respectively. k
is
the
wavevector
mismatching. kj(ωj; T) = n(ωj; T)ωj/c, j = p, s, i denote the propagation
constants of the pump, signal and idler respectively. Let us
consider the case of collinear degenerate down-conversion,
i.e., ω0
s = ω0
i = ω0
p/2, the deviations from the central fre-
quencies are given by Ωs,i = ωs,i −ω0
p/2. Without loss
of generality, k can be Taylor expanded to its first-order
terms According to the theoretical computation, the center
wavelength of signal and idler photon are 1560.0 and
1559.9 nm, respectively. And the bandwidth of signal
and idler photon are both 2.4 nm. The coincidence width
of the joint spectrum is 0.43 nm. According to the cri-
teria of frequency entanglement introduced in [48], the
degree of frequency entanglement was determined to be
R = s/δ c = 5.58 . The HOM visibility is estimated to
be 99.9 % with a FWHM dip width of 1.48 ps. k(ωs, ωi) ≡(k0
p(T) −k0
s (T)
−k0
i (T) ± 2π/Λ)
−(k
′
p(ω0
p; T) −k
′
s(ω0
p/2; T))Ωs
−(k
′
p(ω0
p/2; T) −k
′
s(ω0
p/2; T))Ωi, (8) 4 Experimental results Based on the Sellmeier equation of PPKTP and its temper-
ature dependence given in [44–46], the crystal temperature
can be optimized to achieve the phase-matching condition
(k0
p(T) −k0
s (T) −k0
i (T)) ± 2π/Λ = 0.i 4.1 Generation of the 780‑nm pump source By applying PDH locking to the SHG cavity, the gener-
ated second harmonic laser at 780 nm as a function of the
fundamental laser power is experimentally investigated
and shown in Fig. 4. At an incident fundamental power of
1.41 W (the power of 1560 nm laser is attenuated by the
optical components, such as EO modulator, optical isola-
tor, for about 64.2 % before it arrives at the SHG cavity), a
power as high as 742 mW of 780-nm laser is generated and
the corresponding second harmonic generation efficiency
reaches 52.6 %. The theoretical simulation is also shown in
Fig. 4a by black line. It can be seen that, the experimental p
To quantify a frequency-entangled source, two figures
of merit are used, which are the frequency indistinguish-
ability and the degree of frequency entanglement. The fre-
quency indistinguishability refers to the similarity between
the spectral distributions of the down-converted signal and
idler photons, which can be readily measured by the visi-
bility of the HOM interferometric measurement [39]. In the
theoretical form, it is given by the relative overlap function V =
dωsdωi|A(ωs, ωi)A(ωi, ωs)|
dωsdωi|A(ωs, ωi)|2
. (9) 3 1 An efficient source of frequency anti-correlated entanglement at telecom wavelength Page 5 of 8 128 Fig. 3 Figure in theory. a the pump spectrum function, b the phase-matching function, c the joint spectrum of the photon pairs, d spectra of both
signal and idler photon Fig. 3 Figure in theory. a the pump spectrum function, b the phase-matching function, c the joint spectrum of the photon pairs, d spectra of both
signal and idler photon detectors with a measurement bandwidth of 5 MHz. With
5 mW of optical power incident into each detector, the
homodyne output is then analyzed by a spectrum analyzer
(ROHDE&SCHWARZ FSH4) with a RBW of 30 kHz and
a VBW of 300 Hz. As shown in Fig. 5, it has a rather high
excess intensity noise, which remains 20 dB above the shot
noise level until the analyzing frequency of 3 MHz. result fits well with the theoretical curve when the power
of the fundamental laser is low. While further increasing
the power of the fundamental laser, the difference between
the experimental and theoretical results becomes obvious. 4.1 Generation of the 780‑nm pump source It can be explained by transform of the generated 780-nm
laser into 1560 nm when the power in the cavity is above
the threshold of the parameter down-conversion [50]. The
long-term stability of 780-nm output is monitored. The
second harmonic cavity can be locked stably for as long
as 70 h can be readily achieved and the result is given in
Fig. 4b. 4.2 Characterization of generated photon pairs
at 1560 nm With the above generated 780-nm laser as the pump and
the type-II PPKTP crystal as the SPDC crystal, the down-
converted signal and idler photons are then generated with
orthogonal polarizations. As shown in Fig. 1, after filter-
ing out the residual 780-nm laser by a serials of dichroic
mirrors and filters, the frequency-entangled photon pairs The intensity noise of the fundamental fiber laser always
has a considerable excess intensity noise [51], it may make
an effect on the generated the 780-nm laser and the prop-
erties of the subsequent frequency entanglement states. We
first measure the intensity noise spectrum of the 780-nm
laser by using a pair of home-made balanced homodyne 1 3 F. Hou et al. 128 Page 6 of 8 128 Page 6 of 8 evaluated to be 40 %. The small pump waist ωp0 = 48 µm ,
which corresponds to a focusing parameter of ξ = 1.08 ,
should be the reason why such low heralding efficiency
was achieved [52]. 4.2.1 Spectral measurement of the photon pairs
Remaining only one of the monochromators in the setup
(as shown in Fig. 1) while the other output of the FPBS
being set to the single photon detector, the singles spec-
tra of both signal and idler are shown in Fig. 6a. Through
Gaussian fitting, we achieved the center wavelengths
of the signal and idler to be 1560.23 ± 0.03 nm and
Fig. 4 a The theoretical and experimental power of second harmonic
laser at 780 nm as the function of the power of fundamental laser at
1560 nm. b The long-term stability of 780-nm output
Fig. 5 The measured noise power of the generated 780-nm source
(black line). The red line gives the relevant shot noise limit, and the
blue line is the electronics noise of the detector. The measurement
parameters: the power of 780-nm laser is 5 mW; the resolution band-
width of the spectrum analyzer is 30 kHz while the video bandwidth
is 300 Hz Fig. 5 The measured noise power of the generated 780-nm source
(black line). The red line gives the relevant shot noise limit, and the
blue line is the electronics noise of the detector. The measurement
parameters: the power of 780-nm laser is 5 mW; the resolution band-
width of the spectrum analyzer is 30 kHz while the video bandwidth
is 300 Hz evaluated to be 40 %. 4.2 Characterization of generated photon pairs
at 1560 nm The small pump waist ωp0 = 48 µm ,
which corresponds to a focusing parameter of ξ = 1.08 ,
should be the reason why such low heralding efficiency
was achieved [52]. 4.2.2 HOM interference measurement After extracting the accidental coincidences and normal-
izing, the result of the HOM interference measurement is
shown by red points in Fig. 7. To estimate the error of vis-
ibility, we measured the HOM coincidence distribution for
10 times. Out of the measurements, we got the standard
deviation of the visibility results. The HOM interference
visibility was measured to be 95 ± 3 %, with a FWHM
coherence time-width of 1.28 ± 0.02 ps. The measurement results have a very good agreement
with the theoretical simulations, which shows that the high
excess intensity noise of pump laser has no influence on the
generated frequency anti-correlated entanglement. (b) (b) 5 Conclusions In summary, we have generated a frequency anti-correlated
entangled biphoton source at 1560 nm and experimentally
characterized its quantum properties. The results shows that
such source has a frequency entanglement degree of 6.19,
while the frequency indistinguishability is determined as
around 95 ± 3 % by HOM interference measurement. The
agreement between the measurements and theory have
also shown that the high intensity noise of pump laser has
no influence on the generated frequency anti-correlated
entanglement. Acknowledgments This work is supported by the National Natural
Science Foundation of China (11174282, 91336108, 61127901), the
Youth Talent Support Plan of the Organization Department of China,
the Key Fund of the Western light Talent Cultivation Plan of the CAS,
China, the Cross and Cooperative Science and Technology Innovation
Team Project of the CAS, China. Fig. 6 The experimental results of the single photon spectra (a) and
the joint spectrum (b) Fig. 6 The experimental results of the single photon spectra (a) and
the joint spectrum (b) -3
-2
-1
0
1
2
3
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Theoretical
Experimental
Normalized coincidence counts
Relative time delay (ps)
Fig. 7 HOM interference result. Black line is the theoretical HOM
interference curve, red dots are the experimental results -3
-2
-1
0
1
2
3
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Theoretical
Experimental
Normalized coincidence counts
Relative time delay (ps) Open Access This article is distributed under the terms of the Crea-
tive Commons Attribution 4.0 International License (http://creativecom-
mons.org/licenses/by/4.0/), which permits unrestricted use, distribution,
and reproduction in any medium, provided you give appropriate credit
to the original author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made. Theoretical
Experimental 4.2.1 Spectral measurement of the photon pairs Remaining only one of the monochromators in the setup
(as shown in Fig. 1) while the other output of the FPBS
being set to the single photon detector, the singles spec-
tra of both signal and idler are shown in Fig. 6a. Through
Gaussian fitting, we achieved the center wavelengths
of the signal and idler to be 1560.23 ± 0.03 nm and
1560.04 ± 0.03 nm, while the 3-dB bandwidths of the sig-
nal and idler are both 3.22 ± 0.01 nm. We further meas-
ured the joint spectrum of the photon pairs, and the result-
ant two-photon spectral intensity is shown in Fig. 6b. The
color bar on the right hand of the figure shows the cor-
responding coincidence counts for different colors. We
can see that the joint spectrum has an anti-diagonal dis-
tribution. The coincidence width of the joint spectrum
is δ c = 0.52 ± 0.01 nm. According to Ref. [48], the
degree of frequency entanglement was determined to be
R = s/δ c = 6.19. The discrepancy between the experi-
mental result and the theoretical result is mainly caused by
the resolution bandwidth of the monochromators. It gener-
ated a broaden in the bandwidth of signal and idler pho-
ton spectrum as well as the coincidence width of the joint
spectrum. Fig. 4 a The theoretical and experimental power of second harmonic
laser at 780 nm as the function of the power of fundamental laser at
1560 nm. b The long-term stability of 780-nm output are coupled into a FPBS. The fiber coupling efficiency is
measured to be 74 %, by linking the two output ports of
FPBS directly to two single photon detectors and subse-
quent coincidence measurement device, both the single
photon count rate and the coincidence rate are measured. In our experiment, the two APD are used in the gated mode
triggered by an external 75-MHz square-wave signal from
a waveform generator (Tektronix AFG3252), the detection
efficiency is 20 %. The rate of the singles were then meas-
ured to be 370 kHz, while the coincidence rate was around
22 kHz. The heralding efficiency µs(i) for the signal (idler)
is defined to be the ratio between coincidences C to the
number of singles (detector noise subtracted) in the idler
(signal) detector Si(s) corrected by the detection efficiency
ηs(i). 4.2.1 Spectral measurement of the photon pairs In our experiment, the heralding efficiency of µs(i) is 1 3 3 An efficient source of frequency anti-correlated entanglement at telecom wavelength Page 7 of 8 128 Page 7 of 8 128 1554
1556
1558
1560
1562
1564
1566
0.2
0.4
0.6
0.8
1.0
signal results
idler results
Gauss ft of signal
Gauss ft of idler
Normalized coincidence count
Wavelength (nm)
(a)
(b)
Fig. 6 The experimental results of the single photon spectra (a) and
the joint spectrum (b)
0.8
1.0
1.2
Theoretical
Experimental
nce counts 1554
1556
1558
1560
1562
1564
1566
0.2
0.4
0.6
0.8
1.0
signal results
idler results
Gauss ft of signal
Gauss ft of idler
Normalized coincidence count
Wavelength (nm)
(a)
(b) 1. D.N. Klyshko, Photons and Nonlinear Optics (Gordon and
Breach, New York, 1988)
2. A. Zeilinger, Rev. Mod. Phys. 71, S288 (1999)
3. R. Horodecki, P. Horodecki, M. Horodecki, K. Horodecki, Rev.
Mod. Phys. 81, 865 (2009)
4. D. Bouwmeester, A. Ekert, A. Zeilinger, The Physics of Quantum
Information (Springer, New York, 2000)
5. V. Giovannetti, S. Lloyd, L. Maccone, Nature 412, 417 (2001)
6. V. Giovannetti, S. Lloyd, L. Maccone, Phys. Rev. Lett. 87,
117902 (2001) References 1. D.N. Klyshko, Photons and Nonlinear Optics (Gordon and
Breach, New York, 1988)
2. A. Zeilinger, Rev. Mod. Phys. 71, S288 (1999)
3. R. Horodecki, P. Horodecki, M. Horodecki, K. Horodecki, Rev. Mod. Phys. 81, 865 (2009)
4. D. Bouwmeester, A. Ekert, A. Zeilinger, The Physics of Quantum
Information (Springer, New York, 2000)
5. V. Giovannetti, S. Lloyd, L. Maccone, Nature 412, 417 (2001)
6. V. Giovannetti, S. Lloyd, L. Maccone, Phys. Rev. Lett. 87,
117902 (2001) 1. D.N. Klyshko, Photons and Nonlinear Optics (Gordon and
Breach, New York, 1988) 2. A. Zeilinger, Rev. Mod. Phys. 71, S288 (1999) Fig. 7 HOM interference result. Black line is the theoretical HOM
interference curve, red dots are the experimental results 1 3 128 Page 8 of 8 F. Hou et al. 7. V. Giovannetti, S. Lloyd, L. Maccone, Phys. Rev. A 65, 022309
(2002) 31. R. Quan, M. Wang, F. Hou, Z. Tai, T. Liu, S. Zhang, Appl. Phys. B: Lasers Opt. 118, 431 (2015) (
)
8. T.B. Bahder, W.M. Golding, AIP Conf. Proc. 734, 395 (2004) 32. R.J. Thompson, M. Tu, D.C. Aveline, N. Lundblad, L. Maleki,
Opt. Express 11, 1709–1713 (2003) 9. A. Valencia, G. Scarcelli, Y. Shih, Appl. Phys. Lett. 85, 2635
(2004) 33. C. Yang, S. Xu, S. Mo, C. Li, Z. Feng, D. Chen, Z. Yang, Z. Jiang, Opt. Express 21, 12546–12551 (2013) (
)
10. V. Giovannetti, S. Lloyd, L. Maccone, Science 306, 1330 (2004) 34. X. Tian, Y. Li, Q. Liu, K. Zhang, Chin. Phys. B 18, 2324–04
(2009) 11. V. Giovannetti, S. Lloyd, L. Maccone, J.H. Shapiro, F.N.C. Wong, Phys. Rev. A 70, 043808 (2004) 35. R.W.P. Drever, J.L. Hall, F.V. Kowalski et al., Appl. Phys. B 31,
97–105 (1983) g
y
12. F. Hou, R. Dong, R. Quan, Y. Zhang, Y. Bai, T. Liu, S. Zhang, T. Zhang, Adv. Space Res. 50, 1489 (2012) g
p
13. C. Ho, A. Lamas-Linares, C. Kurtsiefer, New J. Phys. 11, 045011
(2009) 36. Y.H. Kim, W.P. Grice, Opt. Lett. 30, 908 (2005) 37. C.K. Hong, Z.Y. Ou, L. Mandle, Phys. Rev. Lett. 59 (
)
14. J. Wang, Z. Tian, J. Jing, H. Fan, 1501. arXiv:01478v2 (2015) 38. W.J. Kozlovsky, C.D. Nabors, R.L. Byer, IEEE J. Q. Electron. 24, 913–919 (1988) 15. A. Yabushita, T. Kobayashi, Phys. Rev. A 69, 013806 (2004) 39. W.P. Grice, I.A. Walmsley, Phys. Rev. A 56, 1627–1634 (1997) 16. A.F. References Abouraddy, M.B. Nasr, B.E.A. Saleh, A.V. Sergienko, M.C. Teich, Phys. Rev. A 65, 053817 (2002) 40. V. Giovannetti, L. Maccone, J.H. Shapiro, F.N.C. Wong, Phys. Rev. A 66, 043813 (2002) y
17. A.V. Sergienko, B.E.A. Saleh, M.C. Teich, Opt. Lett. 29, 2429
(2004) 41. G. Fujii, N. Namekata, M. Motoya, S. Kurimura, S. Inoue, Opt. Express 15, 12769 (2007) (
)
18. L. Dorilian, L. Novotny, Opt. Lett. 37, 4077 (2012) 42. L. Mandel, E. Wolf, Optical Coherence and Quantum Optics
(Cambridge University Press, Cambridge, 1995). (Chap. 22.4) 19. M.B. Nasr, B.E.A. Saleh, A.V. Sergienko, M.C. Teich, Phys. Rev. Lett. 91, 083601 (2003) 20. M.C. Booth, B.E.A. Saleh, M.C. Teich, Opt. Commun. 284,
2542 (2011) 43. R. Erdmann, D. Branning, W. Grice, I.A. Walmsley, Phys. Rev. A
62, 053810 (2000) 21. T.B. Pittman, Y.H. Shih, D.V. Strekalov, A.V. Sergienko, Phys. Rev. A 52, R3429 (1995) 44. K. Fradkin, A. Arie, A. Skliar, G. Rosenman, Appl. Phys. Lett. 74, 914 (1999) 45. T.Y. Fan, C.E. Huang, B.Q. Hu, R.C. Eckardt, Y.X. Fan, R.L. Byer, R.S. Feigelson, Appl. Opt. 26, 2390 (1987) 22. P. Zerom, K. Wai, C. Chan, J.C. Howell, R.W. Boyd, Phys. Rev. A 84, 061804(R) (2011) 23. S. Jeff, M. Michiel, Rev. Sci. Instrum. 72, 2855 (2001 46. S. Emanueli, A. Arie, Appl. Opt. 42, 6661 (2003) 23. S. Jeff, M. Michiel, Rev. Sci. Instrum. 72, 2855 (2001) ,
,
pp
p
,
(
)
47. C.K. Law, J.H. Eberly, Phys. Rev. Lett. 92, 127903 (2004) 24. S. Brasselet, V.L. Floc’h, F. Treussart, J.F. Roch, J. Zyss, E. Bo
zung-Appert, A. Ibanez, Phys. Rev. Lett. 92, 207401 (2003) 47. C.K. Law, J.H. Eberly, Phys. Rev. Lett. 92, 127903 24. S. Brasselet, V.L. Floc’h, F. Treussart, J.F. Roch, J. Z 48. Y.M. Mikhailova, P.A. Volkov, M.V. Fedorov, Phys. Rev. A. 78,
062327 (2008) 25. B. Dayan, A. Peer, A.A. Friesem, Y. Silberberg, Phys. Rev. Lett. 93, 023005 (2004) 49. M.V. Fedorov, M.A. Efremov, P.A. Volkov, J.H. Eberly, Phys. B:
At. Mol. Opt. Phys. 39, S467 (2006) 26. L. Zhang, C. Silberhorn, I.A. Walmsley, Phys. Rev. Lett. 100,
110504 (2008) 50. A.G. White, P.K. Lam, M.S. Taubman, M.A.M. Marte, S. Schil-
ler, D.E. McClelland, H.-A. Bachor, Phys. Rev. A. 55, 4511
(1997) 27. J. Nunn, L.J. Wright, C. Söller, L. Zhang, I.A. Walmsley, B.J. Smith, Opt. Express 21, 15971 (2013) 28. O. Kuzucu, M. Fiorentino, M.A. Albota, F.C. Wong, F.X. References Kart-
ner, Phys. Rev. Lett. 94, 083601 (2005) 51. E. Desurvire, J.R. Simpson, J. Lightwave Technol. 7(5), 835–845
(1989) 29. M. Avenhaus, M.V. Chekhova, L.A. Krivitsky, G. Leuchs, C. Sil-
berhorn, Phys. Rev. A 79, 043836 (2009) 52. T. Guerreiro, A. Martin, B. Sanguinetti, N. Bruno, H. Zbinden,
R.T. Thew, Opt. Express 21, 27641–27651 (2013) 30. Y. Zhang, F. Hou, Y. Bai, T. Liu, S. Zhang, R. Dong, Acta Phys. Sin. 62, 144206 (2013) 1 3 3
|
https://openalex.org/W2526541291
|
https://europepmc.org/articles/pmc5041572?pdf=render
|
English
| null |
Gender-modulated risk of coronary heart disease, diabetes and coronary mortality among Turks for three major risk factors, and residual adiposity risk
|
BMC endocrine disorders
| 2,016
|
cc-by
| 7,615
|
© 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Can et al. BMC Endocrine Disorders (2016) 16:54
DOI 10.1186/s12902-016-0134-6 Can et al. BMC Endocrine Disorders (2016) 16:54
DOI 10.1186/s12902-016-0134-6 Abstract Background: We determined the proportion of the effects of body mass index (BMI) or its categories on
cardiometabolic outcomes mediated through systolic blood pressure (SBP), total cholesterol and fasting glucose. Methods: Cox regression analyses were performed for incident outcomes among Turkish Adult Risk Factor study
participants in whom the three mediators had been determined (n = 2158, age 48.5 ± 11 years). Over a mean 10.2-years’
follow-up, new coronary heart disease (CHD) developed in 406, diabetes in 284 individuals, and 149 CHD deaths
occurred. Results: Hazard ratios (HR) of BMI for incident diabetes were no more than marginally attenuated by the 3 mediators
including glucose, irrespective of gender. Compared to “normal-weight”, sex- and age-adjusted RRs for incident CHD of
overweight and obesity were 1.40 and 2.24 (95 % CI 1.68; 2.99), respectively, in gender combined. Only three-tenths of
the excess risk was retained by BMI in men, six-tenths in women. No mediation of glycemia was discerned in males, in
contrast to greatest mediation in females. HR of age-adjusted continuous BMI was a significant but modest contributor
to CHD mortality in each gender. While the BMI risk of CHD death was abolished by mediation of SBP in men, HR
strengthened to over two-fold in women through mediation of fasting glucose. Conclusions: Mediation of adiposity by 3 traditional factors exhibited among Turkish adults strong gender dependence
regarding its magnitude for CHD risk and the mediation by individual risk factors. Retention of the large part of risk for
diabetes in each sex and for CHD in women likely reflects underlying autoimmune activation. Keywords: Autoimmune activation, Body mass index, Coronary heart disease, Coronary mortality, Type-2 diabetes,
Gender difference Keywords: Autoimmune activation, Body mass index, Coronary heart disease, Coronary mortality, Type-2 diabetes,
Gender difference Gender-modulated risk of coronary heart
disease, diabetes and coronary mortality
among Turks for three major risk factors,
and residual adiposity risk Günay Can1*, Altan Onat2, Eray Yurtseven1, Yusuf Karadeniz3, Tuğba Akbaş-Şimşek4, Ayşem Kaya5
and Hüsniye Yüksel2 Günay Can1*, Altan Onat2, Eray Yurtseven1, Yusuf Karadeniz3, Tuğba Akbaş-Şimşek4, Ayşem Kay
and Hüsniye Yüksel2 * Correspondence: gunaycan09@yahoo.fr
1Departments of Public Health, Istanbul University, Yazıcı sok. 18/5,
Kocamustafapaşa, 34098 Istanbul, Turkey
Full list of author information is available at the end of the article Background across races and by gender. In order to appropriately
target the individual risk factors, the effect size of the
mediating factors and the excess residual risk of adiposity
are highly relevant from the public health viewpoint. Rise in the proportion of overweight and obese individuals
represents the most prominent recent cardiovascular risk
factor in Turkey and worldwide. Many studies have shown
that cardiovascular risk of excess adiposity is mediated
mainly by traditional risk factors such as systolic blood
pressure (SBP), levels of total cholesterol and glucose, or
diabetes [1–4]. The magnitude of the mediation may vary The role of BMI in cardiovascular disease in the Asia-
Pacific region has been studied in a meta-analysis [5],
and the association of overweight with increased coron-
ary heart disease (CHD) risk independent of BP and
cholesterol level has been investigated in another meta-
analysis [6]. The most extensive quantification of the
effects of high BMI, and of overweight and obesity on
CHD mediated through SBP, serum cholesterol, and * Correspondence: gunaycan09@yahoo.fr
1Departments of Public Health, Istanbul University, Yazıcı sok. 18/5,
Kocamustafapaşa, 34098 Istanbul, Turkey
Full list of author information is available at the end of the article Can et al. BMC Endocrine Disorders (2016) 16:54 Page 2 of 9 glucose, has been in the Global Burden of Disease
(GBD) study [7], analyzing data from 97 prospective
cohort studies. The HR for each 5 kg/m2 higher BMI
was 1.27 for CHD, after adjustment for confounders. Additional adjustment for the three metabolic risk fac-
tors reduced the HRs to 1.15 indicating that 46 % of the
excess risk of BMI for CHD was mediated by these fac-
tors. SBP was the most important mediator, accounting
for 31 % of the excess risk on CHD. Compared with nor-
mal weight, both overweight and obesity were associated
with significantly increased CHD risk, with 50 % of the
excess risk of overweight and 44 % of the excess risk of
obesity for CHD being mediated by the selected 3 medi-
ators. Authors concluded that, though nearly one half of
excess risk for CHD due to high BMI was mediated
through 3 metabolic risk factors, maintaining optimal
body weight was needed for the full benefits. excluded. The study was approved by the Ethics Commit-
tee of the Istanbul University Medical Faculty. Written
informed consent for participation was obtained from all. Measurement of risk factors BP was measured with an aneroid sphygmomanometer
(Erka, Bad Tölz, Germany) in the sitting position on the
right arm, and the mean of two recordings 5 min apart
was recorded. Height was measured without shoes using
a measuring stick and weight without shoes in light in-
door clothes using a scale. Waist circumference was
measured at the level midway between the lower rib
margin and the iliac crest. Cigarette smoking status was
categorized into never, former and current smokers. Anyone who reported the use of alcoholic beverages
once a week or more was considered as alcohol user. Physical activity was graded by the participant himself
into four categories of increasing order with the aid of a
scheme [9]. The share of Middle Eastern populations in the GBD
meta-analysis comprised less than 0.5 % of the total par-
ticipants and hardly 1 % of the CHD events. Of the
57161 CHD events, 88.6 % were recorded in Western
cohorts. Therein elicited major findings might have lim-
ited applicability for Turkish adults, since differences
may exist in how BMI affects the mediating metabolic
risks in diverse populations. Moreover, sex-stratified
analyses were not provided, an aspect of critical import-
ance among Turks [8]. Thus we undertook the task of
clarifying the effects of the mediating risk factors for
BMI on CHD risk among Turks. Our aim was to deter-
mine, separately in the sexes, the residual risk of inci-
dent nonfatal and fatal CHD of overweight and obesity
after excluding the effects of mediation of blood pres-
sure (BP), glucose and total cholesterol levels. Finally,
the “net” BMI effect was examined regarding incident
diabetes as well. Blood samples were collected, spun at 1000 g, shipped
to Istanbul and stored in deep-freeze at −75 °C, until ana-
lyzed within weeks. Serum concentrations of glucose, total
cholesterol, fasting triglycerides, low-density lipoprotein
(LDL)- and high-density lipoprotein (HDL)-cholesterol
were determined using Cobas 500 autoanalyzer (Roche
Diagnostic GmbH, Germany). Serum concentrations of
apolipoprotein (apo) B and apoA-I were measured nephe-
lometrically by BN ProSpec analyzer (Siemens Healthcare
Diagnostics, Germany). Definitions Participants were classified into three categories: BMI
≤25 kg/m2, overweight (BMI 25.0–29.9 kg/m2), obesity
(BMI ≥30.0 kg/m2). Hypertension was defined as a blood
pressure ≥140 mmHg and/or ≥90 mmHg, and/or use of
antihypertensive medication. Individuals with metabolic
syndrome (MetS) were identified when 3 out of the 5
criteria of the National Cholesterol Education Program
(ATP III) [10] were met, modified for prediabetes (fast-
ing glucose 100–125 mg/dl [11] and further for abdom-
inal obesity using as cutpoint ≥95 cm in men, as
assessed in the Turkish Adult Risk Factor study [12]. Diabetes was diagnosed with the criteria of the American
Diabetes Association, namely by self-report of antidiabetic
medication, or a plasma fasting glucose ≥126 mg/dl [13]. Background Data were obtained by history of the past years via a ques-
tionnaire, physical examination of the cardiovascular sys-
tem, sampling of blood and recording a resting 12-lead
electrocardiogram. Data analysis Descriptive parameters were shown as mean ± standard
deviation or in percentages. For several other variables
with skewed distribution, values derived from log-
transformed (geometric) means were used. An SD value
of (for example) 1.7 indicates the factor needed to multi-
ply or divide the mean value to obtain the limits of the
SD. ANOVA analyses and pairwise comparisons with
post hoc Tukey HSD were made to detect significance
across more than two groups; two-sided t-tests and
Pearson’s chi-square tests were used to analyze the dif-
ferences between means and proportions of two groups. g
Table 2 shows the prediction of incident CHD by BMI
categories in Cox regression, by gender and adjusted by
the three mediating factors in 3 models. In the model
adjusted only for sex, age and smoking status, over-
weight disclosed an RR of 1.40 compared to normal-
weight, whereas obesity had an HR 2.24 (95 % CI 1.68;
2.99) in gender combined. The addition of systolic BP,
fasting glucose plus LDL- and HDL-cholesterol attenu-
ated the HR of overweight to non-significance, but obes-
ity retained strong significance (RR 1.56). This showed
gender disparity inasmuch as while obesity in men was
no longer significant, it was significant in women with
an RR 1.90 (95 % CI 1.20; 3.02), retaining 61 % of the
excess CHD risk independent of fasting glucose, systolic
BP and total cholesterol. Cox proportional hazard regression analyses of the
BMI categories for outcomes were performed in models
that adjusted basically for (sex and) age and smoking
status. Participants with outcome conditions at baseline
were excluded from regression analyses. The relative risk
(RR) for the three adiposity categories and the hazard
ratio (HR)s for the mediating factors (systolic BP, total
cholesterol and fasting glucose) were estimated and
expressed for each in terms of 1-SD increment. We
adjusted for sex, age, smoking status, and additionally
for the 3 mediators in regard to analyses involving WC. Residual adiposity risk was estimated as percentage of
excess risk mediated by the 3 risk factors deducted from
the excess risk of the BMI category, which we expressed
also as percentage of excess risk mediated [7, 15]. A
value of p < 0.05 on the two-tail test was considered sta-
tistically significant. Statistical analyses were performed
using SPSS-10 for Windows. Study sample The TARF study is a prospective cohort study on the
prevalence of cardiac disease and risk factors in adults in
Turkey carried out biennially since 1990 in 59 communi-
ties scattered throughout all geographical regions of the
country [9]. It comprises a random sample of the Turkish
adult population, representatively stratified for sex, age,
geographical regions and for rural–urban distribution. New random recruitment forming 15 % of the study sam-
ple was made in 2002/’03. Participants numbering 2287 –
of whom 129 had CHD at baseline– composed the cohort
of the current study. BMI, BP and fasting glucose determi-
nations were available in the surveys encompassing the
10 years from 1997/’98 to 2007. Seventy-three % of partic-
ipants had a baseline in the survey 1997/’98. The sample
did not contain individuals who had been deceased prior
to baseline, and 390 participants with no follow-up were Diagnosis of CHD was based on the presence of angina
pectoris, of a history of myocardial infarction with or
without accompanying Minnesota codes of the ECG [14]
or on a history of myocardial revascularization. Typical
angina and, in women, age >45 years were prerequisite Page 3 of 9 Can et al. BMC Endocrine Disorders (2016) 16:54 Can et al. BMC Endocrine Disorders (2016) 16:54 Page 3 of 9 for a diagnosis when angina was isolated. ECG changes
of “ischemic type” of greater than minor degree (Codes
1.1–2, 4.1–2, 5.1–2, 7.1) were considered as myocardial
infarct sequelae or myocardial ischemia, respectively. CHD death comprised death from heart failure of coron-
ary origin or fatal coronary event. Death was ascertained
via information from first-degree relatives, records of
local health personnel and/or the nation-wide Identity
Participation System. Baseline characteristics of the sample free of CHD are
presented in Table 1, stratified to BMI categories and
gender. As also seen in Fig. 1, obesity existed only in
19 % of men, but in 44.4 % of women –an odds of 2.3-
fold. Among the 18 variables examined, all but apoA-I
and in men alcohol usage, were significantly different
across the BMI categories. Though HDL-cholesterol was
significantly lower in obesity than in normal-weight in
each gender, apoA-I concentrations were similar. Never
smokers
were
significantly
more
prevalent,
current
smokers less prevalent in obesity than in normal-weight,
regardless of sex. Pulse pressure was higher in obese
women than obese men (55.3 vs. 50.6 mmHg) Data analysis Substituting abdominal obesity for the two BMI cat-
egories (Additional file 1: Table S1) replicated the
changes in magnitude and sex difference stated for over-
all adiposity. Whereas only 48 % of the CHD risk was
retained by abdominal obesity in men, attenuating the
RR to borderline significance, 64 % of the risk was
retained in women, rendering persistence of independent
significance in RR by abdominal obesity, Table 3 shows prediction of CHD mortality and inci-
dent diabetes by continuous BMI values in each gender,
adjusted for SBP, total cholesterol and fasting glucose. For incident diabetes, sex- and age-adjusted BMI was as-
sociated with an HR 1.73 per 1 SD increment. When all
3 mediators were added to the model, the HR of BMI
was reduced merely as little as less than 3 %: to 2.05 in
men and to 1.52 in women. Results The study sample consisted of 2287 participants (of
whom 1199 female) at age 48.5 ± 11 years at baseline. Mean age was similar in the sexes. Of the 616 subjects
in the category BMI ≤25 kg/m2, only 67 had BMI
<20 kg/m2. Follow-up ranged from 2 to 16 years with a
mean of 10.2 ± 4.6 years. Total follow-up amounted to
21,980 person-years. Incident CHD manifested in 406,
incident diabetes in 284 persons, and CHD mortality
was identified in 149 (6.9 %) individuals (yielding a CHD
mortality rate of 6.56 [8.0 for men, 5.0 for women] per
1000 person-years). CHD mortality was, however, determined only by a
significant HR 1.12 by the sex- and age-adjusted BMI. This effect was fully mediated in men by total choles-
terol and glycemia and was more than offset by SBP. Women, in contrast, disclosed an inverse mediation ef-
fect by the 3 risk factors, so that HR of BMI rose to 1.26
(95 % CI 1.00; 1.60). Replacing in model 3, Table 3, HDL-C with apoA-I
(values were missing in 18 % of the sample) showed ten-
dency to protect against CHD in women (p = 0.062) along Can et al. BMC Endocrine Disorders (2016) 16:54 Page 4 of 9 with mild attenuation of the RR of obesity (p = 0.053),
while apoA-I was not protective in men (p = 0.48), age,
Discussion
This population-based prospective study on middle-a
Table 1 Baseline characteristics of the study sample stratified to BMI categories and gender (n = 2158)
BMI categories
ANOVA p-v
BMI ≤25
Overweight
Obesity
n
Mean
SD
Mean
SD
Mean
SD
Men
1015
n = 350
n = 473
n = 192
Age, years
1015
47,3
12,9
48,9
11,3
50,5
11,3
,008
Phys. activity gr., I-IV
1013
2,59
1,07
2,40
1,02
2,31
1,00
,004
Waist circumf., cm
1012
83,5
7,1
96,3
6,0
106,8
8,7
<0,001
Systolic BP, mmHg
1015
117,5
17,5
128,0
19,7
139,8
24,9
<0,001
Diastolic BP, mmHg
1015
75,0
10,5
82,1
11,4
89,2
14,6
<0,001
T. cholesterol, mg/dl
1015
173,8
35,5
185,4
36,6
194,8
38,8
<0,001
HDL cholest, mg/dl
1011
41,1
13,1
36,3
10,6
35,5
10,0
<0,001
LDL-cholest, mg/dl
994
108,8
31,0
114,8
31,4
119,3
31,7
,001
Triglycerides, mg/dl
1012
121,7
70
172,0
100,6
194,6
112
<0,001
Apolipoprot.A-I, mg/dl
838
131,9
28
128,7
24,1
131
27,6
,262
Apolipoprot. B, mg/dl
824
103,8
31,2
110,4
33,8
111,5
30,9
,015
Fast. Results glucose, mg/dl
1015
94,7
24,3
97,9
29,8
102,8
31,2
,007
B mass index, kg/m2
1015
22,5
1,9
27,4
1,35
32,6
2,5
<0,001
Diabetes, yes n, %
1015
10
2,9
25
5,3
19
9,9
0,002
MetS, yes n, %
1015
41
11,7
260
55,0
142
74,0
<0,001
Hypertension, n, %
1015
54
15,4
171
36,2
109
56,8
<0,001
Current, former smoker, n
1011
216; 55
62.2; 15.9
218; 111
46.1; 23.5
81; 47
42.4; 24.6
<0,001
Alcohol usage, yes, n, %
1008
58
16,8
91
19,3
28
14,7
0,332
Women
1143
234
402
507
Age, years
1143
45,0
12,5
47,6
12,1
50,4
10,6
<0,001
Phys. activity gr., I-IV
1135
2,24
,71
2,18
,67
2,08
,67
,010
Waist circumf., cm
1137
77,3
7,7
88,2
7,9
100,5
9,4
<0,001
Systolic BP, mmHg
1143
118,9
20,7
129,8
24
143,0
26,7
<0,001
Diastolic BP, mmHg
1143
75,0
11,2
82,2
12,7
88,7
14,7
<0,001
T. cholesterol, mg/dl
1143
176,9
37,4
190,7
37,2
198,2
38,7
<0,001
HDL cholest, mg/dl
1139
47,9
13,7
44,7
11,7
43,8
12, 6
<0,001
LDL-cholest, mg/dl
1115
108
32,8
120,
31,3
123,4
34,4
<0,001
Triglycerides, mg/dl
1143
110,7
80,1
131,7
78,9
155,4
88,8
<0,001
Apolipoprot.A-I, mg/dl
937
144,1
31,9
142,6
28,3
144,3
28,8
,692
Apolipoprot. B, mg/dl
944
103,7
30,7
110,4
37,3
116,3
45,0
,002
Fast. glucose, mg/dl
1143
96,3
26,7
98
23,9
102,6
30,4
,005
B mass index, kg/m2
1143
22,6
1,89
27,6
1,4
34,5
3,8
<0,001
Diabetes, yes n, %
1139
7
3,0
15
3,7
53
7
<0,001
MetS, yes n, %
1143
35
15,0
164
40,8
373
73,6
<0,001
Hypertension, n, %
1143
43
18,4
145
36,1
301
59,4
<0,001
Current, former smoker, n, %
1142
77; 9
32,9; 3,8
83; 23
20,6; 5.7
51; 20
10.1; 4.0
<0,001
Alcohol usage, yes, n, %
1138
7
3,0
2
,5
0
0
<0,001 Table 1 Baseline characteristics of the study sample stratified to BMI categories and gender (n = 2158) with mild attenuation of the RR of obesity (p = 0.053),
while apoA-I was not protective in men (p = 0.48), age,
SBP and total cholesterol retaining their significant pre-
dictive abilities. with mild attenuation of the RR of obesity (p = 0.053),
while apoA-I was not protective in men (p = 0.48), age,
SBP and total cholesterol retaining their significant pre-
dictive abilities. Diabetes risk is overwhelmingly determined by BMI and
little by glucose Age-standardized adult diabetes prevalence globally rose
since 1980 more rapidly in women than men [18]. The
development of diabetes is recognized to be largely co-
determined by BMI with which our findings concur. Noteworthy is that added adjustment for fasting glucose
hardly attenuated the HR of BMI for diabetes in either
or both sexes. This may be related to lipoprotein[Lp}(a)-
activated autoimmunity [19] which determines glycemia
and concomitantly mediates rise in BMI. In the meta-
analysis by Singh et al. [20] RRs for 5 kg/m2 higher BMI
for ages 55–64 was 2.32 (2.04–2.63) for diabetes and
ranged for all adult ages from 3 to 1.4. RRs for the esti-
mated effect of BMI were larger in Western cohorts as
compared with Asian cohorts in adults <55 years old. Of
interest was that the effect of BMI on incident diabetes
in Turkish women was only half that found in men. This
may be related to the gender difference existing on
Lp(a)-induced autoimmunity [19]. Of further interest
was that the mediation of systolic BP and total choles-
terol (alike of glucose) was virtually negligible. Fig. 1 Depicts percentages of excess risks of obesity in men and women
for incident coronary heart disease (CHD), type-2 diabetes and overall
mortality. CHD risk is indicated separately also for overweight. It is evident
that obesity in females, as distinct from males, retains substantially greater
risk for CHD mortality and CHD, unmediated by the studied 3 traditional
risk factors findings dependent on type of cardiometabolic condition
and gender. A) BMI determined the development of type-
2 diabetes at a 1.5- to 2-fold HR, more so in men than
women, and was virtually not mediated by a combination
of SBP, total cholesterol and glucose levels. B) Regarding
incident CHD, overweight displayed a non-significant
tendency to confer risk that was little modified in either
gender by the 3 mediators. C) Compared with normal-
weight, obesity imparted a significant ~2-fold CHD risk. In men, ¾ of this excess risk was mediated especially by
SBP, and by total cholesterol and glucose. In women, in
contrast, 60 % of the excess risk was retained by BMI,
leaving the remainder to the mediation of the major risk
factors. D) BMI, though a significant modest predictor of
CHD mortality, was mediated by the conventional risk
factors in males, but retained substantially greater inde-
pendent risk in females. Discussion This population-based prospective study on middle-aged
Turkish adults seeking the direct and mediated effect of
adiposity on cardiometabolic risk demonstrated following Can et al. BMC Endocrine Disorders (2016) 16:54 Page 5 of 9 Page 5 of 9 Page 5 of 9 Fig. 1 Depicts percentages of excess risks of obesity in men and women
for incident coronary heart disease (CHD), type-2 diabetes and overall
mortality. CHD risk is indicated separately also for overweight. It is evident
that obesity in females, as distinct from males, retains substantially greater
risk for CHD mortality and CHD, unmediated by the studied 3 traditional
risk factors Resistin and tumor necrosis factor TNF-α are implicated
in inducing atherogenic adipokines, such as plas-
minogen activating inhibitor-1 and interleukin-6, and
inhibiting adiponectin. TNF-α activates also nuclear
factor-(NF-)kB which may mediate hypertension and
endothelial dysfunction. In obese patients, macrophage
and lymphocyte infiltration in adipose tissue may strongly
contribute to obesity-related metabolic dysfunction and
chronic inflammation [17]. Diabetes risk is overwhelmingly determined by BMI and
little by glucose These findings, collectively, sug-
gest that other determinant factors are mediated by BMI/
obesity for the risk of diabetes and CHD and, in women,
for the risk of CHD by obesity. CHD risk of obesity in
men was largely conferred by SBP-total cholesterol, and
little by obesity. CHD risk of overweight little modified by the three
mediators Elevated peripheral vascular resistance and renal salt
retention due to higher sympathetic nervous system ac-
tivity, angiotensin-aldosterone activity and insulin levels
[21] can lead to hypertension in people with adiposity
which leads also to dyslipidemia. Moreover, enhanced
low-grade inflammation may render insulin resistance
and diabetes [22]. We concur with the global meta-
analysis [7] that the association between adiposity and
CHD is not completely explained by the three mediators
in men and underline that it is far from being explained
in women. Noteworthy is that the overall obesity category was pre-
dominated by females whereas the referent “normal-weight”
category was so by males (Additional file 2: Figure S1). Overweight imparted modest age-adjusted CHD risk
(1.33-fold the “normal” weight); and this was attenuated
mainly by systolic BP. The attenuation via BP and serum
total cholesterol was 45 % in the meta-analysis of the
BMI-CHD Collaborators [6] and was one-half (of the un-
adjusted RR 1.26) via the three mediating risk factors in
the meta-analysis of the Chronic Diseases Collaboration Adipose tissue, particularly tissue from visceral-fat de-
posits, secretes potential mediators in the development
of chronic diseases. Obesity is characterized by abnormal
adipokine production and the activation of several proin-
flammatory signaling pathways, resulting in the induc-
tion of several biological markers of inflammation [16]. Can et al. BMC Endocrine Disorders (2016) 16:54 Page 6 of 9 Table 2 Cox regression models for the prediction of incident CHD by BMI categories and mediators, by gender
Total
Men
Women
HR
95 % CI
HR
95 % CI
HR
95 % CI
404/2153a
185/1011a
219/1142a
Sex, female
0.93
0.73; 1.18
Age, 11 years
1.78
1.64; 1.96
1.82
1.59; 2.08
1.76
1.56; 2.00
Current vs. never smoking
1.60
1.24; 2.07
1.58
1.11; 2.25
1.65
1.13; 2.42
Former vs. never smoking
1.28
0.91; 1.79
1.25
0.83; 1. CHD risk of overweight little modified by the three
mediators 09
1.32
0.67; 2.58
Overweight
1.40
1.05; 1.85
1.44
1.01; 2.05
1.40
0.87; 2.24
Obesity
2.24
1.68; 2.99
2.00
1.34; 3.00
2.48
1.59; 3.87
Model 1
Sex, female
0.83
0.68; 1.02
Age, 11 years
1.57
1.43; 1.73
1.56
1.34; 1.78
1.57
1.37; 1.78
Overweight
1.18
0.89; 1.57
1.17
0.82; 1.67
1.19
0.74; 1.92
Obesity
1.61
1.20; 2.17
1.37
0.89; 2.10
1.79
1.14; 2.80
Systolic BP, 25 mmHg
1.37
1.24; 1.51
1.52
1.27; 1.82
1.31
1.17; 1.48
Model 2
Sex, female
0.87
0.71; 1.07
Age, 11 years
1.64
1.51; 1.80
1.71
1.49; 1.96
1.61
1.40; 1.84
Overweight
1.26
0.95; 1.67
1.31
0.92; 1.86
1.27
0.79; 2.03
Obesity
1.87
1.41; 2.50
1.69
1.12; 2.55
2.10
1.35; 3.25
Total cholesterol, 40 mg/dl
1.23
1.11; 1.37
1.25
1.07; 1.45
1.23
1.07; 1.42
Fasting glucose, 25 mg/dl
1.12
1.04; 1.20
1.02
0.92; 1.13
1.20
1.05; 1.32
Model 3
399/2098a
183/987a
216/1111a
Sex, female
0.92
0.72; 1.18
Age, 11 years
1.62
1.45; 1.78
1.61
1.39; 1.86
1.61
1.38; 1.84
Current vs. never smoking
1.67
1.28; 2.17
1.68
1.17; 2.42
1.61
1.09; 2.38
Former vs. never smoking
1.30
0.93; 1.83
1.23
0.82; 1.72
1.47
0.75; 2.90
Overweight
1.19
0.90; 1.59
1.19
0.82; 1.72
1.31
0.81; 2.11
Obesity
1.56
1.15; 2.12
1.30
0.83; 2.04
1.90
1.20; 3.02
LDL-cholesterol, 32 mg/dl
1.14
1.04; 1.26
1.13
0.98; 1.31
1.16
1.02; 1.32
HDL-cholesterol, 12 mg/dl
0.89
0.80; 0.99
0.96
0.82; 1.12
0.83
0.72; 0.96
Fasting glucose, 25 mg/dl
1.12
1.04; 1.21
1.05
0.92; 1.19
1.20
1.08; 1.32
Systolic BP, 25 mmHg
1.38
1.25; 1.52
1.52
1.27; 1.82
1.32
1.17; 1.49
Overweight and obesity prevailed in 45.6 and 18.9 % in men, in 35.2 and 44.4 % in women
Referent in adiposity categories was BMI ≤25 kg/m2
aIncident cases/whole sample Diabetes prevailed in 100 (46 M/54 W) participants at baseline on models for the prediction of incident CHD by BMI categories and mediators, by gender Overweight and obesity prevailed in 45.6 and 18.9 % in men, in 35.2 and 44.4 % in women
Referent in adiposity categories was BMI ≤25 kg/m2 aIncident cases/whole sample Diabetes prevailed in 100 (46 M/54 W) participants at baseline
Bold numbers denote significant values [7]. RR of the age-adjusted CHD risk in the current study
was also close to the former meta-analysis. [7]. RR of the age-adjusted CHD risk in the current study
was also close to the former meta-analysis. CHD risk of overweight little modified by the three
mediators difference insofar as the RR was by one-third higher in
men, but nearly two-fold in women. Though mediation
by the 3 risk factors (with 55 % of the excess risk) was
closely similar to that in the global cohorts, gender dis-
parity in the current study was more prominent in the
mediation by the 3 risk factors: whereas 70 % of the
excess BMI risk was thus mediated in men, 60 % was
retained by BMI in women. A gender difference of the
BMI-mediated risk was not reported in three meta-
analyses [5–7] in which sex-stratified analysis was, how-
ever, not reported. ctors largely in men, modestly in women Excess CHD risk, compared with normal weight, showed
a steep rise in the obesity category relativeto the over-
weight one; sex-, age- and smoking-adjusted RRs in
combined gender were, namely, 2.24 and 1.40, respect-
ively. In comparison to the meta-analysis on global
cohorts [7], this basic excess risk displayed gender Can et al. BMC Endocrine Disorders (2016) 16:54 Page 7 of 9 Table 3 Cox regression models for the prediction of sex- and- age-adjusted CHD mortality and incident diabetes by 1SD of body
mass index adjusted also for mediators, by gender Table 3 Cox regression models for the prediction of sex- and- age-adjusted CHD mortality and incident diabetes by 1SD of body
mass index adjusted also for mediators, by gender Table 3 Cox regression models for the prediction of sex- and- age-adjusted CHD mortality and incident diabetes by 1SD of body
mass index adjusted also for mediators, by gender
Total
Men
Women
HR (95 % CI)
p
HR (95 % CI)
p
HR (95 % CI)
p
CHD mortality
146/2158
86/1015
60/1143
Body mass index, 5 kg/m2
1.12 (1.10–1.14)
<0.001
1.11 (1.09–1.13)
0.001
1.14 (1.11–1.17)
<0,001
Adj. Systolic BP
1.14 (0.95–1.37)
0.16
0.96 (0.72–1.28)
0.81
1.31 (1.03–1.66)
0,025
Adj. total cholesterol
1.23 (1.05–1.45)
0.010
1.05 (0.81–1.37)
0.70
1.14 (1.11–1.17)
<0,001
Adj. fasting glucose
1.22 (1.03–1.44)
0.021
1.04 (0.80–1.36)
0.76
1.38 (1.12–1.72)
0,003
Adj. SBP & T Chol. 1.11 (0.93–1.34)
0.25
0.94 (0.71–1.26)
0.70
1.27 (0.99–1.62)
0,051
Adj. SBP & FPG
1.12 (0.93–1.35)
0.21
0.93 (0.70–1.24)
0.63
1.30 (1.03–1.64)
0,025
Adj. T Chol & FPG
1.19 (1.00–1.41)
0.050
1.11 (0.78–1.33)
0.90
1.34 (1.07–1.67)
0,008
Adjusted for all 3 factors
1.10 (0.91–1.32)
0.31
0.91 (0.68–1.22)
0.55
1.26 (1.00–1.60)
0,050
Incident diabetes
284/2136
146/1014
138/1122
Body mass index, 5 kg/m2
1.73 (1.54–1.93)
<0.001
2.11 (1.75–2.54)
0.001
1.56 (1.35–1.79)
<0,001
Adj. Systolic BP
1.71 (1.52–1.93)
<0.001
2.13 (1.73–2.63)
0.001
1.54 (1.32–1.79)
<0,001
Adj. total cholesterol
1.71 (1.53–1.91)
0.001
2.06 (1.70–2.49)
0.001
1.56 (1.35–1.79
0,001
Adj. fasting glucose
1.71 (1.52–1.91)
0.001
2.10 (1.73–2.55)
0.001
1.55 (1.34–1.78
0,001
Adj. SBP & T Chol. 1.70 (1.50–1.92)
0.001
2.09 (1.69–2.58)
0.001
1.53 (1.32–1.79)
0,001
Adj. SBP & FPG
1.68 (1.49–1.90)
<0.001
2.09 (1.69–2.58)
0.001
1.52 (1.31–1.77)
<0,001
Adj. ctors largely in men, modestly in women T Chol & FPG
1.69 (1.51–1.90)
0.001
2.03 (1.67–2.48)
0.001
1.55 (1.34–1.78)
0,001
Adjusted for all 3 factorsa
1.67 (1.48–1.89)
<0.001
2.05 (1.65–2.54)
0.001
1.52 (1.31–1.77)
<0,001
Female sex was protected against diabetes at HR 0.58 (0.45; 0.75)
1 SD considered here for: systolic BP 25 mmHg, total cholesterol 40 mg/dl, glucose 25 mg/dl
aGlucose was the only other significant determinant of diabetes
Numbers in bold denote significant values Female sex was protected against diabetes at HR 0.58 (0.45; 0.75)
1 SD considered here for: systolic BP 25 mmHg, total cholesterol 40 mg/dl, glucose 25 mg/dl
aGlucose was the only other significant determinant of diabetes
Numbers in bold denote significant values The retained CHD risk by BMI may be due to path-
ways of systemic inflammation, endothelial dysfunction,
and thrombogenic factors. Alike the meta-analysis find-
ings, the largest mediation was found for systolic BP,
followed
by
total
cholesterol
and
least for
fasting
glucose. Obesity mediation by the individual risk factors
was markedly different across sexes inasmuch as gly-
cemia did not mediate any BMI risk in men, in contrast
to demonstrating the largest mediation (with 14 %) in
women. This distribution of obesity-mediated risk across
the sexes as well as the retention of the majority of the
obesity risk in women support the notion of autoimmune
process activated by obesity being far more common in
females than males [19]. It is the autoimmune activation
(induced by oxidative stress) and thrombo-embolic events
rather than BP or cholesterolemia that lead to fatal and
non-fatal CHD. fasting glucose, implicating that impaired fasting glucose
(IFG) protected Turkish women against CHD death. This
is in agreement with our previous report [23] that IFG sta-
tus in non-diabetic people without MetS is associated with
a less adverse cardiovascular risk profile and reduced future
CHD risk, lest modulated by the developed MetS, especially
in women in whom serum Lp(a), linked inversely to
HOMA, is the likely modulator. The modest magnitude of HR herein is in line with no
effects of overweight on cardiovascular mortality [24], or
with overall mortality in a meta-analysis [25]. In con-
trast, in the Prospective Studies Collaboration [26] 5
units of BMI revealed 40 % excess vascular mortality risk
above the optimal BMI 22.5-25 kg/m2; below this range,
BMI was inversely related to overall mortality. BMI modest determinant of CHD mortality Excess risk of age- and smoking-adjusted obesity was
larger by one-half in women than men. This is at some
variance from the findings of a large pooled analysis on
1.4 million individuals [20] in which effects by age, sex
or global region of major metabolic risk factors on car-
diovascular disease and diabetes were evaluated. At the
age group 55–64 years, the RR was largest for systolic The sex- and age-adjusted continuous BMI variable was a
significant modest contributor to HR for CHD mortality;
yet this risk was fully abolished by mediation of systolic BP
in men, indicating this factor assumed the main determin-
ant of CHD mortality. In contrast, the HR was strength-
ened to over two-fold in women through mediation of Page 8 of 9 Page 8 of 9 Can et al. BMC Endocrine Disorders (2016) 16:54 BP with respect to hypertensive HD or stroke, for BMI
regarding diabetes or CHD, and at much lower effect
size for fasting glucose in regard to CHD or stroke. Pro-
portional effects declined with age, were generally con-
sistent by sex, and differed little by region. However,
effects of BMI on diabetes were larger in Western com-
pared with Asian cohorts in younger adults [20]. CHD risk. In men, this was extensively mediated while in
women, 60 % of the excess CHD risk and virtually the en-
tire risk for CHD death were retained by BMI. These find-
ings can be explained by autoimmune activation occurring
in a greater proportion of females, a process mainly in-
duced by oxidative damage to Lp(a) in a setting of excess
adiposity. In regard to BMI-mediated CHD risk of fasting glu-
cose and systolic BP, a clear sex disparity was evident:
Women had a significant risk through glucose (by 15 %
higher than men), while men had such through systolic
BP (by 20 % higher). We have evidence that these
features are largely gene-dependent. The TT genotype of
the CYP19A1 polymorphism encoding the aromatase
enzyme involved in the final step of estrogen synthesis
showed, namely, interaction with narrow waist girth for
hypertension only in men, independent of age and BMI
[27]. In Caucasian US women, an interaction between
the risk allele of the FTO rs8050136 polymorphism
mediated by BMI and low physical activity yielded
increased cardiovascular disease risk [28]. Authors’ contributions GC designed the study, performed the statistical analyses and contributed to
drafting of the manuscript, AO obtained funding, conceived the study and
revised the manuscript critically, EY performed the statistical analyses
analyzed and contributed to the discussion; AK performed the biochemical
analyses and revised the manuscript critically; YK and TA collected data and
revised the manuscript critically; HY contributed to drafting and revision of
the manuscript. All authors read and approved the final manuscript. Limitations and strength Our basic regression models did not adjust for certain
confounding factors such as socioeconomic state, physical
activity grade, or diet; this might have affected the BMI-
mediated CHD risk, though not to a great extent. Analysis
stratified to sex which substantially generated novel know-
ledge constitutes a major strength. Despite the fact that,
due to lack of an oral glucose tolerance test, some cases of
actual diabetes may have been missed, analyzing the BMI-
mediated effects also for diabetes and CHD mortality in
the same cohort represents further strength. Received: 9 March 2016 Accepted: 21 September 2016 Received: 9 March 2016 Accepted: 21 September 2016 Received: 9 March 2016 Accepted: 21 September 2016 Gender, fatal or nonfatal CHD or diabetes modulated in
middle-aged Turkish adults the mediation of BMI by 3
traditional risk factors, or the retention of residual excess
risk by BMI, respectively. Hardly any mediation was
observed
with
respect
to
diabetes. Compared
with
normal-weight, obesity imparted a significant over 2-fold Funding We thank the Turkish automotive firm TOFAŞ, Istanbul, for partial financial
support of the Turkish Adult Risk Factor study. Implications relate both to the prevention of CHD and
DM. Coronary prevention programs should take into
account the role of sex and be modified for sex. While
addressing the mediators with measures including antihy-
pertensive and lipid-lowering medication may largely be
effective in men, addressing the mediators in women will
alleviate only slightly the prevention problem. Major im-
provement in obesity and related proinflammatory state is
required for the CHD prevention. As regards prevention of
type-2 diabetes, our findings indicate that the overwhelm-
ing portion of the risk can be reduced only by weight loss
and improvement in the related oxidative stress (mediated
by Lp(a)-activated autoimmunity), regardless of gender. The study was approved by the Ethics Committee of the Istanbul University
Medical Faculty. The study was approved by the Ethics Committee of the Istanbul University
Medical Faculty. Consent to participate was obtained from each individual. Availability of data and materials The datasets during and/or analysed during the current study available from
the corresponding author on reasonable request. Consent for publication
l
bl Consent for publication
Not applicable. Additional files Additional file 1: Figure S1. Comparative sex distribution of the
proportions in the three adiposity categories is shown. Though men
prevail in “normal weight” and overweight categories, women
predominate in obesity by over two-fold. (DOCX 15 kb)
Additional file 2: Table S1. Cox regression models for the prediction
of incident CHD by presence of abdominal obesity and three mediators,
by gender. (DOCX 21 kb) Additional file 1: Figure S1. Comparative sex distribution of the
proportions in the three adiposity categories is shown. Though men
prevail in “normal weight” and overweight categories, women
predominate in obesity by over two-fold. (DOCX 15 kb) Additional file 1: Figure S1. Comparative sex distribution of the
proportions in the three adiposity categories is shown. Though men
prevail in “normal weight” and overweight categories, women
predominate in obesity by over two-fold. (DOCX 15 kb) Additional file 2: Table S1. Cox regression models for the prediction
of incident CHD by presence of abdominal obesity and three mediators,
by gender. (DOCX 21 kb) Author details 1Departments of Public Health, Istanbul University, Yazıcı sok. 18/5,
Kocamustafapaşa, 34098 Istanbul, Turkey. 2Departments of Cardiology,
Cerrahpaşa Medical Faculty, Istanbul University, Istanbul, Turkey. 3Department
of Endocrinology and Metabolism, Ataturk University Faculty of Medicine,
Erzurum, Turkey. 4Bağcılar Educational Hospital, Istanbul, Turkey. 5Departments of Biochemistry Laboratory, Institute of Cardiology, Istanbul
University, Istanbul, Turkey. Competing interests g
The authors declare that they have no competing interests. Ethics approval and consent to participate The study was approved by the Ethics Committee of the Istanbul University
Medical Faculty. Can et al. BMC Endocrine Disorders (2016) 16:54 Body mass index and cardiovascular disease in the Asia-Pacific
Region: an overview of 33 cohorts involving 310 000 participants. Int J
Epidemiol. 2004;33:751–8. 24. Flegal KM, Graubard BI, Williamson DF, Gail MH. Cause-specific excess deaths
associated with underweight, overweight, and obesity. JAMA. 2007;298:2028–37. 25. Flegal KM, Kit BK, Orpana H, Graubard BI. Association of all-cause mortality
with overweight and obesity using standard body mass index categories: a
systematic review and meta-analysis. J Am Med Assn. 2013;309:71–82. 6. Bogers RP, Bemelmans WJ, Hoogenveen RT, The BMI-CHD Collaboration
Investigators. Association of overweight with increased risk of coronary
heart disease partly independent of blood pressure and cholesterol levels: a
meta-analysis of 21 cohort studies including more than 300 000 persons. Arch Intern Med. 2007;167:1720–8. 26. Prospective Studies Collaboration, Whitlock G, Lewington S, Sherliker P,
Clarke R, Emberson J, Halsey J, Qizilbash N, Collins R, Peto R. Body-mass
index and cause-specific mortality in 900 000 adults: collaborative analyses
of 57 prospective studies. Lancet. 2009;373:1083–96. 7. Lu Y, Hajifathalian K, Ezzati M, Woodward M, Rimm EB, Danaei G. The Global
Burden of Metabolic Risk Factors for Chronic Diseases Collaborating Group
(BMI-mediated Effects). Metabolic mediators of the effects of body-mass-
index, overweight, and obesity on coronary heart disease and stroke: a
pooled analysis of 97 prospective cohorts with 1.8 million participants. Lancet. 2014;383:970–83. 27. Çoban N, Onat A, Güçlu-Geyik F, Can G, Erginel-Unaltuna N. Sex- and
obesity-specific association of CYP19A1 gene variant with apolipoprotein B
and hypertension. Arch Med Res. 2015;46:564–71. 28. Ahmad T, Chasman DI, Mora S, Paré G, Cook NR, Buring JE, Ridker PM, Lee J-M. The fat-mass and obesity-associated (FTO) gene, physical activity, and risk of
incident cardiovascular events in white women. Am Heart J. 2010;160:1163–9. 8. Onat A, Hergenç G. Low-grade inflammation and dysfunction of high-
density lipoprotein and its apolipoproteins as a major driver of
cardiometabolic risk. Metabolism. 2011;60:499–512. 9. Onat A, Avcı GŞ, Şenocak M, Örnek E, Gözükara Y. Plasma lipids and their
interrelation in Turkish adults. J EpidemCommun Health. 1992;46:470–6. 10. Expert Panel on detection, evaluation and treatment of high blood
cholesterol in adults (Adult Treatment Panel III) Executive Summary of the
Third Report of the National Cholesterol Education Program (NCEP). JAMA
2001;285:2486–97. 11. Grundy SM, Brewer HB, Cleeman JI, Smith SC, Lenfant C. Definition of
metabolic syndrome: report of the National Heart, Lung, and Blood
Institute/American Heart Association conference on scientific issues related
to definition. Circulation. 2004;109:433–8. 12. Can et al. BMC Endocrine Disorders (2016) 16:54 Can et al. BMC Endocrine Disorders (2016) 16:54 Page 9 of 9 Can et al. BMC Endocrine Disorders (2016) 16:54 2. Song X, Jousilahti P, Stehouwer CDA, Söderberg S, Onat A, Laatikainen T,
Yudkin JS, Dankner R, Morris R, Tuomilehto J, Qiao Q. Cardiovascular and all-
cause mortality in relation to various anthropometric measures of obesity in
Europeans. Nutr Metab Cardiovasc Dis. 2015;25(3):295–304. age-specific quantitative effects of metabolic risk factors on cardiovascular
diseases and diabetes: a pooled analysis. PLoS One. 2013;8:e65174. age-specific quantitative effects of metabolic risk factors on cardiovascular
diseases and diabetes: a pooled analysis. PLoS One. 2013;8:e65174. 21. Poirier P, Giles TD, Bray GA, et al. Obesity and cardiovascular disease:
pathophysiology, evaluation, and effect of weight loss: an update of the
1997 American heart association scientific statement on obesity and heart
disease from the obesity committee of the council on nutrition, physical
activity, and metabolism. Circulation. 2006;113:898–918. 3. Jousilahti P, Tuomilehto J, Vartiainen E, Pekkanen J, Puska P. Body weight,
cardiovascular risk factors, and coronary mortality. 15-year follow-up of middle-
aged men and women in eastern Finland. Circulation. 1996;93:1372–9. 3. Jousilahti P, Tuomilehto J, Vartiainen E, Pekkanen J, Puska P. Body weight,
cardiovascular risk factors, and coronary mortality. 15-year follow-up of middle-
aged men and women in eastern Finland. Circulation. 1996;93:1372–9. 22. Van Gaal LF, Mertens IL, De Block CE. Mechanisms linking obesity with
cardiovascular disease. Nature. 2006;444:875–80. 4. Batty GD, Shipley MJ, Jarrett RJ, Breeze E, Marmot MG, Davey SG. Obesity
and overweight in relation to disease-specific mortality in men with and
without existing coronary heart disease in London: the original whitehall
study. Heart. 2006;92:886–92. 4. Batty GD, Shipley MJ, Jarrett RJ, Breeze E, Marmot MG, Davey SG. Obesity
and overweight in relation to disease-specific mortality in men with and
without existing coronary heart disease in London: the original whitehall
study. Heart. 2006;92:886–92. 23. Onat A, Aydın M, Can G, Çakmak HA, Köroğlu B, Kaya A, Ademoğlu E. Impaired fasting glucose: pro-diabetic, “atheroprotective” and modified by
metabolic syndrome. World J Diabetes. 2013;4:210–8. 5. Ni Mhurchu C, Rodgers A, Pan WH, Gu DF, Woodward M, Collaboration
APCS. Body mass index and cardiovascular disease in the Asia-Pacific
Region: an overview of 33 cohorts involving 310 000 participants. Int J
Epidemiol. 2004;33:751–8. 5. Ni Mhurchu C, Rodgers A, Pan WH, Gu DF, Woodward M, Collaboration
APCS. Can et al. BMC Endocrine Disorders (2016) 16:54 Can et al. BMC Endocrine Disorders (2016) 16:54 Onat A, Uyarel H, Hergenç G, Karabulut A, Albayrak S, Can G. Determinants
and definition of abdominal obesity as related to risk of diabetes, metabolic
syndrome and coronary disease in Turkish men: a prospective cohort study. Atherosclerosis. 2007;191:182–90. 13. Genuth S, Alberti KG, Bennett P, Buse J, Defronzo R, et al. Expert committee on the
diagnosis and classification of diabetes mellitus. Follow-up report on the diagnosis
of diabetes mellitus: the expert committee on the diagnosis and classification of
diabetes mellitus. Diabetes Care. 2003;26:3160–7. PMID: 14578255. 14. Rose G, Blackburn H, Gillum RF, Prineas RJ. Cardiovascular survey methods,
2nd ed. Geneva: WHO; 1982. p. 124–7. 15. Lin DY, Fleming TR, Gruttola V. Estimating the proportion of treatment
effect by a surrogate marker. Stat Med. 1997;16:1515–27. 16. Hotamisligil GS, Shargill NS, Spiegelman BM. Adipose expression of tumor
necrosis factor-α: direct role in obesity-linked insulin resistance. Science. 1993;259:87–91. 17. Fuentes E, Fuentes F, Vilahur G, Badimon L, Palomo I. Mechanisms of chronic
state of inflammation as mediators that link obese adipose tissue and metabolic
syndrome. Mediators Inflamm. 2013; 136584. doi: 10.1155/2013/136584 References
W . Wormser D, Kaptoge S, Di Angelantonio E, Wood AM, Pennells L, Thompson A,
et al. Separate and combined associations of body-mass index and abdominal
adiposity with cardiovascular disease: collaborative analysis of 58 prospective
studies. Lancet. 2011;377:1085–95. Page 9 of 9 Page 9 of 9 y
24.
Flegal KM, Graubard BI, Williamson DF, Gail MH. Cause-specific excess deaths
associated with underweight, overweight, and obesity. JAMA. 2007;298:2028–37. Submit your next manuscript to BioMed Central
and we will help you at every step: Submit your next manuscript to BioMed Central
and we will help you at every step: 18. Danaei G, Finucane MM, Lu Y, For the Global Burden of Metabolic Risk
Factors of Chronic Diseases Collaborating Group (Blood glucose), et al. National, regional, and global trends in fasting plasma glucose and diabetes
prevalence since 1980: systematic analysis of health examination surveys
and epidemiological studies with 370 country-years and 2.7 million
participants. Lancet. 2011;378:31–40. • We accept pre-submission inquiries
• Our selector tool helps you to find the most relevant journal
• We provide round the clock customer support
• Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research
Submit your manuscript at
www.biomedcentral.com/submit
and we will help you at every step: 19. Onat A, Can G. Enhanced pro-inflammatory state and autoimmune
activation: a breakthrough to understanding chronic diseases. Curr Pharm
Design. 2014;20:575–84. 20. Singh GM, Danaei G, Farzadfar F, The Global Burden of Metabolic Risk
Factors of Chronic Diseases Collaborating Group.Asia-Pacific Cohort Studies
Collaboration (APCSC); The Diabetes Epidemiology: Collaborative analysis of
Diagnostic criteria in Europe (DECODE); The Emerging Risk Factor
Collaboration (ERFC); The Prospective Studies Collaboration (PSC). The
|
https://openalex.org/W3094816384
|
https://www.repository.cam.ac.uk/bitstream/1810/314830/1/Gray%20matter%20covariations%20and%20core%20symptoms%20of%20autism%20the%20EU-AIMS%20Longitudinal%20European%20Autism%20Project.pdf
|
English
| null |
Gray matter covariations and core symptoms of autism: the EU-AIMS Longitudinal European Autism Project
|
Molecular autism
| 2,020
|
cc-by
| 10,354
|
Open Access Open Access Gray matter covariations and core symptoms
of autism: the EU‑AIMS Longitudinal European
Autism Project Ting Mei1* , Alberto Llera1,2, Dorothea L. Floris1, Natalie J. Forde1, Julian Tillmann3, Sarah Durston4,
Carolin Moessnang5, Tobias Banaschewski6, Rosemary J. Holt7, Simon Baron‑Cohen7, Annika Rausch1,
Eva Loth8, Flavio Dell’Acqua8, Tony Charman3, Declan G. M. Murphy8, Christine Ecker8,9,
Christian F. Beckmann1,10 and Jan K. Buitelaar1,2* on behalf of the EU-AIMS LEAP group © The Author(s) 2020. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Mei et al. Molecular Autism (2020) 11:86
https://doi.org/10.1186/s13229-020-00389-4 Abstract Background: Voxel-based morphometry (VBM) studies in autism spectrum disorder (autism) have yielded diverging
results. This might partly be attributed to structural alterations being associating with the combined influence
of several regions rather than with a single region. Further, these structural covariation differences may relate to
continuous measures of autism rather than with categorical case–control contrasts. The current study aimed to
identify structural covariation alterations in autism, and assessed canonical correlations between brain covariation
patterns and core autism symptoms. Methods: We studied 347 individuals with autism and 252 typically developing individuals, aged between 6 and
30 years, who have been deeply phenotyped in the Longitudinal European Autism Project. All participants’ VBM maps
were decomposed into spatially independent components using independent component analysis. A generalized
linear model (GLM) was used to examine case–control differences. Next, canonical correlation analysis (CCA) was
performed to separately explore the integrated effects between all the brain sources of gray matter variation and two
sets of core autism symptoms. Results: GLM analyses showed significant case–control differences for two independent components. The first
component was primarily associated with decreased density of bilateral insula, inferior frontal gyrus, orbitofrontal
cortex, and increased density of caudate nucleus in the autism group relative to typically developing individuals. The second component was related to decreased densities of the bilateral amygdala, hippocampus, and
parahippocampal gyrus in the autism group relative to typically developing individuals. The CCA results showed
significant correlations between components that involved variation of thalamus, putamen, precentral gyrus, frontal,
parietal, and occipital lobes, and the cerebellum, and repetitive, rigid and stereotyped behaviors and abnormal
sensory behaviors in autism individuals. Limitations: Only 55.9% of the participants with autism had complete questionnaire data on continuous parent-
reported symptom measures. *Correspondence: t.mei@donders.ru.nl; Jan.Buitelaar@radboudumc.nl
1 Department of Cognitive Neuroscience, Donders Institute for Brain,
Cognition and Behaviour, Radboud University Nijmegen Medical Centre,
Nijmegen, The Netherlands
Full list of author information is available at the end of the article *Correspondence: t.mei@donders.ru.nl; Jan.Buitelaar@radboudumc.nl
1 Department of Cognitive Neuroscience, Donders Institute for Brain,
Cognition and Behaviour, Radboud University Nijmegen Medical Centre,
Nijmegen, The Netherlands
Full list of author information is available at the end of the article Background and incorporate an alternative that adheres more
closely to the hypothesis of autism as a disconnection
syndrome [13], thus providing greater sensitivity for
between-group effects. For this purpose, we identify
inter-regional
sources
of
structural
covariation
using independent component analysis (ICA) [15], a
data-driven unsupervised approach that allows the
identification of interconnected brain regions across the
whole brain. It has previously been applied successfully
to identify covariance of brain morphometry in patients
with psychiatric disorders [16–18]. Second, we move
beyond the categorical autism case–control comparison
towards exploring associations between brain structure
and symptom dimensions or profiles of autism. Although
former studies have used univariate approaches to
explore the relationship between brain substrates
and clinical phenotypes [6, 19], such associations are
potentially the consequence of integrated effects across
multiple symptoms dimensions and brain regions,
rather than simple associations between a specific brain
region and a specific symptom dimension. To study such
multidimensional associations multivariate methods are
effective [20, 21] and here we achieve this integration
using Canonical Correlation Analysis (CCA) [22]. g
Autism spectrum disorder (henceforth autism) is an early
onset neurodevelopmental condition characterized by
core deficits in social interaction and communication,
along with restrictive interests and behavior, and
sensory abnormalities [1]. Magnetic resonance imaging
(MRI) studies have increased our understanding of the
neuroanatomical underpinnings of autism and show
that autism is associated, at the group level, with brain
structural changes [2]. However, many results are not
robust across different studies. For example, two studies
using the same large-scale open access Autism Brain
Imaging Data Exchange (ABIDE) dataset [3] came to
different conclusions with regard to the volume of the
pallidum [4, 5]. Also, across whole brain approaches
investigating cortical (i.e., cortical thickness and surface
area) and subcortical (i.e., volume) features have been
inconsistent; two large-scale pooled estimate analytical
studies observed diverging results of cortical changes in
autism [6, 7]. Similarly, autism studies quantifying voxel-
wise gray matter (GM) density also found divergent
results of GM differences between autism diagnosed
and control individuals; for instance, meta-analyses
reported diverse changes of GM morphometry in autistic
individuals on average, reporting either increased or
decreased density of right inferior temporal gyrus in
autism [8, 9]. Even when taking age into account, studies
still observed different structural brain alterations in
children and adolescents with autism [10, 11]. Background In summary, we investigate alterations in GM
morphometric covariations in a deeply phenotyped large
European autism case–control sample [23, 24] that allows
us to improve our understanding of correlated structural
brain alterations in autism. Subsequently, we focus on
the covariation between the identified structural features
and symptom behavior profiles among individuals with
autism. A commonality to all these studies is their reliance
on mass-univariate statistics. This approach identifies
alterations in isolated regions or voxels but ignores
possible relationships between them. The brain is
a complex system of interconnected networks, and
research into the neural basis of autism has moved
away from focusing on local abnormalities into
conceptualizing autism as a disorder of alterations in
structural and functional brain connectivity [12]. This
implies that structural brain alterations in autism likely
reflect the combined influence of several regions and are
not confined to one specific region [13, 14]. The present
paper aims to advance prior work on brain structural
neural correlates of autism in two ways. First, we aim
to move away from the standard univariate approach © The Author(s) 2020. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco
mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Page 2 of 13 (2020) 11:86 Mei et al. Molecular Autism (2020) 11:86 Mei et al. Molecular Autism Conclusions: Covaried areas associated with autism diagnosis and/or symptoms are scattered across the whole
brain and include the limbic system, basal ganglia, thalamus, cerebellum, precentral gyrus, and parts of the frontal,
parietal, and occipital lobes. Some of these areas potentially subserve social-communicative behavior, whereas others
may underpin sensory processing and integration, and motor behavior. Keywords: Autism, Magnetic resonance imaging, Voxel-based morphometry, Independent component analysis,
Canonical correlation analysis GM density estimation Voxel-based morphometry (VBM) is a spatially unbiased
whole-brain approach that extracts voxel-wise GM
density (the amount of GM at a voxel) estimations. We
performed VBM analyses using the CAT12 toolbox
[30] in SPM12 (Wellcome Department of Imaging
Neuroscience,
London,
UK). T1-weighted
images
were automatically segmented into GM, white matter,
and cerebrospinal fluid and affine registered to the
MNI template to improve segmentation. All resulting
segmented GM maps were then used to generate a
study-specific template and registered to MNI space via
a high-dimensional, nonlinear diffeomorphic registration
algorithm (DARTEL) [31]. A Jacobian modulation step
was included using the flow fields to preserve voxel-
wise information on local tissue volume. Images were
smoothed with a 4 mm full-width half-max (FWHM)
isotropic Gaussian kernel. MRI data acquisition
ll All participants were scanned on 3 T MRI scanners
(University of Cambridge: Siemens Verio; King’s
College London: GE Medical Systems Discovery MR
750; Mannheim University: Siemens TimTrio; Radboud
University: Siemens Skyra; Rome University: GE Medical
Systems Sigma HDxTt; Utrecht University: Philips
Medical Systems Achieva/Ingenia CX). High-resolution
structural T1-weighted images were acquired with full
head coverage, at 1.2 mm thickness with 1.1 × 1.1 mm
in-plane resolution.For all other scanning parameters,
please see Additional file 1: Subsection 1. In the present study, we selected participants
with available structural MRI data. All images were
inspected visually and participants were excluded in
cases of brain injury or structural abnormalities (e.g.,
enlarged ventricles or cysts), excessive head motion,
or preprocessing failure (n = 29). We excluded the
participants from the Rome site due to low sample size
(n = 1). We also excluded participants without full-
scale intelligence quotient (FSIQ) data in the further
statistical analyses (n = 5). This resulted in a sample of
599 participants from 5 sites, including 347 individuals
with autism and 252 typically developing (TD) controls. Demographic and clinical information is shown in
Table 1. Participantsh The participants were selected from the first wave of the
European Autism Interventions—A Multicentre Study for
Developing New Medications (EU-AIMS) Longitudinal
European Autism Project (LEAP) dataset, which is
a large multicenter study that aims to identify and
validate biomarkers for autism [24]. In total, six centers
are involved: Institute of Psychiatry, Psychology and
Neuroscience, King’s College London, United Kingdom;
Autism Research Centre, University of Cambridge, Page 3 of 13 Mei et al. Molecular Autism (2020) 11:86 Mei et al. Molecular Autism (2020) 11:86 Mei et al. Molecular Autism United Kingdom; Radboud University Medical Centre,
Nijmegen, the Netherlands; University Medical Centre
Utrecht, the Netherlands; Central Institute of Mental
Health, Mannheim, Germany; and University Campus
Bio-Medico, Rome, Italy. Each participant underwent
clinical,
cognitive,
and
MRI
assessment. Autism
diagnoses were confirmed by clinicians according to
the Diagnostic and Statistical Manual-IV (DSM-IV),
International Statistical Classification of Diseases and
Related Health Problems 10th Revision (ICD-10),
or DSM-5. The study was approved by local ethical
committees in each participating center, and written
informed consent was provided by all participants and/or
their legal guardians (for those < 18 years old). For further
details on experimental design and clinical assessments,
see [23, 24]. Disorder (ADHD), we included comorbidity with ADHD
as an additional covariate in analyses. ADHD symptoms
were assessed with the ADHD DSM-5 rating scale that
includes symptom scales of inattention and hyperactivity/
impulsivity symptoms. The ADHD DSM-5 rating scale
was based on either parent-report or self-report scores;
self-report scores were only used when parent-reports
were unavailable. The categorical output of the ADHD
rating scale was used in this study. The summary for each
of these clinical measures can be found in Table 1. Clinical measures
W
d
h
A We considered calibrated severity scores
g Parent report SRS scores were available for 284 participants with autism and 135 TD individuals
h RBS scores were available for 277 participants with autism and 133 TD individuals
i SSP scores were available for 201 participants with autism and 115 TD individuals
g,h,i In all questionnaires, the scores of the autism group only were used in our study, and they are all parent-rated
Demographic
Autism, n = 347
TD, n = 252
t/χ2
p value
Mean
SD
Mean
SD
Age, yearsa
16.79
5.56
16.92
5.71
− 0.270
0.788
FSIQa,b
99.40
18.94
104.88
18.26
− 3.549
p < 0.001
FSIQ ≥ 75
104.29
14.95
109.02
13.07
− 3.883
p < 0.001
FSIQ < 75
66.61
5.44
63.69
9.20
1.399
0.172
n
%
n
%
Sex, male/femalec
253/94
72.9/27.1
163/89
64.7/35.3
4.658
0.031
ADHD rating scaled, with/
without ADHD
139/160
46.5/53.5
21/180
10.4/89.6
71.750
p < 0.001
Symptom profiles
Mean
SD
Mean
SD
ADIe
Social interaction
16.80
6.66
Communication
13.50
5.62
RRB
4.32
2.67
ADOSf
Social affect
6.04
2.59
RRB
4.73
2.78
SRS T-scoreg
70.59
12.06
47.24
8.79
RBSh
15.76
13.42
2.20
8.28
SSPi
138.62
27.28
175.97
16.18 Table 1 Participant characteristics
TD, typically developing; SD, standard deviation; FSIQ, full-scale intelligence quotient; ADHD, Attention Deficit Hyperactivity Disorder; ADI, Autism Diagnostic
Interview-Revised; RRB restricted repetitive behaviors; ADOS Autism Diagnostic Observational Schedule 2; SRS Social Responsiveness Scale 2nd Edition; RBS
Demographic
Autism, n = 347
TD, n = 252
t/χ2
p value
Mean
SD
Mean
SD
Age, yearsa
16.79
5.56
16.92
5.71
− 0.270
0.788
FSIQa,b
99.40
18.94
104.88
18.26
− 3.549
p < 0.001
FSIQ ≥ 75
104.29
14.95
109.02
13.07
− 3.883
p < 0.001
FSIQ < 75
66.61
5.44
63.69
9.20
1.399
0.172
n
%
n
%
Sex, male/femalec
253/94
72.9/27.1
163/89
64.7/35.3
4.658
0.031
ADHD rating scaled, with/
without ADHD
139/160
46.5/53.5
21/180
10.4/89.6
71.750
p < 0.001
Symptom profiles
Mean
SD
Mean
SD
ADIe
Social interaction
16.80
6.66
Communication
13.50
5.62
RRB
4.32
2.67
ADOSf
Social affect
6.04
2.59
RRB
4.73
2.78
SRS T-scoreg
70.59
12.06
47.24
8.79
RBSh
15.76
13.42
2.20
8.28
SSPi
138.62
27.28
175.97
16.18 Table 1 Participant characteristics after preprocessing, whereas none of them were observed
obvious artifacts, and they were consequently included in
the final analyses. Clinical measures
W
d
h
A We used the Autism Diagnostic Interview-Revised
(ADI) [25] and the Autism Diagnostic Observational
Schedule 2 (ADOS) [26] to quantify past (ever and
previous 4-to-5 years) and current autism symptoms of
the social interaction, communication, and restricted
repetitive behaviors (RRB) domains. We used T-scores
(age- and sex-adjusted) of the Social Responsiveness
Scale 2nd Edition (SRS) [27] in the autism group to assess
severity of autistic traits/symptoms and the Repetitive
Behavior Scale-Revised (RBS) [28] to measure repetitive
and rigid behaviors associated with autism. Moreover,
sensory processing abnormalities of autism were assessed
with the Short Sensory Profile (SSP) [29]. To examine
associations between clinical features in autism and
brain measures, we created two sets of clinical measures:
(1) the subscale scores of ADI-R and ADOS, both
instruments were rated by qualified examiners, and (2)
the total scores of SRS, RBS, and SSP, we included parent-
rated reports only and limited the analyses to within the
autism group. Further, concerning the potential effect
of comorbidity with Attention Deficit Hyperactivity A full quality control report was generated by the
CAT SPM pipeline for each participant that included
visualizations of the segmentation, which were evaluated
visually, and quantitative quality measures including
mean correlation from sample homogeneity module
and weighted overall image quality ratings that were
additionally used to detect and exclude images of
insufficient quality for inclusion in analysis. We visually
checked the images with the mean correlation smaller
than three standard deviations from the sample mean. Accordingly, 5 participants required visual inspection Mei et al. Molecular Autism (2020) 11:86 Page 4 of 13 Mei et al. Molecular Autism Table 1 Participant characteristics
TD, typically developing; SD, standard deviation; FSIQ, full-scale intelligence quotient; ADHD, Attention Deficit Hyperactivity Disorder; ADI, Autism Diagnostic
Interview-Revised; RRB, restricted, repetitive behaviors; ADOS, Autism Diagnostic Observational Schedule 2; SRS, Social Responsiveness Scale 2nd Edition; RBS,
Repetitive Behavior Scale-Revised; SSP, Short Sensory Profile
a Statistical differences were assessed by two-sample t test
b In Schedule A, B, and C (FSIQ ≥ 75), there are 302 participants with autism and 229 participants with TD. Schedule D (FSIQ < 75) comprised 45 participants with
autism and 23 TD individuals
c Sex difference was examined by the chi-square test
d ADHD rating scores were available for 500 participants, including 299 individuals with autism and 201 TD individuals
e ADI scores were available for 332 participants
f ADOS scores were available for 339 participants. Clinical measures
W
d
h
A orders (i.e., number of components) tend to extract
spatially sparser maps, and hence participant loading
vectors variance would be explained by smaller subsets
of participants. Therefore, 100 components were
chosen to capture as much variation as possible while
remaining statistically powered. However, the choice
of model order is heuristic, we thereby examined
the dependence on model order using different
dimensionalities; more precisely, in addition to the
100 dimensional factorization, we also considered
an automatic dimension estimation approach as
implemented in MELODIC-ICA and a 50 dimensional
independent component factorization. Statistical approach level. Canonical variates are generated independently for
the brain and the behavioral datasets according to the
product of the canonical coefficients (learned through
CCA) and the original datasets. We referred to each pair
of canonical variates as CCA mode [38]. The canonical
coefficients separately represent weights for each brain
variable and each symptom measure. For interpretation
of the contribution of each independent source and
each clinical measure to the CCA mode, their canonical
coefficients were corrected as noted in [39]. In this work
we focused on the first canonical mode since it provides
the strongest associations and we denoted it as the
main CCA mode. Furthermore, for each CCA analysis
the statistical significance of the main CCA mode was
determined by permutation testing (10,000 permutations,
Bonferroni corrected p < 0.05/number of CCA modes). Here, we performed two separate CCA analyses to link
the independent components participants’ contributions
to subsets of behavioral measures; in the first CCA
analyses (CCA1) we included the subscales of ADI and
ADOS as clinical measures and in the second (CCA2)
we used total scores of SRS, RBS, and SSP. The reliability
of the CCA results presented as well as its dependence
on the number of participants were tested using a leave-
one-out cross-validation approach (Additional file 1:
Subsection 4). For completeness we first performed a standard mass-
univariate statistical analyses directly on the GM
densities. To that end we used a generalized linear model
(GLM) to detect group differences (autism vs. TD) on
GM densities using the FMRIB Software Library v6.0
(FSL) [33]. Participants’ VBM maps were considered
as the dependent variable and diagnostic group as the
independent factor, with age, sex, FSIQ, and scan site as
covariates. Significance was assessed using permutation
testing (5000 permutations) and correction for multiple
comparisons was achieved using Threshold-Free Cluster
Enhancement (TFCE, two-tailed, threshold at p < 0.05)
[34, 35]. Next, we considered the results of the ICA factorization
of the VBM data. A GLM was used to examine differences
between autistic and TD individuals, by using each
participant’s loading to each component as a dependent
variable, diagnostic group as independent variable, and
age, sex, FSIQ, and scan site as regressors. To avoid
the results of case–control differences being biased
by IQ, we repeated the same procedure by excluding
participants in Schedule D (FSIQ < 75, more details see
Additional file 1: Subsection 2). Structural ICA decomposition All participants’ VBM data were simultaneously
decomposed into 100 spatially independent sources
of spatial variation using MELODIC-ICA [15]. Such
ICA decomposition provides, at each independent
component, a brain map reflecting a pattern of
GM density covariation across participants, and a
participant’s loading vector reflecting the contribution
of each participant to each component. Higher model Mei et al. Molecular Autism (2020) 11:86 Mei et al. Molecular Autism (2020) 11:86 Page 5 of 13 Mass‑univariate statisticsh The standard mass-univariate GLM analysis of the
VBM data comparing cases and controls did not show
significant group differences for voxel-wise GM volumes
after multiple comparison correction. We provide the
uncorrected statistical results (p < 0.05) in Additional
file 1: Subsection 5. Statistical approach Considering that autism
is highly comorbid with ADHD [36], we also controlled
for comorbidity with ADHD by adding a dummy-
coded variable (with/without ADHD) to the original
GLM analyses (autism vs. TD) in the case–control ICA
analysis. Additionally, we also conducted analyses that
included a dimensional score (instead of categorical
code) of the two ADHD subscales as an additional
covariate in the group effect analyses (see Additional
file 1: Subsection 3). Furthermore, we separately checked
the effects of age squared, age-by-group, age squared-
by-group, sex-by-group interactions, anxiety, depression
comorbidity, medication use and image quality on case–
control differences of structural covariance. The results
can be found in Additional file 1: Subsection 3. Multiple
comparison correction was implemented using false
discovery rate (FDR) (p < 0.05) [37]. Group effect on ICA decompositionh The structural data ICA decomposition provided a
set of 100 independent spatial sources, each of which
is connected to a vector that depicts the degree of
each participant’s contribution to the corresponding
components. For clarity, we further refer to these vectors
as the participant loadings. Post-hoc GLM analyses
of these participant loadings showed case–control
differences at nine independent components (ICs)
(p < 0.05, i.e., IC10, IC13, IC14, IC15, IC23, IC28, IC31,
IC48, and IC 99, see Additional file 1: Subsection 6). Of
these,
two
components,
IC10
( β = −0.147 ,
p = 8.850 × 10–5, effect size [Cohen’s d] d = − 0.358)
and IC14 ( β = −0.132 , p = 5.450 × 10–4, d = − 0.321),
survived multiple comparison correction (FDR corrected,
p < 8.072 × 10–4). These results were not driven by age,
sex, or scan site. We also explored independently the relationships
between each estimated brain component and subscales
of ADI and ADOS, SRS, RBS, and SSP in the autism
group using GLM analyses and again correction for
multiple comparisons was implemented with the FDR
method (p < 0.05). Then, to simultaneously explore the
relationship between all the brain structural phenotypes
estimated through ICA and all symptom phenotypes
in the autism group, we used CCA. In the considered
scenario, CCA is able simultaneously to learn linear
projections (via canonical coefficients) of the brain
structural sources and the behavioral measures that
maximize the correlation between them at the participant Page 6 of 13 Mei et al. Molecular Autism (2020) 11:86 Mei et al. Molecular Autism (2020) 11:86 independent of the model order choice. For details, see
Additional file 1: Subsection 7. In Fig. 1, we present summary images reflecting the
brain areas involved in the structural variances occurring
at these two components. The top row of Fig. 1 shows
that IC10 primarily relates to structural variation in the
bilateral insula, inferior frontal gyrus (IFG), orbitofrontal
cortex (OFC), and caudate nuclei. Among these brain
regions, the bilateral caudate exhibits alterations in the
opposite direction to the others. Given the negative beta
coefficient obtained from the GLM analysis between
participant loadings at IC10 and the diagnosis group
labels, individuals with autism demonstrate increased
GM densities in the bilateral caudate and decreased
densities in the bilateral insula, IFG, and OFC. The
bottom row of Fig. Group effect on ICA decompositionh 1 shows that IC14 mainly involves
variations in the bilateral amygdala, hippocampus, and
parahippocampal gyrus (PHG). Similarly, according to
the sign of the beta values obtained through the GLM,
the autism group shows decreased densities in the areas
involved in IC14. i
To validate the ICA results not being biased by low IQ
participants, an additional validation was performed by
taking FSIQ into account to exclude the participants in
Schedule D from the ICA factorization. This showed that
a unique IC, corresponding to the original IC10, survived
FDR correction (see Additional file 1: Subsection 3). Further, the effect of comorbidity with ADHD on brain
structural variations was determined using data from 500
participants (for detailed demographic information, see
Additional file 1: Subsection 3). This analysis showed that
IC14 remained significantly associated with the autism
group (FDR corrected, p = 9.669 × 10–4). However, IC10
was no longer associated with autism (p = 0.004). Further, post-hoc GLM analyses of the relationships
between brain ICs and symptom ratings did not
provide any significant associations (Additional file 1:
Subsection 8). The robustness of the ICA results to the model order
choice was evaluated by considering, in addition to the
original 100-dimensional factorization, an automatic
dimensionality estimation procedure resulting in a
91-dimensional factorization, and a 50-dimensional
factorization. We observed that the main components
reported (IC10 and IC14) are highly reproducible Relating gray matter spatial variation patterns
to symptoms profilesi The regions involved in IC14 were mentioned
above (canonical weight: − 0.312). Since higher scores
of the ADI and ADOS reflect more severe autism symptoms, positive values of IC16 suggest that higher
loading on this component is related to more severe
symptoms in autism, and negative values of IC61, IC89,
and IC14 meant that lower loadings on these three ICs
are associated with more severe symptoms. In Fig. 2a,
participants were color coded according to their ADOS-
RRB scores to illustrate how the ADOS-RRB score drives
the canonical correlation. In CCA2 we linked SRS, RBS, and SSP scores to the
brain measures of 194 individuals with autism, which is
55.9% of all participants with autism (lower number due
to missing questionnaire data). We found a Bonferroni
corrected (p = 0.05/3 = 0.017) significant relationship
(Fig. 2b, r = 0.840, permutation p = 0.002, Fig. 2d). In this
main CCA mode, IC82, IC99, and IC100 were the highest
contributors to the correlation with behavior profiles, and
SSP score loaded most on the association with the brain
measures in the autism group (Fig. 3c, d). IC82 mainly Fig. 2 The first row shows the scatterplot of the main CCA mode of the brain structural covariations versus the symptom profiles for CCA1 and CCA
2 respectively. One dot per participant in each graph is coded with gradient colors according to the scores of ADOS RRB (a) and SSP (b), respectively. The second row shows the histograms of the null distribution of correlation values obtained from the main CCA mode at 10,000 random
participants’ permutations in the autism sample with ADI and ADOS scores (c), and with SRS, RBS, and SSP scores (d). The true r-value is marked by
a red cross. ADI, Autism Diagnostic Interview-Revised; ADOS, Autism Diagnostic Observational Schedule 2; SRS, Social Responsiveness Scale 2nd
Edition; RBS, Repetitive Behavior Scale-Revised; SSP, Short Sensory Profile; RRB, restricted and repetitive behaviors i i Fig. 2 The first row shows the scatterplot of the main CCA mode of the brain structural covariations versus the symptom profiles for CCA1 and CCA
2 respectively. One dot per participant in each graph is coded with gradient colors according to the scores of ADOS RRB (a) and SSP (b), respectively. Relating gray matter spatial variation patterns
to symptoms profilesi As a final step, we applied CCA to examine the
associations between the 100 components and the two
sets of clinical measures among the autism cases only. The
CCA1 (linking ADI and ADOS subscale scores to brain Fig. 1 The components showed significant case–control differences. The component maps were thresholded at 3 < |Z| < 5 . IC10, component
number 10; IC14, component number 14 Fig. 1 The components showed significant case–control differences. The component maps were thresholded at 3 < |Z| < 5 . IC10, component
number 10; IC14, component number 14 Page 7 of 13 Mei et al. Molecular Autism (2020) 11:86 Mei et al. Molecular Autism measures), involved 325 autism participants and showed
a Bonferroni corrected (p = 0.05/5 = 0.010) significant
relationship (Fig. 2a,c, r = 0.701, permutation p = 0.008). In this main CCA mode, IC16, IC61, IC89, and IC14
were the highest contributors to the correlation with
autism symptoms, and the ADOS RRB subscale loaded
most on the association with the brain measures (Fig. 3a,
b). Among the four components, IC16 mainly involved
density variations in bilateral thalamus and putamen
(canonical weight: 0.447), IC61 in right lateral occipital
and left superior parietal lobe (canonical weight: − 0.366),
and IC89 in the right precentral gyrus (canonical weight:
− 0.333). For details, see Additional file 1: Subsection 9. Note that IC14 is among the components previously
reported showing linear significant case–control group
effects. The regions involved in IC14 were mentioned
above (canonical weight: − 0.312). Since higher scores
of the ADI and ADOS reflect more severe autism measures), involved 325 autism participants and showed
a Bonferroni corrected (p = 0.05/5 = 0.010) significant
relationship (Fig. 2a,c, r = 0.701, permutation p = 0.008). In this main CCA mode, IC16, IC61, IC89, and IC14
were the highest contributors to the correlation with
autism symptoms, and the ADOS RRB subscale loaded
most on the association with the brain measures (Fig. 3a,
b). Among the four components, IC16 mainly involved
density variations in bilateral thalamus and putamen
(canonical weight: 0.447), IC61 in right lateral occipital
and left superior parietal lobe (canonical weight: − 0.366),
and IC89 in the right precentral gyrus (canonical weight:
− 0.333). For details, see Additional file 1: Subsection 9. Note that IC14 is among the components previously
reported showing linear significant case–control group
effects. Relating gray matter spatial variation patterns
to symptoms profilesi The second row shows the histograms of the null distribution of correlation values obtained from the main CCA mode at 10,000 random
participants’ permutations in the autism sample with ADI and ADOS scores (c), and with SRS, RBS, and SSP scores (d). The true r-value is marked by
a red cross. ADI, Autism Diagnostic Interview-Revised; ADOS, Autism Diagnostic Observational Schedule 2; SRS, Social Responsiveness Scale 2nd
Edition; RBS, Repetitive Behavior Scale-Revised; SSP, Short Sensory Profile; RRB, restricted and repetitive behaviors Mei et al. Molecular Autism (2020) 11:86 Page 8 of 13 Mei et al. Molecular Autism Fig. 3 The top row shows the corrected canonical coefficients (weights) of the main CCA mode for the CCA1 analyses (ADI&ADOS), and the bottom
row for the CCA2 analyses (SRS&RBS&SSP). a, c display the degree that each brain component contributed to the main CCA mode in each analysis. The two components with significant group effects are displayed in red. b, d display the degree that each symptom profile contributes to each
analysis. CCA, canonical correlation analysis; ADI, Autism Diagnostic Interview-Revised; ADOS, Autism Diagnostic Observational Schedule 2; SRS,
Social Responsiveness Scale 2nd Edition; RBS, Repetitive Behavior Scale-Revised; SSP, Short Sensory Profile Fig. 3 The top row shows the corrected canonical coefficients (weights) of the main CCA mode for the CCA1 analyses (ADI&ADOS), and the bottom
row for the CCA2 analyses (SRS&RBS&SSP). a, c display the degree that each brain component contributed to the main CCA mode in each analysis. The two components with significant group effects are displayed in red. b, d display the degree that each symptom profile contributes to each
analysis. CCA, canonical correlation analysis; ADI, Autism Diagnostic Interview-Revised; ADOS, Autism Diagnostic Observational Schedule 2; SRS,
Social Responsiveness Scale 2nd Edition; RBS, Repetitive Behavior Scale-Revised; SSP, Short Sensory Profile Fig. 3 The top row shows the corrected canonical coefficients (weights) of the main CCA mode for the CCA1 analyses (ADI&ADOS), and the bottom
row for the CCA2 analyses (SRS&RBS&SSP). a, c display the degree that each brain component contributed to the main CCA mode in each analysis. The two components with significant group effects are displayed in red. b, d display the degree that each symptom profile contributes to each
analysis. Discussionh The present study used a reliable approach to quantify
inter-individual
differences
in
GM
morphometry
covariations in a deeply phenotyped large sample of
individuals with and without autism. The standard,
univariate VBM analysis did not show significant
case–control differences. We then utilized an ICA
decomposition of all participants GM density images,
and similarly performed a case–control post-hoc
statistical analyses. This analysis showed that autism
was significantly associated with alterations in two
independent sources of GM density covariations. These
findings corroborated our hypothesis that alterations
in brain morphometry in autism are associated with
the combined influence of several regions rather than
with a single region. In a following step, we applied
CCA to explore multivariate associations between
sets of continuous measures of core symptoms and
sets of ICA-derived morphometry measures within
the autism group, and were able to identify significant
relationships between brain components and symptom
profiles. Notably, one of the components which showed
significant case–control differences was also among the
highest loading components in the CCA. The brain regions with high loadings on either of these
two components, i.e., insula, amygdala, hippocampus and
PHG have lower densities in autism and have earlier been
associated with autism [9, 43]. The opposite direction of
the alteration of the caudate nucleus in autism has also
previously been found [44]. This is not the case for the
IFG and OFG, which showed lower densities in autism
in our study, where prior studies found mixed results
[45, 46]. Importantly, the brain regions identified by our
analyses have earlier been implicated in the neurobiology
and/or neurocognition of autism. In IC10, structural
and/or functional alterations of the insula, IFG, and
OFC have been associated with social and non-social
cognitive impairments in autism [46–49]. A meta-
analysis reported abnormal functional activations of
the insula, IFG, and OFG during social cognition tasks
in autism [50]. Additionally, variance of the caudate
nucleus volume was found to correlate with the severity
of RRB symptoms in autism [44]. Together with deviant
structural and functional connectivity between frontal
cortical areas and striatum in autism [47, 51, 52],
structural covariation in striatum and frontal areas may
underlie atypical functional fronto-striatal connectivity,
and this has been associated with repetitive behavior and
executive functioning impairments in autism [3, 45]. Relating gray matter spatial variation patterns
to symptoms profilesi CCA, canonical correlation analysis; ADI, Autism Diagnostic Interview-Revised; ADOS, Autism Diagnostic Observational Schedule 2; SRS,
Social Responsiveness Scale 2nd Edition; RBS, Repetitive Behavior Scale-Revised; SSP, Short Sensory Profile interpretation (referring to uncorrected coefficients)
of the CCA2 weights ranks IC14 as the third strongest
contributor to this canonical correlation (Additional
file 1: Subsection 10).h involved variations in the bilateral cerebellum (canonical
weight: 0.414), IC99 in the left lateral occipital and
parietal lobe, and bilateral precentral gyrus (canonical
weight: 0.277), and IC100 in the left inferior frontal gyrus
and right middle frontal lobe (canonical weight: 0.262). For details, see Additional file 1: Subsection 9. Similarly,
lower loadings on these three ICs were related to more
severe symptoms. In Fig. 2b, each participant was color
coded according to their SSP score, and it shows how SSP
score drives the correlation. In this case, both IC10 and
IC14 ranked outside the top 20 of the 100 components,
suggesting that these two components with significant
case–control difference have no strong contribution to
the CCA2 correlation. However, for completeness, direct i
The CCA robustness analyses indicated that the
main CCA modes of both CCA analyses were reliably
estimated in a leave-one-subject out setting (Additional
file 1: Subsection 4). In CCA1, the weights of the main
CCA mode of each leave-one-out analysis correlated
on average above 0.94 with the weights of original
main CCA mode in brain loadings and above 0.95 in
behavior phenotypes when the sample was bigger than
122 participants. In CCA2, the weights of the main CCA
mode related on average above 0.92 in brain loadings Page 9 of 13 Mei et al. Molecular Autism (2020) 11:86 Page 9 of 13 Mei et al. Molecular Autism (2020) 11:86 brain areas. This is in line with a previous study that
used a similar approach [40]. It is further in line with
organizing principles of the brain that regions tend to be
more interconnected when they are located close to each
other [41, 42]. However, when we compared the regions
loading on the two components, the covarying regions of
each component distribute in different brain locations. This suggests that neuroanatomic alterations underlying
autism are more widely distributed at the whole brain
level. It is of note that, when accounting for ADHD
comorbidity, IC14 remained significant but IC10 did not. Relating gray matter spatial variation patterns
to symptoms profilesi This suggests that IC14 is more specifically related to
autism associated structural variations, even after linearly
accounting for ADHD effects, while IC10 might reflect
variations associated with both autism and ADHD.h and above 0.96 in behavior profiles when the sample
was bigger than 111 participants. Both CCA analyses
are no reproducible for sample sizes smaller than
(approximately) 100 participants. Discussionh In
the present study, the density of caudate nuclei increase
simultaneously with densities decreasing on other areas
in autism, which fits with the results of a few functional
studies that indicate inverse functional changes of
these areas [53]. Particularly, the special pattern of GM
densities changes in frontal and striatal areas might serve
an important role in autism-related symptoms. We found nonsignificant case–control difference in
univariate analysis in current study, which is inconsistent
with previous studies (e.g., [8, 9]). In light of the
substantial biological heterogeneity among autistic
individuals, it is expected that findings differ across
datasets, especially those of smaller size. Specifically,
the current study included the individuals with a relative
wide age and IQ range, and did not exclude co-occurring
psychiatric symptoms. Additionally, integrating the
findings from multivariate analyses in our study, the
absence of diagnostic differences at single region/voxel
underscores the detection sensitivity of group effect of
a multivariate approach which evidently verified our
hypothesis. Our findings showed two covarying sets of brain
areas that structurally differed between cases and
controls. While one source of GM density covariation,
IC10, mainly related to the bilateral insula, IFG, OFG,
and caudate, another source, IC14, primarily involved
the bilateral amygdala, hippocampus, and PHG. The
brain regions within each component are anatomically
clustered and symmetrical, which indicates that the
independent structural covariation alteration in the GM
of individuals with autism is concentrated in nearby In IC14, we found decreased densities of amygdala
and hippocampus, where the structural alterations
have previously been related to social deficits in
autism [6, 54]. The amygdala, hippocampus, and
PHG subserve cognitive and emotional functions that
were found abnormal in individuals with autism [50, Page 10 of 13 Mei et al. Molecular Autism (2020) 11:86 Mei et al. Molecular Autism (2020) 11:86 55, 56]. In addition to being involved in emotion and
face processing, the three areas have been proposed
as structures critical for working memory in autism
[57]. Furthermore, these cognitive domains exert
bidirectional effects on each other, with atypical social-
emotional processing influencing memory performance
in individuals with autism, and memory being involved
in complex information processing and executive
functioning, which in turn affects social cognition [57,
58]. Strengths and limitationsh The strengths of our study are the use of a large deeply
phenotyped sample, bottom-up data-driven analyses,
a multivariate approach for examining brain-symptom
associations, and a large set of continuous measures of
core autism symptoms. There are several limitations
in our study. First, only 55.9% of the autism group had
complete questionnaire data on continuous parent-
reported symptom measures, which may have lowered
statistical power for this analysis. Second, due to
the clinical characteristic of autism, the participants
showed significant differences in the proportion of sex
distribution. Despite this we regressed out the sex effect
in the group difference analyses, future studies may
consider balancing the sex distributions in case and
control groups; however this may be difficult to achieve. Third, albeit we ran a series of sensitivity and control
analyses, the effect of other potential sources of variance,
such as the complex effects of medication use that may
influence our findings. Our multivariate correlation analyses moved from
the case–control comparison to the use of continuous
symptoms
among
individuals
with
autism
and
identified two prominent relationships between all
structural brain covariances and symptoms in autism. Three of the four brain components that ranked top
in this analysis did not show case–control differences,
while there was one component (IC14) that differed
between cases and controls and also significantly
correlated with the severity of autism symptoms
assessed by ADI and ADOS. The brain areas loading
high on the brain components identified in the CCA
are somewhat different from those implicated in the
case–control analyses. These former brain areas are
the thalamus, putamen, precentral gyrus, and parietal
and occipital lobes in CCA1, and the cerebellum,
frontal lobe, and again precentral gyrus, and parietal
and occipital lobes in CCA2. These are foremost areas
of the brain implicated in the processing and higher
order integration of sensory information and motor
functions. This makes sense since repetitive, rigid and
stereotyped behaviors and abnormal sensory behaviors
seem to drive the brain-behavior associations much
more than the measures on social-communication
symptoms. Note that variance within the different
autism symptom domains (social-communication,
repetitive behaviors and sensory abnormalities) was
similar and cannot explain the differential symptom-
brain associations. Supplementary information Supplementary information accompanies this paper at https://doi. org/10.1186/s13229-020-00389-4. Additional file 1: Subsection 1. Acquisition parameters. Subsection 2. Demographic information of each schedule. Subsection 3. Sensitivity
analyses. Subsection 4. Leave-one-out (LOO) validation of CCA
results. Subsection 5. Group differences at voxel-wise gray matter vol
umes. Subsection 6. Independent components with case-control
differences. Subsection 7. Robustness assessment of ICA model orders. Subsection 8. GLM results of the association between brain components
and symptom profiles. Subsection 9. Components with highest loadings
in CCA. Subsection 10. Uncorrected main CCA mode loadings of each
component. Additional file 1: Subsection 1. Acquisition parameters. Subsection 2. Demographic information of each schedule. Subsection 3. Sensitivity
analyses. Subsection 4. Leave-one-out (LOO) validation of CCA
results. Subsection 5. Group differences at voxel-wise gray matter vol
umes. Subsection 6. Independent components with case-control
differences. Subsection 7. Robustness assessment of ICA model orders. Subsection 8. GLM results of the association between brain components
and symptom profiles. Subsection 9. Components with highest loadings
in CCA. Subsection 10. Uncorrected main CCA mode loadings of each
component. Overall, the results of our multivariate analyses on
case–control differences and on continuous measures of
symptom severity among those with autism demonstrate
the complexity of the brain morphometry correlates of
autism. Brain areas involved are scattered across the
whole brain and include the limbic system, basal ganglia,
thalamus, cerebellum, precentral (motor) gyrus, and
parts of the frontal, parietal, and occipital lobes. Discussionh In sum, given the potential functional interactions
between these three brain areas, the structural
covariance alterations of the amygdala, hippocampus,
and PHG found in our study, may underlie or contribute
to abnormal functional connections of these areas, and
thus underlie poor performance on social cognition
and memory tasks in individuals with autism. Conclusions We
demonstrate
brain
morphometry
differences
between individuals with autism and typical controls in
the inter-regional covariation of the insula, frontal area,
caudate, amygdala, hippocampus, and PHG. Further, we
highlight associations between covariation in density
of the thalamus, putamen, precentral gyrus, frontal,
parietal, and occipital lobes, and the cerebellum, and
core autism symptoms, in particular repetitive behaviors
and abnormal sensory behavior. Future studies may link
our morphometry findings with data on brain function
obtained from cognitive tests and/or functional and
resting-state MRI, and with genomics data. Funding
h
k This work is primarily supported by the EU-AIMS consortium (European
Autism Interventions), which receives support from Innovative Medicines
Initiative Joint Undertaking Grant No.115300, the resources of which are
composed of financial contributions from the European Union’s Seventh
Framework Programme (Grant No. FP7/2007-2013), from the European
Federation of Pharmaceutical Industries and Associations companies’ in-kind
contributions; and by the AIMS-2-TRIALS consortium (Autism Innovative
Medicine Studies-2-Trials), which has received funding from the Innovative
Medicines Initiative 2 Joint Undertaking under Grant Agreement No. 777394,
and this Joint Undertaking receives support from the European Union’s
Horizon 2020 research and innovation programme and EFPIA and AUTISM
SPEAKS, Autistica, SFARI. TM is supported by a China Scholarship Council Grant
(No 201806010408). This work has been further supported by the European
Union Seventh Framework Programme Grant Nos. 602805 (AGGRESSOTYPE)
(to JKB), 603016 (MATRICS) (to JKB), and 278948 (TACTICS) (to JKB); European
Community’s Horizon 2020 Programme (H2020/2014–2020) Grant Nos. 643051 (MiND) (to JKB), 642996 (BRAINVIEW) (to JKB) and 847818 (CANDY) (to
JKB and CFB); the Netherlands Organization for Scientific Research VICI Grant
No. 2020/TTW/00836465 (to CFB); Wellcome Trust Collaborative Award Grant
No. 215573/Z/19/Z (to CFB); the Autism Research Trust (to SBC). Received: 25 May 2020 Accepted: 5 October 2020 Consent for publication Independent component; ICA: Independent component analysis; ICD-
10: International Statistical Classification of Diseases and Related Health
Problems 10th Revision; ID: Intellectual disability; IFG: Inferior frontal
gyrus; OFC: Orbitofrontal cortex; MRI: Magnetic resonance imaging; PHG:
Parahippocampal gyrus; RBS: Repetitive Behavior Scale-Revised; RRB:
Restricted and repetitive behaviors; SA: Social affect; SD: Standard deviation;
SRS: Social Responsiveness Scale 2nd Edition; SSP: Short Sensory Profile; TD:
Typically developing; TFCE: Threshold-Free Cluster Enhancement; VBM: Voxel-
based morphometry. Consent for publication was obtained from all participants prior to the study. Acknowledgements We thank all participants and their families for participating in this study. We gratefully acknowledge the contributions of all members of the
EU-AIMS LEAP group: Jumana Ahmad, Sara Ambrosino, Bonnie Auyeung,
Tobias Banaschewski, Simon Baron-Cohen, Sarah Baumeister, Christian
F. Beckmann, Sven Bölte, Thomas Bourgeron, Carsten Bours, Michael
Brammer, Daniel Brandeis, Claudia Brogna, Yvette de Bruijn, Jan K. Buitelaar,
Bhismadev Chakrabarti, Tony Charman, Ineke Cornelissen, Daisy Crawley,
Flavio Dell’Acqua, Guillaume Dumas, Sarah Durston, Christine Ecker, Jessica
Faulkner, Vincent Frouin, Pilar Garcés, David Goyard, Lindsay Ham, Hannah
Hayward, Joerg Hipp, Rosemary Holt, Mark H. Johnson, Emily J.H. Jones, Prantik
Kundu, Meng-Chuan Lai, Xavier Liogier D’ardhuy, Michael V. Lombardo, Eva
Loth, David J. Lythgoe, René Mandl, Andre Marquand, Luke Mason, Maarten
Mennes, Andreas Meyer-Lindenberg, Carolin Moessnang, Nico Mueller, Declan
G.M. Murphy, Bethany Oakley, Laurence O’Dwyer, Marianne Oldehinkel, Bob
Oranje, Gahan Pandina, Antonio M. Persico, Annika Rausch, Barbara Ruggeri,
Amber Ruigrok, Jessica Sabet, Roberto Sacco, Antonia San José Cáceres,
Emily Simonoff, Will Spooren, Julian Tillmann, Roberto Toro, Heike Tost, Jack
Waldman, Steve C.R. Williams, Caroline Wooldridge, and Marcel P. Zwiers. References 1. American Psychiatric Association. Diagnostic and statistical manual of
mental disorders. 5th ed. Arlington: Author; 2013. 2. Ecker C, Bookheimer SY, Murphy DG. Neuroimaging in autism spectrum
disorder: brain structure and function across the lifespan. Lancet Neurol. 2015;14(11):1121–34. 3. Di Martino A, Yan CG, Li Q, Denio E, Castellanos FX, Alaerts K, et al. The autism brain imaging data exchange: towards a large-scale
evaluation of the intrinsic brain architecture in autism. Mol Psychiatry. 2014;19(6):659–67. 4. Turner AH, Greenspan KS, van Erp TGM. Pallidum and lateral ventricle
volume enlargement in autism spectrum disorder. Psychiatry Res
Neuroimaging. 2016;252:40–5. 5. Zhang W, Groen W, Mennes M, Greven C, Buitelaar J, Rommelse N. Revisiting subcortical brain volume correlates of autism in the ABIDE
dataset: effects of age and sex. Psychol Med. 2018;48(4):654–68. Author details 1 Department of Cognitive Neuroscience, Donders Institute for Brain,
Cognition and Behaviour, Radboud University Nijmegen Medical Centre,
Nijmegen, The Netherlands. 2 Karakter Child and Adolescent Psychiatry
University Centre, Nijmegen, The Netherlands. 3 Department of Psychology,
Institute of Psychiatry, Psychology and Neuroscience, King’s College
London, London, UK. 4 Department of Psychiatry, Brain Center Rudolf
Magnus, University Medical Center Utrecht, Utrecht, The Netherlands. 5 Department of Psychiatry and Psychotherapy, Central Institute of Mental
Health, Medical Faculty Mannheim, University of Heidelberg, Mannheim,
Germany. 6 Department of Child and Adolescent Psychiatry, Central Institute
of Mental Health, Medical Faculty Mannheim, University of Heidelberg,
Mannheim, Germany. 7 Autism Research Centre, Department of Psychiatry,
University of Cambridge, Cambridge, UK. 8 Department of Forensic
and Neurodevelopmental Sciences, Institute of Psychiatry, Psychology
and Neuroscience, King’s College London, London, UK. 9 Department of Child
and Adolescent Psychiatry, University Hospital, Goethe University, Frankfurt am
Main, Germany. 10 Centre for Functional MRI of the Brain, University of Oxford,
Oxford, UK. Competing interests p
g
JKB has been a consultant to, advisory board member of, and a speaker for
Janssen Cilag BV, Eli Lilly, Shire, Lundbeck, Roche, and Servier. He is not an
employee of any of these companies, and not a stock shareholder of any of
these companies. He has no other financial or material support, including
expert testimony, patents or royalties. CFB is director and shareholder in
SBGNeuro Ltd. TB served in an advisory or consultancy role for Lundbeck,
Medice, Neurim Pharmaceuticals, Oberberg GmbH, Shire, and Infectopharm. He received conference support or speaker’s fee by Lilly, Medice, and Shire. He received royalties from Hogrefe, Kohlhammer, CIP Medien, and Oxford
University Press. TC has received consultancy from Roche and received book
royalties from Guildford Press and Sage. DGM has been a consultant to, and
advisory board member, for Roche and Servier. He is not an employee of any
of these companies, and not a stock shareholder of any of these companies. The present work is unrelated to the above grants and relationships. The
other authors report no biomedical financial interests or potential conflicts of
interest. Authors’ contributions JT, SD, CM, TB, RH, SB-C, AR, EL, TC, DGM, CE, CFB, JKB designed the EU-AIMS
study, developed data acquisition and/or analysis protocols. TM ran analyses,
visualized the findings, wrote the first draft, and revised the draft; AL
conceptualized and supervised the analysis, and revised the manuscript;
DLF ran the VBM analysis and revised the draft; JT & NJF revised the draft;
CFB supervised and revised the manuscript; JKB supervised and revised the
manuscript. All authors read and approved the final manuscript. Abbreviations ADHD: Attention Deficit Hyperactivity Disorder; ADI: Autism Diagnostic
Interview-Revised; ADOS: Autism Diagnostic Observational Schedule 2;
CCA: Canonical correlation analysis; DSM-IV: Diagnostic and Statistical
Manual-IV; EU-AIMS: European Autism Interventions—A Multicentre Study
for Developing New Medications; FDR: False discovery rate; FSIQ: Full-scale
intelligence quotient; GLM: Generalized linear model; GM: Gray matter; IC: Page 11 of 13 Page 11 of 13 Mei et al. Molecular Autism (2020) 11:86 Mei et al. Molecular Autism (2020) 11:86 Consent for publication Availability of data and materials Data collected in EU-AIMS LEAP are stored and curated at the central EU-AIMS
database at the Pasteur Institute in Paris. The database is open to members of
the wider scientific community upon request and submission of a paper and
data analytic proposal. 6. van Rooij D, Anagnostou E, Arango C, Auzias G, Behrmann M, Busatto
GF, et al. Cortical and subcortical brain morphometry differences
between patients with autism spectrum disorder and healthy individuals
across the lifespan: results from the ENIGMA ASD working group. Am J
Psychiatry. 2018;175(4):359–69. Ethics approval and consent to participate The disrupted connectivity hypothesis
of autism spectrum disorders: time for the next phase in research. Biol
Psychiatry Cogn Neurosci Neuroimaging. 2016;1(3):245–52. 37. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a
practical and powerful approach to multiple testing. J R Stat Soc Ser B
(Methodolo). 1995;57(1):289–300. y
y
g
g g
13. Geschwind DH, Levitt P. Autism spectrum disorders: developmental
disconnection syndromes. Curr Opin Neurobiol. 2007;17(1):103–11. 14. Picci G, Gotts SJ, Scherf KS. A theoretical rut: revisiting and critically
evaluating the generalized under/over-connectivity hypothesis of autism. Dev Sci. 2016;19(4):524–49. 38. Smith SM, Nichols TE, Vidaurre D, Winkler AM, Behrens TE, Glasser
MF, et al. A positive-negative mode of population covariation links
brain connectivity, demographics and behavior. Nat Neurosci. 2015;18(11):1565–7. 15. Beckmann CF, Smith SM. Probabilistic independent component analysis
for functional magnetic resonance imaging. IEEE Trans Med Imaging. 2004;23(2):137–52. 39. Haufe S, Meinecke F, Gorgen K, Dahne S, Haynes JD, Blankertz B, et al. On the interpretation of weight vectors of linear models in multivariate
neuroimaging. Neuroimage. 2014;87:96–110. 16. Xu L, Groth KM, Pearlson G, Schretlen DJ, Calhoun VD. Source-based
morphometry: the use of independent component analysis to identify
gray matter differences with application to schizophrenia. Hum Brain
Mapp. 2009;30(3):711–24. 40. Di X, Biswal BB. Similarly expanded bilateral temporal lobe volumes in
female and male children with autism spectrum disorder. Biol Psychiatry
Cogn Neurosci Neuroimaging. 2016;1(2):178–85. 41. Beul SF, Grant S, Hilgetag CC. A predictive model of the cat cortical
connectome based on cytoarchitecture and distance. Brain Struct Funct. 2015;220(6):3167–84. 17. Hafkemeijer A, Moller C, Dopper EG, Jiskoot LC, van den Berg-Huysmans
AA, van Swieten JC, et al. Differences in structural covariance brain
networks between behavioral variant frontotemporal dementia and
Alzheimer’s disease. Hum Brain Mapp. 2016;37(3):978–88. 42. Oligschlager S, Xu T, Baczkowski BM, Falkiewicz M, Falchier A, Linn G, et al. Gradients of connectivity distance in the cerebral cortex of the macaque
monkey. Brain Struct Funct. 2019;224(2):925–35. 18. Vanasse TJ, Fox PM, Barron DS, Robertson M, Eickhoff SB, Lancaster JL,
et al. BrainMap VBM: An environment for structural meta-analysis. Hum
Brain Mapp. 2018;39(8):3308–25. 43. Nickl-Jockschat T, Habel U, Michel TM, Manning J, Laird AR, Fox PT, et al. Brain structure anomalies in autism spectrum disorder–a meta-analysis
of VBM studies using anatomic likelihood estimation. Hum Brain Mapp. 2012;33(6):1470–89. 19. Ecker C, Suckling J, Deoni SC, Lombardo MV, Bullmore ET, Baron-Cohen S,
et al. Ethics approval and consent to participate Ethical approval for this study was obtained through ethics committees at
each site. Ethical approval for this study was obtained through ethics committees at
each site. 7. Bedford SA, Park MTM, Devenyi GA, Tullo S, Germann J, Patel R,
et al. Large-scale analyses of the relationship between sex, age and Page 12 of 13 Page 12 of 13 Mei et al. Molecular Autism (2020) 11:86 Mei et al. Molecular Autism (2020) 11:86 31. Ashburner J. A fast diffeomorphic image registration algorithm. Neuroimage. 2007;38(1):95–113. intelligence quotient heterogeneity and cortical morphometry in autism
spectrum disorder. Mol Psychiatry. 2019;25(3):614–28. intelligence quotient heterogeneity and cortical morphometry in autism
spectrum disorder. Mol Psychiatry. 2019;25(3):614–28. p
y
y
8. DeRamus TP, Kana RK. Anatomical likelihood estimation meta-analysis
of grey and white matter anomalies in autism spectrum disorders. Neuroimage Clin. 2015;7:525–36. y
y
8. DeRamus TP, Kana RK. Anatomical likelihood estimation meta-analysis
of grey and white matter anomalies in autism spectrum disorders. Neuroimage Clin. 2015;7:525–36. 32. Groves AR, Smith SM, Fjell AM, Tamnes CK, Walhovd KB, Douaud G, et al. Benefits of multi-modal fusion analysis on a large-scale dataset: life-span
patterns of inter-subject variability in cortical morphometry and white
matter microstructure. Neuroimage. 2012;63(1):365–80. 9. Carlisi CO, Norman LJ, Lukito SS, Radua J, Mataix-Cols D, Rubia K. Comparative multimodal meta-analysis of structural and functional brain
abnormalities in autism spectrum disorder and obsessive-compulsive
disorder. Biol Psychiatry. 2017;82(2):83–102. 33. Jenkinson M, Beckmann CF, Behrens TE, Woolrich MW, Smith SM. Fsl
Neuroimage. 2012;62(2):782–90. 34. Eklund A, Nichols TE, Knutsson H. Cluster failure: why fMRI inferences for
spatial extent have inflated false-positive rates. Proc Natl Acad Sci USA. 2016;113(28):7900–5. 10. Liu J, Yao L, Zhang W, Xiao Y, Liu L, Gao X, et al. Gray matter abnormalities
in pediatric autism spectrum disorder: a meta-analysis with signed
differential mapping. Eur Child Adolesc Psychiatry. 2017;26(8):933–45. 35. Winkler AM, Ridgway GR, Douaud G, Nichols TE, Smith SM. Faster
permutation inference in brain imaging. Neuroimage. 2016;141:502–16. f
11. Cai J, Hu X, Guo K, Yang P, Situ M, Huang Y. Increased left inferior temporal
gyrus was found in both low function autism and high function autism. Front Psychiatry. 2018;9:542. 36. Simonoff E, Pickles A, Charman T, Chandler S, Loucas T, Baird G. Psychiatric
disorders in children with autism spectrum disorders: prevalence,
comorbidity, and associated factors in a population-derived sample. J Am
Acad Child Adolesc Psychiatry. 2008;47(8):921–9. y
y
12. Vasa RA, Mostofsky SH, Ewen JB. Ethics approval and consent to participate Brain anatomy and its relationship to behavior in adults with autism
spectrum disorder: a multicenter magnetic resonance imaging study. Arch Gen Psychiatry. 2012;69(2):195–209. 44. Langen M, Bos D, Noordermeer SD, Nederveen H, van Engeland H,
Durston S. Changes in the development of striatum are involved in
repetitive behavior in autism. Biol Psychiatry. 2014;76(5):405–11. 20. Smith SM, Nichols TE. Statistical challenges in “big data” human
neuroimaging. Neuron. 2018;97(2):263–8. 21. Mulders P, Llera A, Tendolkar I, van Eijndhoven P, Beckmann C. Personality
profiles are associated with functional brain networks related to
cognition and emotion. Sci Rep. 2018;8(1):13874. 45. Foster NE, Doyle-Thomas KA, Tryfon A, Ouimet T, Anagnostou E, Evans AC,
et al. Structural gray matter differences during childhood development
in autism spectrum disorder: a multimetric approach. Pediatr Neurol. 2015;53(4):350–9. 22. Hotelling H. Relations between two sets of variates. Biometrika. 1936;28:312–77. 46. Sato W, Kochiyama T, Uono S, Yoshimura S, Kubota Y, Sawada R, et al. Reduced gray matter volume in the social brain network in adults with
autism spectrum disorder. Front Hum Neurosci. 2017;11:395. 23. Loth E, Charman T, Mason L, Tillmann J, Jones EJH, Wooldridge C, et al. The EU-AIMS Longitudinal European Autism Project (LEAP): design and
methodologies to identify and validate stratification biomarkers for
autism spectrum disorders. Mol Autism. 2017;8:24. 47. Kana RK, Keller TA, Minshew NJ, Just MA. Inhibitory control in high-
functioning autism: decreased activation and underconnectivity in
inhibition networks. Biol Psychiatry. 2007;62(3):198–206. 24. Charman T, Loth E, Tillmann J, Crawley D, Wooldridge C, Goyard D, et al. The EU-AIMS Longitudinal European Autism Project (LEAP): clinical
characterisation. Mol Autism. 2017;8:27. y
y
48. Wadsworth HM, Maximo JO, Donnelly RJ, Kana RK. Action simulation and
mirroring in children with autism spectrum disorders. Behav Brain Res. 2018;341:1–8. 25. Rutter M, Le Couteur A, Lord C. Autism diagnostic interview-revised. Los
Angeles: Western Psychological Services; 2003. 49. Nomi JS, Molnar-Szakacs I, Uddin LQ. Insular function in autism:
Update and future directions in neuroimaging and interventions. Prog
Neuropsychopharmacol Biol Psychiatry. 2019;89:412–26. 26. Lord C, Rutter M, DiLavore PC, Risi S, Gotham K, Bishop S. Autism
diagnosticobservation schedule, Second Edition (ADOS-2) manual (part
I): modules 1–4. Torrance: Western Psychological Services; 2012. 50. Patriquin MA, DeRamus T, Libero LE, Laird A, Kana RK. Neuroanatomical
and neurofunctional markers of social cognition in autism spectrum
disorder. Hum Brain Mapp. 2016;37(11):3957–78. y
g
27. Constantino JN, Gruber CP. Social responsivenes
Angeles: Western Psychological Services; 2012. y
g
27. Ethics approval and consent to participate Constantino JN, Gruber CP. Social responsiveness scale. 2nd ed. Los
Angeles: Western Psychological Services; 2012. Angeles: Western Psychological Services; 2012. 51. Langen M, Leemans A, Johnston P, Ecker C, Daly E, Murphy CM, et al. Fronto-striatal circuitry and inhibitory control in autism: findings from
diffusion tensor imaging tractography. Cortex. 2012;48(2):183–93. 28. Bodfish JW, Symons FJ, Parker DE, Lewis MH. Varieties of repetitive
behavior in autism: comparisons to mental retardation. J Autism Dev
Disord. 2000;30(3):237–43. 29. Tomchek SD, Dunn W. Sensory processing in children with and without
autism: a comparative study using the short sensory profile. Am J Occup
Ther. 2007;61(2):190–200. 52. Di Martino A, Kelly C, Grzadzinski R, Zuo XN, Mennes M, Mairena MA,
et al. Aberrant striatal functional connectivity in children with autism. Biol
Psychiatry. 2011;69(9):847–56. 30. Gaser C, Dahnke R. https://dbm.neuro.uni-jena.de/cat/. Accessed 2 Sept
2018. 53. Clements CC, Zoltowski AR, Yankowitz LD, Yerys BE, Schultz RT, Herrington
JD. Evaluation of the social motivation hypothesis of autism: a systematic
review and meta-analysis. JAMA Psychiatry. 2018;75(8):797–808. Mei et al. Molecular Autism (2020) 11:86 Page 13 of 13 Page 13 of 13 56. Klapwijk ET, Aghajani M, Colins OF, Marijnissen GM, Popma A, van Lang
ND, et al. Different brain responses during empathy in autism spectrum
disorders versus conduct disorder and callous-unemotional traits. J Child
Psychol Psychiatry. 2016;57(6):737–47. 57. Boucher J, Mayes A, Bigham S. Memory in autistic spectrum disorder.
Psychol Bull. 2012;138(3):458–96.
58. Barendse EM, Hendriks MP, Jansen JF, Backes WH, Hofman PA, Thoonen
G, et al. Working memory deficits in high-functioning adolescents with
autism spectrum disorders: neuropsychological and neuroimaging
correlates. J Neurodev Disord. 2013;5(1):14. 54. Baribeau DA, Dupuis A, Paton TA, Hammill C, Scherer SW, Schachar RJ,
et al. Structural neuroimaging correlates of social deficits are similar in
autism spectrum disorder and attention-deficit/hyperactivity disorder:
analysis from the POND Network. Transl Psychiatry. 2019;9(1):72. 55. Aoki Y, Cortese S, Tansella M. Neural bases of atypical emotional face
processing in autism: A meta-analysis of fMRI studies. World J Biol
Psychiatry. 2015;16(5):291–300. 54. Baribeau DA, Dupuis A, Paton TA, Hammill C, Scherer SW, Schachar RJ,
et al. Structural neuroimaging correlates of social deficits are similar in
autism spectrum disorder and attention-deficit/hyperactivity disorder:
analysis from the POND Network. Transl Psychiatry. 2019;9(1):72.
55. Aoki Y, Cortese S, Tansella M. Neural bases of atypical emotional face
processing in autism: A meta-analysis of fMRI studies. World J Biol
Psychiatry. 2015;16(5):291–300.
56. Klapwijk ET, Aghajani M, Colins OF, Marijnissen GM, Popma A, van Lang
ND, et al. Different brain responses during empathy in autism spectrum
disorders versus conduct disorder and callous-unemotional traits. J Child
Psychol Psychiatry. 2016;57(6):737–47. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from: •
fast, convenient online submission
•
thorough peer review by experienced researchers in your field
•
rapid publication on acceptance
•
support for research data, including large and complex data types
•
gold Open Access which fosters wider collaboration and increased citations
maximum visibility for your research: over 100M website views per year
•
At BMC, research is always in progress. Learn more biomedcentral.com/submissions
Ready to submit your research
Ready to submit your research ? Choose BMC and benefit from:
? Choose BMC and benefit from:
|
https://openalex.org/W3164256180
|
https://link.springer.com/content/pdf/10.1007/s42978-021-00108-2.pdf
|
English
| null |
Monitoring the Performance of Alpine Skiers with Inertial Motion Units: Practical and Methodological Considerations
|
Journal of science in sport and exercise
| 2,021
|
cc-by
| 7,140
|
Journal of Science in Sport and Exercise (2021) 3:249–256
https://doi.org/10.1007/s42978-021-00108-2 Journal of Science in Sport and Exercise (2021) 3:249–256
https://doi.org/10.1007/s42978-021-00108-2 REVIEW ARTICLE Abstract Although reliable feedback is crucial to improving the performance of competitive alpine skiers, the coach’s eye may not be
sensitive enough to detect small, but highly significant “mistakes”. Monitoring of the performance of alpine ski racers by
inertial motion units (IMU) has proven to be of value in this context and here we summarize practical and methodological
aspects of this approach. Methodologically, the IMUs employed should combine high sampling frequencies with minimal
signal drift. The sensors should be positioned to sense the movement of the bones in a given body segment while being pro-
tected as much as possible against impact with the ski gates. The data obtained, often synchronized with input from Global
Satellite Navigation Systems (GNSS), are usually refined utilizing advanced biomechanical models and other computerized
approaches. In practice, the combination of inertial sensors and GNSS allows accurate monitoring of skiing kinematics
(technique) and the movement of the skier’s center-of-mass, also allowing analysis of both whole-body vibrations (WBV)
and loss of mechanical energy. Presentation of the findings to coaches and athletes can be facilitated by synchronizing them
with video recordings. Recent advances in IMU technology, including miniaturization, wireless communication, direct stor-
age of data in the cloud, and processing with artificial intelligence may allow these sensors, in-combination with GNSS, to
become real-time virtual alpine ski coaches, perhaps the next step in the development of this sport. Keywords Biomechanics · GPS · GLONASS · Kinematics · IMU · Technique Keywords Biomechanics · GPS · GLONASS · Kinematics · IMU · Technique Monitoring the Performance of Alpine Skiers with Inertial Motion
Units: Practical and Methodological Considerations Matej Supej1,2 · H‑C Holmberg2,3,4 Received: 30 June 2020 / Accepted: 7 January 2021 / Published online: 25 May 2021
© The Author(s) 2021 Received: 30 June 2020 / Accepted: 7 January 2021 / Published online: 25 May 2021
© The Author(s) 2021 * Matej Supej
Matej.Supej@fsp.uni-lj.si
1
Faculty of Sport, University of Ljubljana, Gortanova ulica
22, 1000 Ljubljana, Slovenia
2
Swedish Winter Sports Research Centre, Mid Sweden
University, Östersund, Sweden
3
Department of Physiology and Pharmacology, Biomedicum
C5, Karolinska Institutet, Stockholm, Sweden
4
China Institute of Sport and Health Science, Beijing Sport
University, Beijing, China Kinematic Parameters several sports but also in sports medicine and rehabilitation
[67]. IMUs can consist of either a single multi-axis sensor
(typically incorporating 3-axes with respect to accelera-
tion, angular velocity and orientation, in total as many as
9-axes), e.g., for analyzing jumping performance [32, 36,
37], or multiple sensors that record kinematic parameters
of the arms, legs and/or entire body during various forms of
locomotion [29, 68, 69].fii In general, the coaches of alpine skiers focus on their tech-
nique in an attempt to improve performance. Because ski-
ing speeds are high, on average ranging from 54 km/h dur-
ing slalom to 94 km/h during downhill skiing [18], even
movements of different body segments that are too small
for coaches to see can be crucial. The capability to objec-
tively monitor skiing technique, including joint angles, the
inclination of the skier and of his/her body segments (e.g.,
the shank), angular velocities and the positions of differ-
ent body segments while turning and the movement of
the skier’s center of mass over a large part of or even an
entire course is certainly one of the most important fea-
tures of IMU sensors [3, 50, 51]. This information can be
of immense value in connection with attempts to improve
skiing technique [3, 27, 50, 51] or, e.g., assessment of the
impact of skiing equipment on ski safety [72]. Indeed, due to the difficulties involved in filming outdoors
on snow with camcorders (e.g., multiple camcorders must
be positioned to provide unobstructed visibility of the skiers
from at least two angles and appropriate calibration proce-
dures applied), systems based on IMUs are being applied
even more extensively in connection with alpine skiing than
with a number of other sports [3, 27, 51]. The major chal-
lenges associated with such usage include the necessity of
capturing large amounts of complex data on snow (during
multiple runs through several gates or over the entire course)
in a manner that minimizes post-processing, thereby provid-
ing results within a short period of time (in fact, immedi-
ately, if possible). In the case of alpine skiing, inertial sen-
sors were first used as a complement to GNSS technology
[38] and thereafter to monitor whole-body kinematics [3,
27, 51] and, later, more specific aspects of performance
[9, 12, 13, 40, 53, 62, 63]. Kinematic Parameters With inertial sensors, many of
the limitations described above can be overcome, but, like
all approaches, these also have their limitations (discussed
below in the section on methodological considerations). Another important feature of IMU sensors is that they
can be worn on the body in a manner that allows measure-
ment of joint angles in all three planes, e.g., flexion–exten-
sion, abduction–adduction and internal–external rotation
of the knee joint [72]. Although, because of the manner
in which Euler angles are calculated [21], such measure-
ments of abduction–adduction and internal–external rota-
tion during skiing on-snow are less accurate than those
of flexion–extension, they nonetheless provide valuable
information (discussed further in Methodological Consid-
erations). Furthermore, some IMU sensors can be used to
target specific joint angles, for example, at the knee, hip
and/or trunk [13, 14]. Clearly, IMU technology provides a tool for monitoring
all aspects of skiing technique [3, 51], thereby providing a
basis for improving performance [16, 46, 49, 59, 60]. At
the same time, it is crucial that the feedback provided by
this technology is as accurate as possible. Accordingly, this
review focuses on the scientific literature presently avail-
able that deals with practical and methodological consid-
erations related to the evaluation of competitive skiing with
IMU technology. In this respect, the kinematic and kinetic
parameters that can be monitored by IMU technology will
be discussed in the section on practical considerations; while
under methodological considerations, sample size, the valid-
ity, reliability and limitations of such sensors, and their use
together with supplementary technology will be presented. For monitoring skiing technique, IMU sensors on the
body are usually synchronized with the GNSS system,
which enables accurate positioning of the skier on the
course [3, 27, 50, 51]. Alternatively, this position can be
provided by the IMU sensors themselves if magnets are
placed next to the gates, providing overall accuracy and
precision of 0.24 and 0.09 cm, respectively [9]. Knowing
the position of the skier allows more detailed monitoring
of the most important parameters of performance, such
as the movement of the center-of-mass (CoM), trajectory
of skis, turning radius, timing, speed between the gates
and energy losses [16, 44, 45, 49, 59, 60, 70]. Kinematic Parameters Although
such parameters can also be monitored with the GNSS sys-
tem alone [65], the combined use provides more accurate
monitoring of the movements of the CoM and skis [3, 11,
19, 27, 35, 50, 51], particularly key parameters for evalu-
ating performance, such as turning radius, timing, speed
between the gates and energy losses. Introduction coaches must rely on their own visual evaluation, which may
not be sensitive enough to detect small, but significant dif-
ferences in technique. The performance of competitive ski racers, who must adapt
to a variety of slopes and courses under varying conditions
of weather and snow using different types of equipment,
can be improved in many different ways, both major and
minor. In alpine skiing competitions, fractions of a second
can separate the winner from second-best [58], so the effi-
cacy of their training is clearly a key determinant of success. In this context, reliable feedback is crucial, but most of their In this connection, the latest advances in wearable tech-
nology can provide detailed information concerning skiing
technique, racing tactics (e.g., the choice of trajectory), and
performance on-snow, as well as numerous other aspects of
alpine skiing, such as the performance of equipment (skis
and poles), the occurrence of vibrations and their transmis-
sion to the skier’s body. Indeed, one study has already high-
lighted the potential of Global Navigation Satellite Systems
(GNSS) in this respect [65]. However, to our knowledge,
monitoring of alpine skiers utilizing other wearable technol-
ogy, such as inertial motion units (IMU), has not yet been
characterized in detail. * Matej Supej
Matej.Supej@fsp.uni-lj.si 1
Faculty of Sport, University of Ljubljana, Gortanova ulica
22, 1000 Ljubljana, Slovenia To date, optical motion capture systems based on video
filming remain the golden standard in connection with the
three-dimensional kinematic analysis of skiing technique,
trajectory and performance. At the same time, IMU sys-
tems are being employed more and more extensively for
such purposes, not only in connection with performance in 4
China Institute of Sport and Health Science, Beijing Sport
University, Beijing, China (0121 3456789)
3 Journal of Science in Sport and Exercise (2021) 3:249–256 250 Practical Considerations Biomechanical analysis of the techniques used by competi-
tive alpine skiers, as well as various other aspects of their
performance, is often performed for research purposes, as
well as to provide feedback that can both help the athletes
improve their performance and lower their risk of injury [17,
19, 22, 40, 42, 43, 48, 49, 51, 52, 57, 59, 60]. Timing is crucial in connection with competitive alpine
skiing and must be measured accurately and easily over
many sections of the course [54] or even between two 1 3 Journal of Science in Sport and Exercise (2021) 3:249–256 251 consecutive gates [57]. With the appropriate placement
of strong magnets, IMU sensors can also provide gate-to-
gate times or the time that elapses between any two sec-
tions of the course [12]. Usage of GNSS for this purpose
is dependent on good signal reception by satellites and
is thus less accurate in the vicinity of trees or buildings. With both GNSS [57] and IMU sensors, the multiple com-
parisons of gate-to-gate times provided enable much more
detailed analysis than is possible with the standard usage
of photocells. with GNSS to provide a better model of the movement of
the CoM or to capture the full-body 3D-kinematics pro-
vided also by camcorders, as discussed earlier. Most energy dissipation is due to ski-snow friction, espe-
cially during the performance of the technical disciplines
of alpine skiing (i.e., slalom and giant slalom) [64], but
direct measurement of this friction is difficult in practice
and prone to significant error. For example, comparison of
this determination using a differential GNSS to camcorder-
based measurement indicated that the former approach has a
precision of 42 N, with a mean ± standard deviation between
the two procedures of 1 ± 96 N. In light of the facts that a)
due to involvement of the second-order derivative of time
after filtering, camcorder-based estimation of ground reac-
tion forces is already associated with an average error in the
precision of approximately 150 N; and b) during a carv-
ing turn the average ground reaction forces transverse and
parallel to the outer ski range from 17 to 61 N, respectively
[31], the applicability of such approaches in connection with
competitive skiing is limited. Methodological Considerations Although the advantages of utilizing IMU sensors for analy-
sis of alpine skiing are considerable, like all approaches,
this one has certain limitations, specifically with respect to
sample size, the validity and reliability of measurements,
and placement of the sensors. Aerodynamic drag and ski–snow friction exert a pro-
found influence on skiing performance. More specifically,
almost 50% of the difference in the performance of down-
hill skiers is due to air drag, while in the case of giant sla-
lom the corresponding value is approximately 15% [30, 64,
65]. Both camcorder-based systems and GNSS technology
have already been shown to measure aerodynamic drag dur-
ing alpine skiing reliably [33, 64]. Although IMU sensors
have not yet been applied for measuring aerodynamic drag,
this technology could potentially be used in combination Kinetic Parameters Obviously, the forces acting on any object under observa-
tion are fundamental kinetic parameters. In principle, IMU
sensors do not measure forces directly but instead measure
acceleration, among other parameters as clarified above in
the introduction [1]. In connection with alpine skiing, 3D
kinematic systems utilizing camcorders are often employed
to estimate ground reaction forces (GRF) [39, 41, 55, 56,
58]. At the same time, these GRFs have already been deter-
mined using appropriate placement of accelerometers [63]
and this approach, in particular with the usage of multiple
IMU sensors on different body segments, can be expected
to become more and more common. Practical Considerations As a good alternative, total
energy losses in direct relationship to skiing performance
can be monitored in a manner that only requires the first-
order derivative of time [49, 59, 60]. In addition to the information provided by standard cam-
corder-based systems, IMU technology enables monitoring
of whole-body vibrations (WBV) and the transmission of
WBV produced during skiing [10, 15, 40, 53, 62, 63, 66]. By selecting technology in accordance with the ISO 8041
International Standards [25], the WBV acting on the skier
can be evaluated on the basis of general recommendations
for the exposure of workers to such vibrations (International
Standard, ISO 2631-1 [24]). Indeed, WBV were found to
pose a potential risk for low-back pain, which is experienced
by many alpine skiers [63]. Finally, IMU technology enables the collection of much
more data during alpine skiing than can in practice be pro-
vided by camcorder-based systems. This more extensive data
can be utilized, e.g., for more detailed statistical assessment
of parameters connected with turning [72] and to “teach”
artificial intelligence [6] to recommend approaches to
improving skiing technique [35]. Since aerodynamic drag is not the major determinant of
performance in these technical disciplines [64], this fac-
tor should be considered separately from friction. Previous
approaches to the evaluation of aerodynamic drag involved
subtracting drag-related loss of energy from the total energy
losses to obtain the loss of energy due to friction. Instead
of camera- or solely satellite-based systems, inertial sen-
sors could be used for this purpose, preferably in combina-
tion with a real-time kinematic GNSS system to attain high
accuracy over a large area, as some investigators have done
previously [27, 50, 72]. Sample Size One of the major limitations associated with the 3D-kine-
matic studies on alpine skiing reported to date is the rela-
tively small number of subjects involved [22, 58]. These are
sometimes only case studies [45, 47] or involve, for example,
monitoring of 5–15 skiers performing one or two giant sla-
lom turns [49, 60] or a maximum of four slalom turns [16]. IMU sensors can potentially be utilized to monitor large 1 3 3 Journal of Science in Sport and Exercise (2021) 3:249–256 252 at the same time, minimize potential artefacts due to sensor
movement, e.g., as a consequence of muscle contractions [3,
13, 51, 70]. They also provide instructions about ensuring
that the sensors are fixed on the body segments as firmly
as possible while minimizing interference with athletic
movement/performance. Unfortunately, some of this advice
can be difficult to follow in connection with alpine skiing. For example, it is difficult to protect sensors on the shanks
and arms against impacts with the gates and, furthermore,
because of the tight ski boots, it is practically impossible to
place sensors directly on the foot. numbers of turns, even with respect to full-body 3D-kine-
matics [3, 27, 51]. Nevertheless, alpine skiing is somewhat special, involv-
ing extreme difficulties with respect to performing field
measurements on a large number of subjects with the same
or comparable skiing conditions for all runs (e.g., the prob-
lem of gutters and holes on the ski slope) [28, 61]. To a cer-
tain extent, this latter problem can be attenuated by smooth-
ing/maintaining the track between runs. In addition, accurate
geodetic GNSS systems can help ensure that the course gates
are placed in exactly the same locations on several days in
a row [73]. In any case, snow conditions, which can signifi-
cantly affect the level of friction between the skis and ground
[4], length of the course, shape of the turns, speeds, etc.,
must be carefully monitored, especially when investigating
performance [22, 58]. On the other hand, according to a review [7], the errors
associated with stereophotogrammetry, a golden standard
for measurement of full-body 3D kinematics, are quite sub-
stantial, even under laboratory conditions. The paper con-
clude, as other authors also have, that: “minor rotations out
of this plane are strongly affected by errors introduced at
the anatomical landmark identification level and are prone
to misinterpretation”. Validity and Reliability For instance, when a IMU suit is used to monitor full-body
3D movement, the biomechanical relationships between
body segments can also be used to compensate for drift [34].l GRF can be determined with force plates, but these are
typically cumbersome and may disturb the ski bindings
significantly [31]. On the other hand, determination of
GFR with camera-based systems, the golden standard for
kinematic measurements in the laboratory, is imprecise,
particularly when there are high-frequency oscillations in
these forces, as during alpine skiing [31]. Although to our
knowledge, usage of IMU sensors to determine GRFs during
alpine skiing has not yet been adequately validated, mod-
eling indicates that these sensors may provide better estima-
tions of these forces than camera-based systems. In part, this
is because the usage of IMU sensors does not require dual
data differentiation, which augments noise considerably,
necessitating the filtering of data [26, 55, 56]. In addition,
the sampling frequencies associated with the recording of
IMU data are sufficiently high to allow high-frequency oscil-
lations in GRF to be monitored. y
g
p
[
]
Nevertheless, the angle of knee flexion during alpine ski-
ing indicated by sensors was found to differ by 4.9° ± 0.4°
and 0.6° ± 1.0° for the left and right leg from the same angle
as measured by the optical systems considered to be labora-
tory gold standards [27]. In another investigation, Zhang and
colleagues [71] reported that measurements by inertial and
optical systems differ most with respect to external/internal
rotation and abduction/adduction of the knee, leading these
investigators to suggest that these differences were primar-
ily due to differences in the anatomical reference frames
utilized by these systems. Moreover, the type of inertial
sensor employed was found to be of lesser importance. Fur-
thermore, in another laboratory study with an instrumental
gimbal, the errors in measurements of 3D joint kinematics
by inertial sensors were demonstrated to be approximately
3° in all three directions [2]. Validity and Reliability When they first came into use, the drift associated with the
collection of 3D body kinematic data by inertial sensor-
based systems led the scientific community to question the
validity of their measurements. In the worst-case scenario,
when a single IMU sensor on a forced pendulum with no
filtering is utilized, the drift following 35 s of swinging at
a frequency of 0.5 or 2 Hz is expected to range from 2.1°
to 4.2°, respectively [51]. More recently, the manufacturers
of sensors have begun to employ various types of filtering
(e.g., quaternion-based complementary filtering, dynamic
optimization) and fusion algorithms (e.g., constraining mag-
netometer estimates) to minimize or eliminate drift [5, 8]. For instance, when a IMU suit is used to monitor full-body
3D movement, the biomechanical relationships between
body segments can also be used to compensate for drift [34]. Nevertheless, the angle of knee flexion during alpine ski-
ing indicated by sensors was found to differ by 4.9° ± 0.4°
and 0.6° ± 1.0° for the left and right leg from the same angle
as measured by the optical systems considered to be labora-
tory gold standards [27]. In another investigation, Zhang and
colleagues [71] reported that measurements by inertial and
optical systems differ most with respect to external/internal
rotation and abduction/adduction of the knee, leading these
investigators to suggest that these differences were primar-
ily due to differences in the anatomical reference frames
utilized by these systems. Moreover, the type of inertial
sensor employed was found to be of lesser importance. Fur-
thermore, in another laboratory study with an instrumental
gimbal, the errors in measurements of 3D joint kinematics
by inertial sensors were demonstrated to be approximately
3° in all three directions [2]. When they first came into use, the drift associated with the
collection of 3D body kinematic data by inertial sensor-
based systems led the scientific community to question the
validity of their measurements. In the worst-case scenario,
when a single IMU sensor on a forced pendulum with no
filtering is utilized, the drift following 35 s of swinging at
a frequency of 0.5 or 2 Hz is expected to range from 2.1°
to 4.2°, respectively [51]. More recently, the manufacturers
of sensors have begun to employ various types of filtering
(e.g., quaternion-based complementary filtering, dynamic
optimization) and fusion algorithms (e.g., constraining mag-
netometer estimates) to minimize or eliminate drift [5, 8]. Sample Size Furthermore, the errors associated
with laboratory measurements of rotations and abduction/
adduction angles by stereophotogrammetry appear to be of
the same magnitude as the differences between stereophoto-
grammetry and full-body inertial motion capture [71]. This
is one reason why IMU systems are being employed more
and more extensively in connection with several sports, as
well as with sports medicine and rehabilitation [67]. GNSS in Combination with IMUs The full-body 3D kinematic data provided by IMUs alone,
usually including the relative movement of body segments dur-
ing the performance of sports like alpine skiing [3, 27, 51], can
be related to the athlete´s position by the simultaneous use of a
GNSS. When these two approaches are combined, the quality
of the GNSS is the primary determinant of the global accuracy
of the entire system. Although Real-Time Kinematics GNSS The producers of inertial motion capture suits provide
advice concerning how best to position the sensors on the
body to monitor relative movements of body segments and, 1 3 1 3 253 Journal of Science in Sport and Exercise (2021) 3:249–256 systems are the golden standard with respect to alpine skiing
[65], when the GNSS system itself is not very precise or satel-
lite visibility is poor, IMUs can provide a valuable complement
[3, 38]. improvements in performance can have a considerable
impact on the athlete’s success. Nevertheless, our under-
standing of the determinants of alpine skiing performance
is still relatively limited [22] and the use of IMU sensors,
most often in combination with GNSS systems [65], has
the potential to improve this situation considerably. So far,
the use of these systems individually or in combination
provides coaches and other experts with a large amount
of information concerning different kinematic and kinetic
parameters related to performance and skiing technique
[3, 20, 27, 49, 51, 52, 60, 64]. In recent years, there has
also been a significant reduction in the size and weight of
wearable devices [23], allowing less disruptive and more
comfortable monitoring of alpine skiing. Synchronization of IMU Measurements with Video
Recordings When monitoring of full-body kinematics by IMU involves
multiple IMU sensors, 3D animation of the human body can
be used to provide feedback to the athletes. However, when
fewer IMUs are used, it is advisable to facilitate analysis by
combining the sensor data obtained with video recording (as is
sometimes done with GNSS data [65]), especially since both
skiers and their coaches are accustomed to analyzing video
recordings. To achieve the full potential of this symbiosis, the
IMU monitoring and video recording must be synchronized,
which can be accomplished with a variety of devices or simply
on the basis of isolated, clearly distinguishable body move-
ments, such as leg squats or hits with the skis to the ground
[51]. In light of the widespread use of smartphones and enor-
mous increase in the number of iOS and Android sports
applications available, we can expect smartphones with
built-in or complementary external IMU units to be used
more and more extensively in conjunction with GNSS sys-
tems to monitor alpine skiing. Such use will be promoted
by the development of more user-friendly interaction with
measuring equipment, including wireless communication,
direct storage of data in the cloud, and processing by artifi-
cial intelligence that allows more rapid (perhaps even real-
time) and accurate data analysis and feedback [35]. Such
systems could result in real-time virtual biomechanical
coaches, which may represent the next step in the evolu-
tion (revolution) of alpine skiing. Conclusions IMU sensors are being used more and more extensively in
connection with research on alpine skiing and have made
great contributions to our overall understanding of com-
petitive skiing. They can provide classic 3D kinematic
measurements as least as reliably as camcorders. At pre-
sent, the use of more complex systems consisting of many
IMUs is limited mainly to scientific research, but simpler
systems are already beginning to replace, at least in part,
the use of GNSS systems and, of course, photocells and
video cameras by coaches. In addition to monitoring the
technique and performance of skiers, IMU technology can
also be of considerable value in connection with testing
equipment (e.g., skis, ski boots), which is, of course, also
closely related to performance. Gradually, coaches are beginning to use technologies
more often, even as these technologies become more com-
plex and capable of providing more detailed and accurate
feedback. Perhaps in the near future, advanced technology
will be able to perform many of the functions now carried
out by alpine skiing coaches, but there is little doubt that Additional Limitations and Recommendations In addition to the limitations that have already been presented,
there are some more worth to mention in connection of using
IMUs for monitoring alpine skiing performance. Namely,
when using full-body inertial motion capture, a relatively
expensive set is still needed to be used and only one skier can
be monitored at a time using a single system. Furthermore, to
ensure that none of the sensors has moved while ascending on
the chairlift or gondola or if jostled while riding the lift, a cali-
bration process before each run is advisable that takes some
additional time. To execute a trustworthy full-body calibration
procedure and follow the instructions by the manufacturers,
it is recommended to build a levelled platform on the start to
assure proper calibration. Although IMU sensors are getting
smaller, with their weight being as low as 10 g (e.g. MTx,
Xsens Technologies, Enschede, the Netherlands), their distri-
bution over the body and wiring may also have some influence
on the performance of elite athletes. On the other hand, even
wireless IMU systems exist (e.g. MVN Avida, Xsens Tech-
nologies, Enschede, the Netherlands), the technical specifica-
tions for such systems are often inferior compared to wired
ones and in practice, wireless systems may be prone to more
connection problems. Compliance with Ethical Standards 10. Fasel B, Lechot C, Spörri J, Müller E, Aminian K. Body vibration
and its transmission in alpine ski racing. International Symposium
on 3d Analysis of Human Movement. 2016.fi Conflicts of interest The authors declare that this research was con-
ducted without any commercial or financial relationships that could be
construed as a potential conflict of interest. 11. Fasel B, Spörri J, Gilgien M, Boffi G, Chardonnens J, Müller E,
Aminian K. Three-dimensional body and centre of mass kinemat-
ics in alpine ski racing using differential GNSS and inertial sen-
sors. Remote Sens. 2016;8(8):671. https://doi.org/10.3390/rs808
0671. Consent for publication The work described has not been published
before and is not under consideration for publication anywhere else. The publication has been approved by both authors, as well as by the
responsible authorities at the institute where the work has been car-
ried out. 12. Fasel B, Spörri J, Kröll J, Müller E, Aminian K. A magnet-based
timing system to detect gate crossings in alpine ski racing. Sen-
sors. 2019;19(4):940 https://doi.org/10.3390/s19040940. 13. Fasel B, Spörri J, Schütz P, Lorenzetti S, Aminian K. An inertial
sensor-based method for estimating the athlete’s relative joint
center positions and center of mass kinematics in alpine ski rac-
ing. Front Physiol. 2017;8:850. https://doi.org/10.3389/fphys.
2017.00850. Open Access This article is licensed under a Creative Commons Attri-
bution 4.0 International License, which permits use, sharing, adapta-
tion, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons licence, and indicate if changes
were made. The images or other third party material in this article are
included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in
the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a
copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. 14. Fasel B, Spörri J, Schütz P, Lorenzetti S, Aminian K. Validation
of functional calibration and strap-down joint drift correction for
computing 3D joint angles of knee, hip, and trunk in alpine skiing. PLoS ONE. 2017;12(7):e0181446. https://doi.org/10.1371/journ
al.pone.0181446.l 15. Federolf P, Von Tscharner V, Haeufle D, Nigg B, Gimpl M, Mül-
ler E. Future Perspectives As mentioned prevoiusly, the difference between a first
and second place in an alpine skiing race can be no more
than one-hundredth of a second [18], so even very small 1 1 3 Journal of Science in Sport and Exercise (2021) 3:249–256 254 such technology will become a routine digital/virtual assis-
tant trainer who helps skiers reach the winners’ podium in
connection with the most elite level of competition, such as
the Olympic Games and World Championships. 5. Cotton RJ, Rogers J. Wearable monitoring of joint angle and mus-
cle activity. IEEE Int Conf Rehabil Robot. 2019;2019:258–63. https://doi.org/10.1109/icorr.2019.8779538. p
g
6. Cust EE, Sweeting AJ, Ball K, Robertson S. Machine and deep
learning for sport-specific movement recognition: a systematic
review of model development and performance. J Sports Sci. 2019;37(5):568–600. https://doi.org/10.1080/02640414.2018.
1521769. Acknowledgements We are thankful to several coaches and members
of various elite ski teams for feedback concerning the measurements
and analyses that we have provided to them. 7. Della Croce U, Leardini A, Chiari L, Cappozzo A. Human move-
ment analysis using stereophotogrammetry. Part 4: assessment
of anatomical landmark misplacement and its effects on joint
kinematics. Gait Posture. 2005;21(2):226–37. https://doi.org/10.
1016/j.gaitpost.2004.05.003.i Author Contributions MS and HCH contributed to all parts of this
work, including conception, design, and writing. Both authors approved
the final version of this manuscript for publication. j g
p
8. Dorschky E, Nitschke M, Seifer AK, van den Bogert AJ, Eskofier
BM. Estimation of gait kinematics and kinetics from inertial sen-
sor data using optimal control of musculoskeletal models. J Bio-
mech. 2019;95:109278. https://doi.org/10.1016/j.jbiomech.2019.
07.022. Funding Open Access funding provided by Mid Sweden University. The study was financially supported by the Swedish Olympic Committe
and the Slovenian Research Agency within the research program fund
P5-0147 Kinesiology of monostructural, polystructural and conven-
tional sports. 9. Fasel B, Gilgien M, Sporri J, Aminian K. A new training assess-
ment method for alpine ski racing: estimating center of mass
trajectory by fusing inertial sensors with periodically available
position anchor points. Front Physiol. 2018;9:1203. https://doi.
org/10.3389/fphys.2018.01203. Compliance with Ethical Standards Vibration exposure in alpine skiing and consequences for
muscle activation levels. In: Müller SL E, Stöggl T, editors. Sci-
ence and skiing IV. Maidenhead: Meyer & Meyer (UK) Sport Ltd,
Aspen Snowmass; 2009. pp. 19–25. 16. Federolf PA. Quantifying instantaneous performance in alpine ski
racing. J Sports Sci. 2012;30(10):1063–8. https://doi.org/10.1080/
02640414.2012.690073. 4. Colbeck SC. A review of the friction of snow skis. J Sports Sci.
1994;12(3):285–95. References Wearable sensors: modalities,
challenges, and prospects. Lab Chip. 2018;18(2):217–48. https://
doi.org/10.1039/c7lc00914c. 40. Spörri J, Kröll J, Fasel B, Aminian K, Müller E. The use of body
worn sensors for detecting the vibrations acting on the lower back
in alpine ski racing. Front Physiol. 2017;8:522. https://doi.org/10.
3389/fphys.2017.00522. g
24. ISO 2631-1. Mechanical vibration and shock. Evaluation of
human exposure to whole-body vibration. Part 1: General require-
ments. 2008. 41. Spörri J, Kröll J, Fasel B, Aminian K, Müller E. Standing height
as a prevention measure for overuse injuries of the back in alpine
ski racing: a kinematic and kinetic study of giant slalom. Orthop
J Sports Med. 2018;6(1):2325967117747843. https://doi.org/10.
1177/2325967117747843. 25. ISO-8041. Human response to vibration—measuring instrumenta-
tion. 2005. 26. Klous M, Müller E, Schwameder H. Three-dimensional knee joint
loading in alpine skiing: a comparison between a carved and a
skidded turn. J Appl Biomech. 2012;28(6):655–64. 42. Spörri J, Kröll J, Gilgien M, Müller E. Sidecut radius and the
mechanics of turning—equipment designed to reduce risk of
severe traumatic knee injuries in alpine giant slalom ski rac-
ing. Br J Sports Med. 2016;50(1):14–9. https://doi.org/10.1136/
bjsports-2015-095737. 27. Krüger A, Edelmann-Nusser J. Application of a full body inertial
measurement system in alpine skiing: a comparison with an opti-
cal video based system. J Appl Biomech. 2010;26(3):516–21.l 28. Lešnik B, Axelsson EP, Supej M. Influence of the start num-
ber on elite alpine skiing competitors’ results. Kinesiol Slov. 2013;27(6):3–24. 43. Spörri J, Kröll J, Gilgien M, Müller E. How to prevent injuries
in alpine ski racing: what do we know and where do we go from
here? Sports Med. 2017;47(4):599–614. https://doi.org/10.1007/
s40279-016-0601-2. 29. López-Nava I, Muñoz-Meléndez A. Wearable inertial sen-
sors for human motion analysis a review. IEEE Sensors J. 2016;16(22):7821–34. https://doi.org/10.1109/JSEN.2016.26093
92. 44. Sporri J, Kroll J, Schwameder H, Muller E. The role of path
length- and speed-related factors for the enhancement of
section performance in alpine giant slalom. Eur J Sport Sci. 2018;18(7):911–9. https://doi.org/10.1080/17461391.2018.
1453870. 30. Luethi SM, Denoth J. The influence of aerodynamic and anthro-
pometric factors on speed in skiing. Int J Sport Biomech. 1987;3:345–52. 31. Lüthi A, Federolf PA, Fauve M. Oberhofer K, Rhyner HU,
Ammann WJ, Stricker G, Schiefermuller C, Schwameder H, Mul-
ler E. Determination of forces in carving using three independ-
ent methods. In: Müller E, Bacharach D, Klika R, Lindinger S,
Schwameder H, editors. Science and Skiing III. Oxford: Meyer &
Meyer Sport; 2005. pp. References 1. Aroganam G, Manivannan N, Harrison D. Review on wearable
wechnology wensors used in consumer sport applications. Sen-
sors. 2019;19(9):1983 https://doi.org/10.3390/s19091983.i 17. Gilgien M, Crivelli P, Spörri J, Kröll J, Müller E. Characterization
of course and terrain and their effect on skier speed in World Cup
alpine ski racing. PLoS ONE. 2015;10:e0118119. https://doi.org/
10.1371/journal.pone.0118119. 2. Brennan A, Zhang J, Deluzio K, Li Q. Quantification of iner-
tial sensor-based 3D joint angle measurement accuracy using an
instrumented gimbal. Gait Posture. 2011;34(3):320–3. 18. Gilgien M, Reid R, Raschner C, Supej M, Holmberg HC. The
training of Olympic alpine ski racers. Front Physiol. 2018;9:1772. https://doi.org/10.3389/fphys.2018.01772. 3. Brodie M, Walmsley A, Page W. Fusion motion capture: a
prototype system using inertial measurement units and GPS
for the biomechanical analysis of ski racing. Sports Technol. 2008;1(1):17–28. 19. Gilgien M, Spörri J, Chardonnens J, Kröll J, Limpach P, Müller
E. Determination of the centre of mass kinematics in alpine ski-
ing using differential global navigation satellite systems. J Sport
Sci. 2015;33(9):960–9. https://doi.org/10.1080/02640414.2014.
977934. 4. Colbeck SC. A review of the friction of snow skis. J Sports Sci. 1994;12(3):285–95. 4. Colbeck SC. A review of the friction of snow skis. J Sports Sci. 1994;12(3):285–95. 1 3 Journal of Science in Sport and Exercise (2021) 3:249–256 255 technology. J Sports Sci. 2016;34(9):843–51. https://doi.org/10.
1080/02640414.2015.1075057. 20. Gilgien M, Spörri J, Chardonnens J, Kröll J, Müller E. Determina-
tion of external forces in alpine skiing using a differential global
navigation satellite system. Sensors. 2013;13(8):9821–35. 38. 38. Skaloud J, Limpach P. Synergy of CP-DGPS, accelerometry and
magnetic sensors for precise trajectography in ski racing. 16th
International Technical Meeting of the Satellite Division of The
Institute of Navigation (ION GPS/GNSS 2003). 2003. g
y
21. Grood ES, Suntay WJ. A joint coordinate system for the clinical
description of three-dimensional motions: application to the knee. J Biomech Eng. 1983;105(2):136–44. g
22. Hébert-Losier K, Supej M, Holmberg HC. Biomechanical factors
influencing the performance of elite alpine ski racers. Sports Med. 2014;44(4):519–33. https://doi.org/10.1007/s40279-013-0132-z. 39. Spörri J, Kröll J, Fasel B, Aminian K, Müller E. Course setting as
a prevention measure for overuse injuries of the back in alpine ski
racing: a kinematic and kinetic study of giant slalom and slalom
Orthop. J Sports Med. 2016;4:2325967116630719. https://doi.org/
10.1177/2325967116630719. 23. Heikenfeld J, Jajack A, Rogers J, Gutruf P, Tian L, Pan T, Li R,
Khine M, Kim J, Wang J, Kim J. References 96–106. 45. Spörri J, Kröll J, Schwameder H, Müller E. Turn characteristics
of a top world class athlete in giant slalom: a case study assessing
current performance prediction concepts. Int J Sports Sci Coach. 2012;7:647–59. https://doi.org/10.1260/1747-9541.7.4.647. 46. Spörri J, Kröll J, Schwameder H, Müller E. The role of path
length- and speed-related factors for the enhancement of
section performance in alpine giant slalom. Eur Sport Sci. 2018;18(7):911-9. https://doi.org/10.1080/17461391.2018.14538
70. 32. McGinnis RS, Cain SM, Davidson SP, Vitali RV, Perkins NC,
McLean SG. Quantifying the effects of load carriage and fatigue
under load on sacral kinematics during countermovement verti-
cal jump with IMU-based method. Sports Eng. 2016;19:21–34. https://doi.org/10.1007/s12283-015-0185-3. 47. Spörri J, Kröll J, Schwameder H, Schiefermüller C, Müller E. Course setting and selected biomechanical variables related to
injury risk in alpine ski racing: an explorative case study. Br J
Sports Med. 2012;46(15):1072–7. https://doi.org/10.1136/bjspo
rts-2012-091425. 33. Meyer F, Le Pelley D, Borrani F. Aerodynamic drag modeling
of alpine skiers performing giant slalom turns. Med Sci Sports
Exerc. 2012;44(6):1109–15. https://doi.org/10.1249/MSS.0b013
e3182443315. 48. Spörri J, Schiefermüller C, Müller E. Collecting kinematic data on
a ski track with optoelectronic stereophotogrammetry: a methodo-
logical study assessing the feasibility of bringing the biomechan-
ics lab to the field. PLoS ONE. 2016;11(8):e0161757. https://doi.
org/10.1371/journal.pone.0161757. 34. Mohamed Refai MI, van Beijnum B-J, Buurke JH, Saes M,
Bussmann JBJ, Meskers CG, van Wegen E, Kwakkel G, Veltink
PH. Portable Gait Lab: Zero Moment Point for Minimal Sensing
of Gait*. 2019 41st Annual International Conference of the IEEE
Engineering in Medicine and Biology Society (EMBC). 2019. pp
2077–2081. https://doi.org/10.1109/EMBC.2019.8857314 49. Supej M. Differential specific mechanical energy as a qual-
ity parameter in racing alpine skiing. J Appl Biomech. 2008;24(2):121–9. p
g
35. Nemec B, Petric T, Babic J, Supej M. Estimation of alpine
skier posture using machine learning techniques. Sensors. 2014;14(10):18898–914. https://doi.org/10.3390/s141018898. 50. Supej M. A step forward in 3D measurements in alpine skiing:
a combination of an inertial suit and dGPS technology. In: Mül-
lerSL E, Stöggl T, editors. Science and skiing IV. Maidenhead:
Meyer & Meyer Sport (UK) Ltd, Aspen Snowmass; 2009. pp. 497–504. 36. Picerno P, Camomilla V, Capranica L. Countermovement jump
performance assessment using a wearable 3D inertial measure-
ment unit. J Sports Sci. 2011;29(2):139–46. https://doi.org/10.
1080/02640414.2010.523089. 51. Supej M. 3D measurements of alpine skiing with an inertial sen-
sor motion capture suit and GNSS RTK system. J Sports Sci. 2010;28(7):759–69. https://doi.org/10.1080/02640411003716934. 37. References Setuain I, Martinikorena J, Gonzalez-Izal M, Martinez-Ramirez
A, Gómez M, Alfaro-Adrián J, Izquierdo M. Vertical jumping bio-
mechanical evaluation through the use of an inertial sensor-based 52. Supej M. Gate-to-gate synchronized comparison of velocity
retrieved from a high-end global navigation satellite system in 1 3 Journal of Science in Sport and Exercise (2021) 3:249–256 256 alpine skiing. The 28 International Conference on Biomechanics
in Sports. 2012. 64. Supej M, Sætran L, Oggiano L, Ettema G, Šarabon N, Nemec B,
Holmberg HC. Aerodynamic drag is not the major determinant of
performance during giant slalom skiing at the elite level. Scandi-
navian J Med Sci Sports. 2013;23(1):e38–47. 53. Supej M. Vibrations in recreational alpine skiing: a pilot study. The Conference proceedings of the 31st International Symposium
on Biomechanics in Sport. 2013.f 65. Supej M, Spörri J, Holmberg HC. Methodological and practical
considerations associated with assessment of alpine skiing per-
formance using global navigation satellite systems. Front Sports
Acts Living. 2020;1:74. https://doi.org/10.3389/fspor.2019.00074. p
54. Supej M, Cernigoj M. Relations between different technical and
tactical approaches and overall time at men’s world cup giant sla-
lom races. Kinesiol Slov. 2006;12:63–9. 55. Supej M, Hébert-Losier K, Holmberg HC. Impact of the steepness
of the slope on the biomechanics of World Cup slalom skiers. Int J
Sports Physiol Perform. 2015;10:361–8. https://doi.org/10.1123/
ijspp.2014-0200.l 66. Tarabini M, Saggin B, Scaccabarozzi D. Whole-body vibra-
tion exposure in sport: four relevant cases. Ergonomics. 2015;58(7):1143–50. https://doi.org/10.1080/00140139.2014.
961969.l 67. Teufl W, Miezal M, Taetz B, Frohlich M, Bleser G. Validity of
inertial sensor based 3D joint kinematics of static and dynamic
sport and physiotherapy specific movements. PLoS ONE. 2019;14(2):e0213064. https://doi.org/10.1371/journal.pone.02130
64. 56. Supej M, Holmberg HC. How gate setup and turn radii influ-
ence energy dissipation in slalom ski racing. J Appl Biomech. 2010;26(4):454–64. 57. Supej M, Holmberg HC. A new time measurement method using
a high-end global navigation satellite system to analyze alpine
skiing. Res Q Exerc Sport. 2011;82(3):400–11. https://doi.org/
10.1080/02701367.2011.10599772. 68. Walmsley CP, Williams SA, Grisbrook T, Elliott C, Imms
C, Campbell A. Measurement of upper limb range of motion
using wearable sensors: a systematic review. Sports Med Open. 2018;4(1):53. https://doi.org/10.1186/s40798-018-0167-7. 58. Supej M, Holmberg HC. Recent kinematic and kinetic advances
in olympic alpine skiing: Pyeongchang and beyond. Front Physiol. 2019;10:111. https://doi.org/10.3389/fphys.2019.00111. 69. Weygers I, Kok M, Konings M, Hallez H, De Vroey H, Claeys K. Inertial sensor-based lower limb joint kinematics: a methodologi-
cal systematic review. Sensors (Basel). 2020;20(3):673. References https://
doi.org/10.3390/s20030673. p
g
p y
59. Supej M, Kipp R, Holmberg HC. Mechanical parameters as pre-
dictors of performance in alpine World Cup slalom racing. Scand
J Med Sci Sports. 2011;21(6):e72–81. 70. Yu G, Jang YJ, Kim J, Kim JH, Kim HY, Kim K, Panday SB. Potential of IMU sensors in performance analysis of professional
alpine skiers. Sensors (Basel). 2016;16(4):463. https://doi.org/10.
3390/s16040463. 60. Supej M, Kugovnik O, Nemec B. DGPS measurement system
in alpine skiing track and center of mass estimation. In: Jiang Y,
Baca A, Zhang H, editors. Proceeding of first joint international
pre-olympic conference of sports sciences and sports engineering
computer science in sports. Liverpool: World Academic Union;
2008. pp. 120–5. 71. Zhang JT, Novak AC, Brouwer B, Li Q. Concurrent validation of
Xsens MVN measurement of lower limb joint angular kinematics. Physiol Meas. 2013;34(8):N63. 61. Supej M, Nemec B, Kugovnik O. Changing conditions on the
slalom ski course affect competitors’ performances. Kinesiology. 2005;37:151–8. 72. Zorko M, Nemec B, Babic J, Lesnik B, Supej M. The waist width
of skis influences the kinematics of the knee joint in alpine skiing. J Sports Sci Med. 2015;14(3):606–19. 62. Supej M, Ogrin J. Transmissibility of whole-body vibrations
and injury risk in alpine skiing. J Sci Med Sport. 2019;22(Suppl
1):S71-s77. https://doi.org/10.1016/j.jsams.2019.02.005. 73. Zorko M, Nemec B, Matjačić Z, Olenšek A, Tomazin K, Supej M. Wide skis as a potential knee injury risk factor in alpine Skiing. Front Sports Act Living. 2020. https://doi.org/10.3389/fspor.2020.
00007. 63. Supej M, Ogrin J, Holmberg HC. Whole-body vibrations associ-
ated with alpine skiing: a risk factor for low back pain? Front
Physiol. 2018;9:204. https://doi.org/10.3389/fphys.2018.00204. 1 3 3
|
https://openalex.org/W4379378247
|
https://zenodo.org/records/8005781/files/EJAR0607.pdf
|
English
| null |
DEVELOPMENT OF THE SCIENTIFIC BASED COMPOSITION AND TECHNOLOGY OF THE "ANTIVIRUS TABLET"
|
Zenodo (CERN European Organization for Nuclear Research)
| 2,023
|
cc-by
| 1,876
|
Volume 3 Issue 6, June 2023 ISSN 2181-2020 Innovative Academy Research Support Center
= 8.1 | SJIF = 5.685
www.in-academy.uz
DEVELOPMENT OF THE SCIENTIFIC BASED COMPOSITION
AND TECHNOLOGY OF THE "ANTIVIRUS TABLET"
1A.M. Usubbaev
2Sh.M.Usubbaeva
Tashkent Pharmaceutical Institute
Uzbekistan, Tashkent, oybek 45
е-mail: shahnozau.m@mail.ru
https://www.doi.org/10.5281/zenodo.8005781 DEVELOP
AND
ARTICLE INFO
Received: 27th May 2023
Accepted: 04th June 2023
Online: 05th June 2023
KEY WORDS
Rometin, antiviral, substance,
tablet, technological indicator,
fractional content, scattering
density, dispersibility. ABSTRACT For the first time, on the basis of local raw materials, the
technological properties of the active substance were
studied using the methods and tools presented in the
literature in order to develop the scientifically based
moderate composition and technology of the antiviral
tablet "Rometin". The conducted experiments serve to
theoretically justify the selection of appropriate tablet
preparation technology and the type and quantity of
auxiliary substances that should be added to the tablet
composition. Based on the results of scientific research, the
scientifically based composition and technology of
"Rometin" tablet was developed. One of the urgent tasks of the pharmaceutical science is to develop and introduce into
medical practice the technologies of non-toxic, import-substituting drugs, which are prepared
on the basis of new, local raw materials, using the rich natural resources of Uzbekistan. The domestic pharmaceutical market has a large assortment of ready-made drugs used
in the prevention and treatment of viral diseases. However, creation of import-substitute anti-
virus drugs with high bioefficiency based on local raw materials and organization of
production by local enterprises is one of the urgent problems of the pharmaceutical industry
today. Taking this into account, we found it necessary to develop the composition and
technology of an anti-viral pill form using the rich natural resources of our Republic. Scientists of the Institute of Bioorganic Chemistry of the Academy of Sciences of the
Republic of Uzbekistan named after O.S. Sodikov obtained the compound "Rometin" formed
by megacin with polyvinylpyrrolidone. Preliminary pharmacological studies have shown that
the active substance "Rometin" is a drug with a highly effective antiviral effect at a dose of 100
mg [1]. Purpose of work: development of scientifically based composition and technology of
"Rometin" tablet with antiviral effect. Studying the physico-chemical and technological
properties of the active substance "Rometin", choosing the type and amount of auxiliary
substances added to the tablet as a result of the experiments was determined as the goal of
the research. Experience part: in the development of tablet technology, in order to theoretically
justify the type, quantity and preparation technology of auxiliary substances added to the Volume 3 Issue 6, June 2023 ISSN 2181-2020 Page 39 EURASIAN JOURNAL OF ACADEMIC RESEARCH
I
i
A
d
R
h S
C tablet, the technological properties of the substance were studied using the methods and
laboratory equipment presented in the literature [2,3,5]. The obtained results are presented in Table 1, from which it can be seen that the active
substance has negative indicators in terms of technological properties such as dispersion,
dispersion density, density coefficient, which indicates that it is not possible to obtain
presslab tablets directly from such substances. Table 1 Table 1 Technological properties of the active substance "Rometin". Technological properties of the active substance "Rometin". №
Learned indicators
Unit of measure
The results obtained
1
Fractional content, µm:
+2500
-2500 +1000
-1000 +500
-500 +250
-250 +150
-150
%
11.2
66.4
14.5
5.6
2.3
-
3
Spreading density
kg/m3
440
4
Spreadability
10-3kg/s
2.4
5
Natural deviation angle
Gradus
44
6
Compressibility
N
90
7
Density coefficient
-
3.7
8
Residual humidity, 70ºС
%
5.8 The next stage of our work is devoted to the selection of the type and amount of
appropriate excipients to be added to the tablet, as well as the development of a moderate
technology. , magnesium stearate, stearic acids, 6 different compositions are prepared under
standard conditions, and pressed using wet granulation method in the presence of purified
water, ethyl alcohol, 3-5-7% starch slime, sugar paste, 2, 3% methylcellulose gels as a binding
agent. masses were prepared. The obtained results are presented in Table 2. Sample tablets were prepared from the prepared ingredients in a manual hydropress,
and they were evaluated according to their appearance, breaking strength, and disintegration. The results of the experiment showed that it is appropriate to use 96% ethyl alcohol as a
binder in the preparation of tablets using the method of wet granulation. Because tablets
prepared with the presence of purified water, different concentrations of starch mucilage and
methylcellulose gels did not meet the requirements in terms of appearance, hardness and
disintegration. The results of the conducted experiment showed that the most suitable binder
is 96% ethyl alcohol. The obtained results are presented in Table 3 [4, 5]. Table 2 Volume 3 Issue 6 June 2023
ISSN 2181 2020
Page 40
Tested compositions for making tablets by wet granulation method
Components
Purified water, ethyl alcohol, 3, 5, 7% starch mucilage, 64% sugar
syrup, 1, 2% MTS gel. EURASIAN JOURNAL OF ACADEMIC RESEARCH
I
i
A
d
R
h S
C №1
№2
№3
№4
№ 5
№ 6
Active substance, g;
0.100
0.100
0.100
0.100
0.100
0.100 compositions for making tablets by wet granulation method EURASIAN JOURNAL OF ACADEMIC RESEARCH
Innovative Academy Research Support Center
UIF = 8.1 | SJIF = 5.685
www.in-academy.uz
Microcrystalline Cellulose
(MCTS) "Cotton Cellulose"
-
-
-
-
0.100
Sodium bicarbonate
-
-
-
-
0.067
-
Sucrose, g;
-
0.185
-
-
-
-
Milk sugar, g;
0.185
-
-
-
-
Citric acid, g;
-
-
-
-
0.03
-
Potato starch, g;
0.100
0.100
0.100
0.100
-
0.100
Calcium stearate, g;
-
-
-
-
0.003
-
Stearic acid, g;
0.003
-
-
-
-
-
Magnesium stearate, g;
-
0.003
-
-
-
0.003
Average mass, g
0.300
0.300
0.300
0.300
0.300
0.300 EURASIAN JOURNAL OF ACADEMIC RESEARCH At the next stage of our work, pressable masses of 6 different compositions were
prepared with the above auxiliary substances, in the presence of 96% ethyl alcohol. The
quality of the tablets made from the prepared masses was evaluated and the technological
properties of the pressed masses were studied using the methods presented in the literature. The obtained results are presented in Table 3, from which it can be seen that the technological
properties of the pressed masses prepared using the method of wet granulation have changed
sharply in a positive direction compared to those of the substance. p y
p
p
Table 3 The results of studying the technological properties of pressable masses prepared
with 96% ethyl alcohol as a binder
№
Determinable
indicators and
units of
measurement
The results obtained
№1
№2
№3
№4
№5
№6
1
Fractional content,
µm, %
+ 1000
- 1000 + 500
- 500 + 315
- 315 + 250
- 250 + 100
- 100 + 50
- 50
0.4
19.6
35.3
20.0
7.7
4.0
13.0
1.0
24.0
25.5
23.4
10.7
6.5
8.9
2.0
28.3
31.9
18.3
7.6
4.9
7.0
3.0
30.5
33.8
15.5
8.3
2.0
6.9
1.0
23.2
29.0
23.0
9.9
6.4
7.5
1.5
31.5
30.8
13.8
10.3
7.1
5.0
2
Spreadability,
kg/s*10-3
7.35
7.40
7.5
7.8
8.23
6.4
3
Angle of natural
deviation, grad. EURASIAN JOURNAL OF ACADEMIC RESEARCH
I
i
A
d
R
h S
C 36.0
36.2
35.9
36.6
36.9
38.1 Volume 3 Issue 6, June 2023 EURASIAN JOURNAL OF ACADEMIC RESEARCH
Innovative Academy Research Support Center
UIF = 8.1 | SJIF = 5.685
www.in-academy.uz
4
Spreading density,
kg/m3
686.9
686.0
670.0
677.0
669.0
665.0
5
Compressibility, N
100
95
69
90
50
100
6
Density coefficient
2.54
2.49
2.55
2.3
2.1
2.28
7
Residual moisture
(700C), %
3.5
3.0
3.1
3.5
3.2
3.4
8
Appearance of the
tablet
brown,
large
holder
brown,
large
holder
dark
brown
large
holly
dark
brown
large
holly
pale
yellow
color
small
hollar
brown
large
holly EURASIAN JOURNAL OF ACADEMIC RESEARCH As can be seen from the results presented in the table, the most appropriate composition
was found to be 5. Because the tablets obtained according to other compositions did not meet
the requirements for their appearance - color and disintegration in liquid. the requirements for their appearance - color and disintegration in liquid. Based on the conducted experiments, the following composition and technology of
"Rometin" tablets were proposed: Rometin substance
0.100 g
MKTS
0.100 g
Sodium bicarbonate 0.067 g
Citric acid
0.03 g
Calcium stearate
0.003 g
The average mass of the tablet is 0.300g
Technological process Rometin substance, MKTS "Cotton cellulose", sodium bicarbonate and citric acid are
separately sifted through a sieve with a hole diameter of 150 μm, mixed thoroughly, then
moistened with 96% ethyl alcohol until a moderately wet mass is formed, the wet mass is
spread thinly on parchment paper, 40-50°C temperature on a drying rack to moderate
residual moisture. It is then passed through a 1000 µm sieve, granulated and coated with a
mixture of calcium stearate sieved through a 100 µm sieve. The finished mass was pressed in
a mold with a diameter of 9 mm, with an average mass of 0.3 g, in a tablet machine of the
German company "Erveka". During the pressing process, the constancy of the average mass of
the tablets was observed. The pressing process was normal, the tablets did not stick to the
molds and punches. Summary: Thus, for the first time, the technological properties of the active substance
"Rometin" with antiviral effect were studied with the help of the methods and equipment
presented in the literature. As a result of the conducted scientific research, the scientifically
based type and amount of auxiliary substances included in the composition of the "Rometin"
tablet were determined. EURASIAN JOURNAL OF ACADEMIC RESEARCH
I
i
A
d
R
h S
C The appropriate composition and technology of the tablet have been Volume 3 Issue 6, June 2023 ISSN 2181-2020 Volume 3 Issue 6, June 2023 ISSN 2181-2020 Page 42 EURASIAN JOURNAL OF ACADEMIC RESEARCH
Innovative Academy Research Support Center
UIF = 8.1 | SJIF = 5.685
www.in-academy.uz
developed. The proposed composition and technology provide the opportunity to obtain high-
quality tablets in modern tablet machines[4,5]. References: 1. Borzunov E. E.Issledovaniya v oblasti physico-khimicheskoy mechanic tabletirovaniya
lekarstvennyx poroshkoobraznyx veshchestv [Text]: Autoreferat dis. na soiskanie uchenoy
stepi doctora farmatsevticheskikh nauk. (790) / Lvov., Gos. med., in-t. - Lviv: [b. i.], 1972. - 40
p. 1. Borzunov E. E.Issledovaniya v oblasti physico-khimicheskoy mechanic tabletirovaniya
lekarstvennyx poroshkoobraznyx veshchestv [Text]: Autoreferat dis. na soiskanie uchenoy
stepi doctora farmatsevticheskikh nauk. (790) / Lvov., Gos. med., in-t. - Lviv: [b. i.], 1972. - 40
p. 2. Bulygina I.V., Vorobeva N.V., Egorova S.N., Gubaidullin A.T. Comparative analysis of the
crystalline form of substance thioctoic acid // Sovremennye problemy nauki i obrazovaniya. –
2015. – No. 1-1.; 3. Makhkamov S.M. Accuracy in tablet quality indicators // Chemistry and Pharmacy. -
Tashkent, 1995. - #3.- B. 24-28. 4. Gosudarstvennaya pharmacopoeia USSR.- XI izd. - Moscow.: Medicine, 1989.- Vyp. 2. - S. 154. 5. Kambarov Kh.J, Usubbaev A.M. Development and technology of tablets "Tselitel
Vostoka"//Vestnik pharmacy.-2012.-№ 2 (56) 6. A.M. Usubbayev, Sh.M. Usubbayeva Determination of physico-chemical and
technological parameters of “Lagovin” substation-2021-№2 Volume 3 Issue 6, June 2023 ISSN 2181-2020
Page 43 Volume 3 Issue 6, June 2023 ISSN 2181-2020 Volume 3 Issue 6, June 2023 Page 43
|
W3041073884.txt
|
https://www.medecinesciences.org/articles/medsci/pdf/2019/12/msc190016.pdf
|
fr
|
Améliorer le ciblage tissulaire des anticorps thérapeutiques par de nouveaux formats
|
Médecine/sciences/MS. Médecine sciences
| 2,019
|
cc-by
| 5,148
|
médecine/sciences 2019 ; 35 : 1106-13
médecine/sciences
> De nombreuses pathologies cérébrales neurodégénératives ou tumorales devraient pourvoir bénéficier des progrès thérapeutiques
impressionnants des anticorps médicaments.
Malheureusement, en raison de leur très faible
passage dans le cerveau, de nombreux développements cliniques d’anticorps dont la cible thérapeutique se situe dans le parenchyme cérébral
ont été arrêtés par manque d’efficacité. La barrière hémato-encéphalique (BHE), douée de propriétés extrêmement sélectives et restrictives,
est à l’origine de la faible pénétration cérébrale
des molécules de haute masse moléculaire, telles
que les anticorps thérapeutiques. L’objectif de
cette revue est de présenter les propriétés de la
BHE et les dernières avancées dans le domaine
de l’ingénierie de nouveaux formats d’anticorps
susceptibles d’améliorer leur passage intracérébral. <
Bien que de très nombreux anticorps thérapeutiques
aient été mis sur le marché depuis une décennie pour
traiter principalement des cancers et des maladies
inflammatoires, seuls 5 % d’entre eux sont indiqués
dans des maladies du système nerveux central (SNC). En
France, le natalizumab et l’alemtuzumab sont indiqués
dans la sclérose en plaque (SEP) et le bévacizumab
vient d’obtenir l’autorisation de mise sur le marché
(AMM) des autorités américaines de la FDA (food and
drug administration) dans le traitement d’une tumeur
cérébrale, le glioblastome. Si l’on s’intéresse au mécanisme d’action de ces anticorps thérapeutiques dans
ces maladies du SNC, on s’aperçoit qu’il ne repose
pas sur leur passage du sang vers le cerveau mais en
bloquant dans le sang soit l’extravasation cérébrale
des leucocytes dans la SEP, soit en inhibant la fixation
du VEGF (vascular endothelial growth factor) à son
récepteur, ce facteur de croissance agissant positivement sur la croissance tumorale dans le glioblastome.
De nombreux anticorps à visée thérapeutique ont été,
ou sont encore actuellement testés dans des essais
1106
Les nouveaux formats d’anticorps
Améliorer
le ciblage tissulaire
des anticorps
thérapeutiques
par de nouveaux
formats
L’exemple de la barrière
hémato-encéphalique
Pierre Lafaye1, Dominique Lesuisse2,
Xavier Declèves3
1
Institut Pasteur, Plateforme
d’ingénierie des anticorps,
Paris, France.
2
Sanofi, Département des
Maladies Rares et Neurologiques,
Groupe Barrières Cérébrales,
Paris, France.
3
Université Paris Descartes,
Faculté de Pharmacie de Paris
et Inserm UMRS-1144,
Barrière hémato-encéphalique :
Physiopathologie et Thérapie,
Paris, France.
xavier.decleves@parisdescartes.fr
cliniques dans diverses pathologies
et notamment celles du SNC comme la maladie d’Alzheimer (MA). Malheureusement, la plupart de ces anticorps thérapeutiques ont vu leur
développement arrêté après les essais de phases II ou III par manque
d’efficacité ou du fait de la survenue d’effets secondaires au niveau
cérébral, comme des vasculopathies ayant entraîné des microhémorragies cérébrales et des œdèmes vasogéniques. La raison principale de
l’absence sur le marché d’anticorps thérapeutiques ayant pour cibles
des protéines du parenchyme cérébral est le maintien dans ces pathologies de barrières cérébrales intactes, limitant ainsi très efficacement
leur entrée dans le cerveau. On estime ainsi la fraction biodisponible
pour le tissu cérébral d’anticorps administrés par voie intraveineuse
à moins de 0,1 % de la dose injectée [1], ce qui est la plupart du
temps insuffisant pour espérer un effet thérapeutique. Le parenchyme
cérébral est en effet très efficacement protégé des molécules et cellules sanguines circulantes par deux barrières entre le sang et le tissu
cérébral : (1) la barrière hémato-liquidienne (BHL), entre le sang et
le liquide céphalo-rachidien (LCR), au niveau des espaces méningés
sous-arachnoïdiens et des plexus choroïdes localisés au niveau des
ventricules cérébraux, et (2) la barrière hémato-encéphalique (BHE)
qui sépare le parenchyme cérébral du sang au niveau de la plupart des
structures cérébrales de la substance blanche et grise. Bien que la BHL
soit environ 100 fois plus perméable que la BHE au passage passif de
m/s n° 12, vol. 35, décembre 2019
https://doi.org/10.1051/medsci/2019223
190016_Decleves_Synthese.indd 1106
05/12/2019 09:26:23
Pieds
des
astrocytes
Jonctions serrées
Péricyte
B
20 µm
SANG
REVUES
A
Cellule endothéliale
Membrane
basale
composés quelle que soit leur taille moléculaire, sa surface d’échange
est 100 fois plus faible [2]. Les molécules qui entrent plus facilement
dans le LCR à travers la BHL que dans le parenchyme cérébral à travers la BHE ont, de plus, une capacité de diffusion très limitée du LCR
vers le parenchyme cérébral. On estime la distance de diffusion d’une
molécule biologique depuis le LCR vers le parenchyme cérébral à moins
de 1 mm avec une concentration qui diminue de façon logarithmique à
partir de la surface du LCR. Même si l’on s’affranchit du passage d’un
médicament du sang dans le LCR en l’administrant directement dans
le LCR par voie intrathécale, le fort taux de renouvellement (3 à 4 fois
par jour) des 140 ml de LCR chez l’homme et sa convection (mouvement des fluides cérébraux) des ventricules vers les sinus veineux des
espaces sous-arachnoïdiens, entraînent une sortie rapide du médicament du LCR vers le sang. Ainsi, même si l’administration d’un médicament biologique par voie intraveineuse permet un meilleur passage
de celui-ci dans le LCR à travers la BHL, la faible surface d’échange de
cette barrière et les mécanismes rapides de clairance du LCR ne permettent pas une diffusion intracérébrale satisfaisante. Afin d’espérer
une efficacité thérapeutique des anticorps ayant pour cibles des
protéines intracérébrales dans des pathologies du SNC à BHE conservée, leur passage à travers la BHE doit donc être considérablement
augmenté. Nous présentons dans cette revue les progrès actuels à la
fois dans la compréhension des mécanismes moléculaires de transport à travers la BHE et dans les possibilités d’ingénierie de nouveaux
formats d’anticorps ciblant la BHE qui pourraient permettre de pallier
ces problèmes de distribution cérébrale des médicaments biologiques.
Propriétés de la BHE
La barrière hémato-encéphalique (BHE) est une structure essentielle
du cerveau puisqu’elle maintient l’homéostasie indispensable au bon
déroulement des différentes activités cérébrales. Cette barrière est
m/s n° 12, vol. 35, décembre 2019
190016_Decleves_Synthese.indd 1107
SYNTHÈSE
Figure 1. Structure histologique de la barrière hémato-encéphalique (BHE). A. Schéma d’un capillaire cérébral en coupe transversale montrant les
différentes structures cellulaires composant la BHE. B. Visualisation par microscopie confocale d’un vaisseau cérébral dans le cortex chez le rat. Les
astrocytes sont révélés par un immunomarquage de la GFAP (glial fibrillary acidic protein) (en jaune) et la membrane luminale des cellules endothéliales par un immunomarquage de la P-glycoprotéine (en magenta). Les noyaux cellulaires (cellules endothéliales, cellules murales, neurones,
cellules gliales) sont marqués par un intercalant de l’ADN (en cyan) (image reproduite avec la permission de Bruno Saubaméa, UMR-S 1144).
constituée d’une couche continue de cellules endothéliales jointives reposant sur une lame basale incrustée
de péricytes1 et solidement gainée par les pieds des
astrocytes qui recouvrent pratiquement totalement la
surface des microvaisseaux cérébraux (Figure 1) [3].
Plus d’un siècle après l’introduction du concept de BHE
par les médecins allemands Paul Ehrlich (qui a reçu en
1908 le prix Nobel de Physiologie ou Médecine, partagé
avec Elie Metchnikoff pour leurs travaux sur l’immunité)
et Edwin Goldman (un chirurgien disciple d’Ehrlich), de
très nombreux travaux ont contribué à décrire l’organisation anatomique, cellulaire et fonctionnelle de cette
barrière. Grâce aux progrès récents [4] réalisés dans
le domaine de la biologie cellulaire et moléculaire,
de la biochimie et de la neurocinétique des échanges
entre le sang et le cerveau, les principaux mécanismes
de transports au travers de la BHE ont été élucidés
(Figure 2). Ces mécanismes jouent un rôle majeur dans
la régulation et le maintien de l’homéostasie cérébrale
pour de nombreuses molécules endogènes (sucres,
acides aminés, métabolites, électrolytes, nucléosides,
métaux, vitamines, hormones, etc.), mais également
dans la protection du cerveau vis-à-vis de de composants présents dans la circulation générale et potentiellement neurotoxiques, qu’ils soient endogènes ou
xénobiotiques. La particularité des cellules constituant
la BHE par rapport aux cellules endothéliales périphériques réside dans la présence de jonctions intercellulaires serrées qui limitent très efficacement le passage
1
Les péricytes sont des cellules localisées au niveau de la lame basale de l’endothélium des capillaires. Ils l’entourent par de longs prolongements cytoplasmiques.
1107
05/12/2019 09:26:24
Via un récepteur
Diffusion passive transcellulaire
Transporteurs actifs membranaires
Diffusion passive
transcellulaire
Sang
Superfamille des SLC
Glucose, acides aminés,
vitamines, ions, purines
Superfamille des ABC
Xénobiotiques lipophiles
neutres ou chargés
Petites molécules
lipophiles
Jonctions
intercellulaires
serrées
ATP
Transcytose
Via un récepteur
Via une adsorption
Insuline, transferrine, leptine
Protéines basiques
ADP
Cerveau
Figure 2. Mécanismes généraux du passage des composés à travers la barrière hémato-encéphalique.
paracellulaire (entre les cellules) des composés, y compris ceux de très
faible masse moléculaire (de quelques dizaines de daltons). Le franchissement de la BHE par les médicaments ne peut donc être réalisé
qu’au travers des cellules endothéliales (par la voie transcellulaire)
et dépend de leurs propriétés physico-chimiques et/ou de leur affinité
pour les systèmes de transport qui sont exprimés sur la face luminale
(face sanguine) ou abluminale (face cérébrale) des cellules endothéliales constituant la BHE. Les transporteurs actifs de médicaments de
type solute carrier (SLC) et ATP-binding cassette (ABC) jouent un rôle
prépondérant dans ces échanges membranaires de molécules de faible
masse moléculaire (moins de 4 000 Da), les transporteurs SLC permettant principalement l’influx de composés du sang vers le cerveau et les
transporteurs ABC s’opposant à leur entrée dans le cerveau [5].
La transcytose est un mode particulier de passage à travers des membranes biologiques de molécules de plus haute masse moléculaire
(peptides, polypeptides, protéines). Elle repose essentiellement sur
deux mécanismes : la transcytose par adsorption et la transcytose
dépendant de récepteurs (Figure 2). Dans le cas de la transcytose par
adsorption (TMA), la formation des vésicules intracellulaires ayant
pour origine la membrane plasmique résulte de phénomènes d’attractions électrostatiques. Les protéines cationiques pénètrent ainsi dans
le cerveau via ce mécanisme qui est déclenché par une interaction
électrostatique entre des peptides ou des protéines polycationiques
et des microdomaines anioniques exprimés à la membrane des cellules
endothéliales du cerveau, telles que des glycoprotéines contenant des
résidus d’acide sialique à la membrane plasmique luminale, et des
héparane sulfates du côté abluminal. Le(s) mécanisme(s) exact(s)
d’interaction des protéines basiques avec la membrane, leur translocation dans les cellules, et leur échappement endosomal dans le
cytosol permettant une sortie ultérieure, restent peu documentés. Les
molécules internalisées se déplacent vers le pôle opposé de la cellule
et traversent à nouveau la membrane. Ce trafic dans le cytoplasme
1108
peut entraîner l’accumulation des protéines endocytées
dans le compartiment endosomal et/ou lysosomal où
elles seront recyclées ou dégradées. La TMA nécessite,
en général, une interaction avec des composés exprimés du côté luminal des cellules. Le mécanisme exact
n’est cependant pas complètement compris et savoir
si des récepteurs spécifiques sont impliqués ou si une
adsorption dépendant d’énergie est nécessaire reste à
déterminer.
La transcytose impliquant des récepteurs (ou TMR)
nécessite, quant à elle, l’interaction de la macromolécule avec un récepteur exprimé à la surface luminale
des cellules endothéliales de la barrière. Ce mécanisme
permet par exemple à des molécules de très haute
masse moléculaire, comme les lipoprotéines, l’insuline,
la transferrine et la leptine, de franchir la membrane.
Ces mécanismes de transcytose, TMA ou TMR, sont particulièrement étudiés afin d’améliorer le passage de
médicaments biologiques de haute masse moléculaire
au travers de la BHE.
Stratégies d’optimisation de la pénétration
cérébrale d’anticorps thérapeutiques
L’efficacité d’un anticorps thérapeutique dépend fortement de ses propriétés pharmacocinétiques. Ces
propriétés reposent sur plusieurs caractéristiques,
dont leur activité de liaison à leur cible, leur charge,
leur degré de glycosylation et leur capacité de recyclage. Les anticorps sont des protéines de haute masse
moléculaire, d’environ 150 kDa. Ils sont polaires et se
déplacent lentement entre les cellules endothéliales
m/s n° 12, vol. 35, décembre 2019
190016_Decleves_Synthese.indd 1108
05/12/2019 09:26:24
A
1
2
3
REVUES
Légendes des couleurs :
Anti-récepteur de l’insuline
Anti-récepteur de la transferrine
Anti-bêta-amyloïde
Iduronate sulfatase
4
B
5
6
SYNTHÈSE
Légendes des couleurs :
Anti-récepteur de la transferrine
Anti-bêta-amyloïde
Anti-BACE1
7
Figure 3. Nouveaux formats d’anticorps thérapeutiques à pénétration cérébrale renforcée. BACE1 : beta-secretase 1.
vasculaires périphériques du sang vers les espaces interstitiels des
tissus. Les concentrations les plus élevées d’anticorps sont ainsi
observées dans des tissus très richement perfusés et présentant un
réseau vasculaire lâche, tels que la moelle osseuse, la rate ou le foie.
La pénétration des anticorps est beaucoup plus limitée dans des tissus
comme le cerveau, protégé par la BHE, et les tumeurs dont les réseaux
de vascularisation sont très peu en contact avec la circulation lymphatique. Une des stratégies principales pour accroître la concentration d’anticorps thérapeutiques dans le tissu tumoral consiste à cibler
un ou plusieurs antigènes spécifiquement surexprimés par les tumeurs
[4]. Nous exposerons ici essentiellement les résultats obtenus et les
essais en cours visant à faciliter le passage des anticorps au travers
de la BHE et accroître ainsi l’exposition cérébrale à ces anticorps. Nous
nous limiterons aux anticorps administrés par voie systémique (ce qui
représente la majorité des cas). Nous ne traiterons donc pas des administrations par les autres voies (intrathécale, intracérébrale et intranasale notamment) ni les rares cas de formulations nanoparticulaires
ou d’administrations facilitées par des ultrasons.
Alors que les concentrations d’immunoglobulines (IgG) dans les
organes richement perfusés sont généralement comprises entre 10
et 50 % de leur taux plasmatique [6], celles dans le cerveau ne
représentent, en moyenne, que de 0,01 % à 0,4 % du taux plasmatique [1]. L’une des façons d’augmenter l’exposition cérébrale d’un
anticorps administré par voie systémique est donc de lui adjoindre
un « module » spécifique qui lui permettra de traverser la BHE par
transcytose.
Plusieurs stratégies ont été développées au cours des dernières
décennies pour administrer des protéines au cerveau. L’une d’elles
repose sur l’ajout de charges positives aux protéines ciblant la TMA.
L’équipe de William Pardridge, physiologiste américain de l’université de Californie à Los Angeles, avait en effet montré, dès 1987, que
les IgG cationisées par l’ajout de résidus polyamines, présentaient
une meilleure distribution cérébrale que des IgG natives [7]. Toutem/s n° 12, vol. 35, décembre 2019
190016_Decleves_Synthese.indd 1109
fois, l’affinité de ces anticorps cationisés est souvent
réduite, ce qui peut limiter leurs activités biologiques.
Le cytosol des cellules, qui est un milieu réducteur, peut
également altérer la structure des IgG en réduisant les
ponts disulfures intra-domaines, en particulier celui du
domaine variable, entraînant une modification conformationnelle avec, in fine, une perte de reconnaissance
de l’antigène par l’anticorps. La TMA n’est par ailleurs
pas spécifique d’un type de tissu ou d’une cellule. Des
effets secondaires touchant différentes cibles peuvent
donc apparaître en cas d’utilisation thérapeutique.
Enfin, ces mécanismes d’endocytose qui dépendent des
charges des molécules sont souvent considérés comme
étant toxiques.
D’autres stratégies ont été élaborées afin de cibler
les mécanismes de TMR. Elles consistent à utiliser des
modules transporteurs qui sont inspirés de ceux présents au niveau de la BHE elle-même et des mécanismes
qui lui permettent d’importer les protéines endogènes
qui lui sont indispensables. Les transporteurs les plus
étudiés aujourd’hui sont ceux de l’insuline et de la
transferrine. Des travaux pionniers, conduits par William Pardridge [8], ont montré l’intérêt de ces transporteurs pour véhiculer des protéines dans le cerveau
et ont inspiré une variété de formats d’anticorps multispécifiques qui ont permis d’augmenter leur exposition
cérébrale, parfois d’un facteur important. Ces nouveaux
formats sont, par exemple, des anticorps contre les
récepteurs de l’insuline ou de la transferrine auxquels
est fusionnée une protéine thérapeutique qui peut-être
une enzyme, un facteur de croissance, un anticorps
ou un de ses fragments (comme par exemple un scFv
[single chain fragment variable] ou un Fab).
1109
05/12/2019 09:26:24
Quelques constructions appartenant à cette catégorie (Figure 3A) sont
en développement clinique, comme des protéines de fusion constituées
de l’enzyme lysosomale iduronate sulfatase fusionnée avec un anticorps
contre le récepteur de l’insuline [9] (Figure 3A, -1-) ou avec un anticorps contre le récepteur de la transferrine [10] (Figure 3A, -2-), avec
lesquelles des études cliniques de phase II sont en cours pour traiter la
mucopolysaccharidose2. Plusieurs protéines de fusion de ce type ont
été construites avec des anticorps anti-bêta-amyloïde soit sous forme
de fragments ScFv (Figure 3A, -3-) [11] ou de fragments F(ab’)2 de ces
anticorps [12] (Figure 3A, -4-). Elles ont montré un intérêt dans des
modèles murins ou en imagerie de la maladie d’Alzheimer [13]. Des protéines de fusion avec d’autres protéines thérapeutiques ont été décrites,
comme celle fusionnée avec l’érythropoiëtine [14], ou avec une protéine
inhibitrice du TNF-a (tumor necrosis factor alpha) [15] ou avec des
enzymes lysosomales [16].
Une deuxième catégorie de constructions (Figure 3B) consiste en des
anticorps thérapeutiques dans lesquels un paratope3 spécifique a été
incorporé. Ce paratope peut reconnaître le récepteur de la transferrine,
soit sous forme de bispécifiques bivalents (Figure 3B, -6-)comme les
DVD (dual variable domain) [17] ou les TBTI (tetravalent bispecific
tandem IgG) [18] ou monovalents (Figure 3B, -7-) [19], ou résultant
de la fusion d’un domaine scFv (Figure 3B, -8-) [20]. Là encore, des
anticorps se sont révélés prometteurs dans la maladie d’Alzheimer [un
anticorps contre la protéine beta-amyloïde (Figure 3B, -6,8-) et un
anticorps contre l’enzyme BACE1 (beta-secretase 1) (Figure 3B, -7-) qui
est responsable des coupures de la protéine à l’origine des fragments
toxiques].
La pénétration des anticorps dans les tissus repose sur leur affinité pour
leurs cibles. Ce processus, qui a été décrit pour la distribution des anticorps au sein des tumeurs [21], a été également observé pour l’exposition d’anticorps dirigés contre le récepteur de la transferrine dans le
cerveau. Un laboratoire américain a ainsi montré que des mutations
touchant le paratope d’anticorps spécifiques de ce récepteur, affectant leur affinité pour leur cible, pouvaient augmenter, dans certaines
conditions, leur exposition dans le parenchyme cérébral [18]. Un grand
groupe pharmaceutique européen a également décrit le fait que certains
formats d’anticorps monovalents, spécifiques du récepteur de la transferrine (Figure 3B, -7-), qui ont une avidité réduite pour leur cible, présentent des expositions cérébrales augmentées [20]. Le potentiel de ces
avancées permettant d’accroître les expositions cérébrales d’anticorps
dirigés contre des récepteurs exprimés par la BHE est très prometteur
dans toutes les indications concernant le système nerveux central, que
ce soient les maladies neurodégénératives ou les désordres centraux
liés à certaines maladies génétiques, ou le glioblastome, les lymphomes
cérébraux et les métastases cérébrales. Quelques formats sont d’ores
et déjà en développement. Ils ont fait la preuve de leur innocuité.
Néanmoins, pour certains, comme les anticorps contre les récepteurs de
l’insuline ou de la transferrine, un ajustement très fin de leurs propriétés
reste nécessaire afin d’éviter, pour les uns, des problèmes de toxicité
2
3
1110
Maladie rare du stockage lysosomal due à des mutations alléliques du gène codant l’enzyme.
Partie de l’anticorps reconnaissant l’épitope porté par l’antigène.
liés à leurs effets sur la glycémie [22] ou des désordres
pancréatiques [23], ou, pour les autres, des désordres
sanguins ou médullaires [24]. L’avenir nous dira si ce
potentiel livre toutes ses promesses.
De nouveaux formats d’anticorps ont fait récemment leur
apparition. Ils sont issus d’immunoglobulines d’autres
espèces animales que l’homme. Les camélidés (chameaux, dromadaires, lamas, alpagas, etc.) ont la particularité de produire des anticorps conventionnels
possédant des chaînes lourdes et légères mais aussi
deux sous-classes d’IgG formées uniquement de dimères
de chaînes lourdes [25]. Le paratope de ces anticorps
à domaine unique est constitué d’un seul domaine VH,
appelé VHH ou nanobodyTM. Ces anticorps homodimériques dont la chaîne lourde se caractérise par l’absence
de domaine CH1, révèlent ainsi que le répertoire anticorps ne repose pas uniquement sur une combinatoire
VH-VL. Des anticorps à domaine unique ont aussi été
trouvés dans les poissons cartilagineux comme le requin.
Leurs fragments (VNAR) présentent le même type de propriétés que les VHH [26]. Un grand nombre d’anticorps à
domaine unique ayant une affinité élevée vis-à-vis d’un
antigène ont déjà été isolés par la technique du phage
display4 ou par ribosome display5. En raison de leur pouvoir de diffusion élevé lié à leur petite taille (15 kDa), de
leur sélectivité et de l’absence de réactions immunologiques lorsqu’ils sont injectés dans des animaux d’autres
espèces ou chez l’homme, ces VHH combinent les avantages des anticorps conventionnels à ceux des petites
molécules [27]. Ils sont aussi naturellement très stables :
ils supportent des températures élevées et la réduction
de leur pont disulfure intra-domaine n’entraîne pas de
perte de reconnaissance de l’antigène qu’ils ciblent.
Cette dernière propriété permet ainsi d’utiliser ces VHH,
après leur transfection à l’intérieur d’une cellule, comme
marqueur intracellulaire [28]. Environ 2/3 des VHH ont un
point isoélectrique basique. Ils sont donc naturellement
cationisés ce qui leur confère la capacité de traverser
facilement la BHE in vivo. Lafaye et al. ont en effet
montré que des VHH ayant un point isoélectrique basique
et dirigés contre la protéine fibrillaire gliale acide (glial
fibrillary acidic protein, ou GFAP), un marqueur spécifique des astrocytes, étaient capables, après leur injection intra-carotidienne chez la souris, de se lier à cette
protéine dans le cytoplasme des astrocytes [29]. Cette
capacité des VHH de traverser la barrière a été confir4
Cette technique permet l’expression de peptides ou de protéines exogènes à la
surface des particules de phages, afin de sélectionner et d’amplifier un polypeptide
capable d’interagir avec la molécule cible choisie.
5
Le ribosome display est une technique de sélection des ligands (ici un anticorps)
où le lien entre le phénotype et le génotype est réalisé par un complexe ternaire
formé entre la protéine traduite, le ribosome et l’ARN messager correspondant.
m/s n° 12, vol. 35, décembre 2019
190016_Decleves_Synthese.indd 1110
05/12/2019 09:26:24
Conclusion
Le cerveau est probablement l’ultime frontière pour les anticorps, aussi
bien pour le diagnostic que pour la thérapie. Pour qu’une approche
cérébrale de la thérapie par anticorps soit efficace, la définition des
cibles est cruciale. Ceci implique une meilleure compréhension des
mécanismes physiopathologiques impliqués dans le développement
des pathologies touchant le cerveau. L’étude des mécanismes permettant le passage des anticorps au travers de la BHE nécessite également d’explorer de nouvelles pistes, comme par exemple l’ouverture
transitoire de la BHE en utilisant des ultrasons, ce qui pourrait faciliter
l’entrée de ces anticorps au sein du cerveau [31, 32]. ‡
SUMMARY
New formats for improving brain drug delivery of antibodies: the
blood-brain barrier case
Many neurodegenerative or tumor brain pathologies should be able to
benefit from the impressive medicinal advances that represent therapeutic antibodies. Unfortunately, many failures have been observed
with antibodies whose targets are in the brain parenchyma due to their
very low brain distribution. The blood-brain barrier (BBB) that exhibits
extremely selective and restrictive properties is responsible for the
low brain penetration of high-molecular mass molecules including
therapeutic antibodies. The objective of this article is to present the
properties of the BBB and the latest advances in the engineering of new
antibody formats to possibly improve their brain distribution. ‡
LIENS D’INTÉRÊT
Les auteurs déclarent n’avoir aucun lien d’intérêt concernant les données publiées dans
cet article.
RÉFÉRENCES
1. Pepinsky RB, Shao Z, Ji B, et al. Exposure levels of anti-LINGO-1 Li81 antibody in the central
nervous system and dose-efficacy relationships in rat spinal cord remyelination models after
systemic administration. J Pharmacol Exp Ther 2011 ; 339 : 519-29.
m/s n° 12, vol. 35, décembre 2019
190016_Decleves_Synthese.indd 1111
REVUES
2. Strazielle N, Ghersi-Egea JF. Physiology of blood-brain interfaces in relation
to brain disposition of small compounds and macromolecules. Mol Pharm
2013 ; 10 : 1473-91.
3. Abbott NJ. Blood-brain barrier structure and function and the challenges for
CNS drug delivery. J Inherit Metab Dis 2013 ; 36 : 437-49.
4. Husain B, Ellerman D. Expanding the boundaries of biotherapeutics with
bispecific antibodies. BioDrugs 2018 ; 32 : 441-64.
5. Chaves C, Shawahna R, Jacob A, et al. Human ABC transporters at blood-CNS
interfaces as determinants of CNS drug penetration. Curr Pharm Des 2014 ;
20 : 1450-62.
6. Lobo ED, Hansen RJ, Balthasar JP. Antibody pharmacokinetics and
pharmacodynamics. J Pharm Sci 2004 ; 93 : 2645-68.
7. Kumagai AK, Eisenberg JB, Pardridge WM. Absorptive-mediated endocytosis
of cationized albumin and a beta-endorphin-cationized albumin chimeric
peptide by isolated brain capillaries. Model system of blood-brain barrier
transport. J Biol Chem 1987 ; 262 : 15214-9.
8. Pardridge WM. Delivery of biologics across the blood-brain barrier with
molecular trojan horse technology. BioDrugs 2017 ; 31 : 503-19.
9. Pardridge WM, Boado RJ, Giugliani R, Schmidt M. Plasma pharmacokinetics of
valanafusp alpha, a human insulin receptor antibody-iduronidase fusion protein,
in patients with mucopolysaccharidosis type I. BioDrugs 2018 ; 32 : 169-76.
10. Sonoda H, Morimoto H, Yoden E, et al. A Blood-brain-barrier-penetrating
anti-human transferrin receptor antibody fusion protein for neuronopathic
mucopolysaccharidosis II. Mol Ther 2018 ; 26 : 1366-74.
11. Sumbria RK, Zhou QH, Hui EK, et al. Pharmacokinetics and brain uptake
of an IgG-TNF decoy receptor fusion protein following intravenous,
intraperitoneal, and subcutaneous administration in mice. Mol Pharm 2013 ;
10 : 1425-31.
12. Sehlin D, Fang XT, Meier SR, et al. Pharmacokinetics, biodistribution and
brain retention of a bispecific antibody-based PET radioligand for imaging
of amyloid-beta. Sci Rep 2017 ; 7 : 17254.
13. Syvanen S, Fang XT, Hultqvist G, et al. A bispecific Tribody PET radioligand
for visualization of amyloid-beta protofibrils - a new concept for
neuroimaging. NeuroImage 2017 ; 148 : 55-63.
14. Chang R, Al Maghribi A, Vanderpoel V, et al. Brain penetrating bifunctional
erythropoietin-transferrin receptor antibody fusion protein for Alzheimer’s
disease. Mol Pharm 2018 ; 15 : 4963-73.
15. Chang R, Knox J, Chang J, et al. Blood-brain barrier penetrating biologic TNFalpha inhibitor for Alzheimer’s disease. Mol Pharm 2017 ; 14 : 2340-9.
16. Boado RJ, Lu JZ, Hui EK, et al. Insulin receptor antibody-alpha-Nacetylglucosaminidase fusion protein penetrates the primate blood-brain
barrier and reduces glycosoaminoglycans in Sanfilippo type B fibroblasts.
Mol Pharm 2016 ; 13 : 1385-92.
17. Karaoglu Hanzatian D, Schwartz A, Gizatullin F, et al. Brain uptake
of multivalent and multi-specific DVD-Ig proteins after systemic
administration. mAbs 2018 ; 10 : 765-77.
18. TM. Do, I. Arnould, J. Beninga, et al. Brain exposure and therapeutic efficacy
of multivalent bispecific anti-TfRC antibodies. Abstracts from the 22nd
International Symposium on signal transduction at the blood-brain barriers.
Fluids Barriers CNS 2019 ; 16 (suppl 2) : 29.
19. Yu YJ, Zhang Y, Kenrick M, et al. Boosting brain uptake of a therapeutic
antibody by reducing its affinity for a transcytosis target. Sci Transl Med
2011 ; 3 : 84ra44.
20. Niewoehner J, Bohrmann B, Collin L, et al. Increased brain penetration and
potency of a therapeutic antibody using a monovalent molecular shuttle.
Neuron 2014 ; 81 : 49-60.
21. Adams GP, Schier R, McCall AM, et al. High affinity restricts the localization
and tumor penetration of single-chain fv antibody molecules. Cancer Res
2001 ; 61 : 4750-5.
22. Boado RJ, Hui EK, Lu JZ, Pardridge WM. Glycemic control and chronic dosing
of rhesus monkeys with a fusion protein of iduronidase and a monoclonal
antibody against the human insulin receptor. Drug Metab Dispos 2012 ; 40 :
2021-5.
23. Ohshima-Hosoyama S, Simmons HA, Goecks N, et al. A monoclonal antibodyGDNF fusion protein is not neuroprotective and is associated with proliferative
pancreatic lesions in parkinsonian monkeys. PLoS One 2012 ; 7 : e39036.
24. Pardridge WM, Boado RJ, Patrick DJ, et al. Blood-brain barrier transport,
plasma pharmacokinetics, and neuropathology following chronic treatment
of the rhesus monkey with a brain penetrating humanized monoclonal
antibody against the human transferrin receptor. Mol Pharm 2018 ; 15 :
5207-16.
25. Hamers-Casterman C, Atarhouch T, Muyldermans S, et al. Naturally occurring
antibodies devoid of light chains. Nature 1993 ; 363 : 446-8.
SYNTHÈSE
mée récemment avec des VHH dirigés contre les deux principales lésions
de la maladie d’Alzheimer : les plaques amyloïdes intra-cérébrales et
les dégénérescences neurofibrillaires (DNF) intra-neuronales. Ces VHH
marqués par un fluorochrome vert ont en effet été testés in vivo dans
deux modèles murins de maladie d’Alzheimer. Après injection par voie
intraveineuse, les VHH traversent la BHE et interagissent avec les plaques
amyloïdes et les DNF [30] ; une étude pharmacocinétique utilisant le
VHH dirigé contre les plaques amyloides montre que, deux heures après
l’injection intraveineuse, 0,5 % de la dose injectée est retrouvé dans le
cerveau. Le marquage intracérébral que permettent ces 3 VHH repose sur
un ensemble de propriétés que les anticorps conventionnels n’ont pas :
(1) la capacité de traverser la BHE et d’entrer dans les cellules, car ils
sont cationisés ; (2) une petite taille permettant leur diffusion dans le
tissu cérébral ; (3) la reconnaissance spécifique d’un antigène intracellulaire grâce à leur très grande stabilité ; et, enfin, (4) leur capacité à
transporter des molécules pour le diagnostic ou la thérapie.
1111
05/12/2019 09:26:24
RÉFÉRENCES
26. Feng M, Bian H, Wu X, et al. Construction and next-generation sequencing analysis of a large
phage-displayed VNAR single-domain antibody library from six naïve nurse sharks”. Antibody
Therapeutics 2019 ; 2 h 1-11.
27. Nguyen VK, Desmyter A, Muyldermans S. Functional heavy-chain antibodies in Camelidae. Adv
Immunol 2001 ; 79 : 261-96.
28. Traenkle B, Rothbauer U. Under the microscope: Single-domain antibodies for live-cell imaging
and super-resolution microscopy. Front Immunol 2017 ; 8 : 1030.
29. Li T, Bourgeois JP, Celli S, et al. Cell-penetrating anti-GFAP VHH and corresponding fluorescent
fusion protein VHH-GFP spontaneously cross the blood-brain barrier and specifically recognize
astrocytes: application to brain imaging. FASEB J 2012 ; 26 : 3969-79.
30. Li T, Vandesquille M, Koukouli F, et al. Camelid single-domain antibodies:
a versatile tool for in vivo imaging of extracellular and intracellular brain
targets. J Control Release 2016 ; 243 : 1-10.
31. Carpentier A, Canney M, Vignot A, et al. Clinical trial of blood-brain barrier
disruption by pulsed ultrasound. Sci Transl Med 2016 ; 8 : 343re2.
32. Santin MD, Debeir T, Bridal SL, et al. Fast in vivo imaging of amyloid plaques
using mu-MRI Gd-staining combined with ultrasound-induced blood-brain
barrier opening. NeuroImage 2013 ; 79 : 288-94.
TIRÉS À PART
X. Declèves
pour une meilleure
visibilité des résultats
de la recherche
en santé publique
es résultats de la recherche en santé publique
souffrent en France d’un réel manque de visibilité. Ceci concerne aussi bien le monde
académique (hors santé publique) que le
grand public et les décideurs. Pour pallier ce
déficit, l’IReSP a créé un bulletin à large diffusion intitulé
« Questions de santé publique », largement inspiré du
bulletin mensuel d’information de l’INED « Populations et
sociétés ». L’objectif éditorial est de porter à la connaissance d’un large public (enseignants, étudiants, journalistes, décideurs, milieux de la recherche, associations,
public concerné) les informations les plus récentes
concernant des questions importantes de santé publique,
rédigées de façon facilement lisible et compréhensible
pour des non spécialistes, en garantissant que les informations publiées sont validées scientifiquement. La
publication concerne des faits et non des positions. Audelà de la présentation de résultats, les qualités pédagogiques de Questions de santé publique permettent au
lecteur de mieux comprendre comment sont formulées et
abordées les questions de santé publique et quelles sont
les limites de ces études.
01 49 85 03 45
Questions de santé publique
EDP Sciences
17 avenue du Hoggar
91944 Les Ulis
France
Réservé aux abonnés de M/S
Recevez gratuitement et régulièrement
Questions de santé publique
en renvoyant ce document soigneusement rempli.
Questions de santé publique est une publication de l’Institut de Recherche en Santé Publique.
Directeur de la publication : Corinne Alberti.
Rédactrice en chef : Kodja Yetongnon. Comité de relecture : Lorraine Cousin, Jean-Marie Gagliolo, Coline Terroba. Réalisation : EDP Sciences.
1112
m/s n° 12, vol. 35, décembre 2019
190016_Decleves_Synthese.indd 1112
09/12/2019 16:11:13
|
|
https://openalex.org/W2339040768
|
https://europepmc.org/articles/pmc4943526?pdf=render
|
English
| null |
Adolescent smoking and tertiary education: opposing pathways linking socio‐economic background to alcohol consumption
|
Addiction
| 2,016
|
cc-by
| 6,820
|
ABSTRACT Background and Aims
If socio-economic disadvantage is associated with more adolescent smoking, but less partic-
ipation in tertiary education, and smoking and tertiary education are both associated with heavier drinking, these may
represent opposing pathways to heavy drinking. This paper examines contextual variation in the magnitude and direc-
tion of these associations. Design
Comparing cohort studies. Setting
United Kingdom. Participants
Participants
were from the 1958 National Child Development Study (NCDS58; n = 15 672), the British birth cohort study
(BCS70;
n
=
12
735)
and
the
West
of
Scotland
Twenty-07
1970s
cohort
(T07;
n
=
1515). Measurements
Participants self-reported daily smoking and weekly drinking in adolescence (age 16 years) and
heavy drinking (> 14/21 units in past week) in early adulthood (ages 22–26 years). Parental occupational class
(manual versus non-manual) indicated socio-economic background. Education beyond age 18 was coded as tertiary. Models were adjusted for parental smoking and drinking, family structure and adolescent psychiatric distress. Findings
Respondents from a manual class were more likely to smoke and less likely to enter tertiary education
(e.g. in NCDS58, probit coefficients were 0.201 and –0.765, respectively; P < 0.001 for both) than respondents from
a non-manual class. Adolescent smokers were more likely to drink weekly in adolescence (0.346; P < 0.001) and
more likely to drink heavily in early adulthood (0.178; P < 0.001) than adolescent non-smokers. Respondents who
participated in tertiary education were more likely to drink heavily in early adulthood (0.110 for males, 0.182 for fe-
males; P < 0.001 for both) than respondents with no tertiary education. With some variation in magnitude, these
associations were consistent across all three cohorts. Conclusions
In Britain, young adults are more likely to drink
heavily both if they smoke and participate in tertiary education (college and university) despite socio-economic back-
ground being associated in opposite directions with these risk factors. Keywords
Alcohol, education, life-course, pathways, smoking, socio-economic position. Correspondence to: Michael J. Green, MRC/CSO Social and Public Health Sciences Unit, University of Glasgow, 200 Renfield Street, Glasgow G2 3QB, UK. E-mail: michael.green@glasgow.ac.uk
Submitted 21 August 2015; initial review completed 20 November 2015; final version accepted 15 February 2016 Submitted 21 August 2015; initial review completed 20 November 2015; final version accepted 15 February 2016 RESEARCH REPORT doi:10.1111/add.13365 Adolescent smoking and tertiary education: opposing
pathways linking socio-economic background to alcohol
consumption Adolescent smoking and tertiary education: opposing
pathways linking socio-economic background to alcohol
consumption
Michael J. Green1, Alastair H. Leyland1, Helen Sweeting1 & Michaela Benzeval1,2
MRC/CSO Social and Public Health Sciences Unit, University of Glasgow, Glasgow, UK1 and Institute for Social and Economic Research, University of Essex, Colchester, UK2
RESEARCH REPORT
doi:10.1111/add.13365 Adolescent smoking and tertiary education: opposing
pathways linking socio-economic background to alcohol
consumption
Michael J. Green1, Alastair H. Leyland1, Helen Sweeting1 & Michaela Benzeval1,2
MRC/CSO Social and Public Health Sciences Unit, University of Glasgow, Glasgow, UK1 and Institute for Social and Economic Research, University of Essex, Colchester, UK2
RESEARCH REPORT
doi:10.1111/add.13365 © 2016 The Authors. Addiction published by John Wiley & Sons Ltd on behalf of Society for the Study of Addiction
Addiction, 111, 1457–1465
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium,
provided the original work is properly cited. Addiction, 111, 1457–1465 Correspondence to: Michael J. Green, MRC/CSO Social and Public Health Sciences Unit, University of Glasgow, 200 Renfield Street, Glasgow G2 3QB, UK.
E-mail: michael.green@glasgow.ac.uk
Submitted 21 August 2015; initial review completed 20 November 2015; final version accepted 15 February 2016 Tertiary education pathway The second pathway examined here is tertiary education
(meaning post-secondary school education undertaken
for example in universities or further/vocational education
colleges). Young people from more advantaged back-
grounds are more likely to enter tertiary education [15],
and students in tertiary education drink more heavily than
similar-aged peers outside tertiary education [16–18]. Thus, tertiary education could be a pathway promoting
heavier drinking which operates more frequently among
those
from
a
more
advantaged
socio-economic
background. INTRODUCTION more common among more advantaged adolescents. Opposing pathways could result in no association between
SEP and drinking, or associations in either direction. Deve-
loping a better understanding of the stratification of path-
ways leading to (heavy) drinking could lead to more
effective and targeted interventions or policies to prevent
it. This paper therefore explores two probable opposing
pathways between parental socio-economic position and
drinking in adolescence and early adulthood—smoking
and tertiary education. Analyses are undertaken in three
different cohorts to assess how the findings vary across
time and place. Socio-economic inequalities in excessive alcohol consump-
tion are inconsistent in both adolescence [1–3] and early
adulthood [3–6], which are key developmental periods for
drinking [7–9]. Some studies show no relationship, others
show positive and yet others negative associations [1,4]. It has been suggested that these inconsistent findings result
from pathways associated with socio-economic position
(SEP) working in opposing directions [4]; while some path-
ways leading to increased drinking are more common
among more disadvantaged adolescents, others may be 1458 Michael J. Green et al. Aim and hypotheses The aim of this paper is to investigate two pathways
(smoking and tertiary education) between SEP and drink-
ing in adolescence and early adulthood. We compare three
UK cohorts representing different historical and geograph-
ical contexts. Specifically, we hypothesize that: Smoking pathway might be expected between background SEP and outcomes
such as smokingor tertiaryeducation. Additionally, stronger
associations between background SEP and smoking and ter-
tiary education might be expected in T07: smoking, as a
coping mechanism or feature of social life, may have been
especially likely for young people from disadvantaged back-
grounds here as manual industries declined [19] and jobs
became concentrated in the South of the United Kingdom
[21]; Scotland has also tended to have higher overall rates
of participation in tertiary education than elsewhere in the
United Kingdom, but with wider inequalities [22]. Young people from a disadvantaged SEP are more likely to
smoke, and to start smoking earlier [10–12]. Smoking, in
turn, is often described as a ‘gateway drug’, associated with
onset of alcohol use and alcohol problems [13]. Previous
analysis of West of Scotland data (also analysed here) found
that late adolescent heavy drinkers from disadvantaged
backgrounds tended to have smoked prior to drinking
heavily, whereas those from more advantaged back-
grounds had rarely smoked [14]. This suggests that
smoking may be a pathway operating more frequently
among those from a disadvantaged socio-economic back-
ground, although it is not yet clear whether this pattern
extends into early adulthood or whether it would be repli-
cated in other contexts. Heterogeneity might also be expected in associations
between these mediating factors and drinking. Alcohol
has become more available in the United Kingdom between
the two time-periods examined [23], increasing opportuni-
ties for consumption. If smoking or tertiary education in-
crease individual motivation to drink, the association
may be stronger in more recent cohorts, where motiva-
tions could be acted upon more easily. This strengthening
of association might still be expected despite temporal
trends in the prevalence of these mediating factors, such
as declining smoking rates [23] or increasing participation
in tertiary education [15], assuming that a change in the
prevalence of the mediator does not change the nature of
its effect on drinking motivation. © 2016 The Authors. Addiction published by John Wiley & Sons Ltd on behalf of Society for the Study of Addiction Contextual variation Contextual heterogeneity may occur either in the associa-
tions between SEP and these mediating factors (smoking
and tertiary education) or in the associations between
those mediators and drinking. Therefore, these pathways
are explored with data from three different cohorts: the
UK 1958 National Child Development Study (NCDS58),
the 1970 British Birth Cohort Study (BCS70) and the
1970s cohort of the West of Scotland Twenty-07 Study
(T07). NCDS58 and BCS70 include people from across
Great Britain born 12 years apart, comparing different
historical contexts within the same geographical area. T07 respondents were from approximately the same
time-period as BCS70, but from the specific geographic
context in and around Glasgow, a large urban city which
had been experiencing rapid deindustrialization. • a disadvantaged socio-economic background will be as-
sociated with a higher likelihood of adolescent smoking
which, in turn, will be associated with heavier drinking
in adolescence and early adulthood; • an advantaged socio-economic background will be asso-
ciated with a higher likelihood of participation in tertiary
education which, in turn, will be associated with heavier
drinking in early adulthood; • socio-economic background will be associated more
strongly with smoking and tertiary education in more re-
cent cohorts, and most strongly in T07; and • smoking and tertiary education will be associated more
strongly with drinking in the two more recent cohorts. METHODS Variation in associations between SEP and mediating
factors might be expected in the United Kingdom between
the two time-periods examined, as labour markets shifted
from manual to non-manual occupations [19,20] and in-
come distributions became more unequal [15,21]. Thus,
some indicators of SEP may indicate greater relative disad-
vantage in more recent cohorts, so stronger associations Participants NCDS58 follows children born within Great Britain in
1958 [24]. This paper primarily uses data from follow-up
surveys in adolescence (mean age = 16.0 years) and early
adulthood (mean age = 23.6 years, in 1981), although Pathways linking socio-economic background to alcohol consumption 1459 after age 18. Background SEP was indicated by parental
occupational class, coded according to the British Registrar
General’s classification [28] as either non-manual (I, II and
III non-manual) or manual (III manual, IV and V) using
the highest status from couple parents. Sensitivity analyses
utilized measures of income (contrasting the lowest tertile
of equivalized household income with all others) and pa-
rental education (contrasting parents who had left school
by age 16 with those who remained longer). data from earlier surveys were also used for weighting and
imputation. A total of 15 672 respondents had valid data
for analysis, having participated in either the adolescent
or early adult follow-up. This comprised 84.4% of the total
sample of 18 558 (17 415 baseline respondents plus 1143
immigrants and others not interviewed at baseline; base-
line response rate = 98.8%). BCS70 is similar to NCDS58, following a cohort born
within Great Britain in 1970 (Northern Irish births were
also included at baseline but never followed-up, so excluded
here [24]). Data were taken primarily from follow-ups in
adolescence (1986; mean age = 16.1 years) and early
adulthood (1996; approximate age = 26 years), with data
from earlier surveys used for weighting and imputation. Valid data for analysis from either the adolescent or early
adult follow-ups were available for 12 735 respondents
(66.7% of the total sample of 18 488; 16 568 at baseline
plus 1920 immigrants and others missed at baseline; base-
line response rate = 95.8%). Parental smoking, parental drinking, family structure
and adolescent psychiatric distress were considered possible
confounders. Parental smoking was reported by parents
during adolescent surveys (and also reported by adolescents
in BCS70). For parental drinking: in NCDS58 interviewers
at age 7 indicated whether the family suffered from prob-
lems with alcoholism; in BSC70 parent or child reports of
a parent drinking ‘three or four times a week’ or more or
on ‘most days’ were coded as heavy parental drinking;
and for T07 either parent consuming more than 14 units
(women) and 21 units (men) was coded as heavy parental
drinking. Family structure distinguished between single-
and two-parent families. Participants Adolescent psychiatric distress
was indicated by the 12-item General Health Questionnaire
in BCS70 and T07, with scores of 3 or more indicating psy-
chiatric distress [29,30]. For NCDS58, psychiatric distress
was indicated by scores of 2 or more on the neuroticism
component of the Rutter behavioural scale [31]. T07 has followed three cohorts of people from in and
around Glasgow for 20 years [25]. The youngest cohort,
analysed here, had a mean age of 15.7 years at baseline
in 1987. The baseline sample (n = 1515; response
rate = 85%) was representative of the population within
the sampled area [26], and all cases were included in the
analysis. Data were primarily from baseline and a 1994 fol-
low-up in early adulthood (n = 1181; mean age = 21.7
years). © 2016 The Authors. Addiction published by John Wiley & Sons Ltd on behalf of Society for the Study of Addiction Measures Figure 1 depicts the analysis model, tested using structural
equation modelling (SEM) in Mplus 7 [32]. The positive
and negative signs indicate the hypothesized directions of
association for smoking and tertiary education pathways
linking SEP and drinking. The additional associations in
the model are not the focus, and hence no specific hypoth-
eses were made about them. Models were estimated using a
robust weighted least-squares estimator (WLSMV) with
Probit parameter estimates. Standard errors were calcu-
lated with bootstrapping. SEM provides model-fit statistics
for
the
overall
model
and
facilitates
cross-group Drinking was self-reported in all cohorts in adolescence
and early adulthood. Weekly drinking in adolescence was
based on reported alcohol consumption within the last
week (NCDS58, BCS70) or drinking frequency (BCS70,
T07). In BCS70, data from the question on frequency were
preferred over past week consumption (which may have
been atypical), but the latter was used if frequency data
were missing (n = 332). In early adulthood, respondents
in all three cohorts reported their past week’s drinking,
and numbers of alcohol units were derived. Drinking in ex-
cess of 14 units for women and 21 units for men [27] was
coded as heavy drinking. Figure 1 Analysis model and hypothesized direction of effects All cohort members self-reported smoking in adoles-
cence. As daily smoking would indicate an established
habit, the closest indication of daily smoking available
within each cohort was utilized: smoking 10 or more ciga-
rettes weekly in NCDS58; six or more weekly in BCS70;
and seven or more weekly in T07. Precise wordings of
questions on smoking and drinking are included in
Supporting information, Table S1. Based on detailed histories of economic activity from
age 16, respondents were coded as participating in tertiary
education if they had reported being in full-time education Figure 1 Analysis model and hypothesized direction of effects Michael J. Green et al. 1460 Figure 2 shows probit regression coefficients and boot-
strapped standard errors from the confounder-adjusted
analysis model (confidence intervals and P-values are pre-
sented in Supporting information, Table S3). Models that
did not include confounder adjustment had similar results,
although coefficients were a little larger (not shown). Sep-
arate estimates are provided where Wald tests for cohort
or gender differences were significant (P < 0.05). Results
from sensitivity analyses with income or education as mea-
sures of SEP were largely consistent (see Supporting infor-
mation, Tables S4 and S5). comparisons of parameter estimates. Smoking pathway Adolescent smoking was more likely for respondents from
manual than non-manual households. The association
was stronger in T07 than in NCDS58 and BCS70
(P < 0.05, although this cohort difference was not repli-
cated for parental education and income). Adolescent smoking was associated with more adoles-
cent weekly drinking. This association was stronger in
BCS70 and T07 than in NCDS58 (P < 0.05). Adolescent
smoking was associated independently with heavy drink-
ing in early adulthood, but between-cohort differences in
this association were not significant. Descriptive statistics and missing data Table 1 displays descriptive statistics and information on
missing data from within each cohort. Adolescent daily
smoking was lower in the more recent cohorts. Adolescent
weekly drinking was higher in BCS70 than NCDS58 and
particularly low in T07. Respondents in NCDS58 were
most likely to come from manual households, with little dif-
ference between BCS70 and T07 in this regard. Participa-
tion in tertiary education was higher in the more recent
cohorts and highest in the Scottish cohort. Heavy drinking
in early adulthood was highest in T07 and lowest in BCS70
respondents. Measures A cross-classified
grouping variable based on gender and cohort was used
to examine gender and cohort differences. Thresholds for
categorical dependent variables were permitted to vary by
gender and cohort, and a Wald test was used to examine
differences in coefficients (gender differences were tested
first, then cohort differences). Respondents within each cohort had missing data, so
multiple-imputation and inverse probability weighting
were employed [33]. These adjust for missing values to
the extent that they can be predicted by observed variables
[34]. Weighting adjusted for differences between the analy-
sis and baseline samples of NCDS58 and BCS70, with
weights for these cohorts calculated using relevant baseline
variables (respondents who were male, had low birth
weight, came from single-parent families or, in BCS70,
came from disadvantaged families, were more likely to have
dropped out; results not shown). Weighting was unneces-
sary for T07, as all respondents had some valid data. Mul-
tiple imputation (25 imputations) was used to obtain full
data on all variables for respondents in the analysis samples
within each cohort. Imputation models included additional
SEP indicators and variables often associated with smoking
and drinking (Supporting information, Table S2 provides
details). Weights were included in the imputation models
and used to weight the analyses of the imputed data [33]. RESULTS Respondents from manual compared to non-manual
households were less likely to participate in tertiary educa-
tion. The association was strongest in NCDS58 and
weakest in BCS70 (P < 0.05, although this cohort differ-
ence was not replicated for parental education or income). Tertiary education was associated with heavier adult
drinking, and this association was stronger for females
than for males (P < 0.05). © 2016 The Authors. Addiction published by John Wiley & Sons Ltd on behalf of Society for the Study of Addiction DISCUSSION Pathways linking SEP and drinking in adolescence and
early adulthood were investigated in three UK cohort stud-
ies. In all cohorts, socio-economic disadvantage was associ-
ated with higher chances of smoking in adolescence, and
adolescent smoking was associated with heavier drinking
in adolescence and early adulthood. However, disadvan-
taged adolescents were less likely to participate in tertiary
education, and tertiary education was also associated with
heavier drinking in early adulthood, especially for females. Both pathways leading to heavier drinking were associated
with SEP, but operated in opposing directions. Despite some
variation in magnitude, these opposing pathways were ob-
served consistently across the three studies, for three differ-
ent measures of SEP and for males and females, suggesting
that they represent consistent phenomena. This analysis Response rates in adolescence (81.5–100.0%) and
early adulthood (70.7–80.0%) were reasonable. However,
additional item-non-response among participants in each
survey (particularly evident in BCS70; attributed to a
teachers’ strike which coincided with the adolescent in-
school surveys) meant that there were relatively low pro-
portions of respondents with complete data on all analysis
variables: 36.2% in NCDS58, 19.5% in BCS70 and 61.7%
in T07. Nevertheless, as Table 1 indicates, sample propor-
tions for most characteristics remained similar after
weighting and imputation, indicating that missing data
rates did not differ substantially in terms of the analysis
variables. Addiction, 111, 1457–1465 Addiction, 111, 1457–1465 Pathways linking socio-economic background to alcohol consumption 1461 Table 1 Descriptive statistics and missing data. Table 1 Descriptive statistics and missing data. aUnweighted data. In order to facilitate comparisons with weighted/imputed data, percentages are based on those with valid responses, except those for
missing categories which use the analysis sample as the denominator. bPercentages are based on weighted average results across 25 imputed data sets.
cPercentages are based on average results across 25 imputed data sets. dIn NCDS58 Parental Heavy Drinking was ascertained by interviewers during
interviews at age 7, whereas in BCS70/T07 it was reported at age 16 by parents and/or respondents. eThere are more valid responses than those participating
in the early adulthood survey here because supplementary data from an intervening interview at age 18 were also used to indicate participation. DISCUSSION NCDS58
BCS70
T07
15 672
12 735
1515
Observeda
Weighted and
Imputedb
Observeda
Weighted and
Imputedb
Observeda
Imputedc
Analysis, n
n
%
n
%
n
%
n
%
n
%
n
%
Gender
Male
8032
51.3
8102
51.7
6279
49.3
6648
52.2
737
48.6
737
48.6
Female
7640
48.7
7570
48.3
6456
50.7
6087
47.8
778
51.4
778
51.4
Adolescent measures (age 16)
Participated in
adolescence
No
1307
8.3
2362
18.5
0
0.0
Yes
14 365
91.7
10 373
81.5
1515
100.0
Daily smoking
No
8752
73.1
11 331
72.3
5269
81.1
10 048
78.9
1273
84.5
1280
84.5
Yes
3217
26.9
4341
27.7
1224
18.9
2687
21.1
234
15.5
235
15.5
Missing
3703
23.6
6242
49.0
8
0.5
Weekly drinking
No
6497
54.1
8526
54.4
3068
47.8
6062
47.6
1424
94.3
1429
94.3
Yes
5509
45.9
7146
45.6
3345
52.2
6673
52.4
86
5.7
86
5.7
Missing
3666
23.4
6322
49.6
5
0.3
Parental occupational
class
Non-manual
5538
49.6
7711
49.2
4430
65.3
7475
58.7
891
59.8
904
59.7
Manual
5633
50.4
7961
50.8
2350
34.7
5260
41.3
598
40.2
611
40.3
Missing
4501
28.7
5955
46.8
26
1.7
Parental smoking
No
3232
27.8
4341
27.7
4121
41.8
5145
40.4
398
28.4
408
26.9
Yes
8377
72.2
11 331
72.3
5734
58.2
7590
59.6
1002
71.6
1107
73.1
Missing
4063
25.9
2880
22.6
115
7.6
Parental heavy
drinkingd
No
11 467
98.9
15 500
98.9
6560
68.5
8800
69.1
1162
83.5
1262
83.3
Yes
124
1.1
172
1.1
3015
31.5
3935
30.9
230
16.5
253
16.7
Missing
4081
26.0
3160
24.8
123
8.1
Family structure
Single-parent
1026
8.8
1395
8.9
561
10.3
1477
11.6
202
13.7
211
13.9
Two-parent
10 660
91.2
14 277
91.1
4872
89.7
11 258
88.4
1273
86.3
1304
86.1
Missing
3986
25.4
7302
57.3
40
2.6
Adolescent psychiatric
distress
No
10 129
82.4
12 851
82.0
3548
71.7
9233
72.5
1193
84.7
1285
84.8
Yes
2161
17.6
2821
18.0
1402
28.3
3502
27.5
215
15.3
230
15.2
Missing
3382
21.6
7785
61.1
107
7.1
Early adulthood measures (aged 22–26)
Participated in
early adulthood
No
3135
20.0
3732
29.3
334
22.0
Yes
12 537
80.0
9003
70.7
1181
78.0
Tertiary education
participation
No
9945
79.3
12 538
80.0
6235
70.1
9195
72.2
885
63.7
983
64.9
Yes
2592
20.7
3134
20.0
2658
29.9
3540
27.8
504
36.3
532
35.1
Missing
3135
20.0
3842
30.2
126e
8.3
Heavy drinking in
early adulthood
No
9366
74.8
11 644
74.3
6935
78.8
9972
78.3
714
61.4
914
60.3
Yes
3160
25.2
4028
25.7
1861
21.2
2763
21.7
448
38.6
601
39.7
Missing
3146
20.1
3939
30.9
353
23.3
Additional information on missing data
Participated in
adolescence and early
adulthood
No
4442
28.3
6094
47.9
334
22.0
Yes
11 230
71.7
6641
52.1
1181
78.0
Complete data on all
analysis variables
No
9995
63.8
10 252
80.5
581
38.3
Yes
5677
36.2
2483
19.5
934
61.7 Addiction, 111, 1457–1465 016 The Authors. © 2016 The Authors. Addiction published by John Wiley & Sons Ltd on behalf of Society for the Study of Addiction Addiction, 111, 1457–1465 Residual associations between SEP and drinking Despite the smoking pathway, results consistently showed
more frequent drinking among more advantaged adoles-
cents, prior to entry into tertiary education. This finding
suggests that there may be other pathways associated with
socio-economic advantage, besides tertiary education,
which lead to heavier drinking. For example, alcohol may
be more available in more advantaged homes and families
[41,53]. DISCUSSION Addiction published by John Wiley & Sons Ltd on behalf of Society for the Study of Addiction 1462 62
Michael J. Green et al. Fi
2 P
bit
ffii
t (
d t
d d
) f
l i
d l Figure 2 Probit coefficients (and standard errors) from analysis model adulthood. It is presumably not the actual education
but experiences associated with it that account for this. Drinking may be a coping response to transitional chal-
lenges, may be valued for social goals or overestimation
of how much peers drink may inflate perceived behav-
ioural norms [46,47]. However, these factors may also
apply to those transitioning into work and other adult
roles. Perhaps increasing independence and freedom,
combined with low parental monitoring, few adult re-
sponsibilities and close involvement with peers in similar
situations, contribute to students’ higher drinking levels
[48–50]. This association was stronger for females than
males. Given associations between education and egali-
tarian gender-role attitudes [51], tertiary education may
have been associated with attenuation of the general
population trend (when data were collected) for females
in this age group to drink less than males. Heavy drink-
ing rates among UK men and women aged 16–24 con-
verged during the 1990s [52], so the stronger effect of
education for females could be historical. If not, it could
be increasingly important, as female participation in ter-
tiary education has increased in more recent cohorts
[22]. does not demonstrate causality, although there are possible
causal links (described below). Smoking pathway Associations between adolescent smoking and drinking are
consistent with previous research [13,35] and indicate
that smoking may link socio-economic disadvantage and
heavier drinking. This could be, in part, because the phys-
iological effects of nicotine stimulate drinking [36,37],
although reverse causation may contribute to this associa-
tion if alcohol also disinhibits smoking behaviour. Further,
there may be common pathways leading to both tobacco
and alcohol use that are associated with socio-economic
disadvantage [14]. Smoking and drinking behaviours
may both represent coping strategies used by young people
as they face the wide range of stressors associated with a
disadvantaged SEP [38]. Other pathways associated with
socio-economic disadvantage that lead to greater chances
of developing both smoking and drinking behaviours may
include lack of alternative activities [39], lower quality pa-
rental monitoring [38,40], increased exposure via parents
and peers who smoke and drink heavily [3,41–43] or a
greater likelihood of externalizing behaviour [43–45]. If
there are common pathways, it is important to understand
their relative importance. Interventions addressing com-
mon pathways may be especially effective in tackling both
smoking and drinking behaviours among young people
from a disadvantaged SEP. © 2016 The Authors. Addiction published by John Wiley & Sons Ltd on behalf of Society for the Study of Addiction Addiction, 111, 1457–1465 Tertiary education pathway Associations between tertiaryeducation and heavier drink-
ing are also observed commonly [16,17] and may link
socio-economic advantages to heavier drinking in early Addiction, 111, 1457–1465 Pathways linking socio-economic background to alcohol consumption 1463 Declaration of interests None. Limitations Measurement differences may account for some of the dif-
ferences in findings between the cohorts. The self-report
drinking measures were not ideal. They would not have
captured the full complexity of drinking patterns (e.g. epi-
sodic binge drinking might not have been well repre-
sented), and may have led to under-reporting. T07,
where reports were given in the home (but not with par-
ents present) had lower prevalence of adolescent drinking
than NCDS58 and BCS70 where adolescent measures
were administered in schools. Age differences in reporting
may be important as heavy drinking in early adulthood, es-
pecially among students, can be age-limited, with recovery
to moderate levels within a few years [46]. BCS70 mea-
surements were at age 26 rather than 22–23 and had
the lowest prevalence of heavy drinking. Age was con-
founded too strongly with cohort to be included in these
models, but the consistency of the association between ter-
tiary education and heavy drinking measured at these dif-
ferent ages suggests that it was not just age-limited
drinking. Additionally, the broad heading of tertiary educa-
tion may mask heterogeneity in patterns between universi-
ties and further/vocational education colleges. Acknowledgements We are grateful to survey participants and staff in all
three studies. The West of Scotland Twenty-07 Study is
funded by the UK Medical Research Council (MRC;
MC_A540_53462). We are grateful to the Centre for Lon-
gitudinal Studies, Institute of Education and to the UK Data
Archive for making the NCDS58 and BCS70 data available. However, they bear no responsibility for the analysis or in-
terpretation of these data. M.G. was supported by a doc-
toral training fellowship from the Chief Scientist Office
(CSO) of the Scottish Government Health Directorates
(DTF/11/16) and by the MRC (MC_UU_12017-13). A.L. has funding from the MRC (MC_UU_12017/5) and the
CSO
(SPHSU2). H.S. has
funding
from
the
MRC
(MC_UU_12017/3). M.B. has support from the University
of Essex and the UK Economic and Social Research
Council. References 1. Hanson M. D., Chen E. Socioeconomic status and health be-
haviors in adolescence: a review of the literature. J Behav
Med 2007; 30: 263–85. 2. Blum R. W., Beuhring T., Shew M. L., Bearinger L. H., Sieving
R. E., Resnick M. D. The effects of race/ethnicity, income, and
family structure on adolescent risk behaviors. Am J Public
Health 2000; 90: 1879–84. 3. Droomers M., Schrijvers C. T. M., Casswell S., Mackenbach J. P. Occupational level of the father and alcohol consumption dur-
ing
adolescence;
patterns
and
predictors. J
Epidemiol
Community Health 2003; 57: 704–10. 4. Wiles N. J., Lingford-Hughes A., Daniel J., Hickman M., Farrell
M., Macleod J., et al. Socio-economic status in childhood and
later alcohol use: a systematic review. Addiction 2007; 102:
1546–63. Contextual heterogeneity disadvantaged, and have additional health detriments from
smoking). Interventions focused on pathways common to
smoking and drinking may tend to benefit more disadvan-
taged young people. Further research might explore a
wider range of pathways and contexts in order to under-
stand more clearly the pathways that lead to heavy drink-
ing in different socio-economic circumstances, and lead
towards designing interventions tailored to people’s differ-
ent backgrounds. Pathways for further study might in-
clude: availability of alcohol (within the home and society
more broadly); family drinking practices; the role of social
norms; externalizing behaviour; and challenges associated
with transitions into adulthood. Hypotheses regarding contextual variation in the strength
of smoking and tertiary education pathways were only par-
tially verified: differences in the associations between socio-
economic disadvantage and smoking or tertiary education
were only present for parental occupational class, and not
replicated in analyses using income or parental education. Additionally, even for parental occupational class the asso-
ciation between background SEP and tertiary education
was strongest rather than weakest in NCDS58 (although
it was stronger in T07 than BCS70). There was a stronger association between adolescent
smoking and adolescent drinking in the two more recent
cohorts, as hypothesized, but if this was due to increases
in alcohol availability then it is not clear why a similar dif-
ference was not seen for associations with early adult
drinking. Perhaps changes in societal availability matter
more for those below the legal drinking age, who are
generally more constrained in their opportunities for
consumption. © 2016 The Authors. Addiction published by John Wiley & Sons Ltd on behalf of Society for the Study of Addiction Conclusions B., Buka S. L. Socioeconomic status
over the life course and stages of cigarette use: initiation, reg-
ular use, and cessation. J Epidemiol Community Health 2003;
57: 802–8. 30. Banks M. H. Validation of the General Health Questionnaire in
a young community sample. Psychol Med 1983; 13: 349–53. 12. Green M. J., Leyland A. H., Sweeting H., Benzeval M. Socio-
economic
position
and
early
adolescent
smoking
development: evidence from the British Youth Panel Survey
(1994–2008). Tob Control 2016; 25: 203–10. 31. Rutter M. A Children’s Behaviour Questionnaire for comple-
tion by teachers: preliminary findings. J Child Psychol
Psychiatry 1967; 8: 1–11. 32. Muthén L. K., Muthén B. O. Mplus User’s Guide. Los Angeles,
CA: Muthén & Muthén; 1998–2012. 13. Mathers M., Toumbourou J. W., Catalano R. F., Williams J.,
Patton G. C. Consequences of youth tobacco use: a review
of prospective behavioural studies. Addiction
2006; 101:
948–58. 33. Seaman S. R., White I. R., Copas A. J., Li L. Combining multi-
ple imputation and inverse-probability weighting. Biometrics
2012; 68: 129–37. 14. Green M. J., Leyland A. H., Sweeting H., Benzeval M. Socio-
economic position and adolescent trajectories in smoking,
drinking, and psychiatric distress. J Adolesc Health
2013;
53: 202–8. 34. Clarke P., Hardy R. Methods for handling missing data. In:
Pickles A., Maughan B., Wadsworth M., editors. Epidemiolog-
ical Methods in Life Course Research. Oxford: Oxford
University Press; 2007, pp. 157–80. 15. Machin S., Vignoles A. Educational inequality: the widening
socio-economic gap. Fisc Stud 2004; 25: 107–28. 35. Jackson K. M., Sher K. J., Cooper M. L., Wood P. K. Adolescent
alcohol and tobacco use: onset, persistence and trajectories of
use across two samples. Addiction 2002; 97: 517–31. 16. Kypri K., Cronin M., Wright C. S. Do university students drink
more hazardously than their non-student peers? Addiction
2005; 100: 713–4. 36. Barrett S. P., Tichauer M., Leyton M., Pihl R. O. Nicotine in-
creases alcohol self-administration in non-dependent male
smokers. Drug Alcohol Depend 2006; 81: 197–204. 17. Carter A., Brandon K., Goldman M. The college and noncol-
lege experience: a review of the factors that influence
drinking behaviour in young adulthood. J Stud Alcohol Drugs
2010; 71: 742–50. 37. Lê A. D., Wang A., Harding S., Juzytsch W., Shaham Y. Nicotine increases alcohol self-administration and reinstates
alcohol seeking in rats. Psychopharmacology (Berl)
2003;
168: 216–21. 18. Chen C. M., Dufour M. C., Yi H. Conclusions Interventions focused on reducing excess drinking during
tertiary education may be effective at reducing the preva-
lence of heavy drinking in early adulthood but target a
population with more advantaged socio-economic back-
grounds, while ignoring those who drink heavily outside
tertiary
education
(who
will
tend
to
be
more 5. Power C., Matthews S. Origins of health inequalities in a na-
tional population sample. Lancet 1997; 350: 1584–9. 5. Power C., Matthews S. Origins of health inequalities in a na-
tional population sample. Lancet 1997; 350: 1584–9. 6. Pietila A. M., Hentinen M., Myhrman A. The health behav-
iour of northern Finnish men in adolescence and adulthood. Int J Nurs Stud 1995; 32: 325–38. 6. Pietila A. M., Hentinen M., Myhrman A. The health behav-
iour of northern Finnish men in adolescence and adulthood. Int J Nurs Stud 1995; 32: 325–38. Addiction, 111, 1457–1465 Michael J. Green et al. 1464 7. Kandel D. B., Logan J. A. Patterns of drug use from adoles-
cence to young adulthood: I. Periods of risk for initiation,
continued use, and discontinuation. Am J Public Health
1984; 74: 660–6. 25. Benzeval M., Der G., Ellaway A., Hunt K., Sweeting H., West P.,
et al. Cohort profile: West of Scotland Twenty-07 Study—
Health in the Community. Int J Epidemiol 2009; 38: 1215–23. 26. Der G. A comparison of the West of Scotland Twenty-07
Study sample and the 1991 census SARs. MRC Medical Soci-
ology Unit Working Paper. Glasgow: MRC Medical Sociology
Unit; 1998. 8. Viner R. M., Ross D., Hardy R., Kuh D., Power C., Johnson A., et
al. Life course epidemiology: recognising the importance of ad-
olescence. J Epidemiol Community Health 2015; 69: 719–20. 27. Royal College of Physicians (RCP), Royal College of Psychia-
trists (RCPsych), Royal College of General Practitioners
(RGCP) Alcohol and the Heart in Perspective: Sensible Limits
Reaffirmed. London: RCP, RCPsych, RCGP; 1995. 9. Sawyer S. M., AfifiR. A., Bearinger L. H., Blakemore S.-J., Dick
B., Ezeh A. C., et al. Adolescence: a foundation for future
health. Lancet 2012; 379: 1630–40. 10. Tyas S. L., Pederson L. L. Psychosocial factors related to ado-
lescent smoking: a critical review of the literature. Tob
Control 1998; 7: 409–20. 28. Office of Population Censuses and Surveys Classification of
Occupations. London: HMSO; 1980. 29. Goldberg D., Williams P. A User’s Guide to the General Health
Questionnaire. Windsor: NFER-Nelson; 1988. 11. Gilman S. E., Abrams D. Conclusions Alcohol consumption among
young adults ages 18-24 in the United States: results from the
2001–2002 NESARC Survey. Alcohol Res Health 2004; 28:
269–80. 38. Evans G. W. The environment of childhood poverty. Am
Psychol 2004; 59: 77–92. 39. Stock C., Ejstrud B., Vinther-Larsen M., Schlattmann P., Curtis
T., Grønbæk M., et al. Effects of school district factors on alco-
hol consumption: results of a multi-level analysis among
Danish adolescents. Eur J Public Health 2011; 21: 449–55. 19. Côté J., Bynner J. Changes in the transition to adulthood in the
UK and Canada: the role of structure and agency in emerging
adulthood. J Youth Stud 2008; 11: 251–68. 20. Ashton D., Bynner J. Labour market, employment and skills. In:
Wadsworth M.,., Bynner J., editors. ACompanion to Life Course
Studies: The Social and Historical Context of the British Birth
Cohort Studies. Oxford: Routledge; 2011, pp. 122–47. 40. Hayes L., Smart D., Toumbourou J. W., Sanson A. Parenting
Influences on Adolescent Alcohol Use. Melbourne: Australian
Institute of Family Studies; 2004. 41. Green G., Macintyre S., West P., Ecob R. Like parent like child? Associations between drinking and smoking behaviour of par-
ents and their children. Br J Addict 1991; 86: 745–58. 21. Pemberton H. Economic policy and practice. In: Wadsworth
M., Bynner J., editors. A Companion to Life Course Studies:
The Social and Historical Context of the British Birth Cohort
Studies. Oxford: Routledge; 2011, pp. 91–121. 42. West P., Sweeting H., Ecob R. Family and friends’ influences
on the uptake of regular smoking from mid-adolescence to
early adulthood. Addiction 1999; 94: 1397–411. 22. Iannelli C. Inequalities in entry to higher education: a com-
parison over time between Scotland and England and Wales. High Educ Q 2007; 61: 306–33. 43. Kim M. J., Fleming C. B., Catalano R. F. Individual and social
influences on progression to daily smoking during adoles-
cence. Pediatrics 2009; 124: 895–902. 23. Roberts K. Leisure: Stability and change. In: Wadsworth M.,
Bynner J., editors. A Companion to Life Course Studies: The
Social and Historical Context of the British Birth Cohort Stud-
ies. Oxford: Routledge; 2011, pp. 174–202. 44. Redonnet B., Chollet A., Fombonne E., Bowes L., Melchior M. Tobacco, alcohol, cannabis and other illegal drug use among
young adults: the socioeconomic context. Drug Alcohol Depend
2012; 121: 231–9. 24. Plewis I., Calderwood L., Hawkes D., Nathan G. © 2016 The Authors. Addiction published by John Wiley & Sons Ltd on behalf of Society for the Study of Addiction Conclusions National Child
Development Study and 1970 British Cohort Study: Technical
Report: Changes in the NCDS and BCS70 Populations and Sam-
ples over Time. London: Centre for Longitudinal Studies; 2004. 45. Reiss F. Socioeconomic inequalities and mental health prob-
lems in children and adolescents: a systematic review. Soc
Sci Med 2013; 90: 24–31. Addiction, 111, 1457–1465 2016 The Authors. Addiction published by John Wiley & Sons Ltd on behalf of Society for the Study of Addiction Addiction, 111, 1457–1465 Pathways linking socio-economic background to alcohol consumption Pathways linking socio-economic background to alcohol consumption 1465 53. Forsyth A., Barnard M. Preferred drinking locations of Scot-
tish adolescents. Health Place 2000; 6: 105–15. 53. Forsyth A., Barnard M. Preferred drinking locations of Scot-
tish adolescents. Health Place 2000; 6: 105–15. 46. Schulenberg J. E., Maggs J. L. A developmental perspective on
alcohol use and heavy drinking during adolescence and the
transition to young adulthood. J Stud Alcohol 2002 Suppl;
14: 54–70. 47. Helmer S., Mikolajczyk R., Pischke C., Zeeb H., van Hal G.,
Vriesacker B., et al. Personal and perceived peer use of alcohol
among university and college students in seven European
countries: the SNIPE study. Eur J Public Health 2013; 23: N1. © 2016 The Authors. Addiction published by John Wiley & Sons Ltd on behalf of Society for the Study of Addiction Addiction, 111, 1457–1465 Supporting Information Additional Supporting Information may be found in the
online version of this article at the publisher’s web-site: 48. Arnett J. J. Emerging adulthood: a theory of development from
the late teens through the twenties. Am Psychol 2000; 55:
469–80. Table S1 Drinking and Smoking questions by cohort. Table S2 Additional measures included in imputation
models. Table S3 Probit Coefficients, standard errors, 95% confi-
dence intervals and P-Values from models with Parental
Occupation. Table S4 Probit Coefficients, standard errors, 95% confi-
dence intervals and P-Values from models with Parental
Education. Table S5 Probit Coefficients, standard errors, 95% confi-
dence intervals and P-Values from models with Income. 49. Arnett J. J. The developmental context of substance use in
emerging adulthood. J Drug Issues 2005; 35: 235–54. 50. White H. R., McMorris B. J., Catalano R. F., Fleming C. B.,
Haggerty K. P., Abbott R. D. Increases in alcohol and mari-
juana use during the transition out of high school into
emerging adulthood: the effects of leaving home, going to col-
lege, and high school protective factors. J Stud Alcohol 2006;
67: 810–22. Table S3 Probit Coefficients, standard errors, 95% confi-
dence intervals and P-Values from models with Parental
Occupation. Table S4 Probit Coefficients, standard errors, 95% confi-
dence intervals and P-Values from models with Parental
Education. 51. Sweeting H., Bhaskar A., Benzeval M., Popham F., Hunt K. Changing gender roles and attitudes and their implications
for well-being around the new millenium. Soc Psychiatry
Psychiatr Epidemiol 2014; 49: 791–809. Table S5 Probit Coefficients, standard errors, 95% confi-
dence intervals and P-Values from models with Income. 52. Robinson S., Harris H. Smoking and Drinking Among Adults,
2009: A Report on the 2009 General Lifestyle Survey. New-
port: Office for National Statistics; 2011. Addiction, 111, 1457–1465
|
https://openalex.org/W4287921614
|
https://acp.copernicus.org/articles/22/14401/2022/acp-22-14401-2022.pdf
|
English
| null |
Quantifying the drivers of surface ozone anomalies in the urban areas over the Qinghai-Tibet Plateau
| null | 2,022
|
cc-by
| 14,848
|
Quantifying the drivers of surface ozone anomalies
in the urban areas over the Qinghai-Tibet Plateau Hao Yin1,2,⋆, Youwen Sun1,2,⋆, Justus Notholt3, Mathias Palm3, Chunxiang Ye4, and Cheng Liu2,5,6,7
1Key Laboratory of Environmental Optics and Technology, Anhui Institute of Optics and Fine
Mechanics, HFIPS, Chinese Academy of Sciences, Hefei 230031, China
2Department of Precision Machinery and Precision Instrumentation,
University of Science and Technology of China, Hefei 230026, China
3University of Bremen, Institute of Environmental Physics, P.O. Box 330440, 28334 Bremen, Germany
4State Key Laboratory of Environmental Simulation and Pollution Control, College of Environmental Sciences
and Engineering, Peking University, Beijing 100871, China
5Anhui Province Key Laboratory of Polar Environment and Global Change,
University of Science and Technology of China, Hefei 230026, China
6Center for Excellence in Regional Atmospheric Environment, Institute of Urban Environment,
Chinese Academy of Sciences, Xiamen 361021, China
7Key Laboratory of Precision Scientific Instrumentation of Anhui Higher Education Institutes,
University of Science and Technology of China, Hefei 230026, China
⋆ Hao Yin1,2,⋆, Youwen Sun1,2,⋆, Justus Notholt3, Mathias Palm3, Chunxiang Ye4, and Cheng Liu2,5,6,7
1Key Laboratory of Environmental Optics and Technology, Anhui Institute of Optics and Fine
Mechanics, HFIPS, Chinese Academy of Sciences, Hefei 230031, China
2Department of Precision Machinery and Precision Instrumentation,
University of Science and Technology of China, Hefei 230026, China
3University of Bremen, Institute of Environmental Physics, P.O. Box 330440, 28334 Bremen, Germany
4State Key Laboratory of Environmental Simulation and Pollution Control, College of Environmental Sciences
and Engineering, Peking University, Beijing 100871, China
5Anhui Province Key Laboratory of Polar Environment and Global Change,
University of Science and Technology of China, Hefei 230026, China
6Center for Excellence in Regional Atmospheric Environment, Institute of Urban Environment,
Chinese Academy of Sciences, Xiamen 361021, China
7Key Laboratory of Precision Scientific Instrumentation of Anhui Higher Education Institutes,
University of Science and Technology of China, Hefei 230026, China
⋆These authors contributed equally to this work. Correspondence: Cheng Liu (chliu81@ustc.edu.cn) and Youwen Sun (ywsun@aiofm.ac.cn)
Received: 14 June 2022 – Discussion started: 26 July 2022
R
i
d 8 S
b
2022
A
d 7 O
b
2022
P bli h d 9 N
b
2022 University of Science and Technology of China, Hefei 230026, China ⋆These authors contributed equally to this work. Correspondence: Cheng Liu (chliu81@ustc.edu.cn) and Youwen Sun (ywsun@aiofm.ac.c Correspondence: Cheng Liu (chliu81@ustc.edu.cn) and Youwen Sun (ywsun@aiofm.ac.cn) Cheng Liu (chliu81@ustc.edu.cn) and Youwen Sun (ywsun@aiofm.ac.cn) Abstract. Atmos. Chem. Phys., 22, 14401–14419, 2022
https://doi.org/10.5194/acp-22-14401-2022
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License. Atmos. Chem. Phys., 22, 14401–14419, 2022
https://doi.org/10.5194/acp-22-14401-2022
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License. Research article Atmos. Chem. Phys., 22, 14401–14419, 2022
https://doi.org/10.5194/acp-22-14401-2022
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License. Published by Copernicus Publications on behalf of the European Geosciences Union 1
Introduction How-
ever, there are significant uncertainties in the emission in-
ventories and in the models themselves, and shutting down
an emission inventory in CTMs may cause a large nonlin-
ear effect, which inevitably influences the accuracy, perfor-
mance and efficiency of CTMs (Vu et al., 2019; Zhang et al.,
2020). Mathematical and statistical models such as the multi-
ple linear regression (MLR) model and general additive mod-
els (GAMs) have also been used in many studies to quantify
the influence of meteorological factors (Li et al., 2019; K. Li
et al., 2020; Yin et al., 2021a; Yin et al., 2022; Zhai et al.,
2019). Machine learning (ML) is a well-known field that has been
developing rapidly in recent years. ML is a fusion of statis-
tics, data science and computing that experiences use across a
very wide range of applications (Grange et al., 2018). Unlike
most ML models, such as artificial neural networks, whose
working mechanisms are hard to understand, the random
forest (RF) model is not a “black-box” method, its predic-
tion process can be explained, investigated and understood
(Gardner and Dorling, 2001; Grange et al., 2018; Grange and
Carslaw, 2019; Shi et al., 2021). Recently, the RF model-
based meteorological normalization technique has been pro-
posed and used to decouple the influence of meteorology on
atmospheric pollution. For example, Vu et al. (2019) have
used this technique to demonstrate that the clean air ac-
tion plan implemented in 2013 was highly effective in re-
ducing the anthropogenic emissions and improving air qual-
ity in Beijing. Shi et al. (2021) have used this technique to
quantitatively evaluate changes in ambient NO2, ozone, and
PM2.5 concentrations arising from these emission changes in
11 cities globally during the COVID-19 lockdowns. Surface ozone (O3) is a major air pollutant that threatens
human health and vegetation growth (Jerrett et al., 2009; Yin
et al., 2021b). Surface ozone over the QTP is generated either
from its local anthropogenic and natural precursors such as
nitrogen oxides (NOx), volatile organic compounds (VOCs)
and carbon monoxide (CO) via a chain of photochemical re-
actions or transported from long-distance regions by down-
welling from the stratosphere. Surface ozone level is sensi-
tive to local emissions, meteorological conditions and trans-
port. 1
Introduction ogy. Separation of anthropogenic and meteorological drivers
is very important as it conveys to us exactly which processes
drive the observed ozone anomaly and therefore the correct
conclusions can be drawn on whether an emission mitigation
policy is effective. The Qinghai-Tibet Plateau (QTP) (27–45◦N, 70–105◦E),
with an average altitude of 4000 m a.s.l. (above sea level),
is the highest plateau in the world. It is known as the “Roof
of the World” and the “Third Pole” (Qiu, 2008; Yang et al.,
2013; Yin et al., 2017). The QTP has an area of approxi-
mately 2.5 × 106 km2 and accounts for about one quarter of
China’s territory (Duo et al., 2018). The QTP is the source
region of five major rivers in Asia, i.e., the Indus, Ganges,
Brahmaputra, Yangtze and Yellow rivers, which provide wa-
ter resources to more than 1.4 billion people (Immerzeel et
al., 2010). The QTP has been verified to be a critical re-
gion for regulating Asian monsoon climate and hydrologi-
cal cycle, and it is thus an important ecological barrier for
the whole of Asia (Loewen et al., 2007; Yanai et al., 1992). The QTP has long been regarded as a pristine region ow-
ing to its low population and industrial levels (Zhu et al.,
2013). Because of its unique features of landform, ecosystem
and monsoon circulation pattern, the QTP has been regarded
as a region that is sensitive to anthropogenic impact, and is
referred to as an important indicator of regional and global
climate change (Qiu, 2008). The exogenous and local atmo-
spheric pollutants have the potential to accelerate the melting
of glaciers, damage air quality, water sources and grasslands,
and threaten climate on regional and global scales (Yin et al.,
2017; X. F. Yin et al., 2019; Sun et al., 2021d; Pu et al., 2007;
Kang et al., 2016). Therefore, improved knowledge of the
evolutions and drivers of atmospheric pollutants in the QTP
is of great importance for understanding the local ecological
situation and formulating regulatory policies. p
y
Chemical transport models (CTMs) are widely used to
evaluate the influences of meteorology and anthropogenic
emission on atmospheric pollution levels (Hou et al., 2022;
Sun et al., 2021a; Yin et al., 2020; Yin et al., 2019). H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas https://doi.org/10.5194/acp-22-14401-2022 Quantifying the drivers of surface ozone anomalies
in the urban areas over the Qinghai-Tibet Plateau Improved knowledge of the chemistry and drivers of surface ozone over the Qinghai-Tibet
Plateau (QTP) is significant for regulatory and control purposes in this high-altitude region in the Himalayas. In
this study, we investigate the processes and drivers of surface ozone anomalies (defined as deviations of ozone
levels relative to their seasonal means) between 2015 and 2020 in urban areas over the QTP. We separate quanti-
tatively the contributions of anthropogenic emissions and meteorology to surface ozone anomalies by using the
random forest (RF) machine-learning model-based meteorological normalization method. Diurnal and seasonal
surface ozone anomalies over the QTP were mainly driven by meteorological conditions, such as temperature,
planetary boundary layer height, surface incoming shortwave flux, downward transport velocity and inter-annual
anomalies were mainly driven by anthropogenic emission. Depending on region and measurement hour, diur-
nal surface ozone anomalies varied over −27.82 to 37.11 µg m−3, whereas meteorological and anthropogenic
contributions varied over −33.88 to 35.86 µg m−3 and −4.32 to 4.05 µg m−3 respectively. Exceptional meteo-
rology drove 97 % of surface ozone non-attainment events from 2015 to 2020 in the urban areas over the QTP. Monthly averaged surface ozone anomalies from 2015 to 2020 varied with much smaller amplitudes than their
diurnal anomalies, whereas meteorological and anthropogenic contributions varied over 7.63 to 55.61 µg m−3
and 3.67 to 35.28 µg m−3 respectively. The inter-annual trends of surface ozone in Ngari, Lhasa, Naqu, Qamdo,
Diqing, Haixi and Guoluo can be attributed to anthropogenic emissions in 95.77 %, 96.30 %, 97.83 %, 82.30 %,
99.26 % and 87.85 %, and meteorology in 4.23 %, 3.70 %, 2.17 %, 3.19 %, 0.74 % and 12.15 % respectively. The
inter-annual trends of surface ozone in other cities were fully driven by anthropogenic emission, whereas the
increasing inter-annual trends would have larger values if not for the favorable meteorological conditions. This
study can not only improve our knowledge with respect to spatiotemporal variability of surface ozone but also
provide valuable implications for ozone mitigation over the QTP. Published by Copernicus Publications on behalf of the European Geosciences Union. 14402 14402 1
Introduction Meteorological conditions affect surface ozone level in-
directly through changes in natural emissions of its precur-
sors or directly via changes in wet and dry removal, dilution,
chemical reaction rates and transport flux. Emissions of air
pollutants affect surface ozone level by perturbing the abun-
dances of hydroperoxyl (HO2) and alkylperoxyl (RO2) rad-
icals, which are the key atmospheric constituents in the for-
mation of ozone. Some previous studies have presented the
variability and analyzed qualitatively the drivers of surface
ozone over the QTP at a specific site or region (Xu et al.,
2016; X. Yin et al., 2019; Yin et al., 2017; Zhu et al., 2004). However, none of these studies has quantitatively separated
the contributions of anthropogenic emission and meteorol- In this study, we investigate the evolutions, implications
and drivers of surface ozone anomalies (defined as deviations
of ozone levels relative to their seasonal means) from 2015
to 2020 in the urban areas over the QTP. Compared with pre-
vious studies, which focused on surface ozone over the QTP,
this study involves a larger area and a longer time span. Most
importantly, this study separates quantitatively the contribu-
tions of anthropogenic emission and meteorology to surface
ozone anomalies by using the RF model- based meteoro-
logical normalization method. This study can not only im-
prove our knowledge with respect to spatiotemporal variabil-
ity of surface ozone but also provide valuable implication
for ozone mitigation over the QTP. We introduce detailed https://doi.org/10.5194/acp-22-14401-2022 Atmos. Chem. Phys., 22, 14401–14419, 2022 H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas 14403 Figure 1. Geolocations of each city over the Qinghai-Tibet Plateau (QTP). The base map of the figure was created using the Basemap
package in Python. Figure 1. Geolocations of each city over the Qinghai-Tibet Plateau (QTP). The base map of the figure was created using the Basemap
package in Python. Figure 1. Geolocations of each city over the Qinghai-Tibet Plateau (QTP). The base map of the figure was created using the Basemap
package in Python. Figure 1. Geolocations of each city over the Qinghai-Tibet Plateau (QTP). The base map of the figure w
package in Python. nature cycle of ozone in pristine atmosphere. The latter rep-
resents not only the upper limit of ozone photochemistry
contribution to its budget, also demanding knowledge for
the sake of ozone pollution management. As illustrated in
Fig. 1
Introduction 1, the QTP (latitude range: 26◦00′–39◦47′, longitude
range: 73◦19′–104◦47′) covers the Kunlun Mountain, the A-
erh-chin Mountain and the Qilian Mountain in the north, the
Pamir Plateau and the Karakorum Mountains in the west, the
Himalayas in the south, and the Qinling Mountains and the
Loess Plateau in the east. The 12 cities are the most popu-
lated areas over the QTP. All these cities except for Aba and
Diqing are located in the Tibet or Qinghai provinces. Aba
and Diqing are in the Sichuan and Yunnan provinces respec-
tively. The area of these cities ranges from 7.7 to 430 thou-
sand km2, the altitude ranges from 2.3 to 4.8 km a.s.l. and the
population ranges from 0.12 to 2.47 million. The residents
within the 12 cities number about 3.85 million and account
for about 51 % of the population over the QTP. descriptions of the surface ozone and meteorological field
dataset in Sect. 2. The method for separating contributions
of meteorology and anthropogenic emission is presented in
Sect. 3. Section 4 analyzes spatiotemporal variabilities of
surface ozone from 2015 to 2020 in each city over the QTP. The performance of the RF model used for surface ozone
prediction over the QTP is evaluated in Sect. 5. We discuss
the implications and the drivers of surface ozone anomalies
from 2015 to 2020 in each city over the QTP in Sect 6. We
conclude this study in Sect. 7. https://doi.org/10.5194/acp-22-14401-2022 2.2
Meteorological data 2.2
Meteorological data based ozone measurements have been widely used in many
studies for the evaluation of regional atmospheric pollution
and transport over China (Lu et al., 2019a, 2020, 2021; Sun
et al., 2021c, d; Yin et al., 2021a, b, 2022). The CNEMC net-
work has deployed 33 measurement sites in 12 cities over the
QTP (Table 1). The number of measurement sites in each city
varies from 1 to 6. All surface ozone time series at each mea-
surement site are provided by active differential absorption
ultraviolet (UV) analyzers. For all the 33 measurement sites,
hourly surface ozone data since 2015 are available. We first
removed unreliable measurements at all measurement sites
in each city by using the filter criteria following our previous
studies (Lu et al., 2018, 2020; Sun et al., 2021b, d; Yin et
al., 2021a, b), then averaged all measurements in each city
to generate a city representative dataset. All investigations in
this study are performed on such a city representative basis. Meteorological fields used in this study are from the Modern-
Era Retrospective analysis for Research and Applications
Version 2 (MERRA-2) dataset (Gelaro et al., 2017). The
MERRA-2 dataset is produced by the NASA Global Mod-
eling and Assimilation Office and it can provide time series
of many meteorological variables with a spatial resolution of
0.5◦×0.625◦(GMAO, 2022). The boundary layer height and
surface meteorological variables are available per hour and
other meteorological variables are available every 3 hours. It has been verified that the MERRA-2 meteorological fields
over the QTP are in good agreement with the observations
(Wang and Zeng, 2012; Xie et al., 2017). This MERRA-2
dataset has been extensively used in evaluations of regional
atmospheric pollution formation and transport over China
(Carvalho, 2019; Kishore Kumar et al., 2015; Song et al.,
2018; Zhou et al., 2017; Li et al., 2019; K. Li et al., 2020;
Yin et al., 2022; Zhai et al., 2019). The filter criteria can be summarized as follows. Hourly
observed data points were first transformed into Z scores
via Eq. 2.2
Meteorological data (1) and the observed data were then removed if
the corresponding Zi value met one of the following con-
ditions: (1) Zi is larger or smaller than the previous one
(Zi−1) by 9 (|Zi −Zi−1| > 9), (2) the absolute value of Zi
is greater than 4 (|Zi| > 4) or (3) the ratio of the Z value
to the third-order center moving average is greater than
2
3Zi
Z
+Z +Z
> 2
. Table 1. Geolocations of each city over the QTP. Population statistics are available from the 2020 nationwide population census issued by
the National Bureau of Statistics of China. Table 1. Geolocations of each city over the QTP. Population statistics are available from the 2020 nationwide population census issued by
the National Bureau of Statistics of China. city over the QTP. Population statistics are available from the 2020 nationwide population census issued by
s of China. the National Bureau of Statistics of China. Name
Latitude
Longitude
Number
Altitude
Population
Area
of site
(km)
(million)
(thousand
km2)
Ngari
32.5◦N
80.1◦E
2
4.5
0.12
345.0
Shigatse
29.3◦N
88.9◦E
3
4.0
0.80
182.0
Lhasa
29.7◦N
91.1◦E
6
3.7
0.87
31.7
Shannan
29.2◦N
91.8◦E
2
3.7
0.35
79.3
Naqu
31.5◦N
92.1◦E
3
4.5
0.50
430.0
Nyingchi
29.6◦N
94.4◦E
2
3.1
0.23
117.0
Qamdo
31.1◦N
97.2◦E
3
3.4
0.76
110.0
Diqing
27.8◦N
99.7◦E
2
3.5
0.39
23.9
Haixi
37.4◦N
97.4◦E
1
4.8
0.47
325.8
Guoluo
34.5◦N
100.3◦E
1
4.3
0.21
76.4
Xining
36.6◦N
101.7◦E
5
2.3
2.47
7.7
Aba
32.9◦N
101.7◦E
3
3.8
0.82
84.2 2.1
Surface ozone data The QTP covers an area of 2.5 million square meters and
has a population of around 3 million, with most of them
living in several cities. During the in-depth study of the at-
mospheric chemistry over the Tibetan Plateau, @Tibet field
campaign, ozone photochemistry and its roles in ozone bud-
get are of great interest in both background atmosphere and
in QTP urban areas. The former represents the influence of
anthropogenic emission and cross-boundary transport on the Hourly surface ozone data in the urban areas over the QTP
are available from the China National Environmental Moni-
toring Center (CNMEC) network (http://www.cnemc.cn/en/,
last access: 26 November 2021). The CNMEC network- Atmos. Chem. Phys., 22, 14401–14419, 2022 https://doi.org/10.5194/acp-22-14401-2022 14404 H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas Table 2. List of predictive variables fed into the RF model. Parameters
Description
Unit
Meteorological variables by MERRA-2 dataset
Tsurface
Surface air temperature
◦C
U10m
Zonal wind at 10 m height
m s−1
V10m
Meridional wind at 10 m height
m s−1
PBLH
Planetary boundary layer height
m
CLDT
Total cloud area fraction
unitless
PRECTOT
Total precipitation
kg m−2 s−1
OMEGA
Vertical pressure velocity at PBLH
Pa s−1
SWGDN
Surface incoming shortwave flux
W m−2
QV
Specific humidity at 2 m height
kg kg−1
TROPT
Tropospheric layer pressure
Pa
Time information
Year
Year since 2015
–
Month
Month of the year
–
day
Day of the month
–
Hour
Hour of the day
– Table 2. List of predictive variables fed into the RF model. and lower the regression residual. As a result, the relation-
ship between the measured and bootstrap resampled surface
ozone monthly mean time series can be expressed as V (t,b) = b0 + b1t + b2 cos
2πt
12
+ b3 sin
2πt
12
+ b4 cos
4πt
12
+ b5 sin
4πt
12
(2)
F(t,a,b) = V (t,b) + ε(t),
(3) (3) where F(t,a,b) and V (t,b) represent the measured and fitted
surface ozone time series respectively. The parameters b0–
b5 contained in the vector b are coefficients obtained from
the bootstrap resampling regression with V (t,b). The b0 is
the intercept, the b1 is the annual growth rate and b1/b0 is
the inter-annual trend discussed below. The parameters b2–
b5 describe the seasonality, t is the measurement time in the
month elapsed since January 2015, and ε(t) represents the
residual between the measurements and the fitting results. The autocorrelation in the residual can increase the uncer-
tainty in calculation of inter-annual trend. In this study, we
have followed the procedure of Santer et al. (2008) and in-
cluded the uncertainty arising from the autocorrelation in the
residual. where F(t,a,b) and V (t,b) represent the measured and fitted
surface ozone time series respectively. The parameters b0–
b5 contained in the vector b are coefficients obtained from
the bootstrap resampling regression with V (t,b). The b0 is
the intercept, the b1 is the annual growth rate and b1/b0 is
the inter-annual trend discussed below. 3.2
Random forest model We have established a decision tree-based random for-
est (RF) machine-learning model to describe the relation-
ships between hourly surface ozone concentrations (response
variables) and their potential driving factors (predictive vari-
ables) in the urban areas over the QTP. As summarized in
Table 2, predictive variables used in this study include time
variables such as year 2015 to 2020, month 1 to 12, day of
the year from 1 to 365, hour of the day from 0 to 23 and me-
teorological parameters such as wind, temperature, pressure,
cloud fraction, rainfall, vertical transport, radiation and rela-
tive humidity. These time variables were selected as proxies
for emissions as pollutant emissions vary by the time of day,
day of the week and season (Grange et al., 2018). The detailed descriptions of RF machine-learning model
can be found in Breiman (2001). Briefly, the RF model is an
ensemble model consisting of hundreds of individual deci-
sion tree models. Each individual decision tree model uses
a bootstrap aggregating algorithm to randomly sample re-
sponse variables and their predictive variables with a replace-
ment from a training dataset. In this study, a single regres-
sion decision tree is grown in different decision rules based
on the best fitting between surface ozone measurements and
their predictive variables. The predictive variables are se-
lected randomly to give the best split for each tree node. The
predicted surface ozone concentrations are given by the fi-
nal decision as the outcome of the weighted average of all
individual decision trees. By averaging all predictions from H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas The parameters b2–
b5 describe the seasonality, t is the measurement time in the
month elapsed since January 2015, and ε(t) represents the
residual between the measurements and the fitting results. The autocorrelation in the residual can increase the uncer-
tainty in calculation of inter-annual trend. In this study, we
have followed the procedure of Santer et al. (2008) and in-
cluded the uncertainty arising from the autocorrelation in the
residual. bootstrap samples, the bagging process decreases variance
and thus helps the model to minimize overfitting. As shown in Fig. 2, the whole dataset was randomly di-
vided into (1) a training dataset to establish the RF model
and (2) a testing dataset (not included in model training) to
evaluate the model performance. The training dataset was
randomly selected from 70 % of the whole data and the re-
maining 30 % was taken as the testing dataset. The hyperpa-
rameters for the RF model in this study were configured fol-
lowing those in Vu et al. (2019) and Shi et al. (2021) and are
summarized as follows: the maximum tree of a forest is 300
(n_tree = 300), the number of variables for splitting the de-
cision tree is 4 (mtry = 4) and the minimum size of terminal
nodes is 3 (min_node_size = 3). As the meteorological vari-
ables differ in units and magnitudes, which could lead to un-
stable performance of the model, we therefore uniformized
all meteorological variables via Eq. (1) before using them in
the RF model. This pre-processing procedure can also speed
up the establishment of the RF model. 3.1
Quantifying seasonality and inter-annual variability We quantify the seasonality and inter-annual variability of
surface ozone from 2015 to 2020 in each city over the QTP
by using a bootstrap resampling method. The principle of
such a bootstrap resampling method was described in detail
in Gardiner et al. (2008). Many studies have verified the ro-
bustness of Gardiner’s methodology in modeling the season-
ality and inter-annual variabilities of a suite of atmospheric
species (Sun et al., 2020; Sun et al., 2018, 2021a, b, d). In this study, we used a second Fourier series plus a lin-
ear function to fit surface ozone monthly mean time series
from 2015 to 2020 over the QTP. The usage of measure-
ments on monthly basis can improve the fitting correlation 2
3Zi
Zi−1+Zi+Zi+1 > 2
. 2
Zi−1+Zi+Zi+1 > 2
. zk = xk −uk
σk
,
(1) zk = xk −uk
σk
,
(1) (1) where uk and σk are the average and 1σ standard devia-
tion (SD) of xk, and zk is the pre-processed value for pa-
rameter xk. Atmos. Chem. Phys., 22, 14401–14419, 2022 https://doi.org/10.5194/acp-22-14401-2022 14405 3.3
Separation of meteorological and anthropologic
contributions This process ensures that all kinds of weather
conditions around the measurement time have been consid-
ered in the model predictions, which eliminate the influence
of abnormal meteorological conditions and get concentra-
tions under the averaged meteorological conditions. QV, TROPH). Specifically, meteorological variables at a spe-
cific selected hour of a particular day in the input dataset
were generated by randomly selecting from the meteorolog-
ical data during 1980 to 2020 at that particular hour of dif-
ferent dates within a 4-week period (i.e., 2 weeks before and
2 weeks after that selected date). For example, the new in-
put meteorological data at 18:00 LT, 15 February 2018, are
randomly selected from the meteorological data at 18:00 LT
on any date from 1 to 29 February of any year during 1980
to 2020. This selection process was repeated 1000 times to
generate a final input dataset. The 1000 meteorological data
were then fed into the RF model to predict surface ozone con-
centration. The 1000 predicted ozone concentrations were
then averaged as Eq. (4) to calculate the final meteorological
normalized concentration (O3,dew) for that particular hour,
day and year. This process ensures that all kinds of weather
conditions around the measurement time have been consid-
ered in the model predictions, which eliminate the influence
of abnormal meteorological conditions and get concentra-
tions under the averaged meteorological conditions. O3,anomalies = O3,individual −O3,mean,
(5) (5) where O3,mean in each city are approximated by the seasonal-
ity plus the intercept described in Eq. (1). As a result, the dif- where O3,mean in each city are approximated by the seasonal-
ity plus the intercept described in Eq. (1). As a result, the dif-
ference O3,meteo between O3,individual and O3,dew calculated
as Eq. (6) is the portion of anomalies induced by changes
in meteorology. The difference O3,emis between O3,anomalies
and O3,meteo calculated as Eq. (7) represents the portion of
anomalies induced by changes in anthropogenic emission. O3,meteo = O3,individual −O3,dew
(6)
O3,emis = O3,anomalies −O3,meteo. (7) (6)
(7) O3,dew =
1
1000
1000
X
i=1
O3,i,pred ,
(4) ( )
(7) (4) By applying the meteorological normalization method, we
finally separate the contributions of meteorology and an-
thropogenic emissions to the surface ozone anomalies in
each city over the QTP. Positive O3,meteo and O3,emis indi-
cate that changes in meteorology and anthropogenic emis-
sion cause surface ozone concentration above their seasonal
mean values respectively. 3.3
Separation of meteorological and anthropologic
contributions In order to separate the contributions of meteorology and
anthropological emission to surface ozone anomalies in
each city over the QTP, we have decoupled meteorology-
driven anomalies by using the RF model-based meteorologi-
cal normalization method. The meteorological normalization
method was first introduced by Grange et al. (2018) and im-
proved by Vu et al. (2019) and Shi et al. (2021). To decouple
the meteorological influence, we first generated a new input
dataset of predictive variables, which includes original time
variables and resampled meteorological variables (Tsurface,
U10, V10, PBLH, CLDT, PRECTOT, OMEGA, SWGDN, https://doi.org/10.5194/acp-22-14401-2022 Atmos. Chem. Phys., 22, 14401–14419, 2022 14406
H. Yin et al.: Quantifying the drivers of surface ozone anomalies in
Figure 2. Flowchart for separation of meteorology and anthropological contributions. H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas 14406 14406 Figure 2. Flowchart for separation of meteorology and anthropological contributions. i.e., the annual growth rate of surface ozone and the fitting
residual should be close to zero. But this is not realistic
in the real world. Any year-to-year difference in either an-
thropogenic emission or meteorology could result in anoma-
lies. We calculate surface ozone anomalies (O3,anomalies) in
each city over the QTP by subtracting their seasonal mean
values (O3,mean) from all hourly surface ozone measure-
ments (O3,individual) through Eq. (5) (Hakkarainen et al.,
2016, 2019; Mustafa et al., 2021). QV, TROPH). Specifically, meteorological variables at a spe-
cific selected hour of a particular day in the input dataset
were generated by randomly selecting from the meteorolog-
ical data during 1980 to 2020 at that particular hour of dif-
ferent dates within a 4-week period (i.e., 2 weeks before and
2 weeks after that selected date). For example, the new in-
put meteorological data at 18:00 LT, 15 February 2018, are
randomly selected from the meteorological data at 18:00 LT
on any date from 1 to 29 February of any year during 1980
to 2020. This selection process was repeated 1000 times to
generate a final input dataset. The 1000 meteorological data
were then fed into the RF model to predict surface ozone con-
centration. The 1000 predicted ozone concentrations were
then averaged as Eq. (4) to calculate the final meteorological
normalized concentration (O3,dew) for that particular hour,
day and year. 4.1
Overall ozone level Statistical summary and box plot of surface ozone concen-
tration (units: µg m−3) in each city over the QTP from 2015
to 2020 are presented in Table 3 and Fig. S1 in the Sup-
plement respectively. The average of surface ozone be-
tween 2015 and 2020 in each city over the QTP varied
over 50.67 ± 29.57 to 90.38 ± 28.83 µg m−3, and the me-
dian value varied over 53.00 to 90.00 µg m−3. In compari-
son, the average surface ozone between 2015 and 2020 in the
Beijing-Tianjin-Hebei (BTH), the Fenwei Plain (FWP), the
Yangtze River Delta (YRD) and the Pearl River Delta (PRD)
in densely populated and highly industrialized eastern China
was 140.76, 132.16, 125.09 and 119.82 µg m−3 respectively. The average surface ozone between 2011 and 2015 at the
suburb Nam Co station in the southern-central part of the
QTP was 47.00 ± 12.43 µg m−3 (X. Yin et al., 2019). As a
result, surface ozone levels in the urban areas over the QTP
are much lower than those in urban areas in eastern China
but higher than those in the suburban areas over the QTP. Among all cities over the QTP, the highest and lowest sur-
face ozone concentrations occur in Haixi and Aba, with mean
values of 90.38 ± 28.83 and 50.67 ± 28.83 µg m−3 respec-
tively. Generally, surface ozone concentrations in Qinghai
province are higher than those in Tibet province. We also
presented the percentile variation of surface ozone concen-
tration (units: µg m−3) in each city over the QTP from 2015
to 2020 in Fig. S2. The percentile variation modes of surface
ozone concentration in all cities over the QTP are similar. In
this study, only mean plus standard variance of surface ozone
concentration rather than its percentile variation in each city
was investigated. This prevailing method has been used in a
number of studies to describe the variabilities of atmospheric
compositions over the QTP (Li et al., 2020; Liu et al., 2021;
Ma et al., 2020; Xu et al., 2016, 2018; Yin et al., 2017, 2019). 4
Variabilities of surface ozone over the QTP The ambient air quality standard issued by the Chinese
government regularized that the critical value (Class 1 limit)
for the maximum 8 h average ozone level is 160 µg m−3. With this rule, we summarize the number of non-attainment
day per year in each city over the QTP in Table 3. The num-
ber of non-attainment days per city and per year over the QTP
is only 2 between 2015 and 2020. Ozone non-attainment
events over the QTP typically occur in spring or summer. In
comparison, the number of non-attainment days per city and
per year over the BTH, FWP, YRD and PRD is much larger,
with values of 78, 36, 82 and 45 between 2015 and 2020 re-
spectively, and all ozone non-attainment events over these re-
gions occur in summer. The number of non-attainment days
in Ngari in 2020, Lhasa in 2016 and 2017, Shannan in 2017
and 2018, Haixi in 2015 and 2019, and Xining in 2017 are 13,
10, 20, 12, 10, 14, 16 and 17 d respectively. The number of
non-attainment days in all other cities over the QTP are less
than 10 d. In particular, surface ozone concentrations in Aba,
Naqu and Diqing in all the years between 2015 and 2020
are less than the Class 1 limit of 160 µg m−3. There are
only 1 and 2 non-attainment days in Nyingchi and Qamdo
between 2015 and 2020 respectively. 4.1
Overall ozone level 4
Variabilities of surface ozone over the QTP 3.3
Separation of meteorological and anthropologic
contributions Similarly, negative O3,meteo and
O3,emis indicate that changes in meteorology and anthro-
pogenic emissions cause surface ozone concentration below
their seasonal mean values respectively. By applying the meteorological normalization method, we
finally separate the contributions of meteorology and an-
thropogenic emissions to the surface ozone anomalies in
each city over the QTP. Positive O3,meteo and O3,emis indi-
cate that changes in meteorology and anthropogenic emis-
sion cause surface ozone concentration above their seasonal
mean values respectively. Similarly, negative O3,meteo and
O3,emis indicate that changes in meteorology and anthro-
pogenic emissions cause surface ozone concentration below
their seasonal mean values respectively. where O3,i,pred is the surface ozone concentration predicted
by using the ith meteorological data randomly selected from
the meteorological data at the specific selected hour on any
date from 1 to 29 February of any year in 1980 to 2020. O3,dew represents surface ozone concentration under the
mean meteorological conditions at the specific selected hour
between 1980 and 2020. If the seasonal variabilities of anthropogenic emission and
meteorology are constant over the year, the variability of
the surface ozone can be exactly reproduced by Eq. (2), https://doi.org/10.5194/acp-22-14401-2022 Atmos. Chem. Phys., 22, 14401–14419, 2022 14407 H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas
Table 3. Statistical summary of surface ozone concentration (units: µg m−3) in each city over the QTP from 2015 to 2020. al summary of surface ozone concentration (units: µg m−3) in each city over the QTP from 2015 to 2020. Table 3. Statistical summary of surface ozone concentration (units: µg m−3) in each city over the QTP from 2015 to 202
City
Mean
Standard
Median
The number of non-attainment day
deviation
2015
2016
2017
2018
2019
2020
Ngari
74.18
34.26
73.50
0
0
8
9
1
13
Shigatse
79.25
31.62
82.00
0
5
0
5
5
2
Lhasa
77.90
32.63
78.67
10
20
2
5
0
0
Shannan
77.55
30.75
78.00
0
2
12
10
2
3
Naqu
52.43
26.27
53.00
0
0
0
0
0
0
Nyingchi
67.30
28.30
68.00
0
0
1
0
0
0
Qamdo
64.23
31.47
62.00
0
2
0
0
0
0
Diqing
57.50
27.64
54.50
0
0
0
0
0
0
Haixi
90.38
28.83
90.00
14
0
0
0
16
2
Guoluo
82.98
33.29
86.00
3
0
3
3
0
0
Xining
63.50
36.02
60.00
0
2
17
6
3
3
Aba
50.67
29.57
47.00
0
0
0
0
0
0 4
Variabilities of surface ozone over the QTP 4.3
Seasonal variability The timings of the diurnal cycles in all cities over the
QTP were shifted by 1 to 2 h later in winter than those dur-
ing the rest of the year, most likely due to the later time
of sunrise. Yin et al. (2017) also observed such a shift in
the diurnal cycle at the suburb Nam Co station. The diur-
nal cycles of surface ozone in the urban areas over the QTP
spanned a large range of −43.73 % to 47.12 % depending
on region, season and measurement time. The minimum and
maximum surface ozone levels in the urban areas over the
QTP varied over (22.89 ± 15.55) to (68.96 ± 18.27) µg m−3
and (57.77±21.56) to (102.08±15.14) µg m−3 respectively. On average, surface ozone levels in the urban areas over the
QTP have mean values of (72.41±33.83) µg m−3 during the
daytime (08:00–19:00 LT) and (60.89 ± 32.25) µg m−3 dur-
ing the evening (20:00–08:00 LT). The diurnal cycles of sur-
face ozone in all cities over the QTP are generally consistent
with the results reported in eastern China and the suburban
areas over the QTP (X. Yin et al., 2019; Yin et al., 2017;
Zhao et al., 2016; Shen et al., 2014). Monthly averaged time series of surface ozone in each city
over the QTP between 2015 and 2020 are shown in Fig. 4. Surface ozone levels in all cities over the QTP showed pro-
nounced seasonal features. Seasonal cycles of surface ozone
in most cities present a unimodal pattern with a seasonal
peak occurring around March-July and a seasonal trough oc-
curring around October–December. Specifically, maximum
surface ozone levels occur in spring over Diqing, Lhasa,
Naqu, Nyingchi, Qamdo, Shannan, Shigatse and Aba, and
occur in summer over Ngari, Xining, Guoluo and Haixi;
minimum surface ozone levels in Nyingchi and Diqing oc-
cur in autumn, and in other cities they occur in winter. The
minimum and maximum surface ozone levels between 2015
and 2020 over the QTP varied over (29.21 ± 19.03) to
(60.45 ± 31.35) µg m−3 and (71.25 ± 26.53) to (112.46 ±
28.92) µg m−3 respectively (Table S1 in the Supplement). The peak-to-trough contrast in Diqing, Naqu, Nyingchi and
Aba were smaller than those in other cities. 4.2
Diurnal variability Diurnal cycles of surface ozone in each season and each city
over the QTP are presented in Fig. 3. Overall, diurnal cycle of
surface ozone in each city over the QTP presents a unimodal
pattern in all seasons. For all cities in all seasons, high lev-
els of surface ozone occur in the daytime (09:00 to 20:00 LT)
and low levels of surface ozone occur at nighttime (21:00 to
08:00 LT). As seen from Fig. 3, surface ozone levels usually
increase over time starting at 08:00 to 11:00 LT in the morn-
ing, reach the maximum values at 15:00 to 18:00 LT in the
afternoon, and then decrease over time till the minimum val-
ues at 08:00 or 09:00 LT the next day. Atmos. Chem. Phys., 22, 14401–14419, 2022 https://doi.org/10.5194/acp-22-14401-2022 14408
H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas
Figure 3. Diurnal cycle of surface ozone (units: µg m−3) in each season and each city over the QTP. The vertical error bar is 1σ standard
variation (SD) within that hour. H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas 14408 Figure 3. Diurnal cycle of surface ozone (units: µg m−3) in each season and each city over the QTP. The vertical error bar is 1σ standard
variation (SD) within that hour. H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas 14409 Figure 4. Monthly mean time series of surface ozone (units: µg m−3) between 2015 and 2020 in each city over the QTP. The vertical error
bar is 1σ standard variation (SD) within that month. Figure 4. Monthly mean time series of surface ozone (units: µg m−3) between 2015 and 2020 in each city over the QTP. The vertical error
bar is 1σ standard variation (SD) within that month. 4.4
Inter-annual variability The inter-annual variability of surface ozone between 2015
and 2020 in each city over the QTP fitted by the bootstrap
resampling method is presented in Figs. 5 and S3, and is also
summarized in Table S1. Generally, the measured and fitted
surface ozone concentrations in each city over the QTP are
in good agreement with a correlation coefficient (R) of 0.68–
0.92 (Fig. S4). The measured features in terms of season-
ality and inter-annual variability can be reproduced by the
bootstrap resampling model. However, owing to the year-
to-year deference in anthropogenic emission and meteorol-
ogy, both inter-annual variability and fitting residual were
not zero in all cities. The inter-annual trends in surface ozone
level from 2015 to 2020 over the QTP spanned a large range
of (−2.43 ± 0.56) to (7.55 ± 1.61) µg m−3 yr−1, indicating
a regional representation of each dataset. The inter-annual
trends of surface ozone levels in most cities including Diqing,
Naqu, Ngari, Nyingchi, Shannan, Shigatse, Xining, Abzhou
and Haixi showed positive trends. The largest increasing
trends were presented in Diqing and Nagri, with values of
(5.31 ± 1.28) and (7.55 ± 1.61) µg m−3 yr−1 respectively. In
contrast, surface ozone levels in Lhasa, Qamdo and Guoluo
presented negative trends, with values of (−1.62 ± 0.76),
(−2.43±0.56) and (−2.36±0.81) µg m−3 yr−1 respectively. We evaluate the performance of the RF model in predict-
ing hourly surface ozone level in each city over the QTP
using the metrics of Pearson correlation coefficient (R), the
root means square error (RMSE), and the mean absolute er-
ror (MAE). They are commonly used metrics for evaluation
of machine-learning model predictions, and are defined as
Eqs. (8)–(10) respectively. R =
n
nP
i=0
xiyi −
nP
i=0
xi ·
nP
i=0
yi
s
n
nP
i=0
x2
i −
nP
i=0
xi
2
·
s
n
nP
i=0
y2
i −
nP
i=0
yi
2
(8)
RMSE =
v
u
u
u
t
nP
i=1
(xi −yi)2
n
(9)
MAE =
nP
i=1
|xi −yi|
n
,
(10) (8) (9) MAE =
P
i=1
y
n
,
(10) (10) where xi and yi are the ith concurrent measured and pre-
dicted data pairs respectively. The n is the number of mea-
surements. The R value represents the fitting correlation be- where xi and yi are the ith concurrent measured and pre-
dicted data pairs respectively. 4.3
Seasonal variability Owing to re-
gional deference in meteorology and anthropogenic emis-
sion, the seasonal cycle of surface ozone in the urban areas
over the QTP is also regional dependent. Atmos. Chem. Phys., 22, 14401–14419, 2022 https://doi.org/10.5194/acp-22-14401-2022 Atmos. Chem. Phys., 22, 14401–14419, 2022 H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas 14410 Figure 5. Inter-annual trends of surface ozone levels between 2015 and 2020 in the urban areas over the QTP. Blue dots are the monthly
averaged surface ozone measurements. The seasonality and inter-annual variability in each city fitted by using a bootstrap resampling model
with a second Fourier series (red dots) plus a linear function (black line) are also shown. Figure 5. Inter-annual trends of surface ozone levels between 2015 and 2020 in the urban areas over the QTP. Blue dots are the monthly
averaged surface ozone measurements. The seasonality and inter-annual variability in each city fitted by using a bootstrap resampling model
with a second Fourier series (red dots) plus a linear function (black line) are also shown. tween the measurements and predictions. The RMSE value
measures the relative average difference between the mea-
surements and predictions. The MAE value measures the ab-
solute average difference between the measurements and pre-
dictions. The units of RMSE and MAE are same as the mea-
sured data, namely µg m−3. where temperature term (T2 m) is the most important variable. For all cities, the aggregate importance of time information
is greater than 50 %. In all cities over the QTP, the meteoro-
logical variables such as temperature (T2 m), relative humid-
ity (QV), vertical pressure velocity (OMEGA) and planetary
boundary layer height (PBLH) play significant roles when
explaining surface ozone concentrations. For other variables,
although they are not decisive variables in the RF model pre-
dictions, they are not negligible in predicting surface ozone
in all cities over the QTP. Although time information is the
most important variable in the RF model predictions, it can
be used very precisely, and thus the RF model measurement
discrepancy in all cities could be from other predictive vari-
ables rather than from time information. Comparisons between the model predictions and measure-
ments for the testing data (not included in model training)
in each city over the QTP are shown in Fig. S5. Overall, the
RF model predictions and surface ozone measurements are in
good agreement, showing high R and low RMSE and MAE
for testing dataset in each city over the QTP (Fig. S5). De-
pending on cities, the R values varied over 0.85 to 0.94, the
RMSE over 10.24 to 17.55 µg m−3 and MAE over 7.32 to
12.76 µg m−3. H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas The R, RMSE and MAE are independent of
city and surface ozone level. The results affirm that our model
performs very well in predicting surface ozone levels and
variabilities in each city over the QTP. 4.4
Inter-annual variability The n is the number of mea-
surements. The R value represents the fitting correlation be- https://doi.org/10.5194/acp-22-14401-2022 Atmos. Chem. Phys., 22, 14401–14419, 2022 Atmos. Chem. Phys., 22, 14401–14419, 2022 H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas 14411 Figure 6. Diurnal cycles of surface ozone anomalies (O3,anomalies, blue dots and lines) along with the meteorology-driven portions (O3,meteo,
red dots and lines) and the anthropogenic-driven portions (O3,emis, black dots and lines) in each city over the QTP. Bold curves and the
shadows are diurnal cycles and the 1σ standard variations respectively. Figure 6. Diurnal cycles of surface ozone anomalies (O3,anomalies, blue dots and lines) along with the meteorology-driven portions (O3,meteo,
red dots and lines) and the anthropogenic-driven portions (O3,emis, black dots and lines) in each city over the QTP. Bold curves and the
shadows are diurnal cycles and the 1σ standard variations respectively. scales are summarized. We then present an in-depth analysis
of each driver in Sect. 6.4. We further investigated the drivers of surface ozone non-
attainment events from 2015 to 2020 in each city over the
QTP. All ozone non-attainment events were classified as
meteorology-dominated or anthropogenic-dominated events
according to which one has a larger contribution to the ob-
served surface ozone non-attainment events. The statistical
results are listed in Table S2. Except for 1 d in Ngari in 2018,
1 d in Shigatse in 2016 and 1 d in Haixi in 2019, which
were dominated by anthropogenic emission, all other sur-
face ozone non-attainment events from 2015 to 2020 over
the QTP were dominated by meteorology. Exceptional mete-
orology drove 97 % of surface ozone non-attainment events
from 2015 to 2020 in the urban areas over the QTP. For
the meteorology-dominated surface ozone non-attainment
events, meteorological and anthropogenic contributions var-
ied over 32.85 to 55.61 µg m−3 and 3.67 to 7.23 µg m−3 re-
spectively. For the anthropogenic-dominated surface ozone
non-attainment events, meteorological and anthropogenic
contributions varied over 7.63 to 10.53 µg m−3 and 15.63 to
35.28 µg m−3 respectively. https://doi.org/10.5194/acp-22-14401-2022 6
Drivers of surface ozone anomalies We further investigate the importance of each input vari-
able in the RF model for predicting surface ozone level in
each city over the QTP. As shown in Fig. S6, time infor-
mation such as hour term (Hour), year term (Year) or sea-
sonal term (Month) are the most important variables in the
RF model predictions in all cities except Xining and Haixi, In this section, we explore the drivers of surface ozone
anomalies between 2015 and 2020 over the QTP. We first
present descriptively the contributions of anthropogenic
emission and meteorology to surface ozone anomalies over
the QTP in Sects. 6.1 to 6.3, where statistics on different time Atmos. Chem. Phys., 22, 14401–14419, 2022 https://doi.org/10.5194/acp-22-14401-2022 H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas 14412 Figure 7. Seasonal cycles of surface ozone anomalies (O3,anomalies, blue dots and lines) along with the meteorology-driven por-
tions (O3,meteo, red dots and lines) and the anthropogenic-driven portions (O3,emis, black dots and lines) in each city over the QTP. Bold
curves and the shadows are monthly mean values and the 1σ standard variations respectively. Figure 7. Seasonal cycles of surface ozone anomalies (O3,anomalies, blue dots and lines) along with the meteorology-driven por-
tions (O3,meteo, red dots and lines) and the anthropogenic-driven portions (O3,emis, black dots and lines) in each city over the QTP. Bold
curves and the shadows are monthly mean values and the 1σ standard variations respectively. 6.2
Seasonal scale ozone anomalies in July and September in Ngari respectively. In all cities, seasonal cycles of meteorological contributions
are more consistent with those of surface ozone anomalies
over the QTP. In some cases, surface ozone anomalies would
have larger values if not for the unfavorable meteorological
conditions, e.g., surface ozone anomalies in June in Ngari
and in December in Shannan, Guoluo and Aba. Figure 7 presents seasonal cycles of surface ozone anoma-
lies between 2015 and 2020 along with the meteorology-
driven and anthropogenic-driven portions in each city over
the QTP. In all cities, the monthly averaged surface ozone
anomalies between 2015 and 2020 varied with much smaller
amplitudes than their diurnal anomalies. Noticeable anoma-
lies include pronounced positive anomalies in December in
Nagri, in May in Lhasa, Shannan and Qamdo, in July in
Haixi, in June in Guoluo, and negative anomalies in July in
Lhasa, Nyingchi and Guoluo. Both meteorological and an-
thropogenic contributions are regional dependent and show
large variations throughout the year. Depending on region
and month, meteorological and anthropogenic contributions
varied over −4.54 to 3.31 µg m−3 and −2.67 to 3.35 µg m−3
between 2015 and 2020 respectively. 6.1
Diurnal scale Figure 6 presents diurnal cycles of surface ozone anoma-
lies between 2015 and 2020, along with the meteorology-
driven and anthropogenic-driven portions in each city over
the QTP. In all cities, the anthropogenic contributions are
almost constant, but the meteorological contributions show
large variations throughout the day. Depending on region
and measurement hour, diurnal surface ozone anomalies on
average varied over −27.82 to 37.11 µg m−3 between 2015
and 2020, whereas meteorological and anthropogenic con-
tributions varied over −33.88 to 35.86 µg m−3 and −4.32 to
4.05 µg m−3 respectively. The least contrast between meteo-
rological contribution and anthropogenic contribution occurs
in Haixi. The diurnal cycles of meteorological contribution
are consistent with those of surface ozone anomalies. High
levels of meteorological contributions occur during the day-
time (09:00 to 20:00 LT) and low levels of meteorological
contributions occur in at nighttime. As a result, diurnal sur-
face ozone anomalies in each city over the QTP were mainly
driven by meteorology. Atmos. Chem. Phys., 22, 14401–14419, 2022 https://doi.org/10.5194/acp-22-14401-2022 H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas 14413 Figure 8. Annual mean surface ozone anomalies (O3,anomalies, blue dots and lines) along with meteorology-driven portions (O3,meteo, red
dots and lines) and anthropogenic-driven portions (O3,emis, black dots and lines) in each city over the QTP. Bold curves and the shadows are
annual mean values and the 1σ standard variations respectively. Figure 8. Annual mean surface ozone anomalies (O3,anomalies, blue dots and lines) along with meteorology-driven portions (O3,meteo, red
dots and lines) and anthropogenic-driven portions (O3,emis, black dots and lines) in each city over the QTP. Bold curves and the shadows are
annual mean values and the 1σ standard variations respectively. respectively in Diqing, Naqu, Nagri and Haixi, increased
by 11.13 µg m−3 in Shannan in 2016 relative to 2015 and
increased by 20.85 µg m−3 in Ngari in 2017 relative to 2016. As a result, all above reductions or increments in surface
ozone level were mainly driven by anthropogenic emission. In contrast, surface ozone anomalies in Lhasa in 2017
and 2020, and in Shigatse and Nyingchi in 2019, were
mainly driven by meteorology. anthropogenic emission, where the increasing inter-annual
trends would have larger values if not for the favorable me-
teorological conditions. As a result, the inter-annual trends
of surface ozone anomalies in all cities over the QTP were
dominated by anthropogenic emission. 6.3
Multi-year scale Annual mean surface ozone anomalies between 2015
and
2020,
along
with
meteorology-driven
and
anthropogenic-driven
portions
in
each
city
over
the
QTP are presented in Fig. 8. Surface ozone in Diqing,
Naqu, Nagri, Haixi and Shannan shows larger year-to-year
variations than that in other cities. Annual mean surface
ozone levels in Diqing, Naqu, Nagri and Haixi showed
significant reductions of 2.10, 10.32, 6.87 and 15.97 µg m−3
respectively, Shannan showed an increment of 9.12 µg m−3
and other cities showed comparable values in 2016 relative
to 2015. The largest year-to-year difference occurred in
Ngari during 2016 to 2017, which has an increment of
25.25 µg m−3. The results show that anthropogenic con-
tributions decreased by 1.85, 7.14, 5.65 and 15.98 µg m−3 Seasonal surface ozone anomalies between 2015 and 2020
in all cities over the QTP were mainly driven by me-
teorology. For example, meteorology caused decrements
of 3.05 µg m−3 in July and 4.27 µg m−3 in September in
Diqing, whereas anthropogenic emission caused increments
of 0.64 and 1.34 µg m−3 in respective months. Aggregately,
we observed −2.41 and −2.89 µg m−3 of seasonal surface Atmos. Chem. Phys., 22, 14401–14419, 2022 https://doi.org/10.5194/acp-22-14401-2022 6.4
Discussions These wet
meteorological conditions inhibit biogenic emissions, slow
down ozone chemical production, and facilitate the ventila-
tion of ozone and its precursors (Gong and Liao, 2019; Jiang
et al., 2021; Lu et al., 2019a, b; Ma et al., 2019), and there-
fore contribute to ozone decrease. anomalies in each city during the ozone non-attainment
days. As tabulated in Table S8, temperature is the dominant
meteorological variable responsible for the meteorology-
dominated ozone non-attainment events, especially in Shi-
gatse, Lhasa, Shannan, Haixi and Guoluo. In addition, the
OMEGA is an important meteorological variable in most
cities, especially in Guoluo, where the correlation is up
to 0.69. For other meteorological variables, winds (U10 m,
V10 m) and TROPH also have noticeable contributions to
some ozone non-attainment events. The NOx and VOCs are the main precursors of sur-
face ozone. The monthly and annual averaged anthropogenic
emissions of NOx and VOCs in each city over the QTP
extracted from the Multi-resolution Emission Inventory for
China (MEIC) between 2015 and 2017 are presented in Ta-
bles S9–S12. Major anthropogenic emissions in each city
over the QTP are from the transport sector and residential
sector, including burning emissions of coal, post-harvest crop
residue, yak dung and religious incense (Chen et al., 2009;
Kang et al., 2016, 2019; Li et al., 2017). The NOx and VOCs
emissions have decreased in Diqing, Naqu, Nagri in 2016
relative to 2015. These reductions of NOx and VOCs emis-
sions have jointly driven the changes of ozone in these cities. Although NOx emissions increased in Haixi during 2015
to 2016, VOCs emissions have significantly decreased by
6.82 t. As a result, the decreases in ozone in Haixi in 2016
relative to 2015 were attributed to VOCs reductions during
the same period. The U10 m and V10 m represent the metrics for evaluat-
ing the horizontal transport. In most of the cities over the
QTP, noticeable ozone versus horizontal wind correlations
are observed, indicating that horizontal transport is an im-
portant contributor to surface ozone (Shen et al., 2014; Zhu
et al., 2004). The QTP region, as a whole, is primarily regu-
lated by the interplay of the Indian summer monsoon and the
westerlies, and the atmospheric environment over the QTP
is heterogeneous. 6.4
Discussions Typically, all cities over the QTP are formed in flat val-
leys, with surrounding mountains rising to more than 5.0 km
a.s.l., and maintain continuous expansion and development
over time. Inhibited by surrounding mountains, regional-
dependent emissions and mountain peak-valley meteorologi-
cal systems result in regional representation of surface ozone
level and their drivers on diurnal, seasonal and inter-annual
scales. Table S3 summarizes the inter-annual trends of surface
ozone anomalies, meteorological and anthropogenic contri-
butions from 2015 to 2020 in each city over the QTP. Ex-
cept for Guoluo, Qamdo and Lhasa, which show decreas-
ing trends, anthropogenic contributions in all other cities
showed increasing trends from 2015 to 2020. With respect
to meteorology contribution, Ngari, Naqu, Diqing and Haixi
showed increasing trends from 2015 to 2020 and all other
cities showed decreasing trends. The inter-annual trends of
surface ozone anomalies in Ngari, Lhasa, Naqu, Qamdo,
Diqing, Haixi and Guoluo can be attributed to anthropogenic
emissions in 95.77 %, 96.30 %, 97.83 %, 82.30 %, 99.26 %
and 87.85 %, and to meteorology in 4.23 %, 3.70 %, 2.17 %,
3.19 %, 0.74 % and 12.15 % respectively. The inter-annual
trends of surface ozone in other cities were fully driven by Correlations between O3,meteo and each meteorological
anomaly are summarized for all time, for the diurnal scale,
for the seasonal scale and for the multi-year scale in Ta-
bles S4–S7. We find that all-time scales of meteorology-
driven surface ozone anomalies in each city are positively re-
lated to anomalies of temperature, planetary boundary layer
height (PBLH), surface incoming shortwave flux (SWGDN),
downward transport velocity at the PBLH (OMEGA), and https://doi.org/10.5194/acp-22-14401-2022 Atmos. Chem. Phys., 22, 14401–14419, 2022 H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas 14414 tropopause height (TROPH). Among all these positive cor-
relations, the correlations with temperature, PBLH and
SWGDN in all cities are higher than those with OMEGA
and TROPH. As high temperature and SWGDN facilitate the
formation of ozone via the increase in chemical reaction rates
or biogenic emissions, the meteorology-driven surface ozone
anomalies have the highest correlations with the changes in
temperature and SWGDN. 6.4
Discussions Mount Everest is representative of the Hi-
malayas on the southern edge of the Tibetan Plateau and is
close to South Asia, where anthropogenic atmospheric pollu-
tion has been increasingly recognized as disturbing the high
mountain regions (Decesari et al., 2010; Maione et al., 2011;
Putero et al., 2014). The northern QTP, including Xining,
Haixi and Guoluo, is occasionally influenced by regional pol-
luted air masses (Xue et al., 2011; Zhu et al., 2004), in par-
ticular, the impacts of anthropogenic emissions from central
and eastern China in the summer (Xue et al., 2011). The cities
over the inland QTP are distant from both South Asia and
northwestern China; this area has been found to be influenced
by episodic long-range transport of air pollution from South
Asia (Lüthi et al., 2015), evidenced by the study of aerosol
and precipitation chemistry in these cities (Cong et al., 2010). I
d
t
d t
i
hi h
ifi
t
l
i
l
i The correlations of the monthly and annual averaged an-
thropogenic contributions against the NOx and VOCs emis-
sions are summarized in Table S13. The correlations of
the monthly averaged anthropogenic contributions against
anthropogenic NOx and VOCs emissions are within the
range 0.35–0.81 and 0.33–0.83 respectively. For the annual
averaged statistics, the correlations against NOx and VOCs
emissions are within the range 0.15–0.94 (except for Ny-
ingchi and Diqing), and 0.34–0.98 (except for Haixi) respec-
tively. For all cities except for Shannan, Qamdo and Haixi,
both the NOx and VOCs emissions are consistent with the
anthropogenic contributions. although only NOx emissions
in Qamdo and Haixi and VOCs emissions in Shannan are
consistent with anthropogenic contributions. In general, the
changes in NOx and VOCs emissions in the MEIC are able to
explain the variabilities of both monthly and annual averaged
anthropogenic contributions. 6.4
Discussions Possible reasons for the ozone in-
creases with the increase in PBLH include lower NO concen-
tration at the urban surface owing to the deep vertical mixing,
which then limits ozone destruction and increases ozone con-
centrations (He et al., 2017), and more downward transport
of ozone from the free troposphere, where the ozone concen-
tration is higher than the near-surface concentration (Sun et
al., 2009). Large OMEGA and high tropopause height also
facilitate downward transport of stratospheric ozone, result-
ing in a high surface ozone level. The QTP has been iden-
tified as a hot spot for stratospheric–tropospheric exchange
(Cristofanelli et al., 2010; Škerlak et al., 2014), where the
surface ozone is elevated from the baseline during the spring
owing to frequent stratospheric intrusions. Generally, surface
ozone anomalies are negatively related to humidity, rainfall
and total cloud fraction in each city over the QTP. These wet
meteorological conditions inhibit biogenic emissions, slow
down ozone chemical production, and facilitate the ventila-
tion of ozone and its precursors (Gong and Liao, 2019; Jiang
et al., 2021; Lu et al., 2019a, b; Ma et al., 2019), and there-
fore contribute to ozone decrease. tropopause height (TROPH). Among all these positive cor-
relations, the correlations with temperature, PBLH and
SWGDN in all cities are higher than those with OMEGA
and TROPH. As high temperature and SWGDN facilitate the
formation of ozone via the increase in chemical reaction rates
or biogenic emissions, the meteorology-driven surface ozone
anomalies have the highest correlations with the changes in
temperature and SWGDN. Possible reasons for the ozone in-
creases with the increase in PBLH include lower NO concen-
tration at the urban surface owing to the deep vertical mixing,
which then limits ozone destruction and increases ozone con-
centrations (He et al., 2017), and more downward transport
of ozone from the free troposphere, where the ozone concen-
tration is higher than the near-surface concentration (Sun et
al., 2009). Large OMEGA and high tropopause height also
facilitate downward transport of stratospheric ozone, result-
ing in a high surface ozone level. The QTP has been iden-
tified as a hot spot for stratospheric–tropospheric exchange
(Cristofanelli et al., 2010; Škerlak et al., 2014), where the
surface ozone is elevated from the baseline during the spring
owing to frequent stratospheric intrusions. Generally, surface
ozone anomalies are negatively related to humidity, rainfall
and total cloud fraction in each city over the QTP. 7
Conclusions In this study, we have investigated the evolutions, implica-
tions and the drivers of surface ozone anomalies (defined as
deviations of ozone levels relative to their seasonal means)
between 2015 and 2020 in the urban areas over the QTP. Diurnal, seasonal and inter annual variabilities of surface
ozone in 12 cities over the QTP are analyzed. The average In order to determine which specific meteorological vari-
ables are responsible for the meteorology-dominated ozone
non-attainment events over the QTP, we have investigated the
correlations between each meteorological variable and ozone Atmos. Chem. Phys., 22, 14401–14419, 2022 https://doi.org/10.5194/acp-22-14401-2022 H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas 14415 surface ozone between 2015 and 2020 in each city over the
QTP varied over (50.67 ± 29.57) to (90.38 ± 28.83) µg m−3,
and the median value varied over 53.00 to 90.00 µg m−3. Overall, the diurnal cycle of surface ozone in each city over
the QTP presents a unimodal pattern in all seasons. For all
cities in all seasons, high levels of surface ozone occur in
the daytime (09:00 to 20:00 LT) and low levels of surface
ozone occur at nighttime (21:00 to 08:00 LT). Seasonal cy-
cles of surface ozone in most cities present a unimodal pat-
tern with a seasonal peak occurring around March–July and
a seasonal trough occurring around October–December. The
inter-annual trends in surface ozone level from 2015 to 2020
over the QTP spanned a large range of (−2.43 ± 0.56) to
(7.55 ± 1.61) µg m−3 yr−1, indicating a regional representa-
tion of each dataset. tions. This study can not only improve our knowledge with
respect to spatiotemporal variability of surface ozone but
also provide valuable implications for ozone mitigation over
the QTP. Code and data availability. All
other
data
are
available
on request of the corresponding author (Youwen Sun, yw-
sun@aiofm.ac.cn). Supplement. The supplement related to this article is available
online at: https://doi.org/10.5194/acp-22-14401-2022-supplement. Author contributions. HY designed the study and wrote the pa-
per. YS supervised and revised this paper. JN, MP, CY and CL pro-
vided constructive comments. We have established a RF regression model to describe
the relationships between hourly surface ozone concentra-
tions (response variables) and their potential driving factors
(predictive variables) in the urban areas over the QTP. 7
Conclusions The
RF model predictions and surface ozone measurements are
in good agreement, showing high R and low RMSE and
MAE in each city over the QTP. Depending on the city, the
R values varied over 0.85 to 0.94, the RMSE over 10.24 to
17.55 µg m−3 and the MAE over 7.32 to 12.76 µg m−3. The R, RMSE and MAE are independent of city and sur-
face ozone levels. The results affirm that our model performs
very well in predicting surface ozone levels and variabilities
in each city over the QTP. Competing interests. The contact author has declared that none
of the authors has any competing interests. Disclaimer. Publisher’s note: Copernicus Publications remains
neutral with regard to jurisdictional claims in published maps and
institutional affiliations. y
Q
We have separated quantitatively the contributions of an-
thropogenic emission and meteorology to surface ozone
anomalies by using the RF model-based meteorological
normalization method. Diurnal and seasonal surface ozone
anomalies over the QTP were mainly driven by meteorol-
ogy, and inter-annual anomalies were mainly driven by an-
thropogenic emission. Depending on the region and the mea-
surement hour, diurnal surface ozone anomalies varied over
−30.55 to 34.01 µg m−3 between 2015 and 2020, whereas
meteorological and anthropogenic contributions varied over
−20.08 to 48.73 µg m−3 and −27.18 to 1.92 µg m−3 re-
spectively. Unfavorable meteorology drove 97 % of surface
ozone non-attainment events between 2015 and 2020 in
the urban areas over the QTP. Monthly averaged surface
ozone anomalies varied with much smaller amplitudes than
their diurnal anomalies, whereas meteorological and anthro-
pogenic contributions varied over 7.63 to 55.61 µg m−3 and
3.67 to 35.28 µg m−3 between 2015 and 2020 respectively. The inter-annual trends of surface ozone anomalies in Ngari,
Lhasa, Naqu, Qamdo, Diqing, Haixi and Guoluo can be at-
tributed to anthropogenic emissions in 95.77 %, 96.30 %,
97.83 %, 82.30 %, 99.26 % and 87.85 %, and to meteorology
in 4.23 %, 3.70 %, 2.17 %, 3.19 %, 0.74 % and 12.15 % re-
spectively. The inter-annual trends of surface ozone anoma-
lies in other cities were fully driven by anthropogenic emis-
sion, where the increasing inter-annual trends would have
larger values if not for the favorable meteorological condi- Special issue statement. References L., Song, C. B., Liu, B. W., Guo, F. Z., Dai,
T. J., Li, L. X., Liu, B. S., Bi, X. H., Zhang, Y. F., and Feng, Y. C.: Revealing Drivers of Haze Pollution by Explainable Machine
Learning, Environ. Sci. Technol. Lett., 9, 112–119, 2022. Immerzeel, W. W., van Beek, L. P. H., and Bierkens, M. F. P.: Cli-
mate Change Will Affect the Asian Water Towers, Science, 328,
1382–1385, 2010. Duo, B., Cui, L. L., Wang, Z. Z., Li, R., Zhang, L. W., Fu, H. B.,
Chen, J. M., Zhang, H. F., and Qiong, A.: Observations of atmo-
spheric pollutants at Lhasa during 2014–2015: Pollution status
and the influence of meteorological factors, J. Environ. Sci., 63,
28–42, 2018. Jerrett, M., Burnett, R. T., Pope, C. A., Ito, K., Thurston, G.,
Krewski, D., Shi, Y., Calle, E., and Thun, M.: Long-Term Ozone
Exposure and Mortality, New Engl. J. Med., 360, 1085–1095,
https://doi.org/10.1056/NEJMoa0803894, 2009. Jiang, Z., Li, J., Lu, X., Gong, C., Zhang, L., and Liao, H.:
Impact of western Pacific subtropical high on ozone pollu-
tion over eastern China, Atmos. Chem. Phys., 21, 2601–2613,
https://doi.org/10.5194/acp-21-2601-2021, 2021. Gardiner, T., Forbes, A., de Mazière, M., Vigouroux, C., Mahieu,
E., Demoulin, P., Velazco, V., Notholt, J., Blumenstock, T., Hase,
F., Kramer, I., Sussmann, R., Stremme, W., Mellqvist, J., Strand-
berg, A., Ellingsen, K., and Gauss, M.: Trend analysis of green-
house gases over Europe measured by a network of ground-based
remote FTIR instruments, Atmos. Chem. Phys., 8, 6719–6727,
https://doi.org/10.5194/acp-8-6719-2008, 2008. Kang, S. C., Huang, J., Wang, F. Y., Zhang, Q. G., Zhang, Y. L., Li,
C. L., Wang, L., Chen, P. F., Sharma, C. M., Li, Q., Sillanpaa,
M., Hou, J. Z., Xu, B. Q., and Guo, J. M.: Atmospheric Mercury
Depositional Chronology Reconstructed from Lake Sediments
and Ice Core in the Himalayas and Tibetan Plateau, Environ. Sci. Technol., 50, 2859–2869, 2016. Gardner,
M. and
Dorling,
S.:
Artificial
Neural
Network-
Derived Trends in Daily Maximum Surface Ozone Con-
centrations, J. Air Waste Manage. Assoc., 51, 1202–1210,
https://doi.org/10.1080/10473289.2001.10464338, 2001. Kang, S. C., Zhang, Q. G., Qian, Y., Ji, Z. M., Li, C. L., Cong,
Z. Y., Zhang, Y. L., Guo, J. M., Du, W. T., Huang, J., You, Q. L., Panday, A. K., Rupakheti, M., Chen, D. L., Gustafsson, O.,
Thiemens, M. H., and Qin, D. H.: Linking atmospheric pollution
to cryospheric change in the Third Pole region: current progress
and future prospects, Natl. References Gong, C. and Liao, H.: A typical weather pattern for ozone pollution
events in North China, Atmos. Chem. Phys., 19, 13725–13740,
https://doi.org/10.5194/acp-19-13725-2019, 2019. Breiman,
L.:
Random
Forests,
Mach. Learn.,
45,
5–32,
https://doi.org/10.1023/A:1010933404324, 2001. Grange, S. K. and Carslaw, D. C.: Using meteorological normali-
sation to detect interventions in air quality time series, Sci. Total
Environ., 653, 578–588, 2019. Carvalho, D.: An Assessment of NASA’s GMAO MERRA-
2 Reanalysis Surface Winds, J. Climate, 32, 8261–8281,
https://doi.org/10.1175/JCLI-D-19-0199.1, 2019. Grange, S. K., Carslaw, D. C., Lewis, A. C., Boleti, E., and
Hueglin, C.: Random forest meteorological normalisation mod-
els for Swiss PM10 trend analysis, Atmos. Chem. Phys., 18,
6223–6239, https://doi.org/10.5194/acp-18-6223-2018, 2018. Chen, D., Wang, Y., McElroy, M. B., He, K., Yantosca, R. M., and
Le Sager, P.: Regional CO pollution and export in China simu-
lated by the high-resolution nested-grid GEOS-Chem model, At-
mos. Chem. Phys., 9, 3825–3839, https://doi.org/10.5194/acp-9-
3825-2009, 2009. Hakkarainen, J., Ialongo, I., and Tamminen, J.: Direct space-
based
observations
of
anthropogenic
CO2
emission
ar-
eas from OCO-2, Geophys. Res. Lett., 43, 11400–11406,
https://doi.org/10.1002/2016GL070885, 2016. Cong,
Z.,
Kang,
S.,
Zhang,
Y.,
and
Li,
X.:
Atmo-
spheric
wet
deposition
of
trace
elements
to
cen-
tral
Tibetan
Plateau,
Appl.Geochem.,
25,
1415–1421,
https://doi.org/10.1016/j.apgeochem.2010.06.011, 2010. Hakkarainen, J., Ialongo, I., Maksyutov, S., and Crisp, D.: Anal-
ysis of Four Years of Global XCO2 Anomalies as Seen by
Orbiting Carbon Observatory-2, Remote Sens.-Basel, 11, 7,
https://doi.org/10.3390/rs11070850, 2019. Cristofanelli, P., Bracci, A., Sprenger, M., Marinoni, A., Bonafè,
U., Calzolari, F., Duchi, R., Laj, P., Pichon, J. M., Roc-
cato, F., Venzac, H., Vuillermoz, E., and Bonasoni, P.: Tro-
pospheric ozone variations at the Nepal Climate Observatory-
Pyramid (Himalayas, 5079 m a.s.l.) and influence of deep strato-
spheric intrusion events, Atmos. Chem. Phys., 10, 6537–6549,
https://doi.org/10.5194/acp-10-6537-2010, 2010. He, J., Gong, S., Yu, Y., Yu, L., Wu, L., Mao, H., Song, C.,
Zhao, S., Liu, H., Li, X., and Li, R.: Air pollution characteris-
tics and their relation to meteorological conditions during 2014–
2015 in major Chinese cities, Environ. Pollut., 223, 484-496,
https://doi.org/10.1016/j.envpol.2017.01.050, 2017. Decesari, S., Facchini, M. C., Carbone, C., Giulianelli, L., Ri-
naldi, M., Finessi, E., Fuzzi, S., Marinoni, A., Cristofanelli, P.,
Duchi, R., Bonasoni, P., Vuillermoz, E., Cozic, J., Jaffrezo, J. L., and Laj, P.: Chemical composition of PM10 and PM1 at
the high-altitude Himalayan station Nepal Climate Observatory-
Pyramid (NCO-P) (5079 m a.s.l.), Atmos. Chem. Phys., 10,
4583–4596, https://doi.org/10.5194/acp-10-4583-2010, 2010. Hou, L. L., Dai, Q. 7
Conclusions This article is part of the special is-
sue “In-depth study of the atmospheric chemistry over the Tibetan
Plateau: measurement, processing, and the impacts on climate and
air quality (ACP/AMT inter-journal SI)”. It is not associated with a
conference. Acknowledgements. This work is jointly supported by the Na-
tional Natural Science Foundation of China (U21A2027), the
Strategic Priority Research Program of the Chinese Academy of
Sciences (No. XDA23020301), the Key Research and Development
Project of Anhui Province (202104i07020002), the Major Projects
of High Resolution Earth Observation Systems of National Science
and Technology (05-Y30B01-9001-19/20-3), the Youth Innovation
Promotion Association, CAS (No. 2019434) and the Sino-German
Mobility programme (M-0036) funded by the National Natural Sci-
ence Foundation of China (NSFC) and Deutsche Forschungsge-
meinschaft (DFG). Financial support. This research has been supported by the Na-
tional Natural Science Foundation of China (U21A2027), the
Strategic Priority Research Program of the Chinese Academy of
Sciences (no. XDA23020301), the Key Research and Development
Project of Anhui Province (202104i07020002), the Major Projects
of High Resolution Earth Observation Systems of National Science
and Technology (05-Y30B01-9001-19/20-3), the Youth Innovation
Promotion Association, CAS (no. 2019434) and the Sino-German
Mobility programme (M-0036) funded by the National Natural Sci- https://doi.org/10.5194/acp-22-14401-2022 Atmos. Chem. Phys., 22, 14401–14419, 2022 14416 ence Foundation of China (NSFC) and Deutsche Forschungsge-
meinschaft (DFG). ence Foundation of China (NSFC) and Deutsche Forschungsge-
meinschaft (DFG). tyka, G., Pawson, S., Putman, W., Rienecker, M., Schubert, S. D.,
Sienkiewicz, M., and Zhao, B.: The Modern-Era Retrospective
Analysis for Research and Applications, Version 2 (MERRA-2),
J. Climate, 30, 5419–5454, https://doi.org/10.1175/JCLI-D-16-
0758.1, 2017. Review statement. This paper was edited by Steven Brown and
reviewed by two anonymous referees. GMAO – The NASA Global Modeling and Assimilation Of-
fice: Modern-Era Retrospective analysis for Research and
Applications, Version 2, https://gmao.gsfc.nasa.gov/reanalysis/
MERRA-2/, last access: 1 April 2022. H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas 14417 Lüthi, Z. L., Škerlak, B., Kim, S. W., Lauer, A., Mues, A.,
Rupakheti, M., and Kang, S.: Atmospheric brown clouds
reach the Tibetan Plateau by crossing the Himalayas, Atmos. Chem. Phys., 15, 6007–6021, https://doi.org/10.5194/acp-15-
6007-2015, 2015. Kishore Kumar, G., Kishore Kumar, K., Baumgarten, G., and
Ramkumar, G.: Validation of MERRA reanalysis upper-
level
winds
over
low
latitudes
with
independent
rocket
sounding
data,
J. Atmos. Sol.-Terr. Phy.,
123,
48–54,
https://doi.org/10.1016/j.jastp.2014.12.001, 2015. Li, K., Jacob, D. J., Liao, H., Shen, L., Zhang, Q., and Bates, K. H.: Anthropogenic drivers of 2013–2017 trends in summer sur-
face ozone in China, P. Natl. Acad. Sci. USA, 116, 422–427,
https://doi.org/10.1073/pnas.1812168116, 2019. Ma, J., Dörner, S., Donner, S., Jin, J., Cheng, S., Guo, J., Zhang,
Z., Wang, J., Liu, P., Zhang, G., Pukite, J., Lampel, J., and Wag-
ner, T.: MAX-DOAS measurements of NO2, SO2, HCHO, and
BrO at the Mt. Waliguan WMO GAW global baseline station
in the Tibetan Plateau, Atmos. Chem. Phys., 20, 6973–6990,
https://doi.org/10.5194/acp-20-6973-2020, 2020. Li, K., Jacob, D. J., Shen, L., Lu, X., De Smedt, I., and Liao,
H.: Increases in surface ozone pollution in China from 2013
to 2019: anthropogenic and meteorological influences, Atmos. Chem. Phys., 20, 11423–11433, https://doi.org/10.5194/acp-20-
11423-2020, 2020. Ma, M., Gao, Y., Wang, Y., Zhang, S., Leung, L. R., Liu, C., Wang,
S., Zhao, B., Chang, X., Su, H., Zhang, T., Sheng, L., Yao,
X., and Gao, H.: Substantial ozone enhancement over the North
China Plain from increased biogenic emissions due to heat waves
and land cover in summer 2017, Atmos. Chem. Phys., 19, 12195–
12207, https://doi.org/10.5194/acp-19-12195-2019, 2019. Li, M., Zhang, Q., Kurokawa, J., Woo, J. H., He, K. B., Lu, Z. F.,
Ohara, T., Song, Y., Streets, D. G., Carmichael, G. R., Cheng, Y. F., Hong, C. P., Huo, H., Jiang, X. J., Kang, S. C., Liu, F., Su,
H., and Zheng, B.: MIX: a mosaic Asian anthropogenic emis-
sion inventory under the international collaboration framework
of the MICS-Asia and HTAP, Atmos. Chem. Phys., 17, 935–963,
https://doi.org/10.5194/acp-17-935-2017, 2017. Maione, M., Giostra, U., Arduini, J., Furlani, F., Bonasoni, P.,
Cristofanelli, P., Laj, P., and Vuillermoz, E.: Three-year ob-
servations of halocarbons at the Nepal Climate Observatory at
Pyramid (NCO-P, 5079 m a.s.l.) on the Himalayan range, At-
mos. Chem. Phys., 11, 3431–3441, https://doi.org/10.5194/acp-
11-3431-2011, 2011. H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas Li, R., Zhao, Y., Zhou, W., Meng, Y., Zhang, Z., and Fu,
H.: Developing a novel hybrid model for the estimation of
surface 8 h ozone (O3) across the remote Tibetan Plateau
during 2005–2018, Atmos. Chem. Phys., 20, 6159–6175,
https://doi.org/10.5194/acp-20-6159-2020, 2020. Mustafa, F., Bu, L., Wang, Q., Yao, N., Shahzaman, M., Bi-
lal, M., Aslam, R. W., and Iqbal, R.: Neural-network-based
estimation of regional-scale anthropogenic CO2 emissions us-
ing an Orbiting Carbon Observatory-2 (OCO-2) dataset over
East and West Asia, Atmos. Meas. Tech., 14, 7277–7290,
https://doi.org/10.5194/amt-14-7277-2021, 2021. Liu, S., Fang, S. X., Liu, P., Liang, M., Guo, M. R., and Feng,
Z. Z.: Measurement report: Changing characteristics of atmo-
spheric CH4 in the Tibetan Plateau: records from 1994 to 2019 at
the Mount Waliguan station, Atmos. Chem. Phys., 21, 393–413,
https://doi.org/10.5194/acp-21-393-2021, 2021. Pu,
Z. X.,
Xu,
L.,
and
Salomonson,
V. V.:
MODIS-
/Terra
observed
seasonal
variations
of
snow
cover
over
the
Tibetan
Plateau,
Geophys. Res. Lett.,
34,
L06706,
https://doi.org/10.1029/2007GL029262, 2007. Loewen, M., Kang, S., Armstrong, D., Zhang, Q., Tomy, G., and
Wang, F.: Atmospheric transport of mercury to the Tibetan
plateau, Environ. Sci. Technol., 41, 7632–7638, 2007. Putero, D., Landi, T. C., Cristofanelli, P., Marinoni, A., Laj, P.,
Duchi, R., Calzolari, F., Verza, G. P., and Bonasoni, P.: Influence
of open vegetation fires on black carbon and ozone variability in
the southern Himalayas (NCO-P, 5079 m a.s.l.), Environ. Pollut.,
184,
597–604,
https://doi.org/10.1016/j.envpol.2013.09.035,
2014. Lu, X., Hong, J., Zhang, L., Cooper, O. R., Schultz, M. G., Xu, X., Wang, T., Gao, M., Zhao, Y., and Zhang,
Y.: Severe Surface Ozone Pollution in China: A Global
Perspective,
Environ. Sci. Technol. Lett.,
5,
487–494,
https://doi.org/10.1021/acs.estlett.8b00366, 2018. Qiu, J.: The third pole, Nature, 454, 393–396, 2008. Lu, X., Zhang, L., Chen, Y., Zhou, M., Zheng, B., Li, K., Liu, Y.,
Lin, J., Fu, T. M., and Zhang, Q.: Exploring 2016–2017 sur-
face ozone pollution over China: source contributions and me-
teorological influences, Atmos. Chem. Phys., 19, 8339–8361,
https://doi.org/10.5194/acp-19-8339-2019, 2019a. Santer, B. D., Thorne, P. W., Haimberger, L., Taylor, K. E., Wigley,
T. M. L., Lanzante, J. R., Solomon, S., Free, M., Gleckler, P. J.,
Jones, P. D., Karl, T. R., Klein, S. A., Mears, C., Nychka, D.,
Schmidt, G. A., Sherwood, S. C., and Wentz, F. J.: Consistency
of modelled and observed temperature trends in the tropical tro-
posphere, Int. J. Climatol., 28, 1703–1722, 2008. References Sci. Rev., 6, 796–809, 2019. Gelaro, R., McCarty, W., Suárez, M. J., Todling, R., Molod, A.,
Takacs, L., Randles, C. A., Darmenov, A., Bosilovich, M. G., Re-
ichle, R., Wargan, K., Coy, L., Cullather, R., Draper, C., Akella,
S., Buchard, V., Conaty, A., da Silva, A. M., Gu, W., Gi-Kong,
K., Koster, R., Lucchesi, R., Merkova, D., Nielsen, J. E., Par- https://doi.org/10.5194/acp-22-14401-2022 Atmos. Chem. Phys., 22, 14401–14419, 2022 H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas Xu, W., Xu, X., Lin, M., Lin, W., Tarasick, D., Tang, J., Ma, J., and
Zheng, X.: Long-term trends of surface ozone and its influencing
factors at the Mt Waliguan GAW station, China – Part 2: The
roles of anthropogenic emissions and climate variability, Atmos. Chem. Phys., 18, 773–798, https://doi.org/10.5194/acp-18-773-
2018, 2018. Sun, Y., Wang, Y., and Zhang, C.: Vertical observations and analysis
of PM2.5, O3, and NOx at Beijing and Tianjin from towers dur-
ing summer and Autumn 2006, Adv. Atmos. Sci., 27, 123–136,
https://doi.org/10.1007/s00376-009-8154-z, 2009. Sun, Y., Liu, C., Zhang, L., Palm, M., Notholt, J., Yin, H.,
Vigouroux, C., Lutsch, E., Wang, W., Shan, C., Blumenstock,
T., Nagahama, T., Morino, I., Mahieu, E., Strong, K., Lange-
rock, B., De Mazière, M., Hu, Q., Zhang, H., Petri, C.,
and Liu, J.: Fourier transform infrared time series of tropo-
spheric HCN in eastern China: seasonality, interannual variabil-
ity, and source attribution, Atmos. Chem. Phys., 20, 5437–5456,
https://doi.org/10.5194/acp-20-5437-2020, 2020. Xu, W., Lin, W., Xu, X., Tang, J., Huang, J., Wu, H., and Zhang,
X.: Long-term trends of surface ozone and its influencing fac-
tors at the Mt Waliguan GAW station, China – Part 1: Overall
trends and characteristics, Atmos. Chem. Phys., 16, 6191–6205,
https://doi.org/10.5194/acp-16-6191-2016, 2016. Xue, L. K., Wang, T., Zhang, J. M., Zhang, X. C., Deliger, Poon,
C. N., Ding, A. J., Zhou, X. H., Wu, W. S., Tang, J., Zhang,
Q. Z., and Wang, W. X.: Source of surface ozone and reactive
nitrogen speciation at Mount Waliguan in western China: New
insights from the 2006 summer study, J. Geophys. Res.-Atmos.,
116, D07306, https://doi.org/10.1029/2010JD014735, 2011. Sun, Y., Yin, H., Liu, C., Mahieu, E., Notholt, J., Té, Y.,
Lu, X., Palm, M., Wang, W., Shan, C., Hu, Q., Qin, M.,
Tian, Y., and Zheng, B.: The reduction in C2H6 from 2015
to 2020 over Hefei, eastern China, points to air quality im-
provement in China, Atmos. Chem. Phys., 21, 11759–11779,
https://doi.org/10.5194/acp-21-11759-2021, 2021a. Yanai, M. H., Li, C. F., and Song, Z. S.: Seasonal Heating of the Ti-
betan Plateau and Its Effects on the Evolution of the Asian Sum-
mer Monsoon, J. Meteorol. Soc. Jpn., 70, 319–351, 1992. H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas Lu, X., Zhang, L., and Shen, L.: Meteorology and Climate Influ-
ences on Tropospheric Ozone: a Review of Natural Sources,
Chemistry, and Transport Patterns, Curr. Pollut. Rep., 5, 238–
260, https://doi.org/10.1007/s40726-019-00118-3, 2019b. Shen, Z., Cao, J., Zhang, L., Zhao, Z., Dong, J., Wang, L.,
Wang, Q., Li, G., Liu, S., and Zhang, Q.: Characteristics
of surface O3 over Qinghai Lake area in Northeast Ti-
betan Plateau, China, Sci. Total Environ., 500–501, 295–301,
https://doi.org/10.1016/j.scitotenv.2014.08.104, 2014. Lu,
X.,
Zhang,
L.,
Wang,
X.,
Gao,
M.,
Li,
K.,
Zhang,
Y., Yue, X., and Zhang, Y.: Rapid Increases in Warm-
Season
Surface
Ozone
and
Resulting
Health
Impact
in
China Since 2013, Environ. Sci. Technol. Lett., 7, 240–247,
https://doi.org/10.1021/acs.estlett.0c00171, 2020. Shi, Z. B., Song, C. B., Liu, B. W., Lu, G. D., Xu, J. S., Vu, T. V., Elliott, R. J. R., Li, W. J., Bloss, W. J., and Harrison, R. M.:
Abrupt but smaller than expected changes in surface air qual-
ity attributable to COVID-19 lockdowns, Sci. Adv., 7, eabd6696,
https://doi.org/10.1126/sciadv.abd6696, 2021. Lu, X., Ye, X., Zhou, M., Zhao, Y., Weng, H., Kong, H., Li, K.,
Gao, M., Zheng, B., Lin, J., Zhou, F., Zhang, Q., Wu, D., Zhang,
L., and Zhang, Y.: The underappreciated role of agricultural soil
nitrogen oxide emissions in ozone pollution regulation in North
China, Nat. Commun., 12, 5021, https://doi.org/10.1038/s41467-
021-25147-9, 2021. Škerlak, B., Sprenger, M., and Wernli, H.: A global climatology
of stratosphere–troposphere exchange using the ERA-Interim https://doi.org/10.5194/acp-22-14401-2022 Atmos. Chem. Phys., 22, 14401–14419, 2022 H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas 14418 Atmos., 117, D05102, https://doi.org/10.1029/2011JD016553,
2012. data set from 1979 to 2011, Atmos. Chem. Phys., 14, 913–937,
https://doi.org/10.5194/acp-14-913-2014, 2014. data set from 1979 to 2011, Atmos. Chem. Phys., 14, 913–937,
https://doi.org/10.5194/acp-14-913-2014, 2014. Song, Z., Fu, D., Zhang, X., Wu, Y., Xia, X., He, J., Han, X., Zhang,
R., and Che, H.: Diurnal and seasonal variability of PM2.5 and
AOD in North China plain: Comparison of MERRA-2 prod-
ucts and ground measurements, Atmos. Environ., 191, 70–78,
https://doi.org/10.1016/j.atmosenv.2018.08.012, 2018. Xie, Z., Hu, Z., Gu, L., Sun, G., Du, Y., and Yan, X.: Meteorologi-
cal Forcing Datasets for Blowing Snow Modeling on the Tibetan
Plateau: Evaluation and Intercomparison, J. Hydrometeorol., 18,
2761–2780, https://doi.org/10.1175/JHM-D-17-0075.1, 2017. H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas Sun, Y., Yin, H., Liu, C., Zhang, L., Cheng, Y., Palm, M., Notholt,
J., Lu, X., Vigouroux, C., Zheng, B., Wang, W., Jones, N., Shan,
C., Qin, M., Tian, Y., Hu, Q., Meng, F., and Liu, J.: Map-
ping the drivers of formaldehyde (HCHO) variability from 2015
to 2019 over eastern China: insights from Fourier transform
infrared observation and GEOS-Chem model simulation, At-
mos. Chem. Phys., 21, 6365–6387, https://doi.org/10.5194/acp-
21-6365-2021, 2021b. Yang, R. Q., Zhang, S. J., Li, A., Jiang, G. B., and Jing, C. Y.: Al-
titudinal and Spatial Signature of Persistent Organic Pollutants
in Soil, Lichen, Conifer Needles, and Bark of the Southeast Ti-
betan Plateau: Implications for Sources and Environmental Cy-
cling, Environ. Sci. Technol., 47, 12736–12743, 2013. Yin, H., Sun, Y., Liu, C., Zhang, L., Lu, X., Wang, W., Shan,
C., Hu, Q., Tian, Y., Zhang, C., Su, W., Zhang, H., Palm,
M., Notholt, J., and Liu, J.: FTIR time series of strato-
spheric NO2 over Hefei, China, and comparisons with OMI
and GEOS-Chem model data, Opt. Express, 27, A1225–A1240,
https://doi.org/10.1364/OE.27.0A1225, 2019. Sun, Y., Yin, H., Lu, X., Notholt, J., Palm, M., Liu, C., Tian,
Y., and Zheng, B.: The drivers and health risks of unex-
pected surface ozone enhancements over the Sichuan Basin,
China, in 2020, Atmos. Chem. Phys., 21, 18589–18608,
https://doi.org/10.5194/acp-21-18589-2021, 2021c. Yin, H., Sun, Y., Liu, C., Lu, X., Smale, D., Blumenstock,
T., Nagahama, T., Wang, W., Tian, Y., Hu, Q., Shan, C.,
Zhang, H., and Liu, J.: Ground-based FTIR observation
of hydrogen chloride (HCl) over Hefei, China, and com-
parisons with GEOS-Chem model data and other ground-
based FTIR stations data, Opt. Express, 28, 8041–8055,
https://doi.org/10.1364/OE.384377, 2020. Sun, Y., Yin, H., Cheng, Y., Zhang, Q., Zheng, B., Notholt, J.,
Lu, X., Liu, C., Tian, Y., and Liu, J.: Quantifying variability,
source, and transport of CO in the urban areas over the Hi-
malayas and Tibetan Plateau, Atmos. Chem. Phys., 21, 9201-
9222, 10.5194/acp-21-9201-2021, 2021d. Sun, Y., Liu, C., Palm, M., Vigouroux, C., Notholt, J., Hu, Q., Jones,
N., Wang, W., Su, W., Zhang, W., Shan, C., Tian, Y., Xu, X., De
Mazière, M., Zhou, M., and Liu, J.: Ozone seasonal evolution
and photochemical production regime in the polluted troposphere
in eastern China derived from high-resolution Fourier trans-
form spectrometry (FTS) observations, Atmos. Chem. Phys.,
18, 14569–14583, https://doi.org/10.5194/acp-18-14569-2018,
2018. H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas 14419 Yin, X., Foy, B. D., Wu, K., Feng, C., and Zhang, Q.: Gaseous and
particulate pollutants in Lhasa, Tibet during 2013–2017: Spatial
variability, temporal variations and implications, Environ. Pol-
lut., 253, 68–77, https://doi.org/10.1016/j.envpol.2019.06.113,
2019. Zhang, Y. M., Vu, T. V., Sun, J. Y., He, J. J., Shen, X. J., Lin, W. L., Zhang, X. Y., Zhong, J. T., Gao, W. K., Wang, Y. Q., Fu,
T. Y., Ma, Y. P., Li, W. J., and Shi, Z. B.: Significant Changes
in Chemistry of Fine Particles in Wintertime Beijing from 2007
to 2017: Impact of Clean Air Actions, Environ. Sci. Technol., 54,
1344–1352, 2020. Yin, X., Kang, S., de Foy, B., Cong, Z., Luo, J., Zhang, L., Ma,
Y., Zhang, G., Rupakheti, D., and Zhang, Q.: Surface ozone
at Nam Co in the inland Tibetan Plateau: variation, synthe-
sis comparison and regional representativeness, Atmos. Chem. Phys., 17, 11293–11311, https://doi.org/10.5194/acp-17-11293-
2017, 2017. Zhao, S. P., Yu, Y., Yin, D. Y., He, J. J., Liu, N., Qu, J. J., and Xiao, J. H.: Annual and diurnal variations of gaseous and particulate pol-
lutants in 31 provincial capital cities based on in situ air quality
monitoring data from China National Environmental Monitoring
Center, Environ. Int., 86, 92–106, 2016. Yin, X. F., de Foy, B., Wu, K. P., Feng, C., Kang, S. C., and Zhang,
Q. G.: Gaseous and particulate pollutants in Lhasa, Tibet dur-
ing 2013–2017: Spatial variability, temporal variations and im-
plications, Environ. Pollut., 253, 68–77, 2019. Zhou, C., Wang, K., and Ma, Q.: Evaluation of Eight Cur-
rent Reanalyses in Simulating Land Surface Temperature
from 1979 to 2003 in China, J. Climate, 30, 7379–7398,
https://doi.org/10.1175/JCLI-D-16-0903.1, 2017. Zhai, S., Jacob, D. J., Wang, X., Shen, L., Li, K., Zhang, Y., Gui, K.,
Zhao, T., and Liao, H.: Fine particulate matter (PM2.5) trends in
China, 2013–2018: separating contributions from anthropogenic
emissions and meteorology, Atmos. Chem. Phys., 19, 11031–
11041, https://doi.org/10.5194/acp-19-11031-2019, 2019. Zhu, B., Akimoto, H., Wang, Z., Sudo, K., Tang, J., and
Uno,
I.:
Why
does
surface
ozone
peak
in
summer-
time
at
Waliguan?,
Geophys. Res. Lett.,
31,
L17104,
https://doi.org/10.1029/2004GL020609, 2004. Zhu, N. L., Fu, J. J., Gao, Y., Ssebugere, P., Wang, Y. W., and Jiang,
G. H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas Yin, H., Liu, C., Hu, Q., Liu, T., Wang, S., Gao, M., Xu, S., Zhang,
C., and Su, W.: Opposite impact of emission reduction during
the COVID-19 lockdown period on the surface concentrations of
PM2.5 and O3 in Wuhan, China, Environ. Pollut., 289, 117899,
https://doi.org/10.1016/j.envpol.2021.117899, 2021a. Yin, H., Lu, X., Sun, Y., Li, K., Gao, M., Zheng, B., and Liu,
C.: Unprecedented decline in summertime surface ozone over
eastern China in 2020 comparably attributable to anthropogenic
emission reductions and meteorology, Environ. Res. Lett., 16,
124069, https://doi.org/10.1088/1748-9326/ac3e22, 2021b. Vu, T. V., Shi, Z., Cheng, J., Zhang, Q., He, K., Wang, S., and Har-
rison, R. M.: Assessing the impact of clean air action on air qual-
ity trends in Beijing using a machine learning technique, Atmos. Chem. Phys., 19, 11303–11314, https://doi.org/10.5194/acp-19-
11303-2019, 2019. Yin, H., Sun, Y., Notholt, J., Palm, M., and Liu, C.: Spaceborne
tropospheric nitrogen dioxide (NO2) observations from 2005–
2020 over the Yangtze River Delta YRD), China: variabilities,
implications, and drivers, Atmos. Chem. Phys., 22, 4167–4185,
https://doi.org/10.5194/acp-22-4167-2022, 2022. Wang, A. and Zeng, X.: Evaluation of multireanalysis products with
in situ observations over the Tibetan Plateau, J. Geophys. Res.- Atmos. Chem. Phys., 22, 14401–14419, 2022 https://doi.org/10.5194/acp-22-14401-2022 H. Yin et al.: Quantifying the drivers of surface ozone anomalies in urban areas B.: Hexabromocyclododecane in alpine fish from the Tibetan
Plateau, China, Environ. Pollut., 181, 7–13, 2013. Atmos. Chem. Phys., 22, 14401–14419, 2022 https://doi.org/10.5194/acp-22-14401-2022
|
https://openalex.org/W2146234100
|
https://projecteuclid.org/journals/electronic-journal-of-statistics/volume-11/issue-1/Finite-sample-properties-of-tests-based-on-prewhitened-nonparametric-covariance/10.1214/17-EJS1281.pdf
|
English
| null |
Finite sample properties of tests based on prewhitened nonparametric covariance estimators
|
Electronic journal of statistics
| 2,017
|
cc-by
| 45,823
|
Electronic Journal of Statistics
Vol. 11 (2017) 2097–2167
ISSN: 1935-7524
DOI: 10.1214/17-EJS1281 Electronic Journal of Statistics
Vol. 11 (2017) 2097–2167
ISSN: 1935-7524
DOI: 10.1214/17-EJS1281 Finite sample properties of tests
based on prewhitened nonparametric
covariance estimators David Preinerstorfer
CREATES and Department of Economics and Business Economics, Aarhus University
Fuglesangs All´e 4, 8210 Aarhus V, Denmark
e-mail: david.preinerstorfer@econ.au.dk Abstract: We analytically investigate size and power properties of a popu-
lar family of procedures for testing linear restrictions on the coefficient vec-
tor in a linear regression model with temporally dependent errors. The tests
considered are autocorrelation-corrected F-type tests based on prewhitened
nonparametric covariance estimators that possibly incorporate a data-de-
pendent bandwidth parameter, e.g., estimators as considered in Andrews
and Monahan (1992), Newey and West (1994), or Rho and Shao (2013). For design matrices that are generic in a measure theoretic sense we prove
that these tests either suffer from extreme size distortions or from strong
power deficiencies. Despite this negative result we demonstrate that a sim-
ple adjustment procedure based on artificial regressors can often resolve
this problem. MSC 2010 subject classifications: Primary 62F03; Secondary 62J05,
62F35, 62M10, 62M15. MSC 2010 subject classifications: Primary 62F03; Secondary 62J05,
62F35, 62M10, 62M15. Keywords and phrases: Autocorrelation robustness, prewhitening, size
distortion, power deficiency, artificial regressors. Contents 1. Introduction The construction of tests for hypotheses on the coefficient vector in linear re-
gression models with dependent errors is highly practically relevant and has
received lots of attention in the statistics and econometrics literature. The main
challenge is to obtain tests with good size and power properties in situations
where the nuisance parameter governing the dependence structure of the er-
rors is high- or possibly infinite-dimensional and allows for strong correlations. The large majority of available procedures are autocorrelation-corrected F-type
tests, based on nonparametric covariance estimators trying to take into account
the autocorrelation in the disturbances. The tests currently used can roughly be
categorized into two groups, the distinction depending on the choice of the criti-
cal values. The first group of such tests is based on critical values obtained from
an asymptotic framework in which the nonparametric covariance estimators are
consistent, and where the asymptotic distribution of the F-type test statistic un-
der the null hypothesis is a χ2 distribution. Quantiles of this limiting distribution
are then used for testing. Concerning these tests, important contributions in the
econometrics literature are Newey and West (1987), Andrews (1991), Andrews
and Monahan (1992), and Newey and West (1994). It is safe to say that using F-
type tests based on χ2 critical values and the covariance estimators introduced
in the latter two articles currently constitutes the gold standard for the testing
problem under consideration. In contrast to the estimator suggested earlier by
Newey and West (1987) - structurally 2π times a standard kernel spectral den-
sity estimator (Bartlett (1950), Jowett (1955), Hannan (1957), and Grenander
and Rosenblatt (1957) Section 7.9) evaluated at frequency 0 - the covariance
estimators suggested in Andrews and Monahan (1992) and Newey and West
(1994) both incorporate an additional prewhitening step based on an auxiliary
vector autoregressive (VAR) model, as well as a data-dependent bandwidth pa-
rameter. A distinguishing feature of the estimators introduced by Andrews and
Monahan (1992) on the one hand and Newey and West (1994) on the other hand
is the choice of the bandwidth parameter: Andrews and Monahan (1992) used
an approach introduced by Andrews (1991), where the bandwidth parameter is
chosen based on auxiliary parametric models. In contrast to that, Newey and
West (1994) suggested a nonparametric approach for choosing the bandwidth
parameter. Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2165 Contents 2128
B Proofs of results in Section 3.3
. . . . . . . . . . . . . . . . . . . . . 2129
C Proofs of results in Section 4
. . . . . . . . . . . . . . . . . . . . . . 2140
D Proofs of results in Section 5
. . . . . . . . . . . . . . . . . . . . . . 2157
E
Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2160
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2165
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2165 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2128
B Proofs of results in Section 3.3
. . . . . . . . . . . . . . . . . . . . . 2129
C Proofs of results in Section 4
. . . . . . . . . . . . . . . . . . . . . . 2140
D Proofs of results in Section 5
. . . . . . . . . . . . . . . . . . . . . . 2157
E
Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2160
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2165
References . . . . . . . Contents 1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2098
2
The framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2101
3
Tests based on prewhitened covariance estimators . . . . . . . . . . . 2104
3.1
Bandwidth parameters . . . . . . . . . . . . . . . . . . . . . . . 2106
3.1.1
The parametric approach of Andrews and Monahan (1992) 2107
3.1.2
The non-parametric approach of Newey and West (1994) 2108
3.1.3
Data-independent bandwidth parameters
. . . . . . . . 2109
3.2
Assumptions on κ, M and p . . . . . . . . . . . . . . . . . . . . 2109
3.3
Structural properties of prewhitened covariance estimators . . . 2110
4
A negative result and its generic applicability . . . . . . . . . . . . . 2112
5
A positive result, an adjustment procedure and its generic applicability 2119
6
Numerical results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2125
6.1
Example 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2125
6.2
Example 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2126
7
Conclusion
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2127
2097 2097 D. Preinerstorfer 2098 Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Introduction Even though simulation studies have shown that the inclusion of a
prewhitening step and the data-dependent choice of the bandwidth parameter
can improve the finite sample properties of the tests obtained, these more so- Finite sample properties of prewhitened F-type tests 2099 phisticated tests still suffer from size distortions and power deficiencies. For this
reason Kiefer, Vogelsang and Bunzel (2000), Kiefer and Vogelsang (2002), and
Kiefer and Vogelsang (2005) suggested a different asymptotic framework for ob-
taining critical values. Their framework, in which the bandwidth parameter is a
fixed proportion of the sample size, leads to inconsistent covariance estimators
and to a non-standard limiting distribution of the corresponding test statistic
under the null hypothesis, the quantiles of which are used to obtain tests. In
simulation studies it has been observed that these tests still suffer from size dis-
tortions in finite samples, but less so than tests based on χ2 critical values. How-
ever, this is at the expense of some loss in power. Furthermore, simulation results
in Kiefer and Vogelsang (2005) and Rho and Shao (2013) suggest that the finite
sample properties of tests based on non-standard critical values can be improved
by incorporating a prewhitening step. In the latter paper it was also shown that
the asymptotic distribution under the null of the test suggested by Kiefer, Vo-
gelsang and Bunzel (2000) is the same whether or not prewhitening is used. A number of recent studies (Velasco and Robinson (2001), Jansson (2004),
Sun, Phillips and Jin (2008, 2011), Zhang and Shao (2013)) tried to use higher
order expansions to uncover the mechanism leading to size distortions and power
deficiencies of the above mentioned tests. These higher-order asymptotic results
(and also the first-order results discussed above) are pointwise in the sense that
they are obtained under the assumption of a fixed underlying data-generating-
process. Hence, while they inform us about the limit of the rejection probability
and the rate of convergence to this limit for a fixed underlying data-generating-
process, they do not inform us about the size of the test or its limit as sample
size increases, nor about the power function or its asymptotic behavior. 1. Introduction Size
and power properties of tests in regression models with dependent errors were
recently studied in Preinerstorfer and P¨otscher (2016): In a general finite sam-
ple setup and under high-level conditions on the structure of the test and the
covariance model, they derived conditions on the design matrix under which
a concentration mechanism due to strong dependencies leads to extreme size
distortions or power deficiencies. Furthermore, they suggested an adjustment-
procedure to obtain a modified test with improved size and power properties. Specializing their general theory to a covariance model that includes at least all
covariance matrices corresponding to stationary autoregressive processes of or-
der one (AR(1)), they investigated finite sample properties of F-type tests based
on non-prewhitened covariance estimators with data-independent bandwidth pa-
rameters (covering inter alia the procedures in Newey and West (1987), Sections
3-5 of Andrews (1991), Hansen (1992), Kiefer, Vogelsang and Bunzel (2000),
Kiefer and Vogelsang (2002, 2005), Jansson (2002, 2004), but not the methods
considered by Andrews and Monahan (1992), Newey and West (1994) or Rho
and Shao (2013)). In this setup Preinerstorfer and P¨otscher (2016) demonstrated
that these tests break down in terms of their size or power behavior for generic
design matrices. Despite this negative result, they also showed that the adjust-
ment procedure can often solve these problems, if elements of the covariance
model which are close to being singular can be well approximated by AR(1)
covariance matrices. 2100 D. Preinerstorfer Preinerstorfer and P¨otscher (2016), however, did not consider tests based on
prewhitened covariance estimators or data-dependent bandwidth parameters. Therefore the question remains, whether the more sophisticated tests typically
used in practice, i.e., tests based on χ2 critical values and the estimators by
Andrews and Monahan (1992) or Newey and West (1994), and the prewhitened
tests based on non-standard critical values and data-independent bandwidth
parameters, i.e., tests as considered in Rho and Shao (2013), also suffer from
extreme size distortions and power deficiencies, or if prewhitening and the use
of data-dependent bandwidth parameters can indeed resolve or at least sub-
stantially alleviate these problems. In the present paper we investigate finite
sample properties of tests based on prewhitened covariance estimators or data-
dependent bandwidth parameters. 1. Introduction In particular our analysis covers tests based
on prewhitened covariance estimators using auxiliary AR(1) models for the con-
struction of the bandwidth parameter as discussed in Andrews and Monahan
(1992), tests based on prewhitened covariance estimators as discussed in Newey
and West (1994), and prewhitened tests based on non-standard critical values
as discussed in Rho and Shao (2013). We show that the tests considered, albeit
being structurally much more complex, exhibit a similar behavior as their non-
prewhitened counterparts with data-independent bandwidth parameters: First,
we establish conditions on the design matrix under which the tests considered
have (i) size equal to one, or (ii) size not smaller than one half, or (iii) nuisance-
minimal power equal to zero, respectively. We then demonstrate that at least
one of these conditions is generically satisfied, showing that the tests considered
break down for generic design matrices. It is important to stress that this generic negative result does not only apply
to tests based on χ2 critical values, or to tests based on one of the non-standard
critical values mentioned above. The result is applicable to every F-type test
based on one of the nonparametric covariance estimators considered and com-
bined with any (data-independent) critical value 0 < C < ∞. Hence, the prob-
lem described by our generic negative result can not be resolved by simply
adjusting (data-independently) the critical value used. (
)
Motivated by this negative result, we introduce an adjustment procedure. Under the assumption that elements of the covariance model which are close to
being singular can be well approximated by AR(1) covariance matrices, we show
that the adjustment procedure, if applicable, leads to tests that do not suffer
from extreme size distortions or power deficiencies. Finally, it is shown that
the adjustment procedure is applicable under generic conditions on the design
matrix, unless the regression includes the intercept and the hypothesis to be
tested restricts the corresponding coefficient. On a technical level we employ
the general theory developed in Preinerstorfer and P¨otscher (2016). We remark,
however, that the genericity results in particular do not follow from this general
theory. Rather they are obtained by studying and carefully exploiting the specific
structure of the procedures under consideration. The paper is organized as follows: The framework is introduced in Section 2. In Section 3 we introduce the test statistics, covariance estimators, and band-
width parameters we analyze. 1. Introduction In Section 4 we establish our negative result and Finite sample properties of prewhitened F-type tests 2101 its genericity. In Section 5 we discuss the adjustment-procedure and its generic
applicability. Numerical results are presented in Section 6. Section 7 concludes. The proofs are collected in Appendices B-D. Appendix E contains tables for the
numerical results. Consider the linear regression model Consider the linear regression model (1) Y = Xβ + U,
(1) Y = Xβ + U, where X is a (real) n × k dimensional non-stochastic design matrix satisfying
n > 2, rank(X) = k and 1 ≤k < n. Here, β ∈Rk denotes the unknown
regression parameter vector, and the disturbance vector U = (u1, . . . , un)′ is
Gaussian, has mean zero and its unknown covariance matrix is given by σ2Σ. The parameter σ2 satisfies 0 < σ2 < ∞and Σ is assumed to be an element of
a prescribed (non-void) set of positive definite and symmetric n × n matrices
C, which we shall refer to as the covariance model. Throughout we impose the
assumption on C that the parameters σ2 and Σ can be uniquely determined
from σ2Σ. Remark 2.1. The leading case we have in mind is the situation where u1, . . . , un
are n consecutive elements of a weakly stationary process. In such a setup a co-
variance model is typically obtained from a prescribed (non-void) set of spectral
densities F. Assuming that no element of F vanishes identically almost every-
where, the covariance model corresponding to F is then given by C(F) = {Σ(f) : f ∈F} , with Σ(f) =
π
−π
exp(−ιλ(i −j))f(λ)dλ
π
−π
f(λ)dλ
n
i,j=1
,
(2) (2) and where ι denotes the imaginary unit. Every such Σ(f) is positive definite and
symmetric. Furthermore, since Σ(f) is a correlation matrix, σ2 and Σ(f) can
uniquely be determined from σ2Σ(f). As outlined in the Introduction the tests
we focus on in this article are particularly geared towards setups where F is a
nonparametric class of spectral densities, i.e., where the corresponding set C(F)
is rich. A typical example is the class Fξ, which consists of all spectral densities of
linear processes the coefficients of which satisfy a certain summability condition,
i.e., spectral densities of the form f(λ) = (2π)−1
∞
j=0
cj exp(−ιjλ)
2
, where, for a fixed ξ ≥0, the summability condition 0 < ∞
j=0 jξ|cj| < ∞is
satisfied. We observe that C(Fξ) contains in particular all correlation matrices
corresponding to spectral densities of stationary autoregressive moving average
models of arbitrary large order. 2102 D. Consider the linear regression model Preinerstorfer The linear model described in (1) induces a collection of distributions on
(Rn, B(Rn)), the sample space of Y. Denoting a Gaussian probability measure
with mean μ ∈Rn and covariance matrix σ2Σ by Pμ,σ2Σ and denoting the
regression manifold by M = span(X), the induced collection of distributions is
given by Pμ,σ2Σ : μ ∈M, 0 < σ2 < ∞, Σ ∈C
. Pμ,σ2Σ : μ ∈M, 0 < σ2 < ∞, Σ ∈C
. (3) (3) Since every Σ ∈C is positive definite by definition, each element Pμ,σ2Σ of
the set in the previous display is absolutely continuous with respect to (w.r.t.)
Lebesgue measure on Rn. In this setup we shall consider the problem of testing a linear hypothesis on
the parameter vector β ∈Rk, i.e., the problem of testing the null Rβ = r against
the alternative Rβ ̸= r, where R is a q × k matrix of rank q ≥1 and r ∈Rq. Define the affine space M0 = {μ ∈M : μ = Xβ and Rβ = r} and let M1 = M\M0 = {μ ∈M : μ = Xβ and Rβ ̸= r} . Adopting these definitions, the above testing problem can be written as H0 : μ ∈M0, 0 < σ2 < ∞, Σ ∈C
vs. H1 : μ ∈M1, 0 < σ2 < ∞, Σ ∈C, (4) where it is emphasized that the testing problem is a compound one. It is imme-
diately clear that size and power properties of tests in this setup depend in a
crucial way on the richness of the covariance model C. Before we close this section by introducing some further terminological and
notational conventions, we comment on how the fixed-design-assumption and
the Gaussianity-assumption above can be relaxed: 1. We remark that even though our setup assumes a non-stochastic design
matrix, the results immediately carry over to a setting where the data
generating processes of the design and the disturbances are independent
of each other. In such a setup, which covers many relevant scenarios, e.g.,
most simulation examples in our key references Andrews and Monahan
(1992), Newey and West (1994), and Rho and Shao (2013), our results
then deliver size and power properties conditional on the design, which,
in the tradition of conditional inference, might be considered as the more
relevant criterion, because X is observable (e.g., Robinson (1979)). Consider the linear regression model p
y
(1992), Newey and West (1994), and Rho and Shao (2013), our results
then deliver size and power properties conditional on the design, which,
in the tradition of conditional inference, might be considered as the more
relevant criterion, because X is observable (e.g., Robinson (1979)). ,
( g ,
(
))
2. The Gaussianity assumption might seem to be restrictive. However, as in
Section 5.5 of Preinerstorfer and P¨otscher (2016), we mention that the
negative results given in Section 4 of the present paper immediately ex-
tend in a trivial way without imposing the Gaussianity assumption on the
error vector U in (1), as long as the assumptions on the feasible error
distributions are weak enough to ensure that the implied set of distribu-
tions for Y contains the set in Equation (3), but possibly contains also
other distributions. Furthermore, by applying an invariance argument (ex-
plained in Preinerstorfer and P¨otscher (2016) Section 5.5) one can easily Finite sample properties of prewhitened F-type tests 2103 show that all statements about the null-behavior of the procedures under
consideration derived in the present paper carry over to the more general
distributional setup where U is assumed to be elliptically distributed. This
is to be understood as U having the same distribution as mσΣ1/2E, where
0 < σ < ∞, Σ ∈C, E is a random vector uniformly distributed on the
unit sphere Sn−1, and m is a random variable distributed independently
of E and which is positive with probability one. We next collect some further terminology and notation used throughout the
whole paper. A (non-randomized) test is the indicator function of a set W ∈
B(Rn), i.e., the corresponding rejection region. The size of such a test (rejection
region) is the supremum over all rejection probabilities under the null hypothesis
H0, i.e., sup
μ∈M0
sup
0<σ2<∞
sup
Σ∈C
Pμ,σ2Σ(W). sup
μ∈M0
sup
0<σ2<∞
sup
Σ∈C
Pμ,σ2Σ(W). sup
μ∈M0
sup
0<σ2<∞
sup
Σ∈C
Pμ,σ2Σ(W). Throughout the paper we let ˆβX(y) = (X′X)−1 X′y, where X is the design
matrix appearing in (1) and y ∈Rn. The corresponding ordinary least squares
(OLS) residual vector is denoted by ˆuX(y) = y −X ˆβX(y). The subscript X is
omitted whenever this does not cause confusion. Random vectors and random
variables are always written in bold capital and bold lower case letters, respec-
tively. Consider the linear regression model We use Pr as a generic symbol for a probability measure and denote
by E the corresponding expectation operator. Lebesgue measure on Rn will be
denoted by λRn. The Euclidean norm is denoted by ∥·∥, while d(x, A) denotes
the Euclidean distance of the point x ∈Rn to the set A ⊆Rn. For a vector x in
Euclidean space we define the symbol ⟨x⟩to denote ±x for x ̸= 0, the sign being
chosen in such a way that the first nonzero component of ⟨x⟩is positive, and we
set ⟨0⟩= 0. The j-th standard basis vector in Rn is denoted by ej(n). Let B′
denote the transpose of a matrix B and let span (B) denote the space spanned
by its columns. For a linear subspace L of Rn we let L⊥denote its orthogonal
complement and we let ΠL denote the orthogonal projection onto L. The set
of real matrices of dimension m × n is denoted by Rm×n. Lebesgue measure
on this set equipped with its Borel σ-algebra is denoted by λRm×n. We use the
convention that the adjoint of a 1×1 dimensional matrix D, i.e., adj(D), equals
one. Given a vector v ∈Rm the symbol diag(v) denotes the m × m diagonal
matrix with main diagonal v. We define X0 =
X ∈Rn×k : rank(X) = k
, i.e., the set of n × k design matrices of full rank, and whenever k ≥2 we define ˜X0 =
˜X ∈Rn×(k−1) : rank((e+, ˜X)) = k
, which is canonically identified (as a set) with the set of n × k design matrices of
full column rank the first column of which is the intercept e+ = (1, . . . , 1)′ ∈Rn. D. Preinerstorfer 2104 3. Tests based on prewhitened covariance estimators In the present section we formally describe the construction of tests based on
prewhitened covariance estimators. These tests (cf. Remark 3.4 below and the
discussion preceding it) reject for large values of a statistic T(y) =
(Rˆβ(y) −r)′ ˆΩ−1(y)(Rˆβ(y) −r)
if y /∈N ∗(ˆΩ),
0
else,
(5) (5) where where ˆΩ(y) = nR(X′X)−1 ˆΨ(y)(X′X)−1R′, and N ∗(ˆΩ) =
y ∈Rn : ˆΩ(y) is not invertible or not well defined
. N ∗(ˆΩ) =
y ∈Rn : ˆΩ(y) is not invertible or not well defined
. N ∗(ˆΩ) =
y ∈Rn : ˆΩ(y) is not invertible or not well defined
. The quantity ˆΨ appearing in the definition of ˆΩ above denotes a (VAR-) pre-
whitened nonparametric estimator of n−1E(X′UU′X) that incorporates a band-
width parameter which might depend on the data. Such an estimator is com-
pletely specified by three core ingredients: First, a kernel κ : R →R, i.e., an even
function satisfying κ(0) = 1, such as, e.g., the Bartlett or Parzen kernel; second,
a (non-negative) possibly data-dependent bandwidth parameter M; and third, a
deterministic prewhitening order p, i.e., an integer satisfying 1 ≤p ≤n/(k + 1)
(cf. Remark 3.2). Specific choices of M are discussed in detail in Section 3.1. All possible combinations of κ, M and p we analyze are specified in Assumption
1 of Section 3.2. Once these core ingredients have been chosen, one obtains a
prewhitened estimator ˆΨ, which is computed at an observation y following the
Steps (1) - (3) outlined subsequently (cf. also den Haan and Levin (1997)). We
here assume that the quantities involved (e.g., inverse matrices) are well defined,
cf. Remark 3.1 below, and follow the convention in the literature and leave the
estimator undefined at y else. Using this convention ˆΨ(y) is obtained as follows: 1. To prewhiten the data a VAR(p) model is fitted via ordinary least squares
to the columns of ˆV (y) = X′ diag(ˆu(y)). One so obtains the VAR(p) resid-
ual matrix ˆZ(y) ∈Rk×(n−p) with columns ˆZ·(j−p)(y) = ˆV·j(y) −
p
l=1
ˆA(p)
l
(y) ˆV·(j−l)(y)
for j = p + 1, . . . , n. The k × (kp)-dimensional VAR(p)-OLS estimator is given by The k × (kp)-dimensional VAR(p)-OLS estimator is given by ˆA(p)(y) =
ˆA(p)
1 (y), . . . , ˆA(p)
p (y)
= ˆVp(y) ˆV ′
0(y)
ˆV0(y) ˆV ′
0(y)
−1
,
where ˆVp(y) =
ˆV·(p+1)(y), . . . , ˆV·n(y)
∈Rk×(n−p) and the j-th column of
ˆV0(y) ∈Rkp×(n−p) equals
ˆV ′
·j+p−1(y), . . . , ˆV ′
·j+1(y), ˆV ′
·j(y)
′
∈Rkp for j =
1, . . . , n −p. In matrix form we clearly have ˆZ(y) = ˆVp(y) −ˆA(p)(y) ˆV0(y). ˆA(p)(y) =
ˆA(p)
1 (y), . . . , ˆA(p)
p (y)
= ˆVp(y) ˆV ′
0(y)
ˆV0(y) ˆV ′
0(y)
−1
, ˆA(p)(y) =
ˆA(p)
1 (y), . . . N ∗(ˆΩ) =
y ∈Rn : ˆΩ(y) is not invertible or not well defined
. , ˆA(p)
p (y)
= ˆVp(y) ˆV ′
0(y)
ˆV0(y) ˆV ′
0(y)
−1
,
where ˆVp(y) =
ˆV·(p+1)(y), . . . , ˆV·n(y)
∈Rk×(n−p) and the j-th column of
ˆV0(y) ∈Rkp×(n−p) equals
ˆV ′
·j+p−1(y), . . . , ˆV ′
·j+1(y), ˆV ′
·j(y)
′
∈Rkp for j =
1, . . . , n −p. In matrix form we clearly have ˆZ(y) = ˆVp(y) −ˆA(p)(y) ˆV0(y). Finite sample properties of prewhitened F-type tests
2105 2105 2. Then, one computes the quantities 2. Then, one computes the quantities ˇΓi(y) =
1
n−p
n−p
j=i+1 ˆZ·j(y) ˆZ′
·(j−i)(y)
if 0 ≤i ≤n −p −1,
ˇΓ′
−i(y)
if 0 < −i ≤n −p −1, and defines the preliminary estimate and defines the preliminary estimate and defines the preliminary estimate ˇΨ(y) =
n−p−1
i=−(n−p−1)
κ(i/M(y))ˇΓi(y), where in case M(y) = 0 one sets κ(i/M(y)) = 0 for i ̸= 0 and κ(i/M(y)) =
κ(0) for i = 0. where in case M(y) = 0 one sets κ(i/M(y)) = 0 for i ̸= 0 and κ(i/M(y)) =
κ(0) for i = 0. ( )
3. Finally, the preliminary estimate ˇΨ(y) is ‘recolored’ using the transforma-
tion ˆΨ(y) =
Ik −
p
l=1
ˆA(p)
l
(y)
−1
ˇΨ(y)
⎡
⎣
Ik −
p
l=1
ˆA(p)
l
(y)
−1⎤
⎦
′
. Remark 3.1. The construction of ˆΨ(y) outlined above clearly assumes that (i)
ˆA(p)(y) is well defined, which is equivalent to rank( ˆV0(y)) = kp; that (ii) M(y)
is well defined, which depends on the specific choice of M (cf. Section 3.1); and
that (iii) Ik −p
i=1 ˆA(p)
i
(y) is invertible. Remark 3.2. By assumption, all possible VAR orders p we consider must
satisfy p ≤n/(k + 1). This is done to rule out degenerate cases: for if p >
n/(k + 1), then rank( ˆV0(y)) < kp would follow because of ˆV0(y) ∈Rkp×(n−p). Hence the covariance estimator would nowhere be well defined for such a choice,
because (i) in Remark 3.1 would then clearly be violated at every observation
y. Remark 3.2. By assumption, all possible VAR orders p we consider must
satisfy p ≤n/(k + 1). This is done to rule out degenerate cases: for if p >
n/(k + 1), then rank( ˆV0(y)) < kp would follow because of ˆV0(y) ∈Rkp×(n−p). N ∗(ˆΩ) =
y ∈Rn : ˆΩ(y) is not invertible or not well defined
. Hence the covariance estimator would nowhere be well defined for such a choice,
because (i) in Remark 3.1 would then clearly be violated at every observation
y. Remark 3.3. In the present paper we focus on VAR prewhitening based on
the OLS estimator. This is in line with the original suggestions by Newey and
West (1994), as well as with Rho and Shao (2013). Alternatively, for p = 1,
Andrews and Monahan (1992) suggested to use an eigenvalue adjusted version
of the OLS estimator, the adjustment being applied if the matrix Ik −ˆA(1)
1 (y)
is close to being singular. We shall focus on the unadjusted OLS estimator for
the following reasons: Newey and West (1994) reported that the finite sample
properties show little sensitivity to this eigenvalue adjustment. Furthermore, it
is the unadjusted estimator that is often used in implementations of the method
suggested by Andrews and Monahan (1992) in software packages for statistical
and econometric computing (e.g., its implementation in the R (R Core Team
(2016)) package sandwich by Zeileis (2004), or its implementation in EViews,
e.g., Schwert (2009), p. 784.). We remark, however, that one can obtain a neg-
ative result similar to Theorem 4.2, and a positive result concerning an adjust-
ment procedure similar to Theorem 5.4, also for tests based on prewhitened
estimators with eigenvalue adjustment. Furthermore, we conjecture that it is Remark 3.3. In the present paper we focus on VAR prewhitening based on
the OLS estimator. This is in line with the original suggestions by Newey and
West (1994), as well as with Rho and Shao (2013). Alternatively, for p = 1,
Andrews and Monahan (1992) suggested to use an eigenvalue adjusted version
of the OLS estimator, the adjustment being applied if the matrix Ik −ˆA(1)
1 (y)
is close to being singular. We shall focus on the unadjusted OLS estimator for
the following reasons: Newey and West (1994) reported that the finite sample
properties show little sensitivity to this eigenvalue adjustment. Furthermore, it
is the unadjusted estimator that is often used in implementations of the method
suggested by Andrews and Monahan (1992) in software packages for statistical
and econometric computing (e.g., its implementation in the R (R Core Team
(2016)) package sandwich by Zeileis (2004), or its implementation in EViews,
e.g., Schwert (2009), p. 784.). N ∗(ˆΩ) =
y ∈Rn : ˆΩ(y) is not invertible or not well defined
. We remark, however, that one can obtain a neg-
ative result similar to Theorem 4.2, and a positive result concerning an adjust-
ment procedure similar to Theorem 5.4, also for tests based on prewhitened
estimators with eigenvalue adjustment. Furthermore, we conjecture that it is 2106 D. Preinerstorfer possible to prove (similar to Proposition 4.5) the genericity of such a negative
result, and to show that one can (similar to Proposition 5.5) generically resolve
this problem by using the adjustment procedure. We leave the question of which
estimator to choose for prewhitening to future research. In a typical asymptotic analysis of tests based on prewhitened covariance
estimators the event N ∗(ˆΩ) is asymptotically negligible (since ˆΩ converges to a
positive definite, or almost everywhere positive definite matrix). Hence there is
no need to be specific about the definition of the test statistic for y ∈N ∗(ˆΩ),
and one can work directly with the statistic y →(Rˆβ(y) −r)′ ˆΩ−1(y)(Rˆβ(y) −r),
(6) (6) which is left undefined for y ∈N ∗(ˆΩ). In a finite sample setup, however, one
has to think about the definition of the test statistic also for y ∈N ∗(ˆΩ). Our
decision to assign the value 0 to the test statistic for y ∈N ∗(ˆΩ) is of course
completely arbitrary. That this assignment does not affect our results at all is
discussed in detail in the following remark. Remark 3.4. Given that the estimator ˆΩ is based on a triple κ, M, p that
satisfies Assumption 1 introduced below (which is assumed in all of our main
results, and which is satisfied for covariance estimators using auxiliary AR(1)
models for the construction of the bandwidth parameter as considered in An-
drews and Monahan (1992), for covariance estimators as considered in Newey
and West (1994), and for covariance estimators as considered in Rho and Shao
(2013)), it follows from Lemma 3.10 that N ∗(ˆΩ) is either a λRn-null set, or that
it coincides with Rn. In the first case, which is generic under weak dimension-
ality constraints as shown in Lemma 3.11, the definition of the test statistic on
N ∗(ˆΩ) does hence not influence the rejection probabilities, because our model
is dominated by λRn (C contains only positive definite matrices). Therefore, size
and power properties are not affected by the definition of the test statistic for
y ∈N ∗(ˆΩ). N ∗(ˆΩ) =
y ∈Rn : ˆΩ(y) is not invertible or not well defined
. In the second case, i.e., if N ∗(ˆΩ) coincides with Rn, the statistic in
(6) is nowhere well defined, and hence, regardless of which value is assigned to
it for observations y ∈N ∗(ˆΩ), the resulting test statistic is constant, and thus
any test based on it breaks down trivially. 3.1. Bandwidth parameters In the following we describe bandwidth parameters M that are typically used in
Step 2 in the construction of the prewhitened estimator ˆΨ as discussed above:
The parametric approach (based on auxiliary AR(1) models) suggested by An-
drews (1991) and Andrews and Monahan (1992), the nonparametric approach
introduced by Newey and West (1994), and a data-independent approach which
was already investigated in Kiefer and Vogelsang (2005) in simulation studies
and which has recently been theoretically investigated by Rho and Shao (2013). Since the bandwidth parameter M is computed in Step 2 in the construction
of ˆΨ(y), we assume that κ, p and y are given and that Step 1 has already been Finite sample properties of prewhitened F-type tests 2107 successfully completed, i.e., all operations in Step 1 are well defined at y, in
particular ˆZ(y) is available for the construction of M. If not, we leave the band-
width parameter (and hence the covariance estimator) undefined at y. We also
implicitly assume that the quantities and operations appearing in the procedures
outlined subsequently are well defined and leave the bandwidth parameter (and
hence the covariance estimator) undefined else. A detailed structural analysis
of the subset of the sample space where a prewhitened estimator ˆΩ is well de-
fined is then later given in Lemma 3.9 in Section 3.3. Finally, we emphasize
that the bandwidth parameters discussed subsequently all require the choice of
additional tuning parameters. These tuning parameters are typically chosen in-
dependently of y and X, an assumption we shall maintain throughout the whole
paper (but see Remark 3.8 for some generalizations). 3.1.2. The non-parametric approach of Newey and West (1994) Let ω ∈Rk be as in Section 3.1.1 and let w(i) ≥0 for |i| = 0, . . . , n−p−1 be real
numbers such that w(0) = 1. For example, Newey and West (1994) suggested
to use rectangular weights, i.e., w∗(i) =
1
if |i| ≤⌊4(n/100)2/9⌋,
0
else, where ⌊.⌋denotes the floor function. Define for every |i| = 0, . . . n −p −1 ¯σi(y) = ω′ˇΓi(y)ω = (n −p)−1
n−p
j=|i|+1
ω′ ˆZ·j(y) ˆZ′
·(j−|i|)(y)ω. A bandwidth parameter is then obtained via MNW,ω,w,¯c(y) = ¯c2
⎛
⎜
⎝
⎡
⎣
n−p−1
i=−(n−p−1)
|i|¯c1w(i)¯σi(y)
n−p−1
i=−(n−p−1)
w(i)¯σi(y)
⎤
⎦
2
n
⎞
⎟
⎠
¯c3 where ¯c1 is a positive integer, where ¯c2 and ¯c3 are positive real numbers and
where ¯c = (¯c1, ¯c2, ¯c3). These numbers are constants independent of y and X
and have to be chosen by the user. The choice typically depends on the kernel
(for the specific choices we refer the reader to Newey and West (1994), Section
3). As in the previous section, we do not impose any assumptions beyond posi-
tivity (and independence of y and X) on the constants. Furthermore, we shall
denote by MNW the set of all bandwidth parameters that can be obtained as
special cases of the method in the present section, by appropriately choosing -
functionally independently of y and X - a weights vector, numbers w(i) ≥0 for
|i| = 0, . . . , n −p −1, ¯c1 a positive integer, ¯c2 > 0 and ¯c3 > 0. where ¯c1 is a positive integer, where ¯c2 and ¯c3 are positive real numbers and
where ¯c = (¯c1, ¯c2, ¯c3). These numbers are constants independent of y and X
and have to be chosen by the user. The choice typically depends on the kernel
(for the specific choices we refer the reader to Newey and West (1994), Section
3). As in the previous section, we do not impose any assumptions beyond posi-
tivity (and independence of y and X) on the constants. Furthermore, we shall
denote by MNW the set of all bandwidth parameters that can be obtained as
special cases of the method in the present section, by appropriately choosing -
functionally independently of y and X - a weights vector, numbers w(i) ≥0 for
|i| = 0, . . . 3.1.1. The parametric approach of Andrews and Monahan (1992) We shall denote by MAM the set of all bandwidth
parameters that can be obtained as special cases of the method in the present
section, by appropriately choosing - functionally independently of y and X - a
weights vector ω, constants c1 > 0, c2 > 0 and a j ∈{1, 2}. Remark 3.5. Since n, k and q are fixed quantities, the tuning parameters ω,
ci for i = 1, 2 and j might also depend on them, although we do not signify
this in our notation. A similar remark applies to the constants appearing in
Section 3.1.2 and in Section 3.1.3. Although we do not provide any details, we
furthermore remark that one can extend our analysis to bandwidth parameters
as above, but based on estimators other than ˆρi, e.g., all estimators satisfying
Assumption 4 of Preinerstorfer and P¨otscher (2016) such as the Yule-Walker
estimator or variants of the OLS estimator. 3.1.1. The parametric approach of Andrews and Monahan (1992) Let ω ∈Rk be such that ω ̸= 0 and ωi ≥0 for i = 1, . . . , k, i.e., ω is a
weights vector. Based on this weights vector the bandwidth parameter is now
obtained as follows: First, univariate AR(1) models are fitted via OLS to ˆZi·(y)
for i = 1, . . . , k, giving ˆρi(y) =
n−p
j=2
ˆZij(y) ˆZi(j−1)(y)
n−p−1
j=1
ˆZij(y)2
for i = 1, . . . , k,
ˆσ2
i (y) = (n −p −1)−1
n−p
j=2
ˆZij(y) −ˆρi(y) ˆZi(j−1)(y)
2
for i = 1, . . . , k, where we note that n −p −1 > 0 holds as a consequence of n > 2 and 1 ≤p ≤
n
k+1. Then, one calculates where we note that n −p −1 > 0 holds as a consequence of n > 2 and 1 ≤p ≤
n
k+1. Then, one calculates ˆα1(y) =
k
i=1
ωi
4ˆρ2
i (y)ˆσ4
i (y)
(1 −ˆρi(y))6(1 + ˆρi(y))2
k
i=1
ωi
ˆσ4
i (y)
(1 −ˆρi(y))4 ,
ˆα2(y) =
k
i=1
ωi
4ˆρi(y)2ˆσ4
i (y)
(1 −ˆρi(y))8
k
i=1
ωi
ˆσ4
i (y)
(1 −ˆρi(y))4 . Finally, bandwidth parameters are obtained via Finally, bandwidth parameters are obtained via MAM,j,ω,c(y) = c1 (ˆαj(y)n)c2
for
j = 1, 2, where to obtain a bandwidth parameter, one has to fix the constants c1 > 0,
c2 > 0 and j and where c = (c1, c2). Typically the choice of these constants
and the choice of j depends on certain characteristics of κ (for specific choices
see Andrews (1991), Section 6, in particular p. 834). For example, if κ is the
Bartlett kernel one uses c1 = 1.1447, c2 = 1/3 and j = 1, or if κ is the Quadratic-
Spectral kernel one would use c1 = 1.13221, c2 = 1/5 and j = 2. Since we do 2108 D. Preinerstorfer not need such a specific dependence to derive our theoretical results, we do not
impose any further assumptions on these constants beyond being positive (and
independent of y and X). 3.1.2. The non-parametric approach of Newey and West (1994) , n −p −1, ¯c1 a positive integer, ¯c2 > 0 and ¯c3 > 0. 2109 Finite sample properties of prewhitened F-type tests Remark 3.6. (i) The method described here is the ‘real-bandwidth’ approach
suggested in Newey and West (1994), as opposed to the ‘integer-bandwidth’
approach. In the latter approach one would use 1 + ⌊MNW,ω,w,¯c(y)⌋instead of
MNW,ω,w,¯c(y). Both approaches are asymptotically equivalent (Newey and West
(1994), Theorem 2) for most kernels (including the Bartlett kernel which is sug-
gested in Newey and West (1994)). Therefore, they are equally plausible in terms
of their theoretical foundation. For the sake of simplicity and comparability with
the bandwidth parameter as suggested by Andrews and Monahan (1992), which
is not an integer in general, we have chosen to focus on the ‘real-bandwidth’
approach. (ii) Newey and West (1994), p. 637, in principle also allow for ¯c1 = 0 (q = 0
in their notation) in the definition of their estimator. We do not allow for such a
choice. However, note that ¯c1 = 0 implies MNW,ω,w,¯c(y) ≡¯c2n¯c3. This is a data-
independent bandwidth parameter. These parameters are separately treated in
Section 3.1.3. 3.1.3. Data-independent bandwidth parameters Kiefer and Vogelsang (2005) and Rho and Shao (2013) studied properties of tests
based on prewhitened covariance estimators and data-independent bandwidth
parameters. Here one sets M ≡b(n −p) where b ∈(0, 1] is functionally inde-
pendent of y and X. For example, in Rho and Shao (2013) the choice b = 1 is
studied. These approaches all lead to bandwidth parameters MKV > 0, that are
functionally independent of both X and y. We denote the set of such bandwidth
parameters by MKV . 3.2. Assumptions on κ, M and p Different combinations of kernels κ, bandwidth parameters M and VAR orders
p obviously lead to different estimators. We indicate the dependence of the
estimator on these quantities by writing ˆΩκ,M,p. In the present paper we shall
consider estimators ˆΩκ,M,p based on a triple κ, M, p which satisfies the following
assumption: Assumption 1. The triple κ, M, p satisfies: 2. M ∈MAM ∪MNW ∪MKV . Assumption 1. The triple κ, M, p satisfies: 1. κ : R →R is an even function and κ(0) = 1. Furthermore, κ is continuous,
satisfies limx→∞κ(x) = 0, and for every real number s > 0 and every pos-
itive integer J the J × J symmetric Toeplitz matrix with ij-th coordinate
κ((i −j)/s) is positive definite. 1. κ : R →R is an even function and κ(0) = 1. Furthermore, κ is continuous,
satisfies limx→∞κ(x) = 0, and for every real number s > 0 and every pos-
itive integer J the J × J symmetric Toeplitz matrix with ij-th coordinate
κ((i −j)/s) is positive definite. 2. M ∈MAM ∪MNW ∪MKV . 3. p is an integer satisfying 1 ≤p ≤n/(k + 1). 3. p is an integer satisfying 1 ≤p ≤n/(k + 1). 3. p is an integer satisfying 1 ≤p ≤n/(k + 1). Remark 3.7. First, we remark that the positive definiteness assumption in Part
1 of Assumption 1 is natural in our context, because it guarantees that ˆΩκ,M,p
is nonnegative definite whenever it is well defined. Furthermore, it allows us to 2110 D. Preinerstorfer derive simple conditions for positive definiteness of ˆΩκ,M,p (cf. Lemma 3.10). It is well known that many kernels used in practice satisfy the positive defi-
niteness assumption, e.g., the Bartlett, Parzen, and Quadratic-Spectral kernel. Secondly, we note that in principle Assumption 1 does not prohibit a combi-
nation of MAM,1,ω,c ∈MAM with a second order kernel, or the combination of
MAM,2,ω,c ∈MAM with a first order kernel. It also allows for a combination of
elements of MNW with a prewhitening order p > 1 and for the combination of
elements of MKV with a kernel other than the Bartlett kernel. This goes well
beyond the original suggestions in Andrews and Monahan (1992), Newey and
West (1994) and Rho and Shao (2013), but we include these additional possi-
bilities for convenience. We also remark that since we assume throughout that
n > k, the set of VAR orders satisfying the third part of Assumption 1 always
includes the order p = 1. Remark 3.8 (Tuning parameters depending on the design). The tuning param-
eters used in the construction of M ∈MAM ∪MNW ∪MKV , e.g., the weights
vector ω used in the construction of M ∈MAM ∪MNW , are by definition func-
tionally independent of y and X. Assumption 1. The triple κ, M, p satisfies: Requiring that the tuning parameters are
independent of X is not a restriction in all results of the present paper in which
the design matrix X is fixed (i.e., Theorem 4.2, Proposition 5.2, and Theorem
5.4). To see this, suppose that a design matrix X as in (1) is given, that κ and
p satisfy the first and third part of Assumption 1, respectively, and that M is
constructed as in one of the Sections 3.1.1, 3.1.2, 3.1.3, but with a vector of tun-
ing parameters c∗(.), say, that is not constant on X0. The triple κ, M, p hence
does not satisfy Assumption 1. Let ˜
M be the bandwidth parameter that is ob-
tained from M by replacing the vector of tuning parameters c∗(.) by ˜c ≡c∗(X). Clearly, κ, ˜
M, p satisfies Assumption 1, and the test statistics as in Equation (5)
based on ˆΩκ, ˜
M,p and ˆΩκ,M,p, respectively, coincide for this specific X. Lemma 3.11. Assume that the triple κ, M, p satisfies Assumption 1. Then the
following holds: 3.3. Structural properties of prewhitened covariance estimators The study of finite sample properties of a test based on the statistic in Equation
(5) with ˆΩ = ˆΩκ,M,p requires a detailed understanding of definiteness properties
of the covariance estimator ˆΩκ,M,p, and of the structure of the set N ∗(ˆΩκ,M,p). Denoting the subset of the sample space Rn where ˆΩκ,M,p is not well defined by
N(ˆΩκ,M,p), we can write N ∗(ˆΩκ,M,p) = N(ˆΩκ,M,p) ∪
y ∈Rn\N(ˆΩκ,M,p) : det(ˆΩκ,M,p(y)) = 0
. As a first step we study N(ˆΩκ,M,p) in the subsequent lemma, where it is shown
that N(ˆΩκ,M,p) is algebraic. The lemma also characterizes the dependence of
N(ˆΩκ,M,p) on the design matrix, which will later be useful for obtaining our
genericity results. As a first step we study N(ˆΩκ,M,p) in the subsequent lemma, where it is shown
that N(ˆΩκ,M,p) is algebraic. The lemma also characterizes the dependence of
N(ˆΩκ,M,p) on the design matrix, which will later be useful for obtaining our
genericity results. Lemma 3.9. Assume that the triple κ, M, p satisfies Assumption 1. Then, N(ˆΩκ,M,p) = {y ∈Rn : gκ,M,p(y, X) = 0} , Finite sample properties of prewhitened F-type tests 2111 where gκ,M,p : Rn × Rn×k →R is a multivariate polynomial (explicitly con-
structed in the proof). As a consequence N(ˆΩκ,M,p) is an algebraic set. Further-
more, gκ,M,p does not depend on the hypothesis (R, r). where gκ,M,p : Rn × Rn×k →R is a multivariate polynomial (explicitly con-
structed in the proof). As a consequence N(ˆΩκ,M,p) is an algebraic set. Further-
more, gκ,M,p does not depend on the hypothesis (R, r). The subsequent lemma discusses definiteness and regularity properties of
ˆΩκ,M,p and shows that N ∗(ˆΩκ,M,p) is an algebraic subset of Rn. Again the
dependence of this algebraic set on the design is clarified. Given a prewhitening
order p satisfying Part 3 of Assumption 1, we define for every y ∈Rn such that
ˆA(p)(y) is well defined and such that Ik −p
l=1 ˆA(p)
l
(y) is invertible the matrix Bp(y) = R(X′X)−1
Ik −
p
l=1
ˆA(p)
l
(y)
−1
ˆZ(y). (7) (7) Lemma 3.10. Assume that the triple κ, M, p satisfies Assumption 1. Then the
following holds: 1. ˆΩκ,M,p(y) is nonnegative definite if and only if gκ,M,p(y, X) ̸= 0. ˆ 2. ˆΩκ,M,p(y) is singular if and only if gκ,M,p(y, X) ̸= 0 and rank(Bp(y)) <
ˆ 2. 3.3. Structural properties of prewhitened covariance estimators This notion of genericity is obviously
related to situations, where the data-generating process underlying the design
matrix X is assumed to be absolutely continuous w.r.t. λRn×k. In this situation,
a bandwidth parameter M ∈MAM ∪MNW would typically be based on the
weights vector ω = (1, . . . , 1)′ ∈Rk. As a specific result of independent interest
it is also shown that if k = 1 then g∗
κ,M,p(., e+, R) ̸≡0 holds, which means that
in the location model the set N ∗(ˆΩ
M ) is a λRn- null set (2) Under the assumption that k(p + 1) + p + 1MAM (M) ≤n holds, Part 2
establishes genericity of g∗
κ,M,p(., X, R) ̸≡0 in that it shows that the statement
holds for λRn×k- almost every X ∈X0. This notion of genericity is obviously
related to situations, where the data-generating process underlying the design
matrix X is assumed to be absolutely continuous w.r.t. λRn×k. In this situation,
a bandwidth parameter M ∈MAM ∪MNW would typically be based on the
weights vector ω = (1, . . . , 1)′ ∈Rk. As a specific result of independent interest
it is also shown that if k = 1 then g∗
κ,M,p(., e+, R) ̸≡0 holds, which means that
in the location model the set N ∗(ˆΩκ M p) is a λRn- null set. in the location model the set N ∗(ˆΩκ,M,p) is a λRn- null set. (
,
,p)
(3) Under the assumption that k ≥2 and k(p + 1) + p∗+ 1MAM (M) ≤n
holds, Part 3 establishes genericity of g∗
κ,M,p(., (e+, ˜X), R) ̸≡0 by showing that
the statement holds for λRn×(k−1) almost every ˜X ∈˜X0. This is a genericity
statement concerning design matrices the first column of which is the intercept. In contrast to (2) this notion of genericity is related to situations, where the first
column of the design matrix is fixed and the data-generating process underlying
the remaining columns is absolutely continuous w.r.t. λRn×(k−1). In such a setup
the construction of a bandwidth parameter M ∈MAM ∪MNW would typically
be based on the weights vector ω = (0, 1 . . . , 1)′ ∈Rk. 3.3. Structural properties of prewhitened covariance estimators ,
,p
(3) Under the assumption that k ≥2 and k(p + 1) + p∗+ 1MAM (M) ≤n
holds, Part 3 establishes genericity of g∗
κ,M,p(., (e+, ˜X), R) ̸≡0 by showing that
the statement holds for λRn×(k−1) almost every ˜X ∈˜X0. This is a genericity
statement concerning design matrices the first column of which is the intercept. In contrast to (2) this notion of genericity is related to situations, where the first
column of the design matrix is fixed and the data-generating process underlying
the remaining columns is absolutely continuous w.r.t. λRn×(k−1). In such a setup
the construction of a bandwidth parameter M ∈MAM ∪MNW would typically
be based on the weights vector ω = (0, 1 . . . , 1)′ ∈Rk. 3.3. Structural properties of prewhitened covariance estimators If k(p + 1) + p + 1MAM (M) ≤n, then
g∗
κ,M,p(., X, R) ̸≡0 for λRn×k-almost every X ∈X0 1. If n < k(p + 1) + p and q = k, then
g∗
κ,M,p(., X, R) ≡0 for every X ∈X0. 2. If k(p + 1) + p + 1MAM (M) ≤n, then 2. If k(p + 1) + p + 1MAM (M) ≤n, then g∗
κ,M,p(., X, R) ̸≡0 for λRn×k-almost every X ∈X0; g∗
κ,M,p(., X, R) ̸≡0 for λRn×k-almost every X ∈X0; if k = 1 we have in particular g∗
κ,M,p(., e+, R) ̸≡0. ,
,p
3. If k ≥2 and k(p + 1) + p∗+ 1MAM (M) ≤n, where p∗= p + (p mod 2),
then g∗
κ,M,p(., (e+, ˜X), R) ̸≡0 for λRn×(k−1)-almost every ˜X ∈˜X0. Remark 3.12. (1) Part 1 demonstrates that if n is too small in the sense that
n < k(p + 1) + p, then for every X ∈X0 the test statistic in Equation (5) with
ˆΩ = ˆΩκ,M,p vanishes identically if q = k holds, because the estimator ˆΩκ,M,p is
either not well defined or singular at every observation y. This shows that one
can in general not expect that N ∗(ˆΩκ,M,p) is generically a λRn-null set in case
n < k(p + 1) + p. (2) Under the assumption that k(p + 1) + p + 1MAM (M) ≤n holds, Part 2
establishes genericity of g∗
κ,M,p(., X, R) ̸≡0 in that it shows that the statement
holds for λRn×k- almost every X ∈X0. This notion of genericity is obviously
related to situations, where the data-generating process underlying the design
matrix X is assumed to be absolutely continuous w.r.t. λRn×k. In this situation,
a bandwidth parameter M ∈MAM ∪MNW would typically be based on the
weights vector ω = (1, . . . , 1)′ ∈Rk. As a specific result of independent interest
it is also shown that if k = 1 then g∗
κ,M,p(., e+, R) ̸≡0 holds, which means that
in the location model the set N ∗(ˆΩκ,M,p) is a λRn- null set. (2) Under the assumption that k(p + 1) + p + 1MAM (M) ≤n holds, Part 2
establishes genericity of g∗
κ,M,p(., X, R) ̸≡0 in that it shows that the statement
holds for λRn×k- almost every X ∈X0. 3.3. Structural properties of prewhitened covariance estimators ˆΩκ,M,p(y) is singular if and only if gκ,M,p(y, X) ̸= 0 and rank(Bp(y)) < q. 3
ˆΩ
(y) = 0 if and only if g
(y X) ̸= 0 and B (y) = 0 3. ˆΩκ,M,p(y) = 0 if and only if gκ,M,p(y, X) ̸= 0 and Bp(y) = 0. ˆ
ˆ 4. ˆΩκ,M,p(y) is positive definite if gκ,M,p(y, X) ̸= 0 and rank( ˆZ(y)) = k. 5. We have 4. ˆΩκ,M,p(y) is positive definite if gκ,M,p(y, X) ̸= 0 and rank( ˆZ(y)) = k. 5. We have N ∗(ˆΩκ,M,p) =
y ∈Rn : g∗
κ,M,p(y, X, R) = 0
, where g∗
κ,M,p : Rn×Rn×k×Rq×k →R is a multivariate polynomial (explic-
itly constructed in the proof). As a consequence N ∗(ˆΩκ,M,p) is an algebraic
set. Furthermore, g∗
κ,M,p is independent of r. It is a well known fact that an algebraic subset of Rn is either a closed λRn-
null set, or coincides with Rn (for a proof see, e.g., Okamoto (1973)). The latter
case occurs if and only if a (multivariate) polynomial defining the algebraic set
vanishes everywhere. Together with Part 5 of Lemma 3.10 this implies that
N ∗(ˆΩκ,M,p) is either a closed λRn-null set, or coincides with Rn, depending on
whether g∗
κ,M,p(., X, R) ̸≡0 or g∗
κ,M,p(., X, R) ≡0 holds, respectively. In the
latter case, every test based on the test statistic defined in Equation (5) with
ˆΩ = ˆΩκ,M,p trivially breaks down, because in this case the test statistic vanishes
identically on Rn. Obviously, studying size and power properties of tests based
on this test statistic in a sample of size n is only interesting, if we can guarantee
that g∗
κ,M,p(., X, R) ̸≡0 holds for a sufficiently large set of design matrices. That
this is indeed the case is the content of the subsequent lemma. More precisely
it is shown that g∗
κ,M,p(., X, R) ̸≡0 is generically satisfied whenever n exceeds
a certain threshold. It is also shown that the threshold we give can not be
substantially improved. The notion of genericity employed is further discussed
in Remark 3.12 following the lemma. 2112 D. Preinerstorfer 1. If n < k(p + 1) + p and q = k, then
g∗
κ,M,p(., X, R) ≡0 for every X ∈X0. 2. Assumption 2. CAR(1) ⊆C. Remark 4.1. Assumption 2 implies in particular that the singular boundary
of C ⊆Rn×n, i.e., the set of singular matrices in bd C, contains at least the two
elements e+e′
+ and e−e′
−, where e+ = (1, . . . , 1)′ and e−= (−1, 1, . . . , (−1)n)′. We note that these two singular matrices can be approximated by sequences
Λ(ρm) ∈C with ρm →1 and ρm →−1, respectively, where ρm ∈(−1, 1). Since the procedures we study in the present paper are geared towards situ-
ations such as C ⊇Cξ for some ξ ≥0 (cf. Remark 2.1), covariance models which
clearly satisfy the above assumption, Assumption 2 is mild in our context (cf. also the discussion in Section 3.2.2 of Preinerstorfer and P¨otscher (2016)). Under
this assumption and given a hypothesis (R, r), the subsequent theorem provides
four sufficient conditions on the design matrix under which a test based on a test
statistic as in Equation (5) with ˆΩ = ˆΩκ,M,p, together with an arbitrary (but
data-independent) critical value 0 < C < ∞, breaks down in terms of its finite
sample size and/or power properties. More precisely, Conditions (1) and (4) im-
ply that the test has size equal to one, Condition (3) implies that the test has
size not smaller than 1/2, and Condition (2) implies that the nuisance-minimal
rejection probability equals zero at every point μ1 ∈M1. Theorem 4.2. Suppose that the triple κ, M, p satisfies Assumption 1 and
that C satisfies Assumption 2. Let T be the test statistic defined in (5) with
ˆΩ = ˆΩκ,M,p. Let W(C) = {y ∈Rn : T(y) ≥C} be the rejection region, where C
is a real number satisfying 0 < C < ∞. Then the following holds: 1. Suppose g∗
κ,M,p(e+, X, R) ̸= 0 and T(e+ + μ∗
0) > C holds for some (and
hence all) μ∗
0 ∈M0, or g∗
κ,M,p(e−, X, R) ̸= 0 and T(e−+ μ∗
0) > C holds
for some (and hence all) μ∗
0 ∈M0. Then sup
Σ∈C
Pμ0,σ2Σ (W (C)) = 1 holds for every μ0 ∈M0 and every 0 < σ2 < ∞. In particular, the size of
the test is equal to one. holds for every μ0 ∈M0 and every 0 < σ2 < ∞. In particular, the size of
the test is equal to one. 2. 4. A negative result and its generic applicability In the first part of this section we obtain our main negative result concerning
finite sample properties of tests based on prewhitened nonparametric covariance
estimators. For this result to hold, we have to impose a richness assumption on
the covariance model C. Let CAR(1) denote the set of all correlation matrices
corresponding to stationary autoregressive processes of order one, i.e., CAR(1) =
{Λ(ρ) : ρ ∈(−1, 1)}, where Λ(ρ)ij = ρ|i−j| for 1 ≤i, j ≤n. The assumption is
as follows. 2113 Finite sample properties of prewhitened F-type tests Assumption 2. CAR(1) ⊆C. (ii) Obviously, the conclusions of the preceding theorem also apply to any
rejection region W ∗∈B(Rn) which differs from W(C) only by a λRn-null set. (iii) In Part 1 of the theorem the condition g∗
κ,M,p(e+, X, R) ̸= 0 (g∗
κ,M,p(e−,
X, R) ̸= 0) is superfluous, because it is already implicit in T(e+ + μ∗
0) > C > 0
(T(e−+μ∗
0) > C > 0), which is readily seen from the definition of T in Equation
(5). A similar comment applies to Part 3 of the theorem, where the condition
g∗
κ,M,p(e+, X, R) ̸= 0 (g∗
κ,M,p(e−, X, R) ̸= 0) is already implicit in T(e+ + μ∗
0) =
C > 0 (T(e−+ μ∗
0) = C > 0). The conditions are included for the sake of
comparability with Part 2 of the theorem. (iii) In Part 1 of the theorem the condition g∗
κ,M,p(e+, X, R) ̸= 0 (g∗
κ,M,p(e−,
X, R) ̸= 0) is superfluous, because it is already implicit in T(e+ + μ∗
0) > C > 0
(T(e−+μ∗
0) > C > 0), which is readily seen from the definition of T in Equation
(5). A similar comment applies to Part 3 of the theorem, where the condition
g∗
κ,M,p(e+, X, R) ̸= 0 (g∗
κ,M,p(e−, X, R) ̸= 0) is already implicit in T(e+ + μ∗
0) =
C > 0 (T(e−+ μ∗
0) = C > 0). The conditions are included for the sake of
comparability with Part 2 of the theorem. (iv) In case M ∈MKV , the assumption concerning the existence of the
gradient can be dropped in Part 3 of the theorem. This follows from Lemma
C.2 in Appendix C, where it is shown that if M ∈MKV , then the existence of
grad T(e+ +μ∗
0) and grad T(e−+μ∗
0) is already implied by g∗
κ,M,p(e+, X, R) ̸= 0
and g∗
κ,M,p(e−, X, R) ̸= 0, respectively. κ,M,p(
) ̸
(v) Throughout the theorem, Assumption 2 can be replaced by the weaker
assumption that there exist two sequences Λ(ρ(1)
m ) and Λ(ρ(2)
m ) of AR(1) corre-
lation matrices in C, such that ρ(1)
m →−1 and ρ(2)
m →1. Assumption 2. CAR(1) ⊆C. Suppose g∗
κ,M,p(e+, X, R) ̸= 0 and T(e+ + μ∗
0) < C holds for some (and
hence all) μ∗
0 ∈M0, or g∗
κ,M,p(e−, X, R) ̸= 0 and T(e−+ μ∗
0) < C holds
for some (and hence all) μ∗
0 ∈M0. Then inf
Σ∈C Pμ0,σ2Σ (W (C)) = 0 holds for every μ0 ∈M0 and every 0 < σ2 < ∞, and hence holds for every μ0 ∈M0 and every 0 < σ2 < ∞, and hence inf
μ1∈M1 inf
Σ∈C Pμ1,σ2Σ (W (C)) = 0 μ1∈M1 Σ∈C holds for every 0 < σ2 < ∞. In particular, the test is biased. Furthermore,
the nuisance-infimal rejection probability at every point μ1 ∈M1 is zero,
i.e., inf
0<σ2<∞inf
Σ∈C Pμ1,σ2Σ(W (C)) = 0. In particular, the infimal power of the test is equal to zero. D. Preinerstorfer 2114 3. Suppose g∗
κ,M,p(e+, X, R) ̸= 0, T(e+ +μ∗
0) = C and grad T(e+ +μ∗
0) exists
for some (and hence all) μ∗
0 ∈M0, or g∗
κ,M,p(e−, X, R) ̸= 0, T(e−+μ∗
0) =
C and grad T(e−+ μ∗
0) exists for some (and hence all) μ∗
0 ∈M0. Then sup
Σ∈C
Pμ0,σ2Σ (W (C)) ≥1/2 holds for every μ0 ∈M0 and every 0 < σ2 < ∞. In particular, the size of
the test is at least 1/2. holds for every μ0 ∈M0 and every 0 < σ2 < ∞. In particular, the size of
the test is at least 1/2. 4. Suppose that g∗
κ,M,p(., X, R) ̸≡0. Suppose further that e+ ∈M and
Rˆβ(e+) ̸= 0 holds, or e−∈M and Rˆβ(e−) ̸= 0 holds. Then sup
Σ∈C
Pμ0,σ2Σ (W (C)) = 1 holds for every μ0 ∈M0 and every 0 < σ2 < ∞. In particular, the size of
the test is equal to one. holds for every μ0 ∈M0 and every 0 < σ2 < ∞. In particular, the size of
the test is equal to one. holds for every μ0 ∈M0 and every 0 < σ2 < ∞. In particular, the size of
the test is equal to one. Remark 4.3. (i) Lemma C.1 in Appendix C shows that the rejection probabil-
ities Pμ,σ2Σ(W(C)) depend on (μ, σ2, Σ) only through (⟨(Rβ −r)/σ⟩, Σ), where
β is uniquely determined by Xβ = μ. Assumption 2. CAR(1) ⊆C. The concentration
turns out to be such that eventually the measures put roughly equal weight onto
the rejection region and onto its complement, resulting in rejection probabilities
as large as 1/2 under the null. We point out that the proof idea used to establish
Part 3 is inspired by the proof of Theorem 2.18 in Preinerstorfer and P¨otscher
(2017). The last part of the theorem considers the case where one of the vectors
e+ or e−is an element of M that is also ‘involved’ in the hypothesis. It is then
shown that the size of the test is one if the global condition g∗
κ,M,p(., X, R) ̸≡0
is satisfied. We recall that if this condition fails to hold, then the test T based
on ˆΩκ,M,p breaks down in a trivial way, because T is then zero everywhere. Therefore we see that under Assumption 2 one simply can not test a hypothesis
involving e+ ∈M or e−∈M by means of a test T based on ˆΩκ,M,p with κ, M,
p satisfying Assumption 1 (this in particular covers the location model where
X = e+, cf. also Lemma 3.11, Part 2). Pμ∗
0,σ2Σ →Pμ∗
0,σ2e−e′
−weakly as Σ →e−e′
−with Σ ∈C. These limiting measures
are absolutely continuous w.r.t. λμ∗
0+span(e+) and λμ∗
0+span(e−), respectively. As
a consequence we see that the mass of Pμ∗
0,σ2Σ concentrates on ‘neighborhoods’
of certain one-dimensional affine spaces as Σ approximates e+e′
+ or e−e′
−from
within C. From that it is intuitively clear that size and power properties crucially
depend on the behavior of the tests on ‘neighborhoods’ of these spaces. The first
and second part of the theorem provide sufficient conditions under which these
spaces are almost surely (w.r.t. λμ∗
0+span(e+) and λμ∗
0+span(e−)) contained in the
interior or exterior of the rejection region, respectively. The former case then
leads to size distortions, the latter to power deficiencies. The situation in the
third part of the theorem is quite different and more complex. In this case
the one-dimensional affine space supporting the respective limiting measure is
neither almost surely contained in the interior, nor almost surely contained in
the exterior of the rejection region. Rather it is almost surely contained in the
boundary of the rejection region. Assumption 2. CAR(1) ⊆C. In Parts 1 and 2 of
the theorem it is even enough to assume that there exist sequences Σ(i)
m ∈C for
i = 1, 2 with Σ(1)
m →e+e′
+ and Σ(2)
m →e−e′
−. Therefore, in these parts it is only
important that - and not how - these singular matrices can be approximated
from within C. We shall now provide some intuition for Theorem 4.2 (cf. also the discussion
preceding Theorem 5.7 in Preinerstorfer and P¨otscher (2016)). The repeated
appearance of the vectors e+ and e−in the theorem stems from the fact that
both e+e′
+ and e−e′
−are elements of the singular boundary of C ⊇CAR(1) (cf. Remark 4.1). Furthermore, for every μ∗
0 ∈M0 and every 0 < σ2 < ∞we have
that Pμ∗
0,σ2Σ →Pμ∗
0,σ2e+e′
+ weakly as Σ →e+e′
+ with Σ ∈C, and similarly that Finite sample properties of prewhitened F-type tests 2115 Pμ∗
0,σ2Σ →Pμ∗
0,σ2e−e′
−weakly as Σ →e−e′
−with Σ ∈C. These limiting measures
are absolutely continuous w.r.t. λμ∗
0+span(e+) and λμ∗
0+span(e−), respectively. As
a consequence we see that the mass of Pμ∗
0,σ2Σ concentrates on ‘neighborhoods’
of certain one-dimensional affine spaces as Σ approximates e+e′
+ or e−e′
−from
within C. From that it is intuitively clear that size and power properties crucially
depend on the behavior of the tests on ‘neighborhoods’ of these spaces. The first
and second part of the theorem provide sufficient conditions under which these
spaces are almost surely (w.r.t. λμ∗
0+span(e+) and λμ∗
0+span(e−)) contained in the
interior or exterior of the rejection region, respectively. The former case then
leads to size distortions, the latter to power deficiencies. The situation in the
third part of the theorem is quite different and more complex. In this case
the one-dimensional affine space supporting the respective limiting measure is
neither almost surely contained in the interior, nor almost surely contained in
the exterior of the rejection region. Rather it is almost surely contained in the
boundary of the rejection region. Therefore, in contrast to Parts 1 and 2, it is not
only important that the measures concentrate on the respective one-dimensional
space, but also how they concentrate (cf. Remark 4.3 (v)). Assumption 2. CAR(1) ⊆C. Preinerstorfer This statement covers (in particular) the important special case of testing a
restriction on the mean in a location model. We make the following observations
concerning this version of Theorem 4.2: • Since e−e′
−is not necessarily an element of the singular boundary of the co-
variance model considered here, the result just described does not contain
“size equal to one”- or “nuisance-minimal-power equal to zero”-statements
that arise from covariance matrices approaching e−e′
−. Note, however, that
the original Theorem 4.2 implies by a continuity argument that if ε is small
(compared to sample size), then considerable size distortions or power de-
ficiencies will nevertheless be present for covariance matrices in C that are
close to e−e′
−. • Consider the case where e+ ∈M, i.e., the regression contains an intercept,
and where the hypothesis does not involve the intercept, i.e., Rˆβ(e+) = 0:
Then we see that Parts 1-4 of the version of Theorem 4.2 just obtained do
not apply. In fact, in this case we can establish a positive result concerning
a test based on T with ˆΩ = ˆΩκ,M,p, and based on a non-standard critical
value that depends on ε. This positive result, together with its restrictions,
is discussed in Remark 5.3. Given a hypothesis (R, r) the four sufficient conditions provided in the preced-
ing theorem are conditions on the design matrix X. They depend on observable
quantities only. How these conditions can be checked is discussed in the subse-
quent paragraph: The first three parts of the theorem operate under the local
assumption that the multivariate polynomial g∗
κ,M,p(., X, R) does not vanish at
the point e+ or e−, respectively. The multivariate polynomial g∗
κ,M,p(., X, R) is
explicitly constructed in the proof of Lemma 3.10. Therefore, the condition that
it does not vanish at specific data points can readily be checked. Some addi-
tional conditions needed in Parts 1-3 of the theorem are formulated in terms
of T(e+ + μ∗
0) and T(e−+ μ∗
0), which are in fact independent of the specific
μ∗
0 ∈M0 chosen and therefore easy to calculate. Part 3 of the theorem requires
the existence of grad T(e+ + μ∗
0) or grad T(e−+ μ∗
0) (which is immaterial if
M ∈MKV as discussed in the preceding Remark). Again the existence of the
gradients is independent of the specific choice of μ∗
0 ∈M0. Assumption 2. CAR(1) ⊆C. Therefore, in contrast to Parts 1 and 2, it is not
only important that the measures concentrate on the respective one-dimensional
space, but also how they concentrate (cf. Remark 4.3 (v)). The concentration
turns out to be such that eventually the measures put roughly equal weight onto
the rejection region and onto its complement, resulting in rejection probabilities
as large as 1/2 under the null. We point out that the proof idea used to establish
Part 3 is inspired by the proof of Theorem 2.18 in Preinerstorfer and P¨otscher
(2017). The last part of the theorem considers the case where one of the vectors
e+ or e−is an element of M that is also ‘involved’ in the hypothesis. It is then
shown that the size of the test is one if the global condition g∗
κ,M,p(., X, R) ̸≡0
is satisfied. We recall that if this condition fails to hold, then the test T based
on ˆΩκ,M,p breaks down in a trivial way, because T is then zero everywhere. Therefore we see that under Assumption 2 one simply can not test a hypothesis
involving e+ ∈M or e−∈M by means of a test T based on ˆΩκ,M,p with κ, M,
p satisfying Assumption 1 (this in particular covers the location model where
X = e+, cf. also Lemma 3.11, Part 2). Remark 4.4. Suppose that it is known a priori that for some (fixed) ε ∈(0, 1]
the covariance model C does not contain AR(1) correlation matrices Λ(ρ) with
ρ ≤−1 + ε; i.e., instead of Assumption 2 the covariance model C satisfies CAR(1)(ε) = {Λ(ρ) : ρ ∈(−1 + ε, 1)} ⊆C. Inspection of the proof of Theorem 4.2 then shows that a version of Theorem
4.2 holds, in which all references to e−are deleted in Parts 1-4. For example,
Part 4 of this version of Theorem 4.2 reads as follows: “Suppose that g∗
κ,M,p(., X, R) ̸≡0. Suppose further that e+ ∈M and
Rˆβ(e+) ̸= 0 holds. Then Rˆβ(e+) ̸= 0 holds. Then sup
Σ∈C
Pμ0,σ2Σ (W (C)) = 1 holds for every μ0 ∈M0 and every 0 < σ2 < ∞. In particular, the
size of the test is equal to one.” holds for every μ0 ∈M0 and every 0 < σ2 < ∞. In particular, the
size of the test is equal to one.” 2116 D. Finite sample properties of prewhitened F-type tests Finite sample properties of prewhitened F-type tests The preceding theorem has given sufficient conditions on the design matrix,
under which the test considered breaks down in terms of its size and/or power
behavior. However, for a given hypothesis (R, r) there exist elements of X0 ⊆
Rn×k to which the theorem is not applicable. As a consequence, the question
remains to ‘how many’ elements of X0 the theorem can be applied once (R, r)
has been fixed. This question is studied subsequently. It is shown that generically
in the space of all design matrices at least one of the four conditions of Theorem
4.2 applies. The first part of the proposition establishes this genericity result
in the class of all design matrices of full column rank, i.e., X0. The remaining
parts establish the genericity result in case k ≥2 and the first column of X is
the intercept, i.e., X = (e+, ˜X) with ˜X ∈˜X0. Before we state the proposition,
we introduce two assumptions on the kernel κ. The first assumption is satisfied
by all kernels typically used in practice. Assumption 3. The kernel κ is continuously differentiable on the complement
of Δ(κ) ⊆R, a set consisting of finitely many elements. The second assumption, which is used in some statements of the second part
of the genericity result, imposes compactness of the support of the kernel. This
is satisfied by many kernels used in practice, e.g., the Bartlett kernel or the
Parzen kernel, but is not satisfied by the Quadratic-Spectral kernel. Assumption 4. The support of κ is compact. 2. Let k ≥2 and assume further that k(p + 1) + p∗+ 1MAM (M) ≤n, where
p∗= p + (p mod 2). Define Assumption 2. CAR(1) ⊆C. Sufficient conditions
for the existence of the gradient, under the assumption that κ is continuously
differentiable on the complement of a finite number of points, are provided in
Lemma C.2 in Appendix C. These conditions amount to checking whether or
not M(e+) or M(e−), respectively, is an element of a certain set determined
by κ consisting of finitely many points. In contrast to Parts 1-3, the fourth
part of the theorem operates under the global assumption that the multivari-
ate polynomial g∗
κ,M,p(., X, R) is not the zero polynomial. Since the polyno-
mial g∗
κ,M,p(., X, R) is explicitly constructed in the proof of Lemma 3.10, the
global assumption g∗
κ,M,p(., X, R) ̸≡0 can either be checked analytically, or by
using standard algorithms for polynomial identity testing. In addition to this
global assumption, the fourth part needs additional assumptions on the struc-
ture of M and the hypothesis (R, r) which can of course be easily checked by
the user. 2117 Assumption 4. The support of κ is compact. Preinerstorfer 2118 ˜X1 (e−) =
˜X ∈˜X0 : g∗
κ,M,p(e−, (e+, ˜X), R) = 0
, ˜X1 (e−) =
˜X ∈˜X0 : g∗
κ,M,p(e−, (e+, ˜X), R) = 0
,
˜X2 (e−) =
˜X ∈˜X0\˜X1 (e−) :
∄(grad T(e+, ˜
X)(.))|e−+μ∗
0,(e+, ˜
X)
and T(e+, ˜
X)(e−+ μ∗
0,(e+, ˜
X)) = C
. Then, ˜X1 (e−) is a λRn×(k−1)-null set. Furthermore, ˜X2 (e−) is a λRn×(k−1)-
null set under each of the following conditions: Then, ˜X1 (e−) is a λRn×(k−1)-null set. Furthermore, ˜X2 (e−) is a λRn×(k−1)-
null set under each of the following conditions: (a) M ∈MKV . (a) M ∈MKV . (a) M ∈MKV . (b) M ∈MAM and κ satisfies Assumptions 3 and 4. (c) M ∈MNW , p is odd, ωi > 0 for some i > 1 and κ satisfies Assump-
tions 3 and 4. (c) M ∈MNW , p is odd, ωi > 0 for some i > 1 and κ satisfies Assump-
tions 3 and 4. (d) κ satisfies Assumption 3 and ˜X →T(e+, ˜
X)(e−+ μ∗
0,(e+, ˜
X)) ̸≡C on
˜X0\˜X1 (e−). Suppose that the first column of R consists of zeros and that Assumption
2 holds. Then, the set of all matrices ˜X ∈˜X0 such that the first three
parts of Theorem 4.2 do not apply to the design matrix X = (e+, ˜X) is a
subset of ˜X1 (e−) ∪˜X2 (e−) and hence is a λRn×(k−1)-null set if one of the
conditions in (a)-(d) holds; it thus is a ‘negligible’ subset of ˜X0 in view of
the fact that ˜X0 differs from Rn×(k−1) only by a λRn×(k−1)-null set. Suppose that the first column of R consists of zeros and that Assumption
2 holds. Then, the set of all matrices ˜X ∈˜X0 such that the first three
parts of Theorem 4.2 do not apply to the design matrix X = (e+, ˜X) is a
subset of ˜X1 (e−) ∪˜X2 (e−) and hence is a λRn×(k−1)-null set if one of the
conditions in (a)-(d) holds; it thus is a ‘negligible’ subset of ˜X0 in view of
the fact that ˜X0 differs from Rn×(k−1) only by a λRn×(k−1)-null set. 3. Suppose k ≥2, that the first column of R is nonzero and that Assumption
2 holds. Assumption 4. The support of κ is compact. Assumption 4. The support of κ is compact. Assumption 4. The support of κ is compact. The genericity result is now as follows, where several quantities are equipped
with the additional subindex X to stress their dependence on the design matrix. Proposition 4.5. Fix a hypothesis (R, r) such that rank(R) = q. Let κ, M,
p satisfy Assumption 1. For X ∈X0 let TX be the test statistic defined in (5)
with ˆΩ = ˆΩκ,M,p,X and let μ∗
0,X ∈M0,X = {μ ∈span(X) : μ = Xβ, Rβ = r} be
arbitrary (the sets defined below do not depend on the choice of μ∗
0,X). Fix a
critical value C such that 0 < C < ∞. Then, the following holds. 1. Suppose that k(p + 1) + p + 1MAM (M) ≤n, define
X1 (e+) =
X ∈X0 : g∗
κ,M,p(e+, X, R) = 0
X2 (e+) =
X ∈X0\X1 (e+) :
∄(grad TX(.))|e++μ∗
0,X
and TX(e+ + μ∗
0,X) = C
, 1. Suppose that k(p + 1) + p + 1MAM (M) ≤n, define 1. Suppose that k(p + 1) + p + 1MAM (M) ≤n, define pose that k(p + 1) + p + 1MAM (M) ≤n, define
X1 (e+) =
X ∈X0 : g∗
κ,M,p(e+, X, R) = 0 X1 (e+) =
X ∈X0 : g∗
κ,M,p(e+, X, R) = 0 X (
)
X ∈X0\X1 (e+) :
∄(grad TX X2 (e+) =
X ∈X0\X1 (e+) :
∄(grad TX(.))|e++μ∗
0,X
and TX(e+ + μ∗
0,X) = C
, and similarly define X1 (e−) and X2 (e−). Then, X1 (e+) and X1 (e−) are
λRn×k-null sets. If M ∈MKV or if κ satisfies Assumption 3, then X2 (e+)
and X2 (e−) are λRn×k-null sets. If Assumption 2 holds, then the set of
all design matrices X ∈X0 for which the first three parts of Theorem 4.2
do not apply is a subset of (X1 (e+) ∪X2 (e+)) ∩(X1 (e−) ∪X2 (e−)) and
hence is a λRn×k-null set if M ∈MKV or if κ satisfies Assumption 3; it
thus is a ‘negligible’ subset of X0 in view of the fact that X0 differs from
Rn×k only by a λRn×k-null set. y
y
R
2. Let k ≥2 and assume further that k(p + 1) + p∗+ 1MAM (M) ≤n, where
p∗= p + (p mod 2). Define D. Assumption 4. The support of κ is compact. (a) In case M ∈MKV no additional condition is needed for establishing
generic applicability of Theorem 4.2. (b) If M ∈MAM, generic applicability of the negative result follows if the
kernel satisfies Assumptions 3 and 4, which applies to many kernels used
in practice, but not to the Quadratic-Spectral kernel which is emphasized
in Andrews and Monahan (1992). (b) If M ∈MAM, generic applicability of the negative result follows if the
kernel satisfies Assumptions 3 and 4, which applies to many kernels used
in practice, but not to the Quadratic-Spectral kernel which is emphasized
in Andrews and Monahan (1992). (c) In case M ∈MNW , the result shows that the procedure breaks down
generically if p is odd, ωi > 0 for some i > 1 and κ satisfies Assumptions 3
and 4. This seems to be restrictive. However, the recommended procedure
in Newey and West (1994) is obtained by choosing κ the Bartlett kernel,
p = 1 and ω = (0, 1, . . . , 1)′, because in Part 2 the first column of X is the
intercept. Therefore, we see that the recommended procedure in Newey
and West (1994) satisfies this condition. (c) In case M ∈MNW , the result shows that the procedure breaks down
generically if p is odd, ωi > 0 for some i > 1 and κ satisfies Assumptions 3
and 4. This seems to be restrictive. However, the recommended procedure
in Newey and West (1994) is obtained by choosing κ the Bartlett kernel,
p = 1 and ω = (0, 1, . . . , 1)′, because in Part 2 the first column of X is the
intercept. Therefore, we see that the recommended procedure in Newey
and West (1994) satisfies this condition. Summarizing, we see that the Conditions (a)-(c) in the proposition cover the
recommended choices of κ, M and p in Newey and West (1994) and Rho and
Shao (2013). For all procedures that are not covered by Conditions (a)-(c), e.g.,
the procedure in Andrews and Monahan (1992) based on the Quadratic-Spectral
kernel, one can typically obtain the genericity result by applying Condition (d),
which (under Assumption 3) is always satisfied apart from at most one excep-
tional critical value C∗. This is seen as follows: Clearly, Condition (d) depends
on the critical value C. Assumption 4. The support of κ is compact. We see that if Assumption 3 is satisfied, then (d) can be
violated for at most a single 0 < C∗< ∞. If this C∗happens to coincide with
C, the condition is not satisfied and we can not draw the desired conclusion for
this specific value of C. Moreover, we immediately see that the condition must
then be satisfied for any other choice C′, say. Therefore, generic applicability of
the negative result follows for any value C′ ̸= C in that case. This shows that
even if one chooses a triple κ, M, p that does not allow for an application of
(a)-(c), one can not expect to obtain a procedure that has good finite sample
size and power properties, because for all but at most one exceptional critical
value the corresponding test is guaranteed to break down generically. The third
part of the proposition considers the case where the first column of the design
matrix is the intercept, and where the coefficient corresponding to the intercept
is restricted by the hypothesis. In this case it follows that one can either apply
Part 4 of Theorem 4.2, or the test statistic is constant and hence the test breaks
down in a trivial way (cf. Remark 4.6). Assumption 4. The support of κ is compact. Then Theorem 4.2 (Part 4) applies to the design matrix X =
e+, ˜X
for every ˜X ∈˜X0 satisfying g∗
κ,M,p(., X, R) ̸≡0. Remark 4.6. (i) If n < k(p+1)+p and q = k holds, the first part of Lemma 3.11
shows that the test trivially breaks down, since for every element X of X0 the
test statistic TX is then constant on Rn. Therefore, the assumption on n in the
first two parts of the proposition can in general not be substantially improved. (ii) In the second part of the proposition, the analogously defined sets ˜X1 (e+)
and ˜X2 (e+) clearly satisfy ˜X1 (e+) = ˜X0 and ˜X2 (e+) = ∅. (
)
(
)
(
)
(iii) In the third part of the proposition, if X = (e+, ˜X) does not satisfy
g∗
κ,M,p(., X, R) ̸≡0, then the test breaks down in a trivial way, since TX is then
constant. (iii) In the third part of the proposition, if X = (e+, ˜X) does not satisfy
g∗
κ,M,p(., X, R) ̸≡0, then the test breaks down in a trivial way, since TX is then
constant. The first part of the preceding genericity result shows that if M ∈MKV , or
if the kernel satisfies Assumption 3, then Theorem 4.2 can be applied to generic
elements of X0, i.e., to all elements besides a λRn×k-null set. Since all kernels
used in practice, in particular the kernels emphasized in Andrews and Monahan
(1992) and Newey and West (1994), i.e., the Quadratic-Spectral kernel and the
Bartlett kernel, respectively, satisfy Assumption 3, this additional restriction on
κ is practically immaterial. The second part of the proposition considers the
situation where the first column of the design matrix is the intercept, which in
addition is assumed not to be involved in the hypothesis in the sense that the first
column of R is zero. In this situation it is shown that Theorem 4.2 can generically
be applied to design matrices of the form (e+, ˜X) with ˜X ∈˜X0, under certain
sets of conditions on the triple κ, M, p. We first discuss Conditions (a)-(c): Finite sample properties of prewhitened F-type tests 2119 (a) In case M ∈MKV no additional condition is needed for establishing
generic applicability of Theorem 4.2. (a) In case M ∈MKV no additional condition is needed for establishing
generic applicability of Theorem 4.2. 5. A positive result, an adjustment procedure and its generic
applicability As an example, this rules out the case where C is the correla-
tion model corresponding to all stationary autoregressive processes of order less
than or equal to two (cf. Lemma G.2 in Preinerstorfer and P¨otscher (2016)). If
this is not ruled out, however, further obstructions to good size and power prop-
erties can arise along suitable sequences approximating these boundary points
(cf. Section 3.2.3 in Preinerstorfer and P¨otscher (2016)). The possibility of es-
tablishing positive results in settings like that is beyond the scope of the present
paper and will be discussed elsewhere. (iii) The last part of the assumption states that elements of C that are ‘close’
to being singular can be well approximated by AR(1) correlation matrices. This,
together with CAR(1) being a subset of C, readily implies that the singular bound-
ary of C must coincide with
e+e′
+, e−e′
−
. Therefore, we see that the assump-
tion rules out the existence of rank deficient elements of bd(C) with rank strictly
greater than one. As an example, this rules out the case where C is the correla-
tion model corresponding to all stationary autoregressive processes of order less
than or equal to two (cf. Lemma G.2 in Preinerstorfer and P¨otscher (2016)). If
this is not ruled out, however, further obstructions to good size and power prop-
erties can arise along suitable sequences approximating these boundary points
(cf. Section 3.2.3 in Preinerstorfer and P¨otscher (2016)). The possibility of es-
tablishing positive results in settings like that is beyond the scope of the present
paper and will be discussed elsewhere. (iv) We note that Assumption 5 is clearly satisfied for every covariance model
of the form C = CAR(1)∪C♯, where C♯⊆Rn×n is a closed set consisting of positive
definite correlation matrices. As an example, let d ∈N be fixed and let CMA(d)
denote the set of all correlation matrices corresponding to stationary moving
average processes of an order not exceeding d, i.e., CMA(d) =
Σ(fα,δ) : α = (1, α1, . . . , αd)′ ∈Rd+1, δ > 0
, where Σ(fα,δ) denotes the n×n-dimensional correlation matrix corresponding to
the spectral density fα,δ(λ) = δ2
2π| d
j=0 αj exp(−ιλj)|2 (cf. Equation (2)). 5. A positive result, an adjustment procedure and its generic
applicability In the previous section we have established a (generically applicable) negative
result concerning tests based on (5) and a prewhitened covariance estimator
ˆΩκ,M,p. In the present section we first present a positive result concerning these
tests under a non-generic condition on the design matrix. Then we introduce
an adjustment procedure and establish a condition on the design matrix un-
der which the adjustment procedure leads to improved tests. Finally we prove 2120 D. Preinerstorfer that this condition holds generically in the set of all design matrices. Both,
the positive result concerning tests as in (5) based on a prewhitened covari-
ance estimator ˆΩκ,M,p, and the results concerning the adjustment procedure are
established under the following assumption on the covariance model C. Assumption 5. The set C ⊆Rn×n is norm-bounded and satisfies CAR(1) ⊆C. Furthermore, for every sequence Σm ∈C that converges to ¯Σ ∈bd(C) satisfying
rank(¯Σ) < n there exists a corresponding sequence ρm ∈(−1, 1) such that
Λ(ρm)−1/2ΣmΛ(ρm)−1/2 →In as m →∞. Remark 5.1. (i) We first note that Assumption 5 is stronger than Assumption
2. Therefore, under the former assumption the negative result established in
Section 4 concerning tests as in (5) based on a prewhitened covariance estimator
ˆΩκ,M,p does apply a fortiori. As a consequence, if C satisfies Assumption 5, then
positive results concerning size and power properties of tests of the form (5)
can only be established under non-generic assumptions on the design matrix. However, as we shall show, positive results can generically be established for an
adjusted version of such tests. (ii) Boundedness of C is typically satisfied in our setup, as it is always satisfied
if C consists only of correlation matrices. (ii) Boundedness of C is typically satisfied in our setup, as it is always satisfied
if C consists only of correlation matrices. (iii) The last part of the assumption states that elements of C that are ‘close’
to being singular can be well approximated by AR(1) correlation matrices. This,
together with CAR(1) being a subset of C, readily implies that the singular bound-
ary of C must coincide with
e+e′
+, e−e′
−
. Therefore, we see that the assump-
tion rules out the existence of rank deficient elements of bd(C) with rank strictly
greater than one. Finite sample properties of prewhitened F-type tests 2121 Finite sample properties of prewhitened F-type tests Under Assumption 5 we shall subsequently establish a positive result con-
cerning tests based on a test statistic T as in (5) with ˆΩ = ˆΩκ,M,p. In light of
Part (i) of the preceding remark we already know that such a positive result
can only be established under non-generic conditions on the design matrix. In
particular, the subsequent positive result considers the non-generic case where
- besides g∗
κ,M,p(., X, R) ̸≡0, a condition that is generically satisfied under a
mild constraint on n (cf. Lemma 3.11) - the column span of the design matrix
includes the vectors e+ and e−and where Rˆβ(e+) = Rˆβ(e−) = 0 holds. Proposition 5.2. Suppose that the triple κ, M, p satisfies Assumption 1, and
that C satisfies Assumption 5. Let T be the test statistic defined in Equation
(5) with ˆΩ = ˆΩκ,M,p. Let W(C) = {y ∈Rn : T(y) ≥C} be the rejection region,
where C is a real number satisfying 0 < C < ∞. Suppose further that e+, e−∈
M, Rˆβ(e+) = Rˆβ(e−) = 0 and g∗
κ,M,p(., X, R) ̸≡0. Then, the following holds: 1. The size of the rejection region W(C) is strictly less than 1, i.e., sup
μ0∈M0
sup
0<σ2<∞
sup
Σ∈C
Pμ0,σ2Σ (W(C)) < 1. sup
μ0∈M0
sup
0<σ2<∞
sup
Σ∈C
Pμ0,σ2Σ (W(C)) < 1. sup
μ0∈M0
sup
0<σ2<∞
sup
Σ∈C
Pμ0,σ2Σ (W(C)) < 1. Furthermore, inf
μ0∈M0
inf
0<σ2<∞inf
Σ∈C Pμ0,σ2Σ (W(C)) > 0. inf
μ0∈M0
inf
0<σ2<∞inf
Σ∈C Pμ0,σ2Σ (W(C)) > 0. 2. The infimal power is bounded away from zero, i.e., inf
μ1∈M1
inf
0<σ2<∞inf
Σ∈C Pμ1,σ2Σ(W(C)) > 0. 3. For every 0 < c < ∞ inf
μ1∈M1,0<σ2<∞
d(μ1,M0)/σ≥c
Pμ1,σ2Σm(W(C)) →1 μ1∈M1,0<σ2<∞ d(μ1,M0)/σ≥c holds for m →∞and for any sequence Σm ∈C satisfying Σm →¯Σ with
¯Σ a singular matrix. Furthermore, for every sequence 0 < cm < ∞ inf
μ1∈M1,
d(μ1,M0)≥cm
inf
Σ∈C∗Pμ1,σ2
mΣ(W(C)) →1 inf
Σ∈C∗Pμ1,σ2
mΣ(W(C)) →1 inf
Σ∈C∗Pμ1,σ2
mΣ(W(C)) →1 d(μ1,M0)≥cm holds for m →∞whenever 0 < σ2
m < ∞, cm/σm →∞, and C∗is a
(nonempty) closed subset of C. [The very last statement even holds if one
of the conditions e+, e−∈M and Rˆβ(e+) = Rˆβ(e−) = 0 is violated.]
4. For every δ, 0 < δ < 1, there exists a C(δ), 0 < C(δ) < ∞, such that sup
μ0∈M0
sup
0<σ2<∞
sup
Σ∈C
Pμ0,σ2Σ(W(C(δ))) ≤δ. 5. A positive result, an adjustment procedure and its generic
applicability Then
C = CAR(1) ∪cl(CMA(d)) satisfies Assumption 5, because every element of the
closure of CMA(d) is a positive definite correlation matrix (the latter statement
follows from Equation (2), compactness of the unit sphere in Rd+1, and the
Dominated Convergence Theorem). Finite sample properties of prewhitened F-type tests Under the maintained assumptions on the hypothesis and the design Proposi-
tion 5.2 shows that given any level of significance 0 < δ < 1, a critical value can 2122 D. Preinerstorfer be chosen in such a way that the test obtained holds its size, while its nuisance-
minimal power at every point μ1 in the alternative is bounded away from zero. As Theorem 4.2 in combination with Proposition 4.5 shows, this is impossible
for generic elements of the space of all design matrices. Additionally, Part 3 of
the proposition shows that the power approaches one in certain parts of the pa-
rameter space corresponding to the alternative hypothesis. These parts are char-
acterized by ∥(Rβ(1) −r)/σ∥being bounded away from zero and Σ →¯Σ with
¯Σ being singular, or ∥(Rβ(1) −r)/σ∥→∞and Σ →¯Σ with ¯Σ positive definite,
and where in both cases β(1) is the parameter vector corresponding to μ1 (note
that d(μ1, M0) is bounded from above and below by multiples of ∥Rβ(1) −r∥,
where the constants involved are positive and depend only on X, R and r). Remark 5.3. Suppose that instead of Assumption 5 it is known that the co-
variance model satisfies the following variant of Assumption 5 that rules out
AR(1) correlation matrices Λ(ρ) with ρ arbitrarily close to −1: The covariance model C ⊆Rn×n is norm-bounded and there exists
an ε ∈(0, 1] such that CAR(1)(ε) ⊆C (cf. Remark 4.4). Furthermore,
for every sequence Σm ∈C that converges to ¯Σ ∈bd(C) satisfying
rank(¯Σ) < n there exists a corresponding sequence ρm ∈(−1 + ε, 1)
such that Λ(ρm)−1/2ΣmΛ(ρm)−1/2 →In as m →∞. The covariance model C ⊆Rn×n is norm-bounded and there exists
an ε ∈(0, 1] such that CAR(1)(ε) ⊆C (cf. Remark 4.4). Furthermore,
for every sequence Σm ∈C that converges to ¯Σ ∈bd(C) satisfying
rank(¯Σ) < n there exists a corresponding sequence ρm ∈(−1 + ε, 1)
such that Λ(ρm)−1/2ΣmΛ(ρm)−1/2 →In as m →∞. Let T and W(C) be defined as in Proposition 5.2 above, and suppose further
that e+ ∈M, i.e., the regression contains an intercept, that Rˆβ(e+) = 0, i.e.,
the hypothesis does not involve the intercept, and that g∗
κ,M,p(., X, R) ̸≡0. Then one can show (using essentially the same argument as in the proof of
Proposition 5.2) that the Conclusions 1-4 of Proposition 5.2 hold in this setup. Finite sample properties of prewhitened F-type tests In particular, for any given δ ∈(0, 1) there exists a critical value C(δ) = C(δ, ε)
such that the test with rejection region W(C(δ, ε)) has size not greater than δ. This establishes a positive result concerning a test based on the test statistic T
with ˆΩ = ˆΩκ,M,p, and based on the non-standard critical value C(δ, ε), which
in practice can be obtained as explained in the discussion following Theorem
5.4. However, the test obtained critically depends on ε, which in practice will
typically be difficult to choose. If one uses a test with critical region W(C(δ, ε∗))
where ε∗> ε, then the size of this test might exceed δ. In light of this drawback,
it is important to stress that autocorrelation-robust-testing is possible without
imposing such an artificial condition that rules out AR(1) correlation matrices
Λ(ρ) with ρ arbitrarily close to −1: Theorems 5.4 and 5.5 in the present section
show that (at the small cost of including the artificial regressor e−) a generic
positive result can be obtained under the more natural Assumption 5. Theorem
5.4 furthermore shows that including the artificial regressor e−leads to good
power properties of the resulting test for covariance matrices close to Λ(−1). Proposition 5.2 assumes, among others, that X satisfies e+, e−∈span(X),
an assumption which is satisfied only by non-generic elements of the set of all
design matrices. In the following we shall now consider the (generic) situation
where e+, e−∈span(X) is violated. Proposition 5.2 is then clearly not applica-
ble. However, as we shall see, a positive result similar to Proposition 5.2 can be Finite sample properties of prewhitened F-type tests 2123 established for an adjusted test statistic. To explain how the adjustment proce-
dure works, suppose that we have a triple κ, M, p satisfying Assumption 1 which
we want to use for covariance estimation. Suppose further that 1 ≤p ≤
n
k+3
holds, which is typically satisfied. Finite sample properties of prewhitened F-type tests The following theorem now shows that under
certain conditions on the design matrix - which are shown to be generically satis-
fied in Proposition 5.5 below, and which in particular require e+, e−∈span(X)
to be violated - one can work with an adjusted test statistic that has improved
size and power properties, and which is constructed as follows: instead of basing
the construction of the test statistic on the true design matrix X and on R,
we first construct an artificial design matrix ¯X of full column rank satisfying
span( ¯X) = span(X, e+, e−) by adding the vectors e+ and/or e−to X. Fur-
thermore, we construct a corresponding matrix ¯R = (R, 0) by appending zero
vectors to R such that ¯R and ¯X have the same number of columns. Then we
construct a test statistic ¯T as in Equation (5), but with X and R replaced by ¯X
and ¯R, respectively, and where the covariance estimator is based on κ, M and p
as above besides some minor updates in the construction of M described below. The subsequent theorem shows that if e+, e−∈span(X) is violated, if every
e ∈{e+, e−} ∩span(X) satisfies Rˆβ(e) = 0, if rank( ¯X) < n, and if an assump-
tion on ¯X and ¯R analogous to the assumption g∗
κ,M,p(., X, R) ̸≡0 in Proposition
5.2 is satisfied, then for every critical value 0 < C < ∞the test with critical
region ¯W(C) =
y ∈Rn : ¯T(y) ≥C
has the same Properties (1)-(4) as W(C)
in Proposition 5.2. Theorem 5.4. Suppose that the triple κ, M, p satisfies Assumption 1, that p
additionally satisfies 1 ≤p ≤
n
k+3, and that C satisfies Assumption 5. Suppose
that one of the following (mutually exclusive) scenarios applies: 1. e+ ∈M with RˆβX(e+) = 0, e−/∈M and ¯k = k + 1 < n. Let ¯X = (X, e−)
and define ¯R = (R, 0) ∈Rqׯk. 2. e+ /∈M, e−∈M with RˆβX(e−) = 0 and ¯k = k + 1 < n. Let ¯X = (X, e+)
and define ¯R = (R, 0) ∈Rqׯk. 3. e+ /∈M, e−/∈M with rank (X, e+, e−) = k + 2 and ¯k = k + 2 < n. Let
¯X = (X, e+, e−) and define ¯R = (R, 0, 0) ∈Rqׯk. 4. Finite sample properties of prewhitened F-type tests e+ /∈M, e−/∈M with rank (X, e+, e−) = k + 1 and ¯k = k + 1 < n. Let
¯X = (X, e+) and define ¯R = (R, 0) ∈Rqׯk. 1. e+ ∈M with RˆβX(e+) = 0, e−/∈M and ¯k = k + 1 < n. Let ¯X = (X, e−)
and define ¯R = (R, 0) ∈Rqׯk. 1. e+ ∈M with RˆβX(e+) = 0, e−/∈M and ¯k = k + 1 < n. Let ¯X = (X, e−)
and define ¯R = (R, 0) ∈Rqׯk. 2. e+ /∈M, e−∈M with RˆβX(e−) = 0 and ¯k = k + 1 < n. Let ¯X = (X, e+)
and define ¯R = (R, 0) ∈Rqׯk. 3
e
/∈M e
/∈M with rank (X e
e )
k + 2 and ¯k
k + 2 < n Let 2. e+ /∈M, e−∈M with RˆβX(e−) = 0 and ¯k = k + 1 < n. Let ¯X = (X, e+)
and define ¯R = (R, 0) ∈Rqׯk. and define R = (R, 0) ∈Rq
. 3. e+ /∈M, e−/∈M with rank (X, e+, e−) = k + 2 and ¯k = k + 2 < n. Let
¯X = (X, e+, e−) and define ¯R = (R, 0, 0) ∈Rqׯk. 4. e+ /∈M, e−/∈M with rank (X, e+, e−) = k + 1 and ¯k = k + 1 < n. Let
¯X = (X, e+) and define ¯R = (R, 0) ∈Rqׯk. fi
( , )
3. e+ /∈M, e−/∈M with rank (X, e+, e−) = k + 2 and ¯k = k + 2 < n. Let
¯
(
)
¯
(
)
¯k (
)
3. e+ /∈M, e−/∈M with rank (X, e+, e−) = k + 2 and ¯k = k + 2 < n. Let
¯X = (X, e+, e−) and define ¯R = (R, 0, 0) ∈Rqׯk. 4. e+ /∈M, e−/∈M with rank (X, e+, e−) = k + 1 and ¯k = k + 1 < n. Let
¯X = (X, e+) and define ¯R = (R, 0) ∈Rqׯk. 3. e+ /∈M, e−/∈M with rank (X, e+, e−) = k + 2 and ¯k = k + 2 < n. Let
¯X = (X, e+, e−) and define ¯R = (R, 0, 0) ∈Rqׯk. 4. Finite sample properties of prewhitened F-type tests If g∗
κ, ¯
M,p(., ¯X, ¯R) ̸≡
0 (where g∗
κ, ¯
M,p is the function obtained from Lemma 3.10 applied to κ,
¯
M
and p and acting as if ¯X was the underlying design matrix and ( ¯R, r) was the
hypothesis to be tested), or equivalently if N ∗(ˆΩκ, ¯
M,p, ¯
X) ̸= Rn, then in each of
the four scenarios above the Conclusions (1)-(4) of Proposition 5.2 hold with
W(C) replaced by ¯W(C). (ω, 0)′ ∈R¯k; in case M ∈MNW we compute
¯
M as outlined in Section 3.1.2
(using as input ˆZ ¯
X(y) as opposed to ˆZX(y)), with the same constants ¯ci for
i = 1, 2, 3 and the same weights w as used in the construction of M, but with ω
replaced by ¯ω = (ω, 0)′ ∈R¯k. Let ¯W(C) =
y ∈Rn : ¯T(y) ≥C
be the rejection
region where C is a real number satisfying 0 < C < ∞. If g∗
κ, ¯
M,p(., ¯X, ¯R) ̸≡
0 (where g∗
κ, ¯
M,p is the function obtained from Lemma 3.10 applied to κ,
¯
M
and p and acting as if ¯X was the underlying design matrix and ( ¯R, r) was the
hypothesis to be tested), or equivalently if N ∗(ˆΩκ, ¯
M,p, ¯
X) ̸= Rn, then in each of
the four scenarios above the Conclusions (1)-(4) of Proposition 5.2 hold with
W(C) replaced by ¯W(C). The procedure outlined in the preceding theorem is based on an artificial de-
sign matrix ¯X which is obtained from X by adding either one or both elements
of the set {e+, e−} to the columns of X. If the so-obtained matrix ¯X satisfies
g∗
κ, ¯
M,p(., ¯X, ¯R) ̸≡0, then the results from Proposition 5.2 carry over to the re-
jection regions derived from ¯T. In particular the adjusted test statistic ¯T leads
to rejection regions the size of which is bounded away from one and such that
the nuisance minimal power is bounded away from zero. Besides these improve-
ments, the adjustment procedure is extremely convenient from a computational
perspective, as the adjusted test statistic ¯T does not require any additional im-
plementations. It is based on the same algorithm as the calculation of T, only
with a different design matrix. Finite sample properties of prewhitened F-type tests e+ /∈M, e−/∈M with rank (X, e+, e−) = k + 1 and ¯k = k + 1 < n. Let
¯X = (X, e+) and define ¯R = (R, 0) ∈Rqׯk. X = (X, e+, e−) and define R = (R, 0, 0) ∈R
. 4. e+ /∈M, e−/∈M with rank (X, e+, e−) = k + 1 and ¯k = k + 1 < n. Let
¯X = (X, e+) and define ¯R = (R, 0) ∈Rqׯk. Then in all cases ¯X is a matrix of full column rank. Define Then in all cases ¯X is a matrix of full column rank. Define ¯T(y) =
⎧
⎨
⎩
( ¯Rˆβ ¯
X(y) −r)′ ˆΩ−1
κ, ¯
M,p, ¯
X(y)( ¯R ˆβ ¯
X(y) −r)
if y /∈N ∗(ˆΩκ, ¯
M,p, ¯
X),
0
else, else, where ˆΩκ, ¯
M,p, ¯
X (y) is the estimator one would obtain following Steps 1-3 in
Section 3 if ¯X was the underlying design matrix, ( ¯R, r) was the hypothesis to be
tested and where ¯
M is defined as follows: in case M ∈MKV we set ¯
M ≡M;
in case M ∈MAM we compute ¯
M as outlined in Section 3.1.1 (using as input
ˆZ ¯
X(y) as opposed to ˆZX(y), and replacing k by ¯k), with the same constants c1
and c2 and j as used in the construction of M, but with ω replaced by ¯ω = 2124 D. Preinerstorfer (ω, 0)′ ∈R¯k; in case M ∈MNW we compute
¯
M as outlined in Section 3.1.2
(using as input ˆZ ¯
X(y) as opposed to ˆZX(y)), with the same constants ¯ci for
i = 1, 2, 3 and the same weights w as used in the construction of M, but with ω
replaced by ¯ω = (ω, 0)′ ∈R¯k. Let ¯W(C) =
y ∈Rn : ¯T(y) ≥C
be the rejection
region where C is a real number satisfying 0 < C < ∞. 6. Numerical results In this section we discuss numerical results to further illustrate our theoretical
findings. In both of the subsequent examples sample size n equals 100 and the
level of significance is 0.05. We study properties of the following tests (based
on the respective design matrix and hypothesis (R, r)): (i) the test that rejects
if (5) with κ the Quadratic-Spectral kernel, M the corresponding bandwidth
parameter based on auxiliary AR(1) models as suggested in Andrews and Mon-
ahan (1992) (cf. the discussion in Section 3.1.1), and prewhitening order p = 1
exceeds the χ2 critical value; (ii) the test that rejects if (5) with κ the Bartlett
kernel, M the corresponding bandwidth parameter as suggested in Newey and
West (1994) (cf. the discussion in Section 3.1.2), and prewhitening order p = 1
exceeds the χ2 critical value; (iii) the test that rejects if (5) with κ the Bartlett
kernel, data-independent bandwidth parameter M = n−p (cf. the discussion in
Section 3.1.3 and Rho and Shao (2013)), and prewhitening order p = 1 exceeds
the corresponding non-standard critical value (which was obtained from Table
1 in Kiefer and Vogelsang (2002)). In Example 2 we also illustrate the effect
of applying our adjustment procedure to the tests (i) - (iii) (the adjustment is
not applicable in the testing problem considered in Example 1). For the actual
computations we used the implementations of the tests (i) - (iii) (and of their
adjustments, which are computationally of the same structure, apart from the
computation of the critical values, cf. the discussion after Theorem 5.4) provided
by the R (R Core Team (2016)) packages lmtest (Zeileis and Hothorn (2002))
and sandwich (Zeileis (2004)). Finite sample properties of prewhitened F-type tests Then the following holds, where p∗= p +
(p mod 2). 1. If (k + 3)(p∗+ 2) + p −1 + 1MAM (M) ≤n, then for λRn×k-almost every
design matrix X ∈X0 ⊆Rn×k Scenario 3 in Theorem 5.4 applies, and
the Conclusions (1)-(4) of Proposition 5.2 hold for any critical value 0 <
C < ∞with W(C) replaced by ¯W(C) =
y ∈Rn : ¯T(y) ≥C
, where ¯T is
constructed as outlined in Theorem 5.4. 1. If (k + 3)(p∗+ 2) + p −1 + 1MAM (M) ≤n, then for λRn×k-almost every
design matrix X ∈X0 ⊆Rn×k Scenario 3 in Theorem 5.4 applies, and
the Conclusions (1)-(4) of Proposition 5.2 hold for any critical value 0 <
C < ∞with W(C) replaced by ¯W(C) =
y ∈Rn : ¯T(y) ≥C
, where ¯T is
constructed as outlined in Theorem 5.4. 2. Suppose that the first column of R is zero, that k ≥2 and assume that
(k + 2)(p∗+ 2) + p −1 + 1MAM (M) ≤n holds. Then for λRn×(k−1)-almost
every ˜X ∈˜X0 Scenario 1 of Theorem 5.4 applies to X = (e+, ˜X), and
the Conclusions (1)-(4) of Proposition 5.2 hold for any critical value 0 <
C < ∞with W(C) replaced by ¯W(C) =
y ∈Rn : ¯T(y) ≥C
, where ¯T is
constructed as outlined in Theorem 5.4. Finite sample properties of prewhitened F-type tests The theorem shows that for every level of significance 0 < δ < 1, there exists
a critical value C(δ) such that the rejection region ¯W(C(δ)) has size smaller
than δ. The critical value can be determined as follows: First of all, due to cer-
tain invariance properties of ¯T (cf. the proof of Theorem 5.4), the probabilities
Pμ0,σ2Σ( ¯W(C)) do not depend on μ0 and σ2. Hence, for any fixed 0 < C < ∞,
the maximal rejection probability under the null can be approximated numeri-
cally by simulating the rejection probabilities from a finite subset of C, and then
doing a grid search. In a second step C(δ) can be approximated by a line search
exploiting monotonicity of Pμ0,σ2Σ( ¯W(C)) in the critical value. The adjustment procedure described in Theorem 5.4 is applicable and yields
an improved test under the assumption that e+, e−∈span(X) is violated (and
hence the positive result in Proposition 5.2 concerning the unadjusted test does
not apply), that every e ∈{e+, e−} ∩span(X) satisfies Rˆβ(e) = 0, that ¯k < n
and that g∗
κ, ¯
M,p(., ¯X, ¯R) ̸≡0. Given a hypothesis (R, r) these are conditions on
the design matrix X. Our final result now shows (under mild constraints on n)
that these conditions are generically satisfied in the set of all design matrices
X0; and also in ˜X0, the set of all design matrices the first column of which is
the intercept, under the additional condition that the first column of R is zero. Under Assumption 5 we hence see that although rejection regions based on T
generically break down as a consequence of Proposition 4.5, this problem can
generically be resolved by using rejection regions based on the adjusted test
statistic ¯T, unless the regression includes an intercept and the first column of
R is nonzero. Finite sample properties of prewhitened F-type tests 2125 Proposition 5.5. Fix a hypothesis (R, r) with rank(R) = q, suppose that the
triple κ, M, p satisfies Assumption 1, that p additionally satisfies 1 ≤p ≤
n
k+3
and that C satisfies Assumption 5. Then the following holds, where p∗= p +
(p mod 2). Proposition 5.5. Fix a hypothesis (R, r) with rank(R) = q, suppose that the
triple κ, M, p satisfies Assumption 1, that p additionally satisfies 1 ≤p ≤
n
k+3
and that C satisfies Assumption 5. 6.1. Example 1 We also observe that while the tests (i) and (ii) show an almost identical
behavior in terms of their null rejection probabilities for ρ close to 1, the null
rejection probabilities of test (iii), although they eventually get much too large,
remain stable for a wider range of values of ρ. Tables 1, 2, and 3 in Appendix
E contain all simulated rejection probabilities. These tables also contain Monte
Carlo approximations of the rejection probabilities under the alternative (i.e.,
the simulation results for β1 > 0). X = e+, R = 1, r = 0, and covariance model C = CAR(1). We observe that the
triples κ, M, p used in the tests (i) - (iii) satisfy Assumption 1 (cf. Remark 3.7),
and that Part 2 of Lemma 3.11 can be used to verify the applicability of Part 4
of Theorem 4.2. This then shows that the size of each test under consideration
is 1. To illustrate this numerically, we obtained Monte Carlo approximations
of the rejection probabilities of the tests (i) - (iii) under Pe+β1,Λ(ρ), recall that
Λ(ρ) denotes the correlation matrix corresponding to a stationary AR(1) model
with parameter ρ, for ρ ∈{0, ±0.1, ±0.2, ±0.3, . . . , ±0.9, ±0.95, ±0.99, ±0.999,
±0.9999} and β1 ∈{0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2}, based on
2500 replications in each scenario. The results for β1 = 0, i.e., the null rejec-
tion probabilities illustrating the above quoted negative result, are shown in
Figure 1(a). As expected from the theoretical results (cf. also the explanations
after Theorem 4.2), the rejection probabilities are very large when the AR(1)
parameter ρ is close to 1. Furthermore, Figure 1(a) shows that for the tests
(i) and (ii) the rejection probabilities are clearly above 0.05 when ρ is close to
−1. We also observe that while the tests (i) and (ii) show an almost identical
behavior in terms of their null rejection probabilities for ρ close to 1, the null
rejection probabilities of test (iii), although they eventually get much too large,
remain stable for a wider range of values of ρ. Tables 1, 2, and 3 in Appendix
E contain all simulated rejection probabilities. These tables also contain Monte
Carlo approximations of the rejection probabilities under the alternative (i.e.,
the simulation results for β1 > 0). 6.1. Example 1 The first example concerns testing a zero restriction on the mean in a location
model with stationary and Gaussian AR(1) errors, i.e., testing problem (4) with D. Preinerstorfer 2126 Fig 1. (a) Example 1: Null rejection probabilities for the tests (i), (ii), and (iii), corresponding
to “Andrews-Monahan”, “Newey-West”, and “Rho-Shao”, respectively. (b) Example 2: Null
rejection probabilities for the tests (i), (ii), and (iii), corresponding to “Andrews-Monahan”,
“Newey-West”, and “Rho-Shao”, respectively. In both (a) and (b) the dashed black line cor-
responds to 0.05. Fig 1. (a) Example 1: Null rejection probabilities for the tests (i), (ii), and (iii), corresponding
to “Andrews-Monahan”, “Newey-West”, and “Rho-Shao”, respectively. (b) Example 2: Null
rejection probabilities for the tests (i), (ii), and (iii), corresponding to “Andrews-Monahan”,
“Newey-West”, and “Rho-Shao”, respectively. In both (a) and (b) the dashed black line cor-
responds to 0.05. X = e+, R = 1, r = 0, and covariance model C = CAR(1). We observe that the
triples κ, M, p used in the tests (i) - (iii) satisfy Assumption 1 (cf. Remark 3.7),
and that Part 2 of Lemma 3.11 can be used to verify the applicability of Part 4
of Theorem 4.2. This then shows that the size of each test under consideration
is 1. To illustrate this numerically, we obtained Monte Carlo approximations
of the rejection probabilities of the tests (i) - (iii) under Pe+β1,Λ(ρ), recall that
Λ(ρ) denotes the correlation matrix corresponding to a stationary AR(1) model
with parameter ρ, for ρ ∈{0, ±0.1, ±0.2, ±0.3, . . . , ±0.9, ±0.95, ±0.99, ±0.999,
±0.9999} and β1 ∈{0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2}, based on
2500 replications in each scenario. The results for β1 = 0, i.e., the null rejec-
tion probabilities illustrating the above quoted negative result, are shown in
Figure 1(a). As expected from the theoretical results (cf. also the explanations
after Theorem 4.2), the rejection probabilities are very large when the AR(1)
parameter ρ is close to 1. Furthermore, Figure 1(a) shows that for the tests
(i) and (ii) the rejection probabilities are clearly above 0.05 when ρ is close to
−1. 6.2. Example 2 Of course, in a specific application, where one only needs
to obtain a critical value for a single design matrix, the grid can be chosen much
finer. The null rejection probabilities of the adjusted versions of the tests (i)
- (iii) are shown in Figure 2(a). We see that the rejection probabilities of the
adjusted versions of the tests, although they are slightly above 0.05 for large
values of |ρ| (due to the coarseness of the grid used), are mostly between 0.03
and 0.04. Comparing these results to the null rejection probabilities of the un-
adjusted versions of the tests in Figure 1(b), the adjustment procedure, albeit
leading to rejection probabilities clearly below 0.05 in most scenarios considered,
leads to a substantial improvement concerning the null rejection probabilities. Power properties of the adjusted versions of the tests (i) - (iii) are presented
in Figures 2(b)-(d) in the form of contour plots. It can be seen that the three
adjusted tests are very similar concerning their power properties. Furthermore,
fixing ρ and treating β2 as a function argument, the power functions become
“flatter” for larger values of |ρ|. A complete summary of rejection probabilities
is provided in Tables 5, 7, and 9 in Appendix E. (e+, x) (the choice of x is outlined below), R = (0, 1), r = 0, and covari-
ance model C = CAR(1). We proceeded as follows: For each value of ρ ∈
{0, ±0.1, ±0.2, ±0.3, . . . , ±0.9, ±0.95, ±0.99, ±0.999, ±0.9999} a regressor x was
generated by drawing a random vector from the distribution of n consecutive
observations of a stationary Gaussian AR(1) process with parameter ρ. Then
rejection probabilities for the tests (i) - (iii) under Pxβ2,Λ(ρ) were obtained for
β2 ∈{0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2}. In each scenario, i.e., for
each combination of ρ and β2, the Monte Carlo approximations were based on
2500 replications. Since the regressor x was drawn randomly for each value of
ρ, the whole procedure was repeated 28 times, and the rejection probabilities
were then averaged. The average null rejection probabilities obtained for each
ρ ∈{0, ±0.1, ±0.2, ±0.3, . . . , ±0.9, ±0.95, ±0.99, ±0.999, ±0.9999} are given in
Figure 1(b). This figure clearly shows that the tests (i) - (iii) under considera-
tion do not control size, not even approximately. 6.2. Example 2 The null rejection probabilities
become very large for |ρ| close to 1. This effect is more pronounced for ρ close
to −1, which is due to the fact that an intercept is included in the model. A
complete summary of the rejection probabilities (which also contains rejection
probabilities under the alternative) can be found in Tables 4, 6 and 8 in Ap-
pendix E. In addition to the tests (i) - (iii) we also studied the behavior of their
adjusted versions (for the same set of x vectors), the adjusted versions being
obtained following the description in Part 1 of Theorem 5.4. The critical val-
ues were obtained by applying the grid-based Monte Carlo procedure explained
after Theorem 5.4, where we have chosen an equally spaced grid from −.99 to
.99 with 20 grid points, and 1000 replications for simulating the corresponding
rejection probabilities. Of course, in a specific application, where one only needs
to obtain a critical value for a single design matrix, the grid can be chosen much
finer. The null rejection probabilities of the adjusted versions of the tests (i)
- (iii) are shown in Figure 2(a). We see that the rejection probabilities of the
adjusted versions of the tests, although they are slightly above 0.05 for large
values of |ρ| (due to the coarseness of the grid used), are mostly between 0.03
and 0.04. Comparing these results to the null rejection probabilities of the un-
adjusted versions of the tests in Figure 1(b), the adjustment procedure, albeit
leading to rejection probabilities clearly below 0.05 in most scenarios considered,
leads to a substantial improvement concerning the null rejection probabilities. Power properties of the adjusted versions of the tests (i) - (iii) are presented
in Figures 2(b)-(d) in the form of contour plots. It can be seen that the three
adjusted tests are very similar concerning their power properties. Furthermore,
fixing ρ and treating β2 as a function argument, the power functions become
“flatter” for larger values of |ρ|. A complete summary of rejection probabilities
is provided in Tables 5, 7, and 9 in Appendix E. 6.2. Example 2 The second example concerns testing a zero restriction on the slope parame-
ter in a regression model with an intercept, one additional regressor, and with
stationary and Gaussian AR(1) errors, i.e., testing problem (4) with X = Finite sample properties of prewhitened F-type tests 2127 (e+, x) (the choice of x is outlined below), R = (0, 1), r = 0, and covari-
ance model C = CAR(1). We proceeded as follows: For each value of ρ ∈
{0, ±0.1, ±0.2, ±0.3, . . . , ±0.9, ±0.95, ±0.99, ±0.999, ±0.9999} a regressor x was
generated by drawing a random vector from the distribution of n consecutive
observations of a stationary Gaussian AR(1) process with parameter ρ. Then
rejection probabilities for the tests (i) - (iii) under Pxβ2,Λ(ρ) were obtained for
β2 ∈{0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2}. In each scenario, i.e., for
each combination of ρ and β2, the Monte Carlo approximations were based on
2500 replications. Since the regressor x was drawn randomly for each value of
ρ, the whole procedure was repeated 28 times, and the rejection probabilities
were then averaged. The average null rejection probabilities obtained for each
ρ ∈{0, ±0.1, ±0.2, ±0.3, . . . , ±0.9, ±0.95, ±0.99, ±0.999, ±0.9999} are given in
Figure 1(b). This figure clearly shows that the tests (i) - (iii) under considera-
tion do not control size, not even approximately. The null rejection probabilities
become very large for |ρ| close to 1. This effect is more pronounced for ρ close
to −1, which is due to the fact that an intercept is included in the model. A
complete summary of the rejection probabilities (which also contains rejection
probabilities under the alternative) can be found in Tables 4, 6 and 8 in Ap-
pendix E. In addition to the tests (i) - (iii) we also studied the behavior of their
adjusted versions (for the same set of x vectors), the adjusted versions being
obtained following the description in Part 1 of Theorem 5.4. The critical val-
ues were obtained by applying the grid-based Monte Carlo procedure explained
after Theorem 5.4, where we have chosen an equally spaced grid from −.99 to
.99 with 20 grid points, and 1000 replications for simulating the corresponding
rejection probabilities. 7. Conclusion We have shown that tests for (4) based on prewhitened covariance estimators
and possibly data-dependent bandwidth parameters break down in finite sam-
ples in terms of their size or power properties. This breakdown arises already for 2128 D. Preinerstorfer Fig 2. (a) Example 2: Null rejection probabilities for the adjusted versions of the tests (i),
(ii), and (iii), corresponding to “adj. Andrews-Monahan”, “adj. Newey-West”, and “adj. Rho-Shao”, respectively. The dashed black line corresponds to 0.05. (b) Example 2: Contour
plot of rejection probabilities of the adjusted version of test (i). (c) Example 2: Contour plot
of rejection probabilities of the adjusted version of test (ii). (d) Example 2: Contour plot of
rejection probabilities of the adjusted version of test (iii). Fig 2. (a) Example 2: Null rejection probabilities for the adjusted versions of the tests (i),
(ii), and (iii), corresponding to “adj. Andrews-Monahan”, “adj. Newey-West”, and “adj. Rho-Shao”, respectively. The dashed black line corresponds to 0.05. (b) Example 2: Contour
plot of rejection probabilities of the adjusted version of test (i). (c) Example 2: Contour plot
of rejection probabilities of the adjusted version of test (ii). (d) Example 2: Contour plot of
rejection probabilities of the adjusted version of test (iii). comparably simple covariance models such as C = CAR(1). We have also shown
how a simple adjustment procedure can generically solve this problem in many
cases. The test statistic obtained by applying the adjustment procedure is of
the same structural form as the test statistic based on estimators suggested by
Andrews and Monahan (1992) and Newey and West (1994) and the test statistic
in Rho and Shao (2013), but it is based on an artificial design matrix. There-
fore, the adjustment procedure does not only lead to improved size and power
properties, but is also convenient from a computational point of view. For the
adjustment procedure to work, Assumption 5 has to be satisfied, which requires
that elements of the covariance model C that are close to being singular are
well approximated by AR(1) correlation matrices. If and how the adjustment
procedure can be extended to settings where this approximation condition is
not satisfied is currently under investigation. Appendices Additional notation: For the sake of clarity we shall repeatedly stress the
dependence of ˆV , ˆV0, ˆVp, ˆZ, ˆA(p), ˆu, ˆΩκ,M,p and Bp on the design matrix X
by writing ˆVX, ˆV0,X, ˆVp,X, ˆZX, ˆA(p)
X , ˆuX, ˆΩκ,M,p,X and Bp,X in the following 2129 Finite sample properties of prewhitened F-type tests proofs. At various places we shall use the following notation: Given a matrix
M ∈Rm1×m2 and indices 1 ≤i ≤m1 and 1 ≤j ≤m2 we denote by [M]ij = Mij
the ij-th coordinate of M, by [M]·j = M·j the j −th column of M and by
[M]i· = Mi· the i −th row of M. In case m2 = 1 we write [M]i = Mi instead of
[M]i1. Appendix B: Proofs of results in Section 3.3 Proof of Lemma 3.9. Since X is a matrix of full column rank by assumption,
we clearly have det(X′X) ̸= 0. From the definition of ˆΩκ,M,p,X we see that
y ∈N(ˆΩκ,M,p,X), i.e., ˆΩκ,M,p,X(y) is not well defined, if and only if one of the
following conditions is satisfied (cf. Remark 3.1): (I) det
ˆV0,X(y) ˆV ′
0,X(y)
= 0;
(II) det
ˆV0,X(y) ˆV ′
0,X(y)
̸= 0 and det
Ik −p
l=1 ˆA(p)
l,X(y)
= 0;
(III) det
ˆV0,X(y) ˆV ′
0,X(y)
̸= 0 and det
Ik −p
l=1 ˆA(p)
l,X(y)
̸= 0 and
not well defined. (I) det
ˆV0,X(y) ˆV ′
0,X(y)
= 0; Using ˆuX(y) = (In −det(X′X)−1X adj(X′X)X′)y we see that the coordinates
of ¯V0,X(y) := det(X′X) ˆV0,X(y) and of ¯Vp,X(y) := det(X′X) ˆVp,X(y) are values
of certain multivariate polynomials defined on Rn ×Rn×k evaluated at the point
(y, X). Since (I) is equivalent to det(det(X′X)2 ˆV0,X(y) ˆV ′
0,X(y)) = det( ¯V0,X(y) ¯V ′
0,X(y)) = 0, this shows that (I) is equivalent to g1(y, X) = 0, say, where g1 : Rn × Rn×k →
R is a multivariate polynomial which is clearly independent of (R, r). Using
this equivalence, Condition (II) is seen to be equivalent to g1(y, X) ̸= 0 and
det
Ik −p
l=1 ˆA(p)
l,X(y)
= 0. Because of g1(y, X) ̸= 0 we have Ik −
p
l=1
ˆA(p)
l,X(y) =
Ik −ˆVp,X(y) ˆV ′
0,X(y)
ˆV0,X(y) ˆV ′
0,X(y)
−1
D(p), =
Ik −¯Vp,X(y) ¯V ′
0,X(y)
" ¯V0,X(y) ¯V ′
0,X(y)
#−1 D(p), =
Ik −det
" ¯V0,X(y) ¯V ′
0,X(y)
#−1 ¯Vp,X(y) ¯V ′
0,X(y) adj
" ¯V0,X(y) ¯V ′
0,X(y)
#
D(p),
where D(p) = (Ik, . . . , Ik)′ ∈Rkp×k. Using this together with similar arguments
as above we see that pre-multiplying Ik −p
l=1 ˆA(p)
l,X(y) by det
" ¯V0,X(y) ¯V ′
0,X(y)
#
results in a matrix, the entries of which are values of certain multivariate poly-
nomials, defined on Rn × Rn×k, evaluated at the point (y, X). Appendix B: Proofs of results in Section 3.3 It follows that
the second equation in (II) can be replaced by =
Ik −det
" ¯V0,X(y) ¯V ′
0,X(y)
#−1 ¯Vp,X(y) ¯V ′
0,X(y) adj
" ¯V0,X(y) ¯V ′
0,X(y)
#
D(p), g2(y, X) :=
$
det
" ¯V0,X(y) ¯V ′
0,X(y)
#%k det
Ik −
p
l=1
ˆA(p)
l,X(y)
= 0, 130
D Preinerstorfer 2130 D. Preinerstorfer where g2 : Rn ×Rn×k →R is a multivariate polynomial which is independent o
(R, r) either. Summarizing our observations concerning (I) and (II) we see tha where g2 : Rn ×Rn×k →R is a multivariate polynomial which is independent of
(R, r) either. Summarizing our observations concerning (I) and (II) we see that N(ˆΩκ,M,p,X) = {y ∈Rn : g1(y, X)g2(y, X) = 0} ∪{y ∈Rn : g1(y, X)g2(y, X) ̸= 0 and M(y) not w.d.} . The set in the second line of the previous display depends on the specific
bandwidth M. Hence, we have to distinguish three cases: Suppose first that
M ≡MKV ∈MKV , i.e., M is a constant which is functionally independent of
y, X and thus everywhere well defined on Rn. Define gκ,MKV ,p ≡g1g2, so that
gκ,MKV ,p : Rn × Rn×k →R is a multivariate polynomial. Noting that N(ˆΩκ,MKV ,p,X) = {y ∈Rn : gκ,MKV ,p(y, X) = 0} then proves the statement in case M ∈MKV , because gκ,MKV ,p is independent
of (R, r). Next we consider the case M = MAM,j,ω,c ∈MAM, and we start
with j = 1 (the case j = 2 is handled almost identically: omit the square in
the definition of D1 below, the squares from the expression in Equation (9),
and adapt the discussion following Equation (9) and the exponents of the pre-
multiplying factors accordingly). We shall drop the subindices j and c from
MAM,j,ω,c for notational convenience and write MAM,ω instead. We partition {y ∈Rn : g1(y, X)g2(y, X) ̸= 0 and MAM,ω(y) not w.d.} = D1 ∪D2,
(8) (8) where D1 and D2 are disjoint and defined as where D1 and D2 are disjoint and defined as D1 =
y ∈Rn : g1(y, X)g2(y, X) ̸= 0, ∃i∗: ˆρi∗(y) not w.d. or ˆρi∗(y)2 = 1
,
D2 =
y ∈Rn\D1 : g1(y, X)g2(y, X) ̸= 0, ∀i s.t. ωi ̸= 0 : ˆσ2
i (y) = 0
. The equality in (8) is readily seen from the definition of MAM,ω. Appendix B: Proofs of results in Section 3.3 We want to
obtain more suitable characterizations of D1 and D2 and proceed in two steps:
(i) First, we claim that y ∈D1 if and only if g1(y, X)g2(y, X) ̸= 0
and g1(y, X)g2(y, X) ̸= 0
and
k
&
i=1
⎛
⎜
⎝
⎡
⎣
n−p
j=2
[ ˆZX(y)]ij[ ˆZX(y)]i(j−1)
⎤
⎦
2
−
⎡
⎣
n−p−1
j=1
[ ˆZX(y)]2
ij
⎤
⎦
2⎞
⎟
⎠= 0. (9) (9) To see this assume that g1(y, X)g2(y, X) ̸= 0 holds: Suppose that ˆρi∗(y) is not
well defined. The latter occurs if and only if n−p−1
j=1
[ ˆZX(y)]2
i∗j = 0, i.e., all sum-
mands are zero, which immediately implies n−p
j=2 [ ˆZX(y)]i∗j[ ˆZX(y)]i∗(j−1) = 0. Therefore, the factor corresponding to index i∗vanishes and thus the prod-
uct defining the second equation in (9) vanishes. That ˆρ2
i∗(y) = 1 implies
that the product vanishes is obvious. To prove the other direction assume that
g1(y, X)g2(y, X) ̸= 0 and that the product vanishes. This implies that at least
one factor with index i∗, say, equals zero, which implies that either ˆρi∗(y) is Finite sample properties of prewhitened F-type tests 2131 not well defined or ˆρ2
i∗(y) = 1 holds. This proves the claim. Secondly, we recall
that if g1(y, X)g2(y, X) ̸= 0, then ˆZX(y) = ˆVp,X(y) −ˆA(p)
X (y) ˆV0,X(y). Using an
argument as above it is then easy to see that ˆZX(y) pre-multiplied by det
" ¯V0,X(y) ¯V ′
0,X(y)
#
det(X′X)
(10) (10) gives a matrix, the entries of which are values of certain multivariate polynomials
defined on Rn × Rn×k evaluated at the point (y, X). Thus, if we multiply the
second equation in (9) by the (4k)-th power of the expression in the previous
display we see that Equation (9) can equivalently be written as g1(y, X)g2(y, X) ̸= 0 and gAM,1(y, X) = 0, where gAM,1 : Rn × Rn×k →R is a multivariate polynomial that is independent
of (R, r). Summarizing, we have shown that D1 = {y ∈Rn : g1(y, X)g2(y, X) ̸= 0, gAM,1(y, X) = 0} . Appendix B: Proofs of results in Section 3.3 (ii) First we observe that y ∈D2 if and only if g1(y, X)g2(y, X)gAM,1(y, X) ̸= 0
and
k
i=1
ωi
n−p
j=2
[ ˆZX(y)]ij −ˆρi(y)[ ˆZX(y)]i(j−1)
2
= 0,
(11) (11) where we recall that by assumption ω is functionally independent of y and X. Because gAM,1(y, X) ̸= 0 implies that ˆρi(y) is well defined for i = 1, . . . , k, which
is equivalent to n−p−1
l=1
[ ˆZX(y)]2
il ̸= 0 for i = 1, . . . , k, the second equation in
the previous display can be replaced by where we recall that by assumption ω is functionally independent of y and X. Because gAM,1(y, X) ̸= 0 implies that ˆρi(y) is well defined for i = 1, . . . , k, which
is equivalent to n−p−1
l=1
[ ˆZX(y)]2
il ̸= 0 for i = 1, . . . , k, the second equation in
the previous display can be replaced by k
i=1
ωi
n−p
j=2
n−p−1
l=1
[ ˆZX(y)]2
il[ ˆZX(y)]ij −
n−p
l=2
[ ˆZX(y)]il[ ˆZX(y)]i(l−1)[ ˆZX(y)]i(j−1)
2
= 0. We now multiply the function defining this equation by the 6-th power of the
expression in Equation (10) and denote the resulting function by gAM,ω,2(y, X). The statement in Equation (11) is then seen to be equivalent to g1(y, X)g2(y, X)gAM,1(y, X) ̸= 0 and gAM,ω,2(y, X) = 0, where gAM,ω,2 : Rn × Rn×k →R is a multivariate polynomial. We also see that
gAM,ω,2 is independent of (R, r). We conclude that D2 = {y ∈Rn : g1(y, X)g2(y, X)gAM,1(y, X) ̸= 0, gAM,ω,2(y, X) = 0} . D2 = {y ∈Rn : g1(y, X)g2(y, X)gAM,1(y, X) ̸= 0, gAM,ω,2(y, X) = 0} . D2 = {y ∈Rn : g1(y, X)g2(y, X)gAM,1(y, X) ̸= 0, gAM,ω,2(y, X) = 0} . Now let gκ,MAM,ω,p ≡g1g2gAM,1gAM,ω,2. By what has been shown above
gκ,MAM,ω,p : Rn × Rn×k →R is a multivariate polynomial. Furthermore,
gκ,MAM,ω,p does not depend on (R, r). We observe that N(ˆΩκ,MAM,ω,p,X) =
y ∈Rn : gκ,MAM,ω,p(y, X) = 0
. 2132 D. Preinerstorfer This proves the lemma in case M ∈MAM. Finally, we consider M ≡MNW,ω,w ∈
MNW , where we write MNW,ω,w instead of MNW,ω,w,¯c, because the argument
and the resulting polynomial are independent of ¯c. We use a similar argument as
in the previous case. Appendix B: Proofs of results in Section 3.3 (14) Finite sample properties of prewhitened F-type tests
2133 Finite sample properties of prewhitened F-type tests
2133 2133 From Part 2 of the present lemma we know that we can rewrite the second set
to the right as From Part 2 of the present lemma we know that we can rewrite the second set
to the right as y ∈Rn\N(ˆΩκ,M,p,X) : det
$
Bp,X(y)B′
p,X(y)
%
= 0
. (15) (15) For every y ∈Rn\N(ˆΩκ,M,p,X) we have with D(p) = (Ik, . . . , Ik)′ ∈R(kp)×k
that (using the same notation as in the proof of Lemma 3.9) Bp,X(y) can be
written as R(X′X)−1
Ik −¯Vp,X(y) ¯V ′
0,X(y)
$ ¯V0,X(y) ¯V ′
0,X(y)
%−1 D(p)
−1
×
ˆVp,X(y) −¯Vp,X(y) ¯V ′
0,X(y)
$ ¯V0,X(y) ¯V ′
0,X(y)
%−1 ˆV0,X(y)
=R(X′X)−1 det(X′X)−1 =R(X′X)−1 det(X′X)−1 ×
"
det(
$ ¯V0,X(y) ¯V ′
0,X(y)
%
)Ik −¯Vp,X(y) ¯V ′
0,X(y) adj
$ ¯V0,X(y) ¯V ′
0,X(y)
%
D(p)
#−1
×
"
det
$ ¯V0,X(y) ¯V ′
0,X(y)
% ¯Vp,X(y) −¯Vp,X(y) ¯V ′
0,X(y) adj
$ ¯V0,X(y) ¯V ′
0,X(y)
% ¯V0,X(y)
# = det(X′X)−2 × det
"
det(
$ ¯V0,X(y) ¯V ′
0,X(y)
%
)Ik −¯Vp,X(y) ¯V ′
0,X(y) adj
$ ¯V0,X(y) ¯V ′
0,X(y)
%
D(p)
#−1
× R adj(X′X) × adj
"
det(
$ ¯V0,X(y) ¯V ′
0,X(y)
%
)Ik −¯Vp,X(y) ¯V ′
0,X(y) adj
$ ¯V0,X(y) ¯V ′
0,X(y)
%
D(p)
#
×
"
det
$ ¯V0,X(y) ¯V ′
0,X(y)
% ¯Vp,X(y) −¯Vp,X(y) ¯V ′
0,X(y) adj
$ ¯V0,X(y) ¯V ′
0,X(y)
% ¯V0,X(y)
# We therefore see that the coordinates of the matrix ¯Bp,X(y), say, which is ob-
tained by pre-multiplying Bp,X(y) by the factor Fp(y, X), defined as the product
of det(X′X)2 and det
"
det(
$ ¯V0,X(y) ¯V ′
0,X(y)
%
)Ik −¯Vp,X(y) ¯V ′
0,X(y) adj
$ ¯V0,X(y) ¯V ′
0,X(y)
%
D(p)
#
, (for later reference we note that Fp : Rn × Rn×k →R is a multivariate poly-
nomial), are values of certain multivariate polynomials defined on Rn × Rn×k
evaluated at (y, X). Furthermore, we can replace Bp,X(y) in Equation (15) by
¯Bp,X(y) without changing the set. This follows because y /∈N(ˆΩκ,M,p) implies Fp(y, X) = det(X′X)2 det( ¯V0,X(y) ¯V ′
0,X(y))k det
Ik −
p
l=1
ˆA(p)
l,X(y)
̸= 0. Appendix B: Proofs of results in Section 3.3 We observe that if g1(y, X)g2(y, X) ̸= 0, then the function
MNW,ω,w is not well defined if and only if n−p−1
i=−(n−p−1)
w(i)¯σi(y) = 0,
(12) (12) where where ¯σi(y) = (n −p)−1
n−p
j=|i|+1
ω′[ ˆZX(y)]·j
[ ˆZX(y)]·(j−|i|)
′
ω for |i| = 0, . . . n −p −1. Since ω and w are both functionally independent of X and y, we can pre-
multiply Equation (12) by the square of the expression in Equation (10) to see
that the statement g1(y, X)g2(y, X) ̸= 0 and MNW,ω,w not being well defined is
equivalent to g1(y, X)g2(y, X) ̸= 0 and gNW,ω,w(y, X) = 0, where gNW,ω,w : Rn × Rn×k →R is a multivariate polynomial. The function
gNW,ω,w is independent of (R, r). Using these properties, defining gκ,MNW,ω,w,p =
g1g2gNW,ω,w, a function which does not depend on (R, r), and noting that N(ˆΩκ,MNW,ω,w,p,X) =
y ∈Rn : gκ,MNW,ω,w,p(y, X) = 0
, then proves the claim in case M ∈MNW . Proof of Lemma 3.10. To establish Parts 1-4 of the lemma we apply a similar
argument as in the proof of Lemma 3.1 in Preinerstorfer and P¨otscher (2016). We observe that if y /∈N(ˆΩκ,M,p,X), or equivalently gκ,M,p(y, X) ̸= 0, we can
write ˆΩκ,M,p,X(y) as ˆΩκ,M,p,X(y) =
n
n −pBp,X(y)Wn−p(y)B′
p,X(y),
(13) (13) where Wn−p(y) ∈R(n−p)×(n−p) is the symmetric Toeplitz matrix with ones
on the main diagonal, and where for i ̸= j its ij-th coordinate is given by
κ((i −j)/M(y)) whenever M(y) ̸= 0, and by 0 else. Recall that M(y) ≥0. If M(y) = 0 we have Wn−p(y) = In. If M(y) > 0 the matrix Wn−p(y) is
positive definite by Assumption 1. Therefore, in both cases the matrix Wn−p(y)
is positive definite. This immediately establishes Parts 1-4, where rank(R) = q
is used in proving Part 4 (we emphasize that ˆΩκ,M,p,X(y) can be nonnegative
definite, singular, zero or positive definite only if it is well defined, i.e., only if
gκ,M,p(y, X) ̸= 0 holds). It remains to prove Part 5. We recall that N ∗(ˆΩκ,M,p,X) (14) =
N(ˆΩκ,M,p,X) ∪
y ∈Rn\N(ˆΩκ,M,p,X) : det
'
ˆΩκ,M,p,X(y)
(
= 0
. Appendix B: Proofs of results in Section 3.3 If we combine this equivalent expression for (15) with (14) and Lemma 3.9 we
obtain N ∗(ˆΩκ,M,p,X) =
y ∈Rn : gκ,M,p(y, X) det
$ ¯Bp,X(y) ¯B′
p,X(y)
%
= 0
. We next define g∗
κ,M,p(y, X, R) ≡gκ,M,p(y, X) det[ ¯Bp,X(y) ¯B′
p,X(y)]. By Lemma
3.9 we see that g∗
κ,M,p : Rn × Rn×k × Rq×k →R is a multivariate polynomial
that does not depend on r. 2134 D. Preinerstorfer The subsequent technical lemma plays a key role in several constructions in
the proofs of the genericity results. Lemma B.1. Let 1 ≤k < n, n > 2 and let (R, r) be a hypothesis. Suppose that
the triple κ, M, p satisfies Assumption 1. Assume that the tuple (y, X) ∈Rn×X0
satisfies for some t ≥k: (A1) ˆVX(y) has exactly t + 1 nonzero columns with indices 1 = j1 < j2 < . . . <
jt+1 ≤n. (A2) ji+1 −ji ≥p + 1 for i = 1, . . . , t, and n −jt+1 ≥p −1. (A3) If t = k, then rank( ˆVX(y)) = k. Otherwise, (A2) ji+1 −ji ≥p + 1 for i = 1, . . . , t, and n −jt+1 ≥p −1. ˆ (A3) If t = k, then rank( ˆVX(y)) = k. Otherwise, span(
[ ˆVX(y)]·ji : i = 1, . . . , t
)
=
span(
[ ˆVX(y)]·ji : i = 2, . . . , t + 1
) = Rk. span(
[ ˆVX(y)]·ji : i = 1, . . . , t
) Then, the following holds: Then, the following holds: 1. ˆA(p)
X (y) = 0 and rank( ˆZX(y)) = k. 1. ˆA(p)
X (y) = 0 and rank( ˆZX(y)) = k. 2. Under each of the following three conditions we have g∗
κ,M,p(y, X, R) ̸= 0
(or equivalently y /∈N ∗(ˆΩκ,M,p,X)): X ( )
(
( ))
2. Under each of the following three conditions we have g∗
κ,M,p(y, X, R) ̸=
(or equivalently y /∈N ∗(ˆΩκ,M,p,X)): (CKV) M ∈MKV ; (CAM) M ∈MAM, and every row vector of the matrix obtained from ˆZX(y)
by deleting its last column is nonzero [this is in particular satisfied if
n −jt+1 > p −1]; (CNW) M ∈MNW , and either each coordinate of ω′ ˆZX(y) is non-negative,
or each coordinate of ω′ ˆZX(y) is non-positive. 0 =
n
j=1
[ˆuX(y)]jX′
j· =
n
j=1
vj =
k+1
i=1
vji, Appendix B: Proofs of results in Section 3.3 Hence the claim follows. We next show that ˆA(p)
X (y) is well defined. For this
we have to verify that rank( ˆV0,X(y)) = kp (cf. Remark 3.1). The j-th column
(j = 1, . . . , n −p) of ˆV0,X(y) is given by By (A3) rank( ˆVX(y)) = k, which together with (A1) implies that span({vji : i =
1, . . . , k +1}) = Rk. Therefore, it follows from the two equations in the previous
display that {vji : i = 1, . . . , k} and {vji : i = 2, . . . , k + 1}, respectively, spans
Rk. Hence the claim follows. We next show that ˆA(p)
X (y) is well defined. For this
we have to verify that rank( ˆV0,X(y)) = kp (cf. Remark 3.1). The j-th column
(j = 1, . . . , n −p) of ˆV0,X(y) is given by (v′
j+p−1, . . . , v′
j+1, v′
j)′ ∈Rkp,
(16) (16) which is to be interpreted as vj if p = 1, as (v′
j+1, v′
j)′ if p = 2 etc. For l = 1, . . . , p
we define the (kp) × k dimensional auxiliary matrix Dl = el(p) ⊗Ik, where
el(p) denotes the l-th element of the canonical basis of Rp (and ⊗denotes the
Kronecker product). The following claims are immediate consequences of the
structure of ˆVX(y) implied by (A1) - (A2): (I) Dpvj1 is the first column of ˆV0,X(y); (I) Dpvj1 is the first column of ˆV0,X(y);
(II) If t ≥2, then Dlvji for i = 2, . . . , t and l = 1, . . . , p are columns of ˆV0,X(y);
(III) If p ≥2, then Dlvjt+1 for l = 1, . . . , (p −1) are columns of ˆV0,X(y). p j
,
(II) If t ≥2, then Dlvji for i = 2, . . . , t and l = 1, . . . , p are columns of ˆV0,X(y);
(III) If p ≥2, then Dlvjt+1 for l = 1, . . . , (p −1) are columns of ˆV0,X(y). To see Parts (I) and (II), we observe that (A1) and (A2) imply that for i =
1, . . . , t, there are at least p zero columns between the columns vji and vji+1 of
ˆVX(y). Appendix B: Proofs of results in Section 3.3 (CNW) M ∈MNW , and either each coordinate of ω′ ˆZX(y) is non-negative,
or each coordinate of ω′ ˆZX(y) is non-positive. 3. For every Q ∈Rk×k such that rank(Q) = k, the tuple (y, XQ) is an
element of Rn × X0 that satisfies (A1), (A2) and (A3). 4. If k ≥2 and either [ ˆVX(y)]1ji > 0 for i = 2, . . . , t+1 or [ ˆVX(y)]1ji < 0 for
i = 2, . . . , t + 1 holds, then there exists a regular matrix Q ∈Rk×k such
that the first columns of X and XQ, respectively, coincide and such that
g∗
κ,M,p(y, XQ, R) ̸= 0 (or equivalently y /∈N ∗(ˆΩκ,M,p,XQ)). Proof. Denote the column vectors of ˆVX(y) by vi for i = 1, . . . , n. If t > k, then
by (A3) the set {vj1, . . . , vjt} and
vj2, . . . , vjt+1
, respectively, spans Rk. Using
(A3), we now show that this is automatically satisfied in case t = k. To prove
this claim, we first recall that vj = [ˆuX(y)]jX′
j·. We see that ˆuX(y)⊥span(X)
implies Proof. Denote the column vectors of ˆVX(y) by vi for i = 1, . . . , n. If t > k, then
by (A3) the set {vj1, . . . , vjt} and
vj2, . . . , vjt+1
, respectively, spans Rk. Using
(A3), we now show that this is automatically satisfied in case t = k. To prove
this claim, we first recall that vj = [ˆuX(y)]jX′
j·. We see that ˆuX(y)⊥span(X)
implies 0 =
n
j=1
[ˆuX(y)]jX′
j· =
n
j=1
vj =
k+1
i=1
vji, Finite sample properties of prewhitened F-type tests 2135 where the third equality follows from (A1) (t = k). This shows that where the third equality follows from (A1) (t = k). This shows that vj1
=
−
k+1
i=2
vji
vjk+1
=
−
k
i=1
vji. By (A3) rank( ˆVX(y)) = k, which together with (A1) implies that span({vji : i =
1, . . . , k +1}) = Rk. Therefore, it follows from the two equations in the previous
display that {vji : i = 1, . . . , k} and {vji : i = 2, . . . , k + 1}, respectively, spans
Rk. Appendix B: Proofs of results in Section 3.3 Equation (16) together with j1 = 1 then immediately implies Parts (I)
and (II). Now we consider Part (III) and hence assume that p ≥2. We start with
the case l = (p−1). Every column of ˆVX(y) with index greater than jt+1 is zero
by Assumption (A1). By Assumption (A2) we have n −jt+1 ≥p −1. Together,
this implies that the column vjt+1 is followed by at least p−1 zero columns. Since
jt+1 −jt ≥p+1 by Assumption (A2), the column vjt+1 is preceded by at least p
zero columns. The assumption n−jt+1 ≥p−1 is equivalent to n−p ≥jt+1 −1. Hence, denoting the m1 × m2-dimensional zero matrix by 0m1,m2, we can use
Equation (16) with j = jt+1 −1 to see that (v′
jt+1+p−2, . . . , v′
jt+1, v′
jt+1−1)′ =
(01,(p−2)k, v′
jt+1, 01,k)′
if p > 2
(v′
jt+1, 01,k)′
if p = 2, is a column of ˆV0,X(y), where in deriving the equality we made use of the already
established fact that the column vjt+1 of ˆVX(y) is preceded by at least p > 1
zero columns (which implies that v′
jt+1−1 is the zero vector), and followed by at
least p−1 zero columns (which is used in case p > 2). This proves the statement 2136 D. Preinerstorfer concerning Dp−1vjt+1. In case p = 2 we are done. If p > 2, then the statements
concerning Dlvjt+1 for l = 1, . . . , p −2 follow from Equation (16) together with
the equation in the previous display and the fact that vjt+1 is preceded by at
least p zero columns. We now use (I)-(III) and span({vj1, . . . , vjt}) = span(
vj2, . . . , vjt+1
) = Rk
to show that rank( ˆV0,X(y)) = kp: The matrix ˆV0,X(y) is kp×(n−p) dimensional. Therefore, we must show that it has full row rank. Assume existence of a row
vector ξ = (ξ1, . . . , ξp), where ξ′
i ∈Rk for i = 1, . . . , p, such that ξ ˆV0,X(y) = 0
holds. Part (I) shows that 0 = ξDpvj1 = ξpvj1. If t ≥2, then Part (II) applied
with l = p shows that 0 = ξDpvji = ξpvji for i = 2, . . . , t. Appendix B: Proofs of results in Section 3.3 Summarizing the
cases t = 1 and t ≥2 we obtain ξpvji = 0 for i = 1, . . . , t. Because {vj1, . . . , vjt}
spans Rk, it follows that ξp = 0. If p ≥2, Part (III) implies that 0 = ξDlvjt+1 =
ξlvjt+1 for l = 1, . . . , (p −1). If t ≥2 Part (II) implies 0 = ξDlvji = ξlvji for
l = 1, . . . , (p −1) and i = 2, . . . , t. Summarizing again the cases t = 1 and t ≥2
we obtain that for l = 1, . . . , (p −1) we have ξlvji = 0 for i = 2, . . . , t + 1. Because
vj2, . . . , vjt+1
spans Rk, it follows from the previous display that
ξl = 0 for l = 1, . . . , p −1. Since we already know that ξp = 0, we obtain
ξ = 0 and thus rank( ˆV0,X(y)) = kp. Therefore ˆA(p)
X (y) is well defined. To see
that ˆA(p)
X (y) = 0 we observe that every nonzero column of ˆVX(y) besides the
first one is preceded by at least p zero columns. The matrix ˆVp,X(y) is obtained
from ˆVX(y) by deleting the first p ≥1 columns. This together with Equation
(16) immediately implies ˆVp,X(y) ˆV ′
0,X(y) = 0 and thus ˆA(p)
X (y) = 0. Before we
show that y /∈N ∗(ˆΩκ,M,p,X) under the conditions (CKV), (CAM) and (CNW),
respectively, we note that rank( ˆZX(y)) = k. This follows, because ˆA(p)
X (y) = 0
implies ˆZX(y) = ˆVp,X(y), which together with j2 −j1 ≥p + 1 shows that the
vectors vji for i = 2, . . . , t + 1 (which span Rk) are column vectors of ˆZX(y). ji
(
)
( )
Now, as a consequence of Part 4 of Lemma 3.10 positive definiteness of
ˆΩκ,M,p,X(y) and hence y /∈N ∗(ˆΩκ,M,p,X) follows if we can show that ˆΩκ,M,p,X(y)
is well defined. Since ˆA(p)
X (y) = 0 implies invertibility of Ik −p
l=1 ˆA(p)
i,X(y), it
remains to show that M is well defined at y (cf. Remark 3.1). This is trivially
satisfied under Condition (CKV) because M ∈MKV is everywhere well defined. n−p−1
j=1
[ ˆZX(y)]2
ij
for
i = 1, . . . , k, Appendix B: Proofs of results in Section 3.3 Suppose that Condition (CAM) holds, i.e., M ∈MAM and every row vector of
the matrix obtained from ˆZX(y) by deleting the last column is nonzero. That
the latter condition is satisfied if n −jt+1 > p −1 follows because in that case
the last column of ˆZX(y) is the zero vector and rank( ˆZX(y)) = k. Under the as-
sumption that every row vector of the matrix obtained from ˆZX(y) by deleting
the last column is nonzero it is obvious that the denominators in the definition
of ˆρi(y) for i = 1, . . . , k, i.e., n−p−1
j=1
[ ˆZX(y)]2
ij
for
i = 1, . . . , k, Finite sample properties of prewhitened F-type tests 2137 do not vanish. Therefore, ˆρi(y) for i = 1, . . . , k are well defined. Using Assump-
tions (A1) and (A2) together with p ≥1 and ˆZX(y) = ˆVp,X(y), it follows that
there is always at least one zero column between two nonzero columns of ˆZX(y). Therefore, it is clear that the numerators appearing in the definition of ˆρi(y)
for i = 1, . . . , k, i.e., n−p
j=2
[ ˆZX(y)]ij[ ˆZX(y)]i(j−1)
for
i = 1, . . . , k, must vanish. It follows that ˆρi(y) = 0 for i = 1, . . . , k. We finally show that
ˆσi(y) > 0 for i = 1, . . . , k. We note that ˆρi(y) = 0 for i = 1, . . . , k implies must vanish. It follows that ˆρi(y) = 0 for i = 1, . . . , k. We finally show that
ˆσi(y) > 0 for i = 1, . . . , k. We note that ˆρi(y) = 0 for i = 1, . . . , k implies ˆσi(y) = (n −p −1)−1
n−p
j=2
[ ˆZX(y)]2
ij for i = 1, . . . , k. Because of ˆZX(y) = ˆVp,X(y) it follows from Assumptions (A1) and (A2) that
the first column of ˆZX(y) must be zero. Furthermore, we already know that
rank( ˆZX(y)) = k. This implies that the matrix Z∗, say, which is obtained from
ˆZX(y) by deleting the first column, must be of full row rank k. Consequently
all rows of Z∗must be non-zero. The previous display shows that ˆσi(y) for
i = 1, . . . Appendix B: Proofs of results in Section 3.3 , k is, up to a positive factor, the squared Euclidean norm of the i-th
row of Z∗. Therefore ˆσi(y) > 0 for i = 1, . . . , k must hold. Therefore, we have
shown that M(y) is well defined (we even see that M(y) = 0 holds). Now we
consider the case where Condition (CNW) holds, i.e., M ∈MNW and every
coordinate of ω′ ˆZX(y) is non-negative (non-positive). We have to show that n−p−1
i=−(n−p−1)
w(i)¯σi(y) ̸= 0. n−p−1
i=−(n−p−1)
w(i)¯σi(y) ̸= 0. The non-negativity (non-positivity) condition immediately implies ¯σi(y) ≥0
for |i| = 0, 1, . . . , n −p −1. Furthermore, since rank( ˆZX(y)) = k and ω ̸= 0,
by the definition of the weights vector, we have ω′ ˆZX(y) ̸= 0. This implies
¯σ0(y) = (n −p)−1∥ω′ ˆZX(y)∥2 > 0. By assumption w(0) = 1 and w(i) ≥0 for
|i| = 1, . . . , n −p −1. Therefore, the quantity in the previous display does not
vanish and thus M(y) is well defined. This proves the second part of the lemma. ( )
We next prove Part 3. Let Q be a regular k × k dimensional matrix. First,
we obviously have XQ ∈X0, because X ∈X0 and Q is regular. Secondly, since
span(X) = span(XQ), using regularity of Q, we have that ˆuX(y) = ˆuXQ(y). This immediately entails ˆVXQ(y) = (XQ)′ diag(ˆuXQ(y)) = (XQ)′ diag(ˆuX(y)) = Q′X′ diag(ˆuX(y))
= Q′ ˆVX(y). Therefore, the tuple (y, XQ) ∈Rn × X0 satisfies (A1), (A2) and (A3), because
(y, X) does so and Q is regular. Therefore, the tuple (y, XQ) ∈Rn × X0 satisfies (A1), (A2) and (A3), because
(y, X) does so and Q is regular. 2138 D. Preinerstorfer It remains to prove Part 4. We do this by constructing a Q as in Part 3 such
that the tuple (y, XQ) ∈Rn×X0 satisfies Condition (CKV), (CAM) or (CNW),
respectively, if M ∈MKV , M ∈MAM or M ∈MNW , respectively. If M ∈MKV
we can obviously choose Q = Ik. Suppose that M /∈MKV . Let Q ∈Rk×k be
such that rank(Q) = k. We specify this matrix later on. Appendix B: Proofs of results in Section 3.3 From Part 2 of the
present lemma we see that the tuple (y, XQ) satisfies (A1), (A2) and (A3) and
therefore we can conclude from Part 1 of the present lemma that ˆA(p)
XQ(y) = 0,
which implies ˆZXQ(y) = ˆVp,XQ(y). Together with the equation in the previous
display, we see that ˆZXQ(y) = ˆVp,XQ(y) = Q′ ˆVp,X(y). (17)
se Q ∈Rk×k (regular) such that ˆZXQ(y) = ˆVp,XQ(y) = Q′ ˆVp,X(y). (17)
We now want to choose Q ∈Rk×k (regular) such that (17) (i) every row vector of the matrix obtained from ˆZXQ(y) by deleting the last
column is nonzero, and
ˆ (i) every row vector of the matrix obtained from ˆZXQ(y) by deleting the last
column is nonzero, and
ˆ (ii) either every coordinate of ω′ ˆZXQ(y) is non-negative, or every coordinate
of ω′ ˆZXQ(y) is non-positive, (ii) either every coordinate of ω′ ˆZXQ(y) is non-negative, or every coordinate
of ω′ ˆZXQ(y) is non-positive, holds, and that additionally the first columns of X and XQ, respectively, co-
incide. By assumption we either have [ ˆVX(y)]1ji > 0 for i = 2, . . . , t + 1, or we
have [ ˆVX(y)]1ji < 0 for i = 2, . . . , t + 1. Consider the former (latter) case: Let Q(γ) =
⎛
⎜
⎜
⎜
⎜
⎜
⎝
1
γ
γ
γ
. . . γ
0
1
0
0
. . . 0
0
0
1
0
. . . 0
... ... ... ... ... ... 0
0
0
0
. . . 1
⎞
⎟
⎟
⎟
⎟
⎟
⎠
, Q(γ) =
⎛
⎜
⎜
⎜
⎜
⎜
⎝
1
γ
γ
γ
. . . γ
0
1
0
0
. . . 0
0
0
1
0
. . . 0
... ... ... ... ... ... 0
0
0
0
. . . 1
⎞
⎟
⎟
⎟
⎟
⎟
⎠
, which is a regular matrix for every γ ∈R. Post-multiplying X by Q(γ) has the
same effect as adding γ-times the first column of X to all other columns, without
changing the first column. We observe that since ˆVp,X(y) is obtained from ˆVX(y)
by deleting the first p columns, the nonzero columns of ˆVp,X(y) are precisely
the vectors [ ˆVX(y)]·ji for i = 2, . . . Appendix B: Proofs of results in Section 3.3 , t + 1 because (y, X) satisfies Assumptions
(A1) and (A2). Therefore, it is obvious from Equation (17), together with the
assumed [ ˆVX(y)]1ji > 0 for i = 2, . . . , t + 1 ([ ˆVX(y)]1ji < 0 for i = 2, . . . , t +
1), that by choosing γ∗> 0 large enough, we can enforce that all nonzero
columns of ˆZXQ(γ∗)(y) are coordinate-wise positive (negative). Consider (i):
Using Equation (17) we see that Q(γ∗)′[ ˆVX(y)]·j2 is a column of the matrix
obtained by deleting the last column of ˆZXQ(γ∗)(y). This follows from j2 ≥
p + 2 (a consequence of the assumptions j2 −j1 ≥p + 1 and j1 = 1), which
shows that Q(γ∗)′[ ˆVX(y)]·j2 is a column of ˆZXQ(γ∗)(y), together with j2 < j3 ≤
n (a consequence of k ≥2), which shows that it is not the last column of
ˆZXQ(γ∗)(y). Since all coordinates of Q(γ∗)′[ ˆVX(y)]·j2 are positive (negative) by
construction of Q(γ∗), it follows that Q(γ∗) satisfies (i) above. To show (ii)
we recall that ω is nonzero and coordinate-wise nonnegative. Since all nonzero which is a regular matrix for every γ ∈R. Post-multiplying X by Q(γ) has the
same effect as adding γ-times the first column of X to all other columns, without
changing the first column. We observe that since ˆVp,X(y) is obtained from ˆVX(y)
by deleting the first p columns, the nonzero columns of ˆVp,X(y) are precisely
the vectors [ ˆVX(y)]·ji for i = 2, . . . , t + 1 because (y, X) satisfies Assumptions
(A1) and (A2). Therefore, it is obvious from Equation (17), together with the
assumed [ ˆVX(y)]1ji > 0 for i = 2, . . . , t + 1 ([ ˆVX(y)]1ji < 0 for i = 2, . . . , t +
1), that by choosing γ∗> 0 large enough, we can enforce that all nonzero
columns of ˆZXQ(γ∗)(y) are coordinate-wise positive (negative). Consider (i):
Using Equation (17) we see that Q(γ∗)′[ ˆVX(y)]·j2 is a column of the matrix
obtained by deleting the last column of ˆZXQ(γ∗)(y). Appendix B: Proofs of results in Section 3.3 This follows from j2 ≥
p + 2 (a consequence of the assumptions j2 −j1 ≥p + 1 and j1 = 1), which
shows that Q(γ∗)′[ ˆVX(y)]·j2 is a column of ˆZXQ(γ∗)(y), together with j2 < j3 ≤
n (a consequence of k ≥2), which shows that it is not the last column of
ˆZXQ(γ∗)(y). Since all coordinates of Q(γ∗)′[ ˆVX(y)]·j2 are positive (negative) by
construction of Q(γ∗), it follows that Q(γ∗) satisfies (i) above. To show (ii)
we recall that ω is nonzero and coordinate-wise nonnegative. Since all nonzero Finite sample properties of prewhitened F-type tests 2139 columns of ˆZXQ(γ∗)(y) are coordinate-wise positive (negative), it immediately
follows that every coordinate of ω′ ˆZXQ(γ∗)(y) is non-negative (non-positive). By construction the first columns of X and XQ(γ∗) coincide. This proves the
claim. Proof of Lemma 3.11. We start with the first part. Let X ∈X0 ⊆Rn×k and
y ∈Rn be arbitrary but fixed. We show that y ∈N ∗(ˆΩκ,M,p,X), which is equiv-
alent to g∗
κ,M,p(y, X, R) = 0 by Part 5 of Lemma 3.10. If y ∈N(ˆΩκ,M,p,X) ⊆
N ∗(ˆΩκ,M,p,X) we are done. Suppose y /∈N(ˆΩκ,M,p,X) which is equivalent to
gκ,M,p(y, X) ̸= 0 by Lemma 3.9. We claim that rank( ˆZX(y)) < k must hold. As-
suming this claim and using rank(R) = q = k, X ∈X0 which implies rank(X) =
k, and y /∈N(ˆΩκ,M,p,X) which implies rank(Ik −p
l=1 ˆA(p)
l
(y)) = k, it then fol-
lows from the definition of Bp,X(y) in Equation (7) that rank(Bp,X(y)) < q. As
a consequence, Part 2 of Lemma 3.10 then shows that ˆΩκ,M,p,X(y) is singular,
which implies y ∈N ∗(ˆΩκ,M,p,X). To prove rank( ˆZX(y)) < k we note that ˆZX(y) = ˆVp,X(y)
)
In−p −ˆV ′
0,X(y)
ˆV0,X(y) ˆV ′
0,X(y)
−1 ˆV0,X(y)
*
= ˆVp,X(y)Πspan( ˆV ′
0,X(y))⊥. We see from the previous display that rank( ˆZX(y)) = k, i.e., ˆZX(y) hav-
ing full row rank, is equivalent to rank( ˆVp,X(y)) = k and span( ˆV ′
0,X(y)) ∩
span( ˆV ′
p,X(y)) = {0}. Using rank( ˆV0,X(y)) = kp, a consequence of y
/∈
N(ˆΩκ,M,p,X) (cf. Remark 3.1), this implies rank
( ˆV ′
0,X(y) : ˆV ′
p,X(y))
= (p + 1)k. Appendix B: Proofs of results in Section 3.3 (
)
It remains to prove Part 3. Under the present assumptions it is shown in
Part 2 of Proposition 4.5 that for λRn×(k−1) almost every ˜X ∈˜X0 we have
g∗
κ,M,p(e−, (e+, ˜X), R) ̸= 0 and therefore g∗
κ,M,p(., (e+, ˜X), R) ̸≡0. It remains to prove Part 3. Under the present assumptions it is shown in
Part 2 of Proposition 4.5 that for λRn×(k−1) almost every ˜X ∈˜X0 we have
g∗
κ,M,p(e−, (e+, ˜X), R) ̸= 0 and therefore g∗
κ,M,p(., (e+, ˜X), R) ̸≡0. Appendix C: Proofs of results in Section 4 For a definition of the group G (M0) appearing in the following lemma we refer
the reader to Preinerstorfer and P¨otscher (2016) Section 5.1. For a definition of the group G (M0) appearing in the following lemma we refer
the reader to Preinerstorfer and P¨otscher (2016) Section 5.1. Lemma C.1. Assume that the triple κ, M, p satisfies Assumption 1. Assume
further that g∗
κ,M,p(., X, R) ̸≡0. Then, ˆβ and ˆΩκ,M,p satisfy Assumptions 5,
6, and 7 in Preinerstorfer and P¨otscher (2016) with N = N(ˆΩκ,M,p). In fact,
ˆΩκ,M,p (y) is nonnegative definite for every y ∈Rn\N(ˆΩκ,M,p), and is positive
definite λRn-almost everywhere. The test statistic T defined in Equation (5),
with ˆΩ = ˆΩκ,M,p, is invariant under the group G (M0) and the rejection prob-
abilities Pμ,σ2Σ(T ≥C) depend on
"
μ, σ2, Σ
#
∈M × (0, ∞) × C only through
((Rβ −r) /σ, Σ) (in fact, only through (⟨(Rβ −r) /σ⟩, Σ)), where β corresponds
to μ via μ = Xβ. Lemma C.1. Assume that the triple κ, M, p satisfies Assumption 1. Assume
further that g∗
κ,M,p(., X, R) ̸≡0. Then, ˆβ and ˆΩκ,M,p satisfy Assumptions 5,
6, and 7 in Preinerstorfer and P¨otscher (2016) with N = N(ˆΩκ,M,p). In fact,
ˆΩκ,M,p (y) is nonnegative definite for every y ∈Rn\N(ˆΩκ,M,p), and is positive
definite λRn-almost everywhere. The test statistic T defined in Equation (5),
with ˆΩ = ˆΩκ,M,p, is invariant under the group G (M0) and the rejection prob-
abilities Pμ,σ2Σ(T ≥C) depend on
"
μ, σ2, Σ
#
∈M × (0, ∞) × C only through
((Rβ −r) /σ, Σ) (in fact, only through (⟨(Rβ −r) /σ⟩, Σ)), where β corresponds
to μ via μ = Xβ. Proof of Lemma C.1. The assumption g∗
κ,M,p(., X, R) ̸≡0 together with Part 5
of Lemma 3.10 implies that the algebraic set N ∗(ˆΩκ,M,p) is a closed λRn-null
set. Therefore, by Lemma 3.9, it follows that the algebraic set N(ˆΩκ,M,p) ⊆
N ∗(ˆΩκ,M,p) is a closed λRn-null set as well. We claim that ˆΩκ,M,p is continuous
(and obviously well defined by definition) on Rn\N(ˆΩκ,M,p), because it can be
written as a composition of continuous functions on this set: We recall from
Equation (13) the representation ˆΩκ,M,p(y) =
n
n −pBp(y)Wn−p(y)B′
p(y) for every y ∈Rn\N(ˆΩκ,M,p). Appendix B: Proofs of results in Section 3.3 But this is impossible, because the matrix ( ˆV ′
0,X(y) : ˆV ′
p,X(y)) is (n −p) × ((p +
1)k) dimensional, which together with n < (p + 1)k + p implies rank( ˆV ′
0,X(y) :
ˆV ′
p,X(y)) ≤n −p < (p + 1)k. p,
Next, we prove Part 2 of the lemma. Under the present assumptions it is
shown in Part 1 of Proposition 4.5 that for λRn×k almost every X ∈X0 we have
g∗
κ,M,p(e+, X, R) ̸= 0 and therefore in particular g∗
κ,M,p(., X, R) ̸≡0. This proves
the first statement. To show the remaining statement we construct a y such that
g∗
κ,M,p(y, e+, R) ̸= 0. Note first that 2p + 1 + 1MAM (M) ≤n obviously implies
2p + 1 ≤n. Let y ∈Rn satisfy y1 = −1, yp+2 = 1 and yi = 0 else. [Note that
this is feasible, i.e., p+2 ≤n holds, because of 2p+1 ≤n and p ≥1.] We intend
to apply Part 2 of Lemma B.1 with t = k = 1 to the tuple (y, e+). We first have
to show that the tuple (y, e+), which is clearly an element of Rn × X0, satisfies
Assumptions (A1), (A2) and (A3). For this we observe that e+⊥y, which implies
ˆue+(y) = y and therefore ˆVe+(y) = ˆu′
e+(y) = y′. 2140 D. Preinerstorfer Hence (A1) is satisfied, because y1 = −1 ̸= 0 and y has only two nonzero
coordinates. The corresponding indices are j1 = 1 and jt+1 = p + 2. The first
part of Assumption (A2) is therefore obviously satisfied. The second part, i.e.,
n−jt+1 = n−(p+2) ≥p−1, follows immediately from 2p+1 ≤n. Assumption
(A3) follows from y ̸= 0 together with the previous display and t = k. Therefore,
ˆZe+(y) = ˆVp,e+(y) = (0, 1, 0, . . . 0) ∈Rn−p follows as an application of Part 1 of
Lemma B.1. Obviously (CKV) holds if M ∈MKV . Since ˆZe+(y) = (0, 1, 0, . . . 0)
it is also obvious that (CNW) holds if M ∈MNW . Suppose now that M ∈MAM
holds. In this case 1MAM (M) = 1 and therefore 2(p + 1) ≤n holds. The latter
implies n −(p + 2) = n −jt+1 > p −1. Consequently the statement in brackets
in (CAM) shows that the condition is satisfied. Appendix C: Proofs of results in Section 4 We first observe that Bp(.) (which was defined in Equation (7)) is continuous
on Rn\N(ˆΩκ,M,p), because ˆV (.) and hence ˆA(p)(.) and ˆZ(.) are continuous on
Rn\N(ˆΩκ,M,p). By considering each of the cases M ∈MKV , M ∈MNW and
M ∈MAM separately, it is easy to see that M(.) is continuous on Rn\N(ˆΩκ,M,p). The main diagonal entries of Wn(.) are by definition constant on Rn\N(ˆΩκ,M,p). Finite sample properties of prewhitened F-type tests 2141 Therefore, it remains to show that all off-diagonal entries are continuous on
Rn\N(ˆΩκ,M,p). Each of them is of the form κ(i/M(y)) for some fixed |i| =
1, . . . , n−p−1 if M(y) ̸= 0, and 0 if M(y) = 0. Since κ is a continuous function by
Assumption 1, and M(.) is continuous on Rn\N(ˆΩκ,M,p) and satisfies M(y) ≥0,
it remains to check that κ(x) →0 as |x| →∞, which is a part of Assumption
1. This proves the claim. Since ˆβ is well defined and continuous everywhere
on Rn, it follows that both ˆβ and ˆΩκ,M,p are well-defined and continuous on
Rn\N(ˆΩκ,M,p). Clearly, ˆΩκ,M,p is symmetric on Rn\N(ˆΩκ,M,p). This proves Part
(i) of Assumption 5 in Preinerstorfer and P¨otscher (2016). To prove the second
part let y ∈Rn\N(ˆΩκ,M,p), α ̸= 0 and γ ∈Rk. We have to show that αy+Xγ ∈
Rn\N(ˆΩκ,M,p). Note that ˆV (αy + Xγ) = X′ diag(ˆu(αy + Xγ)) = α ˆV (ˆu(y)),
which implies ˆA(p)(αy + Xγ) = ˆA(p)(ˆu(y)) and ˆZ(αy + Xγ) = α ˆZ(y). The lat-
ter immediately leads (considering each of the cases M ∈MKV , M ∈MNW
and M ∈MAM separately) to M(αy + Xγ) = M(y), which in turn implies
Wn−p(αy + Xγ) = Wn−p(y). It then follows from the previous display and
the definition of Bp(y) that ˆΩκ,M,p(αy + Xγ) = α2 ˆΩκ,M,p(y). Therefore, we
clearly have αy + Xγ ∈Rn\N(ˆΩκ,M,p), which proves Part (ii) of Assumption 5
in Preinerstorfer and P¨otscher (2016), and where we have also established the
equivariance property of ˆΩκ,M,p required in Part (iii) of Assumption 5 in Prein-
erstorfer and P¨otscher (2016). That ˆβ satisfies the equivariance property in Part
(iii) of Assumption 5 in Preinerstorfer and P¨otscher (2016) is obvious. It remains
to show that ˆΩκ,M,p is λRn-almost everywhere nonsingular on Rn\N(ˆΩκ,M,p). This is equivalent to y ∈Rn\N(ˆΩκ,M,p) : det(ˆΩκ,M,p(y)) = 0
= N ∗(ˆΩκ,M,p)\N(ˆΩκ,M,p) being a λRn-null set. Appendix C: Proofs of results in Section 4 Hence, Parts 1 and 2
of the present theorem now follow by applying the first two parts of Corollary
5.17 in Preinerstorfer and P¨otscher (2016) and Remark 5.18(i) in Preinerstorfer
and P¨otscher (2016) to Z+ as well as to Z−, and by noting that Part 5 of
Lemma 3.10 shows that the statement e+ ∈Rn\N ∗(ˆΩκ,M,p) translates into
g∗
κ,M,p(e+, X, R) ̸= 0, with a similar translation if e+ is replaced by e−. That
the test is biased in Part 2 of the theorem follows immediately from Part 5 of
Lemma 5.15 in Preinerstorfer and P¨otscher (2016) (note that Assumptions 5 and
6 in Preinerstorfer and P¨otscher (2016) are satisfied by Lemma C.1) showing
that W(C) contains a non-empty open set. To prove Part 4 we apply Theorem
5.19 in Preinerstorfer and P¨otscher (2016). Lemma C.1 shows that ˆβ and ˆΩκ,M,p
also satisfy Assumption 7 in Preinerstorfer and P¨otscher (2016). We consider
the case where e+ ∈M and Rˆβ(e+) ̸= 0. The other case can be handled
analogously. From Remark 5.20 in Preinerstorfer and P¨otscher (2016) we see
that all conditions on the covariance model in Theorem 5.19 in Preinerstorfer
and P¨otscher (2016) are satisfied with ¯Σ = e+e′
+, span(¯Σ) = span(e+) and
Z = e+. Clearly, span(¯Σ) = span(e+) ⊆M and Rˆβ(z) ̸= 0 holds λspan(¯Σ)-a.e. This shows that Equation (33) in Preinerstorfer and P¨otscher (2016) holds in the
present setup. To conclude, it remains to observe that K2 in this equation equals
one. This follows from the discussion preceding Theorem 5.19 in Preinerstorfer
and P¨otscher (2016), because ˆΩκ,M,p is almost everywhere positive definite by
Lemma C.1. Now we consider Part 3 of the theorem. We prove the case where g∗
κ,M,p(e+,
X, R) ̸= 0, T(e+ + μ∗
0) = C and grad T(e+ + μ∗
0) exists for some μ∗
0 ∈M0. The other case works analogously. The statement in the theorem saying that
if grad T(e+ + μ∗
0) exists and T(e+ + μ∗
0) = C holds for some μ∗
0 ∈M0, then
grad T(e+ + μ∗
0) exists and T(e+ + μ∗
0) = C holds for all μ∗
0 ∈M0 follows at
once from invariance of T w.r.t. G(M0), which holds as a consequence of Lemma
C.1. Appendix C: Proofs of results in Section 4 This is obvious, since we have already observed that
N ∗(ˆΩκ,M,p) is a λRn-null set under the maintained assumptions. This proves
the claim concerning Assumption 5. That ˆΩκ,M,p(y) is nonnegative definite for
every y /∈N(ˆΩκ,M,p) (which is equivalent to gκ,M,p(y, X) ̸= 0 by Lemma 3.9)
has been shown in Part 1 of Lemma 3.10. It follows that ˆΩκ,M,p is positive defi-
nite on the complement of N ∗(ˆΩκ,M,p). Hence ˆΩκ,M,p is λRn- almost everywhere
positive definite. This immediately shows that Assumptions 6 and 7 in Preiner-
storfer and P¨otscher (2016) are satisfied. The remaining two claims in the lemma
now follow immediately from what has been established together with Lemma
5.15 Part 3 and Proposition 5.4 in Preinerstorfer and P¨otscher (2016). being a λRn-null set. This is obvious, since we have already observed that
N ∗(ˆΩκ,M,p) is a λRn-null set under the maintained assumptions. This proves
the claim concerning Assumption 5. That ˆΩκ,M,p(y) is nonnegative definite for
every y /∈N(ˆΩκ,M,p) (which is equivalent to gκ,M,p(y, X) ̸= 0 by Lemma 3.9)
has been shown in Part 1 of Lemma 3.10. It follows that ˆΩκ,M,p is positive defi-
nite on the complement of N ∗(ˆΩκ,M,p). Hence ˆΩκ,M,p is λRn- almost everywhere
positive definite. This immediately shows that Assumptions 6 and 7 in Preiner-
storfer and P¨otscher (2016) are satisfied. The remaining two claims in the lemma
now follow immediately from what has been established together with Lemma
5.15 Part 3 and Proposition 5.4 in Preinerstorfer and P¨otscher (2016). Proof of Theorem 4.2. In each part of the theorem we have g∗
κ,M,p(., X, R) ̸≡0. In Part 4 this is an explicit assumption. In the other parts this is implied by
the assumption that g∗
κ,M,p(., X, R) does not vanish at a specific point. As a
consequence Lemma C.1 is applicable in all parts of the theorem. We shall now
apply the first two parts of Corollary 5.17 in Preinerstorfer and P¨otscher (2016)
to prove the first two parts of the present theorem. Lemma C.1 shows that ˆβ
and ˆΩκ,M,p satisfy Assumption 5 in Preinerstorfer and P¨otscher (2016) with
N = N(ˆΩκ,M,p). Furthermore, note that the set N ∗figuring Corollary 5.17 of 2142 D. Preinerstorfer Preinerstorfer and P¨otscher (2016) coincides with N ∗(ˆΩκ,M,p). By Assumption
2 the spaces Z+ = span(e+) and Z−= span(e−) are concentration spaces of C
(cf. Lemma G.1 in Preinerstorfer and P¨otscher (2016)). Appendix C: Proofs of results in Section 4 In a first step we now show that the linear functional on Rn corresponding
to the row vector grad T(μ∗
0 + e+) does not vanish everywhere on span(e+)⊥:
Arguing by contradiction, assume that grad T(μ∗
0 + e+)w = 0 for every w ∈
span(e+)⊥, which is equivalent to grad T(μ∗
0 + e+)′⊥span(e+)⊥and therefore
grad T(μ∗
0 + e+)′ ∈span(e+) holds, i.e., grad T(μ∗
0 + e+) = ce′
+ for some c ∈R. Since T is G(M0) invariant, it holds for every γ ̸= 0 that T(γe+ + μ∗
0) = T(γ(e+ + μ∗
0 −μ∗
0) + μ∗
0) = T(e+ + μ∗
0) = C
Hence on the set R\ {−1} the mapping
α →T(e+ + μ∗
0 + αe+) = T((1 + α)e+ + μ∗
0) = C T(γe+ + μ∗
0) = T(γ(e+ + μ∗
0 −μ∗
0) + μ∗
0) = T(e+ + μ∗
0) = C. Hence on the set R\ {−1} the mapping α →T(e+ + μ∗
0 + αe+) = T((1 + α)e+ + μ∗
0) = C is constant, thus showing that the directional derivative of T at the point e+ +
μ∗
0 in direction e+ is zero. The latter is equivalent to grad T(μ∗
0 + e+)e+ =
c∥e+∥2 = 0, and hence c = 0 holds which implies grad T(μ∗
0 + e+) = 0. To
arrive at a contradiction it remains to show that there is a vector v such that Finite sample properties of prewhitened F-type tests 2143 the directional derivative of T at e+ + μ0 in direction v does not vanish. To
this end, recall that Assumption 5 in Preinerstorfer and P¨otscher (2016) is
satisfied, hence the discussion following that Assumption in Preinerstorfer and
P¨otscher (2016) shows that N ∗(ˆΩκ,M,p) is invariant w.r.t. G(M). Therefore,
e+ /∈N ∗(ˆΩκ,M,p) implies e+ + μ∗
0 /∈N ∗(ˆΩκ,M,p). Since N ∗(ˆΩκ,M,p) is closed by
Lemma 3.10, there exists an open ball Uε of radius ε > 0 centered at e+ + μ∗
0
such that Uε ⊆Rn\N ∗(ˆΩκ,M,p). Additionally, we note that e+ /∈M, because
M ⊆N ∗(ˆΩκ,M,p) always holds (see the discussion in Preinerstorfer and P¨otscher
(2016) after Assumption 5). Therefore, v = ΠM⊥e+/∥ΠM⊥e+∥is well defined
and for 0 ≤|α| < ε we have e+ + μ∗
0 + αv ∈Uε. Assume that 0 ≤|α| < ε. Appendix C: Proofs of results in Section 4 The
OLS estimator ˆβ clearly satisfies Rˆβ(e+ + μ∗
0 + αv) = Rˆβ(e+ + μ∗
0) + αRˆβ(v) = Rˆβ(e+ + μ∗
0), where the second equality follows from v⊥M. Since ˆΩκ,M,p satisfies the equiv-
ariance condition in Assumption 5 of Preinerstorfer and P¨otscher (2016) we can
furthermore write ˆΩκ,M,p(e+ + μ∗
0 + αv) = ˆΩκ,M,p(e+ + αv)
= ˆΩκ,M,p(e+ −ΠMe+ + αv)
= ˆΩκ,M,p(ΠM⊥e+ + αΠM⊥e+/∥ΠM⊥e+∥)
= ˆΩκ,M,p
(1 +
α
∥ΠM⊥e+∥)ΠM⊥e+
= (1 +
α
∥ΠM⊥e+∥)2 ˆΩκ,M,p (ΠM⊥e+)
= (1 +
α
∥ΠM⊥e+∥)2 ˆΩκ,M,p (e+ + μ∗
0) . By definition of T (cf. Equation (5) and recall that e+ + μ∗
0 + αv ∈Uε ⊆
Rn\N ∗(ˆΩκ,M,p)) and the assumed equality T(e++μ∗
0) = C, the relations derived
above allow us to show that T(e+ + μ∗
0 + αv)
=
Rˆβ(e+ + μ∗
0 + αv) −r
′ ˆΩ−1
κ,M,p(e+ + μ∗
0 + αv)
Rˆβ(e+ + μ∗
0 + αv) −r
=(1 +
α
∥ΠM⊥e+∥)−2
Rˆβ(e+ + μ∗
0) −r
′ ˆΩ−1
κ,M,p(e+ + μ∗
0)
Rˆβ(e+ + μ∗
0) −r
=(1 +
α
∥ΠM⊥e+∥)−2T(e+ + μ∗
0)
=(1 +
α
∥ΠM⊥e+∥)−2C holds for every 0 ≤|α| < ε. This implies that the directional derivative of T in
direction v at the point e+ + μ∗
0 equals −2C/∥ΠM⊥e+∥, which is nonzero as a 2144 D. Preinerstorfer consequence of C > 0. In a second step we shall now derive an expansion of T
at points of the form y + μ∗
0 for y satisfying e′
+y ̸= 0: For every h ∈Rn we have T(e+ + μ∗
0 + h) = T(e+ + μ∗
0) + grad T(e+ + μ∗
0)h + Q(h)
(18) (18) where Q(h)/∥h∥→0 as h →0 and h ̸= 0. Recall that T is invariant under the
group G(M0). In particular for every y such that e′
+y ̸= 0 we have where Q(h)/∥h∥→0 as h →0 and h ̸= 0. Recall that T is invariant under the
group G(M0). In particular for every y such that e′
+y ̸= 0 we have T(y + μ∗
0) = T(e′
+y
n e+ + μ∗
0 + Πspan(e+)⊥y) = T(e+ + μ∗
0 +
n
e′
+y Πspan(e+)⊥y), where the first equality holds because of y = Πspan(e+)y + Πspan(e+)⊥y and the
second follows from invariance of T w.r.t. Appendix C: Proofs of results in Section 4 G(M0). This means that whenever
e′
+y ̸= 0 holds, we can combine the equation in the previous display and Equa-
tion (18) with h =
n
e′
+yΠspan(e+)⊥y to see that where the first equality holds because of y = Πspan(e+)y + Πspan(e+)⊥y and the
second follows from invariance of T w.r.t. G(M0). This means that whenever
e′
+y ̸= 0 holds, we can combine the equation in the previous display and Equa-
tion (18) with h =
n
e′
+yΠspan(e+)⊥y to see that T(y+μ∗
0) = T(e+ +μ∗
0)+ n
e′
+y grad T(e+ +μ∗
0)Πspan(e+)⊥y+Q( n
e′
+y Πspan(e+)⊥y)
(19)
holds and that T(y+μ∗
0) = T(e+ +μ∗
0)+ n
e′
+y grad T(e+ +μ∗
0)Πspan(e+)⊥y+Q( n
e′
+y Πspan(e+)⊥y)
(19) (19) holds and that Q(
n
e′
+ym
Πspan(e+)⊥ym)/∥
n
e′
+ym
Πspan(e+)⊥ym∥→0,
(20) (20) for any sequence ym satisfying e′
+ym
̸= 0,
n
e′
+ym Πspan(e+)⊥ym
→0 and
n
e′
+ym Πspan(e+)⊥ym ̸= 0. Now, we choose a sequence ρm ∈(−1, 1) such that
ρm →1 and apply Assumption 2 to obtain Λ(ρm) ∈C for every m. We intend
to show that Pμ∗
0,Λ(ρm)(W(C)) →1/2 along a subsequence. The last state-
ment in Lemma C.1 then implies Pμ0,σ2Λ(ρm)(W(C)) →1/2 for every pair
μ0 ∈M0 and 0 < σ2 < ∞along this subsequence. In Part 3 of Lemma
G.1 in Preinerstorfer and P¨otscher (2016) it is shown that Λ(ρm) →e+e′
+,
Dm = Πspan(e+)⊥Λ(ρm)Πspan(e+)⊥/sm →D, where sm is a sequence of numbers
such that sm > 0, sm →0 and D is regular on span(e+)⊥. Furthermore, it is
shown that Πspan(e+)⊥Λ(ρm)Πspan(e+)/s1/2
m
→0. We can use these relations to
derive three useful facts: (i) Observe that the matrix s−1/2
m
Πspan(e+)⊥Λ(ρm)1/2
is an n × n-dimensional nonnegative square root of the symmetric matrix Dm. Therefore, we can find an orthogonal matrix Um such that for any sequence ym satisfying e′
+ym
̸= 0,
n
e′
+ym Πspan(e+)⊥ym
→0 and
n
e′
+ym Πspan(e+)⊥ym ̸= 0. Now, we choose a sequence ρm ∈(−1, 1) such that
ρm →1 and apply Assumption 2 to obtain Λ(ρm) ∈C for every m. We intend
to show that Pμ∗
0,Λ(ρm)(W(C)) →1/2 along a subsequence. The last state-
ment in Lemma C.1 then implies Pμ0,σ2Λ(ρm)(W(C)) →1/2 for every pair
μ0 ∈M0 and 0 < σ2 < ∞along this subsequence. Appendix C: Proofs of results in Section 4 In Part 3 of Lemma
G.1 in Preinerstorfer and P¨otscher (2016) it is shown that Λ(ρm) →e+e′
+,
Dm = Πspan(e+)⊥Λ(ρm)Πspan(e+)⊥/sm →D, where sm is a sequence of numbers
such that sm > 0, sm →0 and D is regular on span(e+)⊥. Furthermore, it is
shown that Πspan(e+)⊥Λ(ρm)Πspan(e+)/s1/2
m
→0. We can use these relations to
derive three useful facts: (i) Observe that the matrix s−1/2
m
Πspan(e+)⊥Λ(ρm)1/2
is an n × n-dimensional nonnegative square root of the symmetric matrix Dm. Therefore, we can find an orthogonal matrix Um such that s−1/2
m
Πspan(e+)⊥Λ(ρm)1/2Um = D1/2
m holds. The sequence D1/2
m
converges to D1/2 as a consequence of Dm →D to-
gether with the continuity of taking the nonnegative definite symmetric matrix
square root of a symmetric and nonnegative definite matrix. Since Um is or-
thogonal we can choose a subsequence m′ along which Um converges to U, say. Without loss of generality we henceforth assume m′ ≡m. Using the relation in
the previous display we find that s−1/2
m
Πspan(e+)⊥Λ(ρm)1/2 converges to D∗= D1/2U ′,
(21) D∗= D1/2U ′, (21) Finite sample properties of prewhitened F-type tests
2145 2145 and we recall from above that D1/2 ∈Rn×n is regular on span(e+)⊥. (ii) We
note that Λ(ρm) →e+e′
+ implies and we recall from above that D1/2 ∈Rn×n is regular on span(e+)⊥. (ii) We
note that Λ(ρm) →e+e′
+ implies Λ(ρm)1/2 →n−1/2e+e′
+. (iii) We show that D∗e+ = 0 holds: Note that Πspan(e+)⊥Λ(ρm)Πspan(e+)/s1/2
m
→
0 can be rewritten as (iii) We show that D∗e+ = 0 holds: Note that Πspan(e+)⊥Λ(ρm)Πspan(e+)/s1/2
m
→
0 can be rewritten as
s−1/2
m
Πspan(e+)⊥Λ(ρm)1/2
Λ(ρm)1/2Πspan(e+) →0, that the term in brackets converges to D∗by (i) and that the other term con-
verges to n−1/2e+e′
+ by (ii). Therefore, D∗n−1/2e+e′
+ = 0, or equivalently D∗Πspan(e+) = 0. But this implies D∗e+ = 0. Now, we are ready
to show that Pμ∗
0,Λ(ρm)(W(C)) →1/2. Let G be a random n-vector defined on
some underlying probability space such that the probability measure induced by
G on (Rn, B(Rn)) equals P0,In. Consequently, the random vector Λ(ρm)1/2G +
μ∗
0 induces the distribution Pμ∗
0,Λ(ρm) on (Rn, B(Rn)). For notational convenience
we write Gm = Λ(ρm)1/2G. Appendix C: Proofs of results in Section 4 Consequently, we have Pμ∗
0,Λ(ρm) (W(C)) = Pr(T(Gm + μ∗
0) ≥C)
(22)
= Pr
s−1/2
m
[T(Gm + μ∗
0) −T(e+ + μ∗
0)] ≥0
, (22) where we used s−1/2
m
> 0 and T(e+ + μ∗
0) = C in deriving the second equality. Note that e′
+Gm ̸= 0 and ∥
n
e′
+Gm Πspan(e+)⊥Gm∥> 0 on an event of proba-
bility one. Using the expansion developed in Equation (19) we see that with
probability one s−1/2
m
[T(Gm + μ∗
0) −T(e+ + μ∗
0)] can be written as s−1/2
m
)
n
e′
+Gm
grad T(e+ + μ∗
0)Πspan(e+)⊥Gm + Q
n
e′
+Gm
Πspan(e+)⊥Gm
* =
n
e′
+Gm
grad T(e+ + μ∗
0)s−1/2
m
Πspan(e+)⊥Gm + ∥
n
e′
+Gm
s−1/2
m
Πspan(e+)⊥Gm∥ ×
∥
n
e′
+Gm
Πspan(e+)⊥Gm
∥−1Q
n
e′
+Gm
Πspan(e+)⊥Gm
. To derive the almost sure limit as m →∞of the expression in the previous
display we first observe that Gm converges point-wise to n−1/2e+e′
+G because
of (ii). From that it follows that e′
+Gm converges point-wise to √ne′
+G and that
Πspan(e+)⊥Gm converges point-wise to zero. An application of the continuous
mapping theorem hence shows that n
e′
+Gm
Πspan(e+)⊥Gm →0 2146
D. Preinerstorfer D. Preinerstorfer 2146 almost surely as m →∞, which immediately implies almost surely as m →∞, which immediately implies ∥
n
e′
+Gm
Πspan(e+)⊥Gm
∥−1Q
n
e′
+Gm
Πspan(e+)⊥Gm
→0 almost surely as m →∞as a consequence of Equation (20) together with
Q(0) = 0. We also observe that (i) above implies almost surely as m →∞as a consequence of Equation (20) together with
Q(0) = 0. We also observe that (i) above implies Πspan(e+)⊥s−1/2
m
Gm →D∗G point-wise and thus, using the continuous mapping theorem again, we see that point-wise and thus, using the continuous mapping theorem again, we see that n
e′
+Gm
Πspan(e+)⊥s−1/2
m
Gm →
√n
e′
+GD∗G almost surely as m →∞(where the limiting random vector is well defined
almost-surely). This finally shows that s−1/2
m
[T(Gm + μ∗
0) −T(e+ + μ∗
0)] →
√n
e′
+G grad T(e+ + μ∗
0)D∗G, almost surely. We already know from Equation (21) that D∗= D1/2U ′, where
U is an orthogonal matrix. 2. Suppose that M /∈MKV , that κ is continuously differentiable on the non-
void complement of a closed set Δ(κ) ⊆R, and that g∗
κ,M,p(y, X, R) ̸= 0.
Assume further that one of the following conditions is satisfied: Appendix C: Proofs of results in Section 4 Furthermore D1/2 maps Rn onto span(e+)⊥, and
grad T(e+ + μ∗
0) does not vanish everywhere on span(e+)⊥. Hence, we see that
the probability that the limiting random variable in the previous display takes
on the value 0 vanishes because grad T(e+ + μ∗
0)D∗G is a Gaussian random
variable with mean zero and positive variance. Hence, Equation (22) together
with Portmanteau theorem shows that Pμ∗
0,Λ(ρm) (W(C)) →Pr(
√n
e′
+G grad T(e+ + μ∗
0)D∗G ≥0). (23) (23) The covariance between the Gaussian mean-zero random variables grad T(e+ +
μ∗
0)D∗G and e′
+G is given by grad T(e+ + μ∗
0)D∗e+ = 0, grad T(e+ + μ∗
0)D∗e+ = 0, where the equality follows from (iii). Therefore, e′
+G and grad T(e+ + μ∗
0)D∗G
are independent. Since the probability to the right in Equation (23) equals the
probability that the random variables e′
+G and grad T(e+ + μ∗
0)D∗G have the
same sign it is now obvious that the limit equals 1/2. Lemma C.2. Assume that the triple κ, M, p satisfies Assumption 1 and let T
be as in Equation (5) with ˆΩ = ˆΩκ,M,p. Let μ0 ∈M0. Then the following holds. Finite sample properties of prewhitened F-type tests Finite sample properties of prewhitened F-type tests 2147 (a) M(y) ̸= 0 and
i
M(y) /∈Δ(κ) for |i| = 1, . . . , n −p −1. (b) M(y) = 0 and κ has compact support. (b) M(y) = 0 and κ has compact support. Then grad T(μ0 + y) exists. Then grad T(μ0 + y) exists. 3. If M /∈MKV , then for every δ ≥0 we have g
(μ0
y)
3. If M /∈MKV , then for every δ ≥0 we have y ∈Rn : g∗
κ,M,p(y, X, R) ̸= 0 and M(y) = δ
=
y ∈Rn : g∗
κ,M,p(y, X, R) ̸= 0 and g(δ)
κ,M,p(y, X) = 0
, where g(δ)
κ,M,p : Rn × Rn×k →R is a multivariate polynomial (explicitly
constructed in the proof) that does not depend on the hypothesis (R, r). where g(δ)
κ,M,p : Rn × Rn×k →R is a multivariate polynomial (explicitly
constructed in the proof) that does not depend on the hypothesis (R, r). Proof. We first verify Parts 1 and 2. Appendix C: Proofs of results in Section 4 Let us start by deriving a convenient ex-
pression for T(μ0 + y) under the assumption g∗
κ,M,p(y, X, R) ̸= 0. By Lemma
3.10 the assumption g∗
κ,M,p(y, X, R) ̸= 0 is equivalent to y /∈N ∗(ˆΩκ,M,p). An
application of Lemma C.1 shows that ˆΩκ,M,p satisfies Assumption 5 in Prein-
erstorfer and P¨otscher (2016). An application of Part (ii) of this Assumption
shows that μ0+y /∈N ∗(ˆΩκ,M,p). We can therefore use Equation (5) together with
Rˆβ(y + μ0) −r = Rˆβ(y) and ˆΩκ,M,p(μ0 + y) = ˆΩκ,M,p(y) (both following from
Assumption 5 in Preinerstorfer and P¨otscher (2016)) to see that y /∈N ∗(ˆΩκ,M,p)
implies T(μ0 + y) =ˆβ(y)′R′ ˆΩ−1
κ,M,p(y)Rˆβ(y)
=ˆβ(y)′R′
n
n −pBp(y)Wn−p(y)B′
p(y)
−1
Rˆβ(y), where in deriving the second equality we made use of the representation of
ˆΩκ,M,p(y) developed in Equation (13). The function ˆβ(.) is linear and hence
totally differentiable on Rn. Furthermore, in the proof of Lemma 3.10 it is
shown that the coordinates of the matrix Bp(.) are multivariate rational func-
tions (without singularities) on Rn\N ∗(ˆΩκ,M,p). In particular the coordinates
of Bp(.) are continuously partially differentiable on Rn\N ∗(ˆΩκ,M,p). To show
that grad T(μ0 + y) exists at a given point y ∈Rn\N ∗(ˆΩκ,M,p) it is there-
fore sufficient to show that each off-diagonal element (recall that the diago-
nal is constant) of Wn−p(.) is continuously partially differentiable on an open
neighborhood of y in Rn\N ∗(ˆΩκ,M,p). Recall that the i-th off-diagonal element
(i ∈{1, . . . , n −p −1}) of Wn−p(.) evaluated at some y ∈Rn\N ∗(ˆΩκ,M,p) is
given by where in deriving the second equality we made use of the representation of
ˆΩκ,M,p(y) developed in Equation (13). The function ˆβ(.) is linear and hence
totally differentiable on Rn. Furthermore, in the proof of Lemma 3.10 it is
shown that the coordinates of the matrix Bp(.) are multivariate rational func-
tions (without singularities) on Rn\N ∗(ˆΩκ,M,p). In particular the coordinates
of Bp(.) are continuously partially differentiable on Rn\N ∗(ˆΩκ,M,p). To show
that grad T(μ0 + y) exists at a given point y ∈Rn\N ∗(ˆΩκ,M,p) it is there-
fore sufficient to show that each off-diagonal element (recall that the diago-
nal is constant) of Wn−p(.) is continuously partially differentiable on an open
neighborhood of y in Rn\N ∗(ˆΩκ,M,p). Recall that the i-th off-diagonal element
(i ∈{1, . . . Appendix C: Proofs of results in Section 4 , n −p −1}) of Wn−p(.) evaluated at some y ∈Rn\N ∗(ˆΩκ,M,p) is
given by fi(y) :=
κ(i/M(y))
if M(y) ̸= 0
0
else. If M ∈MKV the sufficient condition above is obviously satisfied, because in this
case M > 0 is constant and therefore fi(.) is constant on Rn\N ∗(ˆΩκ,M,p). This
proves Part 1 of the lemma. Consider now Part 2. By considering separately the
cases M ∈MAM and M ∈MNW , we observe that M(.) is continuously partially 2148 D. Preinerstorfer differentiable in an open neighborhood of any element y of Rn\N ∗(ˆΩκ,M,p) sat-
isfying M(y) ̸= 0. We start with Condition (a). Let y satisfy y /∈N ∗(ˆΩκ,M,p)
and M(y) ̸= 0. Fix an i ∈{1, . . . , n −p −1}. By assumption κ is continuously
differentiable on an open neighborhood of i/M(y) /∈Δ(κ). Hence there exists an
open neighborhood U of y in Rn\N ∗(ˆΩκ,M,p) on which M(.) is strictly greater
than zero and such that κ(i/M(.)) is continuously partially differentiable on U. It hence follows that fi(.) is continuously partially differentiable on U because it
coincides with κ(i/M(.)) on this set. To establish existence of the gradient under
Condition (b) let y satisfy y /∈N ∗(ˆΩκ,M,p) and M(y) = 0. Let i be as before and
recall that the support of κ is compact by assumption. Since M is continuous
on Rn\N ∗(ˆΩκ,M,p), there exists an open neighborhood of y in Rn\N ∗(ˆΩκ,M,p)
such that for every point y∗in this neighborhood we either have M(y∗) = 0
or that i/M(y∗) is not contained in the support of κ. It follows that the func-
tion fi is constant equal to 0, and thus is in particular continuously partially
differentiable, on this neighborhood. To prove the third part of the lemma consider first the case M ≡MAM,1,ω ∈
MAM, where we dropped the index c because the argument and the resulting
polynomial do not depend on it. Suppose y satisfies g∗
κ,MAM,1,ω,p(y, X, R) ̸= 0. Then MAM,1,ω(y) is well defined and by definition MAM,1,ω(y) = δ if and only
if ˆα1(y) = n−1(c−1
1 δ)1/c2 =: δ∗holds, where c1 and c2 are positive constants. Appendix C: Proofs of results in Section 4 Finally, we note that the multivariate polynomial g(δ)
κ,MAM,1,ω,p : Rn ×Rn×k →R Finally, we note that the multivariate polynomial g(δ)
κ,MAM,1,ω,p : Rn ×Rn×k →R
does not depend on the hypothesis (R, r). This proves the last part of the lemma
in case M ≡MAM,1,ω ∈MAM. The proof of the case M ≡MAM,2,ω ∈MAM is
almost identical and therefore we omit it. We finally note that similar arguments
can be used to prove the statement in case M ∈MNW , but we omit details. ,
AM,1,ω,p
does not depend on the hypothesis (R, r). This proves the last part of the lemma
in case M ≡MAM,1,ω ∈MAM. The proof of the case M ≡MAM,2,ω ∈MAM is
almost identical and therefore we omit it. We finally note that similar arguments
can be used to prove the statement in case M ∈MNW , but we omit details. Proof of Proposition 4.5. We first prove that the sets X2(e+), X2(e−), ˜X2(e+)
and ˜X2(e−) do not depend on the specific choice of μ∗
0,X ∈M0,X. This follows
from an invariance argument. Consider for example the set X2(e+), which is
by definition a subset of X0\X1(e+). Every element X of this superset satisfies
gκ,M,p(., X, R) ̸≡0. Hence, for every such X the corresponding test statistic
TX is invariant w.r.t. G(M0,X) by Lemma C.1. It now immediately follows
that X1(e+) does not depend on the specific choice of μ∗
0,X ∈M0,X. The same
argument shows that the statement in Part 2 Condition (d) is independent of
the specific choice of μ∗
0,(e+, ˜
X). We shall now prove the three main parts of the
proposition and start with the first: 1) We begin with the statement concerning X1 (e+). Under the maintained
assumptions we know from Part 5 of Lemma 3.10 that g∗
κ,M,p(., ., .) : Rn ×
Rn×k × Rq×k →R is a multivariate polynomial. This immediately implies that
g∗
κ,M,p(e+, ., R) : Rn×k →R is a multivariate polynomial, showing that X ∈Rn×k : g∗
κ,M,p(e+, X, R) = 0 is an algebraic superset of X1 (e+). It hence suffices to show that the set in the
previous display is a λRn×k- null set, or equivalently that g∗
κ,M,p(e+, ., R) ̸≡0. To
this end we shall use Lemma B.1 (with t = k) to construct a matrix X ∈X0 such
that g∗
κ,M,p(e+, X, R) ̸= 0. Appendix C: Proofs of results in Section 4 This can equivalently be written as k
i=1
ωi
4ˆρ2
i (y)ˆσ4
i (y)
(1 −ˆρi(y))6(1 + ˆρi(y))2 = δ∗
k
i=1
ωi
ˆσ4
i (y)
(1 −ˆρi(y))4 , which, after multiplying both sides of the equation by +k
j=1(1 −ˆρj(y))6(1 +
ˆρj(y))2 (which is nonzero), is seen to be equivalent to which, after multiplying both sides of the equation by +k
j=1(1 −ˆρj(y))6(1 +
ˆρj(y))2 (which is nonzero), is seen to be equivalent to k
i=1
ωi
⎡
⎣4ˆρ2
i (y)ˆσ4
i (y)
k
&
j̸=i
(1 −ˆρj(y))6(1 + ˆρj(y))2
⎤
⎦
−
δ∗
k
i=1
ωi
⎡
⎣ˆσ4
i (y)(1 + ˆρi(y))2(1 −ˆρi(y))2
k
&
j̸=i
(1 −ˆρj(y))6(1 + ˆρj(y))2
⎤
⎦= 0. By multiplying both sides of this equation by a suitably large power of the
products of the denominators of ˆρi(y) (which are nonzero), we can write the
preceding equation equivalently as k
i=1
ωi¯p(δ)
i ( ˆZ(y)) = 0 where each ¯p(δ)
i
: Rk×(n−p) →R for i = 1, . . . , k is a multivariate polynomial. In
a final step we multiply both sides of the equation by a suitably large power of
the non-vanishing factor in Equation (10) to obtain an equivalent equation of
the form where each ¯p(δ)
i
: Rk×(n−p) →R for i = 1, . . . , k is a multivariate polynomial. In
a final step we multiply both sides of the equation by a suitably large power of
the non-vanishing factor in Equation (10) to obtain an equivalent equation of
the form Finite sample properties of prewhitened F-type tests Finite sample properties of prewhitened F-type tests
2149 2149 g(δ)
κ,MAM,1,ω,p(y, X) =
k
i=1
ωip(δ)
i (y, X) = 0, where each p(δ)
i
: Rn × Rn×k →R is a multivariate polynomial. Therefore, the
condition where each p(δ)
i
: Rn × Rn×k →R is a multivariate polynomial. Therefore, the
condition g∗
κ,MAM,1,ω,p(y, X) ̸= 0 and MAM,1,ω(y) = δ g∗
κ,MAM,1,ω,p(y, X) ̸= 0 and MAM,1,ω(y) = δ g∗
κ,MAM,1,ω,p(y, X) ̸= 0 and MAM,1,ω(y) = δ
can be equivalently stated as can be equivalently stated as g∗
κ,MAM,1,ω,p(y, X, R) ̸= 0 and g(δ)
κ,MAM,1,ω,p(y, X) = 0. Appendix C: Proofs of results in Section 4 Let H ∈R(k+1)×k be an auxiliary matrix the column
vectors of which span span(¯e+)⊥, where ¯e+ = (1, . . . , 1)′ ∈Rk+1 is the vector
obtained from e+ by selecting the coordinates with indices ji = 1+(i−1)(p+1)
for i = 1, . . . , k + 1 (we shall need a similar construction for e−later on). Note that this selection is feasible because jk+1 = 1 + k(p + 1) ≤n holds as
a consequence of the assumption n −[k(p + 1) + p] −1MAM (M) ≥0 together
with p ≥1. We also note that H does not contain a row consisting of zeros
only. If the construction of M involves a weights vector ω (which is assumed to
be functionally independent of the design) we choose the columns of H in such 150
D P
i
t
f 2150 D. Preinerstorfer a way that Hω = (−1, 1, 0, . . . , 0)′ which is possible because this vector is an
element of span(¯e+)⊥and ω ̸= 0 holds. We now let X ∈Rn×k be the matrix
the non-zero rows of which are precisely Xji· = Hi· for i = 1, . . . , k + 1, i.e., X =
H′
1·, 0k,p, H′
2·, 0k,p, H′
3·, 0k,p, . . . , H′
(k+1)·, 0k,n−k(p+1)−1
′
∈Rn×k,
(24) (24) where 0m1,m2 denotes the m1 ×m2-dimensional zero matrix (here 0k,n−k(p+1)−1
vanishes if n−k(p+1)−1 = 0). Obviously rank(X) = rank(H) = k holds, which
implies X ∈X0. Furthermore, e+⊥span(X) holds, which follows immediately
from ¯e+⊥span(H), because the non-zero columns of X have column indices ji
for i = 1, . . . , k+1 by construction. Therefore, we see that ˆuX(e+) = e+ showing
that ˆVX(e+) = X′ diag(ˆuX(e+)) = X′ diag(e+). Now we apply Lemma B.1 with t = k to (e+, X) ∈Rn × X0. That the tuple
(e+, X) satisfies (A1) of that lemma is obvious from the preceding display and
Equation (24). We also see that (A2) is satisfied because ji+1 −ji = p + 1
for i = 1, . . . k, and k(p + 1) + p + 1MAM (M) ≤n implies n −jk+1 = n −
k(p + 1) −1 ≥p −1. That (A3) is satisfied follows from the preceding display
together with rank(X) = rank(H) = k. Appendix C: Proofs of results in Section 4 To infer g∗
κ,M,p(e+, X, R) ̸= 0 from
Part 2 of Lemma B.1, we consider three cases: First, if M ∈MKV (CKV) is
obviously satisfied and we are done. Secondly, assume that M ∈MAM. In this
case we have by assumption k(p + 1) + p + 1 ≤n which implies n −jk+1 =
n −k(p + 1) −1 > p −1. This shows that (CAM) is satisfied. Thirdly suppose
that M ∈MNW . Since ˆA(p)
X (e+) = 0 follows from Part 1 of Lemma B.1, we see
that ˆZX(e+) = ˆVp,X(e+) and hence that the nonzero columns of ˆZX(e+) are
precisely H′
i· for i = 2, . . . , k+1. By construction we have H2·ω ̸= 0 and Hi·ω = 0
for i = 3, . . . , k +1. This shows that exactly one coordinate of ω′ ˆZX(e+) is non-
zero which implies that (CNW) holds. To show that X1(e−) is a λRn×k−null
set, we can use a similar construction: we replace e+ by e−throughout. If
p is even we then have ¯e−= (−1, 1, −1, . . . , (−1)k+1)′. If p is odd we then
have ¯e−= (−1, −1, . . . , −1)′. Furthermore, if the construction of M involves a
weights vector we choose H such that Hω = (1, 1, 0, . . . , 0)′ if p is even, and
Hω = (−1, 1, 0, . . . , 0)′ if p is odd. The remaining arguments are identical. Now consider X2(e+). Using Part 1 of Lemma C.2 we see that in case M ∈
MKV the set X2(e+) is empty, because (grad TX(.))|e++μ∗
0,X exists whenever
X ∈X0\X1(e+). Next consider the cases where M /∈MKV and where κ satisfies
Assumption 3. From Part 2a of Lemma C.2 we know that for X ∈X0\X1(e+) the
non-existence of (grad TX(.))|e++μ∗
0,X implies either M(e+) = 0 or i/M(e+) ∈
Δ(κ) for some |i| = 1, . . . , n −p −1. The latter two cases can clearly be sum-
marized as M(e+) ∈¯Δ, where ¯Δ = {δ0, δ1, . . . , δm} is a set consisting of finitely
many elements. Appendix C: Proofs of results in Section 4 Therefore, X2(e+) ⊆
X ∈X0\X1(e+) : (grad TX(.))|e++μ∗
0 does not exist
⊆
X ∈X0\X1(e+) : M(e+) ∈¯Δ 2151 Finite sample properties of prewhitened F-type tests
2151 2151 =
m
,
i=0
{X ∈X0\X1(e+) : M(e+) = δi} . =
m
,
i=0
{X ∈X0\X1(e+) : M(e+) = δi} . We use Part 3 of Lemma C.2 to rewrite the latter set as m
,
i=0
X ∈X0 : g∗
κ,M,p(e+, X, R) ̸= 0 and g(δi)
κ,M,p(e+, X) = 0
=
X ∈X0 : g∗
κ,M,p(e+, X, R) ̸= 0 and
m
&
i=0
g(δi)
κ,M,p(e+, X) = 0
, which is clearly a subset of X ∈Rn×k :
m
&
i=0
g(δi)
κ,M,p(e+, X) = 0
. Part 3 of Lemma C.2 shows that +m
i=0 g(δi)
κ,M,p(e+, .) : Rn×k →R is a multivariate
polynomial. We consider two cases: First assume that +m
i=0 g(δi)
κ,M,p(e+, .) ̸≡0. Consequently, the set in the previous display is a λRn×k-null set. It hence
follows that X2(e+) is a λRn×k-null set and we are done. Next, assume that
+m
i=0 g(δi)
κ,M,p(e+, .) ≡0. It follows that there must exist a single index i such that
g(δi)
κ,M,p(e+, .) ≡0 holds [this is easily shown by contradiction]. Part 3 of Lemma
C.2 hence shows that X ∈X0 and g∗
κ,M,p(e+, X, R) ̸= 0, i.e., X ∈X0\X1(e+),
implies M(e+) = δi. Clearly, X2(e+) ⊆{X ∈X0\X1(e+) : TX(e+ + μ∗
0) = C} . If X ∈X0\X1(e+), then (cf. the argument in the beginning of the proof of
Lemma C.2 applied to y = e+) TX(e+ + μ∗
0,X) TX(e+ + μ0,X)
=
ˆβX(e+)′R′
n
n −pBp,X(e+)Wn−p(e+)B′
p,X((e+)
−1
RˆβX(e+). (25) (25) Furthermore, since X ∈X0\X1(e+) implies M(e+) = δi, the matrix Wn−p(e+)
is constant ¯
Wn−p, say, on X0\X1(e+). where ¯Bp,X(e+) and Fp(e+, X) have been defined in the proof of Lemma 3.10,
where it is shown that gκ,M,p(e+, X) ̸= 0 and X ∈X0 (which is weaker than Appendix C: Proofs of results in Section 4 Hence, for X ∈X0\X1(e+), the statement
TX(e+ + μ∗
0,X) = C is equivalent to F 2
p (e+, X)
'
det(X′X)ˆβX(e+)
(′
R′
× adj( ¯Bp,X(e+) ¯
Wn−p ¯B′
p,X((e+)))R
'
det(X′X)ˆβX(e+)
(
=
n
n −p det(X′X)2 det( ¯Bp,X(e+) ¯
Wn−p ¯B′
p,X((e+)))C where ¯Bp,X(e+) and Fp(e+, X) have been defined in the proof of Lemma 3.10,
where it is shown that gκ,M,p(e+, X) ̸= 0 and X ∈X0 (which is weaker than where ¯Bp,X(e+) and Fp(e+, X) have been defined in the proof of Lemma 3.10,
where it is shown that gκ,M,p(e+, X) ̸= 0 and X ∈X0 (which is weaker than 2152 D. Preinerstorfer X ∈X0\X1(e+)) implies Fp(e+, X) ̸= 0. Furthermore, it is shown in the proof
of Lemma 3.10 that [ ¯Bp,.(e+)]ij : Rn×k →R (for 1 ≤i ≤q and 1 ≤j ≤n −p)
is a multivariate polynomial, and that Fp(e+, .) : Rn×k →R is a multivariate
polynomial as well. It is easily seen that the coordinates of det(X′X)ˆβX(e+)
as a function of X are multivariate polynomials. Putting this together we have
shown that X2(e+) ⊆
X ∈Rn×k : pκ,M,p(e+, X, C) = 0
, where pκ,M,p(e+, ., C) : Rn×k →R is a multivariate polynomial. Therefore,
if we can show that pκ,M,p(e+, ., C) ̸≡0 we obtain that X2(e+) is a λRn×k-
null set. In the proof of the part concerning X1(e+) above we have already
constructed an X ∈X0\X1 that satisfies e+⊥span(X) which implies ˆβX(e+) =
0. Together with Equation (25) this shows that TX(e+ + μ∗
0,X) = 0 < C holds
for this specific X. But this immediately shows that pκ,M,p(e+, X, C) ̸= 0 and
we are done. Clearly, we can use an almost identical argument to prove the
statement concerning X2(e−). Under Assumption 2 the set of matrices X ∈X0
for which the first three cases of Theorem 4.2 do not apply is obviously a subset
of (X1(e+)∪X2(e+))∩(X1(e−)∪X2(e−)). Hence the first part of the proposition
follows. 2) We start with the statement concerning ˜X1 (e−). Under the maintained
assumptions we know from Part 5 of Lemma 3.10 that g∗
κ,M,p(., ., .) : Rn ×
Rn×k × Rq×k →R is a multivariate polynomial. Appendix C: Proofs of results in Section 4 This immediately implies that
g∗
κ,M,p(e−, (e+, .), R) : Rn×(k−1) →R is a multivariate polynomial, which shows
that ˜X ∈Rn×(k−1) : g∗
κ,M,p(e−, (e+, ˜X), R) = 0 is an algebraic superset of ˜X1 (e−). It hence suffices to show that the set in the
previous display is a λRn×(k−1)- null set, or equivalently to show that g∗
κ,M,p(e−,
(e+, .), R) ̸≡0. Again, we shall use Lemma B.1 with t = k to construct a
matrix ˜X ∈˜X0 such that g∗
κ,M,p(e−, (e+, ˜X), R) ̸= 0. The situation here is more
complicated than in the first part, because the first column of the design matrix
we seek has to be the intercept. For our construction we need some additional
ingredients: By definition p∗= p + 1 if p is odd, and p∗= p if p is even. If
p is odd set v = ¯e∗
−= (−1, −1, . . . , −1, 1)′ ∈Rk+1, where ¯e∗
−is the vector
obtained from e−by selecting the coordinates j∗
i = ji for i = 1, . . . , k and
j∗
k+1 = j∗
k + p∗+ 1, where ji = 1 + (i −1)(p + 1) for i = 1, . . . , k + 1 was
defined in Part 1 above. This selection is feasible, because by assumption we
have k(p+1)+p∗≤n, which, since p is odd, gives k(p+1)+p+1 ≤n, implying
that j∗
k+1 = (k −1)(p + 1) + p∗+ 2 = k(p + 1) + 2 ≤n, because of p ≥1. If p is
even set v = ¯e−= (−1, 1, −1, . . . , (−1)k+1)′ ∈Rk+1, the vector obtained from
e−by selecting the coordinates j∗
i = ji for i = 1, . . . , k + 1. Next, define z = (−1, k−1, . . . , k−1)′ ∈Rk+1. We claim that v and z satisfy u := Πspan(z)v ̸= 0, x := Πspan(z)⊥v = v −u is
linearly independent of e := (1, . . . , 1)′ ∈Rk+1 and z is orthogonal to e. The Finite sample properties of prewhitened F-type tests 2153 latter property is clearly always satisfied, regardless of whether p is even or odd. We thus only have to verify the first two conditions. We start with the case p
odd. Here we have z′v = 2k−1 ̸= 0 and therefore Πspan(z)v ̸= 0. Appendix C: Proofs of results in Section 4 Furthermore
Πspan(z)⊥v = v −∥z∥−2z′vz can not equal ce for some c ∈R, because the last
and the last but one coordinate of v are unequal. For p even z′v = (1 + k−1s),
where s equals either 0 (if k is even) or 1 (if k is odd), therefore z′v ̸= 0 holds
and thus Πspan(z)v ̸= 0. Furthermore Πspan(z)⊥v = v −∥z∥−2z′vz can not equal
ce for some c ∈R, because the second and third coordinate of v are unequal. This proves the claim. Using these properties, we see that u ∈span(z)\ {0} and u = v −x = v −Πspan(e,x)(v −Πspan(z)v) = v −Πspan(e,x)v = Πspan(e,x)⊥v, where we have used that z is orthogonal to both e and x to derive the third
equality. We shall now define an auxiliary matrix. Let L denote a (k + 1) × k-
dimensional matrix such that L·1 = e, L·2 = x and such that the remaining
k −2 columns L·j for j = 3, . . . , k are linearly independent and orthogonal to
span(e, x, v). Since e and x are linearly independent, we have rank(L) = k. For
later use we observe that Πspan(L)v = Πspan((e,x))v = v −Πspan(e,x)⊥v = v −u = x, where the first equality follows immediately from L·j for j = 3, . . . , k being
linearly independent and orthogonal to span(e, x, v). This immediately shows where the first equality follows immediately from L·j for j = 3, . . . , k being
linearly independent and orthogonal to span(e, x, v). This immediately shows ˆβL(v) = (0, 1, 0, . . . , 0)′ ∈Rk. Define the two k-vectors r−= (1, −1, 0, . . . , 0)′ and r+ = (1, 1, 0, . . . , 0)′. Let
X ∈Rn×k be such that Xj∗
i · = Li· for i = 1, . . . , k + 1, and if the index j /∈
j∗
1, . . . , j∗
k+1
, then let Xj· = r+ if [e−]j = 1, and let Xj· = r−if [e−]j = −1. By
construction the matrix X is of the form X = (e+, ˜X). We claim that ˆβX(e−) =
ˆβL(v). To see this denote the set of indices j ∈{1, . . . , n} \
j∗
1, . . . Appendix C: Proofs of results in Section 4 , j∗
k+1
such
that [e−]j = −1 by I−, and the set of indices j ∈{1, . . . , n} \
j∗
1, . . . , j∗
k+1
such that [e−]j = 1 by I+. The sum of squares S(β) = ∥e−−Xβ∥2 can be
written as Define the two k-vectors r−= (1, −1, 0, . . . , 0)′ and r+ = (1, 1, 0, . . . , 0)′. Let
X ∈Rn×k be such that Xj∗
i · = Li· for i = 1, . . . , k + 1, and if the index j /∈
j∗
1, . . . , j∗
k+1
, then let Xj· = r+ if [e−]j = 1, and let Xj· = r−if [e−]j = −1. By
construction the matrix X is of the form X = (e+, ˜X). We claim that ˆβX(e−) =
ˆβL(v). To see this denote the set of indices j ∈{1, . . . , n} \
j∗
1, . . . , j∗
k+1
such
that [e−]j = −1 by I−, and the set of indices j ∈{1, . . . , n} \
j∗
1, . . . , j∗
k+1
such that [e−]j = 1 by I+. The sum of squares S(β) = ∥e−−Xβ∥2 can be
written as S(β) =
k+1
i=1
([e−]j∗
i −Xj∗
i ·β)2 +
j∈I−
(−1 −r−β)2 +
j∈I+
(1 −r+β)2
=
k+1
i=1
(vi −Li·β)2 +
j∈I−
(−1 −r′
−β)2 +
j∈I+
(1 −r′
+β)2
= ∥v −Lβ∥2 +
j∈I−
(−1 −r′
−β)2 +
j∈I+
(1 −r′
+β)2 If we now plug in β = ˆβL(v) and note that r′
+ ˆβL(v) = 1 and r′
−ˆβL(v) = −1 we
see that If we now plug in β = ˆβL(v) and note that r′
+ ˆβL(v) = 1 and r′
−ˆβL(v) = −1 we
see that
S(ˆβL(v)) =
k+1
(vi
Li ˆβL(v))2 = min ∥v
Lβ∥2 If we now plug in β = ˆβL(v) and note that r′
+ ˆβL(v) = 1 and r′
−ˆβL(v) = −1 we
see that
S(ˆβL(v)) =
k+1
i=1
(vi −Li· ˆβL(v))2 = min
β∈Rk ∥v −Lβ∥2. S(ˆβL(v)) =
k+1
i=1
(vi −Li· ˆβL(v))2 = min
β∈Rk ∥v −Lβ∥2. 2154 D. Preinerstorfer This immediately proves the claim ˆβX(e−) = ˆβL(v). Appendix C: Proofs of results in Section 4 Hence, the residual vector
satisfies This immediately proves the claim ˆβX(e−) = ˆβL(v). Hence, the residual vector
satisfies [ˆuX(e−)]j =
ui
if j = j∗
i for some i = 1, . . . , k + 1
0
else. ]j =
ui
if j = j∗
i for some i = 1, . . . , k + 1 This immediately entails that ˆVX(e−) = X′ diag(ˆuX(e−)) equals This immediately entails that ˆVX(e−) = X′ diag(ˆuX(e−)) equals This immediately entails that ˆVX(e−) = X′ diag(ˆuX(e−)) equals (u1L′
1·, 0k,p, u2L′
2·, 0k,p, . . . . . . , ukL′
k, 0k,p∗, uk+1L′
(k+1)·, 0k,n−[k(p+1)+p∗−p+1]),
(26) (26) where the indices of the nonzero columns of this matrix are precisely j∗
i for
i = 1, . . . , k+1, because the first column of L is e and ui ̸= 0 for i = 1, . . . , k+1,
the latter following since u ∈span(z)\ {0} and zi ̸= 0 for i = 1, . . . , k + 1 by
definition. In deriving the dimension of 0k,n−[k(p+1)+p∗−p+1] we used j∗
k+1 = k(p + 1) + 1 + p∗−p =
k(p + 1) + 2
if p odd
k(p + 1) + 1
if p even. Now we apply Lemma B.1 (with t = k). Clearly, rank(X) = rank(L) = k. From
Equation (26), and the discussion following it, we see that the tuple (e−, X) ∈
Rn ×X0 satisfies Assumption (A1). Assumption (A2) is satisfied, because j∗
i+1 −
j∗
i ≥p + 1 for i = 1, . . . , k, and because we see from the previous display that
n −j∗
k+1 = n −[k(p + 1) + 1 + p∗−p], which together with the assumption
k(p + 1) + p∗+ 1MAM (M) ≤n implies n −j∗
k+1 ≥p −1. Assumption (A3)
is satisfied because rank( ˆVX(e−)) = rank(L) = k. If M ∈MKV we are done. Consider the case M ∈MAM. We show that Condition (CAM) is satisfied. But
this is obvious, because the assumption k(p+1)+p∗+1 ≤n immediately implies
n −j∗
k+1 > p −1. Suppose M ∈MNW . We apply Part 4 of Lemma B.1. For this
we claim that either [ ˆVX(e−)]1j∗
i > 0 for i = 2, . . . , k + 1 or [ ˆVX(e−)]1j∗
i < 0 for
i = 2, . . Appendix C: Proofs of results in Section 4 . , k + 1. Assuming that this claim is true, the lemma shows that there
exists a regular matrix ¯Q ∈Rk×k such that X ¯Q ∈X0, the first column of X ¯Q
is e+ and g∗
κ,M,p(e+, X ¯Q, R) ̸= 0, and we are done. To prove the claim recall
that by construction u ∈span(z)\ {0} holds, which shows that either ui < 0 for
i = 2, . . . , k + 1 or ui > 0 for i = 2, . . . , k + 1. Furthermore, the first column of
L is the vector e = (1, . . . , 1). Equation (26) now shows that [ ˆVX(e−)]1j∗
i = ui
for i = 2, . . . , k + 1. This proves the claim. Now we apply Lemma B.1 (with t = k). Clearly, rank(X) = rank(L) = k. From
Equation (26), and the discussion following it, we see that the tuple (e−, X) ∈
Rn ×X0 satisfies Assumption (A1). Assumption (A2) is satisfied, because j∗
i+1 −
j∗
i ≥p + 1 for i = 1, . . . , k, and because we see from the previous display that
n −j∗
k+1 = n −[k(p + 1) + 1 + p∗−p], which together with the assumption
k(p + 1) + p∗+ 1MAM (M) ≤n implies n −j∗
k+1 ≥p −1. Assumption (A3)
is satisfied because rank( ˆVX(e−)) = rank(L) = k. If M ∈MKV we are done. Consider the case M ∈MAM. We show that Condition (CAM) is satisfied. But
this is obvious, because the assumption k(p+1)+p∗+1 ≤n immediately implies
n −j∗
k+1 > p −1. Suppose M ∈MNW . We apply Part 4 of Lemma B.1. For this
we claim that either [ ˆVX(e−)]1j∗
i > 0 for i = 2, . . . , k + 1 or [ ˆVX(e−)]1j∗
i < 0 for
i = 2, . . . , k + 1. Assuming that this claim is true, the lemma shows that there
exists a regular matrix ¯Q ∈Rk×k such that X ¯Q ∈X0, the first column of X ¯Q
is e+ and g∗
κ,M,p(e+, X ¯Q, R) ̸= 0, and we are done. To prove the claim recall
that by construction u ∈span(z)\ {0} holds, which shows that either ui < 0 for
i = 2, . . . Appendix C: Proofs of results in Section 4 In the latter case one can show, with a similar argument as
in Part 1 of the proof, that ˜X2(e−) ⊆
˜X ∈Rn×(k−1) : pκ,M,p(e−, (e+, ˜X), C) = 0
,
(28) (28) where pκ,M,p(e−, (e+, .), C) : Rn×(k−1) →R is a multivariate polynomial. Either
we have that pκ,M,p(e−, (e+, .), C) ̸≡0 and ˜X2(e−) is a λRn×(k−1)-null set, or
pκ,M,p(e−, (e+, .), C) ≡0 and the superset in the previous display coincides with
˜X0\˜X1(e−). Consider Condition (d). If the function ˜X →T(e+, ˜
X)(μ∗
0,(e+, ˜
X)+e−)
is not constant C on ˜X0\˜X1(e−), then pκ,M,p(e−, (e+, .), C) ̸≡0, and hence the
superset in Equation (28) is a λRn×(k−1)-null set. This shows that ˜X2(e−) is a
null set under Condition (d). For Condition (b) we consider again the inclusion in Equation (27). Either
the superset is a null set and we are done, or there must exist a real number δi
such that g(δi)
κ,M,p(e−, (e+, .)) ≡0. Assume the latter. We exploit a property of
the matrix X = (e+, ˜X) with ˜X ∈˜X0\˜X1(e−) constructed above. In the proof
of Part 2 (CAM) of Lemma B.1 it is shown that for this specific X we have
M(e−) = 0. Therefore, g(0)
κ,M,p(e−, (e+, .)) ≡0, or equivalently M(e−) = 0 for
every design matrix X = (e+, ˜X) with ˜X ∈˜X0\˜X1(e−). But since the kernel
satisfies Assumption 4, the existence of (grad T(e+, ˜
X)(.))|μ∗
0,(e+, ˜
X)+e−for every X ∈X0\X1(e−) then follows from Part 2b of Lemma C.2. Hence, X2(e−) = ∅. Consider Condition (c). We use a similar argument as under Condition (b):
We establish the existence of a sequence of matrices (e+, ˜Xm) with ˜Xm even-
tually in ˜X0\˜X1(e−), such that M(e−) →0 as m →∞. From an argument as
in the proof under Condition (b) this then implies that either ˜X2(e−) is a null
set, or that M(e−) ≡0 on ˜X0\˜X1(e−). But since κ satisfies Assumption 4, it
then follows from Part 2b of Lemma C.2 that in the latter case ˜X2(e−) is empty. This then proves the claim. It remains to construct a sequence ˜Xm as claimed. By assumption ωi > 0 for some i > 1. Assume without loss of generality that
i = 2 (otherwise we have to interchange the columns of the ˜Xm sequence to
be constructed accordingly). Let X = (e+, ˜X) be as constructed above. Appendix C: Proofs of results in Section 4 , k + 1 or ui > 0 for i = 2, . . . , k + 1. Furthermore, the first column of
L is the vector e = (1, . . . , 1). Equation (26) now shows that [ ˆVX(e−)]1j∗
i = ui
for i = 2, . . . , k + 1. This proves the claim. The part of the statement concerning ˜X2(e−) is established by exploiting an
argument similar to the one given in Part 1 of the proof. Firstly, if M ∈MKV ,
then we know from Part 1 of Lemma C.2 that g∗
κ,M,p(e−, (e+, ˜X), R) ̸= 0 implies
existence of grad(T(e+, ˜
X)(.))|e−+μ∗
0,(e+, ˜
X). Therefore, ˜X2(e−) is empty in Case
(a). It remains to prove the remaining three cases, in all of which Assumption 3
holds. Clearly we can also assume that M /∈MKV . We start with the following
observation: Combining Assumption 3 with Part 2a of Lemma C.2 as in Part
1 of the proof, we see that there exists an integer m ≥0 and real numbers The part of the statement concerning ˜X2(e−) is established by exploiting an
argument similar to the one given in Part 1 of the proof. Firstly, if M ∈MKV ,
then we know from Part 1 of Lemma C.2 that g∗
κ,M,p(e−, (e+, ˜X), R) ̸= 0 implies
existence of grad(T(e+, ˜
X)(.))|e−+μ∗
0,(e+, ˜
X). Therefore, ˜X2(e−) is empty in Case +
(a). It remains to prove the remaining three cases, in all of which Assumption 3
holds. Clearly we can also assume that M /∈MKV . We start with the following
observation: Combining Assumption 3 with Part 2a of Lemma C.2 as in Part
1 of the proof, we see that there exists an integer m ≥0 and real numbers Finite sample properties of prewhitened F-type tests
2155 Finite sample properties of prewhitened F-type tests
2155 2155 δ0, . . . , δm, such that ˜X2(e−) ⊆
˜X ∈Rn×(k−1) :
m
&
i=0
g(δi)
κ,M,p(e−, (e+, ˜X)) = 0
. (27) (27) It follows with the same argument as in Part 1 that either ˜X2(e−) is a λRn×(k−1)-
null set, or there exists an index i such that g(δi)
κ,M,p(e−, (e+, .)) ≡0. In the former
case we are done. Appendix C: Proofs of results in Section 4 Let
γm > 0 be a sequence diverging to ∞. Recall that by construction ˜Xj∗
i 1 = xi for
i = 1, . . . , k + 1. Since p is odd, a simple calculation shows that x = Πspan(z)⊥v
equals x =
−1 +
2
k + 1, −1 −2k−1
k + 1, . . . , −1 −2k−1
k + 1, 1 −2k−1
k + 1
′
. D. Preinerstorfer D. Preinerstorfer 2156
D. Preinerstorfer 2156 Let c = 1 + 2k−1
k+1 and note that Let c = 1 + 2k−1
k+1 and note that Let c = 1 + 2k−1
k+1 and note that x + ce = 2(k−1 + 1
k + 1 , 0, . . . , 0, 1)′. Define the k × k dimensional regular matrix Define the k × k dimensional regular matrix Qm = QDm =
⎛
⎜
⎜
⎜
⎜
⎜
⎝
1
c
. . . . . . . . . 0
0
1
0
. . . . . . 0
0
0
1
0
. . . 0
... ...... ... ... ... 0
0
0
0
. . . 1
⎞
⎟
⎟
⎟
⎟
⎟
⎠
diag(1, γm, 1, . . . , 1). Clearly, post-multiplying a matrix with k columns by Qm has the same effect
as adding c times the first column to the second column, then multiplying the
column so obtained by γm and leaving all other columns unchanged. Since L·1 =
e and L·2 = x, the expression for x + ce above shows that the second column of
LQ, has precisely two nonzero elements with indices 1 and k + 1, respectively. Since X = (e+, ˜X) ∈X0 and (e−, X) satisfies (A1)-(A3) as established above,
Part 3 of Lemma B.1 shows that Xm = (e+, ˜X)Qm = (e+, ˜Xm) ∈X0, and that the tuple (e−, Xm) ∈Rn × X0 satisfies (A1), (A2) and (A3). An
application of Lemma B.1 to that tuple then shows that rank( ˆZXm)(e−)) = k,
which together with ω ̸= 0 immediately implies ¯σ0,Xm(e−) ̸= 0. Since ˆZX(e−)
is obtained from ˆVX(e−) by deleting its first p columns, we observe, using the
remark concerning the second column of LQ above together with Equation (26),
that the second row of Q′ ˆZX(e−) has exactly one non-zero coordinate, namely
2uk+1. 3) Let ˜X ∈˜X0 and assume that X = (e+, ˜X) satisfies g∗
κ,M,p(., X, R) ̸≡0.
Obviously, e+ ∈span(X). Note that ˆβX(e+) = e1(k). The first column of R
is non-zero. Therefore RˆβX(e+) ̸= 0. Thus we can (since Assumption 2 holds)
apply Part 4 of Theorem 4.2. Appendix D: Proofs of results in Section 5 Proof of Theorem 5.2. We verify the assumptions of Theorem 5.21 in Preiner-
storfer and P¨otscher (2016) with ˆΩκ,M,p = ˇΩ and ˆβ = ˇβ. Because g∗
κ,M,p(., X, R)
̸≡0 by assumption, and since the triple κ, M, p satisfies Assumption 1, we
can use Lemma C.1 to conclude that ˆβ and ˆΩκ,M,p satisfy Assumptions 5, 6
and 7 in Preinerstorfer and P¨otscher (2016), that ˆΩκ,M,p is almost everywhere
positive definite and that T is invariant w.r.t. G(M0). Assumption 5, Remark
5.1 (Part (iii)) together with Remark 5.14 (ii) in Preinerstorfer and P¨otscher
(2016) now shows that J(C) = span(e+) ∪span(e−) and that all assumptions
on C appearing in Theorem 5.21 in Preinerstorfer and P¨otscher (2016) are sat-
isfied. Because e+, e−∈M is assumed we have J(C) ⊆M. The assumption
Rˆβ(e+) = Rˆβ(e−) = 0 even implies J(C) ⊆M0 −μ0 (for some arbitrary
μ0 ∈M0). Invariance of T w.r.t. G(M0) then shows that Equation (34) in
Theorem 5.21 of Preinerstorfer and P¨otscher (2016) is satisfied. The assump-
tions on ˇΩ appearing in Parts 2 and 3 of that theorem are satisfied, because
ˆΩκ,M,p is positive definite almost everywhere. The theorem now follows from
Theorem 5.21 in Preinerstorfer and P¨otscher (2016), using a standard subse-
quence argument, positive definiteness of every element of C and compactness
of C∗, to obtain the second statement in Part 3 from the corresponding Part of
Theorem 5.21 in Preinerstorfer and P¨otscher (2016). The claim in parenthesis
in Part 3 follows from the corresponding claim in parenthesis in Theorem 5.21
of Preinerstorfer and P¨otscher (2016), together with the observation that the
conditions on e+ and e−have only been used to verify the condition in Equa-
tion (34) of Theorem 5.21 of Preinerstorfer and P¨otscher (2016) (cf. the proof
of Theorem 3.7 in Preinerstorfer and P¨otscher (2016)). Proof of Theorem 5.4. We apply Theorem 5.21 of Preinerstorfer and P¨otscher
(2016) with the estimators ˆΩκ, ¯
M,p, ¯
X = ˇΩ and (Ik, 0)ˆβ ¯
X = ˇβ. Obviously, the test
statistic defined in Equation (28) of Preinerstorfer and P¨otscher (2016) based on
these estimators coincides with the test statistic ¯T as defined in the statement of
the present theorem. Appendix C: Proofs of results in Section 4 Consider ¯σi,Xm(e−)
¯σ0,Xm(e−) =
n−p
j=|i|+1 ω′[ ˆZXm(y)]·j[ ˆZXm(y)]′
·(j−|i|)ω
n−p
j=1 ω′[ ˆZXm(y)]·j[ ˆZXm(y)]′
·jω
(29)
=
n−p
j=|i|+1 ¯ω′
m[Q′ ˆZX(y)]·j[Q′ ˆZX(y)]′
·(j−|i|)¯ωm
n−p
j=1 ¯ω′m[Q′ ˆZX(y)]·j[Q′ ˆZX(y)]′
·j ¯ωm
, (29) for |i| = 1, . . . , n−p−1, where ¯ωm = Dmω/∥Dmω∥. Clearly ¯ωm →(0, 1, 0, . . . , 0). Since the second row of Q′ ˆZ(e+, ˜
X) contains by construction precisely one nonzero
entry, it follows that the limit of (29) must be 0 for i = 1, . . . , n −p −1. Furthermore we have ¯σ0,Xm(e−) →∞. It immediately follows from w(0) = 1
and the definition of M that M(e−) is well defined for m large and that it
converges to 0 as m →∞. The remaining part of the proposition is obvious. g
g
3) Let ˜X ∈˜X0 and assume that X = (e+, ˜X) satisfies g∗
κ,M,p(., X, R) ̸≡0. Obviously, e+ ∈span(X). Note that ˆβX(e+) = e1(k). The first column of R
is non-zero. Therefore RˆβX(e+) ̸= 0. Thus we can (since Assumption 2 holds)
apply Part 4 of Theorem 4.2. Finite sample properties of prewhitened F-type tests 2157 Appendix D: Proofs of results in Section 5 We begin with the proof of the first statement. We
note that for λRn×k- almost every X ∈X0 Case 3 of Theorem 5.4 applies:
Since by assumption (k + 3)(p∗+ 2) + p −1 ≤n and by definition p∗≥1,
we have k + 2 < n. Therefore, the set of matrices X in Rn×k such that
det((X, e+, e−)(X, e+, e−)′) = 0 holds is a λRn×k- null set. Hence, e+, e−/∈MX
and rank((X, e+, e−)) = k + 2 holds for λRn×k- almost every X ∈X0. It re-
mains to verify that g∗
κ, ¯
M,p(., ¯X, ¯R) ̸≡0 for almost every X ∈X0, where
¯X = ¯X(X) = (X, e+, e−), ¯R = (R, 0, 0) and ¯
M is constructed as outlined in
Theorem 5.4. For that it suffices to find a matrix X ∈Rn×k and a vector y ∈Rn
such that g∗
κ, ¯
M,p(y, ¯X, ¯R) ̸= 0. To see this note that the triple κ, ¯
M, p satis-
fies Assumption 1 w.r.t. (the dimensions of) ¯X (cf. the proof of Theorem 5.4). Therefore, Lemma 3.10 shows that (y, X) →g∗
κ, ¯
M,p(y, ¯X, ¯R) is a multivariate
polynomial. If we can find a matrix X and a vector y as above, this implies that
the multivariate polynomial (y, X) →g∗
κ, ¯
M,p(y, ¯X, ¯R) is not the zero polynomial,
and therefore the zero set of this multivariate polynomial is a λRn×Rn×k- null
set. It follows that for λRn×k- almost every X we must have g∗
κ, ¯
M,p(., ¯X, ¯R) ̸≡0
[Assuming the opposite, there exists a set A ∈B(Rn×k) of positive λRn×k-
measure such that g∗
κ, ¯
M,p(., ¯X, ¯R) ≡0 for every X ∈A, which implies that
Rn × A ⊆Rn × Rn×k is a subset of the zero set of (y, X) →g∗
M(y, ¯X, ¯R). But
clearly Rn×A has positive Lebesgue measure, a contradiction.]. In the following
we shall construct such a pair (y, X) as above: Let δ ̸= 0 and define w1, w2, v(δ) ∈Rk+4 as w1 = (1, 1, . . . , 1)′, w2 =
(−1, −1, 1, . . . , 1)′ and v(δ) = (−δ, δ, −(k +1), 1, . . . , 1, 1)′. By construction v(δ)
is orthogonal to w1 and w2. Appendix D: Proofs of results in Section 5 Since the assumptions concerning C in the present theorem
are the same as in Proposition 5.2, we see from the proof of this proposition that
it suffices to verify that ˆΩκ, ¯
M,p, ¯
X and (Ik, 0)ˆβ ¯
X satisfy Assumption 5 in Preiner-
storfer and P¨otscher (2016), that ˆΩκ, ¯
M,p, ¯
X is almost everywhere positive definite
(implying that Assumptions 6 and 7 in Preinerstorfer and P¨otscher (2016) are
satisfied), and that the invariance condition in Equation (34) of Preinerstorfer
and P¨otscher (2016) is satisfied by ¯T. By definition, ˆΩκ, ¯
M,p, ¯
X is the estimator
one would obtain following Steps 1-3 of the construction in Section 3 based on
κ, ¯
M and p, if ¯X was the underlying design matrix (observe that ¯X is of full
column rank) and ( ¯R, r) was the hypothesis to be tested. By assumption, the
triple κ, M, p satisfies Assumption 1 w.r.t. the (dimensions k and n of the)
design matrix X and additionally 1 ≤p ≤n/(k + 3) holds. From 1 ≤¯k −k ≤2,
and the definition of
¯
M it follows that the triple κ,
¯
M, p satisfies Assump-
tion 1 w.r.t. (the dimensions ¯k and n of) ¯X. Furthermore, it is assumed that
g∗
κ, ¯
M,p(., ¯X, ¯R) ̸≡0. Therefore, we can apply Lemma C.1, acting as if ¯X was the 2158 D. Preinerstorfer underlying design matrix, to conclude that ˆβ ¯
X and ˆΩκ, ¯
M,p, ¯
X satisfy Assumption
5 in Preinerstorfer and P¨otscher (2016) with N = N(ˆΩκ, ¯
M,p, ¯
X) and k replaced
by ¯k, X replaced by ¯X and M replaced by ¯M = span( ¯X). Furthermore, Lemma
C.1 shows that ˆΩκ, ¯
M,p, ¯
X is almost everywhere positive definite. We now apply
Part 1 of Proposition 5.23 in Preinerstorfer and P¨otscher (2016) to obtain that
ˆΩκ, ¯
M,p, ¯
X and (Ik, 0)ˆβ ¯
X satisfy (the original) Assumption 5 in Preinerstorfer and
P¨otscher (2016), and that the invariance condition is satisfied. To this end, it
suffices to verify that in each of the four cases we have span(J(C))∩M ⊆M0−μ0
(for some arbitrary μ0 ∈M0). This is obvious in the first three cases. For Case
4 we can use exactly the same argument as in the proof of Part 4 in Theorem
3.8 in Preinerstorfer and P¨otscher (2016). Proof of Proposition 5.5. Appendix D: Proofs of results in Section 5 , vk+4(δ) ¯W ′
k+4·, 0k+2,p−1, 0k+2,n−n∗),
(31) (31) where we recall that all coordinates of v(δ) are nonzero and therefore ˆV ¯
X(y(δ))
has precisely k + 4 nonzero columns. We now intend to apply Lemma B.1 with
t = k + 3, acting as if ¯X ∈Rn×(k+2) was the underlying design, ( ¯R, r) was the
hypothesis to be tested and with the triple κ, ¯
M, p which obviously satisfies
Assumption 1 with respect to ¯X (a matrix with k+2 columns), since by assump-
tion we have 1 ≤p ≤
n
k+3 (Note that due to interpreting ¯X as the underlying
design, k +2 corresponds to the ‘k’ in Lemma B.1). We note first that removing
the first or last row of ¯W does not reduce its rank, because v(δ) (a vector all
coordinates of which are nonzero) is orthogonal to every column of this matrix
(cf. the argument in the beginning of the proof of Lemma B.1). Using p∗≥p
we hence see that Assumptions (A1)-(A3) in Lemma B.1 are satisfied by con-
struction. Now consider the case M ∈MKV . By definition ¯
M is an element of
MKV (acting as if ¯X was the underlying design matrix). Therefore, Condition
(CKV) is satisfied for δ ̸= 0 arbitrary, and g∗
κ, ¯
M,p(y(δ), ¯X, ¯R) ̸= 0 follows. Con-
sider the case where M ∈MAM. Since n −n∗≥1MAM (M) = 1 and because of
jk+4 = n∗−p+1 it follows that n−jk+4 > p−1. Therefore, Condition (CAM) in
Lemma B.1 is satisfied and therefore g∗
κ, ¯
M,p(y(δ), ¯X, ¯R) ̸= 0 for δ ̸= 0 arbitrary. It remains to consider the case where M ∈MNW . It suffices to find a δ∗̸= 0
such that ¯
M(y(δ∗)) is well defined (see the proof of Part 2 of Lemma B.1). The
latter statement is equivalent to the denominator in the fraction appearing in where we recall that all coordinates of v(δ) are nonzero and therefore ˆV ¯
X(y(δ))
has precisely k + 4 nonzero columns. Appendix D: Proofs of results in Section 5 Noting that [w1]i = [w2]i for i ≥3 a dimension-
ality argument implies existence of k normalized vectors w3, . . . , wk+2 ∈Rk+4,
that are functionally independent of δ, linearly independent and orthogonal to
e1(k + 4), e2(k + 4), w1, w2 and v(δ) (for every δ ̸= 0). Recall that e1(k + 4)
and e2(k + 4) are the first two elements of the canonical basis of Rk+4. Hence,
the first two coordinates of wi for i = 3, . . . , k + 2 are zero. These orthogonality
properties readily imply Πspan(w3,...,wk+2,w1,w2)⊥v(δ) = Πspan(w3+w1,...,wk+2+w1,w1,w2)⊥v(δ) = v(δ) (30)
and rank( ¯W) = k + 2 for ¯W = (w3 + w1, . . . , wk+2 + w1, w1, w2). Inserting zero Πspan(w3,...,wk+2,w1,w2)⊥v(δ) = Πspan(w3+w1,...,wk+2+w1,w1,w2)⊥v(δ) = v(δ) (30)
and rank( ¯W) = k + 2 for ¯W = (w3 + w1, . . . , wk+2 + w1, w1, w2). Inserting zero 2159 Finite sample properties of prewhitened F-type tests coordinates and rows, respectively, we shall now suitably embed v(δ) ∈Rk+4
and W = (w3 + w1, . . . , wk+2 + w1) ∈Rk+4×k into Rn and Rn×k. Define y(δ) as (v1(δ), 01,p∗+1, v2(δ), 01,p∗, v3(δ), 01,p∗+1, v4(δ), 01,p∗+1, . . . . . . , vk+4(δ), 01,p−1, 01,n−n∗)′ . . . , vk+4(δ), 01,p−1, 01,n−n∗)′ and X as (W ′
1·, 0k,p∗+1, W ′
2·, 0k,p∗, W ′
3·, 0k,p∗+1, W ′
4·, 0k,p∗+1, . . . , W ′
(k+4)·, 01,p−1, 01,n−n∗)′, where n∗= (k+3)(p∗+2)+p−1, a number that does not exceed n by assumption. We emphasize that by construction X does not depend on δ. Furthermore, if we
delete from e+ and e−those coordinates that correspond to the zero coordinates
that have been inserted to obtain y(δ) from v(δ), we obtain the vectors w1
and w2. Therefore, it follows from Equation (30) that y(δ) is orthogonal to
span( ¯X) = span((X, e+, e−)), that rank(X, e+, e−) = rank( ¯W) = k + 2 and
that for every δ ̸= 0 we have ˆu ¯
X(y(δ)) = y(δ). ˆu ¯
X(y(δ)) = y(δ). As an immediate consequence we obtain ˆV ¯
X(y(δ)) = ¯X′ diag(y(δ))
= (v1(δ) ¯W ′
1·, 0k+2,p∗+1, v2(δ) ¯W ′
2·, 0k+2,p∗, v3(δ) ¯W ′
3·, 0k+2,p∗+1, . . . . . . Appendix D: Proofs of results in Section 5 By definition ¯ω = (ω′, 0, 0)′ and we recall that ˆZ ¯
X = ˆVp, ¯
X(y(δ)) which implies
via Equation (31) that (ω′, 0, 0) ˆZ ¯
X(y(δ)) equals By definition ¯ω = (ω′, 0, 0)′ and we recall that ˆZ ¯
X = ˆVp, ¯
X(y(δ)) which implies
via Equation (31) that (ω′, 0, 0) ˆZ ¯
X(y(δ)) equals (0k,p∗−p+2, v2(δ)ω′W ′
2·, 0k,p∗, v3(δ)ω′W ′
3·, 0k,p∗+1, . . . . . . , vk+4(δ)ω′W ′
k+4·, 0k,p−1, 0k,n−n∗). . . . , vk+4(δ)ω′W ′
k+4·, 0k,p−1, 0k,n−n∗). The only coordinate of this vector that depends on δ is v2(δ)ω′W ′
2· = δ k
i=1 ωi,
the latter equation following from v2(δ) = δ and W2· = (1, . . . , 1). Since
k
j=1 ωi > 0, the denominator appearing in the definition of M(y(δ)) inter-
preted as a function of δ is now seen to be a polynomial of degree 2 in δ. Hence,
there must exist a δ∗̸= 0 such that the denominator does not vanish. It follows
that g∗
κ, ¯
M,p(y(δ∗), ¯X, ¯R) ̸= 0. ,
,p
Concerning the second statement we observe that (k + 2)(p∗+ 2) + p −1 ≤n
implies k + 1 < n, and therefore we have rank( ˜X, e+, e−) = k + 1 for λRn×(k−1)-
almost every ˜X ∈˜X0. By assumption the first column of R is zero. Therefore, for
λRn×(k−1)-almost every X = (e+, ˜X) ∈Rn×k we have e+ ∈MX, RˆβX(e+) = 0
and e−/∈MX, i.e., for λRn×(k−1)-almost every (e+, ˜X) ∈Rn×k Scenario (1)
in Theorem 5.4 applies. As above, it suffices to construct a pair y ∈Rn and
˜X ∈Rn×(k−1) (recall that k ≥2), such that g∗
κ, ¯
M,p(y, ¯X, ¯R) ̸= 0, where ¯X =
(e+, ˜X, e−) ∈Rn×(k+1) and ¯R = (R, 0). Here, the matrix ˜X is n × (k −1)
dimensional. By assumption k♯= k −1 obviously satisfies (k♯+ 3)(p∗+ 2) +
p −1 + 1MAM (M) ≤n. To construct the matrix ˜X we can thus use the same
argument as was used to construct X in the proof of the first statement (k♯
replacing k). The matrix ¯X so obtained has (after a permutation of its columns)
the same structure as has the matrix ¯X constructed in the proof of the first
statement. Appendix D: Proofs of results in Section 5 We can therefore use almost the same arguments to conclude that
g∗
κ, ¯
M,p(y(δ∗), ¯X, ¯R) ̸= 0 for some δ∗̸= 0 and y(δ) as constructed in the first part
of the proof. Appendix D: Proofs of results in Section 5 We now intend to apply Lemma B.1 with
t = k + 3, acting as if ¯X ∈Rn×(k+2) was the underlying design, ( ¯R, r) was the
hypothesis to be tested and with the triple κ, ¯
M, p which obviously satisfies
Assumption 1 with respect to ¯X (a matrix with k+2 columns), since by assump-
tion we have 1 ≤p ≤
n
k+3 (Note that due to interpreting ¯X as the underlying
design, k +2 corresponds to the ‘k’ in Lemma B.1). We note first that removing
the first or last row of ¯W does not reduce its rank, because v(δ) (a vector all
coordinates of which are nonzero) is orthogonal to every column of this matrix
(cf. the argument in the beginning of the proof of Lemma B.1). Using p∗≥p
we hence see that Assumptions (A1)-(A3) in Lemma B.1 are satisfied by con-
struction. Now consider the case M ∈MKV . By definition ¯
M is an element of
MKV (acting as if ¯X was the underlying design matrix). Therefore, Condition
(CKV) is satisfied for δ ̸= 0 arbitrary, and g∗
κ, ¯
M,p(y(δ), ¯X, ¯R) ̸= 0 follows. Con-
sider the case where M ∈MAM. Since n −n∗≥1MAM (M) = 1 and because of
jk+4 = n∗−p+1 it follows that n−jk+4 > p−1. Therefore, Condition (CAM) in
Lemma B.1 is satisfied and therefore g∗
κ, ¯
M,p(y(δ), ¯X, ¯R) ̸= 0 for δ ̸= 0 arbitrary. It remains to consider the case where M ∈MNW . It suffices to find a δ∗̸= 0
such that ¯
M(y(δ∗)) is well defined (see the proof of Part 2 of Lemma B.1). The
latter statement is equivalent to the denominator in the fraction appearing in D. Preinerstorfer 2160 the definition of ¯
M(y(δ∗)) being nonzero, i.e., the definition of ¯
M(y(δ∗)) being nonzero, i.e., n−p−1
i=−(n−p−1)
w(i)¯σi(y(δ∗)) ̸= 0,
where for |i| = 0, . . . , n −p −1
¯σi(y(δ∗)) = (n −p)−1
n−p
j=|i|+1
¯ω′[ ˆZ ¯
X(y(δ∗))]·j[ ˆZ ¯
X(y(δ∗))]′
·(j−|i|)¯ω. n−p−1
i=−(n−p−1)
w(i)¯σi(y(δ∗)) ̸= 0, where for |i| = 0, . . . , n −p −1 ¯σi(y(δ∗)) = (n −p)−1
n−p
j=|i|+1
¯ω′[ ˆZ ¯
X(y(δ∗))]·j[ ˆZ ¯
X(y(δ∗))]′
·(j−|i|)¯ω. Appendix E: Tables In Tables 1, 2, and 3 rows correspond to ρ and columns correspond to β1,
whereas in Tables 4, 5, 6, 7, 8, 9, rows correspond to ρ and columns correspond
to β2. Finite sample properties of prewhitened F-type tests Finite sample properties of prewhitened F-type tests 2161 Table 1
Rejection probabilities for test (i) in Example 1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.5
2
-0.9999
0.18
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.999
0.16
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.99
0.13
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.95
0.08
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.9
0.07
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.8
0.06
0.85
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.7
0.06
0.66
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.6
0.05
0.53
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.5
0.05
0.43
0.93
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.4
0.05
0.33
0.86
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.3
0.06
0.29
0.79
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.2
0.06
0.25
0.69
0.95
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.1
0.06
0.22
0.60
0.92
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0
0.06
0.19
0.54
0.85
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.1
0.06
0.16
0.45
0.79
0.95
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.2
0.06
0.15
0.40
0.69
0.91
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.3
0.06
0.13
0.32
0.60
0.83
0.95
0.99
1.00
1.00
1.00
1.00
1.00
1.00
0.4
0.07
0.12
0.28
0.52
0.74
0.90
0.97
1.00
1.00
1.00
1.00
1.00
1.00
0.5
0.07
0.12
0.25
0.44
0.65
0.82
0.93
0.98
0.99
1.00
1.00
1.00
1.00
0.6
0.07
0.11
0.21
0.37
0.54
0.72
0.86
0.93
0.98
0.99
1.00
1.00
1.00
0.7
0.09
0.10
0.18
0.30
0.41
0.59
0.72
0.85
0.91
0.96
0.98
1.00
1.00
0.8
0.09
0.13
0.18
0.23
0.34
0.46
0.58
0.68
0.78
0.85
0.91
1.00
1.00
0.9
0.15
0.16
0.18
0.23
0.29
0.37
0.43
0.53
0.59
0.67
0.72
0.92
0.98
0.95
0.23
0.22
0.25
0.29
0.32
0.37
0.40
0.45
0.49
0.54
0.61
0.81
0.93
0.99
0.48
0.47
0.48
0.50
0.49
0.52
0.55
0.58
0.59
0.60
0.65
0.77
0.84
0.999
0.75
0.76
0.76
0.78
0.76
0.78
0.78
0.78
0.80
0.81
0.83
0.87
0.92
0.9999
0.89
0.89
0.90
0.89
0.89
0.91
0.91
0.91
0.91
0.93
0.92
0.95
0.97
Table 2
Rejection probabilities for test (ii) in Example 1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.5
2
-0.9999
0.19
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.999
0.17
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.99
0.14
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.95
0.09
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.9
0.09
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.8
0.07
0.86
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.7
0.08
0.68
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.6
0.07
0.55
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.5
0.07
0.46
0.94
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.4
0.07
0.37
0.88
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.3
0.08
0.32
0.80
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.2
0.09
0.28
0.71
0.96
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.1
0.08
0.24
0.63
0.93
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0
0.07
0.21
0.56
0.86
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.1
0.08
0.17
0.47
0.80
0.95
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.2
0.07
0.16
0.42
0.70
0.92
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.3
0.06
0.15
0.34
0.63
0.84
0.95
0.99
1.00
1.00
1.00
1.00
1.00
1.00
0.4
0.07
0.13
0.29
0.54
0.75
0.91
0.98
1.00
1.00
1.00
1.00
1.00
1.00
0.5
0.07
0.13
0.25
0.45
0.66
0.83
0.94
0.98
0.99
1.00
1.00
1.00
1.00
0.6
0.07
0.11
0.22
0.38
0.55
0.73
0.87
0.94
0.98
0.99
1.00
1.00
1.00
0.7
0.10
0.11
0.19
0.30
0.41
0.59
0.72
0.85
0.91
0.96
0.99
1.00
1.00
0.8
0.09
0.13
0.18
0.24
0.34
0.46
0.59
0.69
0.78
0.85
0.91
1.00
1.00
0.9
0.15
0.16
0.18
0.23
0.29
0.38
0.43
0.53
0.59
0.67
0.72
0.92
0.98
0.95
0.23
0.22
0.25
0.29
0.32
0.36
0.40
0.46
0.49
0.54
0.61
0.81
0.93
0.99
0.47
0.47
0.48
0.50
0.49
0.51
0.55
0.58
0.59
0.60
0.64
0.77
0.84
0.999
0.75
0.76
0.76
0.78
0.76
0.78
0.78
0.78
0.80
0.81
0.83
0.87
0.92
0.9999
0.89
0.89
0.90
0.89
0.89
0.91
0.91
0.91
0.91
0.93
0.92
0.95
0.97 Table 1
Rejection probabilities for test (i) in Example 1 2162 D. Finite sample properties of prewhitened F-type tests Preinerstorfer Table 3
Rejection probabilities for test (iii) in Example 1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.5
2
-0.9999
0.02
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.999
0.02
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.99
0.02
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.95
0.03
0.97
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.9
0.04
0.86
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.8
0.04
0.62
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.7
0.05
0.44
0.92
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.6
0.04
0.34
0.84
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.5
0.04
0.29
0.75
0.96
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.4
0.04
0.23
0.67
0.90
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.3
0.05
0.20
0.57
0.85
0.97
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.2
0.05
0.17
0.48
0.79
0.94
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.1
0.04
0.15
0.42
0.73
0.90
0.96
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0
0.04
0.13
0.36
0.65
0.85
0.95
0.98
0.99
1.00
1.00
1.00
1.00
1.00
0.1
0.05
0.11
0.30
0.56
0.78
0.90
0.97
0.99
1.00
1.00
1.00
1.00
1.00
0.2
0.05
0.10
0.25
0.49
0.70
0.84
0.93
0.98
0.99
1.00
1.00
1.00
1.00
0.3
0.04
0.10
0.22
0.42
0.62
0.79
0.91
0.95
0.98
0.99
1.00
1.00
1.00
0.4
0.05
0.09
0.18
0.35
0.53
0.70
0.84
0.92
0.96
0.98
0.99
1.00
1.00
0.5
0.04
0.07
0.16
0.28
0.44
0.62
0.75
0.86
0.92
0.96
0.97
1.00
1.00
0.6
0.05
0.07
0.14
0.24
0.35
0.51
0.64
0.76
0.84
0.91
0.94
1.00
1.00
0.7
0.05
0.05
0.10
0.18
0.26
0.37
0.50
0.63
0.73
0.81
0.87
0.99
1.00
0.8
0.05
0.07
0.10
0.14
0.20
0.27
0.36
0.46
0.57
0.64
0.71
0.95
0.99
0.9
0.08
0.08
0.09
0.11
0.15
0.20
0.24
0.32
0.36
0.42
0.48
0.74
0.88
0.95
0.11
0.11
0.13
0.14
0.17
0.20
0.22
0.26
0.27
0.32
0.36
0.58
0.74
0.99
0.27
0.25
0.28
0.30
0.29
0.31
0.33
0.36
0.38
0.39
0.43
0.56
0.64
0.999
0.56
0.56
0.58
0.58
0.56
0.59
0.61
0.60
0.62
0.63
0.66
0.74
0.81
0.9999
0.78
0.77
0.77
0.77
0.78
0.80
0.81
0.81
0.82
0.83
0.83
0.89
0.92
Table 4
Rejection probabilities for test (i) in Example 2
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.5
2
-0.9999
0.88
0.88
0.88
0.88
0.88
0.88
0.88
0.87
0.88
0.88
0.88
0.88
0.89
-0.999
0.68
0.68
0.69
0.69
0.69
0.70
0.71
0.71
0.72
0.73
0.74
0.77
0.81
-0.99
0.50
0.51
0.51
0.53
0.55
0.58
0.61
0.63
0.66
0.69
0.71
0.83
0.89
-0.95
0.24
0.25
0.29
0.34
0.41
0.47
0.54
0.61
0.68
0.73
0.78
0.93
0.98
-0.9
0.17
0.19
0.25
0.33
0.42
0.53
0.64
0.73
0.80
0.86
0.90
0.99
1.00
-0.8
0.12
0.15
0.25
0.39
0.55
0.70
0.82
0.90
0.95
0.97
0.99
1.00
1.00
-0.7
0.10
0.15
0.30
0.50
0.70
0.84
0.93
0.98
0.99
1.00
1.00
1.00
1.00
-0.6
0.09
0.16
0.35
0.59
0.79
0.92
0.98
0.99
1.00
1.00
1.00
1.00
1.00
-0.5
0.08
0.16
0.39
0.67
0.87
0.96
0.99
1.00
1.00
1.00
1.00
1.00
1.00
-0.4
0.07
0.17
0.44
0.75
0.92
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.3
0.07
0.18
0.46
0.77
0.94
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.2
0.07
0.19
0.50
0.80
0.95
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.1
0.07
0.20
0.54
0.85
0.97
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0
0.07
0.19
0.53
0.84
0.97
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.1
0.07
0.20
0.52
0.83
0.97
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.2
0.07
0.19
0.51
0.82
0.96
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.3
0.07
0.18
0.46
0.76
0.93
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.4
0.08
0.18
0.45
0.74
0.92
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.5
0.08
0.16
0.39
0.67
0.87
0.96
0.99
1.00
1.00
1.00
1.00
1.00
1.00
0.6
0.09
0.16
0.35
0.60
0.80
0.93
0.98
0.99
1.00
1.00
1.00
1.00
1.00
0.7
0.10
0.15
0.29
0.50
0.70
0.84
0.93
0.97
0.99
1.00
1.00
1.00
1.00
0.8
0.12
0.15
0.25
0.40
0.56
0.71
0.83
0.91
0.95
0.98
0.99
1.00
1.00
0.9
0.16
0.19
0.24
0.33
0.43
0.54
0.64
0.73
0.80
0.86
0.91
0.98
1.00
0.95
0.22
0.24
0.26
0.32
0.38
0.45
0.53
0.60
0.67
0.73
0.78
0.93
0.98
0.99
0.28
0.28
0.31
0.34
0.38
0.43
0.48
0.53
0.58
0.63
0.68
0.84
0.91
0.999
0.35
0.35
0.37
0.39
0.42
0.46
0.50
0.54
0.58
0.62
0.65
0.79
0.87
0.9999
0.37
0.37
0.39
0.41
0.44
0.48
0.52
0.56
0.60
0.63
0.67
0.80
0.88 Table 3
Rejection probabilities for test (iii) in Example 1 Table 3 Finite sample properties of prewhitened F-type tests Finite sample properties of prewhitened F-type tests 2163 Table 5
Rejection probabilities for the adjusted version of test (i) in Example 2
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.5
2
-0.9999
0.06
0.06
0.07
0.09
0.12
0.15
0.18
0.22
0.26
0.30
0.34
0.52
0.65
-0.999
0.06
0.06
0.08
0.09
0.12
0.16
0.19
0.23
0.27
0.31
0.35
0.53
0.66
-0.99
0.05
0.05
0.07
0.09
0.12
0.15
0.20
0.24
0.28
0.33
0.38
0.59
0.73
-0.95
0.04
0.04
0.06
0.09
0.13
0.18
0.24
0.30
0.37
0.43
0.49
0.74
0.87
-0.9
0.03
0.04
0.06
0.10
0.16
0.22
0.30
0.38
0.47
0.55
0.62
0.85
0.94
-0.8
0.03
0.05
0.10
0.18
0.30
0.43
0.57
0.69
0.78
0.85
0.90
0.99
1.00
-0.7
0.04
0.06
0.15
0.30
0.49
0.67
0.81
0.90
0.95
0.97
0.99
1.00
1.00
-0.6
0.04
0.07
0.20
0.41
0.63
0.81
0.91
0.97
0.99
1.00
1.00
1.00
1.00
-0.5
0.04
0.09
0.25
0.52
0.76
0.91
0.97
0.99
1.00
1.00
1.00
1.00
1.00
-0.4
0.04
0.10
0.31
0.61
0.84
0.95
0.99
1.00
1.00
1.00
1.00
1.00
1.00
-0.3
0.03
0.10
0.32
0.63
0.86
0.96
0.99
1.00
1.00
1.00
1.00
1.00
1.00
-0.2
0.04
0.12
0.38
0.71
0.91
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.1
0.03
0.12
0.41
0.75
0.94
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0
0.04
0.12
0.40
0.74
0.93
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.1
0.04
0.12
0.39
0.73
0.93
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.2
0.03
0.11
0.36
0.69
0.91
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.3
0.03
0.10
0.31
0.61
0.84
0.95
0.99
1.00
1.00
1.00
1.00
1.00
1.00
0.4
0.03
0.10
0.30
0.60
0.84
0.95
0.99
1.00
1.00
1.00
1.00
1.00
1.00
0.5
0.03
0.09
0.25
0.50
0.74
0.89
0.96
0.99
1.00
1.00
1.00
1.00
1.00
0.6
0.03
0.06
0.18
0.37
0.58
0.77
0.88
0.95
0.98
0.99
1.00
1.00
1.00
0.7
0.03
0.06
0.15
0.30
0.49
0.67
0.81
0.90
0.95
0.98
0.99
1.00
1.00
0.8
0.03
0.05
0.10
0.20
0.32
0.46
0.60
0.72
0.81
0.88
0.93
0.99
1.00
0.9
0.03
0.05
0.07
0.12
0.18
0.25
0.34
0.42
0.51
0.59
0.67
0.89
0.97
0.95
0.04
0.04
0.06
0.09
0.12
0.17
0.22
0.28
0.34
0.40
0.46
0.70
0.85
0.99
0.05
0.05
0.06
0.09
0.11
0.14
0.18
0.22
0.26
0.31
0.35
0.56
0.70
0.999
0.06
0.06
0.07
0.09
0.11
0.14
0.17
0.20
0.24
0.27
0.31
0.48
0.62
0.9999
0.06
0.06
0.07
0.09
0.11
0.14
0.17
0.21
0.24
0.28
0.32
0.49
0.62
Table 6
Rejection probabilities for test (ii) in Example 2
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.5
2
-0.9999
0.87
0.88
0.88
0.87
0.88
0.87
0.87
0.87
0.87
0.88
0.88
0.88
0.89
-0.999
0.68
0.68
0.68
0.68
0.69
0.70
0.70
0.71
0.72
0.73
0.73
0.77
0.80
-0.99
0.50
0.50
0.51
0.53
0.55
0.58
0.60
0.63
0.66
0.69
0.71
0.82
0.89
-0.95
0.25
0.26
0.29
0.34
0.41
0.47
0.55
0.62
0.68
0.73
0.79
0.93
0.98
-0.9
0.17
0.20
0.25
0.33
0.43
0.54
0.64
0.73
0.80
0.86
0.90
0.99
1.00
-0.8
0.13
0.16
0.26
0.40
0.56
0.71
0.82
0.90
0.95
0.98
0.99
1.00
1.00
-0.7
0.10
0.16
0.31
0.51
0.71
0.85
0.94
0.98
0.99
1.00
1.00
1.00
1.00
-0.6
0.10
0.17
0.37
0.61
0.81
0.93
0.98
1.00
1.00
1.00
1.00
1.00
1.00
-0.5
0.09
0.18
0.41
0.69
0.88
0.97
0.99
1.00
1.00
1.00
1.00
1.00
1.00
-0.4
0.09
0.19
0.46
0.76
0.93
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.3
0.08
0.20
0.49
0.78
0.94
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.2
0.08
0.21
0.52
0.82
0.96
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.1
0.08
0.22
0.56
0.86
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0
0.08
0.21
0.55
0.85
0.97
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.1
0.08
0.21
0.54
0.85
0.97
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.2
0.08
0.21
0.53
0.83
0.97
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.3
0.08
0.19
0.48
0.78
0.94
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.4
0.09
0.20
0.47
0.76
0.93
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.5
0.09
0.18
0.42
0.69
0.88
0.97
0.99
1.00
1.00
1.00
1.00
1.00
1.00
0.6
0.10
0.17
0.37
0.62
0.82
0.93
0.98
0.99
1.00
1.00
1.00
1.00
1.00
0.7
0.11
0.16
0.31
0.52
0.71
0.85
0.94
0.98
0.99
1.00
1.00
1.00
1.00
0.8
0.12
0.16
0.26
0.41
0.57
0.72
0.84
0.91
0.95
0.98
0.99
1.00
1.00
0.9
0.16
0.19
0.24
0.33
0.43
0.54
0.65
0.73
0.81
0.86
0.91
0.98
1.00
0.95
0.22
0.24
0.26
0.32
0.38
0.45
0.53
0.60
0.67
0.73
0.78
0.93
0.97
0.99
0.27
0.28
0.30
0.34
0.38
0.43
0.48
0.53
0.58
0.63
0.67
0.84
0.91
0.999
0.35
0.35
0.37
0.39
0.42
0.46
0.50
0.54
0.58
0.62
0.65
0.79
0.87
0.9999
0.37
0.37
0.39
0.41
0.44
0.48
0.52
0.55
0.60
0.63
0.67
0.80
0.87 2164 D. Acknowledgements Financial support by the Austrian Science Fund (FWF) P27398 and by the Dan-
ish National Research Foundation (Grant DNRF 78, CREATES) is gratefully
acknowledged. I am grateful to Benedikt M. P¨otscher for many helpful discus-
sions and for feedback on an earlier version of this manuscript. Furthermore I
would like to thank a referee for helpful comments. Finite sample properties of prewhitened F-type tests Preinerstorfer Table 7
Rejection probabilities for the adjusted version of test (ii) in Example 2
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.5
2
-0.9999
0.06
0.06
0.08
0.09
0.12
0.15
0.18
0.22
0.26
0.29
0.34
0.52
0.64
-0.999
0.05
0.06
0.07
0.09
0.11
0.15
0.18
0.22
0.26
0.30
0.34
0.51
0.65
-0.99
0.05
0.05
0.07
0.09
0.12
0.15
0.19
0.24
0.28
0.33
0.37
0.58
0.73
-0.95
0.04
0.04
0.06
0.09
0.13
0.18
0.24
0.30
0.36
0.42
0.49
0.73
0.87
-0.9
0.03
0.04
0.06
0.10
0.16
0.22
0.30
0.38
0.46
0.54
0.61
0.85
0.94
-0.8
0.03
0.05
0.10
0.18
0.30
0.44
0.57
0.69
0.79
0.85
0.90
0.99
1.00
-0.7
0.03
0.06
0.15
0.29
0.47
0.66
0.80
0.89
0.95
0.97
0.99
1.00
1.00
-0.6
0.04
0.07
0.20
0.40
0.62
0.80
0.91
0.97
0.99
1.00
1.00
1.00
1.00
-0.5
0.03
0.08
0.24
0.49
0.74
0.90
0.97
0.99
1.00
1.00
1.00
1.00
1.00
-0.4
0.04
0.10
0.30
0.59
0.83
0.95
0.99
1.00
1.00
1.00
1.00
1.00
1.00
-0.3
0.04
0.10
0.32
0.62
0.85
0.96
0.99
1.00
1.00
1.00
1.00
1.00
1.00
-0.2
0.04
0.11
0.35
0.67
0.89
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.1
0.03
0.11
0.37
0.71
0.92
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0
0.04
0.11
0.38
0.72
0.92
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.1
0.04
0.12
0.37
0.70
0.92
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.2
0.03
0.10
0.33
0.66
0.89
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.3
0.03
0.09
0.28
0.57
0.81
0.94
0.99
1.00
1.00
1.00
1.00
1.00
1.00
0.4
0.04
0.10
0.30
0.59
0.83
0.95
0.99
1.00
1.00
1.00
1.00
1.00
1.00
0.5
0.03
0.08
0.24
0.49
0.73
0.88
0.96
0.99
1.00
1.00
1.00
1.00
1.00
0.6
0.03
0.06
0.17
0.36
0.57
0.76
0.88
0.94
0.97
0.99
0.99
1.00
1.00
0.7
0.04
0.06
0.15
0.30
0.49
0.67
0.81
0.90
0.95
0.98
0.99
1.00
1.00
0.8
0.03
0.05
0.10
0.19
0.31
0.45
0.59
0.71
0.81
0.88
0.92
0.99
1.00
0.9
0.03
0.05
0.07
0.12
0.18
0.25
0.34
0.42
0.51
0.59
0.67
0.89
0.97
0.95
0.04
0.04
0.06
0.09
0.12
0.17
0.22
0.28
0.34
0.40
0.45
0.70
0.85
0.99
0.05
0.05
0.06
0.09
0.11
0.14
0.18
0.22
0.26
0.31
0.35
0.56
0.70
0.999
0.06
0.06
0.07
0.09
0.11
0.14
0.17
0.20
0.24
0.27
0.31
0.48
0.62
0.9999
0.06
0.06
0.07
0.09
0.11
0.13
0.16
0.20
0.24
0.27
0.31
0.48
0.61
Table 8
Rejection probabilities for test (iii) in Example 2
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.5
2
-0.9999
0.65
0.64
0.65
0.65
0.66
0.66
0.66
0.67
0.68
0.68
0.69
0.72
0.75
-0.999
0.47
0.47
0.47
0.48
0.49
0.50
0.51
0.52
0.53
0.55
0.56
0.62
0.67
-0.99
0.32
0.32
0.33
0.35
0.37
0.40
0.42
0.46
0.48
0.51
0.54
0.67
0.76
-0.95
0.14
0.15
0.18
0.22
0.27
0.33
0.40
0.46
0.52
0.58
0.63
0.83
0.92
-0.9
0.09
0.10
0.14
0.21
0.28
0.37
0.46
0.54
0.62
0.70
0.75
0.92
0.97
-0.8
0.07
0.09
0.17
0.27
0.40
0.53
0.65
0.75
0.83
0.89
0.93
0.99
1.00
-0.7
0.07
0.10
0.20
0.35
0.53
0.68
0.80
0.88
0.93
0.96
0.98
1.00
1.00
-0.6
0.06
0.10
0.24
0.42
0.61
0.77
0.87
0.93
0.97
0.98
0.99
1.00
1.00
-0.5
0.06
0.11
0.27
0.49
0.70
0.84
0.93
0.97
0.99
0.99
1.00
1.00
1.00
-0.4
0.06
0.12
0.31
0.55
0.76
0.89
0.95
0.98
0.99
1.00
1.00
1.00
1.00
-0.3
0.05
0.12
0.32
0.58
0.78
0.90
0.96
0.99
1.00
1.00
1.00
1.00
1.00
-0.2
0.05
0.13
0.34
0.60
0.80
0.92
0.97
0.99
1.00
1.00
1.00
1.00
1.00
-0.1
0.05
0.14
0.38
0.66
0.85
0.94
0.98
1.00
1.00
1.00
1.00
1.00
1.00
0
0.05
0.13
0.37
0.64
0.84
0.94
0.98
0.99
1.00
1.00
1.00
1.00
1.00
0.1
0.05
0.14
0.36
0.64
0.83
0.93
0.98
0.99
1.00
1.00
1.00
1.00
1.00
0.2
0.05
0.13
0.35
0.62
0.82
0.93
0.98
0.99
1.00
1.00
1.00
1.00
1.00
0.3
0.05
0.12
0.31
0.56
0.77
0.90
0.96
0.98
0.99
1.00
1.00
1.00
1.00
0.4
0.06
0.12
0.32
0.56
0.76
0.89
0.95
0.98
0.99
1.00
1.00
1.00
1.00
0.5
0.06
0.12
0.28
0.50
0.70
0.84
0.92
0.96
0.98
0.99
1.00
1.00
1.00
0.6
0.06
0.11
0.24
0.44
0.63
0.78
0.88
0.94
0.97
0.98
0.99
1.00
1.00
0.7
0.07
0.10
0.21
0.36
0.53
0.68
0.80
0.88
0.93
0.96
0.98
1.00
1.00
0.8
0.08
0.10
0.17
0.28
0.42
0.55
0.67
0.77
0.85
0.90
0.94
0.99
1.00
0.9
0.11
0.12
0.17
0.23
0.32
0.41
0.51
0.59
0.67
0.73
0.79
0.93
0.98
0.95
0.13
0.14
0.16
0.20
0.26
0.31
0.38
0.45
0.51
0.57
0.63
0.82
0.91
0.99
0.16
0.17
0.18
0.21
0.24
0.28
0.32
0.37
0.41
0.45
0.50
0.68
0.77
0.999
0.22
0.22
0.24
0.25
0.28
0.32
0.35
0.39
0.42
0.46
0.49
0.64
0.73
0.9999
0.23
0.23
0.25
0.27
0.30
0.33
0.36
0.40
0.44
0.46
0.50
0.64
0.74 Table 7 Finite sample properties of prewhitened F-type tests Finite sample properties of prewhitened F-type tests 2165 Table 9
Rejection probabilities for the adjusted version of test (iii) in Example 2
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1.5
2
-0.9999
0.06
0.06
0.07
0.09
0.11
0.15
0.17
0.21
0.25
0.28
0.32
0.50
0.62
-0.999
0.05
0.06
0.07
0.09
0.11
0.14
0.17
0.21
0.25
0.29
0.32
0.50
0.63
-0.99
0.05
0.05
0.07
0.09
0.12
0.15
0.19
0.24
0.28
0.32
0.37
0.57
0.71
-0.95
0.04
0.04
0.06
0.09
0.13
0.18
0.24
0.30
0.36
0.43
0.49
0.73
0.87
-0.9
0.03
0.04
0.07
0.11
0.16
0.23
0.31
0.39
0.47
0.54
0.61
0.85
0.94
-0.8
0.03
0.05
0.10
0.19
0.31
0.44
0.58
0.70
0.79
0.86
0.90
0.99
1.00
-0.7
0.04
0.06
0.15
0.30
0.49
0.67
0.81
0.90
0.95
0.98
0.99
1.00
1.00
-0.6
0.04
0.07
0.20
0.39
0.61
0.80
0.91
0.97
0.99
1.00
1.00
1.00
1.00
-0.5
0.04
0.09
0.25
0.50
0.75
0.90
0.97
0.99
1.00
1.00
1.00
1.00
1.00
-0.4
0.04
0.10
0.30
0.60
0.83
0.95
0.99
1.00
1.00
1.00
1.00
1.00
1.00
-0.3
0.04
0.10
0.32
0.61
0.85
0.95
0.99
1.00
1.00
1.00
1.00
1.00
1.00
-0.2
0.04
0.12
0.35
0.67
0.89
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
-0.1
0.04
0.12
0.39
0.72
0.92
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0
0.04
0.12
0.38
0.71
0.92
0.99
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.1
0.04
0.12
0.37
0.70
0.91
0.98
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.2
0.03
0.10
0.34
0.66
0.88
0.97
1.00
1.00
1.00
1.00
1.00
1.00
1.00
0.3
0.03
0.10
0.29
0.58
0.82
0.95
0.99
1.00
1.00
1.00
1.00
1.00
1.00
0.4
0.04
0.10
0.29
0.57
0.81
0.94
0.99
1.00
1.00
1.00
1.00
1.00
1.00
0.5
0.04
0.09
0.24
0.48
0.72
0.88
0.96
0.99
1.00
1.00
1.00
1.00
1.00
0.6
0.03
0.07
0.19
0.38
0.59
0.77
0.89
0.95
0.98
0.99
0.99
1.00
1.00
0.7
0.04
0.07
0.15
0.31
0.49
0.67
0.81
0.89
0.95
0.97
0.99
1.00
1.00
0.8
0.03
0.05
0.10
0.20
0.32
0.46
0.60
0.72
0.82
0.88
0.93
0.99
1.00
0.9
0.04
0.05
0.07
0.12
0.18
0.25
0.34
0.42
0.50
0.58
0.66
0.88
0.96
0.95
0.04
0.05
0.06
0.09
0.13
0.17
0.22
0.28
0.34
0.40
0.45
0.70
0.84
0.99
0.05
0.05
0.06
0.08
0.11
0.14
0.17
0.21
0.25
0.30
0.34
0.54
0.68
0.999
0.06
0.06
0.07
0.09
0.11
0.14
0.17
0.20
0.24
0.27
0.30
0.47
0.60
0.9999
0.06
0.06
0.07
0.09
0.11
0.14
0.17
0.20
0.23
0.26
0.31
0.47
0.60 References Andrews, D. W. K. (1991). Heteroskedasticity and Autocorrelation Con-
sistent Covariance Matrix Estimation. Econometrica 59 pp. 817–858. MR1106513 Andrews, D. W. K. and Monahan, J. C. (1992). An Improved Heteroskedas-
ticity and Autocorrelation Consistent Covariance Matrix Estimator. Econo-
metrica 60 pp. 953–966. MR1168742 Bartlett, M. S. (1950). Periodogram Analysis and Continuous Spectra. Biometrika 37 1–16. MR0035934 den Haan, W. J. and Levin, A. T. (1997). A practitioner’s guide to ro-
bust covariance matrix estimation. In Robust Inference, (G. S. Maddala and
C. R. Rao, eds.). Handbook of Statistics 15 299–342. Elsevier. MR1492717 2166 D. Preinerstorfer Grenander, U. and Rosenblatt, M. (1957). Statistical analysis of stationary
time series. John Wiley & Sons, New York. MR0084975 (18,959b) Hannan, E. J. (1957). The Variance of the Mean of a Stationary Pro-
cess. Journal of the Royal Statistical Society. Series B 19 282–285. MR0092333 Hansen, B. E. (1992). Consistent Covariance Matrix Estimation for Dependent
Heterogeneous Processes. Econometrica 60 pp. 967–972. MR1168743 Jansson, M. (2002). Consistent covariance matrix estimation for linear pro-
cesses. Econometric Theory 18 1449–1459. MR1945420 Jansson, M. (2004). The Error in Rejection Probability of Simple Autocorre-
lation Robust Tests. Econometrica 72 937–946. MR2051441 Jowett, G. H. (1955). The Comparison of Means of Sets of Observations from
Sections of Independent Stochastic Series. Journal of the Royal Statistical
Society. Series B 17 208–227. MR0076248 Kiefer, N. M., Vogelsang, T. J. and Bunzel, H. (2000). Simple robust
testing of regression hypotheses. Econometrica 68 695–714. MR1769382 Kiefer,
N. M. and
Vogelsang,
T. J. (2002). Heteroskedasticity-
autocorrelation robust testing using bandwidth equal to sample size. Econo-
metric Theory 18 1350–1366. MR1945416 (2003k:62230) (
)
Kiefer, N. M. and Vogelsang, T. J. (2005). A new asymptotic theory for
heteroskedasticity-autocorrelation robust tests. Econometric Theory 21 1130–
1164. MR2200988 Newey, W. K. and West, K. D. (1987). A simple, positive semidefinite,
heteroskedasticity and autocorrelation consistent covariance matrix. Econo-
metrica 55 703–708. MR890864 (88g:62111) Newey, W. K. and West, K. D. (1994). Automatic Lag Selection in Covari-
ance Matrix Estimation. The Review of Economic Studies 61 pp. 631–653. MR1299308 Okamoto, M. (1973). Distinctness of the eigenvalues of a quadratic form in a
multivariate sample. The Annals of Statistics 763–765. MR0331643 Preinerstorfer, D. and P¨otscher, B. M. (2016). On Size and Power of
Heteroskedasticity and Autocorrelation Robust Tests. Econometric Theory
32 261–358. MR3471120 Preinerstorfer, D. and P¨otscher, B. M. (2017). On the power of invariant
tests for hypotheses on a covariance matrix. References Econometric Theory 33 1–68. MR3574860 Rho, Y. and Shao, X. (2013). Improving the bandwidth-free inference methods
by prewhitening. Journal of Statistical Planning and Inference 143 1912–
1922. MR3095081 Robinson, G. (1979). Conditional properties of statistical procedures. The An-
nals of Statistics 742–755. MR0532239 Schwert, G. (2009). EViews 7 User’s Guide II. Quantitative Micro Software,
LLC, Irvine, California. Sun, Y., Phillips, P. C. B. and Jin, S. (2008). Optimal Bandwidth Selection
in Heteroskedasticity-Autocorrelation Robust Testing. Econometrica 76 175–
194. MR2374985 Finite sample properties of prewhitened F-type tests 2167 Sun, Y., Phillips, P. C. B. and Jin, S. (2011). Power maximization and size
control in heteroskedasticity and autocorrelation robust tests with exponen-
tiated kernels. Econometric Theory 27 1320–1368. MR2868841 R Core Team (2016). R: A Language and Environment for Statistical Com-
puting R Foundation for Statistical Computing, Vienna, Austria. Velasco, C. and Robinson, P. M. (2001). Edgeworth Expansions for Spectral
Density Estimates and Studentized Sample Mean. Econometric Theory 17
497–539. MR1841819 Zeileis, A. (2004). Econometric Computing with HC and HAC Covariance
Matrix Estimators. Journal of Statistical Software 11 1–17. Zeileis, A. and Hothorn, T. (2002). Diagnostic Checking in Regression Re-
lationships. R News 2 7–10. Zhang, X. and Shao, X. (2013). Fixed-smoothing asymptotics for time series. The Annals of Statistics 41 1329–1349. MR3113813
|
https://openalex.org/W2945999031
|
https://europepmc.org/articles/pmc6566999?pdf=render
|
English
| null |
Incorporation and Conduction of Protons in Ca, Sr, Ba-Doped BaLaInO4 with Ruddlesden-Popper Structure
|
Materials
| 2,019
|
cc-by
| 10,928
|
Received: 6 May 2019; Accepted: 20 May 2019; Published: 22 May 2019 Abstract: The new phases BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, Ba2+) with a Ruddlesden-Popper
structure were obtained. It was established that all investigated samples were capable for the water
uptake from the gas phase. The ability of water incorporation was due to not only by the presence of
oxygen vacancies, but also due to the presence of La-O blocks in the structure. The degree of hydration
of the samples was much higher than the concentration of oxygen vacancies and the composition of the
samples appear to be BaLaInO3.42(OH)1.16, BaLa0.9Ca0.1InO3.25(OH)1.4, BaLa0.9Sr0.1InO3.03(OH)1.84,
BaLa0.9Ba0.1InO2.9(OH)2.1. The degree of hydration increased with an increase in the size of the
dopant, i.e., with an increase in the size of the salt blocks. It was proven that doping led to the
increase in the oxygen ionic conductivity. The conductivities for doped samples BaLa0.9M0.1InO3.95
were higher than for undoped composition BaLaInO4 at ~1.5 order of magnitude. The increase in
the conductivity was mainly attributed to the increase of the carrier concentration as a result of the
formation of oxygen vacancies during doping. The proton conductivities of doped samples increased
in the order Ca2+–Sr2+–Ba2+ due to an increase in the concentration of protons. It was established
that all doped samples demonstrated the dominant proton transport below 450 ◦C. Keywords: Ruddlesden-Popper structure; oxygen-ion conductivity; proton conductivity; water uptake Keywords: Ruddlesden-Popper structure; oxygen-ion conductivity; proton conductivity; water uptake Incorporation and Conduction of Protons in Ca, Sr
Ba-Doped BaLaInO4 with
Ruddlesden-Popper Structure Nataliia Tarasova *, Irina Animitsa, Anzhelika Galisheva and Daniil Korona
Institute of Natural Sciences and Mathematics, Ural Federal University, 620000 Yekaterinburg, Russia;
irina.animitsa@urfu.ru (I.A.); jelya95@gmail.com (A.G.); D.V.Korona@urfu.ru (D.K.)
* Correspondence: Natalia.Tarasova@urfu.ru Nataliia Tarasova *, Irina Animitsa, Anzhelika Galisheva and Daniil Korona
Institute of Natural Sciences and Mathematics, Ural Federal University, 620000 Yekaterinburg, Russia;
irina.animitsa@urfu.ru (I.A.); jelya95@gmail.com (A.G.); D.V.Korona@urfu.ru (D.K.)
* Correspondence: Natalia.Tarasova@urfu.ru
Received: 6 May 2019; Accepted: 20 May 2019; Published: 22 May 2019 Materials 2019, 12, 1668; doi:10.3390/ma12101668 materials materials materials 1. Introduction Protonic ceramics are important materials for using as electrolytes in proton-conducting solid oxide
fuel cells (H-SOFCs) [1–6]. The application of the proton-conducting complex oxides in comparison
with oxygen-ion conductors (in oxygen ion-conducting solid oxide fuel cells O-SOFCs) has several
advantages. First, the proton electrolytes have lower activation energies of charge carriers (~0.5 eV)
than the oxygen-ion conductors (~1 eV). Consequently, they have higher conductivities. This allows
us to decrease the operating temperatures from 900–1000 ◦C (O-SOFCs) to 500–700 ◦C (H-SOFCs) [7]
and provides the decrease of the production cost of H-SOFCs at 27–37% [8]. Secondly, the water
vapor in the H-SOFCs is produced at the air-side electrode (the cathode side) and this avoids the
dilution of fuel by steam [9], and, consequently, increases the energy conversion efficiency [10,11]. Therefore, the discovery of a new structure family of proton conductors is important both in terms of
the development of scientific materials and in the applied aspect. The most studied protonic conductors are the complex oxides with perovskite or perovskite-related
structures. Further development of proton conductors suggests investigations of materials with new
types of structures, including Ruddlesden-Popper structure. It is common understanding that
oxygen-deficient compounds are capable of manifesting not only oxygen-ionic conductivity, but also
protonic conductivity in a humid atmosphere. Therefore, perovskite-related structures with oxygen-ion www.mdpi.com/journal/materials www.mdpi.com/journal/materials 2 of 14 Materials 2019, 12, 1668 conductivity can be tested as potentially capable of proton transport. Recently, researchers uncovered
new data about oxygen-ion transport in the complex oxides with a Ruddlesden-Popper structure [12–14]. The phase BaNdInO4 has a structure that is slightly different from K2NiF4 [15]. The BaNdInO4-type
structure (monoclinic, P21/c) consists of alternative stacking of the [NdO] oxide unit and Nd1/4Ba3/4InO3
perovskite unit with the InO6 edge-facing Nd−O units. p
g
g
The new compounds BaYInO4, BaSmInO4, BaHoInO4, BaErInO4, and BaYbInO4 were also
synthesized [16,17], but the total electrical conductivity of BaNdInO4 was the highest among BaRInO4
compounds. Therefore, the investigations were mainly focused on BaNdInO4. The phase BaNdInO4
was described as a mixed oxygen ionic-electronic conductor [12–14] and cation doping with different
metals M = Ca, Sr, Ba, Ce, Ga, Cr, Si, Mg, Zr, Nb, Ta, Ti, and Sn at the Nd-sites or In-sites were used
for increasing the oxide ion conductivity [13,18]. It was shown that oxygen-ion conductivity of Ca2+
and Ti4+ doped compositions was higher than those for other dopants [13,18]. 1. Introduction It was proven that the
increase of oxygen-ion conductivity at acceptor doping is due to the formation of oxygen vacancies. At the same time, the possibility of the formation of the interstitial oxygens in the rock-salt block of
BaNdIn1-xTixO4+δ due to donor doping, which leads to the increase of oxide-ion conductivity, was also
considered [18]. The interstitial oxygen conduction was also discussed in the donor doped K2NiF4-type
LaSrInO4 oxides [19,20]. It should also be mentioned that Kotaro Fujii et al. [17] emphasized that
there is no direct evidence for the interstitial oxygen. In the BaNdInO4-type structure, there is no
rock-salt block, but the A rare earth oxide unit could have an influence on the impurity phases. All this
demonstrates that these phases are very interesting and further investigation is required. Although oxygen-ion transport in the systems based on BaNdInO4 was widely discussed,
no studies of proton transport have been performed. Nevertheless, the possibility of proton transport
for phases with the Ruddlesden-Popper structure was described earlier for oxyfluorides Ba2InO3F
and Ba3In2O5F2 [21]. K2NiF4-type materials of Pr1-xM1+xInO4 (M = Ba2+, Sr2+; x = 0, 0.1) exhibit
proton conductivity and are suitable for the operation of proton-conducting solid oxide fuel cells
(H-SOFCs) at targeted temperatures of 500 to 700 ◦C [22]. The layered perovskites Sr1+xSm1−xAlO4−δ
and Sr1+xPr1−xAlO4-δ also have proton conductivity and high proton transport number at 700 to
900 ◦C [23]. These studies also show that, for a correct discussion, the conductivity measurements for
phases with a layered structure should be carried out in atmospheres with controlled humidity. To develop the novel proton conductors, we focused on the composition of BaLaInO4 as a parent
compound, which has a Ruddlesden-Popper structure. The structural features of the phases of
the composition AIILnInO4 (AA′BO4) were described earlier [24], where the relationship between
transitions of K2NiF4−CaFe2O4 type structures was established. The compound BaLaInO4 crystallizes
in an orthorhombic structure with space group Pbca and with lattice parameters a = 12.933 (3),
b = 5.911 (1), and c = 5.905 (1) Å [24]. In the present work, the new complex oxides BaLa0.9M0.1InO3.95
(M = Ca2+, Sr2+, Ba2+) with a Ruddlesden-Popper structure were obtained via the solid-state method. The structure, water uptake, and proton transport were examined. The influence of the nature of
dopant (alkaline earth metals) on the transport properties were discussed. 2. Materials and Methods The phases BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, Ba2+) and the parent composition BaLaInO4 were
prepared by a solid-state method. The preliminary dried high-purity powders of CaCO3, SrCO3, BaCO3,
In2O3, and La2O3 were used. Prior to weighing, the starting materials were dried. The stoichiometric
powder mixtures were manually milling in an agate mortar and then calcined at 800, 900, 1000, 1100,
1200, and 1300 ◦C for 24 h. Intermediate grindings were made for every following heating step. Phase purity was monitored by powder X-ray diffraction (XRD) on a Bruker Advance D8
diffractometer with CuKα radiation. The crystal structures of dry and hydrated samples were
determined by Rietveld refinement [25]. The dry samples were prepared for XRD by heat treated
at 1100 ◦C for 4 h and then cooled in dry Ar (pH2O = 3.5 × 10−5 atm). The hydrated samples were Materials 2019, 12, 1668 3 of 14 obtained at slow cooling from 1100 to 150 ◦C (1 ◦C/min) under a flow of wet Ar (99.999% purity,
pH2O = 2 × 10−2 atm). Ar atmosphere was used to avoid any carbonization of the samples. TG-analysis was carried out on a STA 409 PC analyzer (Netzsch, Selb, Germany) coupled with a
quadrupole mass spectrometer QMS 403 C Aëolos (Netzsch, Selb, Germany) in order to determine the
degree of hydration. The samples were heated at the rate of 10 ◦C/min in a corundum crucible under
a flow of dry Ar at temperatures ranging from 40 to 1100 ◦C. Before TG-measurements, the powder
samples were initially hydrated as described above. The number of moles of water per one formula
unit was determined and the notation of the total composition as BaLa0.9M0.1InO3.95·nH2O was used
for convenience of comparison. To indicate that H2O incorporation gives OH−-groups, the formulas
were also written in another representation as BaLa0.9M0.1InO3.95−n(OH)2n. The ceramics used for the electrical measurements were prepared by pressing cylindrical pellets
and sintering them at 1300 ◦C for 24 h in dry air. The samples typically had a relative density of around
∼92% (density of the sintered samples was determined by the Archimede method). After polishing,
the platinum paste electrodes were fired at 900 ◦C for 3 h. The ac conductivity of the samples
(two-probe method) was measured over the frequency range of 1 kHz–1 MHz using a Z-1000P
(Elins, RF) impedance spectrometer at an applied voltage of 150 mV. 2. Materials and Methods The least squares refinement
program ZView (Scribner Associates Inc., Southern Pines, NC, USA) was used to fit the impedance
data to an (RQ) equivalent circuit, where R is resistance and Q is the constant phase element. The bulk
resistance was calculated from the impedance data and then the bulk conductivity was calculated and
discussed below. From the fitted resistances and, together with the geometrical dimensions, the total
conductivity can be calculated using the following equation. σ = 1
R · l
S
(1) (1) where l and S are the thickness and surface area of the sample. The measurements of the temperature dependencies of conductivities were performed in the
temperature range from 200 to 1000 ◦C every 10 to 20 ◦C with a cooling rate of 1 ◦C/min. During the
measurement, the sample was held at each temperature until the resistances became constant. The conductivity measurements were carried out under dry and wet air or Ar. The «wet» gas was obtained by bubbling the gas at room temperature first through distilled water
and then through a saturated solution of KBr (pH2O = 2 × 10−2 atm). The «dry» gas was produced by
circulating the gas through P2O5 (pH2O = 3.5 × 10−5 atm). The humidity of gases was measured by the
H2O-sensor (“Honeywell” HIH-3610, Charlotte, NC, USA). 3.1. X-ray Characterization The X-ray diffraction (XRD) results for the parent compound BaLaInO4 and its doped analogs
BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, Ba2+) showed that the samples were single phase (orthorhombic
symmetry, space group Pbca). The obtained lattice parameters for the sample BaLaInO4 were in good
agreement with previously reported data [24]. The substitution of M2+ ions for La+3 ion led to changes
in structural parameters (Table 1). The lattice parameters and cell volumes increased with an increasing
ionic radius of the dopant [26]. It should be noted that a more significant increase in parameters
and cell volume in comparison with the undoped sample was observed for the Ba2+-doped sample. At the same time, for the Ca2+-doped sample (as the closest-sized ion to lanthanum), an increase and
decrease in the lattice parameters was observed. It is known that the formation of oxygen vacancy
is also accompanied by decreases of the lattice parameter because of the size relationship of oxygen
vacancy and the oxide ion [27]. We can assume that, for the Ca2+-doped sample, the overall effect of
the influence of the size of the B-cation and oxygen vacancy occurred and no monotonic change of the
parameters was observed. 4 of 14 Materials 2019, 12, 1668 Table 1. Cell parameters and volume of anhydrous and hydrated samples. Table 1. Cell parameters and volume of anhydrous and hydrated samples. Sample
a, Å
b, Å
c, Å
β, o
Cell Volume, (Å3)
BaLaInO4
12.932(3)
5.906(0)
5.894(2)
90
450.188(2)
BaLa0.9Ca0.1InO3.95
12.967(4)
5.913(3)
5.884(5)
90
451.224(2)
BaLa0.9Sr0.1InO3.95
12.950(2)
5.917(1)
5.897(5)
90
451.911(4)
BaLa0.9Ba0.1InO3.95
13.002(1)
5.919(3)
5.901(3)
90
454.183(7)
BaLaInO4·nH2O
12.717(4)
14.763(4)
7.214(9)
92.92(6)
1352(8)
BaLa0.9Ca0.1InO3.95·nH2O
12.710(8)
14.770(8)
7.218(6)
92.82(2)
1353(6)
BaLa0.9Sr0.1InO3.95·nH2O
12.725(3)
14.791(2)
7.220(4)
92.79(0)
1357(4)
BaLa0.9Ba0.1InO3.95·nH2O
12.741(0)
14.806(5)
7.222(6)
92.75(2)
1360(9)
Materials 2019, 12, x FOR PEER REVIEW
4 of 1
Table 1. Cell parameters and volume of anhydrous and hydrated samples. Sample
a, Å
b, Å
c, Å
β, o
Cell Volume, (Å3)
BaLaInO4
12.932(3)
5.906(0)
5.894(2)
90
450.188(2)
BaLa0.9Ca0.1InO3.95
12.967(4)
5.913(3)
5.884(5)
90
451.224(2)
BaLa0.9Sr0.1InO3.95
12.950(2)
5.917(1)
5.897(5)
90
451.911(4)
BaLa0.9Ba0.1InO3.95
13.002(1)
5.919(3)
5.901(3)
90
454.183(7) The hydration processes led to the changes in the crystal structure for the samples compared with
anhydrous samples. All hydrated samples BaLaInO4·nH2O and BaLa0.9M0.1InO3.95·nH2O belonged to
the monoclinic symmetry (space group P2/m). The increase in the ionic radii of alkaline earth metal led
to the increase in the cell volume of hydrate samples (Table 1). 3.2. Thermal Properties The water uptake was established by TG-analysis in combination with differential scanning
calorimetry (DSC) and mass-spectrometry (MS). Materials 2019, 12, x FOR PEER REVIEW
5 of 14 As is well known, the possibility of water uptake from the gas phase for perovskite-type oxide
conductors is due to the presence of oxygen vacancies in the structure and can be described as follows.
OH
O
O
H
V
)
(2
(2) V••
o + H2O + O×
o ⇔2(OH)•
o
(2)
O
is the oxygen atom in a regular position and
)
OH
(
is the (2)
the where V••
o is the oxygen vacancy, O×
o is the oxygen atom in a regular position, and (OH)•
o is the hydroxyl
group in the oxygen sublattice. Based on the concentration of oxygen vacancies of obtained samples
BaLaInO4[VO]0 and BaLa0.9M0.1InO3.95[VO]0.05, there should be a degree of hydration of 0 and 0.05 mol
H2O per formula unit. As shown in Figure 2, for all the investigated samples, the mass loss was observed,
which is attributed to a loss of water, in accordance with MS data. Water uptake of the samples was
much higher than the concentration of oxygen vacancies and the composition of the samples appears
to be BaLaInO4·0.58H2O (or BaLaInO3.42(OH)1.16, which represents that H2O incorporation gives
OH−-groups, BaLa0.9Ca0.1InO3.95·0.7H2O (or BaLa0.9Ca0.1InO3.25(OH)1.4), BaLa0.9Sr0.1InO3.95·0.92H2O
(or BaLa0.9Sr0.1InO3.03(OH)1.84), and BaLa0.9Ba0.1InO3.95·1.05H2O (or BaLa0.9Ba0.1InO2.9(OH)2.1). o
V
yg
y,
o
yg
g
p
,
o)
OH
(
hydroxyl group in the oxygen sublattice. Based on the concentration of oxygen vacancies of obtained
samples BaLaInO4[VO]0 and BaLa0.9M0.1InO3.95[VO]0.05, there should be a degree of hydration of 0 and
0.05 mol H2O per formula unit. As shown in Figure 2, for all the investigated samples, the mass loss
was observed, which is attributed to a loss of water, in accordance with MS data. Water uptake of the
samples was much higher than the concentration of oxygen vacancies and the composition of the
samples appears to be BaLaInO4∙0.58H2O (or BaLaInO3.42(OH)1.16, which represents that H2O
incorporation
gives
OH−-groups,
BaLa0.9Ca0.1InO3.95∙0.7H2O
(or
BaLa0.9Ca0.1InO3.25(OH)1.4),
BaLa0.9Sr0.1InO3.95∙0.92H2O (or BaLa0.9Sr0.1InO3.03(OH)1.84), and BaLa0.9Ba0.1InO3.95∙1.05H2O (or
BaLa0.9Ba0.1InO2.9(OH)2.1). 200
400
600
800
1000
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.8
2.0
2.2
2.4
2.6
2.8
3.0
5
Ionic current, •10
11 А
T, °C
water uptake
MS (H2O)
TG
4
3
2
1
Figure 2. 3.2. Thermal Properties TG data for hydrated samples BaLaInO4∙nH2O (1), BaLa0.9Ca0.1InO3.95∙nH2O (2),
BaLa0.9Sr0.1InO3.95∙nH2O
(3),
BaLa0.9Ba0.1InO3.95∙nH2O
(4),
and
MS
(H2O)-data
for
BaLa0.9Ba0.1InO3.95∙nH2O (5). Figure 3 represents the comparison of TG-curves, DSC-curves, and MS-curves for the hydrated
mple BaLaInO4∙nH2O. As can be seen, the sample underwent a mass decrease in the temperature
om 200 to 800 °C, which was accompanied by a three endothermic DSC-signals (two well
onounced and one weak signal) at ~320 °C, ~500 °C, and ~700 °C. The MS analysis indicated that
e mass change can be attributed by the water loss. No other volatile components (CO2, O2) were
d Th d
d
l
B L
M I O
h
d h
i
il
TG
DSC
d MS
Figure
2. TG data for hydrated samples BaLaInO4·nH2O (1),
BaLa0.9Ca0.1InO3.95·nH2O (2),
BaLa0.9Sr0.1InO3.95·nH2O(3),BaLa0.9Ba0.1InO3.95·nH2O(4),andMS(H2O)-dataforBaLa0.9Ba0.1InO3.95·nH2O(5). Figure 3 represents the comparison of TG-curves, DSC-curves, and MS-curves for the hydrated
mple BaLaInO4·nH2O. As can be seen, the sample underwent a mass decrease in the temperature
om 200 to 800 ◦C, which was accompanied by a three endothermic DSC-signals (two well pronounced
nd one weak signal) at ~320 ◦C, ~500 ◦C, and ~700 ◦C. The MS analysis indicated that the mass change
n be attributed by the water loss. No other volatile components (CO2, O2) were detected. The doped
mples BaLa0.9M0.1InO3.95 showed the similar TG-curves, DSC-curves, and MS-curves (Figure 2). 200
400
600
800
1000
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.8
2.0
2.2
2.4
2.6
2.8
3.0
5
Ionic current, •10
11 А
T, °C
water uptake
MS (H2O)
TG
4
3
2
1
Figure
2. TG data for hydrated samples BaLaInO4·nH2O (1),
BaLa0 9Ca0 1InO3 95·nH2O (2), Figure 2. TG data for hydrated samples BaLaInO4∙nH2O (1), BaLa0.9Ca0.1InO3.95∙nH2O (2),
BaLa0 9Sr0 1InO3 95∙nH2O
(3),
BaLa0 9Ba0 1InO3 95∙nH2O
(4),
and
MS
(H2O)-data
for
Figure
2. TG data for hydrated samples BaLaInO4·nH2O (1),
BaLa0.9Ca0.1InO3.95·nH2O (2),
BaLa0.9Sr0.1InO3.95·nH2O(3),BaLa0.9Ba0.1InO3.95·nH2O(4),andMS(H2O)-dataforBaLa0.9Ba0.1InO3.95·nH2O(5). ( )
( )
(
)
BaLa0.9Ba0.1InO3.95∙nH2O (5). Figure 3 represents the comparison of TG-curves, DSC-curves, and MS-curves for the hydrated
sample BaLaInO4∙nH2O. As can be seen, the sample underwent a mass decrease in the temperature
from 200 to 800 °C, which was accompanied by a three endothermic DSC-signals (two well
pronounced and one weak signal) at ~320 °C, ~500 °C, and ~700 °C. 3.1. X-ray Characterization BaLaInO4∙nH2O
12.717(4)
14.763(4)
7.214(9)
92.92(6)
1352(8)
BaLa0.9Ca0.1InO3.95∙nH2O
12.710(8)
14.770(8)
7.218(6)
92.82(2)
1353(6)
BaLa0.9Sr0.1InO3.95∙nH2O
12.725(3)
14.791(2)
7.220(4)
92.79(0)
1357(4)
BaLa0.9Ba0.1InO3.95∙nH2O
12.741(0)
14.806(5)
7.222(6)
92.75(2)
1360(9) It should be noted that XRD-patterns for all obtained samples were refined by the Rietveld analysis. Results for anhydrous BaLa0.9Sr0.1InO3.95 and hydrated BaLa0.9Sr0.1InO3.95·nH2O are presented in
Figure 1a,b as an example of fitting. It should be noted that XRD-patterns for all obtained samples were refined by the Rietveld
analysis. Results for anhydrous BaLa0.9Sr0.1InO3.95 and hydrated BaLa0.9Sr0.1InO3.95∙nH2O are presented
in Figure 1a,b as an example of fitting. 20
30
40
50
60
70
80
4 000
2
Intensity
2 000
6 000
0
BaLa0.9Sr0.1InO3.95
(a) 20
30
40
50
60
70
80
4 000
2
Intensity
2 000
6 000
0
BaLa0.9Sr0.1InO3.95
(a)
20
30
40
50
60
70
80
4 000
2
Intensity
2 000
6 000
0
BaLa0.9Sr0.1InO3.95•nH2O
(b)
Figure 1. XRD patterns of BaLa0.9Sr0.1InO3.95 (a) and BaLa0.9Sr0.1InO3.95∙nH2O (b). At the bottom of the
figure, the pattern is the difference between the experimental (red) and the calculated one (dark) after
refinement. Vertical bars show the Bragg angle positions. Figure 1. XRD patterns of BaLa0.9Sr0.1InO3.95 (a) and BaLa0.9Sr0.1InO3.95·nH2O (b). At the bottom of
the figure, the pattern is the difference between the experimental (red) and the calculated one (dark)
after refinement. Vertical bars show the Bragg angle positions. Figure 1. XRD patterns of BaLa0.9Sr0.1InO3.95 (a) and BaLa0.9Sr0.1InO3.95∙nH2O (b). At the bottom of the
figure, the pattern is the difference between the experimental (red) and the calculated one (dark) after
refinement. Vertical bars show the Bragg angle positions. Figure 1. XRD patterns of BaLa0.9Sr0.1InO3.95 (a) and BaLa0.9Sr0.1InO3.95·nH2O (b). At the bottom of
the figure, the pattern is the difference between the experimental (red) and the calculated one (dark)
after refinement. Vertical bars show the Bragg angle positions. Materials 2019, 12, 1668 5 of 14 3.2. Thermal Properties The MS analysis indicated that
the mass change can be attributed by the water loss No other volatile components (CO2, O2) were
Figure 3 represents the comparison of TG-curves, DSC-curves, and MS-curves for the hydrated
sample BaLaInO4·nH2O. As can be seen, the sample underwent a mass decrease in the temperature
from 200 to 800 ◦C, which was accompanied by a three endothermic DSC-signals (two well pronounced
and one weak signal) at ~320 ◦C, ~500 ◦C, and ~700 ◦C. The MS analysis indicated that the mass change
can be attributed by the water loss. No other volatile components (CO2, O2) were detected. The doped
samples BaLa0.9M0.1InO3.95 showed the similar TG-curves, DSC-curves, and MS-curves (Figure 2). 6 of 14 Materials 2019, 12, 1668
g
water into the rock-s 0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
0
200
400
600
800
1000
98.5
99.0
99.5
100.0
0.00
0.05
0.10
0.15
0.20
0.25
0.30
Ionic current, 10
10 А
T, °C
mass loss, %
DSC
MS (H2O)
TG
endo
DSC, mW/mg
Figure 3. TG-(dark line), DSC-(blue line), and MS (H2O)-data (red line) for hydrated sample
BaLaInO4∙nH2O. Figure 3. TG-(dark line), DSC-(blue line), and MS (H2O)-data (red line) for hydrated sample
BaLaInO4·nH2O. Figure 3. TG-(dark line), DSC-(blue line), and MS (H2O)-data (red line) for hydrated sample
BaLaInO4∙nH2O. Figure 3. TG-(dark line), DSC-(blue line), and MS (H2O)-data (red line) for hydrated sample
BaLaInO4·nH2O. As can be seen from Figure 2, the degree of hydration increases in the order of BaLaInO4–
BaLa0.9Ca0.1InO3.95–BaLa0.9Sr0.1InO3.95–BaLa0.9Ba0.1InO3.95. We can suggest that, with an increase in the
size of the dopant, the size of the salt block increases. This contributes to the placement of a larger
number of OH−-groups, which occupy interstitial sites within the rock salt layers. Therefore, we can conclude that investigated samples BaLaInO4 and BaLa0.9M0.1InO3.95 can show
the ability for water uptake not only in accordance with reaction (2), but also due to the presence in
the structure of La-O blocks (rock-salt blocks). 3.2. Thermal Properties x
x OH
BaLaInO
O
H
x
BaLaInO
)
(
2
2
4
2
4
(3)
x
x OH
InO
M
BaLa
O
H
x
InO
M
BaLa
)
(
2
2
95
.3
1.0
9.0
2
95
.3
1.0
9.0
(4)
It is known that the ability of water incorporation is ensured not only by the presence of
oxygen vacancies, but also by the features of the crystal structure [28]. For layered structures, it
can be explained by the ability of water incorporation by the salt La−O blocks, which alternates
with blocks of the perovskite matrix. For example, the formation of hydrated phases based on the
Ruddlesden-Popper-type layered oxide Ba2ZrO4 [29] was proven. During hydration, this phase can
incorporate up to 1.7 mol of water Ba2ZrO4·1.7H2O. The effects of water incorporation into layered
structures were summarized in the dissertation [30]. It was shown that RP phases can adopt up to
two water molecules per formula unit/rock-salt block. At the same time, the large amount of water
molecules in some RP-phases can be explained by the formation of hydrates and, in addition, not all RP
compounds react with water under ambient conditions. This shows that the elemental composition of
an RP compounds has a significant effect on the stability, and atoms with the biggest influence are the
cations present in the rock-salt block. If the majority of cations in the rock-salt block are large cations,
e.g., Ba, K, Rb or Cs, then the compounds have a higher tendency for the intercalation of water into the
rock-salt block [30]. Thus, the results show the ready ability of the RP-structure to incorporate extra water as OH−-
groups, located in the sites of the rock salt layers. This ability can lead to proton transport. In addition,
it is important to note that these results also show that it is possible to regulate the concentration of
intercalated water by the nature of the dopant introduced into the salt layers. 3.3. Electrical Properties
As can be seen from Figure 2,
the degree of hydration increases in the order of
BaLaInO4–BaLa0.9Ca0.1InO3.95–BaLa0.9Sr0.1InO3.95–BaLa0.9Ba0.1InO3.95. We can suggest that, with
an increase in the size of the dopant, the size of the salt block increases. This contributes to the
placement of a larger number of OH−-groups, which occupy interstitial sites within the rock salt layers. 3.2. Thermal Properties Therefore, we can conclude that investigated samples BaLaInO4 and BaLa0.9M0.1InO3.95 can show
the ability for water uptake not only in accordance with reaction (2), but also due to the presence in the Thus, the results show the ready ability of the RP-structure to incorporate extra water as OH−-
groups, located in the sites of the rock salt layers. This ability can lead to proton transport. In addition,
it is important to note that these results also show that it is possible to regulate the concentration of
intercalated water bythe nature of the dopant introduced into the salt layers. As can be seen from Figure 2,
the degree of hydration increases in the order of
BaLaInO4–BaLa0.9Ca0.1InO3.95–BaLa0.9Sr0.1InO3.95–BaLa0.9Ba0.1InO3.95. We can suggest that, with
an increase in the size of the dopant, the size of the salt block increases. This contributes to the
placement of a larger number of OH−-groups, which occupy interstitial sites within the rock salt layers. intercalated water by the nature of the dopant introduced into the salt layers. 3.3. Electrical Properties
F
ll th
l
i
thi
t d
i i
l
t
ti
f
di
t
f th
l
Therefore, we can conclude that investigated samples BaLaInO4 and BaLa0.9M0.1InO3.95 can show
the ability for water uptake not only in accordance with reaction (2), but also due to the presence in the
structure of La-O blocks (rock-salt blocks). y,
,
g
p
responding to the bulk component, was observed. Typical impedance
ms for the BaLa0.9Sr0.1InO3.95 sample, recorded at various temperatures, are given
citance of the observed arc was estimated to be about 10 −11 F∙cm−1, which is typical
ccording to the experimental data obtained for other phases with an RP structure
BaLaInO4 + x
2H2O →BaLaInO4−x
2 (OH)x
(3)
BaLa0.9M0.1InO3.95 + x
2H2O →BaLa0.9M0.1InO3.95−x
2 (OH)x
(4) nce
(3) citance of the observed arc was estimated to be about 10 −11 F∙cm−1, which is typical
ccording to the experimental data obtained for other phases with an RP structure
BaLa0.9M0.1InO3.95 + x
2H2O →BaLa0.9M0.1InO3.95−x
2 (OH)x
(4) cal
ure(4) Thus, the results show the ready ability of the RP-structure to incorporate extra water as
OH−-groups, located in the sites of the rock salt layers. This ability can lead to proton transport. 3.3. Electrical Properties For all the samples in this study, one semicircle, starting from zero coordinates of the complex
Z-Z′ plane and corresponding to the bulk component, was observed. Typical impedance spectroscopy
diagrams for the BaLa0.9Sr0.1InO3.95 sample, recorded at various temperatures, are given in Figure 4. The capacitance of the observed arc was estimated to be about 10−11 F·cm−1, which is typical of a
bulk response. According to the experimental data obtained for other phases with an RP structure
(for example, LaSrInO4 [19,20] and NdBaInO4 [13]), two contributions can be observed in the typical
impedance spectroscopy diagrams. The semicircle in the high-frequency range, corresponding to
the contribution from bulk resistance, and very small low-frequency semicircle, attributed to the
contribution of grain-boundary resistance. It should be noted that the grain-boundary arc was
manifested in the frequency range of 100 mHz to 100 Hz. In comparison with the presented data
for doped BaLaInO4 (Figure 4), the observed one semicircle is realized in the frequency range of
1 to 1000 kHz. Therefore, the arc corresponding to the grain boundary response, which appears at
lower frequencies (in comparison with bulk semicircle), cannot be observed due to the instrumental
limitations of the frequency range. Materials 2019, 12, x FOR PEER REVIEW
7 of 14
(for example, LaSrInO4 [19,20] and NdBaInO4 [13]), two contributions can be observed in the typical
impedance spectroscopy diagrams. The semicircle in the high-frequency range, corresponding to the
contribution from bulk resistance, and very small low-frequency semicircle, attributed to the
contribution of grain-boundary resistance. It should be noted that the grain-boundary arc was
manifested in the frequency range of 100 mHz to 100 Hz. In comparison with the presented data for
doped BaLaInO4 (Figure 4), the observed one semicircle is realized in the frequency range of 1 to 1000
kHz. Therefore, the arc corresponding to the grain boundary response, which appears at lower
frequencies (in comparison with bulk semicircle), cannot be observed due to the instrumental
li
it ti
f th f
20
25 q
y
g
0
5
10
15
20
25
0
5
10
15
1 kHz
100 kHz
480
oC
BaLa0.9Ba0.1InO3.95
460
oC 440
oC
420
oC
Z', k
-Z'', k
400
oC
1000 kHz
Figure 4. Evolution of the impedance spectra with temperature for the sample BaLa0.9Ba0.1InO3.95 in
dry air (pH2O = 3.5 × 10−5 atm). Figure 4. 3 4 Dry Conditions
3.4. Dry Conditions 3.4. Dry Conditions
The temperature dependencies of total conductivities for investigated samples BaLaInO4 and
BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, Ba2+) are shown in Figure 5. For dry air (pH2O = 3.5 × 10−5 atm) (Figure
5), the conductivities for doped samples BaLa0.9M0.1InO3.95 are higher than those for undoped
composition BaLaInO4 at a ~1.5 order of magnitude. To evaluate the contribution of electronic
conductivity of investigated phases, the measurements were performed in the Ar atmosphere (that
is, at lower pO2 where ionic conductivity dominates). The comparison of total conductivities in air
and Ar under dry conditions was shown in Figure 6. The conductivities in Ar atmosphere were lower
than the conductivities, measured in air. The conductivity enhancement in air can be ascribed to the
electronic conductionbecause of formation of holes in air as shown bythe quasi chemical reaction
The temperature dependencies of total conductivities for investigated samples BaLaInO4 and
BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, Ba2+) are shown in Figure 5. For dry air (pH2O = 3.5 × 10−5 atm)
(Figure 5), the conductivities for doped samples BaLa0.9M0.1InO3.95 are higher than those for undoped
composition BaLaInO4 at a ~1.5 order of magnitude. To evaluate the contribution of electronic
conductivity of investigated phases, the measurements were performed in the Ar atmosphere (that is,
at lower pO2 where ionic conductivity dominates). The comparison of total conductivities in air and Ar
under dry conditions was shown in Figure 6. The conductivities in Ar atmosphere were lower than the
conductivities, measured in air. The conductivity enhancement in air can be ascribed to the electronic
conduction because of formation of holes in air, as shown by the quasi-chemical reaction. y
q
h
O
O
V
o
o
2
2
1
2
(5)
V••
o + 1
2O2 ⇔O×
o + 2h•
(5) (5)
(5) For the undoped sample, the conductivity in Ar atmosphere were lower than the conductivities,
measured in air, over the whole temperature range studied. This proves the mixed oxygen ionic-
electronic character of conductivity of BaLaInO4. 3.3. Electrical Properties Evolution of the impedance spectra with temperature for the sample BaLa0.9Ba0.1InO3.95 in
dry air (pH2O = 3.5 × 10−5 atm). 0
5
10
15
20
25
0
5
10
15
1 kHz
100 kHz
480
oC
BaLa0.9Ba0.1InO3.95
460
oC 440
oC
420
oC
Z', k
-Z'', k
400
oC
1000 kHz Figure 4. Evolution of the impedance spectra with temperature for the sample BaLa0.9Ba0.1InO3.95 in
dry air (pH2O = 3.5 × 10−5 atm). Figure 4. Evolution of the impedance spectra with temperature for the sample BaLa0.9Ba0.1InO3.95 in
dry air (pH2O = 3.5 × 10−5 atm). 3.2. Thermal Properties In addition, it is important to note that these results also show that it is possible to regulate the
concentration of intercalated water by the nature of the dopant introduced into the salt layers. Thus, the results show the ready ability of the RP-structure to incorporate extra water as
OH−-groups, located in the sites of the rock salt layers. This ability can lead to proton transport. In addition, it is important to note that these results also show that it is possible to regulate the
concentration of intercalated water by the nature of the dopant introduced into the salt layers. Materials 2019, 12, 1668 7 of 14 7 of 14 3 4 Dry Conditions
3.4. Dry Conditions 0.8
1.0
1.2
1.4
1.6
1.8
2.0
-8
-7
-6
-5
-4
-3
-2
2 [13]
BaLaInO4
BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95
BaLa0.9Ba0.1InO3.95
10
3/Т K
-1
log (S/cm)
pH2O = 3.5•10
-5 atm
1 [13]
0.8
1.0
1.2
1.4
1.6
1.8
2.0
-8
-7
-6
-5
-4
-3
2 [13]
BaLaInO4
BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95
BaLa0.9Ba0.1InO3.95
10
3/Т, K
-1
log (S/cm)
pH2O = 3.5•10
-5 atm
1 [13] Figure 5. The temperature dependencies of the total conductivities (pO2 = 0.21 atm) of BaLaInO4 and
BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, and Ba2+) in dry air and comparison with (1) BaNdInO4 and (2)
Ba1.1Nd0.9InO3.95 in air. Figure 5. The temperature dependencies of the total conductivities (pO2 = 0.21 atm) of BaLaInO4 and
BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, and Ba2+) in dry air and comparison with (1) BaNdInO4 and (2)
Ba1.1Nd0.9InO3.95 in air. p
p
p
BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, and Ba2+) in dry air and comparison with (1) BaNdInO4 and (2)
Ba1.1Nd0.9InO3.95 in air. 0.8
1.0
1.2
1.4
1.6
1.8
2.0
-9
-8
-7
-6
-5
-4
-3
-2
pH2O = 3.5•10
-5 atm
BaLaInO4
BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95
BaLa0.9Ba0.1InO3.95
10
3/Т, K
-1
log (S/cm)
filled signs - air
open signs - Ar
Figure 6. The temperature dependencies of the conductivities of BaLaInO4 and BaLa0.9M0.1InO3.95 (M
= Ca2+, Sr2+, and Ba2+) in dry air (filled signs) and dry Ar (open signs). 0.8
1.0
1.2
1.4
1.6
1.8
2.0
-9
-8
-7
-6
-5
-4
-3
-2
pH2O = 3.5•10
-5 atm
BaLaInO4
BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95
BaLa0.9Ba0.1InO3.95
10
3/Т, K
-1
log (S/cm)
filled signs - air
open signs - Ar
Figure 6. The temperature dependencies of the conductivities of BaLaInO4 and BaLa0.9M0.1InO3.95 (M
= Ca2+, Sr2+, and Ba2+) in dry air (filled signs) and dry Ar (open signs). h
h
b
d
h
d
A
h
f
d
d
h
Figure 6. The temperature dependencies of the conductivities of BaLaInO4 and BaLa0.9M0.1InO3.95
(M = Ca2+, Sr2+, and Ba2+) in dry air (filled signs) and dry Ar (open signs). 3 4 Dry Conditions
3.4. Dry Conditions For Sr2+, Ba2+-doped samples, the conductivity
exhibited the same character, but, for the Ca2+-doped sample, the differences were more significant
at higher temperatures and the conductivity values were comparable at low temperatures, which
indicates that the electron contribution decreases with a falling temperature and this phase exhibited
h d
h
f
d
b l
h
l
l
h
d h
For the undoped sample, the conductivity in Ar atmosphere were lower than the conductivities,
measured in air, over the whole temperature range studied. This proves the mixed oxygen
ionic-electronic character of conductivity of BaLaInO4. For Sr2+, Ba2+-doped samples, the conductivity
exhibited the same character, but, for the Ca2+-doped sample, the differences were more significant at
higher temperatures and the conductivity values were comparable at low temperatures, which indicates
that the electron contribution decreases with a falling temperature and this phase exhibited the dominant Materials 2019, 12, 1668 8 of 14 oxygen–ionic character of conductivity below 500 ◦C. These results also showed that for the Sr2+,
Ba2+-substituted phases, the contribution of the electronic conductivity was greater when compared to
the Ca2+-substituted phase. Materials 2019, 12, x FOR PEER REVIEW
8 of 14
Materials 2019, 12, x FOR PEER REVIEW
8 of 14
2 0.8
1.0
1.2
1.4
1.6
1.8
2.0
-8
-7
-6
-5
-4
-3
-2
2 [13]
BaLaInO4
BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95
BaLa0.9Ba0.1InO3.95
10
3/Т, K
-1
log (S/cm)
pH2O = 3.5•10
-5 atm
1 [13]
Figure 5. The temperature dependencies of the total conductivities (pO2 = 0.21 atm) of BaLaInO4 and
BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, and Ba2+) in dry air and comparison with (1) BaNdInO4 and (2)
Ba1.1Nd0.9InO3.95 in air. Figure 5. The temperature dependencies of the total conductivities (pO2 = 0.21 atm) of BaLaInO4 and
BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, and Ba2+) in dry air and comparison with (1) BaNdInO4 and (2)
Ba1.1Nd0.9InO3.95 in air. 0.8
1.0
1.2
1.4
1.6
1.8
2.0
-8
-7
-6
-5
-4
-3
2 [13]
BaLaInO4
BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95
BaLa0.9Ba0.1InO3.95
10
3/Т, K
-1
log (S/cm)
pH2O = 3.5•10
-5 atm
1 [13]
Figure 5. The temperature dependencies of the total conductivities (pO2 = 0.21 atm) of BaLaInO4 and
BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, and Ba2+) in dry air and comparison with (1) BaNdInO4 and (2)
Ba1.1Nd0.9InO3.95 in air. 3 4 Dry Conditions
3.4. Dry Conditions The M2+-cations are incorporated into the lattice at La3+ sites with the creation of oxygen vacancies as
charge-compensating defects. This doping process can be represented by the following defect reaction. 2MO
La2O3
→2M′
La + 2O×
O + V••
O
(6) (6) Comparing the ionic conductivities of the doped phases, we can note the tendency of its increase in
the order of BaLa0.9Ca0.1InO3.95−BaLa0.9Sr0.1InO3.95−BaLa0.9Ba0.1InO3.95, i.e., in the order of increasing
the ionic radius of dopants. The concentration of the oxygen vacancies was the same for the doped
samples, so the difference in the conductivity of the BaLa0.9M0.1InO3.95 might be attributed to their
mobilities. In this respect, effect of the dopant nature on the conductivity should be considered. p
,
p
y
The authors [13] have investigated NdBaInO4, doped with Ca2+, Sr2+, Ba2+. Their results
show another trend in comparison with the data obtained in this work: σion of Ca2+ > Sr2+ > Ba2+. Their conductivities showed the tendency to increase with a decrease in the ionic radii of the M site ions,
i.e., Ca2+ is the most favorable dopant on Nd3+ sites. The authors [13] indicate that the replacement
of Nd3+ by the cations with the most comparable size may minimize local structural relaxation and
this main reason may explain the enhancement of the oxygen vacancy conductivity in the order of
Ca2+ > Sr2+ > Ba2+. Taking into account the differences in the radii of La3+ and Ca2+, Sr2+, and Ba2+,
we could expect a similar trend in the change in conductivity of doped BaLaInO4. However, this is not
the case. Clearly, other reasons should be considered. The authors [13] have investigated NdBaInO4, doped with Ca2+, Sr2+, Ba2+. Their results
show another trend in comparison with the data obtained in this work: σion of Ca2+ > Sr2+ > Ba2+. Their conductivities showed the tendency to increase with a decrease in the ionic radii of the M site ions,
i.e., Ca2+ is the most favorable dopant on Nd3+ sites. The authors [13] indicate that the replacement
of Nd3+ by the cations with the most comparable size may minimize local structural relaxation and
this main reason may explain the enhancement of the oxygen vacancy conductivity in the order of
Ca2+ > Sr2+ > Ba2+. Taking into account the differences in the radii of La3+ and Ca2+, Sr2+, and Ba2+,
we could expect a similar trend in the change in conductivity of doped BaLaInO4. 3 4 Dry Conditions
3.4. Dry Conditions However, this is not
the case. Clearly, other reasons should be considered. There are not many reports in the literature on the relationships of O2−-transport in phases with
the RP-structure. However, for the ABO3 perovskites, there is a lot of information on the influence of
the nature of A-cations, B-cations, and dopants on the mobility of oxygen vacancies. In the general
case, there are many factors that affect the mobility of oxygen vacancies. For phases with the same
structure and with the same concentration of dopants, the main factors are generally identified. There are not many reports in the literature on the relationships of O2−-transport in phases with
the RP-structure. However, for the ABO3 perovskites, there is a lot of information on the influence of
the nature of A-cations, B-cations, and dopants on the mobility of oxygen vacancies. In the general
case, there are many factors that affect the mobility of oxygen vacancies. For phases with the same
structure and with the same concentration of dopants, the main factors are generally identified. (1) Geometric characteristics. It is well known that the size proportion of the A-cations and
B-cations is an important factor for high ionic conductivity. The relationship of structure parameters
and oxygen-ion transport properties had been studied by Wakamura [31]. The author plotted the
activation energy (Ea) versus lattice volume (V) for numerous ion conductors, and found a simple
inverse correlation between them, according to the expression Ea = Av/V2/3, where Av is a constant. The relationship indicates that the Ea value is decreased with a weakened long-range force due to the
volume expansion. A similar tendency was observed for the ionic conductivity of mixed ionic-electronic
conductors. For example, for Ln0.5M0.5FeO3-δ (M = Sr2+, Ba2+) [32], increasing the mismatch between
La3+ and M2+ radii weakens the metal-oxygen bonding and increases the oxygen mobility. Returning to the analysis of the experimental data, presented in this paper, we can make the
assumption concerning the reason for the enhancement of the oxygen vacancy conductivity in the
order of Ca2+–Sr2+–Ba2+ for doped BaLaInO4. The observed increase in the lattice volume and the
lattice parameters weakens the metal-oxygen bonding, and, as a consequence, increases the oxygen
mobility. It should be noted that the activation energy for oxygen ion conduction in doped BaLaInO4
decreased in the order of Ca2+ (0.86 eV)–Sr2+ (0.85 eV)–Ba2+ (0.82 eV). 3 4 Dry Conditions
3.4. Dry Conditions Thus, the observed increase in the ionic conductivity (Ar atmosphere) for doped phas 0.8
1.0
1.2
1.4
1.6
1.8
2.0
-9
-8
-7
-6
-5
-4
-3
-2
pH2O = 3.5•10
-5 atm
BaLaInO4
BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95
BaLa0.9Ba0.1InO3.95
log (S/cm)
filled signs - air
open signs - Ar
0.8
1.0
1.2
1.4
1.6
1.8
2.0
-9
-8
-7
-6
-5
-4
-3
-2
pH2O = 3.5•10
-5 atm
BaLaInO4
BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95
BaLa0.9Ba0.1InO3.95
10
3/Т, K
-1
log (S/cm)
filled signs - air
open signs - Ar 10
3/Т, K
-1
Figure 6 The temperature dependencies of the conductivities of BaLaInO4 and BaLa0 9M0 1InO3 95 (M
Figure 6. The temperature dependencies of the conductivities of BaLaInO4 and BaLa0.9M0.1InO3.95 (M
= Ca2+, Sr2+, and Ba2+) in dry air (filled signs) and dry Ar (open signs). Figure 6. The temperature dependencies of the conductivities of BaLaInO4 and BaLa0.9M0.1InO3.95
(M = Ca2+, Sr2+, and Ba2+) in dry air (filled signs) and dry Ar (open signs). Figure 6. The temperature dependencies of the conductivities of BaLaInO4 and BaLa0.9M0.1InO3.95 (M
= Ca2+, Sr2+, and Ba2+) in dry air (filled signs) and dry Ar (open signs). Thus the observed increase in the ionic conductivity (Ar atmosphere) for doped phases
)
y
(
g
)
y
( p
g
)
Thus, the observed increase in the ionic conductivity (Ar atmosphere) for doped phases
BaLa0.9M0.1InO3.95 in comparison with undoped composition BaLaInO4 was mainly attributed to the
Thus, the observed increase in the ionic conductivity (Ar atmosphere) for doped phases
BaLa0.9M0.1InO3.95 in comparison with undoped composition BaLaInO4 was mainly attributed to the Figure 6. The temperature dependencies of the conductivities of BaLaInO4 and BaLa0.9M0.1InO3.95 (M
= Ca2+, Sr2+, and Ba2+) in dry air (filled signs) and dry Ar (open signs). Thus the observed increase in the ionic conductivity (Ar atmosphere) for doped phases
)
y
(
g
)
y
( p
g
)
Thus, the observed increase in the ionic conductivity (Ar atmosphere) for doped phases
BaLa0.9M0.1InO3.95 in comparison with undoped composition BaLaInO4 was mainly attributed to the
Thus, the observed increase in the ionic conductivity (Ar atmosphere) for doped phases
BaLa0.9M0.1InO3.95 in comparison with undoped composition BaLaInO4 was mainly attributed to the Materials 2019, 12, 1668 9 of 14 increase of the carrier concentration as a result of the formation of oxygen vacancies during doping. 3 4 Dry Conditions
3.4. Dry Conditions As can be seen, the results obtained for changes in oxygen-ion conductivity are consistent with
the main trends described for perovskites. At the same time, it should be emphasized that the nature
of the dopant has little effect on the oxygen ionic conductivity of doped BaLaInO4. The differences
reached less than 0.3 order at low temperatures. This is due to the fact that the replacement of cations
is carried out in the salt block and the change in geometrical parameters also occurs mainly in the
salt block. The formation of oxygen vacancies and their migration occurs in the perovskite block. Therefore, the nature of the dopant introduced into the salt block does not significantly affect the
oxygen ionic conductivity. (2) Binding energies for dopant−oxygen vacancy pair clusters. It is well-known that interactions
between dopant ions and their charge-compensating defects can lead to the formation of clusters, (2) Binding energies for dopant−oxygen vacancy pair clusters. It is well-known that interactions
between dopant ions and their charge-compensating defects can lead to the formation of clusters, Materials 2019, 12, 1668 10 of 14 which can trap the migrating species. This implies that the oxide ion vacancies might be strongly
bound to the dopant species, and the migration of the oxygen vacancy might depend on the binding
energy of the dopant-oxygen vacancy, which, consequently, leads to lower oxide ion conductivities. As for the calculations for dopant-vacancy clusters in Ca2+-, Sr2+-, and Ba2+-doped NdBaInO4 with
RP-structure, it was shown [13] that the calculated binding energies are comparable to ca.–0.9 eV,
which indicates that trapping for the migrating oxygen vacancies in NdBaInO4 from the Ca2+, Sr2+,
and Ba2+ dopants could be comparable. If we assume that the same tendency is observed for the doped
phase BaLaInO4, then it can also be said that the reason for the cluster formation does not explain the
change in conductivity due to the nature of the dopant. This result shows that the nature of the dopant, by occupying sites within the rock salt layers,
does not have a very significant effect on the oxygen-ion conduction and that the key parameter
governing the oxygen-ion conductivity of these materials is the concentration of oxygen vacancies. Likely, the variation in the concentration of oxygen vacancies of solid solutions, which will be studied
in the future, will reveal new regularities of ionic transport in phases with the RP-structure. 3 4 Dry Conditions
3.4. Dry Conditions The comparison of the conductivities of the investigated samples based on BaLaInO4 with
Nd-structural analogues BaNdInO4 [13] showed the total and ionic conductivities were in good
agreement and comparable for Sr2+ and Ba2+ doped phases at low temperatures, but Nd-structural
analogues exhibited higher conductivities at higher temperatures [13]. Figure 5 shows a comparison of
the temperature dependencies of conductivity for undoped phases BaLaInO4 and BaNdInO4 and Ba2+
doped phases BaLa0.9Ba0.1InO3.95 (or Ba1.1La0.9InO3.95), Ba1.1Nd0.9InO3.95 [13]. 3.5. Wet Conditions The proton conductivity was calculated as the difference between the conductivity in wet and dry
Ar, that is, σH = σ (wet Ar) −σ (dry Ar), and the temperature dependencies are shown in Figure 8. (M = Ca2+, Sr2+, Ba2+) at dry Ar (filled signs) and wet Ar (closed signs). 1.4
1.6
1.8
-8
-7
-6
-5
-4
BaLaInO4
BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95
BaLa0.9Ba0.1InO3.95
10
3/Т, K
-1
log σH
+ (S/cm)
Figure 8. The temperature dependencies of the proton conductivities of BaLaInO4 and
B L
M I O
(M
C
2+ S 2+
d B
2+)
Figure 8. The temperature dependencies of the proton conductivities of BaLaInO4 and BaLa0.9M0.1InO3.95
(M = Ca2+, Sr2+, and Ba2+). Figure 8. The temperature dependencies of the proton conductivities of BaLaInO4 and
Figure 8. The temperature dependencies of the proton conductivities of BaLaInO4 and BaLa0.9M0.1InO3.95
(M = Ca2+, Sr2+, and Ba2+). BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, and Ba2+). The proton transport numbers were calculated according to the formula tH = σH/σ (wet air) and
were plotted as a function of temperatures in Figure 9. It can be seen that the proton transport
numbers increased with decreasing temperature and the investigated samples were nearly pure
proton conductors below 450 °C. As can be seen from Figure 8, the proton conductivities of doped samples increased in the order
Ca2+–Sr2+–Ba2+. We can assume that these differences were due to an increase in the concentration
of protons in the same order. The activation energies decreased in the order of Ca2+ (0.70 eV)–Sr2+
(0.68 eV)–Ba2+ (0.65 eV). These values are greater than the activation energy of proton transport in
doped perovskites ~0.5 eV [33]. As was discussed above in the analysis of TG data, during hydration,
the main part of the protons was in the salt block. This situation is like hydroxides. It is well known
that proton mobility is dependent on the strength of hydrogen bonds with the next-nearest oxygen
atoms and energy barriers are high for proton transport in non-H-bonded systems [34,35]. It is clear
that proton migration in the layer of salt block is a difficult in comparison with the proton migration in
the bulk of perovskite block. 3.5. Wet Conditions Figure 7 shows a comparison of the temperature dependencies of conductivity of investigated
samples in Ar under dry and wet conditions. For wet Ar (pH2O = 2 × 10−2 atm) (Figure 7, curves with
open signs), the conductivities of all samples were higher than those in dry Ar. Clearly, this behavior
was a result of the insertion of water and the formation of proton charge carriers. The conductivity
differences were more significant at lower temperatures as the hydration increased with decreasing
temperature. We can suppose that the nature of dopant does not have significant effects on the type of
dominant charge carrier, and increasing in the conductivity values in wet atmospheres is due to the
appearance of proton carriers for all investigated samples. Materials 2019, 12, x FOR PEER REVIEW
11 of 14 0.8
1.0
1.2
1.4
1.6
1.8
2.0
-9
-8
-7
-6
-5
-4
-3
-2
closed signs - pH2O = 3.5•10
-5 atm
open signs - pH2O = 2•10
-2 atm
BaLaInO4
BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95
BaLa0.9Ba0.1InO3.95
10
3/Т, K
-1
log ion (S/cm)
Figure 7. The temperature dependencies of the ionic conductivities of BaLaInO4 and BaLa0.9M0.1InO3.95
(M = Ca2+, Sr2+, Ba2+) at dry Ar (filled signs) and wet Ar (closed signs). Figure 7. The temperature dependencies of the ionic conductivities of BaLaInO4 and BaLa0.9M0.1InO3.95
(M = Ca2+, Sr2+, Ba2+) at dry Ar (filled signs) and wet Ar (closed signs). 0.8
1.0
1.2
1.4
1.6
1.8
2.0
-9
-8
-7
-6
-5
-4
-3
-2
closed signs - pH2O = 3.5•10
-5 atm
open signs - pH2O = 2•10
-2 atm
BaLaInO4
BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95
BaLa0.9Ba0.1InO3.95
10
3/Т, K
-1
log ion (S/cm) Figure 7. The temperature dependencies of the ionic conductivities of BaLaInO4 and BaLa0.9M0.1InO3.95
(M = Ca2+, Sr2+, Ba2+) at dry Ar (filled signs) and wet Ar (closed signs). Figure 7. The temperature dependencies of the ionic conductivities of BaLaInO4 and BaLa0.9M0.1InO3.95
(M = Ca2+, Sr2+, Ba2+) at dry Ar (filled signs) and wet Ar (closed signs). 11 of 14
nO Materials 2019, 12, 1668
h The proton conductivity was calculated as the difference between the conductivity in wet and dry
Ar, that is, σH = σ (wet Ar) −σ (dry Ar), and the temperature dependencies are shown in Figure 8. (M = Ca2+, Sr2+, Ba2+) at dry Ar (filled signs) and wet Ar (closed signs). The result
4. Conclusions p
g
p
g
proton-conducting oxide in future applications. Further modification of the composition will
improve the H+-conductivity. 4. Conclusions
The new phases BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, and Ba2+) with the Ruddlesden-Popper
structure were obtained. The analysis of the effect of nature of dopant (alkaline earth metals) on the
hydration processes and on the transport properties of BaLaInO4 was carried out. It was established
that all investigated samples were capable of water uptake from the gas phase. The abilityof water
The new phases BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, and Ba2+) with the Ruddlesden-Popper
structure were obtained. The analysis of the effect of nature of dopant (alkaline earth metals) on the
hydration processes and on the transport properties of BaLaInO4 was carried out. It was established
that all investigated samples were capable of water uptake from the gas phase. The ability of water
incorporation was due to, not only the presence of oxygen vacancies, but also due to the presence in
the structure of La-O blocks. The degree of hydration increased with a growth in the size of the dopant,
i.e., with an increase in the size of the salt blocks, that contributed to the placement of a larger number
of OH−-groups, which occupied interstitial sites within the rock salt layers. that all investigated samples were capable of water uptake from the gas phase. The ability of water
incorporation was due to, not only the presence of oxygen vacancies, but also due to the presence in
the structure of La-O blocks. The degree of hydration increased with a growth in the size of the
dopant, i.e., with an increase in the size of the salt blocks, that contributed to the placement of a larger
number of OH−-groups, which occupied interstitial sites within the rock salt layers. It was proven that doping led to the increase of the oxygen ionic conductivity values. The
conductivity values for doped samples BaLa0.9M0.1InO3.95 were higher than for undoped composition
BaLaInO4 at a ~1.5 order of magnitude. The increase in the conductivity values was mainly attributed
to the increase of the carrier concentration as a result of the formation of oxygen vacancies during
doping. The oxygen ionic conductivity of doped samples increased in the order of BaLa0 9Ca0 1InO395–
It was proven that doping led to the increase of the oxygen ionic conductivity values. 3.5. Wet Conditions Moreover, from a general point of view, four key steps for hydration can
be distinguished [36]: these are (1) water adsorption on the oxide surface, (2) proton migration from the
surface to the bulk of the oxide, (3) proton migration in the bulk, and (4) oxide ion vacancy migration
in the bulk. The proton migration from the surface to the bulk is difficult. For investigated systems,
the proton migration from the salt layer is assumed to be the same as the proton migration from the
surface to the bulk of the perovskite structure. Taking into account these differences in the structures of
perovskites and phases with the RP-structure, as well as differences in hydration mechanisms, it can
be concluded that the proton transport in the investigated phases is realized with higher activation
energies compared to the substituted ordinary perovskites. The proton transport numbers were calculated according to the formula tH = σH/σ (wet air)
and were plotted as a function of temperatures in Figure 9. It can be seen that the proton transport
numbers increased with decreasing temperature and the investigated samples were nearly pure proton
conductors below 450 ◦C. The results indicate that new phases BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, Ba2+) might be a
promising proton-conducting oxide in future applications. Further modification of the composition
will improve the H+-conductivity. 12 of 14
12 of 14 Materials 2019, 12, 1668
Materials 2019, 12, x FOR 300
400
500
600
700
0.0
0.2
0.4
0.6
0.8
1.0
BaLaInO4
BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95
BaLa0.9Ba0.1InO3.95
t,
oC
tH
+
Figure 9. The temperature dependencies of the proton transport numbers of BaLaInO4 and
BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, and Ba2+). Figure 9. The temperature dependencies of the proton transport numbers of BaLaInO4 and
BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, and Ba2+). 300
400
500
600
700
0.0
0.2
0.4
0.6
0.8
1.0
BaLaInO4
BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95
BaLa0.9Ba0.1InO3.95
t,
oC
tH
+ Figure 9. The temperature dependencies of the proton transport numbers of BaLaInO4 and
BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, and Ba2+). Figure 9. The temperature dependencies of the proton transport numbers of BaLaInO4 and
BaLa0.9M0.1InO3.95 (M = Ca2+, Sr2+, and Ba2+). formation of proton charge carriers. The proton conductivity was estimated by subtracting the
conductivity data for dry Ar from the corresponding data in wet Ar. The proton conductivities of
doped samples increased in the order Ca2+–Sr2+–Ba2+ due to an increase in the concentration of
Author Contributions: Conceptualization, I.A. and N.T.; methodology, I.A.; investigation, A.G. and D.K.; data
curation, N.T. and A.G.; writing—original draft preparation, N.T.; writing—review and editing, I.A. doped samples increased in the order Ca
Sr
Ba due to an increase in the concentratio
protons. The activation energies for the proton conductivities were 0.70–0.68–0.65 eV, respectiv
Funding: A grant from the President of the Russian Federation (project MK-24.2019.3) supported this resear formation of proton charge carriers. The proton conductivity was estimated by subtracting the
conductivity data for dry Ar from the corresponding data in wet Ar. The proton conductivities of
doped samples increased in the order Ca2+–Sr2+–Ba2+ due to an increase in the concentration of
protons. The activation energies for the proton conductivities were 0.70–0.68–0.65 eV, respectively.
Author Contributions: Conceptualization, I.A. and N.T.; methodology, I.A.; investigation, A.G. and D.K.; data
curation, N.T. and A.G.; writing—original draft preparation, N.T.; writing—review and editing, I.A.
Funding: A grant from the President of the Russian Federation (project MK-24.2019.3) supported this research. References 1. Shim, J.H. Ceramics breakthrough. Nat. Energy 2018, 3, 168–169. [CrossRef] 2. Fabbri, E.; Bi, L.; Pergolesi, D.; Traversa, E. Towards the next generation of solid oxide fuel cells operating
below 600 ◦C with chemically stable proton-conducting electrolytes. Adv. Mater. 2012, 24, 195–208. [CrossRef]
3. Marrony, M.; Dailly, J. Advanced Proton Conducting Ceramic Cell as Energy Storage Device. ECS Trans. 2. Fabbri, E.; Bi, L.; Pergolesi, D.; Traversa, E. Towards the next generation of solid oxide fuel cells operating
below 600 ◦C with chemically stable proton-conducting electrolytes. Adv. Mater. 2012, 24, 195–208. [CrossRef] ,
;
,
;
g
,
;
,
g
p
g
below 600 ◦C with chemically stable proton-conducting electrolytes. Adv. Mater. 2012, 24, 195–208. [CrossRef]
3. Marrony, M.; Dailly, J. Advanced Proton Conducting Ceramic Cell as Energy Storage Device. ECS Trans. 2017, 78, 3349–3363. [CrossRef] 3. Marrony, M.; Dailly, J. Advanced Proton Conducting Ceramic Cell as Energy Storage Device. ECS Trans. 2017, 78, 3349–3363. [CrossRef] 4. Norby, T. Advances in proton ceramic fuel cells, steam electrolyzers, and dehydrogenation reactors based on
materials and process optimizations. ECS Trans. 2017, 80, 23–32. [CrossRef] 5. Colomban, P.; Zaafrani, O.; Slodczyk, A. Proton Content and Nature in Perovskite Ceramic Membranes for
Medium Temperature Fuel Cells and Electrolysers. Membranes 2012, 2, 493–509. [CrossRef] 6. Kochetova, N.; Animitsa, I.; Medvedev, D.; Demin, A.; Tsiakaras, P. Recent activity in the development of
proton-conducting oxides for high-temperature applications. RSC Adv. 2016, 6, 73222–73268. [CrossRef] 7. Duan, C.; Tong, J.; Shang, M.; Nikodemski, S.; Sanders, M.; Ricote, S.; Almansoori, A.; O’Hayre, R. Readily
processed protonic ceramic fuel cells with high performance at low temperatures. Science 2015, 349, 1321–1326. [CrossRef] [PubMed] 8. Dubois, A.; Ricote, S.; Braun, R. Benchmarking the expected stack manufacturing cost of next generation,
intermediate-temperature protonic ceramic fuel cells with solid oxide fuel cell technology. J. Power Sources
2017, 369, 65–77. [CrossRef] 9. De Souza, E.C.C.; Muccillo, R. Properties and applications of perovskite proton conductors. Mater. Res. 2010, 13, 385–394. [CrossRef] 10. Duan, C.; Kee, R.J.; Zhu, H.; Karakaya, C.; Chen, Y.; Ricote, S.; Jarry, A.; Crumlin, E.J.; Hook, D.; Braun, R.; et al. Highly durable, coking and sulfur tolerant, fuel-flexible protonic ceramic fuel cells. Nature 2018, 557, 217–222. [CrossRef] [PubMed] 11. Choi, S.; Kucharczyk, C.J.; Liang, Y.; Zhang, X.; Takeuchi, I.; Ji, H.-I.; Haile, S.M. Exceptional power density
and stability at intermediate temperatures in protonic ceramic fuel cells. Nat. Energy 2018, 3, 202–210. The result
4. Conclusions The conductivity values for doped samples BaLa0.9M0.1InO3.95 were higher than for undoped
composition BaLaInO4 at a ~1.5 order of magnitude. The increase in the conductivity values
was mainly attributed to the increase of the carrier concentration as a result of the formation of
oxygen vacancies during doping. The oxygen ionic conductivity of doped samples increased in the
order of BaLa0.9Ca0.1InO3.95–BaLa0.9Sr0.1InO3.95–BaLa0.9Ba0.1InO3.95, i.e., in the order of increasing the
ionic radius of dopants. The increase in the lattice volume and the lattice parameters weakens the
metal-oxygen bonding, and, as a consequence, increases the oxygen mobility. The activation energy
for oxygen ion conduction decreased in the same order. doping. The oxygen ionic conductivity of doped samples increased in the order of BaLa0.9Ca0.1InO3.95
BaLa0.9Sr0.1InO3.95–BaLa0.9Ba0.1InO3.95, i.e., in the order of increasing the ionic radius of dopants. The
increase in the lattice volume and the lattice parameters weakens the metal-oxygen bonding, and, as
a consequence, increases the oxygen mobility. The activation energy for oxygen ion conduction
decreased in the same order. The conductivity increased in wet conditions as a result of the insertion of water and the
The conductivity increased in wet conditions as a result of the insertion of water and the formation
of proton charge carriers. The proton conductivity was estimated by subtracting the conductivity
data for dry Ar from the corresponding data in wet Ar. The proton conductivities of doped samples
increased in the order Ca2+–Sr2+–Ba2+ due to an increase in the concentration of protons. The activation
energies for the proton conductivities were 0.70–0.68–0.65 eV, respectively. formation of proton charge carriers. The proton conductivity was estimated by subtracting the
conductivity data for dry Ar from the corresponding data in wet Ar. The proton conductivities of
doped samples increased in the order Ca2+–Sr2+–Ba2+ due to an increase in the concentration of
Author Contributions: Conceptualization, I.A. and N.T.; methodology, I.A.; investigation, A.G. and D.K.; data
curation, N.T. and A.G.; writing—original draft preparation, N.T.; writing—review and editing, I.A. 13 of 14 Materials 2019, 12, 1668 13 of 14 Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the
study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to
publish the results. References [CrossRef] 12. Fujii, K.; Shiraiwa, M.; Esaki, Y. Improved oxide-ion conductivity of NdBaInO4 by Sr doping. J. Mater. Chem. A 2015, 3, 11985–11990. [CrossRef] 13. Yang, X.; Liu, S.; Lu, F.; Xu, J.; Kuang, X. Acceptor Doping and Oxygen Vacancy Migration in Layered
Perovskite NdBaInO4-Based Mixed Conductors. J. Phys. Chem. C 2016, 120, 6416–6426. [CrossRef] 14. Shiraiwa, M.; Fujii, K.; Esaki, Y.; Kim, S.J.; Lee, S.; Yashima, M. Crystal Structure and Oxide-Ion Conductivity
of Ba1+xNd1−xInO4−x/2. J. Electrochem. Soc. 2017, 164, F1392–F1399. [CrossRef] 15. Birgeneau, R.J.; Guggenheim, H.J.; Shirane, G. Neutron scattering investigation of phase transitions and
magnetic correlations in the two-dimensional antiferromagnets K2NiF4, Rb2MnF4, Rb2FeF4. Phys. Rev. B
1970, 1, 2211–2230. [CrossRef] 16. Cochrane, A.K.; Telfer, M.; Dixon, C.A.L.; Zhang, W.; Halasyamani, P.S.; Bousquet, E.; Lightfoot, P. NdBaScO4:
Aristotype of a new family of geometric ferroelectrics. Chem. Commun. 2016, 52, 10980–10983. [CrossRef] 17. Fuji, K.; Yashima, M. Discovery and development of BaNdInO4—A brief review. J. Ceram. Soc. Jpn. 2018, 126, 852–859. [CrossRef] 18. Ishihara, T.; Yan, Y.; Sakai, T.; Ida, S. Oxide ion conductivity in doped NdBaInO4. Solid State Ion. 2016, 288, 262–265. [CrossRef] 19. Troncoso, L.; Alonso, J.A.; Fernández-Díaz, M.T.; Aguadero, A. Introduction of interstitial oxygen atoms in
the layered perovskite LaSrIn1−xBxO4+δ system (B = Zr, Ti). Solid State Ion. 2015, 282, 82–87. [CrossRef] 20. Troncoso, L.; Alonso, J.A.; Aguadero, A. Low activation energies for interstitial oxygen conduction in the
layered perovskites La1+xSr1−xInO4+δ. J. Mater. Chem. A 2015, 3, 17797–17803. [CrossRef] 21. Tarasova, N.;
Animitsa, I. Protonic transport in oxyfluorides Ba2InO3F and Ba3In2O5F2 with
Ruddlesden-Popper structure. Solid State Ion. 2015, 275, 53–57. [CrossRef] Materials 2019, 12, 1668 14 of 14 14 of 14 22. Li, X.; Shimada, H.; Ihara, M. Conductivity of New Electrolyte Material Pr1−xM1+xInO4 (M = Ba, Sr) with
Related Perovskite Structure for Solid Oxide Fuel Cells. ECS Trans. 2013, 50, 3–14. [CrossRef] 23. Matsuhira, T.; Kurahashi, Y.; Hasegawa, K.; Ihara, M. Proton Conducting Properties of Sr1+xLn1−xAlO4−δ
(Ln = Pr, Sm) with Layered Perovskite Structure for Solid Oxide Fuel Cells. In Proceedings of the 232nd ECS
Meeting, National Harbor, MD, USA, 1–5 October 2017. 24. Tyitov, Y.O.; Byilyavina, N.M.; Markyiv, V.Y.; Slobodyanik, M.S.; Krajevs’ka, Y.A. Synthesis and crystal
structure of BaLaInO4 and SrLnInO4 (Ln−La, Pr). Dopovyidyi Natsyional'noyi Akademyiyi Nauk Ukrayini
2009, 10, 160–166. 25. Rietveld, H.M. A profile refinement method for nuclear and magnetic structures. J. App. Cryst. 1969, 2, 65–71. [CrossRef] 26. Shannon, R.D. References Revised effective ionic radii and systematic studies of interatomic distances in halides and
chalcogenides. Act. Cryst. 1976, A32, 751–767. [CrossRef] 27. Jedvik, E.; Lindman, A.; Benediktsson, M.; Wahnström, G. Size and shape of oxygen vacancies and protons
in acceptor-doped barium zirconate. Solid State Ion. 2015, 275, 2–8. [CrossRef] 28. Norby, T. Chapter 11. Proton Conductivity in Perovskite Oxides. In Perovskite Oxide for Solid Oxide Fu
Ishihara, T., Ed.; Springer Science+Business Media, LLC: Berlin/Heidelberg, Germany, 2009; pp. 217 28. Norby, T. Chapter 11. Proton Conductivity in Perovskite Oxides. In Perovskite Oxide for Solid Oxide Fuel Cells;
Ishihara, T., Ed.; Springer Science+Business Media, LLC: Berlin/Heidelberg, Germany, 2009; pp. 217–241. 29. Shpanchenko, R.V.; Antipov, E.V.; Kovba, L.M. Ba2ZrO4 and its crystallohydrates. Russ. J. Inorg. Chem. 9. Shpanchenko, R.V.; Antipov, E.V.; Kovba, L.M. Ba2ZrO4 and its crystallohydrates. Russ. J. Inorg. C
1993, 38, 599. 30. Lehtimäki, M. Stability, Cation Ordering and Oxygen Non-Stoichiometry of Some Perovskites and Related
Layered Oxides. Ph.D. Dissertation, Aalto University, Aalto, Finland, 2013. 31. Wakamura, K. Ion conduction in proton- and related defect (super) ionic conductors: Mechanical, electronic
and structure parameters. Solid State Ion. 2009, 180, 1343–1349. [CrossRef] 32. Kharton, V.V.; Marques, F.M.B.; Kilner, J.A.; Atkinson, A. Oxygen Ion-Conducting Materials in Solid State
Electrochemistry I: Fundamentals, Materials and Their Applications; Kharton, V.V., Ed.; WILEY-VCH Verlag
GmbH & Co.: Weinheim, Germany, 2009; ISBN 978-3-527-32318-0. 33. Liang, K.C.; Du, Y.; Nowick, A.S. Fast high-temperature proton transport in nonstoichiometric mixed
perovskites. Solid State Ion. 1994, 69, 117–120. [CrossRef] 4. Wakamura, K. Empirical relationships for ion conduction based on vibration amplitude in perovskite-t
proton and superionic conductors. J. Phys. Chem. Solids 2005, 66, 133–142. [CrossRef] proton and superionic conductors. J. Phys. Chem. Solids 2005, 66, 133–142. [CrossRef]
35. Proton Conductors: Solids, Membranes and Gels—Materials and Devices (Chemistry of Solid State Materials);
Colomban, P. (Ed.) Cambridge University Press: Cambridge, UK, 1992. [CrossRef] 35. Proton Conductors: Solids, Membranes and Gels—Materials and Devices (Chemistry of Solid State Materials);
Colomban, P. (Ed.) Cambridge University Press: Cambridge, UK, 1992. [CrossRef] 36. Jing, Y.; Matsumoto, H.; Aluru, N.R. Mechanistic Insights into Hydration of Solid Oxides. Chem. Mater. 2018, 30, 138–144. [CrossRef] © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
|
https://openalex.org/W1932525912
|
https://ojvr.org/index.php/ojvr/article/download/888/1288
|
English
| null |
Occurrence of porcine cysticercosis in free-ranging pigs delivered to slaughter points in Arapai, Soroti district, Uganda
|
Onderstepoort journal of veterinary research
| 2,015
|
cc-by
| 3,767
|
Occurrence of porcine cysticercosis in free-ranging
pigs delivered to slaughter points in Arapai,
Soroti district, Uganda Poverty, hunger and the need for production of pigs with meagre or zero inputs have made most
farmers release their pigs to range freely, thus creating a pig-human cycle that maintains Taenia
solium, the pig tapeworm and cause of porcine cysticercosis, in the ecosystem. A preliminary
study was designed to establish the prevalence of porcine cysticercosis by postmortem
examination of the tongue and carcass of free-range pigs from February to April 2014 in
Arapai subcounty, Soroti district, eastern Uganda. The tongue of each pig was extended and
examined before deep incisions were made and the cut surfaces were examined. The rest of the
carcasses were examined for cysts. Out of 178 pigs examined, 32 were qualitatively positive
for porcine cysticercosis, representing a prevalence of 18.0%. This high prevalence represents a
marked risk to the communities in the study area of neurocysticercosis, a debilitating parasitic
zoonosis. Proper human waste disposal by use of pit latrines, confinement of free-range pigs
and treatment with albendazole and oxfendazole are recommended. 2Department of Livestock
Health and Entomology,
Ministry of Agriculture,
Animal Industry and
Fisheries, Uganda Page 1 of 5
Original Research Page 1 of 5
Original Research Introduction Pig production has increasingly become an important activity in Uganda, with the pig population
increasing in the last three decades from 0.19 million to 3.6 million (Ministry of Agriculture,
Animal Industry and Fisheries [MAAIF] & Uganda Bureau of Statistics 2009; Uganda Bureau
of Statistics 2013). In comparison to other animal rearing enterprises, pig production requires
minimal inputs and relatively smaller space (Eusebio 1980), which makes pig farming popular. It is thus not surprising that more than 1.1 million families, about 18% of the total households in
Uganda, own pigs (MAAIF & Uganda Bureau of Statistics 2009). Email: This rapid increase in production has been matched by a rapid increase in consumption of pork
within the country, driven not only by population growth but also by a combination of rising
income and changing preferences associated with urbanisation. Uganda has the highest per
capita consumption of pork in sub-Saharan Africa, with a 2011 estimate of 3.4 kg/person/year
representing a ten-fold increase in the last 30 years (Ballantyne 2012). However, programmes
promoting pig production have not emphasised proper management and public health concerns. Poverty and lack of resources have driven farmers or communities to rear pigs extensively with
very minimal inputs, hence exposing them to the risk of porcine cysticercosis. Porcine cysticercosis
is caused by the metacestodes (cysticerci) of the cestode Taenia solium, and it is endemic in Uganda
(World Organization for Animal Health 2014; Waiswa et al. 2009). How to cite this article:
Zirintunda, G. & Ekou, J.,
2015, ‘Occurrence of
porcine cysticercosis in
free-ranging pigs delivered
to slaughter points in Arapai,
Soroti district, Uganda’,
Onderstepoort Journal of
Veterinary Research 82(1),
Art. #888, 5 pages. http://
dx.doi.org/10.4102/ojvr. v82i1.888 Taenia solium is a zonootic tapeworm that is maintained by a pig-human cycle in the ecosystem. The infection is contracted by pigs when they either ingest human faeces containing infective eggs
or when feeding on pasture contaminated with T. solium eggs (Carrique-Mas et al. 2001). If people
consume raw or inadequately cooked pork from infected animals, the larval cysts can develop into
the adult stage tapeworm in their intestines, where gravid proglottids containing infective eggs
detach from the adult tapeworm and are excreted in the faeces (Garcia et al. 2003). In places where
open defaecation is common the faeces containing these infective eggs are consumed directly or from
contaminated pasture by pigs, and the lifecycle is perpetuated (Ito et al. 2006; Lescano et al. 2007). Copyright:
© 2015. The Authors. Licensee: AOSIS
OpenJournals. This work is
licensed under the Creative
Commons Attribution
License. Read online:
Scan this QR
code with your
smart phone or
mobile device
to read online. Copyright:
© 2015. The Authors. Licensee: AOSIS
OpenJournals. This work is
licensed under the Creative
Commons Attribution
License. Copyright:
© 2015. The Authors.
Licensee: AOSIS
OpenJournals. This work is
licensed under the Creative
Commons Attribution
License. Copyright: Humans can also act as an aberrant intermediate host for T. solium if there is faecal-oral
contamination with the infective eggs. In such cases the larval stage can be found in human muscle,
heart, eyes, skin or central nervous system, causing human cysticercosis (Flisser, Rodríguez-
Canul & Willingham 2006). The most serious form of human cysticercosis is when the larval form
develops in the brain, a condition called neurocysticercosis (NCC). Adult T. solium infestation
in humans is associated with subclinical conditions of malnutrition and anasarca due to larval
migration through the tissues (Delgado-Azanero et al. 2007). Human NCC may manifest with
headaches, blindness, hydrocephalus, chronic meningitis and dementia (Carabin et al. 2005). NCC Page 2 of 5
Original Research Source: Authors’ own creation
FIGURE 1: Map of Uganda showing location of Soroti District (shaded red). Source: Authors’ own creation URE 1: Map of Uganda showing location of Soroti District (shaded red). FIGURE 1: Map of Uganda showing location of Soroti District (shaded red). Age increases the risk of being positive for cysticercosis
in pigs where open-air defaecation and free-range pig
raising are practised (Jayashi et al. 2012). Approximately
50 000 individuals die globally every year of NCC caused
by the parasitic intermediate stages of T. solium (Ito et al. 2006). The purpose of the present study was therefore to
estimate the prevalence of porcine cysticercosis amongst
pigs delivered for slaughter in Arapai in the Soroti district
of eastern Uganda. Age increases the risk of being positive for cysticercosis
in pigs where open-air defaecation and free-range pig
raising are practised (Jayashi et al. 2012). Approximately
50 000 individuals die globally every year of NCC caused
by the parasitic intermediate stages of T. solium (Ito et al. 2006). The purpose of the present study was therefore to
estimate the prevalence of porcine cysticercosis amongst
pigs delivered for slaughter in Arapai in the Soroti district
of eastern Uganda. contributes to epilepsy in regions where pigs are free-ranging
and hygiene is poor (Blocher et al. 2011; Rottbeck et al. 2013). The prevalence of cysticercosis was determined to be 11.7%
amongst patients with epilepsy and 2.8% amongst controls
who were normal individuals in families of Burundi (Newell
et al. 1997), indicating that cysticercosis causes epilepsy. contributes to epilepsy in regions where pigs are free-ranging
and hygiene is poor (Blocher et al. 2011; Rottbeck et al. 2013). Copyright: The prevalence of cysticercosis was determined to be 11.7%
amongst patients with epilepsy and 2.8% amongst controls
who were normal individuals in families of Burundi (Newell
et al. 1997), indicating that cysticercosis causes epilepsy. There is evidence of a high prevalence of NCC infecting
people in villages where pigs are raised (Phiri et al. 2003). http://www.ojvr.org doi:10.4102/ojvr.v82i1.888 doi:10.4102/ojvr.v82i1.888 Page 3 of 5
Original Research
Source: Authors’ own creation
FIGURE 2 M
S
ti Di t i t h
i
l
ti
f A
i ( h d d) Page 3 of 5
Original Research Page 3 of 5 Source: Authors’ own creation Source: Authors’ own creation FIGURE 2: Map Soroti District showing location of Arapai (shaded). FIGURE 2: Map Soroti District showing location of Arapai (shaded). Materials and methods masseters and the pterygoid muscles were examined and
two incisions made into each, the cuts being parallel to the
bone and right through the muscle. The pericardium was
examined visually. The heart was incised once lengthwise
through the left ventricle and interventricular septum
to expose the interior and cut surfaces for examination. In addition two deep incisions were made into the left
ventricle. After removal of the peritoneum the muscles of the
diaphragm were examined visually and incised once. The
oesophagus was examined visually. The gracilis muscle was
incised once parallel to the pubic symphisis. The study was done in Arapai subcounty which is located in
the northern part of Soroti district in eastern Uganda. Figures
1 and 2 show the location of Soroti district in Uganda and the
location of Arapai subcounty in Soroti. The 16 most popular slaughter points for pig trade in the
subcounty were selected for inclusion in the study, namely
five slaughter places in Aloet parish, eight in Arapai market,
two in Temele trading centre and one in Apida trading centre. A cross-sectional qualitative study in which the tongue,
cardiac muscles and thigh muscles were examined for the
presence of cysts was performed. The head of a slaughtered
pig was set aside; the tongue was fully extended, dried with
a smooth towel and examined for cysts on both the dorsum
and ventrum before it was incised and the cut surfaces were
examined. Visual inspection of the carcass, its cut surfaces
and the organs within it was done. The external and internal All pigs were examined consistently. Cysts were identified
as oval, about 10 mm × 5 mm or larger, with a delicate, fairly
translucent, white parasite membrane and host capsule. Within the cyst a pale fluid and the scolex, visible as a white
dot within the cyst, usually invaginating midway along
the long axis of the cyst, was considered diagnostic (World
Organization for Animal Health 2014). doi:10.4102/ojvr.v82i1.888 doi:10.4102/ojvr.v82i1.888 http://www.ojvr.org Original Research Original Research Page 4 of 5 TABLE 1: Occurrence of porcine cysticercosis amongst free-range pigs delivered
to various slaughter places in Arapai subcounty, Soroti district, eastern Uganda. TABLE 1: Occurrence of porcine cysticercosis amongst free-range pigs delivered
to various slaughter places in Arapai subcounty, Soroti district, eastern Uganda. the environment and to consume human faeces. Conclusion In conclusion, the prevalence of porcine cysticercosis in
Arapai, Soroti district is very high. Since it is a zonoosis, the
human population in Arapai is at a high risk of the maladies
associated with porcine cysticercosis, like NCC and epilepsy. It is also possible for cysticercosis to occur in people without
brain involvement, and for clinical symptoms to appear to
be absent (Somers et al. 2006). Confinement of pigs should
be adopted to prevent continuous transmission of porcine
cysticercosis (Pouedet et al. 2002). Vaccination of pigs with
crude extracts of T. solium metacestodes and oncosphere
antigens (Flisser et al. 2004; Molinari et al. 1997) may also
be helpful. Treatment with drugs such as albendazole and
oxfendazole is of value, as the cysts may lose their fluid and
collapse. In conclusion, the prevalence of porcine cysticercosis in
Arapai, Soroti district is very high. Since it is a zonoosis, the
human population in Arapai is at a high risk of the maladies
associated with porcine cysticercosis, like NCC and epilepsy. Shey-Njila et al. (2003) found 4.44% by tongue examination
and 27.7% prevalence by ELISA in the same population. Sikasunge et al. (2007) found 12.7% by tongue examination
and 32.1% by ELISA, Pondja et al. (2010) 12.7% by tongue
examination and 34.9% by ELISA, Eshitera et al. (2012)
5.6% by tongue examination and 32.6% by ELISA,
Krecek et al. (2008) 11.9% by tongue examination and
33.3% by EITB, whilst Gonzalez et al. (1990) found 23.4%
by tongue examination, 37.7% by ELISA and 51.9%
by EITB. Shey-Njila et al. (2003) found 4.44% by tongue examination
and 27.7% prevalence by ELISA in the same population. Sikasunge et al. (2007) found 12.7% by tongue examination
and 32.1% by ELISA, Pondja et al. (2010) 12.7% by tongue
examination and 34.9% by ELISA, Eshitera et al. (2012)
5.6% by tongue examination and 32.6% by ELISA,
Krecek et al. (2008) 11.9% by tongue examination and
33.3% by EITB, whilst Gonzalez et al. (1990) found 23.4%
by tongue examination, 37.7% by ELISA and 51.9%
by EITB. Thus the observed prevalence was very high given the low
sensitivity of the test methods employed, which is seriously
concerning since the true prevalence is probably much
higher. Discussion The overall observed porcine cysticercosis prevalence in this
study was 18.0%. This was higher than the prevalence
determined by tongue and necropsy examination previously
reported in eastern and southern provinces of Zambia
(Mwape et al. 2012; Sikasunge et al. 2007), the Teso district of
Western Kenya (Mutua et al. 2007), North West Cameroon
(Shey-Njila et al. 2003), Angónia district of Mozambique
(Pondja et al. 2010), Homa Bay district of Kenya (Eshitera
et al. 2012) and the Eastern Cape Province, South Africa
(Krecek et al. 2008). The prevalence of cysticercosis was similar for Aloet and
Arapai sub-counties and for Temele and Apida slaughter
points; this was possibly because of almost the same
magnitude of those factors that affect porcine cysticercosis,
like latrine coverage and use. Aloet and Arapai slaughter
points had a slightly lower prevalence compared to Temele
and Apida. Aloet is a township with moderate pit latrine
coverage whilst Arapai is a cattle market with some sanitary
facilities. In addition, both places receive pigs for slaughter
from distant places and through middlemen who probably
carry out pre-transit lingual examination of the pigs before
delivery (Nsadha et al. 2014). A prevalence as high as 18.0% based on tests of low sensitivity
should cause concern. The tongue test is 70% sensitive and
100% specific; the enzyme-linked immunosorbent assay
(ELISA) has a sensitivity and specificity of 75%, whilst
enzyme-linked immune-electro transfer blot (EITB) has
a sensitivity and specificity of 100% (Gonzalez et al. 1990). The prevalence may be projected to be about 40% had more
sensitive tests such as EITB been used. Results As shown on Table 1, a total of 178 pigs were examined and
32 were found to be positive for cysticercosis, indicating
an overall prevalence of 18.0% in the study area. Aloet and
Arapai market recorded similar results of 14.9% and 14.6%
respectively, whilst Temele and Apida also had similar
results of 28.6% each. The pigs are not fed commercial rations since most owners
also have very little to eat, and therefore pigs are exposed
to human faeces whilst scavenging. The farmers in Soroti
have no regular strategies to control worms amongst
themselves or amongst their pigs. This could lead to a
higher prevalence of cysticercosis when compared to
other places where deworming programmes have been
implemented. Materials and methods In Soroti
pit latrine coverage is 71% and 94% of pigs are reared
either free-ranging or tethered in bushes where they are at
risk of acquiring porcine cysticercosis (Uganda Bureau of
Statistics 2009; Zirintunda 2011). Most farmers are poor and
food insecure, and equally their pigs lack sufficient food, as
reported in other studies (Adesehinwa, Makinde & Oladele
2003; Chimonyo et al. 2005; Halimani et al. 2012). The free-
ranging pigs are also able to move long distances away from
their owners’ premises where they access eggs of T. solium,
even if their owners are free from infection and have access
to latrines. Place of slaughter
Number of pigs
sampled
Positive
cases
% prevalence
Aloet (5 points)
47
07
14.9
Arapai market (8 points)
89
13
14.6
Temele (2 points)
07
02
28.6
Apida centre (1 point)
35
10
28.6
Total
178
32
18.0 References et al., 2003,
‘The emergence of Taenia solium cysticercosis in eastern and southern Africa as
a serious agricultural problem and public health risk’, Acta Tropica 87(1), 13–23. http://dx.doi.org/10.1016/S0001-706X(03)00051-2 Chimonyo, M., Bhebhe, E., Dzama, K., Halimani, T.E. & Kanengoni, A., 2005,
‘Improving smallholder pig production for food security and livelihood of
the poor in Southern Africa’, Africa Crop Science Conference Proceedings 7,
569–573. Pondja, A., Neves, L., Mlangwa, J., Afonso, S., Fatetine, J., Willingham III, A.L. et al.,
2010, ‘Prevalence and risk factors of porcine cysticercosis in Angonia District,
Mozambique’, PLoS Neglected Tropical Diseases 4(2), e594. http://dx.doi. org/10.1371/journal.pntd.0000594 Delgado-Azanero, W.A., Mosqueda-Taylor, A., Carlos-Bregni, R., Del Muro-Delgado, R.,
Diaz-Franco, M.A. & Contreras-Vidarre, E., 2007, ‘Oral cysticercosis: A collaborative
study of 16 cases’, Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology
Endocrinology 103(4), 528–533. http://dx.doi.org/10.1016/j.tripleo.2006.01.022 Pouedet, M.S., Zoli, A.P., Nguekam, A., Vondou, L., Assana, E., Speybroeck, N. et al.,
2002, ‘Epidemiological survey of swine cysticercosis in two rural communities
of West Cameroon’, Veterinary Parasitology 106(1), 45–54. http://dx.doi. org/10.1016/S0304-4017(02)00035-3 Eshitera, E.E., Githigia, S.M., Kitala, P., Thomas, L.F., Fevre, E.M., Harrison, L.J. et al.,
2012, ‘Prevalence of porcine cysticercosis and associated risk factors in Homa
Bay District, Kenya’, BMC Veterinary Research 8, 234. http://dx.doi.org/10.1186/
1746-6148-8-234 Rottbeck, R., Nshimiyimana, J.F., Tugirimana, P., Dull, U.E., Sattler, J., Hategekimana,
J.C. et al., 2013, ‘High prevalence of cysticercosis in people with epilepsy in
southern Rwanda’, PLoS Neglected Tropical Diseases 7(11), e2558. http://dx.doi. org/10.1371/journal.pntd.0002558 Eusebio, J.A., 1980, Pig Production in the Tropics, Intermediate Tropical Agriculture
Series, Longman Group, London. Shey-Njila, O., Zoli, P.A., Awah-Ndukum, J., Nguekam, A.E., Byambas, P., Dorny, P. et al., 2003, ‘Porcine cysticercosis in village pigs of North-West Cameroon’, Journal
of Helminthology 77(4), 351–354. http://dx.doi.org/10.1079/JOH2003179 Flisser, A., Gauci, C.G., Zoli, A., Ocana, J.M., Rodriguez, A.G., Dominguez, A.J.L. et al.,
2004, ‘Induction of protection against porcine cysticercosis by vaccination with
recombinant oncosphere antigens’, Infection & Immunity 72(9), 5292–5297. http://dx.doi.org/10.1128/IAI.72.9.5292-5297.2004 Sikasunge, C.S., Phiri, I.K., Phiri, A.M., Dorny, P., Siziya, S. & Willingham III, A.L., 2007,
‘Risk factors associated with porcine cysticercosis in selected districts of Eastern
and Southern provinces of Zambia’, Veterinary Parasitology 143(1), 59–66. http://
dx.doi.org/10.1016/j.vetpar.2006.07.023 Flisser, A., Rodríguez-Canul, R. & Willingham III, A.L., 2006, ‘Control of the taeniosis/
cysticercosis complex: Future developments’, Veterinary Parasitology 139(4),
283–292. http://dx.doi.org/10.1016/j.vetpar.2006.04.019 Somers, R., Dorny, P., Nguyen, V.K., Dang, T.C.T., Godderis, B., Craig, P.S. References Molinari, J.L., Rodriguez, D., Tato, P., Soto, R., Arechavaleta, F. & Solano, S., 1997, ‘Field
trial for reducing porcine Taenia solium cysticercosis in Mexico by systematic
vaccination of pigs’, Veterinary Parasitology 69(1–2), 55–63. http://dx.doi. org/10.1016/S0304-4017(96)01102-8 Adesehinwa, A.O.K., Makinde, G.E.O. & Oladele, O.I., 2003, ‘Socio-economic
characteristics of pig farmers as determinant of pig feeding pattern in Oyo state,
Nigeria’, Livestock Research for Rural Development 15(12), viewed 22 September
2014, from http://www.lrrd.org/lrrd15/12/ades1512.htm Mutua, F.K., Randolph, T.F., Arimi, S.M., Kitala, P.M., Githigia, S.M., Willingham, A.L. et
al., 2007, ‘Palpable lingual cysts; a possible indicator of porcine cysticercosis in Teso
District, Western Kenya’, Journal of Swine Health and Production 15(4), 206–212. Ballantyne, P., 2012, ‘Smallholder pigs value chain project to increase rural incomes
in Uganda’, CGIAR, Research Program on Livestock and Fish, viewed 20 June
2014, from http://livestockfish.cgiar.org/12/05/04/smallholder-pigs-value-chain-
project-to-increase-rural-incomes-in-uganda/ Mwape, K.E., Phiri, I.K., Praet, N., Muma, J.B., Zulu, G., Van Den Bossche, P. et al.,
2012, ‘Taenia solium infections in a rural area of Eastern Zambia - A community
based study’, PLoS Neglected Tropical Diseases 6(3), e1594. Blocher, J., Schmurtzhard, E., Wilkins, P.P., Gupton, P.N., Schaffert, M., Aner, H. et al.,
2011, ‘A cross-sectional study of people with epilepsy and neurocysticercosis in
Tanzania: Clinical characteristics and diagnostic approaches’, PLoS Neglected
Tropical Disesases 5(6), e185. http://dx.doi.org/10.1371/journal.pntd.0001185 Newell, E., Vyungimana, F., Geerts, S., Van Kerckhoven, I., Tsang, V.C. & Engels, D.,
1997, ‘Prevalence of cysticercosis in epileptics and members of their families
in Burundi’, Transactions of the Royal Society of Tropical Medicine and Hygiene
91(4), 389–391. http://dx.doi.org/10.1016/S0035-9203(97)90251-0 Carabin, H., Budke, C.M., Cowan, L.D., Willingham III, A.L. & Torgerson, P.R., 2005,
‘Methods for assessing the burden of parasitic zoonoses: Echinococcosis and
cysticercosis’, Trends in Parasitology 21(7), 327–333. http://dx.doi.org/10.1016/j. pt.2005.05.009 Nsadha, Z., Kawuma, P., Doble, L., Kivali, V., Eric, F., Ojok, L. et al., 2014, ‘Diagnostic
efficiency of meat inspection service to detect Taenia solium cysticercostic pork
at Wambizi pig abattoir, Kampala, Uganda: Implications for public health’, Africa
Journal of Animal and Biomedical Sciences 8(1), 17–22. Carrique-Mas, J., Iihoshi, N., Widdowson, M.A., Roca, Y., Morales, G., Quiroga, J. et al., 2001, ‘An epidemiological study of Taenia solium cysticercosis in a rural
population in the Bolivian Chaco’, Acta Tropica 80(3), 229–235. http://dx.doi. org/10.1016/S0001-706X(01)00161-9 Phiri, I.K., Ngowi, H., Afonso, S., Matenga, E., Boa, M., Mukaratirwa, S. Acknowledgements
Competing interests The higher prevalence in Arapai subcounty was thought
to be due to the very low latrine coverage and the free-
ranging rearing system, which enables pigs to scavenge in The authors declare that they have no financial or personal
relationships which may have inappropriately influenced
them in writing this article. doi:10.4102/ojvr.v82i1.888 doi:10.4102/ojvr.v82i1.888 http://www.ojvr.org Authors’ contributions Krecek, R.C., Michael, L.M., Schantz, P.M., Ntanjana, L., Smith, M.F., Dorny, P. et al.,
2008, ‘Prevalence of Taenia solium cysticercosis in swine from a community
based study in 21 villages of the Eastern Cape Province, South Africa’, Veterinary
Parasitology 154(1–2), 38–47. http://dx.doi.org/10.1016/j.vetpar.2008.03.005 G.Z. (Busitema University) was responsible for the study
design and data collection, while J.E. (Busitema University) G.Z. (Busitema University) was responsible for the study
design and data collection, while J.E. (Busitema University)
carried out the data analysis, interpretation, discussion and
drafting of the manuscript. Lescano, A.G., Garcia, H.H., Gilman, R.H., Guezala, M.C., Tsang, V.C.W., Gavidia, C.M. et al., 2007, ‘Swine cysticercosis hotspots surrounding Taenia solium tapeworm
carriers’, American Journal of Tropical Medicine and Hygiene 76(2), 376. carried out the data analysis, interpretation, discussion and
drafting of the manuscript. Ministry of Agriculture, Animal Industry and Fisheries & Uganda Bureau of Statistics,
2009, The National Livestock Census Report 2008, Ministry of Agriculture,
Animal Industry and Fisheries, Entebbe and Uganda Bureau of Statistics,
Kampala. References et al., 2006,
‘Taenia solium taeniasis and cysticercosis in three communities in North Vietnam’,
Tropical Medicine & International Health 11(1), 65–72. http://dx.doi.org/10.1111/
j.1365-3156.2005.01537.x Garcia, H.H., Gonzalez, A.E., Evans, C.A.W. & Gilman, R.H., 2003, ‘Taenia solium
cysticercosis’, Lancet 362(9383), 547–556. http://dx.doi.org/10.1016/S0140-
6736(03)14117-7 Uganda Bureau of Statistics, 2009. ‘Higher local government statistical abstract: Soroti
district’, viewed 20 June 2014, from www.ubos.org/.../Soroti%20stat%20%20
abstract%202009-Final.pdf Gonzalez, A.E., Cama, V., Gilman, R.H., Tsang, V.C, Pilcher, J.B., Chavera, A. et al.,
1990, ‘Prevalence and comparison of serologic assays, necropsy and tongue
examination for the diagnosis of porcine cysticercosis in Peru’, American Journal
of Tropical Medicine and Hygiene 43(2), 194–199. Uganda Bureau of Statistics, 2013, Statistical Abstract 2013, The Republic of Uganda,
Kampala. Halimani, T.E., Muchadeyi, F.C., Chimonyo, M. & Dzama, K., 2012, ‘Some insights into
the phenotypic and genetic diversity of indigenous pigs in southern Africa’, South
African Journal of Animal Science 42(5), 505–510. http://dx.doi.org/10.4314/
sajas.v42i5.13 Waiswa, C., Fèvre, E.M., Nsadha, Z., Sikasunge, C.S. & Willingham III, A.L., 2009, ‘Porcine
cysticercosis in Southeast Uganda: Seroprevalence in Kamuli and Kaliro Districts’,
Journal of Parasitology Research 28, 1–5. http://dx.doi.org/10.1155/2009/375493 Ito, A., Takayanagui, O.M., Sako, Y., Sato, M.O., Odashima, N.S., Yamasaki, H. et al.,
2006, ‘Neurocysticercosis: Clinical manifestation, neuroimaging, serology and
molecular confirmation of histopathologic specimens’, Southeast Asian Journal of
Tropical Medicine and Public Health 37(3), 74–88. World Organization for Animal Health, 2014, Cysticercosis, OIE Terrestrial Manual,
viewed 10 August 2014, from http://www.oie.int/fileadmin/Home/eng/Health_
standards/tahm/2.09.05_CYSTICERCOSIS.pdf Zirintunda, G., 2011, ‘Assessment of human faecal environmental contamination and
the prevalence of porcine cysticercosis in Soroti District, Uganda’, dissertation
submitted in part fulfilment for the degree of Master of Science in International
Animal Health, Royal (Dick) School of Veterinary Studies, Edinburgh. Jayashi, C.M., Arroyo, G., Lightowlers, M.W., Garcia, H.H. & Rodriguez, S., 2012,
‘Seroprevalence and risk factors for Taenia solium cysticercosis in rural pigs of
Northern Peru’, PLoS Neglected Tropical Diseases 6(7), e1733. http://www.ojvr.org
doi:10.4102/ojvr.v82i1.888 doi:10.4102/ojvr.v82i1.888 http://www.ojvr.org
|
https://openalex.org/W2966755025
|
https://ieeexplore.ieee.org/ielx7/8254253/8787225/08787228.pdf
|
English
| null |
Statistical learning for semantic parsing: A survey
|
Big data mining and analytics
| 2,019
|
cc-by
| 17,425
|
BIG DATA MINING AND ANALYTICS
ISSN
2096-0654
01/08
pp217–239
Volume 2, Number 4, December 2019
DOI: 10.26599/BDMA.2019.9020011 BIG DATA MINING AND ANALYTICS
ISSN
2096-0654
01/08
pp217–239
Volume 2, Number 4, December 2019
DOI: 10.26599/BDMA.2019.9020011 Qile Zhu, Xiyao Ma, and Xiaolin Li Qile Zhu, Xiyao Ma, and Xiaolin Li Abstract: A long-term goal of Artificial Intelligence (AI) is to provide machines with the capability of understanding
natural language. Understanding natural language may be referred as the system must produce a correct response
to the received input order. This response can be a robot move, an answer to a question, etc. One way to
achieve this goal is semantic parsing. It parses utterances into semantic representations called logical form, a
representation of many important linguistic phenomena that can be understood by machines. Semantic parsing is a
fundamental problem in natural language understanding area. In recent years, researchers have made tremendous
progress in this field. In this paper, we review recent algorithms for semantic parsing including both conventional
machine learning approaches and deep learning approaches. We first give an overview of a semantic parsing
system, then we summary a general way to do semantic parsing in statistical learning. With the rise of deep
learning, we will pay more attention on the deep learning based semantic parsing, especially for the application of
Knowledge Base Question Answering (KBQA). At last, we survey several benchmarks for KBQA. Key words: deep learning; semantic parsing; Knowledge Base Question Answering (KBQA) representing semantic meaning is called logical forms. For example, the first sentence in Table 1 would map
onto the logical form max.primes\. 1; 10//. We can
treat the logical form as an executable program to get
the desired behavior (e.g., answers to questions, queries
from a database, or an invocation of an API). @ The author(s) 2019. The articles published in this open access journal are distributed under the terms of the
Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/). Qile Zhu, Xiyao Ma, and Xiaolin Li are with National
Science Foundation Center for Big Learning, University of
Florida, Gainesville, FL 32608, USA. E-mail: valder@ufl.edu;
maxiy@ufl.edu; andyli@ece.ufl.edu. @ The author(s) 2019 The articles published in this
Qile Zhu, Xiyao Ma, and Xiaolin Li are with National
Science Foundation Center for Big Learning, University of
Florida, Gainesville, FL 32608, USA. E-mail: valder@ufl.edu;
maxiy@ufl.edu; andyli@ece.ufl.edu.
To whom correspondence should be addressed.
Manuscript received: 2018-09-17; accepted: 2019-04-29 1
Introduction Besides, semantic parsers have been
applied to a large number of applications out of KBQA,
such as navigation systems[9,10], identifying objects in
scene[11,12], and converting natural language to regular
expressions[13]. Deep
learning
demonstrates
state-of-the-art
performance
in
many
areas[14]
including
speech
recognition[15],
image
classification[16],
image
segmentation[17],
Part-Of-Speech (POS) tagging[18],
Named Entity Recognition (NER)[18–20], and Malware
Detection[21]. For example, fully connected neural
network is used[22] to effectively identify entities
in a newswire corpus. The application of character
and word embeddings in Bi-directional Long Short-
Term Memory (LSTM)[18,19] achieved state-of-the-art
performance in several sequence-to-sequence datasets,
such as CoNLL03[23] for NER and Penn Treebank
WSJ[24] for POS tagging. In the field of semantic
parsing, it is also boosted by the strength of deep
learning[25]. Inspired by the rise of statistical learning techniques
in the speech recognition community, more and more
researchers tended to apply statistical algorithms to
NLP problems. It offered a new paradigm: collect
examples of the desired input-output pairs and fit
them into a statistical model. This new paradigm
provides a new way to learn a more general NLP
system. Tasks like documentation classification, part-
of-speech tagging, and syntactic parsing had made
significant improvements. Fields which need more
semantic understanding like question answering also
achieved much progress. We select several representative approaches for
semantic parsing and discuss them in detail covering
from conventional machine learning to deep learning
including
both
supervised
and
weak
supervised
learning. We also provide a taxonomy for these methods
shown in Fig. 1. Statistical machine learning approach requires a
labeled dataset of natural language utterances paired
with annotated logical forms, as shown in Table 1. Over
the last few years, many attempts have been applied
to reduce the amount of supervision from annotated
logical forms[5,6]. This paper is structured as follows: we will present
a general framework for statistical semantic parsing in
Section 2[2]. This framework is gratifying modular and
can be easy to be extended for conventional statistical
learning algorithms with hand-crafted features (Section
3). Section 4 introduces novel deep learning approaches
for semantic parsing. We further discuss the datasets in The first is transforming the supervision from
annotated logical forms to answers. Take the first
question in Table 1 as an example, we use the answer
7 as the ground truth label instead of logical form
max.primes \ . 1; 10//. 1
Introduction Enabling machines to understand natural language
with big challenge and huge promising future, has
attracted so many attention in the last few decades. Meanwhile, semantic parsing, as a subfield of Natural
Language Processing (NLP), aims to map utterances
into a precise representation of its semantic meaning
with a formal language, which is executable and
understandable by machines. Semantic parsing has been
widely adopted for language reasoning and question
answering with knowledge base[1]. Before we introduce
the formal language for semantic parsing, let us give
some examples of some pairs of utterance-action pairs
(Table 1)[2]. We are interested in utterances listed
in Table 1, which may require understanding and
reasoning for natural language. The formal language for Semantic parsing, similar to machine translation,
involves in translating from one semantic representation
into another one. Both of them are dealing with complex
structures, often related in complex ways. However,
the target of semantic parsing is machine-readable
while machine translation is human-readable. Logical
forms play a foundational role in natural language
understanding systems and building an interface for
human and machine. Luna[3],
a natural language Table 1
Examples for utterance-action pairs. Utterance
Action
What is the largest
prime less than 10? 7
What is the highest
mountain in the world? Mount Everest
Call the number
1234567
Open the phone APP
and make a call Table 1
Examples for utterance-action pairs. What is the largest
prime less than 10? What is the highest
mountain in the world? Call the number
1234567 @ The author(s) 2019. The articles published in this open access journal are distributed under the terms of the
Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/). Big Data Mining and Analytics, December 2019, 2(4): 217–239 218 interface to a database about moon rocks,
and
SHRDLU, a toy blocks world environment, could both
answer questions and perform actions[4]. These systems
achieved significant achievements in the early time. They could handle complex linguistic phenomena,
integrate syntax-semantics, and reason in an end-to-end
fashion. Take an example from SHRDLU, although it
was able to process “find a book which is taller than
the one you are holding and put it into the box”[2], it
becomes extremely difficult to generalize beyond the
specific domain and handle general language, since
these systems are based on hand-crafted rules. structured tables[1]. 1
Introduction Compared with the logical
form, this form of data is much easier to obtain
view crowd-sourcing, which may result in a more
difficult learning problem. But Liang et al.[6] showed a
possible way to solve it with modern machine learning
algorithms. Learning semantic parsers from answers is
called weak supervision. Fig. 1
Semantic parsing taxonomies by supervisory signal
and techniques (names are corresponding to Section 4.1). The second is to apply semantic parsers to more
complex domains, e.g., answering questions from
broad-coverage knowledge bases such as Freebase[7]
and Wikidata[8]. Systems leverage knowledge bases to
answer questions are called KBQA system. With the
weak supervision technique, it is easy to collect datasets
via crowdsourcing, and semantic parsers have even
been extended beyond fixed knowledge bases to semi- Fig. 1
Semantic parsing taxonomies by supervisory signal
and techniques (names are corresponding to Section 4.1). 219 Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey .>/
AnBŒf
BŒx
) AŒf .x/;
.</
BŒx
AnBŒf
) AŒf .x/: Section 5 and analyze the challenge and tendency in
Section 6. Section 7 concludes the paper. Section 5 and analyze the challenge and tendency in
Section 6. Section 7 concludes the paper. CCG can be extended to use other combinations
which both handle more complex linguistic phenomena
and ungrammatical languages[31]. These issues can also
be solved by having a more expansive lexicon[32]. 2
Framework In this section, we introduce the components of a
statistical semantic parsing system. The key of a
semantic parser is positing an intermediate logical form
d that links utterance x and action y. In particular, d
captures the semantics of the utterance x and it also
executes to the action y based on the context of c (from
a knowledge base or other information source). An alternative approach which is less complicated
than CCG is to adopt a cruder grammar whose rules
correspond directly to the lambda Dependency-based
Composition Semantics (lambda DCS)[33], join and
intersect. Here is an example to parse the sentence
“prime less than 10” in lambda DCS: Our
semantic
parsing
system
consists
of
five
components[2]: prime
less than
10
N Œprime
N jN Œless
N Œ10
N Œless:10join Executor. Compute the action y given a logical
form d and context c. Grammar. Rules G that used to produce the
candidates derived from logical forms. N Œprime u less:10intersect: N Œprime u less:10intersect:
The join and intersect grammar are
.Join/
NjN Œr N Œz ) N Œr:z;
.Intersect/
N Œz1 N Œz2 ) N Œz1 u z2: Model. A statistical model to model the
distribution p.djx; c/ parametrized by . Parser. Inferences from the model and get high
probability results. However,
these crude rules may result in the
derivation of other logical forms such as 10uless:prime. These incorrect logical forms can be figured out by the
model. Learn. Tune the parameters from the given
example pairs. We
now
use
an
example
to
instantiate
each
component: “What is the largest prime less than 10?” The choice of grammar is the most important
part of a semantic parser because it directly controls
the tradeoff between the precision of derived logical
forms and statistical model’s complexity. CCG is
natural to capture the complicated linguistic features
from well-structured sentences. For applications with
noisy utterance but simple logical forms, a more
flexible grammar like lambda DCS is preferred. To
answer the question above, we can use CCG to get
the logical form x:performBy.x; natalie potman/ ^
characterIn.x; star wars/. Executor. The executor will give us the answer 7. Grammar. The
grammar
parses
utterances
to
possible derivations of logical forms. There are
plenty
of
grammar
formalisms
to
represent
the
semantic meaning[6,26–28]. One is lambda calculus with
Combinatory Categorial Grammar (CCG)[29]. 2
Framework Due to the concavity of the objection
function O./, SGD is best guaranteed to converge to a
local maxim instead of a global one. p.djx; c/ D
exp.score.x; c; d//
P
d 0Dx;c exp.score.x; c; d 0//: p.djx; c/ D
exp.score.x; c; d//
P
d 0Dx;c exp.score.x; c; d 0//: Parser. With a trained model p, a semantic parser
computes the logical form with the highest probability
for an utterance x. Normally, an utterance consists
of a sequence of tokens (words). A standard approach
is to use a chart parser[36], which recursively builds
logical forms for each span of the utterance. The parser
takes each category and span Œi:j .where 0 ⩽i < j ⩽
length.x//, then loops over all the rules in the grammar
to apply each one to build new derivations of each
category over Œi:j. The final logical forms for the
utterance x are collected in the root category over span
Œ0 W length.x/. Until now, we have covered the components of a
semantic parser. In next sections, we will focus on the
learning problem, especially on the model and learner
part of the whole system. 2
Framework Below is
an example of a CCG derivation[2]: prime
less than
10
N Œx:prime.x/
.N nN /=NPŒf:x:f .x/ ^ less.x; y/
NPŒ10
.>/
N nN Œf:x:f .x/ ^ less.x; 10/
.</
:
N Œx:prime.x/ ^ less.x; 10/ prime
less than
10
N Œx:prime.x/
.N nN /=NPŒf:x:f .x/ ^ less.x; y/
NPŒ10
.>/
N nN Œf:x:f .x/ ^ less.x; 10/
.</
:
N Œx:prime.x/ ^ less.x; 10/ When answering questions with a knowledge base,
another issue is how to map the entities in the utterances
to the entities in the knowledge base. This problem is
referred as entity linking problem[34]. Although entity
linking is not strictly as the same as grammars, it is also
a mapping rule. Table 2 shows an example of entity
linking result from the question “what character did
Natalie Portman play in Star Wars?” A novel entity A CCG can be defined by lexicons and a set of
combinators[30]. The lexicon links words and phrases
with their categorial meaning (usually composed of
pairs (words, logical constant)). In CCG trees, each
node is a category. Nones (N) denote sets and None
Phrases (NP) denote objects, here N and NP are
categories in CCG instead of POS tags. It also has
composite categories (.NnN/=NP), which represent
functions of multiple arguments. Rules in CCG are
mostly lexical (e.g., [fun ) NŒ.x/:fun.x/]), which
map particular words in utterances. Also, we have a
forward application and backward application rule: Table 2
Examples of entity linking. Entity in utterance
Entity in knowledge base
NataliePortman
Natalie Portman
Star Wars
Star Wars Episode I: The Phantom Menace Table 2
Examples of entity linking. Entity in utterance
Entity in knowledge base
NataliePortman
Natalie Portman
Star Wars
Star Wars Episode I: The Phantom Menace Big Data Mining and Analytics, December 2019, 2(4): 217–239 220 linking system[35] used a tree-based structured learning
framework based on multiple additive regression trees
to model entity linking as a structured learning problem. which scans the utterance from left to right without
backing up, so new sub-derivations can depend on the
staff before. Model. Due to the lexical ambiguity and different
parsing decisions, there are many possible logical forms
derived from an input utterance. The model produces
the scores of candidate logical forms generated by
the grammar. Before deep learning, log-linear model
(generalizations of logistic regressions) is the common
choice for virtually all semantic parsers. 2
Framework The log-linear
model defines a feature vector .x; c; d/ RF for
each possible logical form d. We can regard each
feature as voting for different derivations d based on
some property of the utterance and the derivation. Model. Due to the lexical ambiguity and different
parsing decisions, there are many possible logical forms
derived from an input utterance. The model produces
the scores of candidate logical forms generated by
the grammar. Before deep learning, log-linear model
(generalizations of logistic regressions) is the common
choice for virtually all semantic parsers. The log-linear
model defines a feature vector .x; c; d/ RF for
each possible logical form d. We can regard each
feature as voting for different derivations d based on
some property of the utterance and the derivation. Learner. Learner also refers to the optimization
problem. The dominant paradigm in machine learning
is to set up an objective function and optimize it; while
another way is to use reinforcement learning[38]. A
standard way is to maximize the likelihood of training
data .xi; ci; yi/n
iD1. One important thing is that we only
have the labeled action yi rather than the correct logical
form d. Under this assumption, we must consider all
logical forms d whose action based on context ci is yi. The corresponding log-likelihood is Oi./ D log
X
dD^jdjci Dyi
p.djxi; ci/: Let RF represent the parameter vector, in which
every dimension defines a weight for each feature in the
feature vector .x; c; d/. The score of every generated
logical form is .x; c; d/ . This score gives an
measure how good the logical form is. Normally, we use
softmax function to obtain the distribution probabilities
over all logical forms: There are numerous approaches to maximize O./. Among them, Stochastic Gradient Descent (SGD) is
the simplest way, which is an iterative algorithm that
updates the parameters over the training data. C 5 Oi./; where is the step size, known as learning rate that
decides how aggressively we want to update parameters
with the gradients. Due to the concavity of the objection
function O./, SGD is best guaranteed to converge to a
local maxim instead of a global one. where is the step size, known as learning rate that
decides how aggressively we want to update parameters
with the gradients. 3.1
Supervised semantic parsing with CCG CCG has two key components,
the lexicon and
combinators. Combinators are a small set of operations
that are predefined, while lexicon pair words and
phrases with CCG categories. So we have the structure
learning problem, where we need to learn the structure
of the CCG grammar. Also, the parsing process
is often ambiguous, so we need another model to
choose the best parse giving these ambiguities. To
solve these two problems jointly, we will introduce a
simple algorithm named structured perceptron and then
an online learning algorithm derived from it. Many
supervised semantic parsing algorithms share the same
idea so that we will introduce a basic algorithm. All the derivations generated by Chart Parser could
be exponentially large. However, we only care about
the derivations with high probability under our model
p. Therefore, beam search is often applied, where we
only keep top-k sub-derivations with top-k scores given
by our model at each step. Although beam search is not
guaranteed to return the K highest scores, it is still an
effective heuristic. In addition, chart parser suffers from incremental
contextual interpretation: the features of a span may
only depend on the sub-derivations of that span instead
of on the parts constructed before. This problem makes
anaphora (resolution of pronoun) challenging to model. A possible solution is to use shift-reduce parsing[37], 3.1.1
Structured perceptron 3.1.1
Structured perceptron When applying SGD to optimize
the log-likelihood with Viterbi approximation, we have dL
dw D ˚.xi; h0; yi/ ˚.xi; h; y/;
h0 D argmaxhw f .xi; h; yi/;
y; h D argmaxy;hw f .xi; h; yi/: dL
dw D ˚.xi; h0; yi/ ˚.xi; h; y/; Similar
to
the
multiclass
setting,
structured
perceptron[39]
is
an
extension
of
the
perceptron
algorithm. The difference is that structured perceptron
deals with the labels which represent a set of structures. These structures can be generated from a given
structured input x[42]. The prediction function of the
structured perceptron is h0 D argmaxhw f .xi; h; yi/; Unfortunately, the hidden variable perceptron is not
guaranteed to converge due to the log-linear version
is non-convex. It is still a good choice because this
algorithm is very simple and easy to implement, and
works well with careful initialization. y D argmaxy0Y.x/˚.x; y0/ w; Algorithm 2
Learning for structured perceptron
Input:
Training set D D f.x1; y1/; .x2; y2/; : : : ; .xN ; yN /g, map
function ˚. Output:
Weight vector w. 1: w .0; 0; : : : ; 0/
2: repeat
3:
for each .x; y/ D do
4:
z argmaxy0Y.x/˚.x; y0/ w
5:
if z ¤ y then
6:
w w C ˚.x; y/ ˚.x; z/
7:
end if
8:
end for
9: until Converged Algorithm 2
Learning for structured perceptron
Input:
Training set D D f.x1; y1/; .x2; y2/; : : : ; .xN ; yN /g, map
function ˚. Output:
Weight vector w. 1: w .0; 0; : : : ; 0/
2: repeat
3:
for each .x; y/ D do
4:
z argmaxy0Y.x/˚.x; y0/ w
5:
if z ¤ y then
6:
w w C ˚.x; y/ ˚.x; z/
7:
end if
8:
end for
9: until Converged Algorithm 1
Learning for generic perceptron
Input:
Training set D
D
f.x1; y1/; .x2; y2/; : : : ; .xN ; yN /g,
number of iterations I. Output:
Weight vector w. 3.1.1
Structured perceptron 1: w .0; 0; : : : ; 0/
2: for i D 1; 2; : : : ; I do
3:
for j D 1; 2; : : : ; N do
4:
if yj xj w ⩽0 then
5:
w w C yj xj
6:
end if
7:
end for
8: end for
Algorithm 2
Learning for structured perceptron
Input:
Training set D D f.x1; y1/; .x2; y2/; : : : ; .xN ; yN /g, map
function ˚. Output:
Weight vector w. 1: w .0; 0; : : : ; 0/
2: repeat
3:
for each .x; y/ D do
4:
z argmaxy0Y.x/˚.x; y0/ w
5:
if z ¤ y then
6:
w w C ˚.x; y/ ˚.x; z/
7:
end if
8:
end for
9: until Converged 3.1.1
Structured perceptron A straightforward algorithm of machine learning, 221 Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey where ˚.x; y0/ maps the pair .x; y0/ to a vector
(features). Algorithm 2 shows the learning algorithm
for structured perceptron. structured perceptron was proposed in Ref. [39]. In this
part, we first introduce the generic perceptron[40]. After
that, we show how structured perceptron works. Generic
perceptron[40]
uses
a
linear
prediction
function to do binary classification. More formally, the
prediction function is Structured perceptron can be connected with a log-
linear model naturally. Recall the log-linear model, we
define a condition probability on x and y: p.xjy/ D
exp.w ˚.x; y//
P
y0 exp.w ˚.x; y0//: f .x/ D
(
1;
if wx C b > 0I
0;
otherwise; p.xjy/ D
e p.w
.x; y//
P
y0 exp.w ˚.x; y0//: To maximize the log-likelihood function, we can use
gradient descent to solve this optimization problem as
follows: where x is the input vector, w is the weight vector, and
b is the bias. As Algorithm 1 shown, perceptron tunes the weight
vector in an online fashion. For each example,
perceptron checks whether it is classified correctly. If not, it will update the weight vector by moving
w towards the current input vector. The perceptron
algorithm is guaranteed to accurately classify each
example in a linear separable training set[40]. L D
n
X
iD1
logp.yijxi/;
dL
dw D
n
X
iD1
˚.xi; yi/ Ep.yjx/˚.xi; y/: When using SGD and Viterbi approximation, we get dL
dw D ˚.xi; yi/ ˚.xi; y/;
y D argmaxyw f .xi; y/: Reference [41] proposed averaged perceptron that
assigns more weights for the examples at the beginning
of the training. With this weight averaging, perceptron
achieves some kind of large margin effect. Binary
perceptron can be easily extended to multiclass
classification. In multiclass setting, for each input x,
the corresponding label y belongs to a finite set Y and
simply chooses the maximum score. To make the model more powerful, we can add
latent variables into the log-linear model (p.yjx/ D
P
h p.y; hjx/)[31,43]. Input: In this subsection, we investigate a unified learning
algorithm[28] to induce semantic parsing which is
related to loss-sensitive structured perceptron[44]. This
algorithm can be applied to both supervised learning
and weak supervision. It jointly estimates the parsing
parameters and induces the lexicon structure. Training sample .xi; Li/, lexicon 0, and parameters . Output: Lexicon 0. 1: Set D 0 [ GENLEX.xi; Li; 0; / 2: Parse xi with model and lexicon and get k highest
scoring parses Y 2: Parse xi with model and lexicon and get k highest
scoring parses Y 3: t D 0 [ S
yY LEX.Y / 3: t D 0 [ S
yY LEX.Y / In the unified learning algorithm, there are two
learning choices to be made. The first is V W Y ) .t; f /; expected in those parse results with the highest scores. So we extract the top-k parses and combine them with
the original set. where
the
validation
function
V
indicates
the
correctness (true of false) of a parse y. The varying
function V
allows us to use different forms of
supervision. The second choice need to specify the
lexical generation procedure (GENLEX[28]), which
takes input .sentence: x; validation function: V or
labeled logical form z; lexicon: ; parameters: /
and produces a overly general set of lexical entries. The
generated set is noisy and needs to be pruned later with
correct lexical entries. The overall algorithm is shown
as Algorithm 3. where
the
validation
function
V
indicates
the
correctness (true of false) of a parse y. The varying
function V
allows us to use different forms of
supervision. The second choice need to specify the
lexical generation procedure (GENLEX[28]), which
takes input .sentence: x; validation function: V or
labeled logical form z; lexicon: ; parameters: /
and produces a overly general set of lexical entries. The
generated set is noisy and needs to be pruned later with
correct lexical entries. The overall algorithm is shown
as Algorithm 3. We can apply structured perceptron as discussed in
the last subsection to update the parameters based on
the parsing results and lexicon. Cycling between Steps 1 and 2 will keep only
the lexical entries that occur in the highest scoring
parses, which leads to producing a compact lexicon. Big Data Mining and Analytics, December 2019, 2(4): 217–239 222 Input: In
summary, the unified learning algorithm forms a greedy,
iterative method for simultaneously finding a compact
lexicon and optimizing the log-likelihood of the training
data[28]. With the online learning schema, we have two steps
for each training sample. One is lexical generation and
the other is update parameters. 3.2
Weak supervised semantic parsing In the supervised learning fashion,
we have the
labeled logical forms,
e.g.,
“show me flights to
Boston (x:flight.x/^to .x; BOSTON/)” and “I need
a flight from Baltimore to Seattle (.x/:flight.x/^
from.x; BALTIMORE/^to.x; SEATTLE/)”. In this form
of learning, we learn directly from pairs of utterances
and logical forms[28]. However, it is hard to label
the logical forms when given thousands of utterances. Another problem is that one utterance may correspond
to several correct logical forms. A possible way to learn
flexibly is weak supervision[5,6], which only requires
executing logical forms within a system and evaluating
the result[45,46]. In this form, both parsing process and
logical forms are latent values. Also, it uses the executor
directly to evaluate the parser, which can be easily
applied to different domains, and it does not suffer from
the labeling problem from expertise. One of the advantages of this algorithm is that it can
provide a relatively compact lexicon, which is a subset
of possible lexical entries. This can be achieved by
alternating between two steps. Step 1 is aimed to search
for a relatively small number of lexical entries, which
are sufficient to parse all training samples successfully. Step 2 is used to update the parameters of the lexical
entries that are selected in Step 1. In Algorithm 4, each utterance in one training sample
is parsed with the current parameters with a new
lexicon 0[GENLEX.xi/. This results in a large set of
result logical forms. However, correct lexical entries are Algorithm 3
Unified learning algorithm
Input:
Training set .xi; Vi/ W i D 1; : : : ; n, number of iterations T. Output:
Parameters and lexicon . 1: Initialize using 0, 0
2: for i D 1; 2; : : : ; T do
3:
for j D 1; 2; : : : ; n do
4:
Step 1: Lexical generation
5:
Step 2: Update parameters
6:
end for
7: end for Algorithm 3
Unified learning algorithm
Input:
Training set .xi; Vi/ W i D 1; : : : ; n, number of iterations T. Output:
Parameters and lexicon . Algorithm 1
Learning for generic perceptron Output:
Weight vector w. 1: w .0; 0; : : : ; 0/
2: repeat
3:
for each .x; y/ D do
4:
z argmaxy0Y.x/˚.x; y0/ w
5:
if z ¤ y then
6:
w w C ˚.x; y/ ˚.x; z/
7:
end if
8:
end for
9: until Converged until Converged Big Data Mining and Analytics, December 2019, 2(4): 217–239 Big Data Mining and Analytics, December 2019, 2(4): 217–239 3.2
Weak supervised semantic parsing 1: Initialize using 0, 0
2: for i D 1; 2; : : : ; T do
3:
for j D 1; 2; : : : ; n do
4:
Step 1: Lexical generation
5:
Step 2: Update parameters
6:
end for
7: end for Algorithm 3
Unified learning algorithm
Input:
Training set .xi; Vi/ W i D 1; : : : ; n, number of iterations T. Output:
Parameters and lexicon . 1: Initialize using 0, 0
2: for i D 1; 2; : : : ; T do
3:
for j D 1; 2; : : : ; n do
4:
Step 1: Lexical generation
5:
Step 2: Update parameters
6:
end for
7: end for In this section, we will focus on KBQA[47,48], one
of the most exciting problems in weak supervising
semantic parsing. A knowledge base K is a set of
assertions .e1; p; e2/, where e denotes an entity (e.g.,
BarackObama) and p is a property (e.g., BirthDate) in
K. There are two challenges for KBQA: (1) knowledge
base has a very large set of logical predicates, which is Parameters and lexicon . 1: Initialize using 0, 0 Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey 223 f.BarackObama;
Honolulu/; : : : g). The lexicon is
generated based on the overlap Fr1 [Fr2, for r1 R1
and r2 R2. With the alignment, Berant13 computed
three types of features. Alignment features: (1) log
of number of entity pairs that occur with the phrase
r1 .jF.r1/j/, and (2) log of number of entity pairs
that occur with the phrase r2 .jF.r2/j/, and (3) log of
number of entity pairs that occur with both r1 and r2
.jF.r1/ [ F.r2/j/. Lexicalized features: conjunction
of phrase and predicate. Text similarity features: phrase
r1 is equal/prefix/suffix of r2 or r1 and r2 overlap. impossible to store them in the parser, and (2) without
labeled logical forms, we cannot generate sufficient
lexical entries as in Section 3. An example of KBQA
is shown in Fig. 2. Most work of KBQA is based on
Freebase[7]. 3.2.1
Berant13 Berant13[49] used a large amount of web text and a
knowledge base to build a coarse alignment between
phrases and predicates. However, this alignment cannot
cover all the predicates. Light verbs (e.g., “go”) and
prepositions are hard due to polysemy. Rare predicates
(e.g., “cover price”) are difficult even given a large
corpus. To improve coverage, Berant13[49] proposed a
new bridging operation that generates predicates based
on adjacent predicates rather than on words. Based on
these features, Berant13 used a log-linear model to
score the logical forms. The bridging operation generates a binary predicate
based on neighboring logical predicates instead of
explicit lexical material. Considering the question
“what is the cover price of X-men?”, the correct
prediction
ComicBookCoverPrice
is
expressed
explicitly but is not in the lexicon generated by
the alignment. In this situation, given a unary z with
type t (X-men), they constructed a logical form b:z
for any predicate b with type (*, t). Based on this
operation, Berant13 computed the bridging feature:
log of the number of pairs that occur with bridging
predicate b. The goal of alignment was to construct a lexicon
L,
a mapping from natural language phrases to
logical predicates accompanied by a set of features. Intuitively, Berant13 aligned a phrase and a predicate
based on whether they co-occur with many of the
same entities. In summary, Berant13 first constructed
a set of typed (Freebase associates each entity
with
a
set
of
types
using
the
type
property)
phrases R1 (e.g.,
“born in” [Person,
Location])
and predicates R2 (e.g., PlaceOfBirth). For each
element r R1 [ R2, they created a mapping F.r/
from a phrase in R1 to related predicates R2 and
vice versa (e.g., F .“bornin” ŒPerson; Location/
D Together with the features from alignment, bridging,
and other syntax based features, Berant13 used a log-
linear model to rank the generated logical forms. 4.1
Overview of deep learning Deep learning allows computational models to learn
representations
of
data
with
multiple
levels
of
abstraction through multiple processing layers[14]. These deep learning based methods have dramatically
improved the state-of-the-art performance in speech
recognition, visual object detection and recognition,
language models, and many other fields such as drug
discovery and genomics. In this subsection, we will
first introduce the feedforward neural network, followed
the architecture of Convolution Neural Network (CNN)
which has brought breakthroughs in processing images. Then we will discuss Recurrent Neural Network (RNN)
and some improvement upon it which has shone a light
on sequential data such as text and speech. Before
generating
logical
forms
when
given
utterances, Berant14 defined a set of templates to
cover:
(1) p:e, (2) p1:pe:2, (3) p:.p1:e1[p2:e2/,
(4) support “unary filter”, and (5) handle aggregation
formulas. From an utterance x, Berant14 found an
entity in x and grow the logical form from it, details
can be referred in the original paper[50]. With the same
strategy, Berant14 also used some rules and templates
to generate utterances from the logical forms (e.g.,
“What NP is VP by d.e/?”, where d.e/ is the Freebase
descriptions for the entity). 4.1.2
CNN CNN[52] is a specialized kind of neural network for
processing data with a known grid-like topology (e.g.,
time-series data, which can be treated as a 1D grid and
image data, which is a 2D grid of pixels). As the name
implies, CNN employs a mathematical operation named
convolution. A convolution layer uses a set of kernels
to attain local features over the whole image sharing the
same parameters, which means that one channel of the
output is a result calculated by one kernel interacts with
the whole image. This results in that CNN has fewer
parameters than MLP. In
vector
space
model,
Berant14
estimated
a
paraphrase score via a weighted combination: vT
xWvc D
k
X
i;jD1
wij vx;ivc;j ; where
W Rkk
is
a
parameter
matrix. The
association
model
aligned
particular
phrases
to
one another, while the vector space model provided a
soft representation for utterances. A typical layer of CNN consists of three stages: a
convolution layer, an activation layer, and a pooling
layer. The activation layer performs the same role as
in MLPs to provide a nonlinear transformation. The
pooling layer applies a pooling function which replaces
the output of the net at a certain location with a 3.2.2
Berant14 One problem of conventional semantic parsers is that
they only use data which pairs natural language with
the KB. This leaves untapped a vast amount of text not
related to the KB. Berant14[50] presented an approach
for semantic parsing based on paraphrasing that was
able to exploit large amounts of text not covered by the
KB (Fig. 3). Fig. 2
An example to answer a question through a
knowledge base[49]. To narrow down the space of logical
predicates, they use (1) course alignment based on Freebase
and a text corpus and (2) a bridging operation that generates
predicates compatible with neighboring predicates. This approach consists of three steps:
(1) use a Fig. 3
An example of paraphrasing for semantic parsing[50]. For each candidate logical form (red), it generated canonical
utterances (purple). The model is trained to paraphrase
the input utterance (green) into the canonical utterances
associated with the correct denotation (blue). Fig. 2
An example to answer a question through a
knowledge base[49]. To narrow down the space of logical
predicates, they use (1) course alignment based on Freebase
and a text corpus and (2) a bridging operation that generates
predicates compatible with neighboring predicates. Fig. 3
An example of paraphrasing for semantic parsing[50]. For each candidate logical form (red), it generated canonical
utterances (purple). The model is trained to paraphrase
the input utterance (green) into the canonical utterances
associated with the correct denotation (blue). Big Data Mining and Analytics, December 2019, 2(4): 217–239 224 simple procedure to construct a manageable set of
candidate logical forms based on the input utterance,
(2) leverage the generated logical forms to generate
canonical utterances based on the text descriptions of
entities and predicates from the KB, and (3) choose
the generated utterance that best paraphrases and input
together with the corresponding logical form. semantic parsing algorithms for both supervised and
weak supervised learning. 4.1.1
Feedforward neural network When both logical forms paired with utterances
were constructed, the problem was reduced to scoring
pairs (c; z) based on a paraphrase model. Berant14
combined traditional association model (determine
whether x and c contain phrases that are likely to
be paraphrases) and a vector space model (based on
word2vec embedding[51]): Feedforward neural network, which is also known as
MultiLayer Perceptron (MLP), is aimed to approximate
some function f , e.g., for a classifier, y D f .x/
maps an input x to a category y. To enhance
the expressive ability of MLP, we can add non-
linear activation function (e.g., sigmoid and ReLU)
after each hidden unit. To train the MLP, we can
apply backpropagation with chain-rule. In summary,
feedforward neural networks are the quintessential deep
learning methods. pr.x; c/Tpr D as.x; c/Tas C vs.x; c/Tvs: The association model links phrases of x and c in
multiple overlapping ways with features: (1) lemma
(xiWj )^lemma(ci0Wj 0), where xiWj denotes spans from x,
(2) pos(xiWj )^pos(ci0Wj 0), (3) lemma(xiWj / D lemma
(ci0Wj 0)?, (4) pos(xiWj / D pos(ci0Wj 0)?, (5) lemma(xiWj )
and lemma(ci0Wj 0) are synonyms?, and (6) lemma(xiWj )
and lemma(ci0Wj 0) are derivations? (“?” will return true
or false of the association model). 4.1.3
RNN From the fomulas, we can get that GRU uses a single
gating unit simultaneously to control the forgetting
factor and the decision to update the hidden state. This
results in that fewer parameters needed in a GRU cell. LSTM cell at time step t are LSTM cell at time step t are summary statistic of the nearby outputs[53]. One popular
pooling function is max pooling[54], which reports the
maximum output within the rectangular neighborhood. Pooling
layer
helps
to
make
the
representation
approximately invariant to small translations of the
input. Another purpose of pooling is to reduce the
dimension of the output and save memory. p
it D .Wiht 1 C Uixt C bi/;
f t D .Wf ht 1 C Uf xt C bf /;
Oct D tanh.Wcht 1 C Ucxt C bc/;
ct D f t ˇ ct 1 C it ˇ Oct;
ot D .Woht 1 C Uoxt C bo/;
ht D ot ˇ tanh.ct/; ot D .Woht 1 C Uoxt C bo/; 4.2
Deep learning for semantic parsing With a basic understanding of different neural nets,
we will introduce how the deep learning combined
with semantic parsing framework. We will give several
elegant algorithms to combine deep learning and
semantic parsing: (1) transform the logical form into
a query graph, (2) use a sequence-to-sequence model
with data recombination to learn logical form directly,
(3) use a sequence-to-sequence model and utilize a
key-variable memory to handle compositionality with
REINFORCE, and (4) a neural semantic parser firstly
maps the utterances to an intermediate representation
and then induces intermediate representations in the
form of predicate-argument structures from data. RNNs
suffer
from
the
gradient
vanishing
or
exploding problem, so it is hard to learn long-term
dependencies. When gradients back propagated over
many stages, they tend to either vanish (most of the
time[56]) or explode (rarely[57]). An improvement of the
basic RNNs is called gated RNNs including LSTM[58]
and Gated Recurrent Unit (GRU[59]). The general idea
of gated RNNs is to create paths through time that
derivatives neither vanish nor explode. 4.1.3
RNN Unlike CNN which is good at processing spatial data
like images or speech, RNNs are a family of neural
networks for processing sequential data. An RNN deals
with a sequence of input x1; x2; : : : ; xt (t is the total
number of time steps). The structure of an RNN is based
on MLPs with the idea that sharing parameters across
different parts of a model and inputs, which makes it
possible to generalize across them[53]. RNNs usually
produce an output at each time step with connections
between last hidden units and current input; and an
RNN cell consists of these parts: where is the element-wise sigmoid function and ˇ is
the element-wise product. xt is the input vector in time
step t, and ht is the hidden state vector. Ui; Uf ; Uc; and
Uo are the weight matrices of different gates for input,
and Wi; Wf ; Wc; and Wo are the weight matrices for
hidden state. bi; bf ; bc; and bo denote the bias vectors. Another popular gated RNN is GRU[59], which
combines the forget and input gates into one named
“update” gate. Besides, it also merges the cell state and
hidden state. The update at time step t is zt D .WzŒht 1; xt/;
rt D .WrŒht 1; xt/;
Oht D tanh.WhŒrt ˇ ht 1; xt/;
ht D .1 zt/ ˇ ht 1 C zt ˇ Oht: p
h0 D b C W ht 1 C Uxt;
ht D tanh.h0/;
ot D c C V ht; ht D .1 zt/ ˇ ht 1 C zt ˇ Oht: where b and c are the bias vectors along with the
weight matrices U , V , and W , for input-to-hidden,
hidden-to-output, and hidden-to-hidden connections,
respectively. RNNs capture information only from
the past and the present input instead of the whole
sequence. In many applications, there is a need to
output a prediction that depends on the whole input
sequence. Bidirectional RNNs were invented to address
the need[55]. Bidirectional RNNs use two separate
RNNs to deal with the sequence in two directions
(forward and backward) and combine the two outputs
together as the final representation. From the fomulas, we can get that GRU uses a single
gating unit simultaneously to control the forgetting
factor and the decision to update the hidden state. This
results in that fewer parameters needed in a GRU cell. 4
Deep Learning for Semantic Parsing In this section, we will first give an overview of deep
learning, and then focus on the deep learning based 225 Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey 4.2.1
STAGG[60] Inspired by Refs. [61, 62], STAGG[60] proposed a
semantic parsing that leverages the knowledge base
tightly forming the parse for an input question. First,
STAGG defined a query graph that is straightforwardly
mapped to a logical form in -calculus. In this way, they LSTM leverages a self-loop mechanism to produce
paths where the gradient can flow for long durations[58]. An LSTM cell is composed of three multiplicative gates
which control the proportions of information to flow to
the next time step. Formally, the formulas to update an Big Data Mining and Analytics, December 2019, 2(4): 217–239 226 reduced semantic parsing to query graph generation,
which can be formulated as a search problem with
staged states and actions. Each state is a candidate parse
in the graph representation and each action defines a
way to grow the graph. In particular, STAGG staged
into three main steps: (1) locate the topic entity in
the question, (2) find the main relationship between
the answer and the topic entity, and (3) expand the
query graph with additional constrains that describe
properties the answer needs or the relations between the
answer and other entities. a single node with the topic entity, a core inferential
chain, or a more complex query graph with constraints. Let A D [fAe; Ap; Ac; Aag be the set of actions. An action can grow a graph by adding some edges and
nodes. Ae picks an entity node; Ap determines the core
inferential chain; and Aa and Ac add constraints and
aggregation nodes, respectively. A valid action set can
be defined in Fig. 5. The order of actions can vary and
be viewed as different ways to prune the search space. STAGG defined a reward function using a log-linear
model and search is done using the best-first strategy
with a priority queue. The three stages are described as
follows: Before we describe the query graph,
we first
introduce a particular entity category named Compound
Value Type (CVT) in Freebase. A CVT node is not a
real-world entity but is used to collect multiple fields of
an event or a special relationship. (1) The topic entities are chosen from an entity
linking system[35], and to tolerate potential mistakes of
the entity linking systems, up to 10 top ranked entities
are considered as the topic entities. 4.2.1
STAGG[60] Their query graph consists of four types of nodes:
grounded
entity
(rounded
rectangle),
existential
variable (circle), lambda variable (shaded circle), and
aggregation function (diamond). Figure 4 shows one
possible query graph for the question “who first voiced
Meg on Family Guy?” using Freebase. Meg Griffin and
Family Guy are two entities represented by two rounded
rectangles. y in a circle is an entity that should exist an
entity describing some casting relations (e.g., character,
actor), where y should be a CVT entity in this case. The shaded circle x is the answer node, which is the
answer to the question. The diamond node argmin
constraints that the answer needs to be the earliest
actor. This query graph is equivalent as the logical form
x:9y:cast (FamilyGuy,y/^ actor.y; x/^ character .y;
MegGriffin/ without the aggregation function. (2) To identify core inferential chain,
STAGG
measured the semantic similarity using CNN between
the query and the candidate chain. They proposed
Siamese neural networks[63] for identifying the core
inferential chain. In this way, STAGG mapped the
question to a pattern by replacing the entity mention
with a generic symbol hei, such as “who first voiced
meg on hei” vs. cast-actor. (3) One simple way to derive the constraint set is
first issuing the core inferential chain as a query to
the KB to find the bindings of variables y0 and x,
and then enumerating all neighboring nodes of these
entities. This often results in an unnecessarily large
pool. STAGG employed simple rules to retain only the
nodes that have some possibility to be constraints. With this new query graph, STAGG generated the
graphs with the following properties: (1) there is one
root entity referred to the topic entity in the graph
and (2) there exists only one lambda variable x as
the answer node, with a directed path (named core
inference chain) from the root to it. To obtain all possible query graphs, STAGG used a
log-linear model to learn a reward function based on
each stage’s reward. 4.2.2
JIA16[64] STAGG[60]
reduced
the
semantic
parsing
into
a
searching problem, this method treated it as a sequence-
to-sequence problem (learn logical form from a
supervised fashion). Meanwhile, JIA16[64] introduced
data recombination, a framework for injecting prior
knowledge into a model. Given a question, STAGG formalized the query
generation process as a search problem with staged
states and actions. Let S D [f∅; Se; Sp; Scg be the
set of states, where each state can be an empty graph, In a sequence-to-sequence model (consists of two
RNNs, an encoder and a decoder), the input utterance Fig. 4
Query graph of question “who first voiced Meg on
Family Guy”[60]. Fig. 5
Legitimate actions to grow a query graph[60]. Fig. 4
Query graph of question “who first voiced Meg on
Family Guy”[60]. Fig. 5
Legitimate actions to grow a query graph[60]. Fig. 5
Legitimate actions to grow a query graph[60]. Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey 227 to generate the tokens at depth 1 of the subtree. When the predicted token is hni, sequence-to-tree
decoded the sequence by conditioning on the hidden
vector. They introduced a parent-feeding connection
where the hidden vector of parent nonterminal is
concatenated with the inputs and fed into RNN to
generate subtrees. Figure 6 shows the decoding tree
as an example of the logical form “AB.C/”, where
y1; : : : ; y6 are predicted tokens, and t1; : : : ; t6 are time
steps. The decoding procedure is as follows: (1) once
input has been encoded, the decoder first generates
y1; : : : ; y4 at depth 1 until token h=si; (2) use the
nonterminal token t3’s hidden vector together with the
input to predict the subtree. x is a sequence of words x1; : : : ; xm from the input
vocabulary; the output is the corresponding logical
form y, which is also a sequence of tokens y1; : : : ; yn
from the output vocabulary. This kind of technique
is popular in machine translation[65,66]. One problem
of the conventional sequence-to-sequence model is
that it has difficulty generalizing to the long tail of
entity names commonly found in semantic parsing
datasets. To overcome this problem, JIA16 used a
copying mechanism based on a Pointer Network[67] to
decide whether to write any word in the vocabulary or to
copy a word from any input word directly to the output. 4.2.2
JIA16[64] Due to the limitation of training data,
JIA16
used a data recombination framework to inject prior
knowledge and generated new examples to train
the sequence-to-sequence model inspired by the data
augmentation technique, which is commonly employed
in computer vision[68] and speech recognition[69]. The
key advantage of this approach is that it allows declaring
desired properties for the specific task. In semantic
parsing, consider an example “what states border
texas?”, it should be easy to generalize to questions
where “texas” is replaced by other state and simply
replace the mention of Texas in the logical form with
the new one. After training, sequence-to-tree predicted the logical
form for utterance q by Og D argmaxg0p.g0jq/; where g0 represents a candidate result. In practice, it
was impossible to iterate over all possible candidates
to get the optimal output. Sequence-to-tree decomposed
the probability p.gjq/ and used greedy search (or beam
search) to generate tokens one by one. For parent-
feeding, sequence-to-tree maintained a nonterminal
queue
to
generate
the
subtrees. Sequence-to-tree
also claimed that attention mechanism boosted the
performance in all benchmarks they used. In addition,
sequence-to-tree developed a simple procedure for
data argumentation. They replaced the identify entities
and
numbers
with
the
type
names
and
unique
IDs. Experiments showed the argumentation also
improved the performance. In general, JIA16 started with a training set D of
.x; y/ pairs, which defines the empirical distribution
Op.x; y/. Then they fit a generative model
Qp.x; y/
to
Op which generalizes beyond the support of
Op. Finally, to train the model p.yjx/, they maximized
the expected value of logp.yjx/. JIA16 applied
a Synchronous Context-Free Grammar (SCFG) to
generate the samples, readers who are interested in the
grammar can check the paper[64]. 4.2.4
Neural symbolic machines[71] Neural Symbolic Machine (NSM) contains (1) a
sequence-to-sequence
model
that
maps
language
utterances to programs and utilizes a key-varivable
memory to handle compositionality and (2) a symbolic
“computer”,
an executor that performs execution Fig. 6
A SEQ2TREE decoding example for the logical form
“AB(C)”[70]. 4.2.3
Sequence-to-tree[70] Sequence-to-sequence model ignores the hierarchical
structure of logical forms. To generate well-formed
output,
it needs to memorize various pieces of
auxiliary information (e.g., bracket pairs). Reference
[70] presented a hierarchical tree decoder to overcome
this problem. The tree decoder generated logical forms
in a top-down manner. In order to represent the tree
structure, sequence-to-tree[70] defined several special
token, e.g., “nonterminal” hni token which indicates
subtrees, hsi and h.i represent the beginning of a
sequence and nonterminal sequence, respectively, and
h=si is the end of sequence. Fig. 6
A SEQ2TREE decoding example for the logical form
“AB(C)”[70]. After encoding input q, the tree decoder used RNNs Big Data Mining and Analytics, December 2019, 2(4): 217–239 228 and helps find good programs by pruning the search
space. To train this machine end-to-end, NSMs[71]
applied REINFORCE to directly optimize the task
reward of this problem. The advantages of this
approach are as follows: (1) save and reuse intermediate
executions results with the combination of a sequence-
to-sequence model and a key-value memory, (2) use
the computer to execute partial programs and prune the
models’ search space, and (3) combine REINFORCE
with an iterative maximum likelihood training process. variables whose value was saved in the machine after
execution, NSMs augmented the seq2seq model with
a key-variable memory, where each entry has two
components: a continuous embedding key vi and a
corresponding variable token Ri referencing to the
value in the machine (Fig. 8). During encoding, NSMs
used an entity linker to link text spans (e.g., “US”)
to KB entities. For each linked entity, NSMs added a
memory entry where the key is the average of GRU
hidden states over the text span, and the variable token
(R1) is the name of a variable in the machine holding
the linked entity as its value. During decoding, when
a full sequence is generated, the result is stored as the
value of a new variable in the machine (keyed by the
hidden state at this step). When a new variable R1 with
key embedding v1 is added into the memory, the token
R1 is added into the decoder vocabulary with v1 as its
embedding. Unlike other approaches, NSMs adopted a Lisp
interpreter as the “executor” and defined a subset of
Lisp programming which is only equivalent to the
subset of -calculus presented in STAGG[60] as shown
in Fig. 7. 4.2.3
Sequence-to-tree[70] With the executor, it can help the model to
produce a list of valid tokens. First, a valid token should
not cause a syntax error, e.g., if the previous token is
“(”, the next token must be a function name. Second,
a valid token should not cause a semantic error, e.g.,
if the previously generated tokens were “Hop r”, the
next available token is restricted to predicates that are
reachable from entities in r. NSM executes non-differentiable operations against
a KB, and thus end-to-end backpropagation is not
possible. Therefore, NSMs used REINFORCE[72] with
some full supervision pre-train. To train from weak
supervision, NSMs proposed an iterative machine
learning
algorithm
(similar
to
hard
Expectation-
Maximization (EM)), where they searched for good
programs given fixed parameters, and then optimized
the probability of the best program found so far. To
do this, NSMs optimized the log-likelyhood objective
function: With the executor, the sequence-to-sequence aims
to map natural language into a program, the basic
structure is a standard seq2seq model with an attention
which is similar with JIA16[64], but extend it with a
key-value memory that allows the model to learn to
represent and refer to program variables. To let the
decoder learn to represent and refer to intermediate J D
X
x
logP.Abest.x/jx/; where Abest is the pseudo-gold program in the current
setting. Algorithm 5 summaries the overall training
schema. For pre-training, NSMs first ran iterative ML
for NML iterations and recorded the best program found
for every input xi. Then NSMs ran REINFORCE, Fig. 7
Interpreter functions of the NSM. r represents a
variable, p is a predicate in Freebase. ⩽⩽⩽and ⩾⩾⩾are defined
on numbers and dates[71]. is the knowledge base and
denotes a set of entities. Fig. 8
Semantic parsing with NSM. A special token “GO” indicates the start of decoding, and “Return” indicates the end of
decoding. Due to the fact that the decoding model never sees the values in the encoder (“US”) here, so it only references them
with the name of the variable (“R1”). The memory bridges these two steps to achieve compositionality[71]. Fig. 8
Semantic parsing with NSM. A special token “GO” indicates the start of decoding, and “Return” indicates the end of
decoding. Due to the fact that the decoding model never sees the values in the encoder (“US”) here, so it only references them
with the name of the variable (“R1”). where states and border are natural language predicates. where states and border are natural language predicates. g
g p
An advantage of FunQL is that abstract predicate can
be reused for both grounded and ungrounded meaning
representation. They considered five types of domain-
general predicates illustrated in Table 3. It is important
to notice that domain-general predicates are often
implicit and represent extra-sentential knowledge (e.g.,
all in the above utterances represents all states in the
domain which is not mentioned in the text). Input: Quenstion-answer pairs D D .xi; yi/, mix ratio ˛, reward
function R./, training iterations NML, NRL, and beam sizes
BML, BRL. Procedure: ocedu e:
Initialize C
x D ∅the best program so far for x and model
. Initialize C
x D ∅the best program so far for x and model
. 1: for n D 1; : : : ; NML do
2:
for .x; y/ in D do
3:
C Decode BML programs given x
4:
for j D 1; : : : ; jCj do
5:
if Rx;y.Cj / > Rx;y.C
x / then
6:
C
x Cj
7:
end if
8:
end for
9:
end for
10:
ML training with DML D .x; C
x /
11: end for 12: Initialize model for REINFORCE 12: Initialize model for REINFORCE
13: for n D 1; : : : ; NRL do
14:
DRL ∅is the RL training set
15:
for .x; y/ in D do
16:
C Decode BRL programs given x
17:
for j D 1; : : : ; jCj do
18:
if Rx;y.Cj / > Rx;y.C
x / then
19:
C
x Cj
20:
end if
21:
end for
22:
for j D 1; : : : ; jCj do
23:
Opj .1 ˛/
pj
P
j0 pj 0 , where pj D P.Cj jx/
24:
if Cj D C
x then
25:
Opj Opj C ˛
26:
end if
27:
DRL DRL [ f.x; Cj ; Opj /g
28:
end for
29:
end for
30:
REINFORCE training with DRL
31: end for In the setting of CHENG17, the task is to learn
semantic parser that maps input utterances x to logical
form z via an intermediate ungrounded representation
U . When G is executed against an executor,
it
outputs denotation (in weak supervision, optional)
y. CHENG17 used FunQL[76] for both grounded and
ungrounded meaning representation which is a variable-
free query language, where each predicate is treated as
a function symbol that modifies an argument list. For
example, FunQL grounded representation for an input
which states do not border texas is 4.2.3
Sequence-to-tree[70] The memory bridges these two steps to achieve compositionality[71]. 229 Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey modeling limitations. Moreover, without any task-
specific knowledge,
the learning is unconstrained
and may result in ill-formed output. CHENG17[73]
proposed a neural semantic parser that alleviates the
aforementioned problems. CHENG17[73] first mapped
utterances to an intermediate representation containing
natural language predicates. However,
instead of
using an external parser[74,75] or CCG grammars,
CHENG17 induced intermediate representations in
the form of predicate-argument structures from data
with a transition-based approach. This transition-
based approach by design yielded recursive semantic
structures, avoiding the problem of generating ill-
formed meaning representations. Another advantage
of transition-based approach is that it does not
require feature decomposition over structures and
thereby enabling the exploration of rich, non-local
features[73]. After CHENG17 got the output from
the transition system,
they grounded (e.g.,
to a
knowledge
base)
with
a
neural
mapping
model
under an assumption that grounded and ungrounded
structures are isomorphic. Details of this assumption
can be referred in the original paper[73]. answer.exclude.state.all/; next to.texas///; where next to is a domain-specific binary predicate that
takes one argument and returns a set of entities as its
denotation. The ungrounded meaning representation for
the same example is in which they normalized the probabilities of the
generated programs in bean to sum to .1 ˛/ and add
˛ to the probability of the best found one C .xi/. The
REINFORCE model always put a reasonable amount
of probability on a program with higher reward during
training. answer.exclude.states.all/; border.texas///; answer.exclude.states.all/; border.texas///; 4.2.5
CHENG17[73] Sequence-to-sequence model for semantic parsing
reduces the need for a domain-specific assumption,
grammar
learning,
and
more
expensive
feature
engineering. But this modeling flexibility brings a
problem that it is no longer possible to interpret how
the meaning composition is performed. Meanwhile,
this knowledge plays a critical role in understanding Big Data Mining and Analytics, December 2019, 2(4): 217–239
Fig. 9
Actions taken by the transition system for generating
the ungrounded meaning representation of the example. Symbols in red indicate domain-general predicates[73]. Big Data Mining and Analytics, December 2019, 2(4): 217–239 230 Data Mining and Analytics, December 2019, 2(4): 217–2 Table 3
List of domain-general predicates[73]. Predicate
Usage
Sub-category
Answer
Denotation wrapper
–
Type
Entity type checking
Stateid, cityid
riverid, etc. All
Querying for an entire
set of entities
–
Aggregation
One-argument meta
predicates for sets
Count, largest,
smallest, etc. Logical
connectives
Two-argument meta
predicates for sets
Intersect,
union, exclude Table 3
List of domain-general predicates[73]. Fig. 9
Actions taken by the transition system for generating
the ungrounded meaning representation of the example. Symbols in red indicate domain-general predicates[73]. Fig. 9
Actions taken by the transition system for generating
the ungrounded meaning representation of the example. Symbols in red indicate domain-general predicates[73]. CHENG17 decomposed the semantic parsing task in
two stages: (1) convert the utterance to an intermediate
representation
and
(2)
ground
the
intermediate
representation to a knowledge base. p.U jx/ D p.a; ujx/ D
T
Y
tD1
p.atja<t; x/p.utja<t; x/.at¤RED/: A
transition-based
system
generates
the
representation by following a derivation tree and
some canonical generation order. For FunQL, each
predicate can be visualized as a non-terminal node
of the tree while each entity as a terminal. Some
special predicates (e.g., all) acted as a terminal directly. CHENG17 proposed a neural transition system based
on RNNGs[77]. They used a buffer to store input tokens
and a stack to store partially generated trees. A key
difference in their approach was that tokens in the
buffer were not fetched or removed in a sequential
order. They allowed the generation algorithm to pick
tokens and combine logical forms in arbitrary orders. They defined three actions in the system: NT; TER, and
RED. CHENG17
encoded
the
input
buffer
with
a
bidirectional LSTM and the output stack with a
stack-LSTM[78]. 4.2.5
CHENG17[73] At each time step, the model used
the representation of the transition system et to predict
an action, where et is the concatenation of the buffer
representation bt and the stack representation st. When
the predicted action is NT or TER, ungrounded term ut
needs to be chosen from the candidate list (two types
of choices in Fig. 9). To select a domain-general term,
CHENG17 used et to compute the distribution over
candidate terms. For natural language term, CHENG17
computed the distribution over terms in the buffer based
on the st. (1) NT.X/ generates a non-terminal predicate (either
an entity or one of the domain-general predicates). The
type of predicate is determined by the placeholder X and
once generated, it is pushed onto the stack followed by
an open bracket. The open bracket will be closed by a
REDUCE operation. Since CHENG17 constrained the network to learn
ungrounded structures that are isomorphic to the target
meaning representation, how to convert ungrounded
representations to grounded ones becomes a simple
lexical mapping problem. To map an ungrounded term
ut to a grounded term gt, CHENG17 computed the
probability of gt given ut with a bi-linear neural
network: (2) TER.X/ generates a terminal entity or the special
predicate all. (3) RED stands for reduce and is used for subtree
completion. It recursively pops elements from the stack
until an open non-terminal node. Then the composite
term representing the entire subtree is pushed back to
the stack. The system will terminate when there were
no open non-terminals left in the stack. p.gtjut/ / exp. Eut Wug Egt
T/; p.gtjut/ / exp. Eut Wug Egt
T/; where
Eut
is the contextual representation of the
ungrounded term given by the Bi-LSTM, Egt is the
grounded term embedding, and Wug is the weight
matrix. Figure 9 shows the transition actions used to
generate the example. The neural model generates the
ungrounded representation U conditioned on utterance
x by recursively calling one of the three actions. U
is defined by a sequence of actions a and a sequence
of term choices (u). The conditional probability
p.U jx/ is For the transition system, the objective function is La D
X
xD
logp.ajx/ D
X
xD
n
X
tD1
logp.atjx/; where D denotes examples in the training data. For
the grounded term sequence g, since the ungrounded
terms are latent, CHENG17 maximized the expected 231 Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey log likelihood of the grounded terms for all examples,
which is a lower bound of the log likelihood logp.gjx/ parser to generate a map f : U ! Z for the target
domain t, which contains a set of training examples
.ui; zi/Nt
iD1. Meanwhile, some attributes are induced as
follows: by Jensen’s Inequality: Lg D
X
xD
X
u
Œp.ujx/ logp.gju; x/ D
X
xD
X
u
Œp.ujx/
k
X
tD1
logp.gtjut/ ⩽
X
xD
logp.gjx/: Zs and Zt are completely non-intersecting. Zs and Zt are completely non-intersecting. Typically, Ns Nt. Typically, Ns Nt. Assume the input utterance distributions in source
domain
and
target
domain
are
independent
and
different. To deal with the issues listed above, SU17 proposed
a paraphrase model and convert cross-domain semantic
parsing into a domain adaptation problem. Generally,
as shown in Fig. 10, the semantic parser for source
domain needs to get canonical utterance from the
logical forms. Taken them and input utterance as the
input, the paraphrase model incorporates with external
language resources to adapt new domains. The final objective is the combination of La and Lg,
denoted as L D La C Lg. 4.2.6
SU17[79] The majority of existing approaches of semantic parsing
focus on in-domain setting, while SU17[79] perform
not well in cross-domain circumstance. Due to the
diversity of language semantic in different domains,
cross-domain still remains a challenging task. For
instance, in Fig. 10, in team domain, the semantic
parser is developed to extract the predicates like team
and season. However, in the social domain, the
semantic parser needs to extract some predicates like
employee. SU17 proposed a sequence-to-sequence model with
soft-attention
for
domain
adaptation
problem. A
sequence-to-sequence model is composed with two
components: encoder and decoder. For the encoder,
they exploited a bi-directional RNN to encode the
input utterance u.u1; u2; : : : ; um) to a sequence of state
vectors h.h1; h2; : : : ; hm). The state vectors of forward
RNN and backward RNN are computed respectively as Although introducing a paraphrasing model is a
widely adopted method, it still brings some open
challenging problems. For example, OVERNIGHT
dataset has eight domains, and 30% to 55% words in
each domain never happen in other domains[79]. Due to
this out-of-vocabulary problem, the paraphrasing model
cannot address with cross-domain well. ! hi D GRUf w. .ui/; ! hi 1/;
hi D GRUf w. .ui/; ! hiC1/: SU17 also proposed an RNN model with attention
as the decoder, which generates one token at a time
step[79]. The attention weights are calculated by encoder
states, and then they are used to compute the next
decoder state dj C1 and the probability distribution
p.cj ju; c1Wj 1/ as follows: To begin with, SU17 defined cross-domain semantic
parsing task as follows: suppose U is a set of input
utterances.ui; zi/Ns
iD1, and Z is the set of logical forms. Given K source domains, the task is to train a semantic Fig. 10
Framework of semantic parsing via paraphrasing. Firstly, the logical forms are converted deterministically into
canonical utterance in natural language. Combing canonical utterance with input utterance as the input, paraphrase model is
trained to learn and transfer from the source-domain to the target-domain. External language resources are applied to facilitate
domain adaptation[79]. Fig. 10
Framework of semantic parsing via paraphrasing. Firstly, the logical forms are converted deterministically into
canonical utterance in natural language. Combing canonical utterance with input utterance as the input, paraphrase model is
trained to learn and transfer from the source-domain to the target-domain. External language resources are applied to facilitate
domain adaptation[79]. 232 Big Data Mining and Analytics, December 2019, 2(4): 217–239 d0 D tanh.W0Œ ! hm; h1/;
uji D vT tanh.W1Hi C W2dj /;
˛ji D
Uji
Pm
i0D1 uji0 ;
h0
j D
m
X
iD1
˛jihi;
dj C1 D GRU.Π.cj /; h0
j ; dj /;
p.cj ju; c1Wj 1/ / exp.U Œdj ; h0
j /: d0 D tanh.W0Π! hm; h1/; d0 D tanh.W0Π! hm; h1/; 5.1
JOBS Based on the analysis above, SU17 proposed a
unit variance standardization after initializing word
embedding. There
are
two
choices,
per-example
standardization which standardizes every row of the
word embedding matrix by simple dividing them with
the standard deviation, and per-feature which does the
same operations on columns instead. This benchmark dataset[85] has 640 queries to a database
of job listing. Each question in it is paired with Prolog-
style query. In Refs. [28, 70], JOBS split the dataset
into 500 training samples and 140 test instances. Best
accuracy of this dataset is 0.9 achieved by Ref. [70]. 4.2.7
HERZIG17[80] In parallel of previous approch[79], Ref. [80] explored
another direction of semantic parsing with multiple
domains. HERZIG17[80] trained a single model for
all the domains,
and attached a domain-specific
encoding to help the parser distinguish between
domains. HERZIG17 also employed a sequence-to-
sequence model with attention as a base structure. To
add the explicit representation of the domain that is
being decoded, HERZIG17 encoded the k-th domain
by a one-hot vector dk RK as an addition input of
the decoder. As an alternate way, HERZIG17 adopted
a similar intuition from neural machine translation[81],
where they added an artificial token at the beginning of
each input sentence to specify the target domain. Since
HERZIG17 used one decoder for multiple domains,
tokens from different domains could be generated
during decoding. HERZIG17 prevented that at test
time by excluding out-of-domain tokens before the
last softmax function. The experiments showed that
training a single model for all domains not only got a
better result but also contained much less parameters. In
addition, HERZIG17 found the single model performed
much better when training data is limited. p.cj ju; c1Wj 1/ / exp.U Œdj ; h0
j /: Given a set of training data .ui; ci/N
iD1, SU17
minimized the loss function 1
N
N
X
iD1
logp.cijui/ with
cross-entropy. In other words, SU17 maximized the log
probability of correct canonical utterances. Besides, taking advantages of the standardize pre-
trained word embedding, SU17 enabled the model to
cope with unseen out-of-domain data. Specifically, they
initialized the parameters of the word embedding layer
with the pre-trained word embedding model in the
source domain. By doing this, SU17 addressed with the
out-of-vocabulary problems. Based on the observations
during the experiments, SU17 claimed that word
embedding initialization faces with two dilemmas: one
is small micro variance, which hinders optimization,
and the other is large macro variance, which hinders
generalization. The comparison between different word
embedding initializations (Table 4) shows that the small
micro variance brings a poor starting point, while large
variance is hard to generalize with the unseen words
during training. 5
Datasets Datasets play a critical role in both conventional
statistical learning and deep learning. We survey some
of the existing datasets and describe their properties
together with the state-of-the-art performance. Table 5
gives an overview of all datasets. 5.7
WebQuestionSP WebQuestionSP dataset[86] is an extension word of
WebQuestions published in 2016. It contains full
semantic parses for a subset of the questions from
WebQuestions. Because 18.5% of the original dataset
were found “not answerable”. This dataset has 3098
question-answer pairs for training and 1639 for testing
with labeled logical form, which was collected using
Google Suggest API. The state-of-the-art result of this
dataset is 63.9% achieved by STAGG and Ref. [82]. 5.8
SPADES Datasets above have a limitation on the number
of training samples. SPADES[87] consists of 93 319
questions derived from CLUEWEB09[88] sentences
published in 2016. Specifically, the questions were
created by randomly removing an entity, and producing
sentence-denotation pairs. The sentences include two
or more entities and although they are not very
compositional, it is still a large-scale dataset for neural
network training. Reference [83] achieves the state-of-
the-art performance with an F1 score 39.9%. Reference
[83] leverages external unstructured text information to
do answer selection, which is not directly related to 5.4
Regexp824 The Regexp824 dataset contains 824 natural language
and regular expression pairs published in 2013[13], one
example is “three letter words starting with ‘X’”. One
challenge of this dataset is that there are many logically
equivalent regular expressions, some aligning better to
the natural language than others. Reference [13] obtains
the best performance with 65.6% accuracy. 5.6
WebQuestions which is better compared with JIA16[64] but slightly
lower than conventional approaches. which is better compared with JIA16[64] but slightly
lower than conventional approaches. WebQuestions[49]
is
another
dataset
on
Freebase
published in 2013. It consists of 5810 question-answer
pairs (no logical form) such as “what do Australia call
their money?” Similar to Free917, the questions are
not very complicated, but unlike Free917, the questions
in WebQuestions are real questions asked by people
over the world on the Web independent from Freebase. These real questions result in that the dataset is more
realistic and varied. In this dataset, STAGG[60] achieves
the best performance with accuracy 58.8%. 5.3
ATIS-3 Published in 2007, the ATIS-3 dataset[31] consists of
5418 utterances paired with logical forms (e.g., “show
me information on american airlines from fort worth
texas to philadelphia”). They are extracted from a
flight booking system. The standard split has 4480
training instances, 480 development instances, and 450
test instances. The utterances in this dataset contain
more disfluencies and flexible word order compared
with Geo880, but they are logically simpler. The best-
reported result is based on semantic parsing and obtains
84.6% accuracy[31]. 5.2
Geo880 Table 4
Comparison between different word embedding
initializations. ES: per-example standardization. FS: per-
feature standardization. EN: per-example normalization. Cosine similarity is computed on a randomly selected. The Geo880 dataset[26], which is published in 1996,
drove nearly a decade of semantic parsing research. It has 880 questions and database query pairs about
US geography (e.g., “what is the highest point in the
largest state”). The utterances are compositional, but
the language is simple and the domain is limited. The
majority of questions include at most one entity. On
this dataset, learning from logical forms[48] and data[6]
both achieve around 90% accuracy. For deep learning
approaches, CHENG17[73] achieves 86.7% accuracy Initialization
L2 norm
Micro
variance
Cosine
similarity
Random
17.3 ˙ 0.45
1.00 ˙ 0.05
0.00 ˙ 0.06
WORD2VEC
2.04 ˙ 1.08
0.02 ˙ 0.02
0.13 ˙ 0.11
WORD2VEC+ES
17.3 ˙ 0.05
1.00 ˙ 0.00
0.13 ˙ 0.11
WORD2VEC+FS
16.0 ˙ 8.47
1.09 ˙ 1.31
0.12 ˙ 0.10
WORD2VEC+EN
1.00 ˙ 0.00
0.01 ˙ 0.00
0.13 ˙ 0.11 233 Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey l.:
Statistical Learning for Semantic Parsing: A Surve Table 5
An overview of all datasets (“–” means there is not official split for this dataset). Dataset name
Training set number
Development set number
Test set number
Best result (%)
Supervision form
JOBS
500
–
140
90.0[70]
Supervision
Geo880
880
–
–
91.1[6]
Supervision
ATIS–3
5418
–
–
84.6[31]
Supervision
Regexp824
824
–
–
65.6[13]
Supervision
Free917
917
–
–
68[13]
Supervision
WebQuestions
5810
–
–
58.8[60]
Weak Supervision
WebQuestionSP
3098
–
1639
63.9[60,82]
Weak Supervision
SPADES
93 319
–
–
39.9[83]
Weak Supervision
SimpleQuestion
75 910
10 845
21 687
78.7[82]
Weak Supervision
WikiTableQuestions
22 033
–
–
37.1[84]
Weak Supervision
OVERNIGHT
13 682
–
–
80.6[79]
Supervision
IFTTT
77 495
5171
4294
74.2[70]
Supervision
which is better compared with JIA16[64] but slightly
5.6
WebQuestions Table 5
An overview of all datasets (“–” means there is not official split for this dataset). s (“–” means there is not official split for this dataset). 6.2
Current and future tendency Although semantic parsing is facing a bunch of
problems, researchers still try to settle them with some
other methods. We discuss some potential directions on
semantic parsing as follows. 5.12
IFTTT Reference [91] created this dataset in 2015. This
dataset extracted a large number of if-this-then-that
recipes from the IFTTT website (http://www.ifttt.com). Recipes are the programs that users specify on the site
with exactly one trigger and one action (whenever the
trigger is activated, the action is performed). Different
from other datasets, in IFTTT, each utterance pairs
with one trigger and on action, e.g., an utterance “turn
on heater when temperature drops below 58 degree”,
a TRIGGER: Weather — Current temperature drops
below — ((Temperature (58)) (Degrees in (f))), and an
ACTION: WeMo Insight Switch — Turn on — ((Which
switch? (“ ”))). The original split which contains 77 495
training samples, 5171 development samples, and 4294 6
Challenges and Future Tendency Another dataset for a large scale training schema is
SimpleQuestion[89] published in 2015. It consists of a
total 108 442 questions written in natural language with
corresponding fact and answer. Facts are extracted from
Freebase. The dataset is split into a training set (75 910
samples), a development set (10 845 samples), and a test
set (21 687 samples). The current best result is achieved
by Ref. [82] with accuracy 78.7%. In this section, we discuss the current challenges and
a future tendency of semantic parsing. Although a
lot of effort has been applied in recent decades, the
performance is still far from satisfaction. We conclude
three challenges and several potential directions for
semantic parsing. 5.11
OVERNIGHT OVERNIGHT[90] dataset consists of 8 domains, each
of them is based on a separate knowledge base with
logical forms written in -DCS. The logical forms
are converted into canonical utterances via a simple
grammar. The input utterances are collected by crowd
sourcing. Different domains are aimed to stress different
linguistic phenomena, which requires the parser more
flexible and general. References [79,80] both achieved
best accuracy around 0.8 on average of the 8 domains. Third, is it a good way to represent semantic
or language in a probability way? Logic forms are
connected with rules and discrete in a natural way and
so does language. Combining logic and probability may
be a potential direction for solving this problem, and it
still needs a lot of effort. 5.10
WikiTableQuestions First, unlike other structure prediction problems, the
number of labels in semantic parsing is much more. However, labeling the semantic parsing data costs
a lot. In supervised learning, datasets need to be
labeled as pairs of utterance and logic form while
in weak supervised learning, the corresponding action
is needed. Meanwhile, although semantic parser can
handle utterances in a narrow and closed domain
setting, it still lacks the ability to parse and understand
open-domain queries from human[92]. The goal of WikiTableQuestions is to extend question
answering beyond knowledge based on HTML tables
on Wikipedia,
which are semi-structured[84]. The
dataset consists of 22 033 question-table-answer pairs
(e.g., “how many runners took 2 minutes at the most
to run 1500 meters?”). Each question can be answered
based on the information inside a given table. People
need to aggregate information across the whole table
with the reasoning based on the question. At test time,
new questions together with new tables are provided to
be answered. The result of this dataset is 37.1%. Second, there is no universal logic form for all
tasks. Syntax and complexity of different logic forms
vary a lot. To learn one model for all is a very
challenging problem. Although there are several models
trained in a multi-task way, it still lacks the ability
of generalization. This is an open problem for all
machine learning problems. Fortunately, researchers
have proposed several ways for it. 5.5
Free917 The Free917[47] published in 2013 has 917 examples
of question-logical form pairs that can be answered
directly via Freebase (e.g., “how many works did
Mozard dedicate to Joseph Haydn?”). The questions
are logically less complex than those in the semantic
parsing datasets above, but introduce the new challenge
of scaling up to many more predicates (need to deal
with a large knowledge base). The state-of-the-art
performance is 68% accuracy[13]. Big Data Mining and Analytics, December 2019, 2(4): 217–239 234 test samples. Best F1 score is 74.2% achieved by Ref. [70]. test samples. Best F1 score is 74.2% achieved by Ref. [70]. semantic parsing. semantic parsing. semantic parsing. 6.2.3
Model-agnostic meta-learning Model-Agnostic Meta-Learning (MAML)[98] aims to
learn the learners (for the tasks) and the meta-learner in
the few-shot meta-learning setup. It considers a model
represented by parameters . When the model adapts
to a new task, the model changes its parameters to
i0. In this way, the model can quickly adapt to a new
task given very few training samples. Reference [99]
reduced the supervised learning problem to a few-shot
learning problem by treating each training example as
a unique pseudo-task and achieved faster converge and
better performance when generating Structured Query
Language (SQL) from wiki queries. [7]
K. Bollacker, C. Evans, P. Paritosh, T. Sturge, and J. Taylor, Freebase: A collaboratively created graph database
for structuring human knowledge, in Proc. 2008 ACM
SIGMOD Int. Conf. Management of Data, Vancouver,
Canada, 2008, pp. 1247–1250. [8]
D. Vrandecic and M. Kr¨otzsch,
Wikidata: A free
collaborative knowledgebase, Commun. ACM, vol. 57, no. 10, pp. 78–85, 2014. [9]
S. Tellex, T. Kollar, S. Dickerson, M. R. Walter, A. G. Banerjee,
S. Teller,
and N. Roy,
Understanding
natural language commands for robotic navigation and
mobile manipulation, in Proc. 25th AAAI Conf. Artificial
Intelligence, San Francisco, CA, USA, 2011. References [1]
P. Pasupat and P. Liang, Compositional semantic parsing
on semi-structured tables, in Proc. 53rd Ann. Meeting of
the Association for Computational Linguistics and the 7th
Int. Joint Conf. Natural Language Processing, Beijing,
China, 2015. Acknowledgment showed huge improvement in semantic template level
performance. Similar
algorithms[94]
are
proposed,
which leveraged a shared feature extraction layer
for slot-filling across multiple domains. In Reference
[95], the authors achieved good performance of a
seq-to-seq model with three multi-task architectures. Transfer learning can also be extended as lifelong
learning, whose goal is to sequentially retain learned
knowledge and to selectively transfer that knowledge
when learning a new task to develop more accurate
hypotheses or policies[96]. This work was partially supported by National Science
Foundation (No. CNS-1842407), National Institutes of
Health (No. R01GM110240), and Industry Members of
NSF Center for Big Learning (http://nsfcbl.org/index. php/partners/). 6.2.1
Transfer learning There are some approaches for semantic parsing via
transfer learning, trained from a source task with a large
amount of labeled data together with a target task with
smaller labeled data. Reference [93] proposed to train
a single multi-task deep learning model with shared
implicit features learning across domains. The approach 235 Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey 6.2.2
Zero-shot learning Zero-shot learning aims to recognize objects whose
instances may not have been seen during training along
with some side information. Some work based on zero-
shot learning has been applied to semantic parsing. Reference [92] proposed a slot-filling approach with
natural language slot label descriptions. With multi-
task stacked Bi-LSTM, this method can alleviate the
need for sufficient labeled training data and bootstrap
a model for domain scaling problems. Similar ideas
are exploited in Ref. [97] to establish a general slot or
concept tagger over multiple domains. Different from
lifelong learning, the zero-shot learning is more strict. [2]
P. Liang, Learning executable semantic parsers for natural
language understanding, Commun. ACM, vol. 59, no. 9, pp. 68–76, 2016. [3]
W. Woods, R. Kaplan, and B. Webber, The Lunar Sciences
Natural Language Information System:
Final Report. Cambridge, MA, USA: Bolt Beranek and Newman Inc.,
1972. [4]
T. Winograd, Understanding Natural Language. New
York, NY, USA: Academic Press, 1972. [5]
J. Clarke, D. Goldwasser, M. W. Chang, and D. Roth,
Driving semantic parsing from the world’s response,
in Proc. 14th Conf. Computational Natural Language
Learning, Uppsala, Sweden, 2010, pp. 18–27. [6]
P. Liang, M. I. Jordan, and D. Klein, Learning dependency-
based compositional semantics,
in Proc. 49th Ann. Meeting of the Association for Computational Linguistics:
Human Language Technologies-Volume 1, Portland, OR,
USA, 2011, pp. 590–599. 7
Conclusion Computer Vision and Pattern Recognition, Las
Vegas, NV, USA, 2016, pp. 770–778. [31] L. S. Zettlemoyer and M. Collins,
Online learning
of
relaxed
CCG
grammars
for
parsing
to
logical
form,
in
Proc. Joint
Conf. Empirical
Methods
in
Natural Language Processing and Computational Natural
Language Learning, Prague, Czech Republic, 2007, pp. 678–687. [17] Y. Li, H. Z. Qi, J. F. Dai, X. Y. Ji, and Y. C. Wei,
Fully convolutional instance-aware semantic segmentation,
in Proc. 2017 IEEE Conf. Computer Vision and Pattern
Recognition, Honolulu, HI, USA, 2017. [18] X. Z. Ma and E. Hovy, End-to-end sequence labeling
via
Bi-directional
LSTM-CNNs-CRF,
arXiv
preprint
arXiv:1603.01354, 2016. [32] T. Kwiatkowski, L. Zettlemoyer, S. Goldwater, and M. Steedman, Inducing probabilistic CCG grammars from
logical form with higher-order unification, in Proc. 2010
Conf. Empirical Methods in Natural Language Processing,
Cambridge, MA, USA, 2010, pp. 1223–1233. [19] G. Lample, M. Ballesteros, S. Subramanian, K. Kawakami,
and C. Dyer, Neural architectures for named entity
recognition, in Proc. 2016 Conf. North American Chapter
of the Association for Computational Linguistics: Human
Language Technologies, San Diego, CA, USA, 2016. [33] P. Liang,
Lambda
dependency-based
compositional
semantics, arXiv preprint arXiv:1309.4408, 2013. [20] Q. L. Zhu,
X. L. Li,
A. Conesa,
and P. Cecile,
GRAM-CNN: A deep learning approach with local
context for named entity recognition in biomedical text,
Bioinformatics, vol. 34, no. 9, pp. 1547–1554, 2018. [34] Z. C. Zheng, F. T. Li, M. L. Huang, and X. Y. Zhu,
Learning to link entities with knowledge base, in Human
Language Technologies:
The 2010 Ann. Conf. North
American Chapter of the Association for Computational
Linguistics, Los Angeles, CA, USA, 2010, pp. 483–491. [21] R. M. Sun, X. Y. Yuan, P. He, Q. L. Zhu, A. K. Chen,
A. Gregio, D. Oliveira, and X. L. Li, Learning fast and
slow: Propaedeutica for real-time malware detection, arXiv
preprint arXiv:1712.01145, 2017. [35] Y. Yang and M. W. Chang,
S-mart:
Novel tree-
based structured learning algorithms applied to tweet
entity linking, in Proc. Association for Computational
Linguistics, Beijing, China, 2015. [22] R. Collobert, J. Weston, L. Bottou, M. Karlen, K. Kavukcuoglu, and P. Kuksa, Natural language processing
(almost) from scratch, J. Mach. Learn. Res., vol. 12, pp. 2493–2537, 2011. [36] R. C. Schank and L. Tesler, A conceptual dependency
parser
for
natural
language,
in
Proc. 1969
Conf. Computational Linguistics, S˚ang-S¨aby, Sweden, 1969, pp. 1–3. [23] E. F. T. K. Sang and F. 7
Conclusion [10] Y. Artzi and L. Zettlemoyer, Weakly supervised learning
of semantic parsers for mapping instructions to actions,
Trans. Assoc. Comput. Linguist., vol. 1, pp. 49–62, 2013. We
have
presented
a
general
semantic
parsing
framework
for
the
problem
of
natural
language
understanding, and then given an introduction of some
representative statistical learning algorithms including
deep learning for semantic parsing. Going forward, we
also collected the datasets used to train semantic parsing
models and provided the state-of-the-art performance. At last, we analyzed several challenges for semantic
parsing and pointed out several potential ways for
further research. [11] C. Matuszek, N. FitzGerald, L. Zettlemoyer, L. Bo, and
D. Fox, A joint model of language and perception for
grounded attribute learning, in Proc. 29th Int. Conf. Machine Learning, Edinburgh, UK, 2012. [12] J. Krishnamurthy and T. Kollar, Jointly learning to parse
and perceive: Connecting natural language to the physical
world, Trans. Assoc. Comput. Linguist., vol. 1, pp. 193–
206, 2013. [13] N. Kushman and R. Barzilay, Using semantic unification Big Data Mining and Analytics, December 2019, 2(4): 217–239 236 to generate regular expressions from natural language,
in Proc. 2013 Conf. North American Chapter of the
Association
for
Computational
Linguistics:
Human
Language Technologies, Atlanta, GA, USA, 2013. parsing with statistical machine translation, in Proc. Main
Conf. Human Language Technology Conf. North American
Chapter of the Association of Computational Linguistics,
New York, NY, USA, 2006, pp. 439–446. [28] L. S. Zettlemoyer and M. Collins, Learning to map
sentences to logical form: Structured classification with
probabilistic categorial grammars, in Proc. 21st Conf. Uncertainty in Artificial Intelligence, Edinburgh, UK,
2005. [14] Y. LeCun, Y. Bengio, and G. Hinton, Deep learning,
Nature, vol. 521, no. 7553, pp. 436–444, 2015. [15] G. Hinton, L. Deng, D. Yu, G. E. Dahl, A.R. Mohamed,
N. Jaitly, A. Senior, V. Vanhoucke, P. Nguyen, T. N. Sainath, et al., Deep neural networks for acoustic modeling
in speech recognition: The shared views of four research
groups, IEEE Signal Process. Mag., vol. 29, no. 6, pp. 82–
97, 2012. [29] M. Steedman, The Syntactic Process. Cambridge, MA,
USA: MIT Press, 2000. [30] Y. Artzi, K. Lee, and L. Zettlemoyer, Broad-coverage
CCG semantic parsing with AMR, in Proc. 2015 Conf. Empirical Methods in Natural Language Processing,
Lisbon, Portugal, 2015, pp. 1699–1710. [16] K. M. He, X. Y. Zhang, S. Q. Ren, and J. Sun, Deep
residual learning for image recognition, in Proc. 2016
IEEE Conf. Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey 237 [57] Y. Bengio, P. Simard, and P. Frasconi, Learning long-
term dependencies with gradient descent is difficult, IEEE
Trans. Neural Netw., vol. 5, no. 2, pp. 157–166, 1994. [42] K. Zhao, Structured Prediction with Perceptron: Theory
and Algorithms, New York, NY, USA: The City University
of New York, 2014. [58] S. Hochreiter and J. Schmidhuber,
Long short-term
memory, Neural Comput., vol. 9, no. 8, pp. 1735–1780,
1997. [43] P. Liang, A. Bouchard-Cˆot´e, D. Klein, and B. Taskar, An
end-to-end discriminative approach to machine translation,
in Proc. 21st Int. Conf. Computational Linguistics and the
44th Ann. Meeting of the Association for Computational
Linguistics, Sydney, Australia, 2006, pp. 761–768. [59] K. Cho, B. van Merri¨enboer, C. Gulcehre, D. Bahdanau,
F. Bougares, H. Schwenk, and Y. Bengio, Learning phrase
representations using RNN encoder-decoder for statistical
machine translation,
in Proc. 2014 Conf. Empirical
Methods in Natural Language Processing, Doha, Qatar,
2014. [44] N. Singh-Miller and M. Collins, Trigger-based language
modeling using a loss-sensitive perceptron algorithm, in
Proc. 2007 IEEE Int. Conf. Acoustics, Speech and Signal
Processing, Honolulu, HI, USA, 2007, pp. 25–28. [60] S. W. T. Yih, M. W. Chang, X. D. He, and J. F. Gao, Semantic parsing via staged query graph generation:
Question answering with knowledge base, in Proc. 53rd
Ann. Meeting of the Association for Computational
Linguistics and the 7th Int. Joint Conf. Natural Language
Processing, Beijing, China, 2015, pp. 1321–1331. [45] Y. Artzi and L. Zettlemoyer, Bootstrapping semantic
parsers from conversations, in Proc. 2011 Conf. Empirical
Methods in Natural Language Processing, Edinburgh, UK,
2011, pp. 421–432. [46] D. L. Chen and R. J. Mooney, Learning to interpret natural
language navigation instructions from observations, in
Proc. 25th
AAAI
Conf. Artificial
Intelligence,
San
Francisco, CA, USA, 2011, pp. 859–865. [61] A. Bordes, S. Chopra, and J. Weston, Question answering
with subgraph embeddings, in Proc. 2014 Conf. Empirical
Methods in Natural Language Processing, Doha, Qatar,
2014. [47] Q. Q. Cai and A. Yates, Large-scale semantic parsing via
schema matching and lexicon extension, in Proc. 51st Ann. Meeting of the Association for Computational Linguistics,
Sofia, Bulgaria, 2013, pp. 423–433. [62] J. Bao, N. Duan, M. Zhou, and T. Zhao, Knowledge-
based question answering as machine translation, in Proc. 52nd Ann. Meeting of the Association for Computational
Linguistics, Baltimore, MD, USA, 2014. [48] T. Kwiatkowski, E. Choi, Y. Artzi, and L. Zettlemoyer,
Scaling
semantic
parsers
with
on-the-fly
ontology
matching, in Proc. 2013 Conf. Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey Empirical Methods in
Natural Language Processing, Seattle, WA, USA, 2013. [63] J. Bromley, I. Guyon, Y. LeCun, E. S¨ackinger, and R. Shah, Signature verification using a “siamese” time delay
neural network, in Proc. 6th Int. Conf. Neural Information
Processing Systems, Denver, CO, USA, 1993, pp. 737–744. [49] J. Berant, A. Chou, R. Frostig, and P. Liang, Semantic
parsing on freebase from question-answer pairs, in Proc. 2013 Conf. Empirical Methods in Natural Language
Processing, Seattle, WA, USA, 2013. [64] R. Jia and P. Liang, Data recombination for neural semantic
parsing, in Proc. 54th Ann. Meeting of the Association
for Computational Linguistics, Berlin, Germany, 2016, pp. 12–22. [50] J. Berant and P. Liang, Semantic parsing via paraphrasing,
in Proc. 52nd
Ann. Meeting of the Association for
Computational Linguistics, Baltimore, MA, USA, 2014,
pp. 1415–1425. [65] D. Bahdanau, K. Cho, and Y. Bengio, Neural machine
translation by jointly learning to align and translate, in
Proc. Int. Conf. Learning Representations, San Diego, CA,
USA, 2015. pp
[51] T. Mikolov, I. Sutskever, K. Chen, G. Corrado, and J. Dean, Distributed representations of words and phrases
and their compositionality, in Proc. 26th Int. Conf. Neural
Information Processing Systems, Lake Tahoe, NV, USA,
2013, pp. 3111–3119. [66] M. T. Luong, H. Pham, and C. D. Manning, Effective
approaches to attention-based neural machine translation,
in Proc. 2015 Conf. Empirical Methods in Natural
Language Processing, Lisbon, Portugal, 2015. [52] Y. LeCun, B. Boser, J. S. Denker, D. Henderson, R. E. Howard, W. Hubbard, and L. D. Jackel, Backpropagation
applied to handwritten zip code recognition, Neural
Comput., vol. 1, no. 4, pp. 541–551, 1989. [67] O. Vinyals, M. Fortunato, and N. Jaitly, Pointer networks,
in Proc. 28th Int. Conf. Neural Information Processing
Systems, Montreal, Canada, 2015, pp. 2692–2700. [53] I. Goodfellow,
Y. Bengio,
and A. Courville,
Deep
Learning. Cambridge, MA, USA: MIT Press, 2016. [68] A. Krizhevsky, I. Sutskever, and G. E. Hinton, Imagenet
classification with deep convolutional neural networks,
in Proc. 25th Int. Conf. Neural Information Processing
Systems, Lake Tahoe, NV, USA, 2012, pp. 1097–1105. [54] Y. T. Zhou, R. Chellappa, A. Vaid, and B. K. Jenkins,
Image restoration using a neural network, IEEE Trans. Acoust. Speech Signal Process., vol. 36, no. 7, pp. 1141–
1151, 1988. [69] N. Jaitly and G. E. Hinton, Vocal Tract Length Perturbation
(VTLP) improves speech recognition, in Proc. 30th Int. Conf. Machine Learning, Atlanta, GA, USA, 2013. [55] M. Schuster and K. K. 7
Conclusion De Meulder, Introduction to the
CoNLL-2003 shared task: Language-independent named
entity recognition, in Proc. 7th Conf. Natural Language
Learning at HLT-NAACL 2003-Volume 4, Edmonton,
Canada, 2003, pp. 142–147. [37] K. Zhao and L. Huang, Type-driven incremental semantic
parsing with polymorphism, in Proc. Human Language
Technologies: The Ann. Conf. North American Chapter of
the ACL, Denver, CO, USA, 2015. [38] R. S. Sutton and A. G. Barto, Reinforcement Learning: An
Introduction. Cambridge, MA, USA: MIT Press, 1998. [24] M. P. Marcus, M. A. Marcinkiewicz, and B. Santorini,
Building a large annotated corpus of English: The Penn
Treebank, Comput. Linguist., vol. 19, no. 2, pp. 313–330,
1993. [39] M. Collins, Discriminative training methods for hidden
Markov models: Theory and experiments with perceptron
algorithms, in Proc. ACL-02 Conf. Empirical Methods
in Natural Language Processing-Volume 10, Stroudsburg,
PA, USA, 2002, pp. 1–8. [25] D. K. Misra and Y. Artzi, Neural shift-reduce CCG
semantic parsing, in Proc. 2016 Conf. Empirical Methods
in Natural Language Processing, Austin, TX, USA, 2016,
pp. 1775–1786. [40] F. Rosenblatt,
The
Perceptron:
A
Perceiving
and
Recognizing Automation. Buffalo, NY, USA: Cornell
Aeronautical Laboratory, 1957. [26] J. M. Zelle and R. J. Mooney, Learning to parse database
queries using inductive logic programming, in Proc. 13th
National Conf. Artificial Intelligence, Portland, OR, USA,
1996, pp. 1050–1055. [41] Y. Freund and R. E. Schapire, Large margin classification
using the perceptron algorithm, Mach. Learn., vol. 37, no. 3, pp. 277–296, 1999. [27] Y. W. Wong and R. J. Mooney, Learning for semantic Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey Weston, Large-
scale simple question answering with memory networks,
arXiv preprint arXiv:1506.02075, 2015. [76] R. J. Kate, Y. W. Wong, and R. J. Mooney, Learning
to transform natural to formal languages, in Proc. 20th
National Conf. Artificial Intelligence, Pittsburgh, PA, USA,
2005, pp. 1062–1068. [90] Y. S. Wang, J. Berant, and P. Liang, Building a semantic
parser overnight, in Proc. 53rd Ann. Meeting of the
Association for Computational Linguistics and the 7th Int. Joint Conf. Natural Language Processing, Beijing, China,
2015, pp. 1332–1342. [77] C. Dyer, A. Kuncoro, M. Ballesteros, and N. A. Smith,
Recurrent neural network grammars, in Proc. NAACL-
HLT, San Diego, CA, USA, 2016. [78] C. Dyer, M. Ballesteros, W. Ling, A. Matthews, and N. A. Smith, Transition-based dependency parsing with stack
long short-term memory, in Proc. 53rd Ann. Meeting of the
Association for Computational Linguistics and the 7th Int. Joint Conf. Natural Language Processing, Beijing, China,
2015. [91] C. Quirk, R. Mooney, and M. Galley, Language to code:
Learning semantic parsers for if-this-then-that recipes,
in Proc. 53rd
Ann. Meeting of the Association for
Computational Linguistics and the 7th Int. Joint Conf. Natural Language Processing, Beijing, China, 2015, pp. 878–888. [79] Y. Su and X. F. Yan, Cross-domain semantic parsing
via paraphrasing, in Proc. 2017 Conf. Empirical Methods
in Natural Language Processing, Copenhagen, Denmark,
2017. [92] A. Bapna, G. T¨ur, D. Hakkani-T¨ur, and L. Heck, Towards
zero-shot frame semantic parsing for domain scaling, in
Proc. INTERSPEECH, Stockholm, Sweden, 2017. [93] D. Hakkani-T¨ur, G. T¨ur, A. Celikyilmaz, Y. N. V. Chen,
J. F. Gao, L. Deng, and Y. Y. Wang, Multi-domain
joint semantic frame parsing using bi-directional RNN-
LSTM, in Proc. 17th Ann. Meeting of the Int. Speech
Communication Association, San Francisco, CA, USA,
2016, pp. 715–719. [80] J. Herzig and J. Berant, Neural semantic parsing over
multiple knowledge-bases, in Proc. 55th Ann. Meeting of
the Association for Computational Linguistics, Vancouver,
Canada, 2017. [81] M. Johnson, M. Schuster, Q. V. Le, M. Krikun, Y. H. Wu, Z. F. Chen, N. Thorat, F. Vi´egas, M. Wattenberg,
G. Corrado, et al., Google’s multilingual neural machine
translation system: Enabling zero-shot translation, Trans. Assoc. Comput. Linguist., vol. 5, pp. 339–351, 2017. [94] A. Jaech, L. Heck, and M. Ostendorf, Domain adaptation
of
recurrent
neural
networks
for
natural
language
understanding, in Proc. 17th Ann. Meeting of the Int. Speech Communication Association, San Francisco, CA,
USA, 2016. [82] M. Yu, W. P. Yin, K. Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey Paliwal, Bidirectional recurrent
neural networks, IEEE Trans. Signal Process., vol. 45, no. 11, pp. 2673–2681, 1997. [70] L. Dong and M. Lapata, Language to logical form
with neural attention, in Proc. 54th Ann. Meeting of
the Association for Computational Linguistics, Berlin,
Germany, 2016. [56] S. Hochreiter,
Untersuchungen
zu
dynamischen
neuronalen
netzen,
Master
dissertation,
Technische
Universitaet M¨unchen, Munich, Germany, 1991. [71] C. Liang, J. Berant, Q. Le, K. D. Forbus, and N. Lao, Big Data Mining and Analytics, December 2019, 2(4): 217–239 Big Data Mining and Analytics, December 2019, 2(4): 217–239
Learning to compose neural networks for question
answering, in Proc. NAACL-HLT, San Diego, CA, USA
2016. 238 Neural symbolic machines:
Learning semantic parsers
on freebase with weak supervision, in Proc. 55th Ann. Meeting of the Association for Computational Linguistics,
Vancouver, Canada, 2017, pp. 23–33. Learning to compose neural networks for question
answering, in Proc. NAACL-HLT, San Diego, CA, USA,
2016. [85] L. R. Tang and R. J. Mooney, Using multiple clause
constructors in inductive logic programming for semantic
parsing, in Proc. 12th European Conf. Machine Learning,
Freiburg, Germany, 2001, pp. 466–477. [72] R. J. Williams,
Simple statistical gradient-following
algorithms for connectionist reinforcement learning, Mach. Learn., vol. 8, nos. 3&4, pp. 229–256, 1992. [73] J. P. Cheng, S. Reddy, V. Saraswat, and M. Lapata,
Learning structured natural language representations for
semantic parsing, in Proc. 55th Ann. Meeting of the
Association for Computational Linguistics, Vancouver,
Canada, 2017, pp. 44–55. [86] W. T. Yih, M. Richardson, C. Meek, M. W. Chang, and J. Suh, The value of semantic parse labeling for knowledge
base question answering, in Proc. 54th Ann. Meeting of
the Association for Computational Linguistics, 2016, pp. 201–206. [74] S. Reddy, M. Lapata, and M. Steedman, Large-scale
semantic parsing without question-answer pairs, Trans. Assoc. Comput. Linguist., vol. 2, pp. 377–392, 2014. [87] Y. Bisk, S. Reddy, J. Blitzer, J. Hockenmaier, and M. Steedman, Evaluating induced CCG parsers on grounded
semantic parsing, in Proc. 2016 Conf. Empirical Methods
in Natural Language Processing, Austin, TX, USA, 2016. [75] S. Reddy, O. T¨ackstr¨om, M. Collins, T. Kwiatkowski,
D. Das, M. Steedman, and M. Lapata, Transforming
dependency structures to logical forms for semantic
parsing, Trans. Assoc. Comput. Linguist., vol. 4, pp. 127–
140, 2016. [88] E. Gabrilovich, M. Ringgaard, and A. Subramanya, Facc1:
Freebase Annotation of ClueWeb Corpora. Google Inc.,
2013. [89] A. Bordes, N. Usunier, S. Chopra, and J. Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey S. Hasan, C. dos Santos, B. Xiang,
and B. W. Zhou, Improved neural relation detection for
knowledge base question answering, in Proc. 55th Ann. Meeting of the Association for Computational Linguistics,
Vancouver, Canada, 2017, pp. 571–581. [95] X. Fan,
E. Monti,
L. Mathias,
and
M. Dreyer,
Transfer learning for neural semantic parsing, in Proc. 2nd Workshop on Representation Learning for NLP,
Vancouver, Canada, 2017. [83] R. Das, M. Zaheer, S. Reddy, and A. McCallum, Question
answering on knowledge bases and text using universal
schema and memory networks,
in Proc. 55th Ann. Meeting of the Association for Computational Linguistics,
Vancouver, Canada, 2017. [96] D. L. Silver, Q. Yang, and L. H. Li, Lifelong machine
learning systems: Beyond learning algorithms, in AAAI
Spring Symp.: Lifelong Machine Learning, Palo Alto, CA,
USA, 2013, p. 5. [84] J. Andreas, M. Rohrbach, T. Darrell, and D. Klein, [97] Y. N. Dauphin, G. T¨ur, D. Hakkani-T¨ur, and L. P. 239 Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A
Heck, Zero-shot learning and clustering for semantic
utterance classification, in Proc. 2nd Int. Conf. Learning
Representations, Banff, Canada, 2014. Heck, Zero-shot learning and clustering for semantic
utterance classification, in Proc. 2nd Int. Conf. Learning
Representations, Banff, Canada, 2014. [99] P. S. Huang, C. L. Wang, R. Singh, W. T. Yih, and X. D. He,
Natural language to structured query generation via meta-
learning, in Proc. NAACL-HLT 2018, New Orleans, LA,
USA, 2018. [98] C. Finn, P. Abbeel, and S. Levine, Model-agnostic meta-
learning for fast adaptation of deep networks, in Proc. 34th
Int. Conf. Machine Learning, Sydney, Australia, 2017, pp. Qile Zhu
is a PhD student in National
Science Foundation (NSF) Center for
Big Learning, Department of Computer
& Information Science & Engineering
(CISE),
Univerisity
of
Florida. His
research interests include natural language
processing
and
understanding,
graph
analysis, and Bayesian methods. Xiyao Ma is a PhD student in NSF
Center for Big Learning, Department of
Electrical
and
Computer
Engineering,
Univerisity of Florida. He is interested in
natural language processing, generative
adversarial
network,
and
adversarial
examples. Qile Zhu
is a PhD student in National
Science Foundation (NSF) Center for
Big Learning, Department of Computer
& Information Science & Engineering
(CISE),
Univerisity
of
Florida. His
research interests include natural language
processing
and
understanding,
graph
analysis, and Bayesian methods. Qile Zhu et al.:
Statistical Learning for Semantic Parsing: A Survey Xiaolin Li is a professor and university
term professor in Department of Electrical
and Computer Engineering (ECE) and
Department of Computer at University of
Florida. He is the founding director of
National Science Foundation Center for
Big Learning (CBL). His research interests
include big data, machine learning, deep
learning, cloud computing, intelligent platforms, HPC, security
& privacy for health, precision medicine, IoT, CV, NLP, robotics,
genomics, and science, engineering, and business. He received
the PhD degree from Rutgers University. He received the
National Science Foundation CAREER Award in 2010, the
Internet2 Innovative Application Award in 2013, NSF I-Corps
Top Team Award in 2015, Top Team (DeepBipolar) in the CAGI
Challenge on detecting bipolar disorder in 2016, and best paper
awards (IEEE ICMLA 2016, IEEE SECON 2016, ACM CAC
2013, and IEEE UbiSafe 2007). Xiyao Ma is a PhD student in NSF
Center for Big Learning, Department of
Electrical
and
Computer
Engineering,
Univerisity of Florida. He is interested in
natural language processing, generative
adversarial
network,
and
adversarial
examples. Xiyao Ma is a PhD student in NSF
Center for Big Learning, Department of
Electrical
and
Computer
Engineering,
Univerisity of Florida. He is interested in
natural language processing, generative
adversarial
network,
and
adversarial
examples.
|
https://openalex.org/W2122231327
|
https://retrovirology.biomedcentral.com/counter/pdf/10.1186/1742-4690-4-85
|
English
| null |
Mortality associated with HIV-1, HIV-2, and HTLV-I single and dual infections in a middle-aged and older population in Guinea-Bissau
|
Retrovirology
| 2,007
|
cc-by
| 7,914
|
Open Acc
Research
Mortality associated with HIV-1, HIV-2, and HTLV-I single and dual
infections in a middle-aged and older population in Guinea-Bissau
Birgitta Holmgren*1,2,3, Zacarias da Silva1, Pernille Vastrup3, Olav Larsen3,
Sören Andersson4, Henrik Ravn3 and Peter Aaby1,3 Address: 1Bandim Health Project, Bissau, INDEPTH Network, Guinea-Bissau, 2Department of Laboratory Medicine, Division of Medical
Microbiology/Virology, Lund University, Lund, Sweden, 3Bandim Health Project, Division of Epidemiology, Statens Serum Institut, Copenhagen,
Denmark and 4Swedish Institute of Infectious Disease Control, Stockholm, Sweden Email: Birgitta Holmgren* - Birgitta_G.Holmgren@med.lu.se; Zacarias da Silva - z.dasilva54@bandim.org; Pernille Vastrup - peva@nykredit.dk;
Olav Larsen - Larsen.chin@dadlnet.dk; Sören Andersson - soren.andersson@smi.ki.se; Henrik Ravn - hjn@ssi.dk;
Peter Aaby - p.aaby@bandim.org * Corresponding author Received: 30 March 2007
Accepted: 27 November 2007 Published: 27 November 2007 Published: 27 November 2007
Retrovirology 2007, 4:85
doi:10.1186/1742-4690-4-85 Retrovirology 2007, 4:85
doi:10.1186/1742-4690-4-85 Retrovirology 2007, 4:85
doi:10.1186/1742-4690-4-85 This article is available from: http://www.retrovirology.com/content/4/1/85 © 2007 Holmgren et al; licensee BioMed Central Ltd. g
;
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. BioMed Central BioMed Central BioMed Central Retrovirology Open Access Background numbers in parenthesis indicating individuals in the
HTLV-I mortality sample. Tables 1 and 2 display the out-
come with regard to HIV/HTLV status and gender. Total
follow-up time was 12,283 and 4,022 years for subjects
included in the HIV and HTLV-I analyses, with a median
follow-up time of 3.8 years (range 0.0–12.4) and 3.0 years
(0.0–10.5) respectively. Median age of participants were
53.2 (interquartile range (IQR) = 45.9–61.0), and 49.4
(IQR = 39.2–58.2), respectively. Guinea-Bissau constitutes a unique area in which the
three major retroviruses HIV-1, HIV-2 and HTLV-I circu-
late in the general population [1-8]. The natural history of
these infections and co-infections is not fully understood
in this area. Previously, we have found that various com-
binations of dual infections of these viruses were more
common in women than in men, a trend particularly
strong for individuals over 45 years of age[4,5,7,8]. Also in
a rural population aged 15 years and over in the same
country, we found a strong association between HIV-2
and HTLV-I in women but not in men [7]. Various factors
could contribute to these age and gender patterns of retro-
viral infections. When behavioural factors were included
in the analyses our observations were not modified [8]. Differential mortality of retrovirus infections between
men and women, compared with retrovirus-negative indi-
viduals, could contribute; if retrovirus-infected men have
higher mortality than retrovirus-infected women, com-
pared with negative individuals, a higher prevalence of
retrovirus infections would be observed among older
women. Few studies have investigated mortality associ-
ated with HIV-1/HIV-2 dual infection [9,10], and it is
unclear whether there are any differences between men
and women. Regarding HIV-1 mortality, there is usually
no difference between men and women [11-14], although
a higher mortality rate for HIV-1-positive men was
observed in a professional cohort in Tanzania [15]. HIV-1
mortality studies from community settings in West Africa
are scarce [10]. For HIV-2 infections no difference in mor-
tality between men and women was observed either [16-
18]. However, none of these studies have examined
whether the male-to-female (M/F) mortality ratio differs
for HIV-infected and uninfected individuals. In addition,
there are few studies addressing HTLV-I-associated mor-
tality in a general population in Africa [4,19]. Participation A total of 2,839 of the 2,944 individuals were eligible for
inclusion in the HIV mortality analyses, and 1,075 of the
1,124 in the HTLV-I analyses. In the HIV analyses, four
were not re-identifiable, 15 were excluded due to confu-
sion as to identity or uncertainty about age, and 86 indi-
viduals left the study before reaching age 35 (Table 1). The
figures for the HTLV-I mortality sample were 1, 11, and 37
individuals (Table 2). A total of 495 (162) individuals
moved and 517 (170) died during the follow-up period, Abstract Background: In Guinea-Bissau HIV-1, HIV-2, and HTLV-I are prevalent in the general population. The natural history of HIV/HTLV-I single and dual infections has not been fully elucidated in this
population. Previous studies have shown that combinations of these infections are more common
in older women than in men. The present study compares mortality associated with HIV-1, HIV-2,
and HTLV-I single and dual infections in individuals over 35 years of age within an urban community-
based cohort in Guinea-Bissau. Results: A total of 2,839 and 1,075 individuals were included in the HIV and HTLV-I mortality
analyses respectively. Compared with HIV-negative individuals, adjusted mortality rate ratios
(MRRs) were 4.9 (95% confidence interval (CI): 2.3, 10.4) for HIV-1, 1.8 (95%CI: 1.5, 2.3) for HIV-
2, and 5.9 (2.4, 14.3) for HIV-1/HIV-2 dual infections. MRR for HTLV-I-positive compared with
HTLV-I-negative individuals was 1.7 (1.1, 2.7). Excluding all HIV-positive individuals from the
analysis, the HTLV-I MRR was 2.3 (1.3, 3.8). The MRR of HTLV-I/HIV-2 dually infected individuals
was 1.7 (0.7, 4.3), compared with HIV/HTLV-I-negative individuals. No statistically significant
differences were found in retrovirus-associated mortality between men and women. Conclusion: HIV-1-associated excess mortality was low compared with community studies from
other parts of Africa, presumably because this population was older and the introduction of HIV-
1 into the community recent. HIV-2 and HTLV-I-associated mortality was 2-fold higher than the
mortality in uninfected individuals. We found no significant differences between the mortality risk
for HIV-2 and HTLV-I single infection, respectively, and HIV-2/HTLV-I dual infection. The higher
prevalence of retroviral dual infections in older women is not explained by differential retrovirus-
associated mortality for men and women. Page 1 of 9
(page number not for citation purposes) Page 1 of 9
(page number not for citation purposes) http://www.retrovirology.com/content/4/1/85 Retrovirology 2007, 4:85 General HIV-associated mortality Sixteen HIV-negative individuals seroconverted during
the follow-up period; 13 to HIV-2, 2 to HIV-1, and 1 dual
HIV-1/HIV-2 infection. Five HIV-2-positive individuals
seroconverted to dual HIV-1/HIV-2 infection. As stated in
the statistical methods these seroconversions were
accounted for in the analyses. Among the 517 deaths, 277 occurred among men (MR/
1000 pyo = 47.9, 95 percent CI: 42.6, 53.9) and 240
among women (MR = 36.9, CI: 32.5, 41.9). Seven of these
deaths occurred among HIV-1-infected (MR = 116, CI:
52.2, 243), 105 deaths among HIV-2-infected (MR = 62.0,
CI: 51.4, 75.3) and five among HIV-1 + 2 dually infected
(MR = 93.6, CI: 39.0, 225). Overall MRRs, adjusted for
current age, current HIV status and sex were 4.9 (2.3, 10.4)
for HIV-1, 1.8 (1.5, 2.3) for HIV-2, and 5.9 (2.4, 14.3) for
HIV-1/HIV-2 dual infections (Table 3). In a sub-analysis, possible confounding variables were
tested such as ethnic group, district, schooling, and work-
ing outside home at the time of entry into the study,
because this information was not available for all individ-
uals. Only working outside home was independently
associated with less risk of dying, the MRR corrected for
age and gender being 0.7 (95 percent CI: 0.5, 1.0). The
information about work was available for 1707 individu-
als. Corrected for age, gender, current HIV-status, and
work, the HIV-1-associated MRR was 4.1 (95 percent CI:
1.9, 8.7), for HIV-2 the MRR was 1.7 (95 percent CI: 1.3,
2.4), and for dual HIV it was 11.7 (95 percent CI: 4.7,
29.2). Hence, we investigated the mortality patterns of HIV-1,
HIV-2, and HTLV-I, single and dual infections as well as
non-retrovirus infections in the capital Bissau. The objec-
tive was also to assess whether the age- and gender-specific
patterns of dual retroviral infections could be related to
mortality patterns. A total of 809 individuals had an HTLV-I status at the time
of entry into the study. Adjusting for HTLV-I status did not
change the estimates. HIV-1 mortality y
For HIV-1 + 2 dual infection, the age adjusted MRR was
4.3 (95 percent CI: 1.1, 17.3) in men and 7.9 (95 percent
CI: 2.5, 24.9) in women, (test of homogeneity, p = 0.51)
(Table 4). Comparing HIV-1-positive with HIV-negative individuals,
we found the age-corrected MRR to be 5.2 (95 percent CI:
1.9, 14.0) in men and 4.7 (95 percent CI: 1.5, 14.6) in
women (test of homogeneity, p = 0.89) (Table 4). When
we considered both HIV-1 single infection and dual infec-
tions as HIV-1 infection, individuals <45 years of age
tended to have stronger excess mortality compared with
HIV-negative individuals than individuals > = 45, though
the differences were not statistically significant (test of
homogeneity, p = 0.41 in men and p = 0.32 in women)
(Table 5). Gender-specific HIV-associated mortality The mortality rate was higher in HIV-negative men than in
women (Table 4), the age-corrected male-to-female (M/F)
MRR being 1.4 (95 percent CI: 1.1, 1.7) in HIV-negative
individuals. Page 2 of 9
(page number not for citation purposes) Page 2 of 9
(page number not for citation purposes) http://www.retrovirology.com/content/4/1/85 Retrovirology 2007, 4:85 Table 1: Outcome and mortality rates with regard to HIV status, irrespective of HTLV-I status. Negative
HIV-1
HIV-2
HIV-1 + HIV-2 dual
Total
Men
Women
Men
Women
Men
Women
Men
Women
All
Previous HIV result from
1987, 1989, 1990, 1992,
1994 or 1996. 1,250
1,260
10
12
156
243
5
8
2,944
Total re-identified
1,247
1,259
10
12
156
243
5
8
2,940
Confusion/age uncertain
1/0
4/0
0/0
0/0
4/1
4/1
0/0
0/0
13/2
Exit study at age <35. 37
43
0
0
2
4
0
0
86
Total participating in
analysis:
1,209
1,212
10
12
149
234
5
8
2,839
Seroconvert during
follow-up time
7
9
2
3
(status after sero-
conversion)
(2 HIV-1,
4 HIV-2 1
HIV-1/2)
(HIV-2)
(HIV-1/2)
(HIV-1/2)
Moved (%*)
198 (16.5)
212 (17.6)
4 (33.3)
1 (8.3)
23 (15.2)
51(21.3)
4 (50)
2 (18.2)
495 (17.4)
Died (%*)
229 (19.1)
171 (14.2)
4 (33.3)
3 (25)
42 (27.8)
63 (26.3)
2 (25)
3 (27.3)
517 (18.2)
Median follow-up time in
years
3.8
3.9
2.5
3.1
3.8
3.7
3.2
2.6
3.8
(range)
(0.0–12.1)
(0.0–12.4)
(0.2–4.3)
(0.4–3.8)
(0.1–11.6)
(0.1–12.0)
(0.3–5.9)
(0.6,4.5)
(0.0–12.4)
Total follow-up time,
years
5084
5398
26.9
33.6
648
1039
23.7
29.7
12283
Mortality rate/1000 pyo
45.0
31.7
149
89.3
64.8
60.6
84.3
101
42.1
(95% CI†)
(39.5,51.3)
(27.3, 36.8)
(55.9, 397)
(28.8, 278)
(47.9, 87.7)
(47.3,77.6)
(21.1, 337)
(32.6, 313)
(38.6, 45.9)
Individuals in general urban population cohorts in Guinea-Bissau with an HIV result from any previous survey during 1987–1996 were included in
the study. The table shows the outcome of the eligible individuals according to HIV status. Some individuals changed HIV status during follow-up,
from HIV-negative to HIV-1 and/or HIV-2-positive, or from HIV-2 to HIV-1/HIV-2 dually positive. This is shown in the middle section of the table. *% of total number participating according to current HIV status. †CI = confidence interval Table 1: Outcome and mortality rates with regard to HIV status, irrespective of HTLV-I status. Individuals in general urban population cohorts in Guinea-Bissau with an HIV result from any previous survey during 1987–1996 were included in
the study. HTLV-I mortality One individual seroconverted during follow-up. Twenty-
two HTLV-I-positive (MR = 69.7, CI: 45.9, 106) and 148
HTLV-I-negative individuals (MR = 39.9, cent CI: 34.0,
46.9) died. Overall age- and sex-adjusted HTLV-I MRR,
comparing HTLV-I-positive with HTLV-I-negative individ-
uals, was 1.7 (1.1, 2.7). The age-adjusted MRRs were 1.9
(CI: 0.9, 4.0) and 1.6 (CI: 0.9, 2.8) for men and women
respectively (test of homogeneity, p = 0.78) (Table 6). Adjusting for possible confounding variables such as
schooling, working outside home, ethnic group, or district
did not change the estimates. When all HIV-positive indi-
viduals were excluded from the analysis, the MRR for
HTLV-I-positive compared with HTLV-I negative individ-
uals was 2.3 (1.3, 3.8) (Table 3). Gender-specific HIV-associated mortality The table shows the outcome of the eligible individuals according to HIV status. Some individuals changed HIV status during follow-up,
from HIV-negative to HIV-1 and/or HIV-2-positive, or from HIV-2 to HIV-1/HIV-2 dually positive. This is shown in the middle section of the table. *% of total number participating according to current HIV status. †CI = confidence interval HIV-1 + HIV-2 dual mortality HIV-1 + HIV-2 dual mortality Discussion Discussion
To our knowledge this is the first study to report compar-
ative mortality rates of HIV-1, HIV-2, and HTLV-I single
and dual infections within the same population in a com-
munity setting in Africa. The population was antiretroviral
treatment naïve at the time of the study; thus these data
reflect the natural development of these infections. Gener-
ally, the HIV-associated mortality was higher compared
with the mortality in non-infected individuals, regardless
of the type of HIV infection and of HTLV-I status. The HIV-
1-associated mortality was 4 to 5-fold increased compared
with the mortality in HIV-negative individuals. This figure
is lower than figures reported from community studies in
other parts of Africa, where 10–20 times higher risks are
frequently found [12,13,20-22], and could be explained
by the differences in the age distribution between our
sample and studies from other areas including mainly
younger adults [12,13,20,21]. Reports on HIV-1-associ-
ated mortality from general populations in West Africa are
scarce [10]. The fact that HIV-1 was introduced recently
into Guinea-Bissau and that the HIV-1 epidemic presum-
ably is still in its early phase could also contribute to the
lower relative mortality compared with the more high
prevalent areas. Another possible explanation could be
variations in disease progression between HIV-1 subtypes
as reported from Senegal [23]. In West Africa, subtype A
seems to be most prevalent [24-26], and studies have indi-
cated that an A/G recombinant form is the most common
in West Africa [26-28], in particular in Guinea-Bissau [28]. We compared mortality for co-infected individuals with
mortality for single infected individuals. Among 1,069
individuals with concurrent HIV and HTLV-I results, the
MRR for individuals infected only with HIV-1 was 8.9 (95
percent CI: 2.1, 38.6), for HIV-2 single infected it was 2.9
(95 percent CI: 2.0, 4.2), and for HTLV-I single infected
2.3 (95 percent CI: 1.3, 3.8). For HIV-1/HIV-2 dual infec-
tions the MRR was 7.0 (95 percent CI: 1.7, 28.7), for HIV-
2/HTLV-I dual infections it was 1.7 (95 percent CI: 0.7,
4.3), and for HIV-1/HIV-2/HTLV-I triple infections it was
6.7 (95 percent CI: 0.9, 50.1). All estimates were com-
pared with HIV and HTLV-I-negative individuals. There
was no significant difference between MRRs for HIV-2 sin-
gle infection, HTLV-I single infection and HIV-2/HTLV-1
dual infection. HIV-2 mortality For HIV-2 the age-adjusted MRR was 1.6 (95 percent CI:
1.1, 2.2) in men and 2.1 (95 percent CI: 1.6, 2.8) in
women, test of homogeneity, p = 0.24 (Table 4). Women
<45 had higher HIV-2 associated excess mortality than
women > = 45 (test of homogeneity, p < 0.005). Though
the pattern was the same for men the difference was not
statistically significant (test of homogeneity, p = 0.18)
(Table 5). Page 3 of 9
(page number not for citation purposes) http://www.retrovirology.com/content/4/1/85 Retrovirology 2007, 4:85 Table 2: Outcome and mortality rates with regard to HTLV status, irrespective of HIV-status. Individuals from the same cohorts as in
table 1 are included, but with an HTLV-1 result from any previous survey Table 2: Outcome and mortality rates with regard to HTLV status, irrespective of HIV-status. Individuals from the same cohorts as in
table 1 are included, but with an HTLV-1 result from any previous survey
Negative
HTLV-I
Total
Men
Women
Men
Women
All
Previous HTLV result from 1990 or 1996. 497
550
20
57
1,124
Total re-identified with a previous HTLV result
496
550
20
57
1,123
Confusion/age uncertain
4/0
5/0
0/0
2/0
11
Exit study at age <35. 21
16
0
0
37
Total participating in analysis:
471
529
20
55
1,075
Moved (%*)
75 (15.9)
80(15.1)
3 (15%)
4 (7.3)
162 (15.1)
Died (%*)
79 (16.8)
69 (13.0)
7 (35%)
15 (27.2)
170 (15.8)
Median follow-up time in years (range)
2.9 (0.0–9.1)
2.9 (0.0–10.5)
3.0 (0.8–7.8)
3.4 (0.8–9.4)
3.0 (0.0–10.5)
Total follow-up time in years
1703
2004
73.4
242
4022
Mortality rate/1000 pyo (95%CI†)
46.4 (37.2, 57.8)
34.4 (27.2 43.6)
95.4 (45.5, 200)
61.9 (37.8, 103)
42.3 (36.4, 49.1)
*% of total number participating. † CI = confidence interval rates with regard to HTLV status, irrespective of HIV-status. Individuals from the same cohorts as in
HTLV-1 result from any previous survey Discussion # Adjusted for age and current HIV status in a Poisson regression, see statistical methods Table 4: Mortality rates (MRs) and mortality rate ratios (MRRs) for HIV, according to age and sex, irrespective of HTLV-I status particularly strong when excluding the HIV-positives from
the analysis. A previous follow-up of individuals aged ≥
50 years in this population did not identify any significant
effect of HTLV-I on survival [4]. A larger number of partic-
ipants and a lower mean age in the present study may
explain this difference. In the rural area of Guinea-Bissau,
a 7-year follow-up of 285 HIV-2-positive and HIV-2-nega-
tive individuals found no increased mortality for HTLV-I. However, mortality was associated with increased HTLV-I
proviral load [19]. Theoretically contracting HIV during reflect that the clinical outcome of HIV-1 and HIV-2 dual
infection may resemble HIV-1 more than HIV-2 [29-32]. However, there were few dually infected in the study, and
larger studies are needed to determine whether these
trends are reproducible. Few other community studies
have addressed mortality related to HIV-1 and HIV-2 dual
infection [10]. HTLV-I-associated mortality was about 2-fold increased
compared with HTLV-I-negative mortality. The effect was ality rate ratios for HIV, irrespective of HTLV-I status. MRs and MRRs below and above 45 years of Table 5: Mortality rates and mortality rate ratios for HIV, irrespective of HTLV-I status. MRs and MRRs below and above 45 years of
age are compared
HIV-1 and HIV-1 + HIV-2 dually
infected*
HIV-2 single infected†
Age group
Mortality rate/1000
pyo
MRR‡
95% CI§
Mortality rate/
1000 pyo
MRR‡
95% CI§
MEN
<45
45.1(6.4–320)
12.0#
1.4,99.7
17.9(4.5,71.6)
4.8#
1.0,23.6
> = 45
176(73.2–422)
4.6
1.9,11.1
74.6(54.7,102)
1.5
1.1, 2.2
Test of homogeneity of effect modification of age
p = 0.41
p = 0.18
WOMEN
<45
46.4(6.5–329)
17.1#
2.1,142
33.4(15.0,74.3)
12.3#
4.0,38.9
> = 45
120(49.8–288)
5.4
2.2,13.1
66.3(51.1,85.9)
1.9
1.4,2.6
Test of homogeneity of effect modification of age
p = 0.32
p < 0.005
*Including both HIV-1 single and HIV-1 + HIV-2 dual infections. †Including only HIV-2 single infections. ‡ MRR = Mortality rate ratio, comparing with HIV-negative. § CI = confidence interval. #The MRRs are different from those in Table 4 as the HIV-1 single and HIV-1 + HIV-2 dually infected have been grouped together in this Poisson
model (see statistical methods) § CI = confidence interval. Discussion Table 3: All retrovirus mortality rates and mortality rate ratios
MR* (95% CI†)
MRR‡(95% CI)
HIV-1
116 (52.2, 243)
4.9§ (2.3, 10.4)
HIV-2
62.0 (51.4, 75.3)
1.8§ (1.5, 2.3)
HIV-1/2 dual
93.6 (39.0, 225)
5.9§ (2.4, 14.3)
HTLV-I all
69.7 (45.9, 106)
1.7# (1.1, 2.7)
HTLV I positive/HIV- negative
77.4 (47.4, 126)
2.3&(1.3, 3.8)
* MR = mortality rate per 1000 person years of observation. † CI = confidence interval. ‡ MRR = Mortality rate ratio. §Comparing with HIV-negative. # Comparing with HTLV-I-negative. & Comparing with HIV and HTLV-I-negative Table 3: All retrovirus mortality rates and mortality rate ratios For HIV-2 we found 1.5 to 2-fold increased mortality
rates. This is comparable with previous observations from
the same urban [17] and rural general populations [16]. The MRR for dual HIV-1/HIV-2 infection in our study was
more equivalent to the risk we found for HIV-1 and could Page 4 of 9
(page number not for citation purposes) http://www.retrovirology.com/content/4/1/85 Retrovirology 2007, 4:85 Table 4: Mortality rates (MRs) and mortality rate ratios (MRRs) for HIV, according to age and sex, irrespective of HTLV-I status
HIV-negative
HIV-1
HIV-2
HIV-1 + HIV-2
Age
group*
MR†(deaths/
pyo)
MR†(deaths/
pyo)
MRR‡
95% CI§
MR†(deaths/
pyo)
MRR‡
95%CI§
MR†(deaths/
pyo)
MRR‡
95%CI§
MEN
35–44
4.7(4/844)
100(1/10)
21.1
2.4,189
17.9(2/112)
3.8
0.7,20.6
0(0/12.1)
45–54
20.9(28/1342)
235(2/8.5)
10.3
2.7,47.3
46.9(9/192)
2.2
1.1,4.8
320(2/6.3)
15.3
3.6,64.3
55–64
45.5 (81/1781)
0 (0/4.6)
81.7(18/220)
1.8
1.1,3.0
0 (0/5.3)
65–74
88.2 (65/737)
268(1/3.7)
3.0
0.4,21.9
124 (13/104)
1.4
0.8,2.6
0 (0/0)
75+
134 (51/380)
0 (0/0)
0(0/19.9)
0 (0/0)
Total:
45.0(229/5084)
148.8(4/26.9)
5.2#
1.9,14.0
64.8(42/648)
1.6#
1.1, 2.2
84.3 (2/23.7)
4.3#
1.1,17.3
WOMEN
35–44
1.9 (2/1040)
0(0/11.3)
33.3(6/180)
17.4
3.5,86.0
97.7(1/10.2)
50.8
4.6,560
45–54
13.3 (18/1346)
227(2/8.8)
17.0
3.9,73.1
19.3(6/311)
1.4
0.6,3.6
57.4(1/17.4)
4.3
0.6,32.1
55–64
31.7 (57/1798)
74.3(1/13.5)
2.3
0.3,16.9
91.8(33/359)
2.9
1.9,4.4
485(1/2.1)
15.3
2.1,110
65–74
63.8 (55/862)
0(0/0)
73.9(11/149)
1.2
0.6,2.2
0 (0/0)
75+
110 (39/352)
0(0/0)
174(7/40.2)
1.6
0.7,3.5
0 (0/0)
Total:
31.7(171/5398)
89.3 (3/33.6)
4.7#
1.5,14.6
60.6(63/1040)
2.1#
1.6,2.8
101(3/29.7)
7.9#
2.5,24.9
*Indicates current age, i.e. individuals can contribute to several age groups. † MR = mortality rate per 1000 person years of observation. ‡ MRR = Mortality rate ratio, comparing with HIV-negative. §CI = confidence interval. Discussion #The MRRs are different from those in Table 4 as the HIV-1 single and HIV-1 + HIV-2 dually infected have been grouped together in this Poisson
model (see statistical methods) Page 5 of 9
(page number not for citation purposes) Retrovirology 2007, 4:85 http://www.retrovirology.com/content/4/1/85 Table 6: Mortality rates (MRs) and mortality rate ratios (MRRs) for HTLV-I according to age and sex, irrespective of HIV status
HTLV-I-negative
HTLV-I-positive
Age group*
MR†(deaths/pyo)
MR†(deaths/pyo)
MRR‡
95%CI§
MEN
35–44
7.0(3/426)
88.2 (1/11.3)
12.5
1.3, 120.3
45–54
29.7(11/371)
0 0/16.0)
55–64
43.6(20/458)
0 (0/26.5)
65–74
78.3(25/319)
396(5/12.6)
5.1
1.9,13.2
75+
155(20/129)
144(1/6.9)
0.9
0.1,6.9
Total:
46.4(79/1703)
95.4(7/73.4)
1.9 #
0.9,4.0
WOMEN
35–44
5.5 (3/545)
99.9 (2/20.0)
18.2
3.0,109
45–54
13.4(5/372)
0 (0/67.0)
55–64
45.9(28/610)
88.2(9/102)
1.9
0.9,4.1
65–74
64.8(23/355)
54.2(2/36.9)
0.8
0.2,3.6
75+
82.3 (10/122)
121(2/16.5)
1.5
0.3, 6.7
Total:
34.0(69/2004)
61.9(15/242)
1.6 #
0.9,2.8
*Indicates current age, i.e. individuals can contribute to several age groups. † MR = mortality rate per 1000 person years of observation. ‡ MRR = Mortality rate ratio, comparing with HTLV-negative. §CI = confidence interval. # Adjusted for age in a Poisson regression, see statistical methods Table 6: Mortality rates (MRs) and mortality rate ratios (MRRs) for HTLV-I according to age and sex, irrespective of HIV status prevalence increases with age in women, while it
decreases with age in men [8]. In the present study we
could not identify any gender differences in HIV and
HTLV-I-associated mortality explaining our previous
observation. However, according to Table 1 more HIV-1
and dual HIV-1/HIV-2 positive-men than women moved
during follow-up compared with HIV-negative individu-
als. This was also seen for HTLV-I-positive men. The indi-
viduals who moved were censored on the date of
migration. Thus informative censoring could occur in our
analyses if the reason for having migrated was associated
with a higher or lower risk of dying. The HIV-1-positive
men who moved were younger than those who did not
move (data not shown) so this could potentially bias our
results. However, if higher HIV-associated mortality in
men would explain the higher prevalence of dual infec-
tions among women, this may rather be due to HIV-2 and
dual infection-related mortality, viewing the epidemio-
logical features of the viruses in this population. Page 6 of 9
(page number not for citation purposes) http://www.retrovirology.com/content/4/1/85 http://www.retrovirology.com/content/4/1/85 viduals. A comparison between age groups is, however,
difficult to interpret as we have no information on when
the individuals contracted their retrovirus infections. But
also in previous follow-up studies of HIV-2 [16,17] HIV-
2-associated mortality was found to be higher among
younger individuals. 1994. Laboratory methods have been described elsewhere
[2-4,17,52,53]. Subjects from both cohorts aged 35 years
and over at the beginning of 1998 were included in a sur-
vey of HIV and HTLV infections in 1998–2000. The rea-
son for this was to assess the dynamics of retroviruses in
older age groups, as previous studies from the area had
suggested a higher susceptibility to HIV-2 in women older
than 45 years of age [17,54]. The study included 3,560
subjects [8]. All individuals who had an HIV or HTLV sta-
tus from any of the previous surveys during 1987–1996
(N = 2,944 and N = 1,124 for HIV and HTLV respectively)
were enrolled in the present mortality analyses (Tables 1
&2). 1994. Laboratory methods have been described elsewhere
[2-4,17,52,53]. Subjects from both cohorts aged 35 years
and over at the beginning of 1998 were included in a sur-
vey of HIV and HTLV infections in 1998–2000. The rea-
son for this was to assess the dynamics of retroviruses in
older age groups, as previous studies from the area had
suggested a higher susceptibility to HIV-2 in women older
than 45 years of age [17,54]. The study included 3,560
subjects [8]. All individuals who had an HIV or HTLV sta-
tus from any of the previous surveys during 1987–1996
(N = 2,944 and N = 1,124 for HIV and HTLV respectively)
were enrolled in the present mortality analyses (Tables 1
&2). The present data indicate that there is no differential mor-
tality between men and women explaining the higher
prevalence of dual retroviral infections in older women,
nor could it be explained by behavioural factors [7,8]. This might provide further indirect support for the
hypothesis of an enhanced biological susceptibility to ret-
roviral infections of older women compared with men
[45-47]. The major change around 45 years of age corre-
sponds to the pre-menopausal and menopausal age. Alter-
ations in female hormones and the associated histological
changes could induce changes in the immunological
defence of the genital tract (reviewed in [48]). http://www.retrovirology.com/content/4/1/85 In animal
models of HIV infection it has been demonstrated that
susceptibility to SIV is related to hormonal status [49-51]. Further epidemiological studies on incidence patterns as
well as biological studies are warranted to determine the
generality and mechanisms of this trend towards
increased HIV prevalence among older women. This trend
could have major public health implications with the
longer survival of HIV-infected individuals, and older
women might constitute a vulnerable group to be empha-
sized in future HIV/AIDS prevention control programmes. Date of entry into the present study was the date corre-
sponding to the date of first participation ever in a screen-
ing. Information on vital status was obtained during the
study performed in 1998–2000. If a participant had died
or moved, a family member or a neighbour provided this
information. Date of exit from the study ranged from
1987 to 2000. Statistical methods Mortality rates (MRs) were defined as the number of
deaths per 1000 person years of observation (pyo). Date
of exit for survivors was defined as the date during the fol-
low-up study in 1998–2000 at which the subject was
encountered alive. For those who had moved or died, fol-
low-up time was censored at the midpoint of the month
when only the month for migration/death was given, and
at the midpoint of the year when only the year was given. If the date of migration or death was missing (N = 156),
follow-up time was censored at the midpoint between
date of entry into the study and the date on which the
information about migration or death was obtained,
because more specific information was not available. Poisson regression was used in the mortality analyses giv-
ing mortality rate ratios (MRRs) and their 95 percent CIs,
controlling for current age during follow-up (35–44,
45–54, 55–64, 65–74, 75+ years) and current HIV status
[55]. This was done for HIV-1, HIV-2, and HIV-1 + HIV-2
dual infections in the same model. Date of first participa-
tion in any screening or date of turning 35 years old, was
used as date of entry in analyses, which ever came last. To
account for possible seroconversion during the follow-up
period, the dataset records for sero-converters were split
into episodes before and after seroconversion, i.e. at date
of first positive sample. To test if men and women had
equal mortality risk associated with retrovirus infections,
we tested the interaction between gender and infection in
the Poisson regression, i.e a homogeneity test. To assess
whether age had an effect on the HIV-related excess mor-
tality, we performed a test of homogeneity with age
divided into below and over 45 years of age. In this anal-
ysis the HIV-1 infection group includes both HIV-1 and Conclusion HIV-1-associated mortality was 4–5-fold, and HIV-2-asso-
ciated mortality was 2-fold higher than HIV-negative mor-
tality. The mortality risk for HIV-1/HIV-2 dual infection
was equivalent to the risk for HIV-1. HTLV-I MRR was
2.3(1.3–3.8)
when
HIV-positive
individuals
were
excluded from the analysis. There was no difference
between men and women explaining the previous finding
of higher prevalence of dual infections in older women. Discussion HIV-2
has been endemic in the country for decades, while HIV-
1 was introduced more recently, as illustrated by inci-
dence data which did not show any sero-conversion to
HIV-2 in individuals already infected with HIV-1 or HTLV-
I [8]. For HIV-2 there was no difference in loss to follow-
up between men and women that could bias the results
with regard to HIV-2-associated mortality. The additional
risk of dying associated with HIV was higher in younger
individuals than in older. Though only statistically signif-
icant for HIV-2 in women, the trend was seen for both
men and women regarding both HIV-1 and HIV-2, prob-
ably reflecting higher background mortality in older indi- follow-up could contribute to higher mortality if the
course of a newly acquired HIV infection were rapid. The
few HIV seroconversions during follow-up that we know
of were adjusted for in the analyses. The mortality risk in
the present study is slightly higher than the 1.5–2 fold
increased HTLV-I mortality found in high prevalent areas
such as southern Japan [33-35]. Diseases usually associ-
ated with HTLV-I infection are Tropical spastic parapare-
sis/HTLV-I-associated myelopathy (TSP/HAM) [36,37]
and adult T-cell leukemia (ATL) [38,39] and occur in 3–4
percent of HTLV-I-infected individuals [40-42]. There are
no data regarding the prevalence and incidence of these
diseases from the area but it seems unlikely that these
would explain all the excess mortality. The high incidence
of Tuberculosis (Tb) in this community [43] might con-
tribute to this increased HTLV-I-associated mortality. Increased prevalence of HTLV-I in Tb patients compared
with healthy individuals has been reported from Brazil
[44]. It could also reflect other causes of disease and death
that have so far not been linked to HTLV-I. In the subset
of data where concurrent HIV and HTLV-I result was avail-
able, we found no significant difference in the mortality
risk for HIV-2/HTLV-I dual infection and the mortality
risk for HIV-2 and HTLV-I single infection respectively. Further follow-up studies of HTLV-I single and HTLV-I/
HIV dually infected individuals are needed, as are studies
identifying diseases associated with HTLV-I infection in
this area. We have previously found that dual retroviral infections
are more common in women than in men, and that the Page 6 of 9
(page number not for citation purposes) Page 6 of 9
(page number not for citation purposes) http://www.retrovirology.com/content/4/1/85 Retrovirology 2007, 4:85 Retrovirology 2007, 4:85 Competing interests y
12. Todd J, Balira R, Grosskurth H, Mayaud P, Mosha F, ka-Gina G,
Klokke A, Gabone R, Gavyole A, Mabey D, Hayes R: HIV-associ-
ated adult mortality in a rural Tanzanian population. AIDS
1997, 11:801-807. The author(s) declare that they have no competing inter-
ests. 13. Nunn AJ, Mulder DW, Kamali A, Ruberantwari A, Kengeya-Kayondo
JF, Whitworth J: Mortality associated with HIV-1 infection over
five years in a rural Ugandan population: cohort study. BMJ
1997, 315:767-771. Authors' contributions The study was planned by BH and PA and executed by BH
and ZS. OL and PA carried out the previous cohort studies,
and SA was responsible for the laboratory test strategies. PV and HR were responsible for the statistical analyses. BH carried out the data management and data analyses
and wrote the first draft. All authors contributed with
interpretation of the data and to the final version of the
paper. 14. Crampin AC, Floyd S, Glynn JR, Sibande F, Mulawa D, Nyondo A,
Broadbent P, Bliss L, Ngwira B, Fine PE: Long-term follow-up of
HIV-positive and HIV-negative individuals in rural Malawi. AIDS 2002, 16:1545-1550. 15. Senkoro KP, Boerma JT, Klokke AH, Ng'weshemi JZ, Muro AS,
Gabone R, Borgdorff MW: HIV incidence and HIV-associated
mortality in a cohort of factory workers and their spouses in
Tanzania, 1991 through 1996. J Acquir Immune Defic Syndr 2000,
23:194-202. 16. Ricard D, Wilkins A, N'Gum PT, Hayes R, Morgan G, Da Silva AP,
Whittle H: The effects of HIV-2 infection in a rural area of
Guinea-Bissau. AIDS 1994, 8:977-982. References 1. Wilkins A, Ricard D, Todd J, Whittle H, Dias F, Paulo DS: The epi-
demiology of HIV infection in a rural area of Guinea-Bissau. AIDS 1993, 7:1119-1122. 1. Wilkins A, Ricard D, Todd J, Whittle H, Dias F, Paulo DS: The epi-
demiology of HIV infection in a rural area of Guinea-Bissau. AIDS 1993, 7:1119-1122. 2. Poulsen AG, Aaby P, Gottschau A, Kvinesdal BB, Dias F, Molbak K,
Lauritzen E: HIV-2 infection in Bissau, West Africa, 1987-1989:
incidence, prevalences, and routes of transmission. J Acquir
Immune Defic Syndr 1993, 6:941-948. 20. Mulder DW, Nunn AJ, Kamali A, Nakiyingi J, Wagner HU, Kengeya-
Kayondo JF: Two-year HIV-1-associated mortality in a Ugan-
dan rural population. Lancet 1994, 343:1021-1023. f
y
3. Larsen O, da Silva Z, Sandstrom A, Andersen PK, Andersson S,
Poulsen AG, Melbye M, Dias F, Naucler A, Aaby P: Declining HIV-
2 prevalence and incidence among men in a community
study from Guinea-Bissau. AIDS 1998, 12:1707-1714. 21. Mulder DW, Nunn AJ, Wagner HU, Kamali A, Kengeya-Kayondo JF:
HIV-1 incidence and HIV-1-associated mortality in a rural
Ugandan population cohort. AIDS 1994, 8:87-92. 22. Borgdorff MW, Barongo LR, Klokke AH, Newell JN, Senkoro KP,
Velema JP, Gabone RM: HIV-1 incidence and HIV-1 associated
mortality in a cohort of urban factory workers in Tanzania. Genitourin Med 1995, 71:212-215. y
4. Melbye M, Poulsen AG, Gallo D, Pedersen JB, Biggar RJ, Larsen O,
Dias F, Aaby P: HTLV-1 infection in a population-based cohort
of older persons in Guinea-Bissau, West Africa: risk factors
and impact on survival. Int J Cancer 1998, 76:293-298. 23. Kanki PJ, Hamel DJ, Sankale JL, Hsieh C, Thior I, Barin F, Woodcock
SA, Gueye-Ndiaye A, Zhang E, Montano M, Siby T, Marlink R, Ndoye
I, Essex ME, Mboup S: Human immunodeficiency virus type 1
subtypes differ in disease progression. J Infect Dis 1999,
179:68-73. p
J
5. Larsen O, Andersson S, da Silva Z, Hedegaard K, Sandstrom A, Nau-
cler A, Dias F, Melbye M, Aaby P: Prevalences of HTLV-1 infec-
tion and associated risk determinants in an urban population
in Guinea-Bissau, West Africa. J Acquir Immune Defic Syndr 2000,
25:157-163. p
J
5. Larsen O, Andersson S, da Silva Z, Hedegaard K, Sandstrom A, Nau-
cler A, Dias F, Melbye M, Aaby P: Prevalences of HTLV-1 infec-
tion and associated risk determinants in an urban population
in Guinea-Bissau, West Africa. J Acquir Immune Defic Syndr 2000,
25:157-163. 24. Acknowledgements This work was supported by Council of Development Research, Ministry
of Foreign Affairs, Denmark; Novo Nordisk Fonden, Denmark; Copenha-
gen University, Denmark; the Swedish International Development Cooper-
ation Agency (SIDA), Department of Research Cooperation (SAREC). We
thank field assistants, laboratory and office staff at the study site for making
this study possible, and Dr Eva Maria Fenyö for helpful comments on the
manuscript. 17. Poulsen AG, Aaby P, Larsen O, Jensen H, Naucler A, Lisse IM, Chris-
tiansen CB, Dias F, Melbye M: 9-year HIV-2-associated mortality
in an urban community in Bissau, west Africa. Lancet 1997,
349:911-914. 18. Norrgren H, da Silva ZJ, Andersson S, Biague AJ, Dias F, Biberfeld G,
Naucler A: Clinical features, immunological changes and mor-
tality in a cohort of HIV-2-infected individuals in Bissau,
Guinea-Bissau. Scand J Infect Dis 1998, 30:323-329. 19. Ariyoshi K, Berry N, Cham F, Jaffar S, Schim L, Jobe O, N'Gom PT,
Larsen O, Andersson S, Aaby P, Whittle H: Quantification of
Human T-lymphotropic virus type I (HTLV-I) provirus load
in a rural West African population: no enhancement of
human immunodeficiency virus type 2 pathogenesis, but
HTLV-I provirus load relates to mortality. J Infect Dis 2003,
188:1648-1651. Ethical considerations A protocol of the study was approved by The Central Sci-
entific-Ethical Committee of Denmark and The Ministry
of Public Health in Guinea-Bissau. Informed verbal con-
sent was obtained from every subject prior to interview
and blood sample. Antiretroviral treatment was not avail-
able in the country at the time of the study. Participants
had access to free medical consultation and basic treat-
ments, HIV counselling and information and free con-
doms. 9. Norrgren H, Bamba S, da Silva ZJ, Andersson S, Koivula T, Biberfeld
G: High mortality and severe immunosuppression in hospi-
talized patients with pulmonary tuberculosis and HIV-2
infection in Guinea-Bissau. Scand J Infect Dis 2001, 33:450-456. 10. Seng R, Gustafson P, Gomez VF, Vieira CS, Teixiera M, Rabna P, Jalo
M, Johansson P, Larsen O, Sandstrom A, Larouze B, Dias F, Norberg
R, Murray JF, Aaby P, Lisse IM, Naucler A, Samb B: Community
study of the relative impact of HIV-1 and HIV-2 on intratho-
racic tuberculosis. AIDS 2002, 16:1059-1066. 11. Sewankambo NK, Wawer MJ, Gray RH, Serwadda D, Li C, Stallings
RY, Musgrave SD, Konde-Lule J: Demographic impact of HIV
infection in rural Rakai district, Uganda: results of a popula-
tion-based cohort study. AIDS 1994, 8:1707-1713. http://www.retrovirology.com/content/4/1/85 Retrovirology 2007, 4:85 HIV-1 + HIV-2 dual infections since there were too few
HIV-1 cases to permit further division into age groups. Analyses were performed in Stata version 8. 7. Holmgren B, Andersson S, Harding E, van der Loeff MS, Vastrup P,
Aaby P, Ariyoshi K, Whittle H: Increased prevalence of HTLV-1
among HIV-2-infected women but not HIV-2-infected men
in rural Guinea-Bissau. J Acquir Immune Defic Syndr 2002,
30:342-350. 8. 8. Holmgren B, da Silva Z, Larsen O, Vastrup P, Andersson S, Aaby P:
Dual infections with HIV-1, HIV-2 and HTLV-I are more
common in older women than in men in Guinea-Bissau. AIDS
2003, 17:241-253. Study area and population The study area is divided into three adjacent districts,
Bandim 1, Bandim 2, and Belem. A community-based
cohort, established in 1987 [2,17,52] and including indi-
viduals >15 years of age from 100 randomly selected
houses, was extended in 1995 to encompass an additional
212 randomly selected houses [3]. Surveys in this cohort
were performed in 1987, 1989, 1992, and 1996
[2,3,17,52]. A second cohort, established in 1990,
included all individuals aged 50 years and over resident in
Bandim 1 and 2 based on a general population census in
1986–1987 and was extended in 1994 to encompass all
individuals aged ≥ 50 years from the entire study area
[4,53]. Surveys in this cohort were performed in 1990 and Page 7 of 9
(page number not for citation purposes) Page 7 of 9
(page number not for citation purposes) References Janssens W, Buve A, Nkengasong JN: The puzzle of HIV-1 sub-
types in Africa. AIDS 1997, 11:705-712. 6. van der Loeff MF, Aaby P, Aryioshi K, Vincent T, Awasana AA, Da
Costa C, Pembrey L, Dias F, Harding E, Weiss HA, Whittle HC: HIV-
2 does not protect against HIV-1 infection in a rural commu-
nity in Guinea-Bissau. AIDS 2001, 15:2303-2310. 25. Brandful JA, Ampofo WK, Janssens W, Adu-Sarkodie Y, Apeagyei F,
Anyomi F, Aidoo S, Yamamoto N, Ishikawa K, Sata T, Kurata T: Page 8 of 9
(page number not for citation purposes) http://www.retrovirology.com/content/4/1/85 Retrovirology 2007, 4:85 43. Genetic and phylogenetic analysis of HIV type 1 strains from
southern Ghana. AIDS Res Hum Retroviruses 1998, 14:815-819. 43. Gustafson P, Gomes VF, Vieira CS, Rabna P, Seng R, Johansson P,
Sandstrom A, Norberg R, Lisse I, Samb B, Aaby P, Naucler A: Tuber-
culosis in Bissau: incidence and risk factors in an urban com-
munity in sub-Saharan Africa. Int J Epidemiol 2004, 33:163-172. Genetic and phylogenetic analysis of HIV type 1 strains from
southern Ghana. AIDS Res Hum Retroviruses 1998, 14:815-819. 26. Ellenberger DL, Pieniazek D, Nkengasong J, Luo CC, Devare S, Mau-
rice C, Janini M, Ramos A, Fridlund C, Hu DJ, Coulibaly IM, Ekpini E,
Wiktor SZ, Greenberg AE, Schochetman G, Rayfield MA: Genetic
analysis of human immunodeficiency virus in Abidjan, Ivory
Coast reveals predominance of HIV type 1 subtype A and
introduction of subtype G. AIDS Res Hum Retroviruses 1999,
15:3-9. 44. J p
44. Marinho J, Galvao-Castro B, Rodrigues LC, Barreto ML: Increased
risk of tuberculosis with human T-lymphotropic virus-1
infection: a case-control study. J Acquir Immune Defic Syndr 2005,
40:625-628. 45. Stuver SO, Tachibana N, Okayama A, Shioiri S, Tsunetoshi Y, Tsuda
K, Mueller NE: Heterosexual transmission of human T cell
leukemia/lymphoma virus type I among married couples in
southwestern Japan: an initial report from the Miyazaki
Cohort Study. J Infect Dis 1993, 167:57-65. 27. Takehisa J, Zekeng L, Ido E, Mboudjeka I, Moriyama H, Miura T,
Yamashita M, Gurtler LG, Hayami M, Kaptue L: Various types of
HIV mixed infections in Cameroon. Virology 1998, 245:1-10. 28
A d
S N
H D
F B b
f ld G Alb
J M l
l 28. Andersson S, Norrgren H, Dias F, Biberfeld G, Albert J: Molecular
characterization of human immunodeficiency virus (HIV)-1
and - 2 in individuals from Guinea-Bissau with single or dual
infections: predominance of a distinct HIV-1 subtype A/G
recombinant in West Africa. Virology 1999, 262:312-320. y
J
f
46. Comparison of female to male and male to female transmis-
sion of HIV in 563 stable couples. European Study Group on
Heterosexual Transmission of HIV. BMJ 1992, 304:809-813. J
47. Holmgren B, Aaby P, Jensen H, Larsen O, da Silva Z, Lisse IM:
Increased prevalence of retrovirus infections among older
women in Africa. Scand J Infect Dis 1999, 31:459-466. gy
29. http://www.retrovirology.com/content/4/1/85 De Cock KM, Odehouri K, Colebunders RL, Adjorlolo G, Lafontaine
MF, Porter A, Gnaore E, Diaby L, Moreau J, Heyward WL, .: A com-
parison of HIV-1 and HIV-2 infections in hospitalized
patients in Abidjan, Cote d'Ivoire. AIDS 1990, 4:443-448. J
48. Beagley KW, Gockel CM: Regulation of innate and adaptive
immunity by the female sex hormones oestradiol and pro-
gesterone. FEMS Immunol Med Microbiol 2003, 38:13-22. 30. Kassim S, Sassan-Morokro M, Ackah A, Abouya LY, Digbeu H, Yesso
G, Coulibaly IM, Coulibaly D, Whitaker PJ, Doorly R, .: Two-year
follow-up of persons with HIV-1- and HIV-2-associated pul-
monary tuberculosis treated with short-course chemother-
apy in West Africa. AIDS 1995, 9:1185-1191. g
49. Marx PA, Spira AI, Gettie A, Dailey PJ VRS Lackner AA,Mahoney
CJ,Miller CJ,Claypool LE,Ho DD,Alexander NJ.: Progesteron
implants enhance SIV vaginal transmisson and early virus
load. Nat Med 1996, 2:1084-1089. 50. Smith SM, Baskin GB, Marx PA: Estrogen protects against vaginal
transmission of simian immunodeficiency virus [In Process
Citation]. J Infect Dis 2000, 182:708-715. 31. Kestens L, Brattegaard K, Adjorlolo G, Ekpini E, Sibailly T, Diallo K,
Gigase PL, Gayle H, De Cock KM: Immunological comparison of
HIV-1-, HIV-2- and dually-reactive women delivering in
Abidjan, Cote d'Ivoire. AIDS 1992, 6:803-807. J
51. Smith SM, Mefford M, Sodora D, Klase Z, Singh M, Alexander N, Hess
D, Marx PA: Topical estrogen protects against SIV vaginal
transmission without evidence of systemic effect. AIDS 2004,
18:1637-1643. 32. Nkengasong JN, Kestens L, Ghys PD, Koblavi-Deme S, Otten RA, Bile
C, Maurice C, Kalou M, Laga M, Wiktor SZ, Greenberg AE: Dual
infection with human immunodeficiency virus type 1 and
type 2: impact on HIV type 1 viral load and immune activa-
tion markers in HIV- seropositive female sex workers in
Abidjan, Ivory Coast. AIDS Res Hum Retroviruses 2000,
16:1371-1378. 52. Poulsen AG, Kvinesdal B, Aaby P, Molbak K, Frederiksen K, Dias F,
Lauritzen E: Prevalence of and mortality from human immun-
odeficiency virus type 2 in Bissau, West Africa. Lancet 1989,
1:827-831. 33. Arisawa K, Soda M, Akahoshi M, Matsuo T, Nakashima E, Tomonaga
M, Saito H: Human T-lymphotropic virus type-I infection, anti-
body titers and cause-specific mortality among atomic-
bomb survivors. Jpn J Cancer Res 1998, 89:797-805. 53. Poulsen AG, Aaby P, Jensen H, Dias F: Risk factors for HIV-2 sero-
positivity among older people in Guinea- Bissau. A search for
the early history of HIV-2 infection. 55.
Clayton D, Hills M: Statistical models in Epidemiology Oxford, Oxford
University Press; 1993. Page 8 of 9 (page number not for citation purposes) http://www.retrovirology.com/content/4/1/85 Scand J Infect Dis 2000,
32:169-175. 54. Aaby P, Ariyoshi K, Buckner M, Jensen H, Berry N, Wilkins A, Richard
D, Larsen O, Dias F, Melbye M, Whittle H: Age of wife as a major
determinant of male-to-female transmission of HIV-2 infec-
tion: a community study from rural West Africa. AIDS 1996,
10:1585-1590. 34. Arisawa K, Sobue T, Yoshimi I, Soda M, Shirahama S, Doi H, Katamine
S, Saito H, Urata M: Human T-lymphotropic virus type-I infec-
tion, survival and cancer risk in southwestern Japan: a pro-
spective cohort study. Cancer Causes Control 2003, 14:889-896. p
y
35. Iwata K, Ito S, Saito H, Ito M, Nagatomo M, Yamasaki T, Yoshida S,
Suto H, Tajima K: Mortality among inhabitants of an HTLV-I
endemic area in Japan. Jpn J Cancer Res 1994, 85:231-237. 55. Clayton D, Hills M: Statistical models in Epidemiology Oxford, Oxford
University Press; 1993. 55. Clayton D, Hills M: Statistical models in Epidemiology Oxford, Oxford
University Press; 1993. 36. Gessain A, Barin F, Vernant JC, Gout O, Maurs L, Calender A, de The
G: Antibodies to human T-lymphotropic virus type-I in
patients with tropical spastic paraparesis. Lancet 1985,
2:407-410. 37. Osame M, Usuku K, Izumo S, Ijichi N, Amitani H, Igata A, Matsumoto
M, Tara M: HTLV-I associated myelopathy, a new clinical
entity. Lancet 1986, 1:1031-1032. y
38. Poiesz BJ, Ruscetti FW, Reitz MS, Kalyanaraman VS, Gallo RC: Isola-
tion of a new type C retrovirus (HTLV) in primary uncul-
tured cells of a patient with Sezary T-cell leukaemia. Nature
1981, 294:268-271. 39. Yoshida M, Osame M, Usuku K, Matsumoto M, Igata A: Viruses
detected in HTLV-I-associated myelopathy and adult T-cell
leukaemia are identical on DNA blotting. Lancet 1987,
1:1085-1086. Publish with BioMed Central and every
scientist can read your work free of charge 40. Murphy EL, Hanchard B, Figueroa JP, Gibbs WN, Lofters WS, Camp-
bell M, Goedert JJ, Blattner WA: Modelling the risk of adult T-cell
leukemia/lymphoma in persons infected with human T-lym-
photropic virus type I. Int J Cancer 1989, 43:250-253. p
p
p
J
41. Kaplan JE, Osame M, Kubota H, Igata A, Nishitani H, Maeda Y, Khab-
baz RF, Janssen RS: The risk of development of HTLV-I-associ-
ated myelopathy/tropical spastic paraparesis among persons
infected with HTLV-I. J Acquir Immune Defic Syndr 1990,
3:1096-1101. 42. Murphy EL, Wilks R, Morgan OS, Hanchard B, Cranston B, Figueroa
JP, Gibbs WN, Murphy J, Blattner WA: Health effects of human
T-lymphotropic virus type I (HTLV-I) in a Jamaican cohort. Int J Epidemiol 1996, 25:1090-1097. http://www.retrovirology.com/content/4/1/85 Publish with BioMed Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical research in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
Submit your manuscript here:
http://www.biomedcentral.com/info/publishing_adv.asp
BioMedcentral Publish with BioMed Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical research in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
Submit your manuscript here:
http://www.biomedcentral.com/info/publishing_adv.asp
BioMedcentral Publish with BioMed Central and every
scientist can read your work free of charge
|
https://openalex.org/W4210329742
|
https://nottingham-repository.worktribe.com/preview/723587/clep-6-071.pdf
|
English
| null |
The epidemiology of irritable bowel syndrome
|
Clinical epidemiology
| 2,014
|
cc-by
| 7,788
|
The epidemiology of irritable bowel syndrome Number of times this article has been viewed
This article was published in the following Dove Press journal:
Clinical Epidemiology
4 February 2014 Caroline Canavan
Joe West
Timothy Card
Division of Epidemiology and Public
Health, University of Nottingham,
Nottingham, UK Abstract: Irritable bowel syndrome (IBS) is a functional condition of the bowel that is diagnosed
using clinical criteria. This paper discusses the nature of the diagnostic process for IBS and how
this impacts epidemiological measurements. Depending on the diagnostic criteria employed,
IBS affects around 11% of the population globally. Around 30% of people who experience the
symptoms of IBS will consult physicians for their IBS symptoms. These people do not have
significantly different abdominal symptoms to those who do not consult, but they do have greater
levels of anxiety and lower quality of life. Internationally, there is a female predominance in
the prevalence of IBS. There is 25% less IBS diagnosed in those over 50 years and there is
no association with socioeconomic status. IBS aggregates within families and the genetic and
sociological factors potentially underlying this are reviewed. Patients diagnosed with IBS are
highly likely to have other functional disease and have more surgery than the general population. There is no evidence that IBS is associated with an increased mortality risk. The epidemiological
evidence surrounding these aspects of the natural history is discussed. K
d
i i bl b
l
d
id
i l
l
li
l hi Correspondence: Caroline Canavan
Division of Epidemiology and Public
Health, University of Nottingham,
Clinical Sciences building, City Hospital
Campus, Hucknall Road, Nottingham, UK
Email Caroline.Canavan@nottingham. ac.uk Keywords: irritable bowel syndrome, epidemiology, prevalence, mortality, natural history Introduction The epidemiology of any condition is an expansive topic, covering many subjects that
individually could warrant devoted review articles. Irritable bowel syndrome (IBS) is
no different. The intention of this review is to provide a brief overview of the funda
mental issues of epidemiological interest relating to IBS. We discuss the diagnostic
process and variation in how IBS is defined, how many of the population have IBS
based upon these definitions, which members of the population are most likely to be
affected, and discuss key aspects of the natural history, including symptom fluctuation
and the association with other functional conditions. In compiling this review, potentially relevant articles were identified through a
literature search using MEDLINE and Embase of English language papers published
since the first formal definition of IBS in 1978 until August 1, 2013. Medical subject
headings (MeSH) used were ‘irritable bowel syndrome’ or ‘IBS’ or ‘irritable colon’
or ‘functional bowel disease’ combined with ‘epidemiology’ or ‘incidence’ or
‘prevalence’ or ‘natural history’ or ‘mortality’. The references of these papers were
reviewed for additional important papers not initially captured. Correspondence: Caroline Canavan
Division of Epidemiology and Public
Health, University of Nottingham,
Clinical Sciences building, City Hospital
Campus, Hucknall Road, Nottingham, UK
Email Caroline.Canavan@nottingham. ac.uk Clinical Epidemiology Clinical Epidemiology Dovepress
open access to scientific and medical research Clinical Epidemiology © 2014 Canavan et al. This work is published by Dove Medical Press Limited, and licensed under Creative Commons Attribution – Non Commercial (unported, v3.0)
License. The full terms of the License are available at http://creativecommons.org/licenses/by-nc/3.0/. Non-commercial uses of the work are permitted without any further
permission from Dove Medical Press Limited, provided the work is properly attributed. Permissions beyond the scope of the License are administered by Dove Medical Press Limited. Information on
how to request permission may be found at: http://www.dovepress.com/permissions.php
Clinical Epidemiology 2014:6 71–80
71 Correspondence: Caroline Canavan
Division of Epidemiology and Public
Health, University of Nottingham,
Clinical Sciences building, City Hospital
Campus, Hucknall Road, Nottingham, UK
Email Caroline.Canavan@nottingham.
ac.uk What is IBS? IBS is a chronic functional disorder of the gastrointestinal system. Patients experience
abdominal pain and altered bowel habit, with either predominantly diarrhea (IBS-D), ) 71 Dovepress Dovepress Canavan et al constipation (IBS-C), or both (IBS-M). There is no definitive
investigation as no biomarker has been found, so IBS is diag
nosed clinically. The earliest reports describing IBS are from
the turn of the 19th and 20th century.1 At this time, diagnosis
was only made by exclusion of malignant, inflammatory, or
infectious disease after full investigation and “extensive unsuc
cessful surgeries.”2 IBS remained “frequently misdiagnosed
and poorly understood”3 into the 1970s, with the problem of
unsuccessful, or unnecessary, surgery continuing.3 The validity of these diagnostic criteria has been studied;
however, this has not been done using conventional mea
surement of sensitivity and specificity in general population
samples because there is no gold standard to allow indepen
dent confirmation of diagnosis, such as a biomarker. Instead
estimations of specificity have been made among those with
organic gastrointestinal disease, and sensitivity amongst cases
who have already presented and been diagnosed (potentially
an atypical minority). All of these criteria have modest speci
ficity to diagnose IBS in those with organic gastrointestinal
disease of about 0.7.10 This reflects the fact that the symptoms
experienced in IBS are common to other gastrointestinal
conditions. Specificity is increased to 0.9 if patients with red
flag symptoms such as anemia, weight loss, and rectal bleed
ing (present in just 3% of patients)11 are more extensively
investigated and IBS is diagnosed by exclusion.10 Sensitivity
across the criteria ranges from 0.4 to 0.9 depending on the
experience of the clinician,10 possibly reflecting variation in
confidence to use criteria for positive diagnosis. In an attempt to standardize and define IBS and reduce
unnecessary surgery, Manning created the first set of formal
criteria that allowed IBS to be positively diagnosed without
the need for extensive investigations to exclude other diagno
ses. These criteria were developed through expert consensus
to create the Rome criteria, now in their third iteration.4 The
Rome criteria also recommend that IBS should be a positive
diagnosis but are more restrictive than the Manning criteria. When assessing the epidemiology of IBS, the diagnostic
criteria employed are important to consider as they reflect
how cases are ascertained and the differences between studies
and over time. What is IBS? Most studies use either the Manning, Rome
II, or Rome III criteria, and the variation between these can
be seen in Table 1. It is argued that the Rome criteria are
only useful for clinical trials and not for clinical practice. The criteria have been compiled by secondary and tertiary
care specialists who undoubtedly see a different population
of patients than are seen in primary care where the criteria
are recommended to be implemented. Clinicians frequently
use other clinical findings (such as bloating and psychologi
cal stress) to assist in diagnosing IBS and do not adhere to
the criteria.5,6 Some even suggest that attempts to sub-type
functional gastrointestinal disease are arbitrary and IBS
should remain a diagnosis of exclusion.6–8 International best
practice guidelines disagree with this stance and promote
positive diagnosis using the Rome criteria.9 How much IBS is there? As outlined above, there is no gold standard case definition of
IBS. Diagnostic criteria have not been standardized over time
and they have a large margin of error in their application.5,8,15
The definition of cases in epidemiological studies is there
fore difficult and there may be limited opportunities for
ascertainment of cases. There is no specific or standardized
therapy, with many of the preparations available over the
counter and some patients not requiring any medication,15,16
so prescription data have limited use. Few patients will
be admitted to hospital with IBS or diagnosed during an
admission,17 and IBS is not considered a cause of death so
these data are also not useful to define cases. The methods
chosen to define and ascertain cases will clearly affect the
number of cases found. As outlined above, there is no gold standard case definition of
IBS. Diagnostic criteria have not been standardized over time
and they have a large margin of error in their application.5,8,15 Table 1 Comparison of the Manning and Rome diagnostic criteria frequently used in epidemiological studies for case ascertainment
Manning (1978)12
Rome I (1989)13
Rome II (1999)14
Rome III (2006)4
2 or more of the following
symptoms:
Abdominal distension
Pain relief with defecation
Frequent stools with pain
Looser stools with pain
Passage of mucus
Sensation of incomplete
evacuation
At least 3 months of continuous
or recurrent abdominal pain:
Relieved with defecation
or
Associated with change in stool consistency
With at least 2 of the following on at least
25% of days:
Altered stool frequency
Altered stool form
Altered stool passage
Passage of mucus
Bloating or abdominal distension
At least 12 weeks in past 12 months
of continuous or recurrent abdominal
pain or discomfort
With at least 2 of the following:
Relief with defecation
Altered stool frequency
Altered stool form
Onset of symptoms more than
12 months before diagnosis
At least 3 days per month in
past 12 weeks of continuous
or recurrent abdominal pain
or discomfort
With at least 2 of the
following:
Relief with defecation
Altered stool frequency
Altered stool form
Onset of symptoms more
than 6 months before
diagnosis Clinical Epidemiology 2014:6 72 submit your manuscript | www.dovepress.com Dovepress Epidemiology of irritable bowel syndrome Dovepress set of diagnostic criteria. This method reflects the underlying
burden of symptoms consistent with a diagnosis of IBS. How much IBS is there? Other
studies ask participants if they have ever received a diagnosis
of IBS, capturing those who have sought medical advice for
their symptoms. The prevalence of IBS within the community
is between 10% and 25%.20,43,44,57,58,62–65 Meta-analysis shows
a pooled estimate of international IBS prevalence of 11.2%
(95% confidence interval [CI] 9.8–12.8),38 with variation by
geographic region; the lowest occurring in South Asia (7.0%)
and the highest in South America (21.0%).38 Incidence Symptoms associated with IBS are commonly experienced
within the population, it has an insidious onset, and frequently
does not result in medical care.16,18 Consequently, there is
discrepancy between incidence of the first occurrence of
symptoms, which occur within the community, and the first
diagnosis of IBS, which will occur after visiting a physician.16
Consequently, few studies calculate the incidence of IBS. One study in the USA using first occurrence of IBS
symptoms within the community estimated the incidence of
new IBS by conducting two population cohort surveys 1 year
apart. Of patients with no IBS symptoms and no diagnosis of
IBS in the baseline survey, 9% had developed symptoms over
the year, an incidence rate of 67 per 1,000 person-years.19
Studies that opt for defining cases as first diagnosis by a
physician produce more conservative estimates of around
two per 1,000 person years.20–22 Considerable heterogeneity exists between studies.38 Some
of this is explained through the differences in study methodol
ogy and sampling, as described above, and the use of different
diagnostic criteria to define IBS, rather than a biological marker,
may also account for some of the variation. The Manning cri
teria account for the highest reported prevalence,13,47 whilst the
Rome iterations are associated with lower estimates of preva
lence that are similar across the iterations.13 Meta-analysis has
shown that prevalence does not vary significantly according to
the calendar year in which studies are conducted.38 Who gets IBS?
Sex In most populations, women report more IBS symptoms than
men, irrespective of the diagnostic criteria employed.68 Rates
in women are approximately 1.5- to 3-fold higher than those
seen in men.12,78,79 Internationally, the overall prevalence of
IBS in women is 67% higher than in men (odds ratio 1.67
[95% CI 1.53–1.82]). This relative difference reflects an
absolute difference in prevalence of just over 5% between
the sexes, with a prevalence in women of 14.0% (95% CI
11.0–16.0) compared with 8.9% in men (95% CI 7.3–10.5).80
In South Asia, South America, and Africa, rates of IBS in
men are much closer to those of women, and in some cases
higher. Consequently, if prevalence is stratified according to
geographic region, no significant sex difference can be seen
in these areas.80 It is possible that this reflects the sex differ
ences in access to health care and symptom normalization
in these regions.81,82 Equal proportions of men and women
with symptoms seek primary care advice.83,84 Abbreviations: CI, confidence interval; NR, not reported. symptoms specifically.69,70 It is therefore likely that areas where
greater stigma is perceived will report lower prevalence, as will
populations where symptoms are more often identified as varia
tions of normal. Those communities in which there is higher
perceived stress or lower perceived quality of life, greater
potential gain from receiving a diagnosis, or fewer barriers to
accessing health care will report higher prevalence.71–74 It is likely that the differences in reported sex-specific
prevalence occur for reasons similar to those discussed for
overall prevalence. It is also probable that the rates reflect
differences in illness behavior that motivates consulting and
referral patterns rather than the underlying condition. It could Global prevalence Table 2 Prevalence of irritable bowel syndrome reported
internationally, the highest and lowest estimated rates for each
country attend primary care for their symptoms are between 10%63
and 70%.75 In the UK, estimates of the proportion who con
sult vary from 30%58 to 50%.57,62,63 Reported consultation
rates of people with symptoms also vary in studies from
Germany, between 10%63 and 50%.40 Asking individuals to
recall if they have consulted a primary care physician about
their symptoms shows that the proportion of patients who
report they have attended primary care varies considerably
by country across Europe. The highest proportion is in Italy,
with 50% consulting; 30% consult in the Netherlands, 20%
in Belgium, and 10% in Switzerland, France, and Spain.63
These differences may again reflect differences in diagnostic
criteria employed, perceived acceptability of symptoms, and
ease of access to primary care, which varies across different
health care systems. These studies are also limited by relying
on individuals’ recall of their behavior. Studies from the USA
report more consistently that 30% of people with symptoms
will consult16,65 and, of these patients, 80% have IBS-D.16 There are no significant differences in gastrointestinal
symptoms between those who consult and those who do
not, but those who consult do report higher pain scores,
greater levels of anxiety, and greater reduction in quality
of life.76,77 submit your manuscript | www.dovepress.com
74 Global prevalence Prevalence estimates for IBS vary greatly internationally,
both within and between countries, as shown in Figure 1
and Table 2. It is plausible that the underlying prevalence of symptoms
in communities internationally is the same, around 11%, and
the variations reflect differences in access to health care66,67 and
acceptability of a diagnosis, both to the physician making the
diagnosis and to the patient in receiving and believing it.5,68
Significant stigma is associated with receiving a diagnosis of
a functional disorder as well as seeking health care for IBS Most studies addressing prevalence of IBS are community
surveys, with the majority from Europe, Southeast Asia, and
North America. Often, postal questionnaires or telephone inter
views invite individuals to self-report symptoms, and these
responses are then assessed by investigators according to one Prevalence
<10%
10% to <15%
15% to <20%
>20%
Figure 1 Worldwide prevalence of irritable bowel syndrome, as reported by country. Figure 1 Worldwide prevalence of irritable bowel syndrome, as reported by country. Clinical Epidemiology 2014:6 Clinical Epidemiology 2014:6 73 Dovepress Dovepress Canavan et al Table 2 Prevalence of irritable bowel syndrome reported
internationally, the highest and lowest estimated rates for each
country
Country
Lowest
estimated
prevalence
95% CI
Highest
estimated
prevalence
95% CI
France
1.123
NR
4.724
4.36–5.04
Thailand
–
–
5.725
NR
Netherlands
–
–
5.826
4.0–9.0
Hong Kong
3.727
2.0–5.2
6.628
NR
Italy
–
–
7.229
6.0–9.0
Iran
–
–
7.130
6.0–13.0
South Africa
–
–
8.131
NR
Norway
–
–
8.432
7.9–9.4
Bangladesh
–
–
8.533
7.0–10.0
Turkey
6.334
NR
10.235
6.0–16.0
Singapore
2.336
0.8–3.9
11.037
9.7–12.3
Israel
2.938
NR
11.438
NR
People’s
Republic of
China
0.838
NR
11.539
NR
Germany
–
–
12.540
10.7–14.5
Australia
4.441
3.6–5.1
13.042
11.0–16.0
Pakistan
–
–
13.343
4.0–62.0
Canada
–
–
13.544
10.2–14.0
Japan
6.145
5.0–7.0
14.046
NR
Spain
3.347
2.1–4.9
14.147
10.0–18
Romania
–
–
14.438
11.9–19.0
Sweden
12.520
9.0–18.0
15.048
NR
South Korea
6.649
2.0–11.0
15.550
12.0–19.0
Malaysia
–
–
15.651
13.0–18.0
Finland
5.152
4.4–5.8
16.452
15.0–17.2
Brazil
–
–
17.053
12.0–23.0
New Zealand
3.354
2.1–4.5
18.854
16.3–21.3
Russia
–
–
19.038
17.0–22.0
Columbia
–
–
19.938
16.7–23.3
USA
3.055
NR
20.413
16.7–24.2
Taiwan
17.556
NR
22.156
NR
UK
6.157
NR
21.658
NR
Greece
–
–
21.459
NR
Peru
15.038
NR
24.038
21.0–28.0
Croatia
–
–
28.238
24.0–32.0
Iceland
17.260
14.4–19.9
30.960
28.0–33.0
Nigeria
–
–
31.661
27.0–36.0
Abbreviations: CI, confidence interval; NR, not reported. Symptom patterns Symptoms experienced by patients with IBS fluctuate over
time. Over short periods of time, as some patients experi
ence resolution of symptoms, others develop new ones,
meaning the prevalence of symptomatic IBS remains stable
over 1–2 years follow-up.20,83,107 In the first 3 months fol
lowing diagnosis, patients experience four distinct episodes
of symptoms per month on average. The longest of these
episodes lasts around 5 days, and most patients experi
ence symptoms on more than half of the days.55 However,
1 year after initial diagnosis 30%–45% of patients will have
prolonged periods that are symptom free, potentially in
remission.20,108 After 10 years, 50%–70% of patients report
persistent symptoms.20,109 In patients who do report IBS symptom resolution, 45%
will subsequently experience other functional gastrointestinal
symptoms.20,110 Up to two-thirds of IBS patients experience
functional dyspepsia, the prevalence of which in patients with
IBS is up to seven times higher than in controls.111–113 What is the natural history?
Misdiagnosis IBS occurs in all age groups, including children85 and the
elderly, with no difference seen in the frequency of sub
types by age.86 However, 50% of patients with IBS report
having first had symptoms before the age of 35 years,87 and
prevalence is 25% lower in those aged over 50 years than in
those who are younger.38 This would suggest that symptoms
remit over time, and is contrary to the belief that IBS is a
chronic lifelong condition, because, if this were the case,
then prevalence should remain constant or increase with age. Patients aged over 50 years also report milder pain, but their
overall quality of life is worse.86 Those aged over 65 years
are also likely to have had their symptoms for longer than
1 year before they consult, whilst those under 65 years report
significantly shorter duration of symptoms.42 At the time of diagnosis of IBS, rates of organic lesions found
on colonoscopy in patients with no red flag symptoms are no
higher than seen in healthy controls, ranging from around
10%101,102 to around 40%.103,104 Even most patients with alarm
symptoms have no organic pathology.105 In contrast to the endoscopy findings at diagnosis, in the
period following diagnosis the rate of inflammatory bowel
disease (IBD) is 9106 to 1621 times higher than the general
population. The average time between diagnosis of IBS and
IBD is 2–3 years,106,21 suggesting that, in some, IBS symptoms
are possibly early signs of IBD before lesions are visible. Colorectal cancer incidence is around 1%21 in the first year fol
lowing diagnosis of IBS, higher than in the general population. After 1 year, the rate returns to the population level. Socioeconomic status One study suggested that IBS was associated with lower
socioeconomic status,65 a finding supported by the theory that
lower income is associated with poorer health care outcomes,
lower overall quality of life, and increased life stressors.88
However, others suggest that the opposite is true and that
being in a higher socioeconomic group during childhood is
associated with higher prevalence of IBS.89,90 Similarly, areas
in which there is a lower proportion of people employed in
manual labor have higher rates of IBS. It is suggested that
this is due to the higher level of stress perceived by people
working in professional and managerial roles.91 This supports
the argument that IBS is a disease of industrialization and
urbanization, and that the higher rates now being reported
in Asia, South America, and Africa are due to increased
affluence in these regions.92 This may be because those with
higher income have greater access to health care and tendency
to seek help and hence receive a diagnosis.93 It could also
reflect differing dietary choices94 or greater internalization
of stress in higher earning groups.95 Prevalence in primary care Many patients do not seek medical attention for symptoms
indicative of IBS.18 Estimates of the proportion who do Clinical Epidemiology 2014:6 submit your manuscript | www.dovepress.com
4 74 Dovepress Epidemiology of irritable bowel syndrome Dovepress be argued that the high rate of IBS symptoms in people within
the community who do not access health care is a measure
of unmet need.68 However, it could also be considered that
symptoms affecting such a high proportion of the population
are a variation of normal function. a twin with IBS.97 Concordance in monozygotic twins (the
proportion of twin pairs who both have IBS) is less than
20%,97,98 and the association seen in familial clustering is
significantly reduced when somatization is adjusted for.99
These findings suggest heredity may be more closely linked
to learned behavior than to genetic factors.100 Co-existing functional conditions g
Fibromyalgia, chronic fatigue syndrome, chronic back pain,
chronic pelvic pain, chronic headache, and temporoman
dibular joint dysfunction exist in approximately half of all
patients with IBS, and occur almost twice as often as in
the general population.78,111,118 These conditions, defined by
their symptoms, have considerable overlap and their eti
ologies are poorly understood.119,120 Some people feel these
conditions should be considered together under the single
umbrella term of ‘functional somatic syndromes’.121 Whilst
these conditions do share some predisposing risk factors, the
triggers for developing each condition vary.120 Multivariate
analysis supports the idea that somatic comorbidities seen
in IBS are distinct disorders and not manifestations of one
somatization disorder.118 Even so, patients who have IBS
and other somatic comorbidities report more severe symp
toms than those patients with IBS alone.111,115,122 In addition,
over one-half of all patients with IBS report depression or
anxiety,86,115,120,123 and these patients experience more severe
somatic symptoms.124 IBS is a significant health care burden, irrespective of setting or
geography, affecting around 11% of the population globally. Accurate case definition remains difficult in IBS due
to the high frequency of symptoms within the community,
variations in diagnostic criteria and the stringency with which
they are implemented, lack of specific histopathological
changes, and lack of a definitive point of onset. This makes
epidemiological studies challenging, with rates dependent
on the methods chosen to define and measure IBS, so the
literature should be considered within this context. Evidence of lower prevalence of IBS in older age groups
suggests that symptoms resolve over time, but this is contra
dicted by natural history studies, which imply chronicity of
symptoms. A significant amount of evidence also suggests
that patients with IBS are more likely to also have other
functional conditions. Evidence suggests that ‘symptom shift
ing’ occurs in a proportion of patients, where resolution of
functional bowel symptoms sees development of functional
symptoms in another system. Older studies suggest a high secondary care burden of
IBS, with it accounting for up to 50% of gastrointestinal
outpatient clinic time. However, these are highly biased
estimates, often reliant on questioning physicians about their
perceived workload. More recent studies would suggest IBS
accounts for closer to 20% of gastroenterology outpatient
work, possibly in part reflecting changes in primary care
referral guidelines. Family studies If all gastrointestinal symptoms resolve, then many
develop symptoms of other functional diseases.20,114 Those
with lower quality of life and higher levels of anxiety
are more likely to suffer with other functional comorbid
conditions.111,115 Together with family studies proposing The relative risk of IBS is twice as high in individuals with
a biological relative with IBS.96 In twin studies, having a
mother or father with IBS is an independent risk factor for an
individual having IBS and a stronger predictor than having Clinical Epidemiology 2014:6 75 Dovepress Dovepress Canavan et al increased, especially in elderly patients with IBS-C, where
it was 7.5 times that of the general population. This possibly
reflects misdiagnosis or ascertainment bias.129 consultation behavior is learnt and highly associated with
somatization,97,99,116 this suggests that IBS is one expression
of an underlying tendency toward functional disease. Co-existing functional conditions Even so, this remains a significant propor
tion and highlights the importance of continuing to improve
diagnosis and management of these individuals, to optimize
their health care utilization and fully assess the impact and
benefit of different management approaches. Mortality IBS is a functional disease that significantly reduces patients’
quality of life, is associated with depression and suicidal ide
ation, and patients have an increased frequency of invasive
procedures and surgery.118,127 Despite this, community-based
studies have shown that IBS is not associated with any increased
mortality.128,129 A large study in the USA of over 4,000 patients,
followed for a total of 30,000 patient-years, and with 1,101
deaths, observed no increased mortality compared with the
general population (hazard rate 1.06 [95% CI 0.86–1.32]).128
A recent much smaller study from the People’s Republic of
China followed 263 patients over 5 years and found the same. They did notice that the incidence of colorectal cancer was Surgery Symptoms of IBS have considerable overlap with organic
gastrointestinal and pelvic pathologies. Consequently,
patients with IBS are at risk of undergoing unnecessary sur
gical procedures due to misdiagnosis.93,125 The likelihood of
having cholecystectomy is between two126 and three127 times
as high in IBS as in the general public, and patients with
IBS are almost twice as likely to have appendectomies63,127
or hysterectomies.127 Disclosure CC is funded through a Population Health Scientist
Fellowship from the Medical Research Council. JW is funded
by a Nottingham University Hospitals National Health
Service Trust/University of Nottingham Senior Clinical
Research Fellowship. The authors report no conflicts of
interest in this work. submit your manuscript | www.dovepress.com submit your manuscript | www.dovepress.com
Dovepress References 1. Chaudhary NA, Truelove SC. The irritable colon syndrome. A study of
the clinical features, predisposing causes, and prognosis in 130 cases. Q J Med. 1962;31:307–322. Clinical Epidemiology 2014:6 submit your manuscript | www.dovepress.com 76 Dovepress Dovepress Epidemiology of irritable bowel syndrome 2. Kruse FH. Functional disorders of the colon: the spastic colon, the irri
table colon, and mucous colitis. Cal West Med. 1933;39(2):97–103. 24. Dapoigny M, Bellanger J, Bonaz B, et al. Irritable bowel syndrome in
France: a common, debilitating and costly disorder. Eur J Gastroenterol
Hepatol. 2004;16(10):995–1001. 3. No authors listed. Irritable bowel syndrome. Br Med J. 1972;1(5794):
197–198. 25. Danivat D, Tankeyoon M, Sriratanaban A. Prevalence of irritable
bowel syndrome in a non-Western population. Br Med J (Clin Res Ed). 1988;296(6638):1710. 4. Longstreth GF, Thompson WG, Chey WD, Houghton LA, Mearin F,
Spiller RC. Functional bowel disorders. Gastroenterology. 2006;130(5):
1480–1491. 26. Boekema P, van Dam van Isselt EF, Bots ML, Smout AJ. Functional
bowel symptoms in a general Dutch population and associations with
common stimulants. Neth J Med. 2001;59(1):23–30. 5. Spiegel BM, Farid M, Esrailian E, Talley J, Chang L. Is irritable bowel
syndrome a diagnosis of exclusion? A survey of primary care providers,
gastroenterologists, and IBS experts. Am J Gastroenterol. 2010;105(4):
848–858. 27. Lau EM, Chan FK, Ziea ET, Chan CS, Wu JC, Sung JJ. Epidemiology of irritable bowel syndrome in Chinese. Dig Dis Sci. 2002;47(11):2621–2624. 6. Mearin F, Lacy BE. Diagnostic criteria in IBS: useful or not? Neurogastroenterol Motil. 2012;24(9):791–801. 28. Kwan AC, Hu WH, Chan YK, Yeung YW, Lai TS, Yuen H. Prevalence
of irritable bowel syndrome in Hong Kong. J Gastroenterol Hepatol. 2002;17(11):1180–1186. 7. Engsbro AL, Begtrup LM, Kjeldsen J, et al. Patients suspected of irri
table bowel syndrome: cross-sectional study exploring the sensitivity
of Rome III criteria in primary care. Am J Gastroenterol. 2013;108(6):
972–980. 29. Usai P, Manca R, Lai MA, et al. Prevalence of irritable bowel syn
drome in Italian rural and urban areas. Eur J Intern Med. 2010;21(4):
324–326. 8. Quigley EM. The ‘con’ case. The Rome process and functional gastro
intestinal disorders: the barbarians are at the gate! Neurogastroenterol
Motil. 2007;19(10):793–797. 30. Amra B, Hoseini-Asl MK, Rahmani AR, Golshan M, Mohamad-Zadeh Z. Correlation between asthma and irritable bowel syndrome in a general
population in Iran in 2003. Respir Med. 2006;100(1):110–114. 9. Quigley EMM, Fried M, Gwee K-A, et al. Irritable Bowel Syndrome:
A Global Perspective. References Milwaukee (WI): World Gastroenterology
Organisation; 2009:1–20. 31. Segal I, Walker AR. The irritable bowel syndrome in the black
community. S Afr Med J. 1984;65(3):72–73. 10. Whitehead WE, Drossman DA. Validation of symptom-based diag
nostic criteria for irritable bowel syndrome: a critical review. Am J
Gastroenterol. 2010;105(4):814–820; quiz 813, 821. 32. Vandvik PO, Lydersen S, Farup PG. Prevalence, comorbidity and
impact of irritable bowel syndrome in Norway. Scand J Gastroenterol. 2006;41(6):650–656. 11. Whitehead WE, Palsson OS, Feld AD, et al. Utility of red flag symp
tom exclusions in the diagnosis of irritable bowel syndrome. Aliment
Pharmacol Ther. 2006;24(1):137–146. 33. Perveen I, Hasan M, Masud MA, Bhuiyan MM, Rahman MM. Irritable
bowel syndrome in a Bangladeshi urban community: prevalence
and health care seeking pattern. Saudi J Gastroenterol. 2009;15(4):
239–243. 12. Manning AP, Thompson WG, Heaton KW, Morris AF. Towards
positive diagnosis of the irritable bowel. Br Med J. 1978;2(6138):
653–654. 34. Celebi S, Acik Y, Deveci SE, et al. Epidemiological features of irritable
bowel syndrome in a Turkish urban society. J Gastroenterol Hepatol. 2004;19(7):738–743. 13. Saito YA, Locke GR, Talley NJ, Zinsmeister AR, Fett SL, Melton LJ 3rd. A comparison of the Rome and Manning criteria for case identification
in epidemiological investigations of irritable bowel syndrome. Am J
Gastroenterol. 2000;95(10):2816–2824. 35. Yilmaz S, Dursun M, Ertem M, Canoruc F, Turhanog˘lu A. The epi
demiological aspects of irritable bowel syndrome in Southeastern
Anatolia: a stratified randomised community-based study. Int J Clin
Pract. 2005;59(3):361–369. 14. Thompson WG, Longstreth GF, Drossman DA, Heaton KW, Irvine EJ,
Müller-Lissner SA. Functional bowel disorders and functional abdomi
nal pain. Gut. 1999;45 Suppl 2:II43–II47. 36. Ho KY, Kang JY, Seow A. Prevalence of gastrointestinal symptoms in
a multiracial Asian population, with particular reference to reflux-type
symptoms. Am J Gastroenterol. 1998;93(10):1816–1822. p
pp
15. Ford AC, Talley NJ. Irritable bowel syndrome. BMJ. 2012;345:
e5836. 37. Gwee KA, Wee S, Wong ML, Png DJ. The prevalence, symptom char
acteristics, and impact of irritable bowel syndrome in an asian urban
community. Am J Gastroenterol. 2004;99(5):924–931. 16. Hungin AP, Chang L, Locke GR, Dennis EH, Barghout V. Irritable
bowel syndrome in the United States: prevalence, symptom patterns
and impact. Aliment Pharmacol Ther. 2005;21(11):1365–1375. 38. Lovell RM, Ford AC. Global prevalence of and risk factors for irri
table bowel syndrome: a meta-analysis. Clin Gastroenterol Hepatol. 2012;10(7):712–721. e4. 17. Thompson WG, Heaton KW, Smyth GT, Smyth C. References Haas J, Phillips K, Sonneborn D, et al. Variation in access to health care
for different racial/ethnic groups by the racial/ethnic composition of an
individual’s contry of residence. Med Care. 2004;42(7):707–714. 67. Grundmann O, Yoon SL. Irritable bowel syndrome: epidemiology,
diagnosis and treatment: an update for health-care practitioners. J Gastroenterol Hepatol. 2010;25(4):691–699. 46. Hongo M. Epidemiology of FGID symptoms in Japanese general
population with reference to life style. J Gastroenterol Hepatol. 2011;
26 Suppl 3:19–22. 68. Quigley EM, Bytzer P, Jones R, Mearin F. Irritable bowel syndrome: the bur
den and unmet needs in Europe. Dig Liver Dis. 2006;38(10): 717–723. 47. Mearin F, Badía X, Balboa A, et al. Irritable bowel syndrome prevalence
varies enormously depending on the employed diagnostic criteria:
comparison of Rome II versus previous criteria in general population. Scand J Gastroenterol. 2001;36(11):1155–1161. 69. Dancey CP, Hutton-Young SA, Moye S, Devins GM. Perceived stigma,
illness intrusiveness and quality of life in men and women with irritable
bowel syndrome. Psychol Health Med. 2002;7(4):381–395. Scand J Gastroenterol. 2001;36(11):1155–1161. 48. Walter SA, Kjellström L, Nyhlin H, Talley NJ, Agréus L. Assessment of
normal bowel habits in the general adult population: the Popcol study. Scand J Gastroenterol. 2010;45(5):556–566. 70. Jones MP, Keefer L, Bratten J, et al. Development and initial validation
of a measure of perceived stigma in irritable bowel syndrome. Psychol
Health Med. 2009;14(3):367–374. 71. Nettleton S. ‘I just want permission to be ill’: towards a sociology of med
ically unexplained symptoms. Soc Sci Med. 2006;62(5):1167–1178. 49. Han SH, Lee OY, Bae SC, et al. Prevalence of irritable bowel syn
drome in Korea: population-based survey using the Rome II criteria. J Gastroenterol Hepatol. 2006;21(11):1687–1692. 72. Farmer J, Iversen L, Campbell NC, et al. Rural/urban differences in
accounts of patients’ initial decisions to consult primary care. Health
Place. 2006;12(2):210–221. 50. Choo KY, Choi MG, Choi H, et al. The prevalences of gastrointestinal
symptoms in a rural community in Korea. Korean J Gastrointest Motil. 2000;6(1):31–43. 73. Cummings KM, Becker MH, Maile MC. Bringing the models together:
an empirical approach to combining variables used to explain health
actions. J Behav Med. 1980;3(2):123–145. 51. Rajendra S, Alahuddin S. Prevalence of irritable bowel syndrome in a
multi-ethnic Asian population. Aliment Pharmacol Ther. 2004;19(6):
704–706. 74. Zola IK. Pathways to the doctor – from person to patient. Soc Sci Med. 1973;7(9):677–689. 52. Hillilä MT, Färkkilä MA. References Prevalence of irritable bowel syndrome
according to different diagnostic criteria in a non-selected adult
population. Aliment Pharmacol Ther. 2004;20(3):339–345. 75. Koloski NA, Talley NJ, Huskic SS, Boyce PM. Predictors of conventional
and alternative health care seeking for irritable bowel syndrome and
functional dyspepsia. Aliment Pharmacol Ther. 2003;17(6): 841–851. 53. Soares RL, dos Santos JM, Rocha VR. Prevalence of irritable bowel
syndrome in a Brazilian Amazon community. Neurogastroenterol Motil. 2005;17(6):883. 76. Creed F. Who needs a doctor for IBS? Gut. 1997;41(3):415–416. 77. Lee V, Guthrie E, Robinson A, et al. Functional bowel disorders in
primary care: factors associated with health-related quality of life and
doctor consultation. J Psychosom Res. 2008;64(2):129–138. 54. Barbezat G, Poulton R, Milne B, Howell S, Fawcett JP, Talley N. Prevalence and correlates of irritable bowel syndrome symptoms in a
New Zealand birth cohort. N Z Med J. 2002;115(1164):U220. 78. Kennedy TM, Jones RH, Hungin AP, O’flanagan H, Kelly P. Irritable
bowel syndrome, gastro-oesophageal reflux, and bronchial hyper-
responsiveness in the general population. Gut. 1998;43(6):770–774. 55. Hahn BA, Saunders WB, Maier WC. Differences between individuals
with self-reported irritable bowel syndrome (IBS) and IBS-like
symptoms. Dig Dis Sci. 1997;42(12):2585–2590. 79. Drossman DA, Funch-Jensen P, Janssens J. Identification of subgroups
of functional bowel disorders. Gastroenterol Int. 1990;(3):159–172. Available from: http://www.scopus.com/record/display.url?eid=2-
s2.0-0025601811&origin=resultslist&zone=contextBox. Accessed
December 17, 2013. 56. Lu CL, Chen CY, Lang HC, et al. Current patterns of irritable bowel syn
drome in Taiwan: the Rome II questionnaire on a Chinese population. Aliment Pharmacol Ther. 2003;18(11–12):1159–1169. 57. Heaton KW, O’Donnell LJ, Braddon FE, Mountford RA, Hughes AO,
Cripps PJ. Symptoms of irritable bowel syndrome in a British urban
community: consulters and nonconsulters. Gastroenterology. 1992;
102(6):1962–1967. 80. Lovell RM, Ford AC. Effect of gender on prevalence of irritable bowel
syndrome in the community: systematic review and meta-analysis. Am J Gastroenterol. 2012;107(7):991–1000. 81. Merzel C. Gender differences in health care access indicators in an urban,
low-income community. Am J Public Health. 2000;90(6):909–916. 58. Jones R, Lydeard S. Irritable bowel syndrome in the general population. BMJ. 1992;304(6819):87–90. 59. Papatheodoridis GV, Karamanolis DG. Prevalence and impact of
upper and lower sastrointestinal Symptoms in the Greek urban general
population. Scand J Gastroenterol. 2005;40(4):412–421. 82. World Health Organization. Why Gender and Health? Geneva: World
Health Organization; 2013. Available from: http://www.who.int/gender/
genderandhealth/en/. Accessed July 18, 2013. 83. Williams RE, Black CL, Kim HY, et al. Determinants of healthcare-
seeking behaviour among subjects with irritable bowel syndrome. References Irritable bowel
syndrome in general practice: prevalence, characteristics, and referral. Gut. 2000;46(1):78–82. 39. Xiong LS, Chen MH, Chen HX, Xu AG, Wang WA, Hu PJ. A population-based epidemiologic study of irritable bowel syndrome in
South China: stratified randomized study by cluster sampling. Aliment
Pharmacol Ther. 2004;19(11):1217–1224. 18. Chang SY, Jones MP. Consulters and nonconsulters in irritable
bowel syndrome: what makes an IBS patient? Pract Gastroenterol. 2003;6:15–26. 19. Talley NJ, Weaver A, Zinsmeister AR, Melton LJ 3rd. Onset and disap
pearence of gastrointestinal symptoms and functional gastrointestinal
disorders. Am J Epidemiol. 1992;136(2):165–177. 40. Icks A, Haastert B, Enck P, Rathmann W, Giani G. Prevalence of
functional bowel disorders and related health care seeking: a population-
based study. Z Gastroenterol. 2002;40(3):177–183. 20. Agreus L, Svärdsudd K, Nyrén O, Tibblin G. Irritable bowel syndrome
and dyspepsia in the general population: overlap and lack of stability
over time. Gastroenterology. 1995;109(3):671–680. 41. Boyce PM, Talley NJ, Burke C, Koloski NA. Epidemiology of the
functional gastrointestinal disorders diagnosed according to Rome II
criteria: an Australian population-based study. Intern Med J. 2006;36(1):
28–36. 21. García Rodríguez LA, Ruigómez A, Wallander MA, Johansson S,
Olbe L. Detection of colorectal tumor and inflammatory bowel disease
during follow-up of patients with initial diagnosis of irritable bowel
syndrome. Scand J Gastroenterol. 2000;35(3):306–311. 42. Talley NJ, Boyce PM, Jones M. Predictors of health care seeking for
irritable bowel syndrome: a population based study. Gut. 1997;41(3):
394–398. 43. Husain N, Chaudhry IB, Jafri F, Niaz SK, Tomenson B, Creed F. A population-based study of irritable bowel syndrome in a non-Western
population. Neurogastroenterol Motil. 2008;20(9):1022–1029. 22. Locke GR 3rd, Yawn BP, Wollan PC, Melton LJ 3rd, Lydick E,
Talley NJ. Incidence of a clinical diagnosis of the irritable bowel syn
drome in a United States population. Aliment Pharmacol Ther. 2004;
19(9):1025–1031. 44. Thompson WG, Irvine EJ, Pare P, Ferrazzi S, Rance L. Functional
gastrointestinal disorders in Canada: first population-based survey using
Rome II criteria with suggestions for improving the questionnaire. Dig
Dis Sci. 2002;47(1):225–235. 23. Bommelaer G, Poynard T, Le Pen C, et al. Prevalence of irritable bowel
syndrome (IBS) and variability of diagnostic criteria. Gastroenterol
Clin Biol. 2004;28(6–7 Pt 1):554–561. Clinical Epidemiology 2014:6 77 Dovepress Dovepress Canavan et al 45. Kumano H, Kaiya H, Yoshiuchi K, Yamanaka G, Sasaki T, Kuboki T. Comorbidity of irritable bowel syndrome, panic disorder, and agora
phobia in a Japanese representative sample. Am J Gastroenterol. 2004;
99(2):370–376. 66. References Could
gastrointestinal disorders differ in two close but divergent social
environments? Int J Health Geogr. 2012;11:5. 111. Hillilä MT, Siivola MT, Färkkilä MA. Comorbidity and use of health-
care services among irritable bowel syndrome sufferers. Scand J
Gastroenterol. 2007;42(7):799–806. 112. Talley NJ, Dennis EH, Schettler-Duncan VA, Lacy BE, Olden KW,
Crowell MD. Overlapping upper and lower gastrointestinal symptoms
in irritable bowel syndrome patients with constipation or diarrhea. Am
J Gastroenterol. 2003;98(11):2454–2459. 92. Gwee KA. Irritable bowel syndrome in developing countries: a
disorder of civilization or colonization? Neurogastroenterol Motil. 2005;17(3):317–324. 93. Cremonini F, Talley NJ. Irritable bowel syndrome: epidemiology,
natural history, health care seeking and emerging risk factors. Gastroenterol Clin North Am. 2005;34(2):189–204. 113. Stanghellini V, Tosetti C, Barbara G, et al. Dyspeptic symptoms and
gastric emptying in the irritable bowel syndrome. Am J Gastroenterol. 2002;97(11):2738–2743. 94. Hulshof KF, Brussaard JH, Kruizinga AG, Telman J, Löwik MR. Socio-
economic status, dietary intake and 10 y trends: the Dutch National
Food Consumption Survey. Eur J Clin Nutr. 2003;57(1):128–137. 114. Vandvik PO, Wilhelmsen I, Ihlebaek C, Farup PG. Comorbidity of irri
table bowel syndrome in general practice: a striking feature with clinical
implications. Aliment Pharmacol Ther. 2004;20(10): 1195–1203. 95. Talley NJ, Phillips SF, Bruce B, Twomey CK, Zinsmeister AR,
Melton LJ. Relation among personality and symptoms in nonulcer
dyspepsia and the irritable bowel syndrome. Gastroenterology. 1990;99(2):327–333. 115. Sperber AD, Atzmon Y, Neumann L, et al. Fibromyalgia in the irritable
bowel syndrome: studies of prevalence and clinical implications. Am
J Gastroenterol. 1999;94(12):3541–3546. 96. Locke GR 3rd, Zinsmeister AR, Talley NJ, Fett SL, Melton LJ 3rd. Familial association in adults with functional gastrointestinal disorders. Mayo Clin Proc. 2000;75(9):907–912. 116. Whitehead WE. Twin studies used to prove that the comorbidity of
major depressive disorder with IBS is NOT influenced by heredity. Am J Gastroenterol. 2007;102(10):2230–2231. 97. Levy RL, Jones KR, Whitehead WE, Feld SI, Talley NJ, Corey LA. Irritable bowel syndrome in twins: heredity and social learning both
contribute to etiology. Gastroenterology. 2001;121(4):799–804. 117. Markowitz M, Harris W, Ricci J, et al. Comorbid conditions in patients
with irritable bowel syndrome: Data from a national IBS awareness
registry. Gastroenterology. 2001;120(5 Suppl 1):A231. 98. Morris-Yates A, Talley NJ, Boyce PM, Nandurkar S, Andrews G. Evidence of a genetic contribution to functional bowel disorder. Am
J Gastroenterol. 1998;93(8):1311–1317. 118. Whitehead WE, Palsson O, Jones KR. References Aliment Pharmacol Ther. 2006;23(11):1667–1675. 60. Olafsdottir LB, Gudjonsson H, Jonsdottir HH, Bjornsson E,
Thjodleifsson B. Natural history of functional gastrointestinal disorders:
comparison of two longitudinal population-based studies. Dig Liver
Dis. 2012;44(3):211–217. 84. Lee OY, Mayer EE, Schmulson M, Chang L, Naliboff B. Gender related dif
ferences in IBS symptoms. Am J Gastroenterol. 2001;96(7): 2184–2196. 61. Okeke EN, Ladep NG, Adah S, Bupwatda PW, Agaba EI, Malu AO. Prevalence of irritable bowel syndrome: a community survey in an
African population. Ann Afr Med. 2009;8(3):177–180. 85. Rasquin A, Di Lorenzo C, Forbes D, et al. Childhood functional
gastrointestinal disorders: child/adolescent. Gastroenterology. 2006;130(5):1527–1537. 62. Wilson S, Roberts L, Roalfe A, Bridge P, Singh S. Prevalence of irritable
bowel syndrome: a community survey. Br J Gen Pract. 2004;54(504):
495–502. 86. Tang YR, Yang WW, Liang ML, Xu XY, Wang MF, Lin L. Age-related
symptom and life quality changes in women with irritable bowel
syndrome. World J Gastroenterol. 2012;18(48):7175–7183. 63. Hungin AP, Whorwell PJ, Tack J, Mearin F. The prevalence, patterns
and impact of irritable bowel syndrome: an international survey of
40,000 subjects. Aliment Pharmacol Ther. 2003;17(5):643–650. 87. Maxwell PR, Mendall MA, Kumar D. Irritable bowel syndrome. Lancet. 1997;350(9092):1691–1695. 64. Goodwin L, White PD, Hotopf M, Stansfeld SA, Clark C. Life
course study of the etiology of self-reported irritable bowel syn
drome in the 1958 British birth cohort. Psychosom Med. 2013;75(2):
202–210. 88. Marmot M, Allen J, Bell R, Bloomer E, Goldblatt P; Consortium for
the European Review of Social Determinants of Health and the Health
Divide. WHO European review of social determinants of health and
the health divide. Lancet. 2012;380(9846):1011–1029. 65. Drossman DA, Li Z, Andruzzi E, et al. US householder survey of
functional gastrointestinal disorders. Prevalence, sociodemography,
and health impact. Dig Dis Sci. 1993;38(9):1569–1580. 89. Mendall MA, Kumar D. Antibiotic use, childhood affluence and irri
table bowel syndrome (IBS). Eur J Gastroenterol Hepatol. 1998;10(1):
59–62. Clinical Epidemiology 2014:6 78 submit your manuscript | www.dovepress.com
Dovepress submit your manuscript | www.dovepress.com Epidemiology of irritable bowel syndrome Dovepress 90. Howell S, Talley NJ, Quine S, Poulton R. The irritable bowel syn
drome has origins in the childhood socioeconomic environment. Am
J Gastroenterol. 2004;99(8):1572–1578. 110. Agréus L, Svärdsudd K, Talley NJ, Jones MP, Tibblin G. Natural history
of gastroesophageal reflux disease and functional abdominal disorders: a
population based study. Am J Gastroenterol. 2001;96(10): 2905–2914. 91. Grodzinsky E, Hallert C, Faresjö T, Bergfors E, Faresjö AO. References Systematic review of the
comorbidity of irritable bowel syndrome with other disorders: what
are the causes and implications? Gastroenterology. 2002;122(4):
1140–1156. 99. Mohammed I, Cherkas LF, Riley SA, Spector TD, Trudgill NJ. Genetic influences in irritable bowel syndrome: a twin study. Am J
Gastroenterol. 2005;100(6):1340–1344. 119. Aaron LA, Buchwald D. A review of the evidence for overlap among
unexplained clinical conditions. Ann Intern Med. 2001;134(9 Pt 2):
868–881. 100. Hopper JL, Bishop DT, Easton DF. Population-based family studies
in genetic epidemiology. Lancet. 2005;366(9494):1397–1406. 120. Hamilton WT, Gallagher AM, Thomas JM, White PD. Risk markers for
both chronic fatigue and irritable bowel syndromes: a prospective case-
control study in primary care. Psychol Med. 2009;39(11): 1913–1921. 101. Ishihara S, Yashima K, Kushiyama Y, et al. Prevalence of organic
colonic lesions in patients meeting Rome III criteria for diagnosis
of IBS: a prospective multi-center study utilizing colonoscopy. J Gastroenterol. 2012;47(10):1084–1090. 121. Aggarwal VR, McBeth J, Zakrzewska JM, Lunt M, Macfarlane GJ. The epidemiology of chronic syndromes that are frequently unex
plained: do they have common associated factors? Int J Epidemiol. 2006;35(2):468–476. 102. Chey WD, Nojkov B, Rubenstein JH, Dobhan RR, Greenson JK,
Cash BD. The yield of colonoscopy in patients with non-constipated
irritable bowel syndrome: results from a prospective, controlled US
trial. Am J Gastroenterol. 2010;105(4):859–865. 122. Longstreth GF. Irritable bowel syndrome. Diagnosis in the managed
care era. Dig Dis Sci. 1997;42(6):1105–1111. 103. Akhtar AJ, Shaheen MA, Zha J. Organic colonic lesions in patients
with irritable bowel syndrome (IBS). Med Sci Monit. 2006;12(9):
CR363–CR367. 123. Przekop P, Haviland MG, Zhao Y, Oda K, Morton KR, Fraser GE. Self-reported physical health, mental health, and comorbid diseases
among women with irritable bowel syndrome, fibromyalgia, or both
compared with healthy control respondents [Erratum appears in
J Am Osteopath Assoc. Jan 2013;113(1):15]. J Am Osteopath Assoc. 2012;112(11):726–735. 104. Gu HX, Zhang YL, Zhi FC, Jiang B, Huang Y. Organic colonic lesions
in 3,332 patients with suspected irritable bowel syndrome and lacking
warning signs, a retrospective case–control study. Int J Colorectal Dis. 2011;26(7):935–940. 124. Lackner JM, Ma CX, Keefer L, et al. Type, rather than number, of
mental and physical comorbidities increases the severity of symptoms
in patients with irritable bowel syndrome. Clin Gastroenterol Hepatol. 2013;11(9):1147–1157. 105. Black TP, Manolakis CS, Di Palma JA. “Red flag” evaluation yield
in irritable bowel syndrome. J Gastrointestin Liver Dis. 2012;21(2):
153–156. 106. Clinical Epidemiology Clinical Epidemiology References Porter CK, Cash BD, Pimentel M, Akinseye A, Riddle MS. Risk of
inflammatory bowel disease following a diagnosis of irritable bowel
syndrome. BMC Gastroenterol. 2012;12:55. 125. Longstreth GF. Avoiding unnecessary surgery in irritable bowel syn
drome. Gut. 2007;56(5):608–610. 126. Kennedy TM, Jones RH. Epidemiology of cholecystectomy and irri
table bowel syndrome in a UK population. Br J Surg. 2000;87(12):
1658–1663. 107. Mearin F, Baró E, Roset M, Badía X, Zárate N, Pérez I. Clinical pat
terns over time in irritable bowel syndrome: symptom instability and
severity variability. Am J Gastroenterol. 2004;99(1):113–121. 127. Longstreth GF, Yao JF. Irritable bowel syndrome and surgery:
a multivariable analysis. Gastroenterology. 2004;126(7):1665–1673. 108. Talley NJ, Zinsmeister AR, Van Dyke C, Melton LJ 3rd. Epidemiology of
colonic symptoms and the irritable bowel syndrome. Gastroenterology. 1991;101(4):927–934. 128. Chang JY, Locke GR 3rd, McNally MA, et al. Impact of functional
gastrointestinal disorders on survival in the community. Am J
Gastroenterol. 2010;105(4):822–832. 109. Ford AC, Forman D, Bailey AG, Axon AT, Moayyedi P. Irritable bowel
syndrome: a 10-yr natural history of symptoms and factors that influ
ence consultation behavior. Am J Gastroenterol. 2008;103(5):1229–
1239; quiz 1240. 129. Tang YR, Wang P, Yin R, Ge JX, Wang GP, Lin L. Five-year follow-up
of 263 cases of functional bowel disorder. World J Gastroenterol. 2013;19(9):1466–1471. Clinical Epidemiology 2014:6 79 Dovepress Dovepress Canavan et al Dovepress submit your manuscript | www.dovepress.com Publish your work in this journal reviews, risk & safety of medical interventions, epidemiology & bio
statical methods, evaluation of guidelines, translational medicine, health
policies & economic evaluations. The manuscript management system
is completely online and includes a very quick and fair peer-review
system, which is all easy to use. Clinical Epidemiology is an international, peer-reviewed, open access
journal focusing on disease and drug epidemiology, identification of
risk factors and screening procedures to develop optimal preventative
initiatives and programs. Specific topics include: diagnosis, prognosis,
treatment, screening, prevention, risk factor modification, systematic Submit your manuscript here: http://www.dovepress.com/clinical-epidemiology-journal Clinical Epidemiology 2014:6 submit your manuscript | www.dovepress.com 80 Dovepress
|
https://openalex.org/W4361248502
|
https://www.researchsquare.com/article/rs-2744263/latest.pdf
|
English
| null |
DsRNA sequencing revealed previously missed terminal sequence of an ssRNA virus that infects dinoflagellate Heterocapsa circularisquama
|
Research Square (Research Square)
| 2,023
|
cc-by
| 1,877
|
DsRNA sequencing revealed previously missed
terminal sequence of an ssRNA virus that infects
dinoflagellate Heterocapsa circularisquama Michiko Takahashi
Kochi University
Yuichi Masuda
Kochi University
Yuto Chiba
University of Tsukuba
Syun-ichi Urayama
University of Tsukuba
Keizo Nagasaki
(
nagasaki@kochi-u.ac.jp
Kochi University Short Report License:
This work is licensed under a Creative Commons Attribution 4.0 International
License. License:
This work is licensed under a Creative Commons Attribution 4.0 International
License. Read Full License Additional Declarations: No competing interests reported. Additional Declarations: No competing interests reported. Version of Record: A version of this preprint was published at Virus Genes on January 10th, 2024. See the
published version at https://doi.org/10.1007/s11262-023-02046-3. Page 1/6 Page 1/6 Abstract Heterocapsa circularisquama RNA virus (HcRNAV) is the only dinoflagellate-infecting RNA virus cultured. However, only two strains of HcRNAV have been registered with complete genome sequences (strains 34
and 109 for UA and CY types, respectively). To extend the genomic information of HcRNAV, we performed
full-genome sequencing of an unsequenced strain of HcRNAV (strain A8) using the fragmented and
primer ligated double-stranded RNA (dsRNA) sequencing (FLDS) method. The complete genome of
HcRNAV A8 with 4457 nucleotides (nt) was successfully determined, and sequence alignment of the
major capsid protein gene suggested that A8 was a UA type strain, consistent with its intraspecific host
specificity. The complete sequence was found to be 80 nt longer at the 5’ terminus than the registered
sequences of HcRNAV strains, suggesting that FLDS is more reliable for determining the terminal
sequence than conventional methods (5’ Rapid Amplification of cDNA End). Our study contributes to a
better understanding of dinoflagellate-infecting viruses with limited sequence data. Full Text Although dinoflagellates are abundant and important components of aquatic environments, a few
dinoflagellate-infecting viruses have been isolated from the environment [1]. Heterocapsa
circularisquama RNA virus (HcRNAV) is the only dinoflagellate-infecting RNA virus isolated to date [2]. HcRNAV has a genome size of 4.3 kb and is divided into two types (UA and CY) based on intraspecies
host infectivity as well as the amino acid sequence of the major capsid protein (MCP) gene [2,3]. At
present, more than 100 sequences of the MCP gene of HcRNAV have been deposited in the NCBI
nucleotide database, whereas the full-genome sequences have been registered for only two strains:
strains 34 and 109 (accession numbers AB218608.1 and AB218609.1, respectively). In the present study, we applied fragmented and primer-ligated dsRNA sequencing (FLDS) to HcRNAV to
obtain its full genomic sequence. FLDS enables complete genome sequencing of both dsRNA and single-
stranded RNA (ssRNA) viruses with high-coverage sequencing, including that of the terminal region,
allowing for the identification of viral quasi-species [4,5]. Additionally, an optimized protocol for FLDS ver. 3 that can be applied to low amounts of dsRNA (10 pg) has been reported [6]; thus, we used FLDS ver. 3
for an unsequenced HcRNAV strain. The HcRNAV strain A8 was isolated in Ago Bay, Mie Prefecture, Japan, in August 2001 (Tomaru Y. personal communication). A cross-reactivity test confirmed that this strain showed the same intraspecies
host specificity as the UA-type (Table S1). The host culture (HU9433-P) was maintained in Daigo’s IMK
Medium (Nihon Pharmaceutical Co., Ltd., Tokyo, Japan) supplemented with 0.2% soil extract under a
14/10-h light/dark fluorescence photocycle (light intensity: 345 μmol photons m-2 s-1) at 22°C. To prepare
the HcRNAV lysate, 200 mL of HcRNAV A8 at an infectivity of 105–106 MPN (most probable number) mL-
1 was mixed with 200 mL of H. circularisquama culture at log phase (104 cells mL-1) and then incubated
for 24 h under the same conditions described above. The viral lysate was filtered through a 0.2 µm pore-
sized polycarbonate filter (Cytiva Whatman, Maidstone, UK). The filters on which the infected cells were Page 2/6 collected were disrupted using liquid nitrogen in a sterilized mortar, and total nucleic acids were extracted
using SDS-phenol. DsRNA was purified using cellulose resin, as described previously [4,6] then treated with DNase I
(Invitrogen, Carlsbad, CA, US) and S1 nuclease (Invitrogen). The dsRNA samples were converted to cDNA
using FLDS ver. Data availability The complete genome sequence of HcRNAV A8 has been deposited in DDBJ/EMBL/GenBank database
under the accession number LC753702. Raw data are available in the DDBJ Sequence Read Archive
(DRA) under the accession number DRR453392. Full Text 3 [6], and the constructed cDNA library was sequenced on an Illumina NovaSeq platform
(Illumina, CA, USA). The sequenced reads were processed as described previously [7]; raw sequencing
data were processed using a custom Perl script (https://github.com/takakiy/FLDS); adapter, low quality,
low complexity, PhiX, and rRNA sequences were removed. Preprocessed reads were subjected to de
novo assembly using the CLC Genomics Workbench (version 11.0; CLC Bio, Aarhus, Denmark). The
resulting assemblies were manually examined and extended using a Tablet viewer [8]. The terminal end
was recognized as the region where more than ten contigs were aligned. Full-length RNA sequence of
HcRNAV A8 was aligned with HcRNAV34 (Accession number: AB218608.1) and HcRNAV109
(AB218609.1) in MEGAX [9]. The full-length genome sequence of HcRNAV A8 was successfully determined to be 4,457 nucleotides
(nt) long, which was 80 nt longer than the registered sequences of HcRNAV (34 and 109) (Fig. 1, Table
S2). In the FLDS, genomic 5’ terminal sequence was determined by 5’ RACE and 3’ RACE, for the positive
strand, and the negative strand, respectively. Thus, the 5’ terminal end of HcRNAV may have been
obtained. We considered two scenarios: 1) genome size of HcRNAV varies for different strains and 2) the
registered strains (HcRNAV34 and 109) were not amplified up to the 5’ terminal sequence by the rapid
amplification of cDNA ends (RACE) method [3]. To determine this, it is necessary to obtain more full-
length sequences derived from HcRNAV isolates. A BLASTn analysis indicated that HcRNAV A8 showed
99.7% and 97.1% of nucleotide identity to HcRNAV34, and HcRNAV109, respectively. Sequence alignment
of the MCP gene showed that A8 is homologous to UA-type strains, consistent with intraspecies host
specificity (Fig. 2, Table S1). In summary, a full-genome sequence, including the termini of HcRNAV, was
determined using FLDS. Our data may accelerate the understanding of dinoflagellate-infecting viruses
with limited whole-sequence information. Competing Interests The authors declare that they have no conflicts of interest. This article does not contain any studies
involving human participants or animals performed by any of the authors. Author Contributions MT and KN conceived, designed, and obtained funding for this work. MT, YM, and YC performed the data
collection and analysis. SU and KN supervised this project. The first draft of the manuscript was written
by MT, and all the authors commented on the previous versions of the manuscript. All authors have read
and approved the final manuscript. Funding This work was supported by the Ministry of Education, Culture, Sports, Science and Technology of Japan
(Grant Numbers JP16H06429, JP16K21723, and JP16H06437), the Japan Society for the Promotion of
Science (Grant number: JP19J00346), ACT-X, Japan Science and Technology Agency (Grant number
JPMJAX21BD), and Kochi University Hospital. Acknowledgments We would like to thank Dr. Yuji Tomaru of the Japan Fisheries Research and Education Agency for
providing HcRNAV strains. We are grateful to Drs. Yoshitake Takao and Daisuke Honda for providing
AuRNAV (RNA virus that infects Thraustochytrids), which was used as a positive control in this study. Page 3/6 8. I. Milne, M. Bayer, L. Cardle, et al, Bioinformatics, Tablet--Next generation sequence assembly
visualization. 26, 401–402 (2010). https://doi.org/10.1093/bioinformatics/btp666
9. S. Kumar, G. Stecher, M. Li, C. Knyaz, K. Tamura, Mol. Biol. Evol., MEGA X: Molecular evolutionary
genetics analysis across computing platforms. 35, 1547–1549 (2018).
https://doi.org/10.1093/molbev/msy096 8. I. Milne, M. Bayer, L. Cardle, et al, Bioinformatics, Tablet--Next generation sequence assembly
visualization. 26, 401–402 (2010). https://doi.org/10.1093/bioinformatics/btp666
9. S. Kumar, G. Stecher, M. Li, C. Knyaz, K. Tamura, Mol. Biol. Evol., MEGA X: Molecular evolutionary
genetics analysis across computing platforms. 35, 1547–1549 (2018). https://doi.org/10.1093/molbev/msy096 8. I. Milne, M. Bayer, L. Cardle, et al, Bioinformatics, Tablet--Next generation sequence assembly
visualization. 26, 401–402 (2010). https://doi.org/10.1093/bioinformatics/btp666 8. I. Milne, M. Bayer, L. Cardle, et al, Bioinformatics, Tablet--Next generation sequence assembly
visualization. 26, 401–402 (2010). https://doi.org/10.1093/bioinformatics/btp666 8. I. Milne, M. Bayer, L. Cardle, et al, Bioinformatics, Tablet--Next generation sequence assembly
visualization. 26, 401–402 (2010). https://doi.org/10.1093/bioinformatics/btp666
9. S. Kumar, G. Stecher, M. Li, C. Knyaz, K. Tamura, Mol. Biol. Evol., MEGA X: Molecular evolutionary
genetics analysis across computing platforms. 35, 1547–1549 (2018). https://doi.org/10.1093/molbev/msy096
Figures
Figure 1
Nucleotides (120) at the 5’-terminal of HcRNAV A8 aligned with reference sequences of HcRNAV (34 and
109) 9. S. Kumar, G. Stecher, M. Li, C. Knyaz, K. Tamura, Mol. Biol. Evol., MEGA X: Molecular evolutionary
genetics analysis across computing platforms. 35, 1547–1549 (2018). https://doi.org/10.1093/molbev/msy096 9. S. Kumar, G. Stecher, M. Li, C. Knyaz, K. Tamura, Mol. Biol. Evol., MEGA X: Molecular evolutionary
genetics analysis across computing platforms. 35, 1547–1549 (2018). https://doi.org/10.1093/molbev/msy096 9. S. Kumar, G. Stecher, M. Li, C. Knyaz, K. Tamura, Mol. Biol. Evol., MEGA X: Molecular evolutionary
genetics analysis across computing platforms. 35, 1547–1549 (2018). https://doi.org/10.1093/molbev/msy096 References 1. J. H. Wisecaver, J. D. Hackett, Annu. Rev. Microbiol., Dinoflagellate genome evolution. 65, 369–387
(2011). https://doi.org/10.1146/annurev-micro-090110-102841 2. Y. Tomaru, N. Katanozaka, K. Nishida, et al, Aquat. Microb. Ecol., Isolation and characterization of
two distinct types of HcRNAV, a single-stranded RNA virus infecting the bivalve-killing microalga
Heterocapsa circularisquama. 34, 207–218 (2004). https://doi.org/10.3354/ame034207 3. K. Nagasaki, Y. Shirai, Y. Takao, et al, Appl. Environ. Microbiol., Comparison of genome sequences of
single-stranded RNA viruses infecting the bivalve-killing dinoflagellate Heterocapsa circularisquama. 71, 8888–8894 (2005). https://doi.org/10.1128/AEM.71.12.8888-8894.2005 4. S.-I. Urayama, Y. Takaki, T. Nunoura, Microbes Environ., FLDS: A comprehensive dsRNA sequencing
method for intracellular RNA virus surveillance. 31, 33–40 (2016). https://doi.org/10.1264/jsme2.ME15171 5. T. Izumi, Y. Morioka, S. I. Urayama, et al, Viruses, DsRNA sequencing for RNA virus surveillance using
human clinical samples. 13, 1310 (2021). https://doi.org/10.3390/v13071310 6. M. Hirai, Y. Takaki, F. Kondo, et al, Microbes Environ., RNA viral metagenome analysis of
subnanogram dsRNA using fragmented and primer ligated dsRNA sequencing (FLDS). 36:. https:
(2021). https://doi.org/10.1264/jsme2.ME20152 7. Y. Chiba, S. Oiki, T. Yaguchi, S. I. Urayama, D. Hagiwara, Virus Evol., Discovery of divided RdRp
sequences and a hitherto unknown genomic complexity in fungal viruses. 7:. https:, veaa101 (2021). https://doi.org/10.1093/ve/veaa101 Page 4/6 Page 4/6 Figure 2 Amino acid alignment of the major capsid protein gene of A8, and registered strainsThe black bars
indicate the variable regions at which complete substitution was observed between the UA and CY types. Amino acid alignment of the major capsid protein gene of A8, and registered strainsThe black bars
indicate the variable regions at which complete substitution was observed between the UA and CY types. Figures Figure 1
Nucleotides (120) at the 5’-terminal of HcRNAV A8 aligned with reference sequences of HcRNAV (34 and
109) Nucleotides (120) at the 5’-terminal of HcRNAV A8 aligned with reference sequences of HcRNAV (34 and
09) Nucleotides (120) at the 5’-terminal of HcRNAV A8 aligned with reference sequences of HcRNAV (34 and
109) Nucleotides (120) at the 5’-terminal of HcRNAV A8 aligned with reference sequences of HcRNAV (34 and
109) Page 5/6 Figure 2
Amino acid alignment of the major capsid protein gene of A8, and registered strainsThe black bars
indicate the variable regions at which complete substitution was observed between the UA and CY types Supplementary Files This is a list of supplementary files associated with this preprint. Click to download. Supplementarydata.docx Page 6/6 Page 6/6
|
https://openalex.org/W4315641262
|
https://www.frontiersin.org/articles/10.3389/fmed.2022.1080892/pdf
|
English
| null |
Deep anterior lamellar keratoplasty following thermokeratoplasty assisted epikeratophkia: A novel two-stage one-graft method to treat acute corneal hydrops
|
Frontiers in medicine
| 2,023
|
cc-by
| 5,720
|
OPEN ACCESS OPEN ACCESS
EDITED BY
Michele Lanza,
University of Campania Luigi Vanvitelli, Italy
REVIEWED BY
Alejandro Tello,
Industrial University of Santander, Colombia
Alessandro Meduri,
University of Messina, Italy
Majid Moshirfar,
University of Utah, United States
Hua-Tao Xie,
Huazhong University of Science
and Technology, China
*CORRESPONDENCE
Li Zhang
trudyzhang@163.com
Yanlong Bi
biyanlong@tongji.edu.cn
†These authors have contributed equally to this
work and share first authorship
SPECIALTY SECTION
This article was submitted to
Ophthalmology,
a section of the journal
Frontiers in Medicine
RECEIVED 26 October 2022
ACCEPTED 29 December 2022
PUBLISHED 12 January 2023
CITATION
Liu C, Huang X, Shen J, Zhang Y, Zhang L and
Bi Y (2023) Deep anterior lamellar keratoplasty
following thermokeratoplasty assisted
epikeratophkia: A novel two-stage one-graft
method to treat acute corneal hydrops. Front. Med. 9:1080892. doi: 10.3389/fmed.2022.1080892 OPEN ACCESS
EDITED BY
Michele Lanza,
University of Campania Luigi Vanvitelli, Italy
REVIEWED BY
Alejandro Tello,
Industrial University of Santander, Colombia
Alessandro Meduri,
University of Messina, Italy
Majid Moshirfar,
University of Utah, United States
Hua-Tao Xie,
Huazhong University of Science
and Technology, China
*CORRESPONDENCE
Li Zhang
trudyzhang@163.com
Yanlong Bi
biyanlong@tongji.edu.cn
†These authors have contributed equally to this
work and share first authorship
SPECIALTY SECTION
This article was submitted to
Ophthalmology,
a section of the journal
Frontiers in Medicine
RECEIVED 26 October 2022
ACCEPTED 29 December 2022
PUBLISHED 12 January 2023
CITATION
Liu C, Huang X, Shen J, Zhang Y, Zhang L and
Bi Y (2023) Deep anterior lamellar keratoplasty
following thermokeratoplasty assisted
epikeratophkia: A novel two-stage one-graft
method to treat acute corneal hydrops. Front. Med. 9:1080892. doi: 10.3389/fmed.2022.1080892 Chunyu Liu†, Xinyu Huang†, Jiaqi Shen, Yushan Zhang, Li Zhang*
and Yanlong Bi* Chunyu Liu†, Xinyu Huang†, Jiaqi Shen, Yushan Zhang, Li Zhang*
and Yanlong Bi* Department of Ophthalmology, School of Medicine, Shanghai Tongji Hospital, Tongji University, Shanghai,
China Department of Ophthalmology, School of Medicine, Shanghai Tongji Hospital, Tongji University, Shanghai,
China Purpose: To evaluate the clinical effects of deep anterior lamellar keratoplasty
(DALK) using a single graft after thermokeratoplasty assisted epikeratophakia for the
treatment of acute corneal hydrops. Methods: This novel surgical procedure was performed on seven eyes of seven
patients between 2019 and 2020. The procedure combines a first-stage surgery
of thermokeratoplasty assisted epikeratophkia with intracameral sterile air injection
and a second-stage surgery of DALK using the same corneal graft for both
procedures. TYPE Original Research
PUBLISHED 12 January 2023
DOI 10.3389/fmed.2022.1080892 TYPE Original Research
PUBLISHED 12 January 2023
DOI 10.3389/fmed.2022.1080892 TYPE Original Research
PUBLISHED 12 January 2023
DOI 10.3389/fmed.2022.1080892 acute corneal hydrops, epikeratophakia, deep anterior lamellar keratoplasty, anterior
segment optical coherence tomography, Descemet membrane OPEN ACCESS Main outcome measures included pre- and postoperative corrected
distance visual acuity (CDVA) and anterior segment optical coherence tomography
(AS-OCT) parameters. Corneal transparency, epithelization, and the presence of
neovascularization, were evaluated at the 1-year follow-up visit. Results: Corneal edema resolved rapidly in six of the seven cases. The mean central
corneal thickness was significantly reduced from baseline to 1 day, 1 week, 1 month,
and 2 months after the first-stage surgery (P < 0.0001). At a mean of 2.1 ± 0.7 months
after the first-stage surgery, DALK was successfully performed in all cases. Six months
later, the mean central corneal thickness was 611 ± 31 µm and the mean thickness of
the recipient’s residual stroma bed was 20 ± 6 µm at the central corneal area. Mean
LogMAR CDVA improved from 1.74 ± 0.34 at baseline to 0.20 ± 0.11 after DALK
(P < 0.0001). No postoperative complications appeared in our case series during the
1-year observation period. doi: 10.3389/fmed.2022.1080892 COPYRIGHT © 2023 Liu, Huang, Shen, Zhang, Zhang and Bi. This is an open-access article distributed under
the terms of the Creative Commons Attribution
License (CC BY). The use, distribution or
reproduction in other forums is permitted,
provided the original author(s) and the
copyright owner(s) are credited and that the
original publication in this journal is cited, in
accordance with accepted academic practice. No use, distribution or reproduction is
permitted which does not comply with
these terms. Conclusion: Very good visual results were obtained with a novel technique
(thermokeratoplasty assisted epikeratophakia followed by DALK using the same
corneal graft) in the treatment of acute corneal hydrops. 01 Frontiers in Medicine frontiersin.org 10.3389/fmed.2022.1080892 10.3389/fmed.2022.1080892 Liu et al. Liu et al. Surgical procedure All surgeries were performed by the same surgeon (YL Bi). The
first-stage surgery was thermokeratoplasty assisted epikeratophakia
combined with injection of sterile air into the anterior chamber. The graft bed was prepared firstly, followed by cauterization of the
edematous cornea using bipolar electrocautery forceps. During this
procedure, the epithelium of the pathological cornea was cauterized
and stripped, and the subepithelial tissue was also cauterized until
the pathological cornea was visibly flatter. At this step, the heat
of the cauterization was controlled within the appropriate range
to minimize corneal stromal scarring and damage to the corneal
endothelium. The corneal cone apex position was determined by slit
lamp microscopy and corneal topography before surgery (the mean
cone bottom diameter of this group of cases was 8.51 ± 0.26 mm). A partial thickness trephination was then made to a depth of about
200 µm with a vacuum trephine of corresponding dimension, and a
peripheral pocket was created for insertion of the edge of the donor
graft. The prepared donor graft was a decellularized homograft which
had been stored at –78◦C in pure sterile glycerin, and 1.5 mm larger
than the previous trephine (the average of this group of cases is
10.07 ± 0.32 mm) was then sutured in place with interrupted 10-
0 nylon. A paracentesis of the anterior chamber was made at the 9
o’clock position to decrease the intraocular pressure, and about 0.1 ml
sterile air was then injected into the anterior chamber to compress
the break in the pathological DM (Figures 1A, B). The eye under
treatment was covered with a bandage contact lens and a compression
bandage. The patient was instructed to rest in the supine position as
much as possible after surgery, avoiding strenuous exercise and eye
rubbing. Introduction had previously undergone corneal surgery or been treated with
rigid gas permeable (RGP) contact lenses. All patients underwent
relevant ophthalmic examinations pre and postoperatively, including
slit-lamp examination and CDVA. The extent of corneal edema,
corneal thickness, the location and size of the DM breaches were
evaluated by anterior segment optical coherence tomography (AS-
OCT) examination. Keratoconus is a chronic ectatic corneal disorder whose etiology,
which seems to involve the inflammatory cascade, and is related to
mechanical trauma from the habit of rubbing the eyes and nocturnal
ocular compression (1, 2), is still the subject of extensive debate, is
characterized by the presence of irregular astigmatism and corneal
thinning, which usually progress during the first three decades of life
(3, 4). Acute corneal hydrops (ACH), which may result in corneal
scars and vision loss, is an uncommon complication of keratoconus
secondary to rapid stromal edema, and even intrastromal clefts
(pseudocysts) formation, secondary to Descemet membrane rupture
(5, 6). ACH is usually self-limiting and generally resolves without
intervention over 2–4 months (7), and may preset complications such
as neovascularization (8), infection, malignant glaucoma, and corneal
perforation (9, 10). Surgical interventions has been suggested in order
to in order to accelerate the resolution of corneal edema. A history of
ACH may also increase the risk of endothelial graft rejection after
penetrating keratoplasty (11). Deep anterior lamellar keratoplasty
(DALK) is now the preferred method to treat deep corneal stromal
lesions. During DALK, the healthy endothelium of the recipient
cornea can be preserved, avoiding postoperative endothelial immune
rejection (12) and chronic endothelial dysfunction (13). The graft
is cell-extracted and preserved by freezing in glycerol, so the
risk of stromal immune rejection is also low after DALK (14). However, when ACH arises in keratoconus, disruption of Descemet’s
membrane (DM) and the endothelium may create difficulties in
performing a complete DALK (11, 15). In this study, we designed
a novel and safe two-stage surgical procedure. The first stage
involves thermokeratoplasty assisted epikeratophakia combined with
intracameral air injection to accelerate the ACH healing and the
second stage is the DALK procedure using the same graft. Ethical approval This
retrospective,
interventional,
non-randomized
observational study was approved by the ethics committee of
the Shanghai Tongji Hospital of Tongji University, Shanghai,
China. After a thorough explanation about the nature of the study,
all patients agreed to participate and provided written informed
consents to participate prior to study entry. Signed consent was
obtained from patients for all clinical photographs that permit
their identification and is archived by the authors. Furthermore,
this study conformed to the ethical standards outlined in the
Declaration of Helsinki. The second-stage surgery was DALK. About 2 months after the
epikeratophakia, DALK was performed using the same corneal graft. The edge of the peripheral stromal pocket and the edge of the
donor graft were identified and the donor graft was then completely
dissociated and temporarily preserved in an empty moist chamber. After initial trephination, the incisal edge was held using fine toothed
forceps and was dissected at about 4/5 thickness. During this process,
the manual dissections were performed carefully three to five times
in most cases to avoid further damaging the DM. The Anwar bubble
technique should never be used on such patients. When the anterior
stroma was dissected (at approximately 50% thickness), a sterile air
bubble was injected into the anterior chamber to avoid rapid aqueous
leakage from the reopening of the healed DM breaks and its further
edema. If modest leakage was observed from the stromal bed, the
manual dissection was conducted more carefully and quickly. In all of
our cases, the same donor grafts were used in the second-stage DALK. The graft was reversed, placed, and trimmed on the cutting-offtable
using a trephine 0.5 mm larger than the recipient bed. During this
procedure the corneal epithelium was carefully protected. The donor
graft was then fixed using 10-0 interrupted nylon sutures (Figure 1C). Frontiers in Medicine Results All
seven
surgeries
were
performed
successfully
without
intraoperative or postoperative complications. After the first stage
surgery, the DM was reattached to the posterior corneal stroma in
all patients. In six patients, the corneal edema subsided rapidly 1
day after surgery and gradually decreased over the next 2 weeks
(Table 1). Edema in one case took 5 days to resolve due to a large
break in the DM (Figure 2). One patient, who had been treated with
thermokeratoplasty for emergency management upon diagnosis of
ACH had a recurrence of edema 2 weeks after surgery and had to
undergo epikeratophakia to release the corneal hydrops (Figure 3). Before the second-stage surgery, all breaks in the DM were almost
healed and the corneal stroma was in a non-edematous condition,
based on the corneal morphologic parameters measured using AS-
OCT. In this case series, the mean period was 2.1 ± 0.7 months from
the completion of epikeratophakia to DALK. During the second stage
DALK, the pre-DM corneal stroma was exposed as much as possible. In two cases, slow aqueous leakage was observed in the deep corneal
stroma when the manual dissection was performed close to the DM. In addition, mild effusion between graft and bed was observed in
one patient 1 day after DALK. After topical anesthesia, a lacrimal
irrigating needle was used to gently open the wound between the
two sutures at 5 o’clock. The wound was pressed down gently, the
interlaminar fluid could be seen flowing out, the double anterior sign
disappeared immediately, and the corneal edema decreased gradually
within 5 days without recurrence. Schematic drawing of the two-stage one-graft method. (A) Acute
corneal hydrops in advanced keratoconus due to rupture of
Descemet’s membrane and subsequent intrastromal fluid
accumulation. (B) First-stage surgery. The pathological cornea was
covered and pressed with an allogenic donor graft and the detached
DM was pushed upward by an injected sterile air bubble. The
disrupted DM was allowed to reset and heal, and the severe edema
subsided rapidly. (C) Approximately 6 months later, the second-stage
DALK was performed using the same donor graft. effusion between graft and bed was observed after DALK. After
topical anesthesia, a lacrimal irrigating needle was used to gently open
the wound between the two sutures at 5 o’clock. The wound was
pressed down gently to help the interlaminar fluid flowing out, the
double anterior sign disappeared immediately. Discussion ACH is a severe complication of keratoconus, caused by
disruption and local detachment of the DM (5, 6). Due to its
elasticity, DM retracts or coils when it breaks under tension, and
the aqueous humor flows into the corneal stroma, causing severe
corneal edema and opacity, and even leading to the formation of
intrastromal clefts (pseudocysts). Classic medical treatment includes
the use of hypertonic sodium chloride and cycloplegics, compressive
bandaging (16, 17), and surgical treatments such as injection of inert
gas (C3F8 or SF6) into the anterior chamber (18), corneal cross-
linking, compressive sutures (19, 20), epikeratophakia or lamellar
keratoplasty (21, 22). The severe corneal edema caused by large
breaks in DM also limits the likelihood of a complete and successful
DALK, which is now the preferred treatment for keratoconus (23). Liu et al. Liu et al. deviation. After performed the first-stage surgery, one way ANOVA
was used to test the Ta, Tr, Tnh, Tth, Tnv, and Ttv values at
preoperative and different time points after surgery (1 day, 1 week,
1 month and 2 month). Before and 6 months after DALK, a
paired samples t-test was used to analyze the preoperative and
postoperative Ta, Tnh, Tth, Tnv, and Ttv values, with P < 0.05
considered statistically significant. FIGURE 1
Schematic drawing of the two-stage one-graft method. (A) Acute
corneal hydrops in advanced keratoconus due to rupture of
Descemet’s membrane and subsequent intrastromal fluid
accumulation. (B) First-stage surgery. The pathological cornea was
covered and pressed with an allogenic donor graft and the detached
DM was pushed upward by an injected sterile air bubble. The
disrupted DM was allowed to reset and heal, and the severe edema
subsided rapidly. (C) Approximately 6 months later, the second-stage
DALK was performed using the same donor graft. Results At 6 months after DALK, the CDVA was restored to LogMAR
0.20 ± 0.11 in this group of cases, and the graft-recipient bed was
well attached on observation of AS-OCT images (Figure 4). The
central and paracentral corneal thickness were significantly reduced
compared with pre-operative measures, and the thickness of central
residual corneal stroma was 20 ± 6 µm (Table 2). AS-OCT examination The corneal morphologic examination was performed using
AS-OCT (Carl Zeiss Meditec, USA). Preoperative measurements
included the thickness of the corneal apex (Ta), the corneal thickness
2 mm nasal and temporal to the corneal apex, as determined by
horizontal (Tnh, Tth) and vertical (Tnv, Ttv) scans (Figure 2). If the
DM was detached, the width of the DM detachment was measured
using different scan directions, and the maximum distance from
the detached membrane to the posterior corneal surface was also
determined. The above measurements and the thickness of the central
recipient bed (Tr) were measured at 1 day, 1 week, 1 month, and
2 months after the first-stage surgery and again at 6 months after the
second-stage surgery. Frontiers in Medicine Patients In this study, the novel procedure was performed on seven
eyes of seven patients between February 2019 and May 2021 at the
department of ophthalmology affiliated with the Tongji University
School of Medicine. The patients were aged between 17 and
21 years (mean 18.9 ± 1.4 years), and all were diagnosed with
ACH immediately after onset of symptoms. Among the cases, the
mean ACH duration was 25.9 ± 6.1 days. None of the patients 02 Frontiers in Medicine frontiersin.org 10.3389/fmed.2022.1080892 Statistical analysis Data were analyzed statistically using SPSS Version 21.0 (SPSS
Inc., IBM, USA), and the results were expressed as mean ± standard Frontiers in Medicine Frontiers in Medicine 03 frontiersin.org Liu et al. Liu et al. 10.3389/fmed.2022.1080892 FIGURE 2
Measurement of corneal thickness using anterior segment optical coherence tomography (AS–OCT) after acute corneal hydrops. Indicated from left to
right are the corneal thickness 2 mm nasal to the corneal apex with a horizontal scan, the corneal thickness at the apex, and the corneal thickness 2 mm
temporal to the corneal apex with a horizontal scan. The yellow arrow indicates a break in Descemet’s membrane. FIGURE 2
Measurement of corneal thickness using anterior segment optical coherence tomography (AS–OCT) after acute corneal hydrops. Indicated from left to
right are the corneal thickness 2 mm nasal to the corneal apex with a horizontal scan, the corneal thickness at the apex, and the corneal thickness 2 mm
temporal to the corneal apex with a horizontal scan. The yellow arrow indicates a break in Descemet’s membrane. ABLE 1 AS-OCT results preoperatively and 2 months after the first-stage surgery. TABLE 1 AS-OCT results preoperatively and 2 months after the first-stage surgery. Pre-operation
1 day
1 week
1 month
2 months
Ta (µm)
1, 870 ± 243
1, 373 ± 221*
1, 186 ± 129*
1, 157 ± 131*
1, 063 ± 125*
Tr (µm)
/
778 ± 120
603 ± 72#
567 ± 21#
520 ± 10#
Tnh (µm)
1, 368 ± 244
1, 284 ± 216*
1, 153 ± 154*
1, 124 ± 142*
1, 042 ± 97*
Tth (µm)
1, 412 ± 263
1, 307 ± 239*
1, 172 ± 150*
1, 136 ± 145*
1, 050 ± 101*
Tnv (µm)
1, 344 ± 269
1, 221 ± 218*
1, 130 ± 180*
1, 104 ± 171*
1, 033 ± 138*
Ttv (µm)
1, 238 ± 258
1, 189 ± 238*
1, 118 ± 200*
1, 080 ± 173*
1, 021 ± 129*
Ta, corneal apex thickness; Tr, central recipient bed thickness; Tnh, corneal thickness 2 mm nasal to the corneal apex, determined with a horizontal scan; Tth, corneal thickness 2 mm temporal to
the corneal apex, determined with a horizontal scan; Tnv, corneal thickness 2 mm nasal to the corneal apex, determined with a vertical scan; Ttv, corneal thickness 2 mm temporal to the corneal
apex, determined with a vertical scan. Statistical analysis *P value < 0.05 compared with the preoperative values. #P value < 0.05 compared with the values measured 1 day after surgery. Ta, corneal apex thickness; Tr, central recipient bed thickness; Tnh, corneal thickness 2 mm nasal to the corneal apex, determined with a horizontal scan; Tth, corneal thickness 2 mm temporal to
the corneal apex, determined with a horizontal scan; Tnv, corneal thickness 2 mm nasal to the corneal apex, determined with a vertical scan; Ttv, corneal thickness 2 mm temporal to the corneal
apex, determined with a vertical scan. *P value < 0.05 compared with the preoperative values. #P value < 0.05 compared with the values measured 1 day after surgery. stromal dissection. In the 6 months of follow-up after DALK, AS-
OCT examinations showed corneal (stroma plus DM) thickness and
CDVA values comparable to those following DALK in keratoconus
patients without ACH (24). We recognized that repairing the break in DM would allow the
corneal edema to subside and the corneal stroma to heal, and could
create conditions supportive of subsequent lamellar keratoplasty
and even DALK. We therefore proposed this novel and effective
procedure which combined epikeratophakia with intracameral air
injection and DALK using the same corneal graft. Intracameral
air injection is an effective therapy that can shorten the period
of corneal edema (22). In the first-stage surgery, the pathological
cornea was covered with an allogenic donor graft and an air bubble
was inserted into the anterior chamber to push the detached DM
back and adjacent to the stroma. This contributed to rapid closure
of the DM wound, after which the severe edema of the recipient
graft bed subsided within 1–2 days, as confirmed by the AS-OCT
examination. Six months later, the DM repair was strengthened by
new collagen deposits secreted by the surrounding healthy corneal
endothelial cells (23), providing a solid structural foundation for the
second-stage DALK surgery in the present study, including manual To
our
knowledge,
this
two-step
procedure
of
thermokeratoplasty
assisted
epikeratophakia
with
intracameral
air injection and DALK has not been reported previously in the
medical literature. Currently, sustainable eye banks remain rare in
developing countries, and a shortage of donor corneas is the most
common problem faced by the corneal surgeon (25). Frontiers in Medicine frontiersin.org Statistical analysis The novel
procedure we have proposed consists of two keratoplasties, but
the same donor graft is used for both procedures, which therefore
imposes no additional donor burden and technically avoids the
need for a fresh donor cornea for penetrating keratoplasty with
associated complications such as endothelial rejection and chronic
endothelial dysfunction (26, 27). In addition, about 2 months after
epikeratophakia, the donor graft epithelium had also reepithelized
from the limbus of the recipient cornea. Consequently, transplanting 04 Frontiers in Medicine frontiersin.org Liu et al. 10.3389/fmed.2022.1080892 FIGURE 3
A patient with ACH treated with the epikeratophakia–DALK combination after a first treatment with thermokeratoplasty. (A) Acute onset ACH, with severe
edema of the corneal stroma. (B) At 2 weeks after thermokeratoplasty a recurrence of edema was evident centrally, and epikeratophakia was performed
immediately. (C,D) At 2 days after epikeratophakia, the corneal haze and edema had noticeably subsided. (E) At 3 months after DALK using the same
graft, the cornea was transparent. FIGURE 4
Slit-lamp photographs and anterior segment optical coherence tomography (AS–OCT) scans show the evolution of a patient with ACH treated using the
novel techniques. (A,E) Acute onset ACH, with severe edema of the corneal stroma evident under slit-lamp examination and with a DM break and
detachment observed by AS-OCT. (B,F) One day after the first-stage surgery, at presentation, with double layers of cornea observed on slit-lamp and
AS-OCT examinations, with no corneal edema and with DM attached. (C,G) One week after the second-stage surgery, the corneal graft, which was the
same graft used in the first-stage operation, had survived and attached to the host’s residual stroma bed with a healed DM and corneal endothelial. (D,H)
One year after the second stage surgery, the cornea was transparent and the logMAR CDVA was 0.096. the same donor graft in the-second stage DALK surgery also
avoided the challenge of regenerating a new corneal epithelial layer,
promoting rapid graft healing (28). At 2 months after epikeratophakia, 1 eye of 1 patient in this
thoroughly dissection during the DALK procedure, and the recipien
corneal stroma bed remained transparent during subsequent follow
up. In conclusion, we propose a novel two-stage procedure for FIGURE 3
A patient with ACH treated with the epikeratophakia–DALK combination after a first treatment with thermokeratoplasty. (A) Acute onset ACH, with severe
edema of the corneal stroma. Statistical analysis (B) At 2 weeks after thermokeratoplasty a recurrence of edema was evident centrally, and epikeratophakia was performed
immediately. (C,D) At 2 days after epikeratophakia, the corneal haze and edema had noticeably subsided. (E) At 3 months after DALK using the same
graft, the cornea was transparent. FIGURE 3
A patient with ACH treated with the epikeratophakia–DALK combination after a first treatment with thermokeratoplasty. (A) Acute onset ACH, with severe
edema of the corneal stroma. (B) At 2 weeks after thermokeratoplasty a recurrence of edema was evident centrally, and epikeratophakia was performed
mmediately. (C,D) At 2 days after epikeratophakia, the corneal haze and edema had noticeably subsided. (E) At 3 months after DALK using the same
graft, the cornea was transparent. A patient with ACH treated with the epikeratophakia–DALK combination after a first treatment with thermokeratoplasty. (A) Acute onset ACH, with severe
edema of the corneal stroma. (B) At 2 weeks after thermokeratoplasty a recurrence of edema was evident centrally, and epikeratophakia was performed
immediately. (C,D) At 2 days after epikeratophakia, the corneal haze and edema had noticeably subsided. (E) At 3 months after DALK using the same
graft, the cornea was transparent. FIGURE 4
Slit-lamp photographs and anterior segment optical coherence tomography (AS–OCT) scans show the evolution of a patient with ACH treated using the
novel techniques. (A,E) Acute onset ACH, with severe edema of the corneal stroma evident under slit-lamp examination and with a DM break and
detachment observed by AS-OCT. (B,F) One day after the first-stage surgery, at presentation, with double layers of cornea observed on slit-lamp and
AS-OCT examinations, with no corneal edema and with DM attached. (C,G) One week after the second-stage surgery, the corneal graft, which was the
same graft used in the first-stage operation, had survived and attached to the host’s residual stroma bed with a healed DM and corneal endothelial. (D,H)
One year after the second stage surgery, the cornea was transparent and the logMAR CDVA was 0.096. FIGURE 4
Slit-lamp photographs and anterior segment optical coherence tomography (AS–OCT) scans show the evolution of a patient with ACH treated using the
novel techniques. (A,E) Acute onset ACH, with severe edema of the corneal stroma evident under slit-lamp examination and with a DM break and
detachment observed by AS-OCT. Frontiers in Medicine Data availability statement TABLE 2 AS-OCT results preoperatively and 6 months after the second-stage surgery. Case
Sex
Age
Eye
logMAR CDVAPre/Post
Ta (µm)Pre/Post
Tnh (µm)Pre/Post
Tth (µm)Pre/Post
Tnv (µm)Pre/Post
Ttv (µm)Pre/Post
Tr (µm)
1
M
18
right
1.52/0.22
1,770/631
1,230/672
1,663/660
1,023/662
1,120/657
15
2
F
19
left
2.30/0.40
2,271/654
1,543/702
1,346/698
1,106/685
923/674
25
3
M
21
right
1.52/0.13
1,880/608
1,787/648
1,530/631
1,700/629
1,305/641
28
4
M
18
left
1.70/0.10
2,082/566
1,503/585
1,556/578
1,606/581
1,692/572
24
5
M
17
left
1.85/0.28
1,621/594
1,302/624
1,140/632
1,550/612
1,321/631
21
6
F
19
left
1.30/0.22
1,592/587
1,040/602
992/610
1,201/620
985/633
11
7
M
20
right
2.00/0.09
1,874/636
1,298/700
1,660/698
1,222/673
1,320/667
18
Mean ± SD
1.74 ± 0.34/0.20 ± 0.11
1,870 ± 243/611 ± 31
1,368 ± 244/648 ± 46
1,412 ± 263/644 ± 45
1,344 ± 269/637 ± 37
1,238 ± 258/639 ± 34
20 ± 6
t value
13.227
14.258
8.097
8.026
6.294
5.548
P value
<0.0001
<0.0001
<0.0001
<0.0001
0.001
0.001
Pre/Post preoperative results and postoperative results (6 months after DALK) The original contributions presented in this study are included in
this article/supplementary material, further inquiries can be directed
to the corresponding authors. Funding This work was supported by the National Natural Science
Foundation of China (82070920) and Shanghai “Science and
Technology Innovation Action Plan” Laboratory Animal Research
Project (201409006500). Conflict of interest The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could be
construed as a potential conflict of interest. Ethics statement The studies involving human participants were reviewed and
approved by the Medical Ethical Committee of Tongji Hospital
affiliated to Tongji University. The patients/participants provided
their written informed consent to participate in this study. Statistical analysis (B,F) One day after the first-stage surgery, at presentation, with double layers of cornea observed on slit-lamp and
AS-OCT examinations, with no corneal edema and with DM attached. (C,G) One week after the second-stage surgery, the corneal graft, which was the
same graft used in the first-stage operation, had survived and attached to the host’s residual stroma bed with a healed DM and corneal endothelial. (D,H)
One year after the second stage surgery, the cornea was transparent and the logMAR CDVA was 0.096. the same donor graft in the-second stage DALK surgery also
avoided the challenge of regenerating a new corneal epithelial layer,
promoting rapid graft healing (28). thoroughly dissection during the DALK procedure, and the recipient
corneal stroma bed remained transparent during subsequent follow-
up. In conclusion, we propose a novel two-stage procedure for
treating ACH, which shows rapid absorption of corneal edema and
DM healing and minimizes the risk of postoperative endothelial
immune rejection. At 2 months after epikeratophakia, 1 eye of 1 patient in this
case series showed a complication in the form of a branch-shaped
neovascularization in the superficial recipient corneal stroma bed. In that patient, all corneal neovascularization was removed after 05 frontiersin.org Liu et al. Liu et al. 10.3389/fmed.2022.1080892 Author contributions YB and LZ: conception and review. YB and YZ: clinical operation. CL, XH, and JS: methodology, formal analysis, and writing. YB:
funding acquisition. All authors contributed to the article and
approved the submitted version. Acknowledgments The authors express their gratitude to the many research
assistants for carrying out the extensive data collection required for
this project and the interns who assisted in the statistical analysis. The authors would like to thank all the reviewers who participated
in the review, as well as MJEditor (www.mjeditor.com) for providing
English editing services during the preparation of this manuscript. Publisher’s note All claims expressed in this article are solely those of the
authors and do not necessarily represent those of their affiliated
organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may
be made by its manufacturer, is not guaranteed or endorsed by the
publisher. frontiersin.org 06 10.3389/fmed.2022.1080892 10.3389/fmed.2022.1080892 Liu et al. References 1. Galvis V, Sherwin T, Tello A, Merayo J, Barrera R, Acera A. Keratoconus:
an
inflammatory
disorder? Eye
(Lond). (2015)
29:843–59. doi:
10.1038/eye.20
15.63 1. Galvis V, Sherwin T, Tello A, Merayo J, Barrera R, Acera A. Keratoconus:
an
inflammatory
disorder? Eye
(Lond). (2015)
29:843–59. doi:
10.1038/eye.20
15.63 16. García-Albisua AM, Davila-Avila N, Hernandez-Quintela E, Oteyza GG, Tapia-
Vazquez A, García-Arzate LD, et al. Visual and anatomic results after sole full-
thickness sutures for acute corneal hydrops. Cornea. (2020) 39:661–5. doi: 10.1097/ICO. 0000000000002265 16. García-Albisua AM, Davila-Avila N, Hernandez-Quintela E, Oteyza GG, Tapia-
Vazquez A, García-Arzate LD, et al. Visual and anatomic results after sole full-
thickness sutures for acute corneal hydrops. Cornea. (2020) 39:661–5. doi: 10.1097/ICO. 0000000000002265 2. Gatinel D, Galvis V, Tello A, Niño C, Rey J, Camacho P, et al. Obstructive
sleep
apnea-hypopnea
syndrome
and
keratoconus:
an
epiphenomenon
related
to
sleep
position? Cornea. (2020)
39:e11–2. doi:
10.1097/ICO.000000000000
2219 17. Garcí,a
de
Oteyza
G,
Bregliano
G,
Sassot
I,
Quintana
L,
Rius
C,
García-Albisua
A. Primary
surgical
options
for
acute
corneal
hydrops:
a
review. Eur
J
Ophthalmol. (2021)
32:781–89. doi:
10.1177/1120672121103
7833 3. Rabinowitz YS, Galvis V, Tello A, Rueda D, García JD. Genetics vs chronic corneal
mechanical trauma in the etiology of keratoconus. Exp Eye Res. (2021) 202:108328. doi:
10.1016/j.exer.2020.108328 18. Polo RC, Iglesias JC, Carrasco MG, Bartolom JMÉ, Fajardo JS, Guisado DG, et al. Management of corneal hydrops using air and micro-punctures. Arch Soc Esp Oftalmol. (2015) 90:198–201. doi: 10.1016/j.oftale.2015.03.016 4. Santodomingo-Rubido J, Carracedo G, Suzaki A, Villa-Collar C, Vincent SJ,
Wolffsohn JS. Keratoconus: an updated review. Cont Lens Anterior Eye. (2022) 45:101559. doi: 10.1016/j.clae.2021.101559 19. Subudhi P, Khan Z, Subudhi BN, Sitaram S. To show the efficacy of
compressive sutures alone in the management of acute hydrops in a keratoconus
patient. BMJ
Case
Rep. (2017) 2017:bcr2016218843. doi: 10.1136/bcr-2016-21
8843 5. Gokul A, Krishnan T, Emanuel P, Saunders M, McGhee C. Persisting extreme acute
corneal hydrops with a giant intrastromal cleft secondary to keratoconus. Clin Exp Optom. (2015) 98:483–6. doi: 10.1111/cxo.12260 20. Chaurasia S, Ramappa M, Murthy S. Rapid resolution of large and non- resolving
corneal hydrops using a modified technique of compression sutures. Semin Ophthalmol. (2022) 37:637–42. doi: 10.1080/08820538.2022.2051055 6. Galvis V, Tello A, Cianci MF, Escaf LC, Parra MM. Giant intrastromal cleft
in corneal hydrops. Int Ophthalmol. (2018) 38:1727–32. doi: 10.1007/s10792-017-0
647-7 21. Palioura S, Chodosh J, Pineda R. References A novel approach to the management
of
a
progressive
Descemet
membrane
tear
in
a
patient
with
keratoglobus
and acute hydrops. Cornea. (2013) 32:355–8. doi: 10.1097/ICO.0b013e31825c
ea80 7. Lockington D, Gaskin JC, McGhee CN, Patel DV. A prospective study of acute corneal
hydrops by in vivo confocal microscopy in a New Zealand population with keratoconus. Br J Ophthalmol. (2014) 98:1296–302. doi: 10.1136/bjophthalmol-2013-304145 22. Händel A, Lüke JN, Siebelmann S, Franklin J, Roters S, Matthaei M, et al. Outcomes
of deep anterior lamellar keratoplasty and penetrating keratoplasty in keratoconic eyes
with and without previous hydrops. Graefes Arch Clin Exp Ophthalmol. (2022) 260:2913–
23. doi: 10.1007/s00417-022-05643-4 8. Basu S, Vaddavalli PK, Ramappa M, Shah S, Murthy SI, Sangwan VS. Intracameral
perfluoropropane gas in the treatment of acute corneal hydrops. Ophthalmology. (2011)
118:934–9. doi: 10.1016/j.ophtha.2010.09.030 9. Grewal S, Laibson PR, Cohen EJ, Rapuano CJ. Acute hydrops in the corneal ectasias:
associated factors and outcomes. Trans Am Ophthalmol Soc. (1999) 97:187–203. 23. Li S, Liu M, Wang Q, Wang T, Shi W. Lamellar keratoplasty following
thermokeratoplasty in the treatment of acute corneal hydrops. Am J Ophthalmol. (2014)
158:26–31. doi: 10.1016/j.ajo.2014.03.011 10. Aldave AJ, Mabon M, Hollander DA, McLeod SD, Spencer WH, Abbott RL. Spontaneous corneal hydrops and perfora-tion in keratoconus and pellucid marginal
degeneration. Cornea. (2003) 22:169–74. doi: 10.1097/00003226-200303000-00019 24. Miyata K, Tsuji H, Tanabe T, Mimura Y, Amano S, Oshika T. Intracameral air
injection for acute hydrops in keratoconus. Am J Ophthalmol. (2002) 133:750–2. doi:
10.1016/S0002-9394(02)01437-X 11. Gaskin JC, Patel DV, McGhee CN. Acute corneal hydrops in keratoconus – new
perspectives. Am J Ophthalmol. (2014) 157:921–8. doi: 10.1016/j.ajo.2014.01.017 25. Lin W, Zhu Y, Chen W. Ultrastructure of the interface between descemet’s
membrane and the posterior stroma. Chin J Optom Ophthalmol Vis Sci. (2019) 21:
97–103. 12. Tan JC, Holland SP, Dubord PJ, Moloney G, McCarthy M, Yeung SN. Evolving
indications for and trends in keratoplasty in British Columbia, Canada, from 2002 to 2011:
a 10-year review. Cornea. (2014) 33:252–6. doi: 10.1097/ICO.0000000000000066 26. Romano V, Iovieno A, Parente G, Soldani AM, Fontana L. Long-term clinical
outcomes of deep anterior lamellar keratoplasty in patients with keratoconus. Am J
Ophthalmol. (2015) 159:505–11. doi: 10.1016/j.ajo.2014.11.033 13. Liu M, Hong J. Risk factors for endothelial decompensation after penetrating. J
Ophthalmol. (2018) 2018:1389486. doi: 10.1155/2018/1389486 27. Pineda R. Corneal transplantation in the developing world: lessons learned
and meeting the challenge. Cornea. (2015) 34(Suppl. 10):S35–40. doi: 10.1097/ICO. 0000000000000567 14. 28. Chen JY, Jones MN, Srinivasan S, Neal TJ, Armitage WJ, Kaye SB, et al.
Endophthalmitis after penetrating keratoplasty. Ophthalmology. (2015) 122:25–30. doi:
10.1016/j.ophtha.2014.07.038 References Li J, Yu L, Deng Z, Wang L, Sun L, Ma H, et al. Deep anterior lamellar keratoplasty
using acellular corneal tissue for prevention of allograft rejection in high-risk corneas. Am
J Ophthalmol. (2011) 152:762–70. doi: 10.1016/j.ajo.2011.05.002 15. Barsam A, Petrushkin H, Brennan N, Bunce C, Xing W, Foot B, et al. Acute corneal
hydrops in keratoconus: a national prospective study of incidence and management. Eye. (2015) 29:469–74. doi: 10.1038/eye.2014.333 28. Chen JY, Jones MN, Srinivasan S, Neal TJ, Armitage WJ, Kaye SB, et al. Endophthalmitis after penetrating keratoplasty. Ophthalmology. (2015) 122:25–30. doi:
10.1016/j.ophtha.2014.07.038 Frontiers in Medicine Frontiers in Medicine 07 frontiersin.org frontiersin.org
|
https://openalex.org/W2785517261
|
https://conflictandhealth.biomedcentral.com/track/pdf/10.1186/s13031-018-0139-z
|
English
| null |
Differences between the antibiotic prescribing pattern of newly arrived refugees in Germany and the German population
|
Conflict and health
| 2,018
|
cc-by
| 2,049
|
© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Differences between the antibiotic
prescribing pattern of newly arrived
refugees in Germany and the German
population Fabian Kahl1*
and Thomas Kühlein2 Abstract The number of refugees arriving in Europe increased dramatically in 2015, challenging the German health system. Amongst others, the treatment of infectious diseases is an important topic in refugee healthcare. A high prevalence
of multi-drug-resistant organisms has been identified among the refugee population. Still, little is known about the
prescription of antibiotic medication for refugees. We conducted a descriptive analysis of all antibiotics prescribed
to newly arrived refugees who were treated as outpatients between 10/01/2014 and 09/30/2015 in Erlangen, an
average sized German town. The City’s invoicing documents were used to collect data on prescriptions written for
refugees. Basic penicillins, aminopenicillins with beta-lactamase inhibitor and cephalosporins constituted the largest
proportion of antibiotics prescribed in the adult refugee group. Of these, both aminopenicillins with beta-
lactamase-inhibitor as well as basic penicillins were prescribed significantly more often compared to non-refugees. We conclude that the high percentage of prescriptions of aminopenicillins with beta-lactamase inhibitor is striking
and should be further investigated. the period from 10/01/2014–09/30/2015. The data basis
we used were anonymised invoicing documents of the
City’s Department of Social Services. Kahl and Kühlein Conflict and Health (2018) 12:3
DOI 10.1186/s13031-018-0139-z Kahl and Kühlein Conflict and Health (2018) 12:3
DOI 10.1186/s13031-018-0139-z * Correspondence: fabian.kahl@fau.de
1Institut für Geschichte und Ethik der Medizin, Friedrich-Alexander Universität
Erlangen-Nürnberg, Glückstraße 10, 91054 Erlangen, Germany
Full list of author information is available at the end of the article Correspondence Since the beginning of the refugee crisis in 2015, medical
care for refugees has posed a challenge to European
healthcare systems, including the German health system. Amongst many other health problems these patients
present with, infectious diseases are salient [1, 2], includ-
ing a high prevalence of multidrug-resistant organisms
(MDRO) like methicillin-resistant Staphylococcus aureus
(MRSA) [3–6]. Little is known about the medical treat-
ment of infectious diseases of refugees in Germany [7],
but the studies mentioned above call for a better surveil-
lance of prescribed antibiotics, as they can be both a
cause and a consequence of a high MDRO prevalence. Therefore, we present here the pattern of antibiotics pre-
scribed to newly arrived refugees in the city of Erlangen,
assessed through a retrospective analysis of all outpatient
prescriptions for newly arrived refugees in Erlangen over For a better comparison with literature reporting on
antibiotic prescriptions in Germany, we calculated the
defined daily dose (DDD) [8] of each antibiotic pre-
scribed and classified the substances in classes according
to the classification of Augustin et al. [9], as also used by
Batzing-Feigenbaum et al. [10]. Table 1 shows all anti-
biotic substances prescribed in this study as classes. We analysed the antibiotic prescribing pattern of two
groups which, due to their size, allowed a meaningful
analysis: Persons from 0 to 14 years, in the following
referred to as “children group”, and persons from 20 to
39 years, in the following referred to as “adult group”. While we did not find major differences in the chil-
dren group (0–14 years) compared to children insured
by the German statutory health insurance (gesetzliche
Krankenversicherung, GKV), as our results (see Table 2)
are equal to the analyses of GKV-insured children by
Augustin et al. [9] and Bätzing-Feigenbaum et al. Correspondence [10], * Correspondence: fabian.kahl@fau.de
1Institut für Geschichte und Ethik der Medizin, Friedrich-Alexander Universität
Erlangen-Nürnberg, Glückstraße 10, 91054 Erlangen, Germany
Full list of author information is available at the end of the article Kahl and Kühlein Conflict and Health (2018) 12:3 Page 2 of 3 Page 2 of 3 Table 1 Overview and Classification of all prescribed antibiotics
Antibiotic group
ATC-classes
Substance
Aminopenicillins with
beta-lactamase inhibitor (BLI)
J01CR02
amoxicillin and enzyme
inhibitor
J01CR04
sultamicillin
Basic penicillins
J01CA04
amoxicillin
J01CE02
phenoxymethylpenicillin
J01CE10
benzathine
phenoxymethylpenicillin
Cephalosporins
J01DC02
cefuroxime
J01DC04
cefaclor
J01DD08
cefixime
J01DD13
cefpodoxime
J01DD14
ceftibuten
Fluoroquinolones
J01MA01
ofloxacin
J01MA02
ciprofloxacin
J01MA12
levofloxacin
Macrolides/lincosamides
J01FA01
erythromycin
J01FA06
roxithromycin
J01FA09
clarithromycin
J01FA10
azithromycin
J01FF01
clindamycin
Nitrofurantoin/fosfomycin/
nitroxolin
J01XE01
nitrofurantoin
J01XX01
fosfomycin
Sulfonamide/trimethoprim
J01EE01
sulfamethoxazole and
trimethoprim
Tetracyclines
J01AA02
doxycycline aminopenicillins with beta-lactamase inhibitors (BLI)
(21%) and cephalosporins (16%). To adults, overall 1059.5 DDD of antibiotics were pre-
scribed. Basic penicillins constituted the greatest part with
44%, followed by aminopenicillins with BLI (21%), cepha-
losporins (13%) and macrolides (10%) (see Table 2). In comparison to GKV-insured patients, refugees re-
ceived broad spectrum antibiotics and aminopenicillins
with BLI more often. Bätzing-Feigenbaum et al. [10]
found a percentage of 25.9% for basic penicillins of all
prescribed antibiotic DDD, while aminopenicillins with
BLI had a rate of 2.9% in the age group 15–69 years
[10]. Another striking difference is found for the pre-
scriptions of tetracyclines, where we found a percentage
of 3.8% at all prescribed DDD in the adult group, while
for the GKV-insured persons, Bätzing-Feigenbaum et al. observed a percentage of 17.7% for tetracyclines [10]. Unfortunately, when analysing the prescriptions, it was
not possible to infer anything about the diagnosis lead-
ing to the prescriptions. Therefore, a specific analysis as
to the cause of the strikingly high percentage of amino-
penicillins with BLI in the adult group could not be car-
ried out. One could assume that the illnesses treated in
the refugee group were dissimilar to the GKV-insured. However, this seems not to be the case. Alberer et al. [2]
and Jung [1] described the distribution of diagnoses in
refugee patients. Beside symptom diagnoses, respiratory
tract infections were the most common. It is well known
that these diagnoses are typical for almost any GP’s
office, so the spectrum of diseases seems unlikely to
account for the deviant antibiotic prescription. Unfortunately, we can only report on a small number
of antibiotic prescriptions. Correspondence It was also not possible to
gain information about the diagnoses leading to the pre-
scriptions. However, we offer the first insight into the
antibiotic prescription pattern of refugees in Germany. we found striking deviations in the adult group: Here, of
478 drugs prescribed in total, 100 were antibiotics. Anti-
biotics therefore accounted for 20.9% of all prescriptions
in this group. The most frequently prescribed class of
antibiotics was basic penicillins (31%), followed by Table 2 Prescriptions of antibiotic classes in the adult and children group
Prescriptions in DDD
adults (20–39 years)
children (0–14 years)
Antibiotic groups
DDD absolute
DDD relative
DDD absolute
DDD relative
Aminopenicillins with beta-lactamase inhibitor
218,5
20,62%
10,50
2,61%
Basic penicillins
466,5
44,03%
200,50
49,81%
Cephalosporins
141,5
13,36%
110,50
27,45%
Fluoroquinolones
21
1,98%
0
0,00%
Macrolides/lincosamides
109
10,29%
55,00
13,66%
Metronidazol
0
0,00%
0
0,00%
Nitrofurantoin/fosfomycin/nitroxolin
28
2,64%
1,00
0,25%
Sulfonamide/trimethoprim
35
3,30%
5,00
1,24%
Tetracyclines
40
3,78%
20,00
4,97%
Others
0
0,00%
0
0,00%
Total
1059,5
100,00%
402,50
100,00% Page 3 of 3 Kahl and Kühlein Conflict and Health (2018) 12:3 Page 3 of 3 Page 3 of 3 Competing interests Competing interests
The authors report no conflict of interest. The authors report no conflict of interest. Funding
f
d Funding
No funding was received for this project. No funding was received for this project. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations. Conclusion Due to the fact, that the prevalence for MRSA in refu-
gees is not only higher compared to the German popula-
tion [4], but exceeds the prevalence of MRSA in other
known risk groups [5], the high prescription rate of ami-
nopenicillins with BLI in the adult group might be rea-
son for alarm. Considering the limitations of the study,
we see that data regarding refugee health is lacking and
a better surveillance of prescription rates is required. Authors’ contributions FK analysed and interpreted the data, TK supervised the study and
contributed to the writing of the manuscript. Both authors read and
approved the final manuscript. Acknowledgements The present work was performed in fulfillment of the requirements for
obtaining the degree „Dr. med.” at the Friedrich-Alexander-Universität
Erlangen-Nürnberg (FAU). We acknowledge support by Deutsche Forschungsgemeinschaft and
Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) within the funding
programme Open Access Publishing. pediatric refugees in an university Children's Hospital in Germany 2015-
2016. Infect Control Hosp Epidemiol. 2016;37:1310–4.
4.
Reinheimer C, Kempf VAJ, Jozsa K, Wichelhaus TA, Hogardt M, O'Rourke F,
Brandt C. Prevalence of multidrug-resistant organisms in refugee patients,
medical tourists and domestic patients admitted to a German university
hospital. BMC Infect Dis. 2017;17(17).
5.
Heudorf U, Albert-Braun S, Hunfeld K-P, Birne F-U, Schulze J, Strobel K, et al.
Multidrug-resistant organisms in refugees: prevalences and impact on
infection control in hospitals. GMS Hyg Infect Control. 2016;11:Doc16.
6.
Reinheimer C, Kempf VAJ, Gottig S, Hogardt M, Wichelhaus TA, O'Rourke F,
Brandt C. Multidrug-resistant organisms detected in refugee patients
admitted to a university hospital, Germany JuneDecember 2015. Euro
Surveill. 2016;21.
7.
Razum O, Bunte A, Gilsdorf A, Ziese T, Bozorgmehr K.
Gesundheitsversorgung von Geflüchteten: Zu gesicherten Daten kommen.
Deutsches Ärzteblatt. 2016;113:130–3.
8.
Fricke U, Günther J, Zawinell A, Zeidan R. Anatomisch-therapeutisch-
chemische Klassifikation mit Tagesdosen für den deutschen
Arzneimittelmarkt: ATC-Index mit DDD-Angaben für den deutschen
Arzneimittelmarkt. 14th ed. Berlin; 2015.
9.
Augustin J, Mangiapane S, Kern W. Antibiotika-Verordnungsprävalenzen im
Jahr 2010. 2012. http://www.versorgungsatlas.de/fileadmin/ziva_docs/
Antibiotika_Bericht_final.pdf. Accessed 15 Jan 2018.
10.
Batzing-Feigenbaum J, Schulz M, Schulz M, Hering R, Outpatient Antibiotic
KWV. Prescription. Dtsch Arztebl Int. 2016;113:454–9. Availability of data and materials The data that support the findings of this study are available from the City of
Erlangen (Amt für Soziales, Arbeit und Wohnen) but restrictions apply to the
availability of these data, which were used under license for the current
study, and so are not publicly available. Data are however available from the
authors upon reasonable request and with permission of the City of
Erlangen (Amt für Soziales, Arbeit und Wohnen). Author details
1 1Institut für Geschichte und Ethik der Medizin, Friedrich-Alexander Universität
Erlangen-Nürnberg, Glückstraße 10, 91054 Erlangen, Germany. 2Allgemeinmedizinisches Institut, Friedrich-Alexander Universität
Erlangen-Nürnberg, Universitätsstraße 29, 91054 Erlangen, Germany. • We accept pre-submission inquiries
• Our selector tool helps you to find the most relevant journal
• We provide round the clock customer support
• Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research
Submit your manuscript at
www.biomedcentral.com/submit
Submit your next manuscript to BioMed Central
and we will help you at every step: Submit your next manuscript to BioMed Central
and we will help you at every step: • We accept pre-submission inquiries
• Our selector tool helps you to find the most relevant journal
• We provide round the clock customer support
• Convenient online submission
• Thorough peer review
• Inclusion in PubMed and all major indexing services
• Maximum visibility for your research
Submit your manuscript at
www.biomedcentral.com/submit
Submit your next manuscript to BioMed Central
and we will help you at every step: Received: 16 July 2017 Accepted: 12 January 2018 Received: 16 July 2017 Accepted: 12 January 2018 Received: 16 July 2017 Accepted: 12 January 2018 Ethics approval and consent to participate Due to the study’s nature, which is a retrospective analysis of anonymised
data without contact to patients, an ethic approval was not required
according to German law. As the study only used routinely-collected, anon-
ymized data, the issue of informed patients’ consent does not apply. Consent for publication
Not applicable. Competing interests References References
1. Jung F. Das Bremer Modell. 2011. http://www.gesundheitsamt.bremen.de/
sixcms/media.php/13/3_GBE_Gesundheitsversorgung_Asylsuchender.pdf. Accessed 15 Jan 2018. 2. Alberer M, Wendeborn M, Löscher T, Seilmaier M. Erkrankungen bei
Flüchtlingen und Asylbewerbern. Dtsch Med Wochenschr. 2016;141:e8–e15. 3. Tenenbaum T, Becker K-P, Lange B, Martin A, Schafer P, Weichert S,
Schroten H. Prevalence of multidrug-resistant organisms in hospitalized 1. Jung F. Das Bremer Modell. 2011. http://www.gesundheitsamt.bremen.de/
sixcms/media.php/13/3_GBE_Gesundheitsversorgung_Asylsuchender.pdf. Accessed 15 Jan 2018. 2. Alberer M, Wendeborn M, Löscher T, Seilmaier M. Erkrankungen bei
Flüchtlingen und Asylbewerbern. Dtsch Med Wochenschr. 2016;141:e8–e15. 3. Tenenbaum T, Becker K-P, Lange B, Martin A, Schafer P, Weichert S, Schroten H. Prevalence of multidrug-resistant organisms in hospitalized
|
https://openalex.org/W3027723503
|
https://europepmc.org/articles/pmc7243303?pdf=render
|
English
| null |
Work- and stress-related musculoskeletal and sleep disorders among health professionals: a cross-sectional study in a hospital setting in Switzerland
|
BMC musculoskeletal disorders
| 2,020
|
cc-by
| 8,234
|
Open Access Work- and stress-related musculoskeletal
and sleep disorders among health
professionals: a cross-sectional study in a
hospital setting in Switzerland Oliver Hämmig Oliver Hämmig © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this article are included in the article's Creative Commons
licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons
licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the
data made available in this article, unless otherwise stated in a credit line to the data. Abstract Background: Musculoskeletal and sleep disorders have been reported to be very common among health care and
hospital workers and particularly nurses. They are assumed or found to be a result of psychological stress and/or
physical strain or pain. However, no other study so far – at least in a hospital setting and for Switzerland – has
considered and investigated musculoskeletal as well as sleep disorders in consequence of or rather in association
with both physical workload and psychological stress. Methods: Cross-sectional survey data of 1232 health professionals were used and analysed. Data were collected in
2015/16 among the health care workforces of three public hospitals and two rehabilitation clinics in the German-
speaking part of Switzerland. Musculoskeletal and sleep disorders were assessed by three items taken from the
Swiss Health Survey, a 2-item measure of accumulated low back, back, neck and shoulder pain and a single-item
measure of problems in getting to sleep or sleeping through. Stratified and adjusted bivariate logistic and
multivariate linear regression analyses were performed to calculate measures of association (adjusted odds ratios, z-
standardized beta coefficients), to control for potential confounders, and to compare different health professions
(nurses, physicians, therapists, other). Results: Almost every fourth of the studied health professionals reported severe or even very severe
musculoskeletal disorders (MSDs) and nearly every seventh severe sleep disorders (SDs). These prevalence rates
were significantly or at least slightly higher among nurses than among physicians and other health care workers. General stress, work stress, physical effort at work, and particularly a painful or tiring posture at work were found to
be clear and strong risk factors for MSDs, whereas only general and work-related stress were found to be
significantly associated with SDs. There was no or only weak association between MSDs and SDs. Conclusions: This study found MSDs to be largely a result of physical workload or rather poor posture at work and
only secondarily a consequence of (general) stress, whereas SDs were revealed to be primarily a consequence of
stress on and particularly off the job. Preventive strategies therefore have to differentiate and combine measures for
the reduction of both psychological stress and physical strain. Keywords: Musculoskeletal disorders, Sleep disorders, Physical workload, Psychological stress, Hospital workers,
Nurses, Switzerland g
Epidemiology, Biostatistics and Prevention Institute, University of Zurich,
Hirschengraben 84, 8001 Zurich, Switzerland Correspondence: oliver.haemmig@uzh.ch
Epidemiology, Biostatistics and Prevention Institute, University of Zurich,
Hirschengraben 84, 8001 Zurich, Switzerland Hämmig BMC Musculoskeletal Disorders (2020) 21:319
https://doi.org/10.1186/s12891-020-03327-w Hämmig BMC Musculoskeletal Disorders (2020) 21:319
https://doi.org/10.1186/s12891-020-03327-w Background single musculoskeletal pains but on multiple and com-
bined neck, shoulder, arm, back, and low back pain. Fur-
thermore, SDs are considered as related with MSDs or
rather as a consequence of (musculoskeletal) pain, as has
been repeatedly reported previously [29, 31]. It is well-known in occupational medicine that musculo-
skeletal injuries and disorders (MSDs) are strongly
work-related and as such one of the most prevalent oc-
cupational diseases in modern societies and working
populations. According to the European Foundation for
the Improvement of Living and Working Conditions,
MSDs are the main occupational disease suffered by
European workers and account for more than 50% of
serious work-related diseases [1]. Furthermore, it has
been widely reported that MSDs are particularly preva-
lent and among the most common health complaints in
health care workers and especially in hospital staff and
among nurses and physical therapists [2–12]. The path model (see Fig. 1) illustrates the assumed
causal paths und underlying hypotheses of the present
study: Psychological stress and physical workload are
expected to strongly determine and equally cause mus-
culoskeletal pain or disorders, which in turn are hypoth-
esized to produce sleep problems. In other words, MSDs
are directly and strongly caused by physical workload
and psychological (work) stress, whereas SDs are ex-
pected to be only or mainly indirectly caused by physical
workload and psychological (work) stress and to result
directly from MSDs. Research in occupational medicine has identified a
number of physical and psychosocial risk factors for the
development of work-related MSDs [13–23]. Previous
studies have shown that MSD are directly caused by
physically demanding work and strenuous working condi-
tions, such as lifting or carrying heavy loads, tiring posi-
tions, awkward posture, or repetitive movements [13–15]. Moreover, MSDs were also found to be related and associ-
ated with psychologically stressful work, i.e. with psycho-
social work factors and work-related stressors such as time
pressure, low job control, little social or supervisor support,
effort-reward imbalance, or work-life conflict [16–23]. This study addressed the following research questions: Are physical workloads and/or psychological stresses
and therefore musculoskeletal and sleep disorders
more common among nurses than among other
health professionals as expected? Can a (similarly) strong association be observed
between both psychological stress and physical
workload on the one hand and MSDs on the other? Background Is there at the same time a much less strong relation
between these conditions and SDs as hypothesized? In spite of such extensive research literature, only very
few studies have investigated MSDs depending on both
physical workload and psychological stress and – if at all
– mostly looked at specific pain (e.g. neck or shoulder
pain, upper limb pain, low back pain) and/or individual
occupational groups (predominantly nurses in direct pa-
tient care) and none for Switzerland. The same applies
to sleep problems or disorders, which were found not
only to be very common among hospital workers and
particularly nurses [24, 25], to have similar (psycho-
social) risk factors as MSDs [26–28] and to be likewise
associated with (work) stress [27, 28], but also to be in-
terrelated with (musculoskeletal) pain [29–31]. And al-
though stress and pain in general and work stress and
musculoskeletal pain in particular have been identified
as predictors or correlates of poor sleep, only few studies
have investigated stress-related and painful MSDs jointly
with sleep disorders (SDs). yp
Do SDs go along with MSDs as assumed? And if so,
are SDs substantially more strongly associated with
MSDs than with psychological stress or physical
workload? © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this article are included in the article's Creative Commons
licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons
licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the
data made available in this article, unless otherwise stated in a credit line to the data. Page 2 of 11 Page 2 of 11 Hämmig BMC Musculoskeletal Disorders (2020) 21:319 Data and study sample y
Secondary cross-sectional survey data were used for this
study, i.e. data that already existed and were collected
between summer 2015 and spring 2016. The survey was
conducted in a health care setting, or more precisely,
among the workforces of initially four public hospitals
and
two
rehabilitation
clinics
in
German-speaking
Switzerland. The original purpose of the survey was to
broadly and exploratively study the working conditions,
workloads and health of healthcare workers and particu-
larly health professionals. The only inclusion criterion
for the survey participation was being employed by one
of the six selected healthcare institutions at the period of
the data collection. In total, 1840 hospital employees
participated voluntarily and anonymously in the full
sample postal survey and filled in a written questionnaire
containing 100 questions on “Work and Health in the
Hospital”. The anonymous data collection did not allow
participants to be identified. The return rate overall was
41% and ranged from 36 to 49% for the participating
hospitals and rehab clinics, which included a university In light of all the studied associations and mentioned
limitations of previous research in this field, this study,
which is situated in a health care setting in Switzerland,
investigates both musculoskeletal and – subsequently –
sleep disorders as a possible result of or rather in associ-
ation with both physical (work) load and psychological
(work) stress. The study thereby takes into account not
only nurses but also physicians, therapists, and other
health professionals working in public hospitals and
rehab clinics. In addition, the study does not focus on Hämmig BMC Musculoskeletal Disorders (2020) 21:319 Page 3 of 11 Fig. 1 Theoretical path model explaining musculoskeletal and sleep disorders Fig. 1 Theoretical path model explaining musculoskeletal and sleep disorders hospital, a cantonal hospital, and two regional or district
hospitals. in a single category, since only 21 midwives participated
in the survey. All kinds of therapists (physiotherapists,
psychotherapists, occupational therapists etc.) were also
merged into one professional group. And medical-
technical staffs (radiologists, lab assistants, etc.) or scien-
tific and academic staffs (psychologists, pharmacists,
biologists, etc.) were summarized in the residual and
heterogeneous category “other health professionals”. Data and study sample Due to the accidental absence of a single-item and 5-
point scaled measure of general psychological stress in
the written questionnaire for one hospital, the present
study was limited to five of the six originally participat-
ing hospitals and rehab clinics, excluding one district
hospital with 273 employees who participated in the sur-
vey. An additional 335 hospital employees and members
from other than health professions were excluded from
the study. The study population was therefore restricted
to a total of 1232 health professionals, including 718
nurses and midwives (58.3% of the study sample), 222
physicians (18.0%), 137 therapists (11.1%), and 155 other
health professionals (12.6%) such as medical technical
staff, scientific staff, etc. Being a health professional and
having answered to the general stress question were the
two selection criteria for getting included in the study
population. More than 87% of the study population were women,
with a share of approximately 95% among nurses and
midwives, nearly 84% among therapists, and more than
65% among physicians. The selected study participants
were mostly highly educated (76.5%), with higher voca-
tional education, a maturity certificate, or even a univer-
sity degree, and were below age 45 (65.7%). Musculoskeletal disorders MSDs were assessed by a question and two items taken
from the periodically conducted Swiss Health Survey:
“Please indicate if and how pronounced you’ve had any
of the following complaints in the last four weeks: a)
back pain or pain in the lower back and b) neck ache or
shoulder pain”. For each of these complaints, the re-
sponse options were “not at all” (score 0), “a little” (1),
or “severe” (2). A summary score was created by adding
up the scores of the two 3-point scaled items to an over-
all MSD scale ranging from 0 to 4 and indicating accu-
mulated and (very) severe MSDs with scores of 3 or 4. Due to small numbers of cases and for anonymity rea-
sons, individual health professions were summarized in
larger professional categories, except for nurses and phy-
sicians, the two numerically largest and most homoge-
neous single health professions in the study population
by far. Such broad categorization or rough classification
into four sufficiently large occupational groups was done
not only for the data analysis and the present study but
largely already in the survey when response categories
for the question “To which profession do you belong?”
were specified. Nurses and midwives were summarized Page 4 of 11 Page 4 of 11 Hämmig BMC Musculoskeletal Disorders (2020) 21:319 Sleep disorders “never” to 4 “permanently”). A combined score has been
calculated with a range of values from 0 to 12 and a
Cronbach’s alpha of .66 (as a measure of internal
consistency). Poor posture at work was analogously mea-
sured with a single item: “Please indicate to what extent
your work includes the following: painful or tiring posture”
(response options from 0 “never” to 4 “permanently”). SDs are defined as difficulties in initiating and/or main-
taining sleep (sleep-onset or maintenance insomnia) and
were assessed by the question “Please indicate if and
how pronounced you’ve had any of the following com-
plaints in the last four weeks: difficulties in falling or
staying asleep”, with response options from 0 “not at
all”, 1 “a little”, to 2 “severe”. General stress First, relative frequencies of hypothesized risk factors or
exposures (general stress, work stress, physical effort at
work, poor posture at work) and prevalence rates of
health outcomes (MSDs, SDs) were calculated, aggre-
gated for all health professionals, and stratified for spe-
cific health professions (nurses, physicians, therapists,
other). General stress, defined as universal psychological stress
(as distinguished from physiological stress), was assessed
by an established and validated single-item measure of
psychological stress symptoms developed in the early
1970s [32]. The questionnaire provided a definition of
stress (“Stress means a situation in which a person feels
tense, restless, nervous and anxious and/or is unable to
sleep at night because his/her mind is troubled all the
time”) and then included the following question: “Do
you feel this kind of stress currently – and to what ex-
tent?” The response options on a 5-point scale ranged
from 0 “not at all” to 4 “very strongly”. Second, crosstabulations and bivariate logistic regres-
sion analyses were performed in order to estimate rela-
tive frequencies and odds ratios as measures of the
relative risks of physically demanding and psychologic-
ally stressful working conditions related to MSDs and
SDs. These bivariate association analyses were adjusted
for control variables such as sex, age, and education. g
Finally, multivariate linear regression analyses were
carried out to calculate z-standardized and multiple ad-
justed path or beta coefficients and to obtain independ-
ent and comparable effects of the assumed main work-
related and stress-related predictors in explaining MSDs
and SDs. In accordance with the theoretical path model
(see Fig. 1) which postulates only direct effects of psy-
chological stress and physical workload on MSDs, but
also and mainly indirect effects (via MSDs) on SDs, step-
wise (and stratified) linear regression analyses were per-
formed for explaining or determining SDs with step 1
representing the main but indirect path and step 2
representing the side but direct path. Work stress Work stress, understood as psychological stress on the
job or related to the job, was assessed by an established
and widely used multiple-item measure of failed reci-
procity or gratification crisis at work, better known as
effort-reward imbalance (ERI). This imbalance is con-
ceptualized as a perceived lack of reward received from
or at work compared to the effort put into work [33]
and was measured by a short version of the ERI ques-
tionnaire consisting of two subscales (effort, reward) and
a total of 16 items [34]. The so-called ERI ratio then was
calculated from the two 10-item and 6-item subscales or
sum scores of “effort” (numerator) and “reward” (de-
nominator), multiplied by a factor that corrects for the
different numbers of items of the two subscales. The ERI
ratio quantifies the amount of imbalance or reward frus-
tration and stress at work. In general, the amount of im-
balance or work stress increases with increasing values
of the ERI ratio. In particular, a ratio of below one indi-
cates an unproblematic imbalance for the benefit of the
reward component while a ratio of more than one indi-
cates an under-rewarded work effort or rather a stressful
high effort / low reward job situation. Results Regarding the first research question: Among nurses and
midwives, the proportion of those who report high phys-
ical effort at work (27%) and regular to permanent poor
work posture (27%) was substantially above the average
of all health professionals and particularly high com-
pared to physicians (1%, resp. 10%) and other health
professionals (5%, resp. 17%) (see Table 1). Moderate to
high work stress as measured by effort-reward imbalance
was widely spread throughout the entire study popula-
tion and clearly and significantly more prevalent among
nurses (72%) than among physicians (65%) and all other
health professionals (54–59%). On the other hand, mod-
erate to (very) strong general stress was also very com-
mon among the considered health professions, but did
not significantly differ between nurses (47%) and all
other health professionals (50–52%), as shown in Table 1. Physical workload y
Physically demanding work was assessed by two separate
measures concerning physical effort and poor posture at
work. Physical effort at work was measured by a scale
consisting of three items on the extent one’s work re-
quires carrying or moving persons, carrying or moving
heavy loads or standing (response options from 0 Hämmig BMC Musculoskeletal Disorders (2020) 21:319 Page 5 of 11 Table 1 Sociodemographic characteristics, prevalence rates of musculoskeletal and sleep disorders, general and job stress, and
physical workload among health professionals
Nurses (and
midwives)
Physicians
Therapists
Other health
professionals
All health
professionals
n = 718
n = 222
n = 137
n = 155
N = 1232
N
%
N
%
N
%
N
%
N
%
Sex
**
Female
675
94.5
145
65.3
115
83.9
136
87.7
1071
87.2
Male
39
5.5
77
34.7
22
16.1
19
12.3
157
12.8
Age
***
< 25 years
72
10.0
2
0.9
5
3.7
2
1.3
81
6.6
25–34 years
221
30.8
72
32.7
52
38.9
55
35.5
400
32.6
35–44 years
173
24.1
74
33.6
38
27.9
41
26.4
326
26.5
45–54 years
164
22.9
45
20.5
26
19.1
38
24.5
273
22.2
55+ years
87
12.1
27
12.3
15
11.0
19
12.3
148
12.1
Education
(highest level achieved)
***
Low (1–4)
27
3.8
0
–
0
–
0
–
27
2.2
Medium (5–6)
213
30.2
3
1.4
12
8.8
30
19.6
258
21.3
High (7–10)
346
49.0
2
0.9
28
20.4
51
33.3
427
35.2
Very high (11–12)
120
17.0
212
97.7
97
70.8
72
47.1
501
41.3
Musculoskeletal disorders
***
None (0)
117
16.5
66
30.3
36
26.3
41
26.5
260
21.4
Minor to moderate (1–2)
398
56.3
115
52.8
74
54.0
81
52.3
668
54.9
Severe to very severe (3–4)
192
27.2
37
17.0
27
19.7
33
21.3
289
23.7
Sleep disorders
n.s. Not at all (0)
340
47.9
109
49.3
56
40.9
66
42.6
571
46.7
A little (1)
273
38.5
88
39.8
63
46.0
68
43.9
492
40.2
Severe (2)
97
13.7
24
10.9
18
13.1
21
13.5
160
13.1
General stress
n.s. Physical workload Not at all to minimal (0–1)
363
53.2
106
49.3
66
49.6
72
48.3
607
51.5
Moderate (2)
229
33.6
69
32.1
44
33.1
55
36.9
397
33.7
Strong to very strong (3–4)
90
13.2
40
18.6
23
17.3
22
14.8
175
14.8
Work stress
***
Very low (ERI ratio ≤0.8)
43
6.5
23
11.3
19
14.7
31
21.1
116
10.2
Low (ERI ratio > 0.8–1)
146
22.0
48
23.6
40
31.0
30
20.4
264
23.1
Moderate (ERI ratio > 1–1.5)
379
57.2
108
53.2
56
43.4
68
46.3
611
53.5
High (ERI ratio > 1.5)
95
14.3
24
11.8
14
10.9
18
12.2
151
13.2
Physical effort at work
***
No / low (0–1)
86
12.6
111
51.2
53
39.6
83
56.8
333
28.2
Medium (2–5)
416
60.8
104
47.9
57
42.5
56
38.4
633
53.6
High (6–12)
182
26.6
2
0.9
24
17.9
7
4.8
215
18.2
Poor posture at work
***
Never to occasionally (0–1)
519
73.3
194
89.8
112
82.4
123
82.6
948
78.4
Regularly (2)
125
17.7
16
7.4
21
15.4
14
9.4
176
14.6
Frequently to permanently (3–4) 64
9.0
6
2.8
3
2.2
12
8.1
85
7.0
Pearson’s chi-square test: *p ≤.05; **p < .01; ***p < .001; n.s. = not significant (p > .05) Table 1 Sociodemographic characteristics, prevalence rates of musculoskeletal and sleep disorders, general and job stress, and
physical workload among health professionals Page 6 of 11 Page 6 of 11 Hämmig BMC Musculoskeletal Disorders (2020) 21:319 Hämmig BMC Musculoskeletal Disorders (2020) 21:319 As expected, strong accumulated MSDs were signifi-
cantly more prevalent among nurses (27%) than among
all other health professionals (17% up to 21%). In con-
trast, severe SDs or problems in falling or staying asleep
were only slightly or not at all more frequent among
nurses (14%) than among physicians (11%), therapists
(13%), or other health professionals (14%), as indicated
in Table 1. disorders were reported to be largely work-related and
stress-related as well as interrelated with one another
[21, 27, 30, 31]. Moreover, they were found to be a result
of both psychological stress and physical strain from
work. However, no studies so far have jointly investi-
gated such disorders in association with both physical
workload and psychological (work) stress, and particu-
larly in a hospital setting in Switzerland. Physical workload Furthermore, relatively low career prevalence
rates compared to 1-year prevalence rates of MSDs were
reported due to recall, reporting, and/or selection bias Associations with MSDs and SDs overall were signifi-
cantly less accentuated for physical workload (physical
effort at work, poor posture at work) than for psycho-
logical stress (general stress, work stress). But whereas
general and work-related stress were found to be strong
risk factors for both MSDs and SDs, physical workloads
and particularly poor work posture were more strongly
associated with MSDs than with SDs (see Table 2). p
(
q
g
)
In this study, as expected and often reported, severe
MSDs – just like high physical workloads – are found to
be significantly more common among nurses than
among other health professionals. But even though the
prevalence of (severe) MSDs in the present study is
found to be significantly higher in nurses than in other
health professions, such prevalence rates differ greatly
not only between health professions but also from study
to study and depending on the question or measure used
or on the observation period (year or career or lifetime
prevalence). A systematic review among health profes-
sionals and based on 23 retrospective and prospective
observational (cross-sectional and cohort) studies [2] re-
vealed a strong variation of 1-year prevalence rates of
work-related MSDs among health professionals from 28
to 96%. Furthermore, relatively low career prevalence
rates compared to 1-year prevalence rates of MSDs were
reported due to recall, reporting, and/or selection bias
(survivor or healthy worker effect) [2]. Against expectations, MSDs and SDs were not linearly
correlated or gradually associated with each other, but as
Table
2
shows,
having
strong
accumulated
MSDs
strongly increased the risk of having equally strong SDs
compared to those without any reported MSDs at all
(OR = 4.7). However, the fairly strong bivariate associ-
ation between MSDs and SDs (β = .21) decreased sub-
stantially in a multivariate association analysis (β = .07)
and turned out to be largely mediated or rather con-
founded by (general) stress, as shown in Table 4. Other findings from bivariate logistic regression ana-
lyses were fully supported by the results of multivariate
linear regression analyses. Physical workload When adjusting for all consid-
ered covariates, indicators of physical workload (physical
effort at work, poor posture at work) better explained
and more strongly predicted MSDs among health pro-
fessionals compared to the stress measures (general
stress, work stress) used (see Table 3). In turn, SDs
among specific health professions or among health pro-
fessionals in total were not or hardly determined by
physical workload, work stress or musculoskeletal pain
but were mainly and strongly associated with general
stress (see Table 4). As assumed, MSDs are found to result from combined
physical workload and psychological stress but in par-
ticular are most strongly associated with poor posture at
work and general stress. This main finding is broadly
consistent with two systematic reviews of 63 longitudinal
(case-control or cohort) studies [14] and of 18 mainly
cross-sectional studies [15]. Costa and Vieira [14] found
evidence for an association or even causation between
biomechanical risk factors (e.g. heavy physical work, fre-
quent lifting, awkward posture) and/or psychosocial risk
factors (e.g. low job control, high psychological work de-
mands, high job dissatisfaction) on the one hand and dif-
ferent specific work-related MSDs (i.e. neck, shoulder,
low back disorders) on the other. Similar to the present
study findings, Long et al. [15] found that among health Physical workload Against this
background, this study was carried out based on self-
reported survey data that were collected among health
care workers and hospital employees in the German-
speaking part of Switzerland. Concerning the second and third research questions:
Psychological stress turned out to be a very strong risk
factor for MSDs and SDs. Whereas the relative risk of
both severe MSDs and severe SDs increased remarkably
up to an odds ratio of above 17 with increasing levels of
general stress and work stress, associations were stron-
gest and gradients were steepest for SDs in association
with general stress and for MSDs in association with
work stress (see Table 2). But even though there is no other comparable obser-
vational study, particularly not for Switzerland, that has
considered and included physical work factors and both
general and work-related psychological stress indicators
as potential risk factors for work-related MSDs and SDs,
the findings of this study only partly support the results
of other studies and are in part contrary to the expecta-
tions or assumptions (see research questions and Fig. 1). But even though there is no other comparable obser-
vational study, particularly not for Switzerland, that has
considered and included physical work factors and both
general and work-related psychological stress indicators
as potential risk factors for work-related MSDs and SDs,
the findings of this study only partly support the results
of other studies and are in part contrary to the expecta-
tions or assumptions (see research questions and Fig. 1). In this study, as expected and often reported, severe
MSDs – just like high physical workloads – are found to
be significantly more common among nurses than
among other health professionals. But even though the
prevalence of (severe) MSDs in the present study is
found to be significantly higher in nurses than in other
health professions, such prevalence rates differ greatly
not only between health professions but also from study
to study and depending on the question or measure used
or on the observation period (year or career or lifetime
prevalence). A systematic review among health profes-
sionals and based on 23 retrospective and prospective
observational (cross-sectional and cohort) studies [2] re-
vealed a strong variation of 1-year prevalence rates of
work-related MSDs among health professionals from 28
to 96%. Discussion In previous studies, work-related musculoskeletal and
sleep disorders were both found to be relatively highly
prevalent among health care and hospital workers in
general and nurses in particular [2–10, 24, 25]. And both Hämmig BMC Musculoskeletal Disorders (2020) 21:319 Page 7 of 11 Table 2 Associations of physically demanding and psychologically stressful work with musculoskeletal and sleep disorders among
health professionals
Severe musculoskeletal disorders (3–4)
Severe sleep disorders (2)
%
aOR1)
95% CI
%
aOR1)
95% CI
Total study population
23.7
13.1
General stress
Not at all to minimal (0–1)
13.0
1
3.8
1
Moderate (2)
30.0
3.01
2.16–4.18
14.5
4.36
2.63–7.24
Strong to very strong (3–4)
48.5
6.88
4.61–10.28
41.1
17.67
10.47–29.82
Number of cases
1140
1146
Work stress
Very low (ERI ratio ≤0.8)
6.0
1
6.1
1
Low (ERI ratio > 0.8–1)
14.9
2.61
1.12–6.06
7.7
1.19
0.49–2.90
Moderate (ERI ratio > 1–1.5)
25.5
5.20
2.36–11.45
11.9
1.97
0.88–4.42
High (ERI ratio > 1.5)
44.6
13.03
5.64–30.08
32.0
6.87
2.96–15.94
Number of cases
1106
1110
Physical effort at work
No / low (0–1)
17.8
1
13.3
1
Medium (2–5)
22.8
1.42
0.99–2.02
11.3
0.97
0.64–1.47
High (6–12)
35.8
2.51
1.62–3.91
16.4
1.72
1.01–2.93
Number of cases
1145
1150
Poor posture at work
Never to sometimes (0–1)
17.8
1
10.7
1
Regularly (2)
40.6
3.04
2.13–4.33
20.0
2.33
1.51–3.60
Often to always (3–4)
58.3
6.26
3.88–10.09
22.6
2.71
1.52–4.82
Number of cases
1171
1176
Musculoskeletal disorders
None (0)
–
–
8.1
1
Minor to moderate (1–2)
–
–
–
8.7
1.19
0.70–2.02
Severe to very severe (3–4)
–
–
–
27.0
4.65
2.73–7.91
Number of cases
–
1190
1) Odds ratios adjusted for sex, age, and education
Bold print = significant (p < = .05) Table 2 Associations of physically demanding and psychologically stressful work with musculoskeletal and sleep disorders among
health professionals Table 2 Associations of physically demanding and psychologically stressful work with musculoskeletal and sleep disorders among Table 2 Associations of physically demanding and psychologically stressful work with musculoskeletal and sleep disorders among 1) Odds ratios adjusted for sex, age, and education
B ld
i t
i
ifi
t (
05) 1) Odds ratios adjusted for sex, age, and education
Bold print = significant (p < = .05) related
physical
strain
and
musculoskeletal
pain. Discussion cases in model
689
577
210
185
288
252
1189
1016
*p ≤.05; **p < .01; ***p < .001; no * = not significant (p > .05) 4 Explaining sleep disorders – results of stepwise multiple linear regression analyses
ent variable:
Nurses (and midwives)
Physicians Table 4 Explaining sleep disorders – results of stepwise multiple linear regression analyses
D
d
t
i bl
N
(
d
id i
)
Ph
i i
Th
i Table 4 Explaining sleep disorders – results of stepwise multiple linear regression analyses Discussion 05; **p < .01; ***p < .001; no * = not significant (p > .05) Table 3 Explaining musculoskeletal disorders – results of multiple linear regression analyses 05; **p < .01; ***p < .001; no * = not significant (p > .05) 26]. But no other study up to now has shown what
the
present
study
has
additionally
found
among
health care workers in general and different health professionals in particular: namely, that general stress
is an independent and even stronger risk factor for
sleep problems than occupational stress. Table 4 Explaining sleep disorders – results of stepwise multiple linear regression analyses
Dependent variable:
Sleep disorders
(single-item scale from 1 ‘not at all’ to 3 ‘severe’)
Nurses (and midwives)
Physicians
Therapists and other
health professionals
All health
professionals
n = 718
n = 222
n = 292
N = 1232
beta coeff. (β)
beta coeff. (β)
beta coeff. (β)
beta coeff. (β)
Step 1
Step 2
Step 1
Step 2
Step 1
Step 2
Step 1
Step 2
(Independent variables:)
Musculoskeletal disorders
(two-item scale from 0 ‘none’ to 4 ‘very severe’)
.24***
.11**
.12
−.01
.20***
.09
.21***
.09**
General stress
(single-item scale from 1 ‘not at all’ to 5 ‘very strong’)
–
.33***
–
.38***
–
.44***
–
.36***
Work stress
(ERI ratio from 0.3 ‘lowest’ to 3.7 ‘highest’)
–
.07
–
.11
–
.01
–
.06
Physical effort at work
(three-item scale from 0 ‘lowest’ to 12 ‘highest’)
–
−.01
–
−.12
–
−.05
–
−.03
Poor posture at work
(single-item scale from 0 ‘never’ to 4 ‘permanently’)
–
.09
–
.05
–
−.02
–
.06
(Control variables:)
Sex (male)
.03
.05
−.04
−.05
.08
.03
.02
.01
Age (< 25, 25–34, 35–44, 45–54, 55+)
.14***
.15***
.11
.13
.17**
.13*
.14***
.14***
Educational level (scale from 1 ‘low’ to 4 ‘very high’)
.02
.02
.00
−.05
−.01
−.05
.02
−.01
Adjusted R square
.070
.211
.006
.150
.053
.217
.057
.203
No. Discussion Therefore, MSDs were not found to be a strong cor-
relate or predictor of SD in a multiple adjusted and
fully specified linear regression analysis. This is rather
unexpected,
since
previous
studies
found
clear
(bivariate) associations between musculoskeletal pain
and poor sleep [30, 31]. And the results also do not
support previous findings and evidence from earlier
studies that psychosocial work factors or job stressors
such as low job control, high job strain, effort-reward
imbalance, or work-life imbalance – among other
things – are related to increased sleep problems [24– care workers (nurses, midwives and physicians) both
psychosocial work factors or job stressors and physical
job demands or work exposures are equally or similarly
strong risk factors for work-related (upper quadrant)
MSD or, more precisely, for neck, shoulder, and upper
back symptoms or complaints. Individual results of the present study are fully in
line with previous findings of other studies, but some
of the study findings are more surprising. Against ex-
pectations, SDs were observed to be much more a
direct consequence of psychological stress particularly
off the job rather than an (in-)direct result of work- Hämmig BMC Musculoskeletal Disorders (2020) 21:319 Page 8 of 11 Table 3 Explaining musculoskeletal disorders – results of multiple linear regression analyses
(Dependent variable:)
Musculoskeletal disorders
(two-item scale from 0 ‘none’ to 4 ‘very severe’)
Nurses (and
midwives)
Physicians
Therapists and other
health professionals
All health
professionals
n = 718
n = 222
n = 292
N = 1232
beta coeff. (β)
beta coeff. (β)
beta coeff. (β)
beta coeff. (β)
(Independent variables:)
General stress
(single-item scale from 1 ‘not at all’ to 5 ‘very strong’)
.14***
.13
.22***
.15***
Work stress
(ERI ratio from 0.3 ‘lowest’ to 3.7 ‘highest’)
.08
.23**
.02
.09**
Physical effort at work
(three-item scale from 0 ‘lowest’ to 12 ‘highest’)
−.01
.07
−.05
−.01
Poor posture at work
(single-item scale from 0 ‘never’ to 4 ‘permanently’)
.35***
.17*
.26***
.31***
(Control variables:)
Sex (male)
−.05
−.13
−.24***
−.13***
Age (< 25, 25–34, 35–44, 45–54, 55+)
.02
.03
−.07
.00
Educational level (scale from 1 ‘low’ to 4 ‘very high’)
−.04
−.03
−.05
−.06
Adjusted R square
.194
.142
.208
.204
No. cases in model
578
186
252
1009
*p ≤. Strengths and limitations The present study has its qualities and weaknesses. One of
the strengths is the heterogeneous and fairly large study
population, which provided sufficient statistical power and
allowed for multivariate and stratified analyses and com-
parisons between different health professions. Another strength is the use of established and vali-
dated single-item or multiple-item measures for the two
distinct stress concepts (general stress, work stress). All
other health- and work-related measures used in this
study (musculoskeletal pains, sleep disorders, physical
workload) could be observed or asked directly in the
survey and were easily understandable and interpretable
and therefore unproblematic in view of validity. This
helps to estimate and, at best, to improve the (internal
and external) validity of the measures and findings. The
consistent statistical approach from univariate and bi-
variate to multivariate analyses ensured the stability and
reliability of the findings. Performing multiple adjusted,
stratified, and stepwise regression analyses made it pos-
sible to test for confounding and mediation in the asso-
ciation analyses. Abbreviations Abbreviations
MSDs: Musculoskeletal disorders; SDs: Sleep disorders; ERI: Effort-reward
imbalance Funding This study received no direct funding from any third-party donor or funding
institution in the public, commercial, or non-profit sectors. However, the cross-sectional design did not allow con-
clusions to be drawn concerning causation in the studied
associations between exposures and outcomes. But the
strong and consistent associations found across different Acknowledgements Special thanks must go to the Swiss National Accident Insurance Fund
(Suva), the Federal Office of Public Health and the State Secretariat for
Economic Affairs (Seco) which made this study possible indirectly by
financially supporting the construction of the questionnaire and the
collection of the survey data used in this study. Conclusions
MSD MSDs among health professionals in this study are
found to be clearly work-related, i.e. to be primarily and
quite strongly associated with physically demanding and
psychologically stressful work and with general stress. In
contrast, SDs have proven not to be work-related. SDs
turned out to be only or mainly associated with general
stress and – against expectations – only weakly associ-
ated with musculoskeletal pain and not at all with phys-
ical strain and psychological stress at work. As a result,
MSDs (unlike SDs) are more prevalent among the phys-
ically burdened (hospital) nurses than among other
health care workers. Preventing work-related MSDs or
rather combined and accumulated (low) back, neck and
shoulder pain therefore requires mainly a reduction of
the physical workload, whereas SDs or more precisely
severe problems in falling or staying asleep can be pre-
vented most effectively by reducing the general stress
level. Author’s contributions The author designed, conceptualised and conducted the study, prepared
and revised the manuscript and performed all statistical analyses. The author
(s) read and approved the final manuscript. Dependent variable: Page 9 of 11 Page 9 of 11 Hämmig BMC Musculoskeletal Disorders (2020) 21:319 Hämmig BMC Musculoskeletal Disorders (2020) 21:319 Although evidence in this study is based on survey
data from the workforces of just a few and not randomly
selected hospitals and rehab clinics in German-speaking
Switzerland and therefore strictly speaking not represen-
tative for all health professionals in Switzerland, findings
of the study are nevertheless expected to be valid and
generalizable. There is no indication that the study find-
ings are substantially and systematically biased due to
selection, misclassification or misinformation. And there
is no plausible reason to believe that the study results
would
be
significantly
different
in
other
parts
of
Switzerland or based on self-reported data from other
hospitals and clinics. health professions with clear dose-response relationships
are at least supporting arguments when deducing caus-
ation from association. A possible selection bias due to self-selection of the par-
ticipating hospitals and rehab clinics and, in a next step, a
rather low response rate among the workforces at the par-
ticipating hospitals and clinics potentially calls the external
validity and generalizability of the study findings into ques-
tion. However, there is no indication or reason to believe
that study participants or rather survey respondents differ
systematically from non-respondents and, hence, that find-
ings are systematically biased as a result of self-selection. If
at all, underestimated prevalence rates (or associations)
may be possible or plausible but hardly any invalid findings
or completely false associations have to be expected. Nevertheless, only about 20% of the total variance of
both MSDs and SDs among health professionals were
explained by psychological (work) stress and/or physical
workloads. In other words, despite strong associations
found and important risk factors identified there are
80% of unexplained variance remaining and therefore
still some important undetected risk or explanatory fac-
tors for MSDs and SDs among health professionals. Al-
though it was not the purpose of the present study to
identify all risk factors or to explain most or as much as
possible of the variance of the outcome variables, further
research is obviously needed in light of such highly
prevalent health problems among highly stressed and
overloaded health care workers. Abbreviations
MSDs: Musculoskeletal disorders; SDs: Sleep disorders; ERI: Effort-reward
imbalance References 1. Giaccone M. Managing musculoskeletal disorders. Dublin: European
Foundation for the Improvement of Living and Working Conditions; 2007. http://www.eurofound.europa.eu/publications/report/2007/managing-
musculoskeletal-disorders-0. 1. Giaccone M. Managing musculoskeletal disorders. Dublin: European
Foundation for the Improvement of Living and Working Conditions; 2007. http://www.eurofound.europa.eu/publications/report/2007/managing-
musculoskeletal-disorders-0. 1. Giaccone M. Managing musculoskeletal disorders. Dublin: European
Foundation for the Improvement of Living and Working Conditions; 2007. http://www.eurofound.europa.eu/publications/report/2007/managing-
musculoskeletal-disorders-0. 21. Bernal D, Campos-Serna J, Tobias A, Vargas-Prada S, Benavides FG, Serra C. Work-related psychosocial risk factors and musculoskeletal disorders in
hospital nurses and nursing aides: a systematic review and meta-analysis. Int
J Nurs Stud. 2015;52(2):635–48. https://doi.org/10.1016/j.ijnurstu.2014.11.003. 2. Anderson SP, Oakman J. Allied health professionals and work-related
musculoskeletal disorders: a systematic review. Saf Health Work. 2016;7(4):
259–67. https://doi.org/10.1016/j.shaw.2016.04.001. 2. Anderson SP, Oakman J. Allied health professionals and work-related
musculoskeletal disorders: a systematic review. Saf Health Work. 2016;7(4):
259–67. https://doi.org/10.1016/j.shaw.2016.04.001. 22. Hämmig O, Knecht M, Läubli T, Bauer GF. Work-life conflict and
musculoskeletal disorders: a cross-sectional study of an unexplored
association. BMC Musculoskelet Disord. 2011;12:60. https://doi.org/10.1186/
1471-2474-12-60. 3. Dong H, Zhang Q, Liu G, Shao T, Xu Y. Prevalence and associated factors of
musculoskeletal disorders among Chinese healthcare professionals working
in tertiary hospitals: a cross-sectional study. BMC Musculoskelet Disord. 2019;
20(1):175. https://doi.org/10.1186/s12891-019-2557-5. 23. Hämmig O. Health and well-being at work: the key role of supervisor
support. SSM Popul Health. 2017;3:393–402. https://doi.org/10.1016/j.ssmph. 2017.04.002. 4. Hafner ND, Milek DM, Fikfak MD. Hospital staff’s risk of developing
musculoskeletal disorders, especially low back pain. Zdr Varst. 2018;57(3):
133–9. https://doi.org/10.2478/sjph-2018-0017. 24. Dong H, Zhang Q, Sun Z, Sang F, Xu Y. Sleep problems among Chinese
clinical nurses working in general hospitals. Occup Med (Lond). 2017a;67(7):
534–9. https://doi.org/10.1093/occmed/kqx124. 5. Davis KG, Kotowski SE. Prevalence of musculoskeletal disorders for nurses in
hospitals, long-term care facilities, and home health care: a comprehensive
review. Hum Factors. 2015;57(5):754–92. https://doi.org/10.1177/
0018720815581933. 25. Dong H, Zhang Q, Sun Z, Sang F, Xu Y. Sleep disturbances among
Chinese clinical nurses in general hospitals and its influencing factors. BMC Psychiatry. 2017b;17(1):241. https://doi.org/10.1186/s12888-017-
1402-3. 6. Harcombe H, Herbison GP, McBride D, Derrett S. Musculoskeletal disorders
among nurses compared with two other occupational groups. Occup Med
(Lond). 2014;64(8):601–7. https://doi.org/10.1093/occmed/kqu117. 6. Harcombe H, Herbison GP, McBride D, Derrett S. Musculoskeletal disorders
among nurses compared with two other occupational groups. Occup Med
(Lond). 2014;64(8):601–7. https://doi.org/10.1093/occmed/kqu117. 26. Chazelle E, Chastang JF, Niedhammer I. Psychosocial work factors and
sleep problems: findings from the French national SIP survey. Availability of data and materials Secondary data used in this study were collected anonymously and on a
voluntary basis and by random and full sample surveys among the
workforces of four public hospitals and two privately operated rehab clinics. Page 10 of 11 Page 10 of 11 Hämmig BMC Musculoskeletal Disorders (2020) 21:319 Hämmig BMC Musculoskeletal Disorders (2020) 21:319 Hämmig BMC Musculoskeletal Disorders (2020) 21:319 Nevertheless, the hospitals and clinics have agreed to participate in the
survey on condition that the collected data may not passed on to any third
parties. Or more precisely, data are not publicly accessible and freely
available since the use and analysis of the pooled data and the publication
of any research findings and study results out of it are restricted by contract
with the participating companies (hospitals, clinics) to the University of
Zurich (Epidemiology, Biostatistics and Prevention Institute) and the
collaborating Careum Research, a division of the Careum Foundation. The
use of the data therefore is strictly limited to the responsible scientists from
the two research institutions. Nevertheless, the hospitals and clinics have agreed to participate in the
survey on condition that the collected data may not passed on to any third
parties. Or more precisely, data are not publicly accessible and freely
available since the use and analysis of the pooled data and the publication
of any research findings and study results out of it are restricted by contract
with the participating companies (hospitals, clinics) to the University of
Zurich (Epidemiology, Biostatistics and Prevention Institute) and the
collaborating Careum Research, a division of the Careum Foundation. The
use of the data therefore is strictly limited to the responsible scientists from
the two research institutions. disorders among registered nurses: evidence from the Thai nurse cohort
study. BMC Nurs. 2017;16:68. https://doi.org/10.1186/s12912-017-0263-x. disorders among registered nurses: evidence from the Thai nurse cohort
study. BMC Nurs. 2017;16:68. https://doi.org/10.1186/s12912-017-0263-x. 11. Vieira ER, Schneider P, Guidera C, Gadotti IC, Brunt D. Work-related
musculoskeletal disorders among physical therapists: a systematic review. J
Back Musculoskelet Rehabil. 2016;29(3):417–28. https://doi.org/10.3233/BMR-
150649. 12. Milhem M, Kalichman L, Ezra D, Alperovitch-Najenson D. Work-related
musculoskeletal disorders among physical therapists: a comprehensive
narrative review. Int J Occup Med Environ Health. 2016;29(5):735–47. https://
doi.org/10.13075/ijomeh.1896.00620. 13. Ngan K, Drebit S, Siow S, Yu S, Keen D, Alamgir H. Risks and causes of
musculoskeletal injuries among health care workers. Occup Med (Lond). 2010;60(5):389–94. https://doi.org/10.1093/occmed/kqq052. Ethics approval and consent to participate The study is observational and not clinical or experimental and did not
involve medical records or human tissues. Data used were collected from
hospitals and clinics, but survey respondents were employees and not
patients. Furthermore, survey participation was completely anonymous and
entirely on a voluntary basis. For these reasons, informed and explicit
consent from hospital employees was not needed or obtained but implicitly
given by participating voluntarily in the survey. 14. da Costa BR, Vieira ER. Risk factors for work-related musculoskeletal
disorders: a systematic review of recent longitudinal studies. Am J Ind Med. 2010;53(3):285–323. https://doi.org/10.1002/ajim.20750. 15. Long MH, Johnston V, Bogossian F. Work-related upper quadrant
musculoskeletal disorders in midwives, nurses and physicians: a systematic
review of risk factors and functional consequences. Appl Ergon. 2012;43(3):
455–67. https://doi.org/10.1016/j.apergo.2011.07.002. g
y
g
y
y
In Switzerland, under national and cantonal laws and for this kind of data
collection, no formal approval or authorisation is required or will be issued,
neither by the national ethics committee nor by the cantonal commissioner
for data protection. These are not even recommended by the medical-
ethical guidelines for scientific integrity of the Central Ethics Committee and
the Swiss Academies of Sciences. 16. Freimann T, Pääsuke M, Merisalu E. Work-related psychosocial factors and
mental health problems associated with musculoskeletal pain in nurses. Pain Res Manag. 2016;2016:9361016. 17. Mehrdad R, Dennerlein JT, Haqhighat M, Aminian O. Association between
psychosocial factors and musculoskeletal symptoms among Iranian nurses. Am J Ind Med. 2010;53(10):1032–9. https://doi.org/10.1002/ajim.20869. 18. Azma K, Hosseini A, Safarian MH, Abedi M. Evaluation of the relationship
between musculoskeletal discomforts and occupational stressors among
nurses. N Am J Med Sci. 2015;7(7):322–7. https://doi.org/10.4103/1947-2714. 161250. Received: 4 September 2019 Accepted: 5 May 2020 20. Bazazan A, Dianat I, Bahrampour S, Talebian A, Zandi H, Sharafkhaneh
A, Maleki-Ghahfarokhi A. Association of musculoskeletal disorders and
workload with work schedule and job satisfaction among emergency
nurses. Int Emerg Nurs. 2019;44:8–13. https://doi.org/10.1016/j.ienj.2019. 02.004. Competing interests 19. Amin NA, Quek KF, Oxley JA, Noah R, Nordin R. Emotional distress as a
predictor of work-related musculoskeletal disorders in Malaysian nursing
professionals. Int J Occup Environ Med. 2018;9(2):69–78. https://doi.org/10. 15171/ijoem.2018.1158. p
g
The author declares that he has no competing interests. Received: 4 September 2019 Accepted: 5 May 2020 Received: 4 September 2019 Accepted: 5 May 2020 References Siegrist J, Starke D, Chandola T, Godin I, Marmot M, Niedhammer I, et al. The measurement of effort-reward imbalance at work: European
comparisons. Soc Sci Med. 2004;58(8):1483–99. References Int Arch
Occup Environ Health. 2016;89(3):485–95. https://doi.org/10.1007/s00420-
015-1087-1. 7. Reed LF, Battistutta D, Young J, Newman B. Prevalence and risk factors
for foot and ankle musculoskeletal disorders experienced by nurses. BMC Musculoskelet Disord. 2014;15:196. https://doi.org/10.1186/1471-
2474-15-196. 7. Reed LF, Battistutta D, Young J, Newman B. Prevalence and risk factors
for foot and ankle musculoskeletal disorders experienced by nurses. BMC Musculoskelet Disord. 2014;15:196. https://doi.org/10.1186/1471-
2474-15-196. 27. Linton SJ, Kecklund G, Franklin KA, Leissner LC, Sivertsen B, Lindberg E, et al. The effect of the work environment on future sleep disturbances: a
systematic review. Sleep Med Rev. 2015;23:10–9. https://doi.org/10.1016/j. smrv.2014.10.010. 8. Stolt M, Suhonen R, Virolainen P, Leino-Kilpi H. Lower extremity
musculoskeletal disorders in nurses: a narrative literature review. Scand J
Public Health. 2016;44(1):106–15. https://doi.org/10.1177/1403494815602989. 8. Stolt M, Suhonen R, Virolainen P, Leino-Kilpi H. Lower extremity
musculoskeletal disorders in nurses: a narrative literature review. Scand J
Public Health. 2016;44(1):106–15. https://doi.org/10.1177/1403494815602989. 28. Nieuwenhuijsen K, Bruinvels D, Frings-Dresen M. Psychosocial work
environment and stress-related disorders, a systematic review. Occup Med
(Lond). 2010;60(4):277–86. https://doi.org/10.1093/occmed/kqq081. 9. Passali C, Maniopoulou D, Apostolakis I, Varlamis I. Work-related
musculoskeletal disorders among Greek hospital nursing professionals: a
cross-sectional observational study. Work. 2018;61(3):489–98. https://doi.org/
10.3233/WOR-182812. 9. Passali C, Maniopoulou D, Apostolakis I, Varlamis I. Work-related
musculoskeletal disorders among Greek hospital nursing professionals: a
cross-sectional observational study. Work. 2018;61(3):489–98. https://doi.org/
10.3233/WOR-182812. 29. Finan PH, Goodin BR, Smith MT. The association of sleep and pain: an
update and a path forward. J Pain. 2013;14(12):1539–52. https://doi.org/10. 1016/j.jpain.2013.08.007. 10. Thinkhamrop W, Sawaengdee K, Tangcharoensathien V, Theerawit T,
Laohasiriwong W, Saengsuwan J, Hurst CP. Burden of musculoskeletal Page 11 of 11 Hämmig BMC Musculoskeletal Disorders (2020) 21:319 Hämmig BMC Musculoskeletal Disorders (2020) 21:319 30. Kelly GA, Blake C, Power CK, O'Keeffe D, Fullen BM. The association between
chronic low back pain and sleep: a systematic review. Clin J Pain. 2011;27(2):
169–81. https://doi.org/10.1097/AJP.0b013e3181f3bdd5. 31. Vinstrup J, Jakobsen MD, Calatayud J, Jay K, Andersen LL. Association of
stress and musculoskeletal pain with poor sleep: cross-sectional study
among 3,600 hospital workers. Front Neurol. 2018;9:968. https://doi.org/10. 3389/fneur.2018.00968. 32. Elo AL, Leppänen A, Jahkola A. Validity of a single-item measure of stress
symptoms. Scand J Work Environ Health. 2003;29(6):444–51. 32. Elo AL, Leppänen A, Jahkola A. Validity of a single-item measure of stress
symptoms. Scand J Work Environ Health. 2003;29(6):444–51. 33. Siegrist J. Adverse health effects of high-effort/low-reward conditions. J
Occup Health Psychol. 1996;1(1):27–41. 34. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
|
https://openalex.org/W4280497624
|
https://www.researchsquare.com/article/rs-1371764/latest.pdf
|
English
| null |
Grain Production Space Reconstruction and Its Influencing Factors in the Loess Plateau
|
International journal of environmental research and public health/International journal of environmental research and public health
| 2,022
|
cc-by
| 11,077
|
Grain Production Space Reconstruction and Its
Influencing Factors in the Loess Plateau Northwest University https://orcid.org/0000-0001-6647-6958 Research Article Keywords: grain production space, reconstruction pathway, spatiotemporal evolution, influencing factors,
Loess Plateau Grain Production Space Reconstruction and Its
Influencing Factors in the Loess Plateau
zhangxuan Qin
Northwest University
Xiaolin Liu
Northwest University
Xiaoyan Lu
Northwest University
Mengfei Li
Northwest University
Fei Li
Northwest University https://orcid.org/0000-0001-6647-6958
Research Article
Keywords: grain production space, reconstruction pathway, spatiotemporal evolution, influencing factors,
Loess Plateau
Posted Date: April 1st, 2022
DOI: https://doi.org/10.21203/rs.3.rs-1371764/v1
License: This work is licensed under a Creative Commons Attribution 4.0 International License. Read
Full License
Version of Record: A version of this preprint was published at International Journal of Environmental Grain Production Space Reconstruction and Its
Influencing Factors in the Loess Plateau
zhangxuan Qin
Northwest University
Xiaolin Liu
Northwest University
Xiaoyan Lu
Northwest University
Mengfei Li
Northwest University
Fei Li
Northwest University https://orcid.org/0000-0001-6647-6958
Research Article
Keywords: grain production space, reconstruction pathway, spatiotemporal evolution, influencing factors,
Loess Plateau
Posted Date: April 1st, 2022
DOI: https://doi.org/10.21203/rs.3.rs-1371764/v1
License: This work is licensed under a Creative Commons Attribution 4.0 International License. Read
Full License
Version of Record: A version of this preprint was published at International Journal of Environmental Posted Date: April 1st, 2022 Posted Date: April 1st, 2022 DOI: https://doi.org/10.21203/rs.3.rs-1371764/v1 DOI: https://doi.org/10.21203/rs.3.rs-1371764/v1 License: This work is licensed under a Creative Commons Attribution 4.0 International License. Read
Full License Version of Record: A version of this preprint was published at International Journal of Environmental
Research and Public Health on May 12th, 2022. See the published version at
https://doi.org/10.3390/ijerph19105876. Page 1/26 Abstract Land systems can be classified spatially as grain production space, ecological service space and urban-rural
development space. Grain production space reconstruction concentrates the principal contradictions in land
system changes. Exploring the process, pathways and influencing factors of grain production space
reconstruction can coordinate the trade-off/synergy among grain production, ecological protection and
economic development. Therefore, the reconstruction characteristics of grain production space in the Loess
Plateau were systematically analyzed in terms of quantity, quality, and spatial pattern in this study. Results
showed that although the quantity of grain production space in the Loess Plateau declined, the quality
improved between 1980 and 2018. The grain production space in the Loess Plateau was increasingly
fragmented and scattered, with more regular shapes in 2018 than 1980. Moreover, the center of gravity of
grain production space moved 8.32 kilometers to the northwest; it moved more significantly to the northwest
when considering the quality of grain production space. The reconstruction of grain production space in
quantity, quality and spatial pattern was mainly realized through the pathways of Grain for Green, Urban
Expansion, Deforestation and Reclamation as well as Land Consolidation; and the pathways of Grain for
Green after 2000 was the principal pathway for grain production space reconstruction. The four
reconstruction pathways of grain production space in the Loess Plateau were the result of a combination of
natural environment and socio-economic factors, but driving factors had different strengths and directions
for each reconstruction pathway. This study offered a comprehensive understanding of grain production
space reconstruction, and provided a new research perspective for coordinating grain production, ecological
protection and high-quality economic development. Introduction The quality of grain production space was a complex internal attribute (Rossiter, 1996), natural
features (fertility, structure, texture and stability etc.) and human activities (fertilization, irrigation, and crop
rotation etc.) directly, indirectly, synergistically or antagonistically affected the quality of grain production
space (Leonard et al., 2020; Kurmangozhinov et al., 2020; Jin et al., 2021). Hence, it was difficult to establish
a unified standard to evaluate the quality of grain production space (Fu and Bai, 2015). Climate changes such
as CO2 increasing, temperature rising (Al-Kayssi et al., 2016), and extreme weather adding (Veljanoska,
2018), and human activities such as land consolidation (Fuentes-Guevara et al., 2022), urban expansion(van
Vliet et al., 2015) and intensive utilization (Johnson et al., 2014) of grain production space profoundly
affected the quality reconstruction. Under the severely influence of human activities, fragmentation was
intensified, shape was complicated, and spatial distribution tended to be unstable of grain production space
(Irwin et al, 2007; Jiang et al., 2019; Ju et al., 2020). In the context of global warming and the squeeze of
grain supply and demand, the center of gravity of grain production space had a tendency to expand toward
high latitudes (Zhang et al., 2017; Cheng et al., 2018). Scholars paid more attention to the changing rules of
landscape pattern and spatial heterogeneity of quantitative changes of grain production space, but there was
little study of the spatial pattern evolution of its quality. Natural environment factors were the basis of grain
production space reconstruction, and controlled the direction of grain production space reconstruction
(Zhang et al., 2018b; Bateman et al., 2016); Socio-economic factors were the core driving force for grain
production space reconstruction, and determined speed and intensity of the reconstruction (Uisso and
Tanrivermis, 2021; Arowolo and Deng, 2018; Song et al., 2018); often the driving role of socio-economic
factors was more significant (Tubiello and Fischer, 2007; Ge et al., 2018). These studies analyzed
characteristics of grain production space reconstruction from different dimensions, and provided a series of
models and analysis frameworks for analyzing grain production space reconstruction. However, it was
necessary to establish a comprehensive framework to systematically analyze the reconstruction of grain
production space in terms of quantity, quality and spatial pattern. Since the 1980s, the quantity of global grain production space shown an insignificant increase (Zhang et al.,
2017; Homer et al., 2020), but there were significant regional differences (Winkler et al., 2021). Introduction The loss of
grain production space mainly occurred in economically developed and densely populated areas, resulting
from urban expansion (Canaz et al., 2017; Bren d’Amour et al., 2017; Huang et al., 2022), or in ecologically
fragile areas through policy guidance to protect the ecological environment, and gradually converted
cultivated land into forest or grassland (Liu et al., 2017; Raghavan and Shrimali, 2015). In areas with relatively
slow economic development, different stakeholders reclaimed land to pursue economic benefits and ensure
grain security, and grain production space increased slightly (Hailu et al., 2020; Ceddia, 2019; Arowolo et al.,
2018). The quality of grain production space was a complex internal attribute (Rossiter, 1996), natural
features (fertility, structure, texture and stability etc.) and human activities (fertilization, irrigation, and crop
rotation etc.) directly, indirectly, synergistically or antagonistically affected the quality of grain production
space (Leonard et al., 2020; Kurmangozhinov et al., 2020; Jin et al., 2021). Hence, it was difficult to establish
a unified standard to evaluate the quality of grain production space (Fu and Bai, 2015). Climate changes such
as CO2 increasing, temperature rising (Al-Kayssi et al., 2016), and extreme weather adding (Veljanoska,
2018), and human activities such as land consolidation (Fuentes-Guevara et al., 2022), urban expansion(van
Vliet et al., 2015) and intensive utilization (Johnson et al., 2014) of grain production space profoundly
affected the quality reconstruction. Under the severely influence of human activities, fragmentation was
intensified, shape was complicated, and spatial distribution tended to be unstable of grain production space
(Irwin et al, 2007; Jiang et al., 2019; Ju et al., 2020). In the context of global warming and the squeeze of
grain supply and demand, the center of gravity of grain production space had a tendency to expand toward
high latitudes (Zhang et al., 2017; Cheng et al., 2018). Scholars paid more attention to the changing rules of
landscape pattern and spatial heterogeneity of quantitative changes of grain production space, but there was
little study of the spatial pattern evolution of its quality. Introduction Land systems, coupled social economy-land use-ecological environment, possessed the multiple functions
(Verburg et al., 2015; Li et al., 2020), such as ensuring grain security, providing ecological services, promoting
economic prosperity, etc. The development and evolution of land systems had a profound impact on the
realization of sustainable goals (Leyk et al., 2020; Obersteiner et al., 2016; Meyfroidt et al. 2018). According
to the dominant function of land systems, land systems can be divided into grain production space, urban-
rural development space and ecological service space (Zhang et al., 2021; Li et al., 2021; Yang et al., 2020). However, it was difficult for land systems to achieve all sustainable development goals simultaneously, and
there were trade-offs among different functions (Gao and Bryan, 2017; Hasegawa et al., 2019; Kang et al.,
2021), resulted in the global land systems underwent changes characterized by grain production space
reconstruction, urban-rural development space expansion, ecological service space contraction (Foley et al.,
2005; Liu et al., 2014&2018). The grain production space reconstruction, referred to the changes in the
quantity, quality and spatial pattern of grain production space caused by the subjective trade-offs of different
stakeholders on land system functions, was closely related to the expansion of urban-rural development
space and the contraction of ecological service space, concentrated the main contradictions of the land
system changes. Research on grain production space reconstruction can not only deepen understanding of
land system changes, but also help coordinate the contradictions among land system functions and promote
the realization of sustainable development goals. Page 2/26 Page 2/26 Since the 1980s, the quantity of global grain production space shown an insignificant increase (Zhang et al.,
2017; Homer et al., 2020), but there were significant regional differences (Winkler et al., 2021). The loss of
grain production space mainly occurred in economically developed and densely populated areas, resulting
from urban expansion (Canaz et al., 2017; Bren d’Amour et al., 2017; Huang et al., 2022), or in ecologically
fragile areas through policy guidance to protect the ecological environment, and gradually converted
cultivated land into forest or grassland (Liu et al., 2017; Raghavan and Shrimali, 2015). In areas with relatively
slow economic development, different stakeholders reclaimed land to pursue economic benefits and ensure
grain security, and grain production space increased slightly (Hailu et al., 2020; Ceddia, 2019; Arowolo et al.,
2018). Analysis framework Economic development, grain production and ecological services were the basic functions provided by land
systems for human society, and there were different degrees of trade-offs among the three. The subjective
trade-offs of different stakeholders in land system functions driven the changes of land use. Meanwhile,
different reconstruction pathways of grain production space led to objective trade-offs between land system
functions. Hence, grain production space was reconstructed in the transmission process of subjective trade-
offs and objective trade-offs. Economic development promoted Urban Expansion, which occupied a large
number of cultivated land and affected grain production. In order to ensure grain security, Deforestation and
Reclamation as well as Land Consolidation restored a large amount of grain production space. But
Deforestation and Reclamation led to the degradation of the natural environment, Grain for Green converted
cultivated land into forests or grasslands to improve ecological services function. In addition, grain
production space reconstruction was reflected in three aspects: quantity, quality, and spatial pattern, and
there were significant differences in the impact of different pathways on grain production space
reconstruction. Therefore, this study assumed there was only one reconstruction pathway in grain production
space, then compared its reconstruction characteristics with the initial year, next obtained the principal
pathway for the reconstruction of grain production space in quantity, quality, and spatial pattern (Fig. 1). Based on the systematic analysis framework of grain production space reconstruction, this study was
organized by the following parts: 1. Assessing the characteristics of the quantity, quality, and spatial pattern reconstruction of grain production
space; 1. Assessing the characteristics of the quantity, quality, and spatial pattern reconstruction of grain production
space; 2. Determination the principal pathway for the reconstruction of grain production space in quantity, quality,
and spatial pattern; 2. Determination the principal pathway for the reconstruction of grain production space in quantity, quality,
and spatial pattern; 3. Defining the influencing factors of each reconstruction pathway. 3. Defining the influencing factors of each reconstruction pathway. Introduction Natural environment factors were the basis of grain
production space reconstruction, and controlled the direction of grain production space reconstruction
(Zhang et al., 2018b; Bateman et al., 2016); Socio-economic factors were the core driving force for grain
production space reconstruction, and determined speed and intensity of the reconstruction (Uisso and
Tanrivermis, 2021; Arowolo and Deng, 2018; Song et al., 2018); often the driving role of socio-economic
factors was more significant (Tubiello and Fischer, 2007; Ge et al., 2018). These studies analyzed
characteristics of grain production space reconstruction from different dimensions, and provided a series of
models and analysis frameworks for analyzing grain production space reconstruction. However, it was
necessary to establish a comprehensive framework to systematically analyze the reconstruction of grain
production space in terms of quantity, quality and spatial pattern. The Loess Plateau was one of the most serious soil erosion areas and the most fragile ecological
environment in China and the world. Since the reform and opening up, climate change and human activities
such as cultivated land reclamation, urban expansion, and returning farmland to forests and grasses,
significantly reconstructed the grain production space in the Loess Plateau. Therefore, this study selected
the Loess Plateau as a case study, and analyzed the characteristics and influencing factors of grain
production space reconstruction by developing a systematic analysis framework, aimed to solve the
following issues: Page 3/26 Page 3/26 (a) What was the principal pathway for the reconstruction of grain production space in quantity, quality, and
spatial pattern? (a) What was the principal pathway for the reconstruction of grain production space in quantity, quality, and
spatial pattern? (b) What were the factors that affected the reconstruction pathways in the Loess Plateau? The research results will provide a new research perspective on rational land systems management,
coordinate grain production, ecological protection and high-quality economic development, and provide a
theoretical basis of ensuring the sustainable and intensive development of grain production space and
ensuring grain security in the Loess Plateau. Study area Page 4/26
The Loess Plateau located between 33°41′N ~ 41°16′N and 100°54′E ~ 114°33′E, is the largest loess
geomorphic unit in the world, including Shanxi, Shaanxi, Henan, Inner Mongolia, and Gansu, Ningxia, Qinghai
7 provinces, with a total area of 635,000 km2 (Fig. 2). It is situated in a semi-arid and semi-humid climate zone, with concentrated precipitation in summer, broken terrain, loose soil, and frequent soil erosion. Hence,
it is one of the regions with the most serious soil erosion and vulnerable ecological environment in the world. Since 1980s, climate change and human activities significantly reconstructed the grain production space in
the Loess Plateau. Therefore, in this study, we selected 332 districts or counties in the Loess Plateau as the
study area. GAEZ model This research used the Global Agro-Ecological Zone (GAEZ) model of grain production potential under rain-
fed conditions (only considering the impact of climate on crop yield) to reflect the quality of grain production
space. The GAEZ model was developed jointly by the Food and Agriculture Organization of the United
Nations (FAO) and Institute of International Applied Systems Analysis (IIASA). The model estimated the
climatological suitability of a crop and then calculated crop potential yield by using a progressively limiting
method. For the detailed computation process of GAEZ, please refer to Global Agro-ecological Zones
(IIASA/FAO, 2010). The GAEZ model was widely used in the world, because of the availability basic data, the
simplicity of the calculation process and more accurate calculation results (Fischer et al., 2001; Xu et al.,
2017; Liao and Wei, 2021). Table 1 Table 1
Definition and data source of factors in influencing factor analysis
Factors
Definition
Data Source
Natural
environment
Average altitude
Average elevation of a unit
(m)
US Space Shuttle Radar Terrain Mission
Average slope
Average slope of a unit
(degrees)
River density
Ratio of river length to unit
area (km/ha)
National Basic Geographic Information System
Average
temperature
Variation of average
temperature in a unit (℃)
China Meteorological Administration
Average
precipitation
Variation of average
precipitation in a unit (mm)
Socio-economic
Total population
Variation of total
population in a unit
China Statistical Yearbook (The missing statistics
were inferred from the GM (1,1) model.)
GDP
Variation of GDP in a unit
(yuan)
Highway
mileage
Variation of highway
mileage in a unit (km)
Vectorized "Atlas of China (1981)" and "Atlas of China
Transportation (2019)" published by Sinomap press
Railway mileage
Variation of railway mileage
in a unit (km)
GAEZ model Definition and data source of factors in influencing factor analysis
Definition
Data Source Data sources The data required for this research included land use data, the data used to estimate the GAEZ model
(climate, soil and terrain data) and various factors in influencing factor analysis. The land use data covered seven provinces in the Loess Plateau in 1980, 2000, and 2018. We obtained from
Resource and Environment Science and Data Center (http://www.resdc.cn). Cultivated land that mainly
provided grain production function was defined as grain production space; construction land that promoted
economic development was divided into urban-rural development space; woodland and grassland that
provided ecological service function was classified as ecological service space. The climate data was downloaded from the China Meteorological Administration (http://data.cma.cn), and
included monthly precipitation, wet day frequency, mean maximum temperature, mean minimum
temperature, sunshine hours, mean relative humidity, mean wind speed, and total solar radiation in 1980
2000, and 2018. The Anusplin interpolation method was used to interpolate data from meteorological
stations. Soil data, including soil type, soil depth, soil water holding capacity and other attributes, was
obtained from the 1:1 million National Soil Data Set (http://www.vdb3.soil.csdb.cn/), which developed by
Chinese Academy of Sciences Resource and Environmental Science Data Center. Terrain data came from the
Digital Elevation Model (DEM) provided by the US Space Shuttle Radar Terrain Mission (SRTM), with a spatial
resolution of 90 m. Influencing factors of grain production space reconstruction had diverse and hierarchical, we argued that the
reconstruction pathways were mainly affected by natural environment factors and socio-economic factors
(Table 1). Page 5/26 Center of gravity migration model The center of gravity model was an important tool for studying the spatial changes of elements in the
process of regional reconstruction. The center of gravity migration can objectively reflect the agglomeration
characteristic and migration law of a certain element in the reconstruction process (Liu et al., 2009). Based
on this, this study used center of gravity migration model to describe the overall change trend and spatial
variation characteristics of grain production space. The basic model was as follows: Xti =
# (1.)
∑n
i=1(Ct×Xi)
∑n
i=1 Ct
Yti =
# (2)
∑n
i=1(Ct×Yi)
∑n
i=1 Ct where, Xti and Yti are the horizontal and vertical coordinates of the grain production space in year t in area i;
Ct is the area in year t in area i; Xi and Yi are the longitude and latitude of the geometric center of gravity in
area i where, Xti and Yti are the horizontal and vertical coordinates of the grain production space in year t in area i;
Ct is the area in year t in area i; Xi and Yi are the longitude and latitude of the geometric center of gravity in
area i. area i. Landscape pattern index Landscape pattern index was used to analyze landscape pattern reconstruction of grain production space,
and selected Patch Density (PD), Patch Area _ Mean (AREA_MN), Aggregation Index (AI), Landscape Division Page 6/26 Page 6/26 Index (DIVISION), Fractal Dimension Index _ Area-Weighted Mean (FRAC_AM) to reflected the characteristics
of the area, distribution, and shape reconstruction in grain production space. Quantity reconstruction of grain production space From 1980 to 2018, the proportion of grain production space in the Loess Plateau in land system remained
above 30%, and the overall trend was slightly declining with a net decrease of 9156km2 (Table 2). Table 2
Changes in quantity of grain production space in the
Loess Plateau from 1980 to 2018
Time
Quantity/km2
Proportion of land system
1980
204331
32.70%
2000
206262
33.01%
2018
195175
31.24% Table 2
Changes in quantity of grain production space in the
Loess Plateau from 1980 to 2018
Time
Quantity/km2
Proportion of land system
1980
204331
32.70%
2000
206262
33.01%
2018
195175
31.24% Although the overall quantity of grain production space showed a slight declining trend, its internal
reconstruction was frequent (Fig. 3). The quantity reconstruction of grain production in the Loess Plateau
had been increasing over time, mainly occurred after 2000. Since 2000, contribution rates of Grain for Green,
Deforestation and Reclamation, Urban Expansion, and Land Consolidation to grain production space
reconstruction were 42.80%, 39.42%, 8.32%, and 4.23%, respectively. Therefore, Grain for Green after 2000
were the principal pathway that affected the quantity reconstruction of grain production space. Spatial econometric regression model As grain production space usually showed the spatial autocorrelation characteristics, this study used the
spatial econometric regression model to measure the influencing factors of grain production space
reconstruction. The spatial econometric regression model of spatial autocorrelation can make up for the
shortcomings of traditional econometric models that ignore spatial correlation (Liu and Long, 2016; Tian et
al., 2020; Hao and Liu, 2016). We were concerned about the two most frequently used spatial econometric
models: Spatial Lag Model (SLM) and Spatial Error Model (SEM). The SLM mainly discusses whether each variable has a diffusion phenomenon (spillover effect) in a region. The spatial lag model (SLM) can be specified as: y = ρWy + Xβ + ϵ# (3) where, y is the dependent variable; X is the independent variable; W is the spatial weight matrix and Wy is a
vector of spatial lag dependent variable. ρ is a spatial regression coefficient that reflects the spatial
dependence of the sample observations, β is the parameter vector of X and ε is the model error term. where, y is the dependent variable; X is the independent variable; W is the spatial weight matrix and Wy is a
vector of spatial lag dependent variable. ρ is a spatial regression coefficient that reflects the spatial
dependence of the sample observations, β is the parameter vector of X and ε is the model error term. The SEM describes spatial disturbance correlation and spatial overall correlation. It can be written as: y = Xβ + ϵ# (4)
ϵ = λWϵ + μ# (5) y = Xβ + ϵ# (4)
ϵ = λWϵ + μ# (5) ϵ = λWϵ + μ# (5) Page 7/26 where, y, X, β and W are the same as those in Eq. (3), λ is the autoregressive parameter, µ is the random error
vector of the normal distribution and ε is the regression residual. Quality reconstruction of grain production space The average quality of grain production space was about 3,896.11 kg/hm2, and its total quality showed a
fluctuating growth trend in the Loess Plateau in the past 40 years. Due to differences in temperature,
precipitation, soil types, etc., the spatial distribution of grain production space quality showed a decreasing
distribution characteristic from southeast to northwest (Fig. 4). Assuming that there was only one reconstruction pathway in grain production space, and comparing the
influence of four reconstruction pathways on quality reconstruction, the results showed that (Table 3): Grain
for Green and Land Consolidation were reconstruction pathways to increase the quality of grain production
space, while Deforestation and Reclamation as well as Urban Expansion were pathways that affected the
quality deterioration of grain production space. However, the pathway of Grain for Green after 2000 had the
greatest impact on quality reconstruction of grain production space in the Loess Plateau. Page 8/26 Table 3
The influence of the reconstruction pathway of grain production space on quality reconstruction of grain
production space in the Loess Plateau
Stage
Reconstruction
pathway
Difference from the
average quality in
1980(kg/hm2)
Stage
Reconstruction
pathway
Difference from the
average quality in
2000(kg/hm2)
1980
—
2000
Grain for Green
16.61
2000
—
2018
Grain for Green
298.61
Urban Expansion
-14.16
Urban Expansion
-62.30
Deforestation
and Reclamation
-39.59
Deforestation
and Reclamation
-155.07
Land
Consolidation
0.03
Land
Consolidation
21.59 Pattern reconstruction of grain production space From 1980 to 2018, PD of grain production space in the Loess Plateau was on an increasing trend, but
AREA_MN has decreased overall (Fig. 5). It showed that the large patches were continuously divided into
small patches, and the area of grain production space was more fragmented. Fragmentation degree
increased more obviously after 2000. The changes in AI and DIVISION were more obvious after 2000, and
showed an obvious reverse development trend (Fig. 5). It indicated that the spatial distribution of grain
production space tended to be scattered in the Loess Plateau after 2000. FRAC_AM showed an increasing
trend from 1980 to 2000, but it represented a decreasing trend from 2000 to 2018. The change range was
between 1.25 and 1.28, and the change was relatively small. FRAC_AM was far from reaching the upper limit
of 1.5, indicating that the shape of the grain production space was complex, but there was a trend toward
regularization (Fig. 5). Therefore, since 1980, the impact of human activities on land use changes became
more and more intense, which leading to grain production space area was more fragmentation, the
distribution tended to be decentralized, and the shape was becoming more regular. Assumed there was only one reconstruction pathway in grain production space, and compared with the initial
year landscape pattern index. We found that reconstruction pathways after 2000 had a more profound
impact on landscape pattern reconstruction than 1980–2000, and Grain for Green after 2000 was the
principal pathway for grain production space reconstruction (Fig. 5). Grain for green led to an increase in PD
and a decrease in AREA_MN, which intensifying the fragmentation of grain production space; made AI
decrease and Division increase, which promoting the spatial distribution to be more scattered; and caused a
decrease in the value of FRAC_AM, which motivating the shape more regular. From 1980 to 2018, the quantity center of gravity of grain production space in the Loess Plateau moved
8.32km to the northwest. The quality center of gravity had the same migration direction, but the migration Page 9/26 Page 9/26 distance was long and reached 86.03km (Fig. 6). Compared with 1980–2000, various reconstruction
pathways after 2000 had a greater impact on the migration of the center of gravity (Fig. 6). The pathway of
Grain for Green from 2000 to 2018 had the greatest impact on the quantity and quality migration in grain
production space. Pattern reconstruction of grain production space Compared with 2000, this pathway caused quantity and quality center of gravity moved
southeast 20.79km and 29.71km, respectively. Driving factors for grain production space reconstruction Firstly, we calculated the Moran’s I value of grain production space reconstruction pathways in the Loess
Plateau. The results showed that (Table 4) there was a significant positive spatial autocorrelation for
different reconstruction pathways. Hence, it is necessary to establish the spatial econometric model to
analyze the driving factors of regional grain production space reconstruction. Then, we conducted traditional
OLS model and performed LMlag, LMerror, R-LMlag and R-LMerror tests of the two spatial econometric
models. As shown in Table 4, LMlag tests and R-LMlag tests were more significant for the reconstruction
pathways for Grain for Green, Deforestation and Reclamation as well as Land Consolidation, but the
statistics of LMerror tests and R-LMerror tests were more significant for Urban Expansion. In terms of the
fitting effect detection of the regression model, the smaller the R2 and Log Likelihood (LogL) and the greater
the Akaike Information Criterion (AIC) and Schwartz Criterion (SC), the better the model fitting effect
(Table 5). Therefore, SLM was employed in the estimation of Grain for Green, Deforestation and Reclamation
as well as Land Consolidation, and SEM was employed in the estimation of Urban Expansion. Table 4
Spatial dependence test of the reconstruction pathway of grain production space in the Loess Plateau
Spatial dependence
test
Grain for
Green
Urban
Expansion
Deforestation and
Reclamation
Land
Consolidation
Moran’s I
0.6241***
0.3051***
0.6062***
0.4542***
Lagrange Multiplier
(lag)
182.375***
30.23***
175.957***
129.322***
Robust LM (lag)
16.462***
1.434
22.989***
27.328***
Lagrange Multiplier
(error)
168.935***
43.76***
153.276***
105.439***
Robust LM (error)
3.022*
14.963***
0.308
3.444*
Note: *, ** and *** indicate statistical significance at 10%, 5% and 1% levels. Driving factors for grain production space reconstruction Table 4 Page 10/26 Table 5 The spatial regression analysis results of grain production space reconstruction and their driving factors in
the Loess Plateau
Driving
factors
Grain for Green
Urban Expansion
Deforestation and
Reclamation
Land Consolidation
OLS
SLM
OLS
SEM
OLS
SLM
OLS
SLM
Average
altitude
0.092c
-0.039a
0.002
0.001
-0.043
-0.020
0.003
0.002
Average
slope
16.920c
9.649c
2.009c
1.633c
10.694c
6.742c
2.012c
0.868c
River
density
132.352
6.684
-23.937
-8.526
104.924
-91.759
-28.056
0.629
Average
temperature
-37.326
-12.481
-1.563
1.766
-26.638
-7.083
7.325c
-2.901
Average
precipitation
1.007c
0.262b
0.067c
-0.063b
0.744c
0.157
-0.020
-0.013
Total
population
-0.288
-0.232
0.085
0.055
0.420
0.285
0.014
0.016
GDP
-0.054
0.008
0.022b
0.034c
-0.094
-0.016
-0.016b
-0.009
Highway
mileage
0.823c
0.486c
0.047b
0.067c
0.768c
0.445c
0.034b
0.024b
Railway
mileage
0.373
0.358
0.227c
0.188c
0.401
0.302
0.058
0.060b
W-Y
0.686c
0.701c
0.727c
Lambda
0.439c
R2
0.375
0.650
0.295
0.395
0.331
0.628
0.146
0.496
LogL
2208.13
2131.41
1569.03
1550.65
2187.50
2110.84
1488.76
1424.08
AIC
4436.26
4284.82
3158.06
3121.32
4394.99
4243.69
2997.51
2870.16
SC
4474.31
4326.68
3196.11
3159.37
4433.04
4285.54
3035.56
2912.02
Note: a: significance at 10%; b: significance at 5%; c: significance at 1%. The pathway of Grain for Green was significantly correlated with average slope, highway mileage, average
precipitation, and average altitude (Table 5), indicating that the reconstruction mainly occurred in areas with
steeper slopes, lower terrain, increased precipitation, and more new highways. Among the above driving
factors, average slope had the most significant impact of this reconstruction. In order to curb soil erosion
and solve a series of ecological problems, the Loess Plateau actively responded to the national policy of
Grain for Green. 64.35% grain production space with a slope above 25° was transformed into forests or Page 11/26 grasslands. Grain for Green also mostly occurred in areas where precipitation increased, the increase of
precipitation provided abundant water for vegetation restoration. Convenient transportation made it easier
for farmers go out, and an increasing number of farmers planted trees instead of grain on their own grain
production space when they went out to find work. The pathway of Urban Expansion had a high correlation with GDP, highway mileage, railway mileage, average
slope, and average precipitation (Table 5). It showed that Urban Expansion mostly occurred in areas with
rapid economic development, rapid transportation network improvement, gentle slopes, and reduced
precipitation. Urban expansion had the most affected driving factors among the four reconstruction
pathways. Driving factors for grain production space reconstruction In particular, GDP change only had a significant effect on Urban Expansion, but had no significant
effect on the other three reconstruction pathways, it indicating that areas with superior economic conditions
were more likely to be transformed into urban-rural development space. Meanwhile, low slope provided the
nature basis for the reconstruction pathway. The pathway of Urban Expansion had a high correlation with GDP, h
slope, and average precipitation (Table 5). It showed that Urban Ex
rapid economic development, rapid transportation network improv
precipitation. Urban expansion had the most affected driving facto
pathways. In particular, GDP change only had a significant effect o
effect on the other three reconstruction pathways, it indicating that
were more likely to be transformed into urban-rural development sp
nature basis for the reconstruction pathway. The pathway of Deforestation and Reclamation was highly correlated with average slope and highway
mileage in a county (Table 5). In the early stage of reform and opening up, grain production had a higher
priority in production and life. But constrained by lower capital and technology input, expanding grain
production scale was an inevitable choice to ensure grain security. Against this background, the government
encouraged farmers to desforest and reclaim on steep slopes. With the increase of highway mileage, the
convenience of grain export increased income, which stimulating farmers to continue to reclaim. The pathway of Land Consolidation had a significant correlation with the average slope and the change of
the transportation network (Table 5). This pathway mainly occurred after 2000, the developed transportation
network was conducive to the migration of rural population to the cities, and promoted the non-
agriculturalization and part-time employment of rural labor, so a large number of homesteads were left
unused. Under the guidance of policies (dynamic balance between farmlands and construction lands,
balance of cultivated land occupation and compensation etc.), a large number of idle land resources had
been revitalized, and effectively increasing the space for grain production. In addition, lower slope was
conducive to agricultural production activities, and a large amount of idle land was more likely to be
reclaimed. In general, the four pathways of grain production space reconstruction in the Loess Plateau were the result of
the combined effects of natural environment and socio-economic driving factors. Driving factors for grain production space reconstruction The spatial econometric
regression model results at the county level indicated that all factors had significant impacts on four
pathways, except for river density, average temperature, and total population. Average slope and highway
mileage had significant effect on all reconstruction pathways, but their direction of action and intensity were
different. Moreover, average altitude and GDP only had a significant impact on one reconstruction pathway
(Table 4). The spatial econometric regression analysis results of Deforestation and Reclamation was not
ideal, because the reconstruction pathway was too concentrated in the Loess Plateau; analysis of the other
three reconstruction pathways was ideal. Discussion Page 12/26 Since the reform and opening up, under the dual pressures of reducing supply and increasing demand of
grain production space, Deforestation and Reclamation was always the most common, effective and direct
pathway to replenish grain production space (Foley et al., 2005; Chazdon, 2008). Large-scale Deforestation
and Reclamation had aggravated soil erosion, and a series of ecological and environmental problems had
become prominent in the Loess Plateau. In order to slow down soil erosion and restore vegetation, the
government implemented Grain for Green Project since 1999. Different from the slow response of other
countries’ land use changes in national policies, the Loess Plateau responded to national policies actively
and quickly (Meyfroidt et al., 2013) and made a great contribution to improve China's forest coverage. Although the pathway of Grain for Green has reduced the number of food production spaces, it has caused
poor-quality grain production space withdrawing from planting field, and agricultural production factors
concentrated in higher-quality areas, thereby improving the quality of grain production space. Under the
unified guidance of the national plan, Grain for Green promoted the shape more regular and the distribution
more concentrated of grain production space. Poor-quality grain production space reclaimed on steep slopes
was returned to forest or grassland. The original grain production space was divided severely, which
intensified the fragmentation of grain production space. Because the number of Grain for Green after 2000
was large and concentrated in the southwest and north of the Loess Plateau, this pathway had a great
impact on the quantity and quality center of gravity of grain production space, and the center of gravity of
grain production moved to the southeast. Therefore, Grain for Green after 2000 were the principal pathway
for grain production space reconstruction in quantity, quality, and spatial pattern in the Loess Plateau. We studied the influencing factors of grain production space reconstruction in the Loess Plateau, and found
that the reconstruction was the result of a combination of natural environment and socio-economic factors. In the influencing factors analysis, river density, average temperature, and total population did not show a
direct correlation. Water source was an important factor restricting grain production in the Loess Plateau, but
with the advancement of science and technology, sprinkler irrigation, drip irrigation and other technologies
had solved the problem of water shortage in grain production. Discussion Therefore, river density did not show a
significant correlation in the process of grain production space reconstruction. In the past hundred years,
climate warming was not a local feature (IPCC, 2014; Newan et al., 2014), so temperature changes had not
triggered the reconstruction in the Loess Plateau. Grain production space reconstruction was due to the
grain pressure caused by population increase rapidly, but the mainly reconstruction areas were spatially
misplaced with densely populated areas. Therefore, in the analysis of the factors affecting the reconstruction
of grain production space, river density, average temperature, and total population did not show a direct
correlation. System, policy, economy, location and other factors drive the subjective trade-offs of land system functions
by different stakeholders, and changing the land use (Karimi and Hockings, 2018). The subjective trade-offs
of government management departments and individual mainly drive grain production space reconstruction
through two mechanisms: group crisis response and individual interest driven (Fig. 7). System changes and
policy adjustments play a leading role in the process of grain production space reconstruction (da Silva et al.,
2017). Ecological protection policies and land management systems can affect directly the grain production
space reconstruction, and macro-control policies and regional development policies can affect indirectly the
reconstruction (Gong and Liu, 2021; Tian et al., 2020; Zewdie et al., 2018). The dissemination of cultural Page 13/26 Page 13/26 knowledge and the transformation of value concepts could drive the reconstruction of grain production
space through the process of transmission and feedback between different levels of politics, economy, and
society (Ma et al., 2019). In the context of rising agricultural production costs and higher income from going
out to work, after weighing their benefits, stakeholders actively change land use to obtain higher income,
thereby reconstructing grain production space. Factors such as the deterioration of the rural ecological
environment and the increase in employment opportunities in cities have led to rural population migration
and structural adjustment, which affecting inevitably the reconstruction of grain production space. Hence,
population change is the most direct driving force for grain production space reconstruction (da Silva et al.,
2017; Ducey et al., 2018). The spatial heterogeneity of location can determine the spatial evolution of
geographical elements. Discussion Being close to big cities is susceptible to the diffusion of capital, technology, talents,
technology, transportation and other factors, which driving the development of secondary and tertiary
industries, then weakening the position of agriculture in economic activities. Superior location conditions can
promote the transformation of grain production space into urban-rural development space (Zhang et al.,
2018a). The promotion of agricultural technology could advance the intensive use of high-quality land and
abandon poor-quality land, so less land will produce more goods and services (Zhang et al., 2018a). Therefore, system, policy, income, population, culture, concept, technology and location affect the
reconstruction of grain production space by different stakeholders. Fig.7 Grain production space reconstruction driving mechanism When the income of the original land management mode is lower
than the new management mode (or the opportunity cost of agricultural production) due to economic, policy,
technology and other reasons, grain production space will be reconstructed (Uzuner and Adewale, 2019). The
increase in population brings about an increase in the demand for grain. Driven by economic benefits, economic benefits (Nguyen et al., 2016). When the income of the original land management mode is lower
than the new management mode (or the opportunity cost of agricultural production) due to economic, policy,
technology and other reasons, grain production space will be reconstructed (Uzuner and Adewale, 2019). The
increase in population brings about an increase in the demand for grain. Driven by economic benefits,
farmers deforest and reclaim on steep slopes. Individual in areas with good location, flat terrain and suitable
climate tend to increase capital and technology input, and transform grain production space into urban-rural
development space to gain more benefits. Social and economic development bring more employment
opportunities and income to individuals, and increases the opportunity cost of agricultural production. Driven
by individual economic interests, farmers abandon their original grain production space for population
migration and restore to natural vegetation in original grain production space (Subedi et al., 2021; Heider et
al., 2021). In addition, the national ecological compensation policies also lead farmers to return farmland to
forests or grasses spontaneously to obtain a higher income than farmland. This research also has some limitations. First of all, grain production space refers to the land with grain
production as its main function. This research defines cultivated land as grain production space, but in
reality, the land has multi-functional attributes, cultivated land and grain production space are not completely
coupled. In future studies, we may strive to adopt more rigorous land division methods, and scientifically and
objectively divide grain production space, ecological service space and urban-rural development space. Secondly, based on different stakeholders, group crisis response mechanism and individual interest driven
mechanism that affect grain production space reconstruction is proposed. In our next study, we will select a
series of representative evaluation indicators to study quantitatively group crisis response mechanism and
individual interest driven mechanism, in order to analyze driving mechanism of grain production space
reconstruction more objectively, comprehensively and scientifically. Fig.7 Grain production space reconstruction driving mechanism In the future, the similar analysis
framework can be used to explore the reconstruction characteristics of other spaces in different regions and
scales, and identify its principal reconstruction pathway. This research also has some limitations. First of all, grain production space refers to the land with grain
production as its main function. This research defines cultivated land as grain production space, but in
reality, the land has multi-functional attributes, cultivated land and grain production space are not completely
coupled. In future studies, we may strive to adopt more rigorous land division methods, and scientifically and
objectively divide grain production space, ecological service space and urban-rural development space. Fig.7 Grain production space reconstruction driving mechanism Fig.7 Grain production space reconstruction driving mechanism Group crisis response mechanism (Fig. 7) is that when grain, development or ecological crises occur on
limited land resources, endangering the survival and development of human society, the government or
management department will take measures to deal with the crises by the group, which is the top-down grain
production space reconstruction (Ariti et al., 2019). In the early stages of economic development or in
ecologically fragile areas, expanding the scale of grain production is an advisable choice to ensure grain
security. Therefore, the government or management department will encourage Deforestation and
Reclamation. The natural environment determines the limited amount of land suitable for agricultural
production. It is necessary to achieve agricultural intensification and reduce the demand for land. In addition,
increasing population requires more production and living land. Therefore, the government and management
departments expand urban-rural development space and increase economic benefits through urban
expansion and land planning. When urban expansion threatens grain security, the government may formulate
different policies (basic farmland preservation areas, dynamic balance between farmlands and construction
lands, etc.) to slow down the shrinkage of grain production space. Meanwhile, the government continues to
increase technical input, a large number of idle lands are reclaimed under the guidance of policies. Hence,
grain production space is restructured. Deforestation and Reclamation and Urban Expansion have destroyed
the stability and integrity of the ecosystem, forcing government to take measures such as Grain for Green to
restore and protect the ecosystem, and grain production space is reconstructed accordingly. Individual interest driven mechanism reconstruct the grain production space from down to top, means that
land users adjust land use in order to pursue their own maximum economic benefits under the background
of changes in land rent, land benefits and opportunity costs (Fig. 7). Individuals always pursue the
maximization of their own economic benefits, while ignoring ecological and social benefits. Therefore,
different stakeholders operate and use land resources in a way that they believe can generate the greatest Individual interest driven mechanism reconstruct the grain production space from down to top, means that
land users adjust land use in order to pursue their own maximum economic benefits under the background
of changes in land rent, land benefits and opportunity costs (Fig. 7). Individuals always pursue the economic benefits, while ignoring ecological and social benefits. Therefore, Page 14/26 Page 14/26 economic benefits (Nguyen et al., 2016). Conclusion The land systems have multiple functions, and there is trade-off/synergy between different functions. Grain
production space reconstruction concentrates the main contradictions of the land system changes. Therefore, this study analyzed the reconstruction and influencing factors of grain production space in the
Loess Plateau. The goal was to answer the question: “what was the principal pathway for the reconstruction
and what were the factors that affected the reconstruction pathways of grain production space?” The results
showed that the quantity of grain production space in the Loess Plateau declined, but the quality improved
between 1980 and 2018. The spatial pattern of grain production space tended to be fragmented, scattered
and regular. Quantity and quality center of gravity moved to the northwest, the quality center of gravity moved
farther. Grain for Green after 2000 was the principal pathway in quantity, quality and spatial pattern
reconstruction of grain production space. The reconstruction pathways of Grain for Green, Urban Expansion,
Deforestation and Reclamation as well as Land Consolidation were the result of a combination of natural
environment and socio-economic factors, but influencing factors had different strengths and directions for
each reconstruction pathway in the Loess Plateau. Therefore, it is necessary to strengthen the management
of the land systems, coordinate the relationship among grain production, ecological protection and high- The land systems have multiple functions, and there is trade-off/synergy between different functions. Grain
production space reconstruction concentrates the main contradictions of the land system changes. Page 15/26 Page 15/26 quality economic development, and promote the sustainable and intensified development of grain production
space in the Loess Plateau. quality economic development, and promote the sustainable and intensified development of grain production
space in the Loess Plateau. Competing interests The authors declare no competing interests. Competing interests The authors declare no competing interests. Declarations Author contribution Zhangxuan Qin: conceptualization, visualization, formal analysis, and writing—original
draft. Xiaolin Liu: methodology and software. Xiaoyan Lu: data collection. Mengfei Li: writing—review and
editing. Fei Li: conceptualization, supervision and funding acquisition. All authors read and approved the final
manuscript. Funding The work was financially supported by the National Natural Science Foundation of China (Grant No. 42171197). Data availability The datasets used or analyzed during the current study are available from the corresponding
author on reasonable request. Consent for publication Not applicable. Competing interests The authors declare no competing interests. References 1. Al-Kayssi AW, Mustafa SH (2016) Impact of elevated carbon dioxide on soil heat storage and heat flux
under unheated low-tunnels conditions. J Environ Manag 182:176-186 1. Al-Kayssi AW, Mustafa SH (2016) Impact of elevated carbon dioxide on soil heat storage and heat flux
under unheated low-tunnels conditions. J Environ Manag 182:176-186 2. Ariti AT, van Vliet J, Verburg PH (2019) The role of institutional actors and their interactions in the land
use policy making process in Ethiopia. J Environ Manag 237:235–24 2. Ariti AT, van Vliet J, Verburg PH (2019) The role of institutional actors and their interactions in the land
use policy making process in Ethiopia. J Environ Manag 237:235–24 3. Arowolo AO, Deng XZ (2018) Land use/land cover change and statistical modelling of cultivated land
change drivers in Nigeria. Reg Environ Change 18(1):247-259 4. Arowolo AO, Deng XZ, Olatunji OA, Obayelu AE (2018) Assessing changes in the value of ecosystem
services in response to land-use/land-cover dynamics in Nigeria. Sci Total Environ 636:597-609 5. Bateman I, Agarwala M, Binner A, Coombes E, Day B, Ferrini S, Fezzi C, Hutchins M, Lovett A, Posen P
(2016) Spatially explicit integrated modeling and economic valuation of climate driven land use change
and its indirect effects. J Environ Manag 181:172-184 6. Bren d’Amour C, Reitsma F, Baiocchi G, Barthel S, Güneralp B, Erb KH, Haberl H, Creutzig F, Seto KC
(2017) Future urban land expansion and implications for global croplands. Proc Natl Acad Sci U.S.A. 114(34):8939-8944 7. Canaz S, Aliefendioglu Y, Tanrivermis H (2017) Change detection using Landsat images and an analysis
of the linkages between the change and property tax values in the Istanbul Province of Turkey. J Environ
Manag 200:446-455 8. Chazdon RL (2008) Beyond deforestation, Restoring forests and ecosystem services on degraded lands. Science 320(5882):1458-1460 8. Chazdon RL (2008) Beyond deforestation, Restoring forests and ecosystem services on degraded lands. Science 320(5882):1458-1460 Page 16/26 9. Cheng W, Gao X, Ma T, Xu X, Chen Y, Zhou C (2018) Spatial-temporal distribution of cropland in China
based on geomorphologic regionalization during 1990-2015. Acta Geogr Sin 73(9):1613-1629 10. da Silva RFB, Batistella M, Moran EF (2017) Socioeconomic changes and environmental policies as
dimensions of regional land transitions in the Atlantic Forest, Brazil. Environ Sci Pol 74:14-22 11. Decey MJ, Johnson KM, Belair EP, Cook BD (2018) The Influence of Human Demography on Land Cover
Change in the Great Lakes States, USA. References Environ Manage 62(6):1089-1107 12. Fischer G, Sun LX (2001) Model based analysis of future land-use development in China. Agr Ecosyst
Environ 85(1-3):163-176 12. Fischer G, Sun LX (2001) Model based analysis of future land-use development in China. Agr Ecosyst
Environ 85(1-3):163-176 13. Foley JA, DeFries R, Asner GP, Barford C, Bonan G, Carpenter SR, Chapin FS, Coe MT, Daily GC, Gibbs HK,
Helkowski JH, Holloway T, Howard EA, Kucharik CJ, Monfreda C, Patz JA, Prentice IC, Ramankutty N,
Snyder PK (2005) Global Consequences of Land Use. Science 309(5734):570-574 14. Fu G, Bai W (2015) Advances and prospects of evaluating cultivated land quality. Resour Sci 37(2):226-
236 15. Fuentes-Guevara MD, Armindo RA, Timm LC, Faria LC (2022) Examining the land leveling impacts on the
physical quality of lowland soils in Southern Brazil. Soil Till Res 215:105217 15. Fuentes-Guevara MD, Armindo RA, Timm LC, Faria LC (2022) Examining the land leveling impacts on the
physical quality of lowland soils in Southern Brazil. Soil Till Res 215:105217 16. Gao L, Bryan BA (2017) Finding pathways to national-scale land-sector sustainability. Nature
544(7649):217-222 16. Gao L, Bryan BA (2017) Finding pathways to national-scale land-sector sustainability. Nature
544(7649):217-222 17. Ge D, Long H, Zhang Y, Ma L, Li T (2018) Farmland transition and its influences on grain production in
China. Land Use Pol 70:94-105 17. Ge D, Long H, Zhang Y, Ma L, Li T (2018) Farmland transition and its influences on grain production in
China. Land Use Pol 70:94-105 18. Gong B, Liu Z (2021) Assessing impacts of land use policies on environmental sustainability of oasis
landscapes with scenario analysis: the case of northern China. Landsc Ecol 36(7):1913-1932 18. Gong B, Liu Z (2021) Assessing impacts of land use policies on environmental sustainability of oasis
landscapes with scenario analysis: the case of northern China. Landsc Ecol 36(7):1913-1932 19. Hailu A, Mammo S, Kidane M (2020) Dynamics of land use, land cover change trend and its drivers in
Jimma Geneti District, Western Ethiopia. Land Use Pol 99:105011 19. Hailu A, Mammo S, Kidane M (2020) Dynamics of land use, land cover change trend and its drivers in
Jimma Geneti District, Western Ethiopia. Land Use Pol 99:105011 20. Hao Y, Liu Y (2016) The influential factors of urban PM2.5 concentrations in China, a spatial econometric
analysis. J Clean Prod 112:1443-1453 20. References Hao Y, Liu Y (2016) The influential factors of urban PM2.5 concentrations in China, a spatial econometri
analysis. J Clean Prod 112:1443-1453 21. Hasegawa T, Havlík P, Frank S, Palazzo A, Valin H (2019) Tackling food consumption inequality to fight
hunger without pressuring the environment. Nat Sustain 2(9):826-833 21. Hasegawa T, Havlík P, Frank S, Palazzo A, Valin H (2019) Tackling food consumption inequality to fight
hunger without pressuring the environment. Nat Sustain 2(9):826-833 22. Heider K, Lopez JMR, Balbo AL, Scheffran J (2021) The state of agricultural landscapes in the
Mediterranean: smallholder agriculture and land abandonment in terraced landscapes of the Ricote
Valley, southeast Spain. Reg Environ Change 21(1):23 22. Heider K, Lopez JMR, Balbo AL, Scheffran J (2021) The state of agricultural landscapes in the
Mediterranean: smallholder agriculture and land abandonment in terraced landscapes of the Ricote
Valley, southeast Spain. Reg Environ Change 21(1):23 23. Homer C, Dewitz J, Jin SM, Xia G, Costello C, Danielson P, Gass L, Funk M, Wickham J, Stehman S, Auch
R, Riitters K (2020) Conterminous United States land cover change patterns 2001-2016 from the 2016
National Land Cover Database. ISPRS J Photogrammetry Remote Sens 162:184-199 23. Homer C, Dewitz J, Jin SM, Xia G, Costello C, Danielson P, Gass L, Funk M, Wickham J, Stehman S, Auch
R, Riitters K (2020) Conterminous United States land cover change patterns 2001-2016 from the 2016
National Land Cover Database. ISPRS J Photogrammetry Remote Sens 162:184-199 24. Huang D, Zhu S, Liu T (2022) Are there differences in the forces driving the conversion of different non-
urban lands into urban use? A case study of Beijing. Environ Sci Pollut R 29(5):6414-6432 24. Huang D, Zhu S, Liu T (2022) Are there differences in the forces driving the conversion of different non-
urban lands into urban use? A case study of Beijing. Environ Sci Pollut R 29(5):6414-6432 25. IIASA/FAO (2010) Global Agro-ecological Zones (GAEZ v3.0). IIASA, Laxenburg, Austria and FAO, Rome,
Italy. http://www.fao.org/nr/gaez/en 25. IIASA/FAO (2010) Global Agro-ecological Zones (GAEZ v3.0). IIASA, Laxenburg, Austria and FAO, Rome,
Italy. http://www.fao.org/nr/gaez/en 26. IPCC Climate change (2014) Impact, Adaptation, and Vulnerability. Cambridge: Cambridge University
Press, in press. [2014-05-06]. http://www.ipcc.ch/report/ar5/wg2/ 26. IPCC Climate change (2014) Impact, Adaptation, and Vulnerability. Cambridge: Cambridge University
Press, in press. [2014-05-06]. http://www.ipcc.ch/report/ar5/wg2/ Page 17/26 Page 17/26 27. Irwin EG, Bockstael NE (2007) The evolution of urban sprawl, Evidence of spatial heterogeneity and
increasing land fragmentation. References Proc Natl Acad Sci U.S.A. 104(52):20672-20677 28. Jiang P, Li M, Lv J (2019) The causes of farmland landscape structural changes in different geographical
environments. Sci Total Environ 685:667-680 29. Jin H, Zhong Y, Shi D, Li J, Lou Y, Li Y, Li J (2021) Quantifying the impact of tillage measures on the
cultivated-layer soil quality in the red soil hilly region, Establishing the thresholds of the minimum data
set. Ecol Indic 130:108013 30. Johnson JA, Runge CF, Senauer B, Foley J, Polasky S (2014) Global agriculture and carbon trade-offs. Proc. Natl Acad Sci U.S.A. 111(34): 12342-12347 31. Ju H, Zuo L, Zhang Z, Zhao X, Wang X, Wen Q, Liu F, Xu J, Liu B, Yi L, Hu S, Sun F, Tang Z (2020) Methods
research on describing the spatial pattern of land use types in China. Acta Geogr Sin 75(1)：143-159 31. Ju H, Zuo L, Zhang Z, Zhao X, Wang X, Wen Q, Liu F, Xu J, Liu B, Yi L, Hu S, Sun F, Tang Z (2020) Methods
research on describing the spatial pattern of land use types in China. Acta Geogr Sin 75(1)：143-159 32. Kang Q, Guo Q, Ding Y, Zhang Y (2021) Tradeoffs/synergies analysis of "Production-Living-Ecological"
functions in Shanxi province. J Nat Resour 36(5):1195-1207 32. Kang Q, Guo Q, Ding Y, Zhang Y (2021) Tradeoffs/synergies analysis of "Production-Living-Ecological"
functions in Shanxi province. J Nat Resour 36(5):1195-1207 33. Karimi A, Hockings M (2018) A social-ecological approach to land-use conflict to inform regional and
conservation planning and management. Landsc Ecol 33(5):691-710 33. Karimi A, Hockings M (2018) A social-ecological approach to land-use conflict to inform regional and
conservation planning and management. Landsc Ecol 33(5):691-710 34. Kurmangozhinov A, Xue W, Li X, Zeng F, Sabit R, Tusun T (2020) High biomass production with abundant
leaf litterfall is critical to ameliorating soil quality and productivity in reclaimed sandy desertification
land. J Environ Manag 263:110373 35. Leonard B, Parker D, Anderson TL (2020) Land quality, land rights, and indigenous poverty. J Dev Econ
143:102435 36. Leyk S, Uhl JH, Connor DS, Braswell AE, Mietkiewicz N, Balch JK, Gutmann M (2020) Two centuries of
settlement and urban development in the United States. Sci Adv 6(23): eaba2937 37. Li F, Qin Z, Liu X, Chen Z, Wei X, Zhang Q, Lei M (2021) Grain production space reconstruction and land
system function tradeoffs in China. Geogr Sustain 2(1):22-30 38. References Ma E, Cai J, Lin J, Han Y, Liao L, Han W (2019) Explanation of land use/cover change from the
perspective of tele-coupling. Acta Geogr Sin 74(3):421-431 46. Meyfroidt P, Chowdhury RR, de Bremond A, Ellis EC, Erb KH, Filatova T, Garrett RD, Grove JM, Heinimann
A, Kuemmerle T, Kull CA, Lambin EF, Landon Y, de Warow YL, Messerli P, Muller D, Nielsen JO, Peterson
GD, Garcia VR, Schluter M, Turner BL, Verburg PH (2018) Middle-range theories of land system change. Global Environ Change 53:52-67 47. Meyfroidt P, Lambin EF, Erb KH, Hertel TW (2013) Globalization of land use, distant drivers of land
change and geographic displacement of land use. Curr Opin Env Sust 5(5):438-444 48. Newan ME, McLaren KP, Wilson BS (2014) Long-term socio-economic and spatial pattern drivers of land
cover change in a Caribbean tropical moist forest, the Cockpit Country, Jamaica. Agrc Ecosyst Environ
186:185-200 49. Nguyen THT, Tran VT, Bui QT, Man QH, Walter TD (2016) Socio-economic effects of agricultural land
conversion for urban development, Case study of Hanoi, Vietnam. Land Use Pol 54:583-592 50. Obersteiner M, Walsh B, Frank S, Havlik P, Cantele M, Liu J, Palazzo A, Herrero M, Lu Y, Mosnier A, Valin H,
Riahi K, Kraxner F, Fritz S, van Vuuren D (2016) Assessing the land resource–food price nexus of the
Sustainable Development Goals. Sci Adv 2(9): e1501499 51. Raghavan R, Shrimali G (2015) Forest cover increase in India, The role of policy and markets. For Pol
Econ 61:70-76 51. Raghavan R, Shrimali G (2015) Forest cover increase in India, The role of policy and markets. For Pol
Econ 61:70-76 52. Rossiter DG (1996) A theoretical framework for land evaluation. Geoderma 72(3-4):165-190 53. Song XP, Hansen MC, Stehman SV, Potapov PV, Tyukavina A, Vermote EF, Townshend JR (2018) Global
land change from 1982 to 2016. Science 560:639-643 53. Song XP, Hansen MC, Stehman SV, Potapov PV, Tyukavina A, Vermote EF, Townshend JR (2018) Global
land change from 1982 to 2016. Science 560:639-643 54. Subedi YR, Kristiansen P, Cacho O, Ojha RB (2021) Agricultural Land Abandonment in the Hill Agro-
ecological Region of Nepal: Analysis of Extent, Drivers and Impact of Change. Environ Manage
67(6):1100-1118 54. Subedi YR, Kristiansen P, Cacho O, Ojha RB (2021) Agricultural Land Abandonment in the Hill Agro-
ecological Region of Nepal: Analysis of Extent, Drivers and Impact of Change. Environ Manage
67(6):1100-1118 55. References Li F, Zhou M, Qin Z (2020) Generated land systems, recognition and prospects of land system science. Environ Rev 28(2):199-207 39. Liao M, Wei F (2021) Analysis of grain production and development potentiality in Central Asia based on
GAEZ method. J Nat Resour 36(3):582-593 40. Liu J, Kuang W, Zhang Z, Xu X, Qin Y, Ning J, Zhou W, Zhang S, Li R, Yan C, Wu S, Shi X, Jiang N, Yu D, Pan
X, Chi W (2014) Spatiotemporal characteristics, patterns and causes of land use changes in China since
the late 1980s. Acta Geogr Sin 69(1):3-14 40. Liu J, Kuang W, Zhang Z, Xu X, Qin Y, Ning J, Zhou W, Zhang S, Li R, Yan C, Wu S, Shi X, Jiang N, Yu D, Pan
X, Chi W (2014) Spatiotemporal characteristics, patterns and causes of land use changes in China since
the late 1980s. Acta Geogr Sin 69(1):3-14 41. Liu J, Ning J, Kuang W, Xu X, Zhang S, Yan C, Li R, Wu S, Hu Y, Du G, Chi W, Pan T, Ning J (2018) Spatio-
temporal patterns and characteristics of land-use change in China during 2010-2015. Acta Geogr Sin
73(5):789-802 41. Liu J, Ning J, Kuang W, Xu X, Zhang S, Yan C, Li R, Wu S, Hu Y, Du G, Chi W, Pan T, Ning J (2018) Spatio-
temporal patterns and characteristics of land-use change in China during 2010-2015. Acta Geogr Sin
73(5):789-802 42. Liu X, Zhao C, Song W (2017) Review of the evolution of cultivated land protection policies in the period
following China's reform and liberalization. Land Use Pol 67:660-669 42. Liu X, Zhao C, Song W (2017) Review of the evolution of cultivated land protection policies in the period
following China's reform and liberalization. Land Use Pol 67:660-669 43. Liu Y, Long H (2016) Land use transitions and their dynamic mechanism in the Huang-Huai-Hai Plain. Acta Geogr Sin 71(4):666-679 43. Liu Y, Long H (2016) Land use transitions and their dynamic mechanism in the Huang-Huai-Hai Plain. Acta Geogr Sin 71(4):666-679 Page 18/26 44. Liu Y, Wang J, Guo L (2009) The Spatial-Temporal Changes of Grain Production and Arable Land in
China. Sci Agric Sin 42(12):4269-4274 45. Ma E, Cai J, Lin J, Han Y, Liao L, Han W (2019) Explanation of land use/cover change from the
perspective of tele-coupling. Acta Geogr Sin 74(3):421-431 45. References Tian J, Wang B, Zhang C, Li W, Wang S (2020) Mechanism of regional land use transition in
underdeveloped areas of China, A case study of northeast China. Land Use Pol 94:104538 56. Tubiello FN, Fischer G (2007) Reducing climate change impacts on agriculture, Global and regional
effects of mitigation, 2000–2080. Technol Forecast Soc Chang 74(7):1030-1056 57. Uisso AM, Tanrivermis H (2021) Driving factors and assessment of changes in the use of arable land in
Tanzania. Land Use Pol 104:105359 58. Uzuner G, Adewale AA (2019) Does asymmetric nexus exist between agricultural land and the housing
market? Evidence from non-linear ARDL approach. Environ Sci and Pollut R 26(8): 7677-7687 58. Uzuner G, Adewale AA (2019) Does asymmetric nexus exist between agricultural land and the housing
market? Evidence from non-linear ARDL approach. Environ Sci and Pollut R 26(8): 7677-7687 59. van Vliet JV, de Groot HLF, Rietveld P, Verburg PH (2015) Manifestations and underlying drivers of
agricultural land use change in Europe. Landsc Urban Plann 133:24-36 59. van Vliet JV, de Groot HLF, Rietveld P, Verburg PH (2015) Manifestations and underlying drivers of
agricultural land use change in Europe. Landsc Urban Plann 133:24-36 60. Veljanoska S (2018) Can Land Fragmentation Reduce the Exposure of Rural Households to Weather
Variability? Ecol Econ 154:42-51 Page 19/26 61. Verburg PH, Crossman N, Ellis EC, Heinimann A, Hostert P, Mertz O, Nagendra H, Sikor T, Erb KH,
Golubiewski N, Grau R, Grove M, Konate S, Meyfroidt P, Parker DC, Chowdhury RR, Shibata H, Thomson A,
Zhen L (2015) Land system science and sustainable development of the earth system, A global land
project perspective. Anthropocene 12:29-41 62. Wästfelt A, Zhang Q (2016) Reclaiming localisation for revitalising agriculture, A case study of peri-urban
agricultural change in Gothenburg, Sweden. J Rural Stud 47:172-185 62. Wästfelt A, Zhang Q (2016) Reclaiming localisation for revitalising agriculture, A case study of peri-urban
agricultural change in Gothenburg, Sweden. J Rural Stud 47:172-185 63. Winkler k, Fuchs R, Rounsevell M, Herold M (2021) Global land use changes are four times greater than
previously estimated. Nat Commun 12(1):2501 63. Winkler k, Fuchs R, Rounsevell M, Herold M (2021) Global land use changes are four times greater than
previously estimated. Nat Commun 12(1):2501 64. Xu X, Wang L, Cai H, Wang L, Liu L, Wang H (2017) The influences of spatiotemporal change of cultivated
land on food crop production potential in China. References Food Secur 9:485-495 64. Xu X, Wang L, Cai H, Wang L, Liu L, Wang H (2017) The influences of spatiotemporal change of cultivated
land on food crop production potential in China. Food Secur 9:485-495 65. Yang H, Sun L, Zhou M, Li F (2022) Trade-off analyses of ecosystem services during the reconstruction
of grain production space in Loess Plateau, A case study of Yulin City. Arid Land Geogr 45(1):226-236 65. Yang H, Sun L, Zhou M, Li F (2022) Trade-off analyses of ecosystem services during the reconstruction
of grain production space in Loess Plateau, A case study of Yulin City. Arid Land Geogr 45(1):226-236 66. Zewdie M, Worku H, Bantider A (2018) Temporal Dynamics of the Driving Factors of Urban Landscape
Change of Addis Ababa During the Past Three Decades. Environ Manage 61(1):132-146 66. Zewdie M, Worku H, Bantider A (2018) Temporal Dynamics of the Driving Factors of Urban Landscape
Change of Addis Ababa During the Past Three Decades. Environ Manage 61(1):132-146 67. Zhang B, Gao J, Gao Y, Cai W (2018a) Land use transition of mountainous rural areas in China. Acta
Geogr Sin 73(3):503-517 67. Zhang B, Gao J, Gao Y, Cai W (2018a) Land use transition of mountainous rural areas in China. Acta
Geogr Sin 73(3):503-517 68. Zhang L, Yao Z, Tang S, Li X, Hao T (2017) Spatiotemporal characteristics and patterns of the global
cultivated land since the 1980s. Acta Geogr Sin 72(7):1235-1247 68. Zhang L, Yao Z, Tang S, Li X, Hao T (2017) Spatiotemporal characteristics and patterns of the global
cultivated land since the 1980s. Acta Geogr Sin 72(7):1235-1247 69. Zhang Q, Yang F, Li F (2021) The grain production space reconstruction in China since the reform and
opening up. J Nat Resour 36(6)：1426-1438 69. Zhang Q, Yang F, Li F (2021) The grain production space reconstruction in China since the reform and
opening up. J Nat Resour 36(6)：1426-1438 70. Zhang Y, Lou H, Ge D, Tu S, Qu Y (2018b) Spatio-temporal characteristics and dynamic mechanism of
farmland functions evolution in the Huang-Huai-Hai Plain. Acta Geogr Sin 73(3):518-534 70. Zhang Y, Lou H, Ge D, Tu S, Qu Y (2018b) Spatio-temporal characteristics and dynamic mechanism of
farmland functions evolution in the Huang-Huai-Hai Plain. Acta Geogr Sin 73(3):518-534 Figures Page 20/26 Figure 1 The systematic analysis framework of grain production space reconstruction The systematic analysis framework of grain production space reconstruction Page 21/26 Figure 2 Overview of the study area Page 22/26 Page 22/26 The quantity reconstruction of grain production space in the Loess Plateau The quantity reconstruction of grain production space in the Loess Plateau The quantity reconstruction of grain production space in the Loess Plateau Figure 4
Spatial distribution of grain production space quality in the Loess Plateau from 1980 to 2018 Figure 4 Spatial distribution of grain production space quality in the Loess Plateau from 1980 to 2018 Spatial distribution of grain production space quality in the Loess Plateau from 1980 to 2018 Page 23/26 Figure 5 Figure 5 Figure 5 The landscape pattern reconstruction of grain production space in the Loess Plateau The landscape pattern reconstruction of grain production space in the Loess Plateau Page 24/26 Figure 6
The quantity (a) and quality (b) migration trajectories of grain production space in the Loess Plateau from
1980 to 2018 Figure 6 The quantity (a) and quality (b) migration trajectories of grain production space in the Loess Plateau from
1980 to 2018 Page 25/26 Figure 7
Grain production space reconstruction driving mechanism Page 26/26
Figure 7
Grain production space reconstruction driving mechanism Figure 7 Grain production space reconstruction driving mechanism Grain production space reconstruction driving mechanism Page 26/26
|
https://openalex.org/W1982467318
|
https://zenodo.org/records/1259895/files/article.pdf
|
English
| null |
O-Crystallin, arginine kinase and ferritin from the octopus lens
|
Biochimica et biophysica acta. Proteins and proteomics
| 1,999
|
public-domain
| 3,767
|
Short sequence-paper
O-Crystallin, arginine kinase and ferritin from the octopus lens1
Rina D. Zinovieva 2, Joram Piatigorsky, Stanislav I. Tomarev *
Laboratory of Molecular and Developmental Biology, National Eye Institute, National Institutes of Health, NIH, Bldg. 6, Room 2A04
6 Center Dr. MSC 2730, Bethesda, MD 20892-2730, USA Short sequence-paper
O-Crystallin, arginine kinase and ferritin from the octopus lens1
Rina D. Zinovieva 2, Joram Piatigorsky, Stanislav I. Tomarev *
Laboratory of Molecular and Developmental Biology, National Eye Institute, National Institutes of Health, NIH, Bldg. 6, Room 2A04,
6 Center Dr. MSC 2730, Bethesda, MD 20892-2730, USA Received 12 January 1999; received in revised form 5 March 1999; accepted 9 March 1999 Abbreviations: PEBP, phosphatidylethanolamine-binding pro-
tein
* Corresponding author. Fax: (301) 4968760;
E-mail: tomarev@helix.nih.gov
1 The nucleotide sequence data reported in this paper have
been submitted to GenBank under accession Nos. AF117271
and AF117272.
2 Present address: N.K. Koltzov Institute of Developmental
Biology, Russian Academy of Sciences, 26 Vavilov St., 117808,
Moscow, Russia. Abstract Three proteins have been identified in the eye lens of the octopus, Octopus dofleini. A 22 kDa protein comprising 3^5% of
the soluble protein of the lens is 35^43% identical to a family of phosphatidylethanolamine-binding proteins of vertebrates. Other members of this family include the immunodominant antigen of the filarial parasite, Onchocerca volvulus, putative
odorant-binding proteins of Drosophila and a protein with unknown function of Caenorhabditis elegans. We have called this
protein O-crystallin on the basis of its abundance in the transparent lens. O-Crystallin mRNA was detected only in the lens. Two tryptic peptides of another octopus lens protein, less abundant than O-crystallin, showed 80% identity to arginine kinase
of invertebrates, a relative of creatine kinase of vertebrates. Finally, ferritin cDNA was isolated as an abundant cDNA from
the octopus lens library. Northern blots showed that ferritin mRNA is not lens-specific. ß 1999 Elsevier Science B.V. All
rights reserved. Keywords: Crystallin; Octopus; Lens; Ferritin; Phosphatidylethanolamine-binding protein The eyes of cephalopod molluscs (squid and octo-
pus) and vertebrates have striking similarities in
overall appearance [1]. One common structure of in-
terest to our laboratory is their cellular lens. Both
cephalopods and vertebrates have used a similar
strategy of recruiting pre-existing proteins with ubiq-
uitous functions as lens structural proteins, called crystallins, that are responsible for the refractive
properties of this transparent tissue [2^4]. Vertebrate
crystallins represent a surprisingly diverse group of
soluble proteins often showing little structural simi-
larity to each other. Some (K- and L/Q-crystallins) are
present in all vertebrate species while others (the tax-
on-speci¢c crystallins) are con¢ned to selected species
or phylogenetic groups [2,5]. In general, the taxon-
speci¢c crystallins are identical or related to metabol-
ic enzymes [6,7]. Cephalopods, like vertebrates, have
also utilized enzymes as lens crystallins. For example,
S-crystallins are related to glutathione S-transferase
(GST) and may account for 70^90% of the soluble
protein of the squid or octopus eye lens [8^11]. 6-
Crystallin
is
related
to
aldehyde
dehydrogenase
(ALDH) and is much more prevalent in the lenses
of octopus than squid, suggestive of an incomplete
taxon speci¢city [10,12]. 6-Crystallin represents up 1 The nucleotide sequence data reported in this paper have
been submitted to GenBank under accession Nos. AF117271
and AF117272. 2 Present address: N.K. Koltzov Institute of Developmental
Biology, Russian Academy of Sciences, 26 Vavilov St., 117808,
Moscow, Russia. 0167-4838 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. Abbreviations: PEBP, phosphatidylethanolamine-binding pro-
tein 2 Present address: N.K. Koltzov Institute of Developmental
Biology, Russian Academy of Sciences, 26 Vavilov St., 117808,
Moscow, Russia. 1 The nucleotide sequence data reported in this paper have
been submitted to GenBank under accession Nos. AF117271
and AF117272. Biochimica et Biophysica Acta 1431 (1999) 512^517 Abstract All rights reserved. PII: S 0 1 6 7 - 4 8 3 8 ( 9 9 ) 0 0 0 6 6 - 7 R.D. Zinovieva et al. / Biochimica et Biophysica Acta 1431 (1999) 512^517 513 Fig. 1. Coomassie blue stained SDS-14% polyacrylamide gel of
water-soluble octopus lens proteins. Approx. 5 Wg of proteins
were used. Identi¢ed proteins are marked on the right. Position
of markers is shown on the left. ti¢ed earlier [14]. After the S- and 6-crystallins, the
third most abundant lens protein on the gel in Fig. 1
has an apparent molecular weight of approx. 22
kDa. Initial sequencing of two tryptic peptides
from this gel-puri¢ed protein band indicated that it
is related to phosphatidylethanolamine-binding pro-
tein (PEBP) isolated from the mammalian brain [15]. This polypeptide represents 3^5% of the total soluble
protein of the octopus lens and therefore is not as
abundant as S- or 6-crystallin. It is not established
exactly how much of a given protein is necessary to
be considered a crystallin. Among the L/Q-crystallins
in vertebrates, or even in the case of individual S-
crystallin polypeptides in the octopus lens (see Fig. 1), some crystallin members represent only a few
percent of the total protein. Thus, since the 22 kDa
protein is one of the most prominent proteins in the
octopus lens we have assumed that it contributes
signi¢cantly to the refractive index of the cytoplasm
and called it O-crystallin. Fig. 1. Coomassie blue stained SDS-14% polyacrylamide gel of
water-soluble octopus lens proteins. Approx. 5 Wg of proteins
were used. Identi¢ed proteins are marked on the right. Position
of markers is shown on the left. To clone the cDNA encoding O-crystallin, we used
a PCR approach. A 5P degenerate primer (5P-CAG/
ACAC/TTGGC/TTG/TGTC/TGTC/TAAC/TATC/
TCCC/A/TGG-3P) designed from one of the tryptic
peptides (NT55, positions 79^88 in the complete ami-
no acid sequence) and a 3P oligo(dT) primer were
used to generate a PCR product with a length of
about 600 bp. Sequencing of a cloned PCR fragment
con¢rmed that it encoded part of O-crystallin since
sequences encoding other tryptic peptides in the 22 to approx. 10% of the soluble protein of the octopus
lens [13]. Since only S-crystallins and 6-crystallin of
the octopus lens are known, we investigated the oth-
er prominent proteins expressed in the lens. A SDS-polyacrylamide gel showing the GST/S-
crystallins and ALDH/6-crystallin is presented in
Fig. 1. Fig. 2. Comparison of amino acid sequence of O-crystallin with amino acid sequences of the C. elegans protein (accession No.
AF016423), immunodominant antigen of O. volvulus [18], Drosophila melanogaster putative odorant-binding protein [17] and bovine
PEBP [15]. The sequence of O-crystallin is shown in full; for other sequences only di¡ering amino acid residues are shown. All but
O-crystallin sequences have extra N-terminal sequences. + marks the end of the O. volvulus sequence. 3 indicates gaps that were in-
troduced to maximize similarity. Regions corresponding to putative ligand-binding sites in the human PEBP [25] are boxed. * marks
residues de¢ning the ligand-binding site in the model of the bovine PEBP [26]. Abstract The 600 bp
PCR fragment was used for screening the octopus
lens cDNA library [10]. Three positive plaques
(among about 2000) were identi¢ed and sequenced. The longest cDNA clone was 802 nt excluding the
poly(A) tail. It encoded a protein with a deduced
molecular mass of 21.3 kDa. Comparison of the de-
duced amino acid sequence of O-crystallin with the
GenBank database revealed a 35^43% identity to a
family of PEBP-related proteins [15,16]. This family
included a Caenorhabditis elegans protein with an
unknown
function
(GenBank
accession
No. AF016423),
Drosophila
putative
odorant-binding
proteins [17] and an immunodominant antigen of
the ¢larial parasite, Onchocerca volvulus [18], among
other members (Fig. 2). All these proteins show
about 25^30% identity with yeast TSF1 protein, a
suppressor of cdc25 mutations in Saccharomyces
cerevisiae; TSF1 is implicated in ras/cdc25 signal
transduction [19]. Fig. 3. Tissue distribution of mRNAs for octopus ferritin and
O-crystallin. 2 or 20 Wg of total RNA were used as indicated
above each lane. Positions of the marker RNAs are shown on
the left. The tissue distribution of mRNAs encoding octo-
pus O-crystallin was investigated by Northern blot
hybridization. O-Crystallin cDNA hybridized exclu-
sively to lens RNA approx. 1 kb in length (Fig. 3). No hybridization was detected with RNA samples
from the other tissues. In this respect O-crystallin is
similar to 6-crystallin which is lens-speci¢c [12] and
S-crystallin which is lens- and cornea-speci¢c [11,20]. Vertebrate PEBP mRNA was detected in poly-
(A)RNA derived from many tissues analyzed by
Northern blot hybridization, with the highest abun-
dance being in mouse and rat testis [21]. Since we
used total RNA rather than poly(A)RNA in our
experiments, we cannot exclude the possibility that
O-crystallin is expressed in other tissues at a low
level. Although O-crystallin is relatively abundant
in the octopus lens, it appears to be absent in the
squid lens as judged by SDS-polyacrylamide gel elec-
trophoresis and Northern blot hybridization (not
shown). Therefore, O-crystallin may be considered
a taxon-speci¢c crystallin in cephalopods. and/or nutrition of lens cells [22,23]. It has been sug-
gested that proteins belonging to the PEBP family
are involved in multiple cellular processes including
lipid, opioid and odorant binding and host-parasite
interaction [24]. Three-dimensional structures of hu-
man and bovine PEBP were resolved recently [25,26]. Several residues and regions in the PEBP molecule
have been implicated in ligand binding (Fig. 2) and
are conserved in O-crystallin. Abstract A number of less prominent bands of protein
are also evident. One is L-actin which has been iden- ig. 2. Comparison of amino acid sequence of O-crystallin with amino acid sequences of the C. elegans protein (accession N
AF016423), immunodominant antigen of O. volvulus [18], Drosophila melanogaster putative odorant-binding protein [17] and bovi
EBP [15]. The sequence of O-crystallin is shown in full; for other sequences only di¡ering amino acid residues are shown. All b
O-crystallin sequences have extra N-terminal sequences. + marks the end of the O. volvulus sequence. 3 indicates gaps that were
roduced to maximize similarity Regions corresponding to putative ligand-binding sites in the human PEBP [25] are boxed * mar Fig. 2. Comparison of amino acid sequence of O-crystallin with amino acid sequences of the C. elegans protein (accession No. AF016423), immunodominant antigen of O. volvulus [18], Drosophila melanogaster putative odorant-binding protein [17] and bovine
PEBP [15]. The sequence of O-crystallin is shown in full; for other sequences only di¡ering amino acid residues are shown. All but
O-crystallin sequences have extra N-terminal sequences. + marks the end of the O. volvulus sequence. 3 indicates gaps that were in-
troduced to maximize similarity. Regions corresponding to putative ligand-binding sites in the human PEBP [25] are boxed. * marks
residues de¢ning the ligand-binding site in the model of the bovine PEBP [26]. Fig. 2. Comparison of amino acid sequence of O-crystallin with amino acid sequences of the C. elegans protein (accession No. AF016423), immunodominant antigen of O. volvulus [18], Drosophila melanogaster putative odorant-binding protein [17] and bovine
PEBP [15]. The sequence of O-crystallin is shown in full; for other sequences only di¡ering amino acid residues are shown. All but
O-crystallin sequences have extra N-terminal sequences. + marks the end of the O. volvulus sequence. 3 indicates gaps that were in-
troduced to maximize similarity. Regions corresponding to putative ligand-binding sites in the human PEBP [25] are boxed. * marks
residues de¢ning the ligand-binding site in the model of the bovine PEBP [26]. R.D. Zinovieva et al. / Biochimica et Biophysica Acta 1431 (1999) 512^517 514 Fig. 3. Tissue distribution of mRNAs for octopus ferritin and
O-crystallin. 2 or 20 Wg of total RNA were used as indicated
above each lane. Positions of the marker RNAs are shown on
the left. kDa protein were identi¢ed (not shown). Abstract It has been proposed
that PEBP may be a membrane-binding protein that
interacts with other proteins and modulates catalytic
activity of enzymes [26]. It has also been suggested
that PEBP might be involved in membrane signal
transduction [25]. Virtually nothing is known about
the functions of octopus O-crystallin. We propose in
view of its relative abundance in the octopus lens
that it has a role in refraction in this species. This,
of course, does not exclude its having another, non-
crystallin role in the octopus lens. We also sequenced two tryptic peptides obtained
from another, less abundant protein with an appar-
ent molecular mass of about 40 kDa from the octo-
pus lens (Fig. 1). This protein was easily identi¢able
in the spectrum of the octopus water-soluble proteins
(see Fig. 1). Analysis of the peptide sequences indi-
cated that this 40 kDa polypeptide is related to argi- It is noteworthy that vertebrate lenses also contain
a signi¢cant amount of lipid-binding protein with a
molecular mass of about 15 kDa belonging to the
family of lipid/retinoic acid-binding proteins [22,23]. This vertebrate lipid-binding protein, called LP2, is
structurally di¡erent from O-crystallin and was pro-
posed to be involved in di¡erentiation, protection 515 R.D. Zinovieva et al. / Biochimica et Biophysica Acta 1431 (1999) 512^517 i
f
i
id
f
f
i i
i h f
i i
f
h
f
b
h
h i Fig. 4. Comparison of amino acid sequences of octopus ferritin with ferritins from the frog Rana catesbiana [37], human H-chain [34],
the gastropod mollusc L. stagnalis [31], S. mansoni [32], and soybean [33]. * mark the residues implicated in metal binding [35]. Fig. 4. Comparison of amino acid sequences of octopus ferritin with ferritins from the frog Rana catesbiana [37], human H-chain [34],
the gastropod mollusc L. stagnalis [31], S. mansoni [32], and soybean [33]. * mark the residues implicated in metal binding [35]. nine kinase of several invertebrates [27,28] and
ATP:guanidino kinase of Schistosoma mansoni [29]. These proteins are related to creatine kinase of ver-
tebrates. 80% identity in amino acid sequence was
observed with arginine kinase isolated from muscles
of the gastropod, Nordotis madaka [28]. Arginine and
creatine kinases belong to a conserved family of
ATP:guanidino phosphotransferases, whose mem-
bers play an important role in energy metabolism. Arginine kinase is widely distributed among inverte-
brates. In vertebrates, only creatine kinase activities
have been detected. Fig. 4. Comparison of amino acid sequences of octopus ferritin with ferritins from the frog Rana catesbiana [37], human H-chain [34],
the gastropod mollusc L. stagnalis [31], S. mansoni [32], and soybean [33]. * mark the residues implicated in metal binding [35]. References [1] A. Packard, Cephalopods and ¢sh: the limits of conver-
gence, Biol. Rev. 47 (1972) 251^307. [18] E. Lobos, M. Altmann, G. Mengod, N. Weiss, W. Rudin,
M. Karam, Identi¢cation of an Onchocerca volvulus cDNA
encoding a low-molecular-weight antigen uniquely recog-
nized by onchocerciasis patient sera, Mol. Biochem. Para-
sitol. 39 (1990) 135^146. [2] G. Wistow, J. Piatigorsky, Recruitment of enzymes as lens
structural proteins, Science 236 (1987) 1554^1556. [3] S.I. Tomarev, R.D. Zinovieva, Squid major lens polypep-
tides are homologous to glutathione S-transferase subunits,
Nature 336 (1988) 86^88. [19] M.L. Tripp, R.A. Bouchard, R. Pinon, Cloning and charac-
terization of NSP1, a locus encoding a component of a
cdc25-dependent, nutrient-responsive pathway in Saccharo-
myces cerevisiae, Mol. Microbiol. 3 (1989) 1319^1327. [4] S.I. Tomarev, J. Piatigorsky, Lens crystallins of inverte-
brates. Diversity and recruitment from detoxi¢cation en-
zymes and novel proteins, Eur. J. Biochem. 235 (1996)
449^465. [20] R.A. Cuthbertson, S.I. Tomarev, J. Piatigotsky, Taxon-spe-
ci¢c recruitment of enzymes as major soluble proteins in the
corneal epithelium of three mammals, chicken and squid,
Proc. Natl. Acad. Sci. USA 89 (1992) 4004^4008. [5] G.J. Wistow, J. Piatigorsky, Lens crystallins: the evolution
and expression of proteins for a highly specialized tissue,
Annu. Rev. Biochem. 57 (1988) 479^504. [21] N. Seddiqi, F. Bollengier, P.M. Alliel, J.-P. Perin, F. Bonnet,
S. Bucquoy, P. Jolles, F. Schoentgen, Amino acid sequence
of the Homo sapiens brain 21^23 kDa protein (neuropoly-
peptide h3), comparison with its counterparts from Rattus
norvegicus and Bos taurus species, and expression of its
mRNA in di¡erent tissues, J. Mol. Evol. 39 (1994) 655^660. [6] G. Wistow, Lens crystallins: gene recruitment and evolution-
ary dynamism, Trends Biochem. Sci. 18 (1993) 301^306. [7] J. Piatigorsky, Gene sharing in lens and cornea: facts and
implications, Prog. Retinal Eye Res. 17 (1998) 145^174. [8] R.J. Siezen, D.C. Shaw, Physicochemical characterization of
lens proteins of the squid Nototodarus gouldi and compari-
son with vertebrate crystallins, Biochim. Biophys. Acta 704
(1982) 304^320. [22] Y. Wen, G.W. Li, P. Chen, E. Wong, I. Bekhor, Lens epi-
thelial cell mRNA. II. Expression of a mRNA encoding a
lipid-binding protein in rat lens epithelial cells, Gene 158
(1995) 269^274. [9] S.I. Tomarev, S. Chung, J. Piatigorsky, Glutathione S-trans-
ferase and S-crystallins of cephalopods: evolution from ac-
tive enzyme to lens-refractive proteins, J. Mol. Evol. 41
(1995) 1048^1056. [23] C. Jaworski, G. References Wistow, LP2, a di¡erentiation-associated
lipid-binding protein expressed in bovine lens, Biochem. J. 320 (1996) 49^54. [10] S.I. Tomarev, R.D. Zinovieva, J. Piatigorsky, Crystallins of
the octopus lens. Recruitment from detoxi¢cation enzymes,
J. Biol. Chem. 266 (1991) 24226^24231. [24] F. Schoentgen, P. Jolles, From structure to function: possi-
ble biological roles of a new widespread protein family bind-
ing hydrophobic ligands and displaying a nucleotide binding
site, FEBS Lett. 369 (1995) 22^26. [11] S.I. Tomarev, R.D. Zinovieva, J. Piatigorsky, Characteriza-
tion of squid crystallin genes. Comparison with mammalian
glutathione S-transferase genes, J. Biol. Chem. 267 (1992)
8604^8612. [25] M.J. Ban¢eld, J.J. Barker, A.C. Perry, R.L. Brady, Function
from structure? The crystal structure of human phospha-
tidylethanolamine-binding protein suggests a role in mem-
brane signal transduction, Structure 6 (1998) 1245^1254. [12] R.D. Zinovieva, S.I. Tomarev, J. Piatigorsky, Aldehyde de-
hydrogenase-derived 6-crystallins of squid and octopus. Specialization
for lens expression, J. Biol. Chem. 268
(1993) 11449^11455. [26] L. Serre, B. Vallee, N. Bureaud, F. Schoentgen, C. Zelwer,
Crystal structure of the phosphatidylethanolamine-binding
protein from bovine brain: a novel structural class of phos-
pholipid-binding proteins, Structure 6 (1998) 1255^1265. [13] S.-H. Chiou, A novel crystallin from octopus lens, FEBS
Lett. 241 (1988) 261^264. [27] C. Dumas, J. Camonis, Cloning and sequence analysis of the
cDNA for arginine kinase of lobster muscle, J. Biol. Chem. 268 (1993) 21599^21605. [14] S.I. Tomarev, R.D. Zinovieva, J. Piatigorsky, Primary struc-
ture and lens-speci¢c expression of genes for an intermediate
¢lament protein and a beta-tubulin in cephalopods, Biochim. Biophys. Acta 1216 (1993) 245^254. [28] T. Suzuki, T. Furukohri, Evolution of phosphagen kinase. Primary structure of glycocyamine kinase and arginine kin-
ase from invertebrates, J. Mol. Biol. 237 (1994) 353^357. [15] F. Schoentgen, F. Saccoccio, J. Jolles, I. Bernier, P. Jolles,
Complete amino acid sequence of a basic 21-kDa protein
from bovine brain cytosol, Eur. J. Biochem. 166 (1987)
333^338. [29] L.D. Stein, D.A. Harn, J.R. David, A cloned ATP:guanidi-
no kinase in the trematode Schistosoma mansoni has a novel
duplicated structure, J. Biol. Chem. 265 (1990) 6582^6588. [16] D.K. Grandy, E. Hanneman, J. Bunzow, M. Shih, C.A. Machida, J.M. Bidlack, O. Civelli, Puri¢cation, cloning,
and tissue distribution of a 23 kDa rat protein isolated by
morphine
a¤nity
chromatography,
Mol. Endocrinol. 4
(1990) 1370^1376. [30] D.L. Friedman, J.F. Hejtmancik, J.N. Hope, M.B. Perry-
man, Developmental expression of creatine kinase isozymes
in mammalian lens, Exp. Eye Res. 49 (1989) 445^457. Abstract In humans, there is a marked
increase in the expression of the BB isoform of crea-
tine kinase in the epithelial cells of the lens with the
onset of sexual maturation [30]. It was proposed that
the localization of creatine kinase in the epithelial
cells
re£ects
its
involvement
in
high
transport
ATPase activity [30]. In the octopus lens, the arginine
kinase-related protein might also be involved in in-
tracellular transport. [12]. We also identi¢ed a group of clones that en-
coded ferritin. The longest cDNA insert was 1103
nt long without counting the poly(A) tail. Ferritin
cDNA was present in all tissues analyzed by North-
ern blot hybridization, giving a single band with a
size around 1.2 kb (Fig. 3). Octopus ferritin showed
67% identity to ferritin isolated from the gastropod
mollusc Lymnaea stagnalis [31] and 55^60% identity
to ferritins of S. mansoni [32], soybean [33] and H-
chain ferritins from vertebrates [34] (Fig. 4). All of
the residues implicated in metal binding of human
heavy chain ferritin are conserved in the octopus
ferritin. Ferritin is an iron storage protein that is
essential for protection of cells against oxidative
stress. In vertebrates, cytoplasmic ferritin is com-
posed of a heavy and light subunits [35] and plays
an important role in normal lens function. Recent
data demonstrated that a mutation in the iron-re-
sponsive element of the human L-ferritin leads to
early bilateral cataract [36]. It is not clear at present
whether octopus ferritin has any special role in the
lens besides storage of iron atoms. Finally, to identify other cDNA clones corre-
sponding to abundant mRNAs, we screened the oc-
topus lens cDNA library with total cDNAs synthe-
sized using lens poly(A)RNA as templates and
characterized the clones giving the strongest signals. Most of the clones encoded octopus S-crystallins We thank Todd Johnson for help with character-
ization of the O-crystallin clone. This work was sup-
ported in part by Russian Foundation for Basic Re-
search Grant 96-04-48447 (to R.D.Z.). R.D. Zinovieva et al. / Biochimica et Biophysica Acta 1431 (1999) 512^517 516 are expressed in di¡erent subsets of olfactory hairs, Neuron
12 (1994) 35^49. References [31] M. Von Darl, P.M. Harrison, W. Bottke, cDNA cloning and
deduced amino acid sequence of two ferritins: soma ferritin
and yolk ferritin, from the snail Lymnaea stagnalis, Eur. J. Biochem. 222 (1994) 353^366. [17] C.W. Pikielny, G. Hasan, F. Rouyer, M. Rosbash, Members
of a family of Drosophila putative odorant-binding proteins R.D. Zinovieva et al. / Biochimica et Biophysica Acta 1431 (1999) 512^517 517 iron storage function and cellular regulation, Biochim. Bio-
phys. Acta 1275 (1996) 161^203. [32] J. Dietzel, J. Hirzmann, D. Preis, P. Symmons, W. Kunz,
Ferritins of Schistosoma mansoni: sequence comparison and
expression in female and male worms, Mol. Biochem. Para-
sitol. 50 (1992) 245^254. [36] C. Beaumont, P. Leneuve, I. Devaux, J.Y. Scoazec, M. Berthier, M.N. Loiseau, B. Grandchamp, D. Bonneau, Mu-
tation in the iron responsive element of the L ferritin mRNA
in a family with dominant hyperferritinaemia and cataract,
Nat. Genet. 11 (1995) 444^446. [33] A.-M. Lescure, D. Proudhoni, H. Pesey, M. Ragland, E.C. Theil, J.-F. Briat, Ferritin gene transcription is regulated by
iron in soybean cell cultures, Proc. Natl. Acad. Sci. USA 88
(1991) 8222^8826. [37] J.R. Didsbury, E.C. Theil, R.E. Kaufman, L.F. Dickey,
Multiple red cell ferritin mRNAs, which code for an abun-
dant protein in the embryonic cell type, analyzed by cDNA
sequence and by primer extension of the 5P-untranslated re-
gions, J. Biol. Chem. 261 (1986) 949^955. [34] D. Boyd, C. Vecoli, D.M. Belcher, S.K. Jain, J.W. Drysdale,
Structural and functional relationships of human ferritin H
and L chains deduced from cDNA clones, J. Biol. Chem. 260
(1985) 11755^11761. [35] P.M. Harrison, P. Arosio, Ferritins: molecular properties,
|
https://openalex.org/W3194249621
|
https://ojs.batangkab.go.id/index.php/ristek/article/download/59/170
|
Indonesian
| null |
MENGEMBANGKAN ENTREPRENEURSHIP MELALUI KECAKAPAN HIDUP BERBASIS POTENSI LOKAL SEBAGAI MODEL PENGUATAN PENDIDIKAN KARAKTER DI SEKOLAH ADIWIYATA
|
Ristek
| 2,019
|
cc-by
| 6,187
|
ABSTRAK Latar belakang dilakukan penelitian ini adalah besarnya angka pengangguran di
Indonesia dipengaruhi oleh rendahnya entrepreneurship populasi penduduk yang muda. Membangun SDM yang berpotensi, berkualitas menyangkut relevansinya, dan kebutuhan
pasar di dunia kerja dan usaha masih menjadi persoalan bagi bangsa ini. SDM kita belum
memiliki kecakapan hidup yang baik. Secara otomatis hal tersebut juga akan berimbas pada
persaingan yang semakin ketat di pasar dunia kerja dan usaha. Padahal kecakapan hidup
generasi bangsa sangat ditentukan oleh pendidikan karakter yang mampu mengembangkan
nilai-nilai etika inti seperti kepedulian, kejujuran, keadilan, tanggung jawab, dan rasa hormat
terhadap diri dan orang lain bersama dengan nilai-nilai kinerja pendukungnya seperti
ketekunan, etos kerja yang tinggi, dan kegigihan--sebagai basis karakter yang baik. Nilai-nilai
karakter itu merupakan pondasi yang kuat bagi pembentukan kecakapan hidup yang menjadi
syarat berkembangnya entrepreneurship tenaga kerja muda Indonesia. Penelitian ini bertujuan untuk mengetahui bagaimana dan seberapa besar pengembangan
entrepreneurship melalui kecakapan hidup berbasis potensi lokal sebagai model dalam penguatan
pendidikan karakter di sekolah adiwiyata. Metode penelitian ini menggunakan penelitian deskriptif dengan pendekatan kualitatif. Data dikumpulkan dengan menggunakan metode survey dengan langkah-langkah meliputi : 1)
merumuskan masalah penelitian dan menentukan tujuan survei; 2) menentukan konsep dan
hipotesa serta menggali kepustakaan; 3) pengambilan sampel; 4) pembuatan instrumen
penelitian; 5) pekerjaan lapangan; 6) pengolahan data; 7) analisa dan pelaporan. Langkah-
langkah
strategis
tersebutdipersiapkan
untuk
mengurangi
pengangguran
dengan
mengembangkan entrepreneurship didasarkan pada nilai-nilai kecakapan hidup atau
kecakapan hidup berbasis potensi lokal serta mengacu kepada penguatan pendidikan karakter
bangsa dalam berbagai kegiatan di sekolah seperti pada pembelajaran di kelas, kegiatan
ekstrakurikuler,
maupun
dalam
kegiatan-kegiatan
lainnya
diharapkan
dapat
menumbuhkembangkan jiwa pantang menyerah, etos kerja yang tinggi, mandiri, kreatif,
inovatif dalam menciptakan karya-karya di segala bidang, sehingga pada akhirnya mampu
menciptakan sumber daya manusia yang mumpuni bagi negara ini. Adapun berbasis potensi
lokal dimaksudkan agar kelak generasi muda dapat menjadi agen of change bagi
pembangunan di daerahnya sendiri. Pengembangan unsur-unsur utama yang ada dalam entrepreneurship, yaitu: penerapan
kreativitas dan inovasi, pemanfaatan peluang, membuat perubahan, dan memberikan nilai
tambah bagi diri sendiri dan orang lain dilaksanakan dalam bingkai sekolah adiwiyata dengan
mempertimbangkan hubungan erat antara unsur-unsur yang ada pada kecakapan hidup dan
pendidikan karakter. MENGEMBANGKAN ENTREPRENEURSHIP MELALUI KECAKAPAN HIDUP
BERBASIS POTENSI LOKAL SEBAGAI MODEL PENGUATAN PENDIDIKAN
KARAKTER DI SEKOLAH ADIWIYATA Karso Mulyo 1 Alamat koresponden:SMP N 3 Tersono, Jl.Sidalang, Tersono,
Batang, Jawa Tengah, Indonesia, 51272
E-mail: karsobatang@gmail.com
Situs: http://pena-batang.blogspot.com/ Alamat koresponden:SMP N 3 Tersono, Jl.Sidalang, Tersono,
Batang, Jawa Tengah, Indonesia, 51272
E-mail: karsobatang@gmail.com
Situs: http://pena-batang.blogspot.com/ ABSTRACT The background of this research is that the size of the unemployment rate in Indonesia is
influenced by the low entrepreneurship of the young population. Building potential human resources,
quality concerning their relevance, and market needs in the world of work and business is still a
problem for this nation. Our human resources do not have good life skills. This will automatically
affect the increasingly fierce competition in the world of work and business markets. Even though the
generation of life skills is largely determined by character education that is able to develop core
ethical values such as caring, honesty, fairness, responsibility, and respect for oneself and others
along with supporting performance values such as perseverance, high work ethic and perseverance -
as a good character base. These character values are a strong foundation for the formation of life
skills which is a condition for the development of young Indonesian entrepreneurship workforce. This study aims to find out how and how much the development of entrepreneurship throug
ills based on local potential as a model in strengthening character education in adiwiyata schoo This research method uses descriptive research with a qualitative approach. Data is collected
using survey methods with steps including: 1) formulating research problems and determining survey
objectives; 2) determine concepts and hypotheses and explore literature; 3) sampling; 4) making
research instruments; 5) field work; 6) data processing; 7) analysis and reporting. These strategic
steps are prepared to reduce unemployment by developing entrepreneurship based on the values of life
skills or life skills based on local potential as well as referring to strengthening national character
education in various activities in schools such as in classroom learning, extracurricular activities, and
in activities others are expected to be able to develop an unyielding spirit, high work ethic,
independent, creative, innovative in creating works in all fields, so that in the end they are able to
create qualified human resources for this country. The local potential based is intended so that later
the younger generation can become agents of change for development in their own regions. The development of the main elements that exist in entrepreneurship, namely: the application of
creativity and innovation, the use of opportunities, making changes, and providing added value for
oneself and others is carried out in the frame of the Adiwiyata school by considering the close
relationship between the elements in the skill life and character education. Karso Mulyo 1 Karso Mulyo 1 Correspondent addres:SMP N 3 Tersono,Sidalang Street, Tersono,
Batang,Central Java , Indonesia, 51272
E-mail: karsobatang@gmail.com
Situs: http://pena-batang.blogspot.com/ Correspondent addres:SMP N 3 Tersono,Sidalang Street, Tersono,
Batang,Central Java , Indonesia, 51272
E-mail: karsobatang@gmail.com
Situs: http://pena-batang.blogspot.com/ ABSTRAK Nilai-nilai kecakapan dasar dan instrumental dalam kecakapan hidup dan
nilai-nilai religius, nasionalis, mandiri, gotong-royong dan integritas dalam pendidikan
karakter dibangun untuk mengembangkan entrepreneurship dengan melaksanakan kegiatan-
kegiatan berbasis kelas, budaya sekolah dan masyarakat seperti (i) Membuat taman, membuat
tempat pemilahan sampah dari bahan bekas, membuat pot bunga dan tanaman dari bahan- RISTEK: Jurnal Riset, Inovasi dan Teknologi Page 81 bahan bekas, (ii) Olah hasil kebun untuk komoditi dan pemasarannya: Minuman Jahe, bibit
tanaman, jajan tradisional, (iii) Gerakan go green dan go clean, gerakan membaca (literasi),
berpacu dalam prestasi, ekstrakurikuler, pemberian penghargaan, publikasi facebook,
blogspot, lewat koran dan media cetak lainnya, (iv) Pembenihan, pembibitan,
pengembangbiakan vegetatif (cangkok, tempel, sisip, stek), produksi pupuk bokhasi, pupuk
organik cair dan Minuman Jahe Merah. Kata kunci : Entrepreneurship, Kecakapan Hidup, Penguatan Pendidikan Karakter. : Entrepreneurship, Kecakapan Hidup, Penguatan Pendidikan Karakter. ABSTRACT The values of basic and
instrumental skills in life skills and religious values, nationalism, independence, mutual cooperation
and integrity in character education are built to develop entrepreneurship by carrying out classroom-
based activities, school culture and society such as (i) Making a garden , making waste sorting sites
from used materials, making flower pots and plants from used materials, (ii) Oat crops for
commodities and marketing them: Ginger drinks, plant seeds, traditional snacks, (iii) Go green and go
clean movements. reading movement (literacy), race in achievement, extracurricular, awarding, RISTEK: Jurnal Riset, Inovasi dan Teknologi Page 82 publication of facebook, blogspot, through newspapers and other print media, (iv) Hatchery,
nurseries, vegetative breeding (grafting, paste, cutting, cuttings), bokhasi fertilizer production, liquid
organic fertilizer and Red Ginger Drink. Keywords: Entrepreneurship, Life Skills, Strengthening Character Education. PENDAHULUAN peserta didik, keadaan sekolah, potensi dan
kebutuhan daerah. Pengenalan keadaan
lingkungan, sosial, dan budaya kepada
peserta didik memungkinkan mereka untuk
lebih akrab dengan lingkungan kehidupan
peserta
didik. Pengenalan
dan
pengembangan
lingkungan
melalui
pendidikan kecakapan hidup diarahkan
untuk menunjang peningkatan kualitas
sumber daya manusia, yang pada akhirnya
dapat meningkatkan kompetensi peserta
didik. Besarnya angka pengangguran di
Indonesia dipengaruhi oleh rendahnya
entrepreneurship populasi penduduk yang
muda. Membangun SDM yang berpotensi,
berkualitas menyangkut relevansinya, dan
kebutuhan pasar di dunia kerja dan usaha
masih menjadi persoalan bagi bangsa ini. SDM kita belum memiliki kecakapan
hidup yang baik. Secara otomatis hal
tersebut
juga
akan
berimbas
pada
persaingan yang semakin ketat di pasar
dunia kerja dan usaha. Padahal kecakapan
hidup generasi bangsa sangat ditentukan
oleh pendidikan karakter yang mampu
mengembangkan
nilai-nilai
etika
inti
seperti kepedulian, kejujuran, keadilan,
tanggung jawab, dan rasa hormat terhadap
diri dan orang lain bersama dengan nilai-
nilai
kinerja
pendukungnya
seperti
ketekunan, etos kerja yang tinggi, dan
kegigihan--sebagai basis karakter yang
baik. Nilai-nilai karakter itu merupakan
pondasi yang kuat bagi pembentukan
kecakapan hidup yang menjadi syarat
berkembangnya entrepreneurship tenaga
kerja muda Indonesia. Bahkan Presiden
Joko Widodo ingin melakukan Gerakan
Nasional
Revolusi
Mental
(GNRM)
melalui salah satu butir Nawacitanya yakni
memperkuat pendidikan karakter bangsa
yang akan diterapkan di seluruh sendi
kehidupan
berbangsa
dan
bernegara,
termasuk di dalam dunia pendidikan. Dalam berbagai kegiatan di sekolah
seperti
pada
pembelajaran
di
kelas,
kegiatan ekstrakurikuler, maupun dalam
kegiatan-kegiatan lainnya pengembangan
nilai-nilai kecakapan hidup diharapkan
dapat menumbuhkembangkan nilai-nilai
karakter seperti jiwa pantang menyerah,
etos kerja yang tinggi, mandiri, kreatif da
ninovatif dalam menciptakan karya-karya
di segala bidang, sehingga pada akhirnya
mampu menciptakan sumber daya manusia
yang
berjiwa
entrepreneurship
yang
tinggi. Suatu negara dapat dikatakan
sebagai
negara
maju
jika
secara
komprehensif
dan
substantif
sistem
pendidikan yang diterapkan berkualitas,
sehingga dapat dilihat dari output yang
dihasilkannya, yakni peserta didik sebagai
generasi muda bangsa yang produktif,
kreatif, inovatif, mandiri, berkepribadian
luhur, atau memiliki kualifikasi secara
kognitif, afektif, maupun psikomotorik
yang bermanfaat bagi kehidupan secara
pribadi maupun universal. Itu artinya
pendidikan menjadi barometer kemajuan
suatu bangsa atau negara tertentu. Pentingnya pendidikan kecakapan
hidup bagi peserta didik mendorong
pemerintah memberikan kewenangan yang
luas
kepada
sekolah
untuk
mengembangkan dan menyelenggarakan
pendidikan
sesuai
dengan
kebutuhan Page 83 RISTEK: Jurnal Riset, Inovasi dan Teknologi Nilai-nilai kemanusiaan, kearifan
lokal bangsa, serta ideologi suatu bangsa
tiap-tiap bangsa atau negara di dunia ini
berbeda-beda
karakternya,
sehingga
pelaksanaan sistem kehidupan berbangsa
dan
bernegara,
termasuk
corak
pendidikannya pun berbeda. PENDAHULUAN Pendidikan
perlu diorganisir secara regulatif dan
konstruktif yang sesuai dengan karakter
yang ada di Indonesia. Oleh sebab itu, di
Indonesia
sendiri
pendidikan
diaktualisasikan
oleh
suatu
satuan
pendidikan tertentu. Anonim (2003:3)
menyebutkan bahwa menurut Undang-
undang Republik Indonesia Nomor 20
tahun 2003 tentang Sistem Pendidikan
Nasional pada pasal 1 ayat 10 menjelaskan
bahwa yang dimaksud dengan satuan
pendidikan
adalah
kelompok
layanan
pendidikan
yang
menyelenggarakan
pendidikan pada jalur formal, nonformal,
dan informal di setiap jenjang maupun
jenis pendidikan. entrepreneurship
melalui
pendidikan
kecakapan hidup agar semakin memiliki
kematangan diri atau pendewasaan diri
menuju kearifan diri dalam menghadapi
dan memecahkan segala problematika
kehidupan
secara
signifikan
dan
komprehensif. Sebagai
rumah
kedua,
sekolah
menjadi
suatu
cerminan
yang
harus
diperhatikan oleh setiap warga sekolah
untuk
berupaya
dalam
mewujudkan
kehidupan sekolah yang menyenangkan,
religius, nasionalis, mandiri, memiliki
semangat gotong royong dan integritas
yang tinggi melalui program-program
sekolah yang memberikan kesempatan
seluas-luasnya kepada peserta didik untuk
meangaktualisasikan dirinya secara bebas,
mandiri, kreatif, dan inovatif dalam wadah
sekolah adiwiyata. Melalui sekolah berbasis adiwiyata
inilah
yang
nantinya
akan
mengintegrasikan antara kedirian peserta
didik dengan realitas kehidupan di suatu
lingkungan yang universal. Oleh karena
itu,
sekolah
berbasis
adiwiyata
diformulasikan
secara
terstruktur,
terencana, konstruktif ke dalam setiap
elemen
system
pendidikan,
seperti
kurikulum sekolah, strategi pembelajaran,
pemberdayaan sarana dan prasarana, dan
evaluasi
pembelajaran
sebagai
suatu
instrumen untuk mengaktualisasikan tujuan
pendidikan nasional dan satuan pendidikan
tertentu. Sekolah
berbasis
adiwiyata
merupakan suatu lingkungan edukasi di
dalam suatu satuan pendidikan tertentu
sebagai tempat yang ideal serta substantif
untuk memperoleh segala macam ilmu
pengetahuan, baik secara skriptualisasi
maupun kontekstualisasi dan nilai-nilai
prilaku yang mulia sebagi landasan yang
fundamental
dalam
mewujudkan
kehidupan yang paripurna sesuai dengan
cita-cita luhur lembaga maupun negara
secara kontinuitas atau berkelanjutan. RISTEK J
l Ri
I
i d
T k
l
Sekolah memiliki karakteristik serta
mekanisme pengaktualisasian nilai-nilai
pendidikan secara substantif-transendentif
berbeda-beda di tiap-tiap daerah. Kondisi
tiap-tiap sekolah dipengaruhi oleh potensi
lokal dan tipologi daerah tersebut, baik
secara geografis, demografis, ekonomis,
kultur, serta keadaan sekolah tersebut. Oleh karena itu, tiap-tiap sekolah memiliki
strategi
atau
model
pengembangan
tersendiri
dalam
menginternalisasikan
nilai-nilai karakter bangsa selaras dengan
tujuan pendidikan nasional, yakni untuk
mengembangkan potensi peserta didik agar
menjadi manusia yang berkepribadian
mulia dalam hal agama, individu, sosial
kemasyarakatan, maupun berbangsa dan
bernegara. Melalui
kecakapan
hidup
berbasis potensi lokal peserta didik dapat
disinggungkan
dengan
segala
macam
bentuk interaksi berbasis kelas, budaya
sekolah dan masyarakat. PENDAHULUAN Peserta didik
diharapkan dapat mengembangkan jiwa Page 84 RISTEK: Jurnal Riset, Inovasi dan Teknologi RISTEK: Jurnal Riset, Inovasi dan Teknologi Berdasarkan
kenyataan
tersebut,
perlu segera dilakukan langkah-langkah
strategis yang secara langsung dapat
mengembangkan
entrepreneurship
didasarkan pada nilai-nilai kecakapan
hidup atau kecakapan hidup berbasis
potensi lokal serta
mengacu kepada
penguatan pendidikan karakter bangsa
dalam berbagai kegiatan di sekolah seperti
pada pembelajaran di kelas, kegiatan
ekstrakurikuler, maupun dalam kegiatan-
kegiatan
lainnya
diharapkan
dapat
menumbuhkembangkan
jiwa
pantang
menyerah, etos kerja yang tinggi, mandiri,
kreatif, inovatif dalam menciptakan karya-
karya di segala bidang, sehingga pada
akhirnya mampu menciptakan sumber
daya manusia yang mumpuni bagi negara
ini. Adapun
berbasis
potensi
lokal
dimaksudkan agar kelak generasi muda
dapat menjadi agen of change bagi
pembangunan di daerahnya sendiri. (2007:1) kreativitas adalah kemampuan
untuk
membuat
ide
baru
dengan
mengkombinasikan,
mengubah,
atau
merekonstruksi ide-ide lama. Sedangkan
Rutherford (2002:297) memandang inovasi
sebagai penerapan dari penemuan suatu
proses produksi baru atau pengenalan akan
suatu
produk
baru. Berbeda
dengan
Sunyoto
(2013:
2)
entrepreneurship
menurutnya adalah suatu sikap untuk
menciptakan sesuatu yang baru serta
bernilai bagi diri sendiri dan orang lain. Definisi berbeda juga diungkap oleh
Marlo(2013:5),
yang
mengartikan
entrepreneurship
sebagai
kemampuan
seseorang untuk peka terhadap peluang dan
memanfaatkan peluang tersebut untuk
melakukan perubahan dari sistem yang
ada. Menurutnya
dalam
dunia
entrepreneurship,
peluang
adalah
kesempatan
untuk
mewujudkan
atau
melaksanakan suatu usaha dengan tetap
memperhitungkan resiko yang dihadapi. Berdasarkan latar belakang di atas,
maka permasalahan dalam laporan ini
dirumuskan sebagai-berikut: Definisi
yang
lebih
ringkas
dikemukakan
oleh
Kasmir(2013:
20)
bahwa
entrepreneurship
merupakan
kemampuan dalam menciptakan sesuatu
yang baru dan berbeda. Pengertian ini
mengandung
maksud
bahwa
seorang
entrepreneur adalah orang yang memiliki
kemampuan untuk menciptakan sesuatu
yang belum pernah ada sebelumnya, atau
bisa juga dengan menciptakan sesuatu
yang berbeda dari yang ada. 1. Bagaimana
pengembangan
entrepreneurship
melalui
kecakapan
hidup berbasis potensi lokal sebagai
model dalam penguatan pendidikan
karakter di sekolah adiwiyata? 2. Seberapa
besar
pencapaian
keberhasilan
pengembangan
entrepreneurship
melalui
kecakapan
hidup berbasis potensi lokal sebagai
model dalam penguatan pendidikan
karakter di sekolah adiwiyata? Dari berbagai definisi di atas, dapat
disimpulkan
bahwa
entrepreneurship
adalah proses penerapan kreatifitas dan
inovasi dalam memanfaatkan peluang
untuk menciptakan perubahan, baik berupa
sesuatu
yang baru ataupun berbeda,
sehingga menghasilkan nilai tambah bagi
diri sendiri dan orang lain. Berangkat dari
definisi ini dapat diperoleh secara rinci
unsur-unsur
utama
yang
ada
dalam
entrepreneurship,
yaitu:
penerapan
kreativitas
dan
inovasi,
pemanfaatan Entrepreneurship Entrepreneurship merupakan suatu
proses penerapan kreativitas dan inovasi
untuk memecahkan dan mencari peluang
dari masalah yang dihadapi oleh setiap
orang
dalam
kehidupan
sehari-hari
(Suryana, 2013: 5). Berdasarkan definisi
ini, inti dari entrepreneurship adalah
kreatifitas dan inovasi. Menurut Friday Page 85 RISTEK: Jurnal Riset, Inovasi dan Teknologi lingkungan;
dan
(10)
kecakapan
menyatukan bangsa. (10) peluang,
membuat
perubahan,
dan
memberikan nilai tambah bagi diri sendiri
dan orang lain. Tujuan pendidikan kecakapan hidup
adalah memfungsikan pendidikan sesuai
dengan fitrahnya, yaitu mengembangkan
potensi manusiawi peserta didik untuk
menghadapi perannya di masa datang. Secara
khusus,
pendidikan
yang
berorientasi kecakapan hidup bertujuan:
(1) mengaktualisasikan potensi peserta
didik sehingga dapat digunakan untuk
memecahkan problema yang dihadapi, (2)
memberikan kesempatan kepada sekolah
untuk mengembangkan pembelajaran yang
fleksibel, sesuai dengan prinsip pendidikan
berbasis luas, dan (3) mengoptimalisasikan
pemanfaatan sumberdaya di lingkungan
sekolah, dengan memberikan peluang
pemanfaatan sumberdaya yang ada di
masyarakat,
sesuai
dengan
prinsip
manajemen berbasis sekolah (Depdiknas,
2002). Kecakapan Hidup Dalam pasal 26 ayat 3 Undang
Undang Republik Indonesia No. 20 tahun
2003 tentang Sistem Pendidikan Nasional
disebutkan bahwa “pendidikan kecakapan
hidup adalah pendidikan yang memberikan
kecakapan personal, kecakapan sosial,
kecakapan intelektual, dan kecakapan
vokasional untuk bekerja atau usaha
mandiri” (Depdiknas, 2003: 59). Kecakapan hidup adalah kecakapan
yang dimiliki seseorang untuk berani
menghadapi
problema
hidup
dan
kehidupan dengan wajar tanpa merasa
tertekan, kemudian secara proaktif dan
kreatif mencari serta menemukan solusi
sehingga
akhirnya
mampu
mengatasinya(Idiran, 2008:3). Kecakapan hidup menurut Nuryanti
dan Mahri (2010) mencakup kecakapan
dasar
dan
kecakapan
instrumental. Kecakapan dasar meliputi : (1) kecakapan
belajar mandiri, (2) kecakapan membaca,
menulis dan menghitung; (3) kecakapan
berkomunikasi; (4) kecakapan berfikir
ilmiah, nalar, rasional, lateral, sistem,
kreatif,
eksploratif,
reasoning,
pengambilan keputusan, dan pemecahan
masalah; (5) kecakapan qolbu/personal; (6)
kecakapan mengelola raga; (7) kecakapan
merumuskan kepentingan dan upaya-upaya
mencapainya;
dan
(8)
kecakapan
berkeluarga
dan
sosial. Kecakapan
instrumental
meliputi:
(1)
kecakapan
memanfaatkan teknologi; (2) kecakapan
mengelola sumber daya; (3) kecakapan
bekerja sama dengan orang lain; (4)
kecakapan memanfaatkan informasi; (5)
kecakapan
menggunakan
sistem;
(6)
kecakapan berwirausaha; (7) kecakapan
kejuruan;
(8)
kecakapan
memilih,
menyiapkan dan mengembangkan karier;
(9) kecakapan menjaga harmoni dengan Ada tiga prinsip mendasar dalam
pengembangan
pendidikan
kecakapan
hidup, yaitu: (1) tidak mengubah sistem
pendidikan yang berlaku saat ini, (2) tidak
harus dengan mengubah kurikulum, sebab
yang justru diperlukan adalah pensiasatan
kurikulum
untuk
diorientasikan
pada
kecakapan hidup, dan (3) etika sosio
religius bangsa dapat diintegrasikan dalam
proses pendidikan. Prinsip-prinsip umum
yang khususnya terkait dengan kebijakan
pendidikan di Indonesia, selain ketiga
prinsip dasar di atas, juga meliputi: (1)
pembelajaran
menggunakan
prinsip learning to know, learning to do,
learning to be, dan learning to live
together, (2)
pelaksanaan
pendidikan
kecakapan hidup
dengan menerapkan
manajemen berbasis sekolah (MBS), (3)
potensi wilayah sekitar sekolah dapat
direfleksikan
dalam
penyelenggaraan
pendidikan,
sesuai
dengan
prinsip
pendidikan kontekstual dan pendidikan
berbasis luas (broad based education, (4) RISTEK: Jurnal Riset, Inovasi dan Teknologi Page 86 Pendidikan karakter yang menjadi
fokus pendidikan di seluruh jenjang
pendidikan diharapkan mampu menjadi
fondasi
utama
dalam
mensukseskan
Indonesia Emas 2025. Nilai-nilai etika inti
yang dikembangkan antara lain kepedulian,
kejujuran, keadilan, tanggung jawab, dan
rasa hormat terhadap diri dan orang lain
bersama
dengan
nilai-nilai
kinerja
pendukungnya seperti ketekunan, etos
kerja yang tinggi, dan kegigihan sebagai
basis karakter yang baik. Kecakapan Hidup Menurut Bashori
(2010), sekolah harus berkomitmen untuk
mengembangkan karakter peserta didik
berdasarkan
nilai-nilai
dimaksud,
mendefinisikannya dalam bentuk perilaku
yang dapat diamati dalam kehidupan
sekolah sehari-hari, mencontohkan nilai-
nilai itu, mengkaji dan mendiskusikannya,
menggunakannya sebagai dasar dalam
hubungan
antarmanusia,
dan
mengapresiasi
manifestasi
nilai-nilai
tersebut di sekolah dan masyarakat. paradigma learning to life and school to
work dapat
dijadikan
dasar
kegiatan
pendidikan, sehingga terjadi pertautan
antara pendidikan dengan kebutuhan nyata
peserta didik, dan (5) penyelenggaraan
pendidikan
senantiasa
diarahkan
agar
peserta didik menuju hidup yang sehat dan
berkualitas; mendapatkan pengetahuan dan
wawasan yang luas; serta memiliki akses
untuk mampu memenuhi standar hidupnya
secara layak (Depdiknas, 2002). Menurut Budiman (2017), Penguatan
Pendidikan Karakter atau PPK adalah
program pendidikan di sekolah untuk
memperkuat
karakter
siswa
melalui
harmonisasi olah hati (etik dan spiritual),
olah rasa (estetik), olah pikir (literasi dan
numerasi) dan olah raga (kinestetik) sesuai
dengan
falsafah
Pancasila. Penguatan
Pendidikan
Karakter
hadir
untuk
menyiapkan Generasi Emas 2045 yang
memiliki kecakapan abad 21. Dengan
menempatkan kembali karakter sebagai
ruh pendidikan di Indonesia, berdampingan
dengan
intelektualitas,
PPK
berperan
dalam pembentukan generasi muda yang
tangguh, cerdas dan berkarakter. Sebagai
bagian dari Gerakan Nasional Revolusi
Mental, PPK menguatkan lima nilai utama
karakter pada siswa pendidikan dasar, di
antaranya: Religius, Nasionalis, Mandiri,
Gotong royong dan Integritas. Karakter
yang kuat membentuk individu menjadi
pelaku perubahan bagi diri sendiri dan
masyarakat sekitarnya. Kerja sama antara
sekolah, keluarga dan masyarakat menjadi
kunci penerapan penguatan pendidikan
karakter. Sebagai
program
prioritas
pendidikan dan kebudayaan, Gerakan PPK
berfokus pada struktur yang sudah ada
dalam sistem pendidikan nasional, yaitu:
program, kurikulum dan kegiatan yang
berbasis pada kelas, budaya sekolah, dan
masyarakat. 1
Hubungan Kecakapan Hidup dengan Dengan melihat hasil pendidikan
selama ini dan tuntutan perubahan zaman
mendorong munculnya paradigma baru
dalam pendidikan bahwa: (1) pendidikan
harus ditujukan bagi kepentingan dan
kebutuhan masyarakat luas, bukan hanya
bagi kelompok masyarakat tertentu, (2)
pendidikan harus berorientasi populistis,
tidak boleh elistis semata, (3) pendidikan
harus mampu memenuhi kebutuhan dan
hajat
hidup
masyarakat
luas,
bukan
sebagian kecil masyarakat, (4) pendidikan
harus kontekstual dan cocok dengan
keadaan dan kebutuhan masyarakat yang
diidamkan,
(5)
pembelajaran
harus
diarahkan untuk menyantuni pembelajar
agar mampu hidup mandiri dan otonom
dalam hidup dan kehidupan masing-
masing, bukan hanya memiliki kepandaian
akademis-intelektual, (6) sekolah harus
senantiasa
terlibat
dalam
perubahan
masyarakat,
dan
(7)
sekolah
juga
merupakan
lembaga
pembentukan 1. Hubungan Kecakapan Hidup dengan
Pendidikan Karakter 1. Kecakapan Hidup Hubungan Kecakapan Hidup dengan
Pendidikan Karakter Pendidikan Karakter RISTEK: Jurnal Riset, Inovasi dan Teknologi Page 87 RISTEK: Jurnal Riset, Inovasi dan Teknologi bingkai
sekolah
adiwiyata
dengan
mempertimbangkan hubungan erat antara
unsur-unsur yang ada pada kecakapan
hidup dan pendidikan karakter. Nilai-nilai
kecakapan dasar dan instrumental dalam
kecakapan hidup dan nilai-nilai religius,
nasionalis, mandiri, gotong-royong dan
integritas
dalam
pendidikan
karakter
dibangun
untuk
mengembangkan
entrepreneurship
dengan
melaksanakan
kegiatan-kegiatan berbasis kelas, budaya
sekolah dan masyarakat seperti (i)
Membuat
taman,
membuat
tempat
pemilahan sampah dari bahan bekas,
membuat pot bunga dan tanaman dari
bahan-bahan bekas, (ii) Olah hasil kebun
untuk
komoditi
dan
pemasarannya:
Minuman Jahe, bibit tanaman, jajan
tradisional, (iii) Gerakan go green dan go
clean,
gerakan
membaca
(literasi),
berpacu dalam prestasi, ekstrakurikuler,
pemberian
penghargaan,
publikasi
facebook, blogspot, lewat koran dan media
cetak
lainnya,
(iv)
Pembenihan,
pembibitan, pengembangbiakan vegetatif
(cangkok, tempel, sisip, stek) produksi
pupuk bokhasi, pupuk organik cair dan
Minuman Jahe Merah. Kerangka berfikir
pengembangan entrepreneurship melalui
kecakapan hidup berbasis potensi lokal
dalam penguatan pendidikan karakter
dapat dituangkan dalam gambar model
pengembangan berikut: kecakapan hidup, yang dibutuhkan oleh
masyarakat
luas
dalam
hidup
dan
kehidupan sehari-hari, bukan penerusan
ilmu
pengetahuan
teoritis-akademis
semata. Paradigma
baru
berbasis
masyarakat luas itulah yang kemudian
mendorong
perlunya
pendidikan
kecakapan hidup di sekolah. bingkai
sekolah
adiwiyata
dengan
mempertimbangkan hubungan erat antara
unsur-unsur yang ada pada kecakapan
hidup dan pendidikan karakter. Nilai-nilai
kecakapan dasar dan instrumental dalam
kecakapan hidup dan nilai-nilai religius,
nasionalis, mandiri, gotong-royong dan
integritas
dalam
pendidikan
karakter
dibangun
untuk
mengembangkan
entrepreneurship
dengan
melaksanakan
kegiatan-kegiatan berbasis kelas, budaya
sekolah dan masyarakat seperti (i)
Membuat
taman,
membuat
tempat
pemilahan sampah dari bahan bekas,
membuat pot bunga dan tanaman dari
bahan-bahan bekas, (ii) Olah hasil kebun
untuk
komoditi
dan
pemasarannya:
Minuman Jahe, bibit tanaman, jajan
tradisional, (iii) Gerakan go green dan go
clean,
gerakan
membaca
(literasi),
berpacu dalam prestasi, ekstrakurikuler,
pemberian
penghargaan,
publikasi
facebook, blogspot, lewat koran dan media
cetak
lainnya,
(iv)
Pembenihan,
pembibitan, pengembangbiakan vegetatif
(cangkok, tempel, sisip, stek) produksi
pupuk bokhasi, pupuk organik cair dan
Minuman Jahe Merah. Kerangka berfikir
pengembangan entrepreneurship melalui
kecakapan hidup berbasis potensi lokal
dalam penguatan pendidikan karakter
dapat dituangkan dalam gambar model
pengembangan berikut: 4. Adiwiyata
sebagai
Penunjang
Pendidikan Karakter dan Kecakapan
Hidup
dalam
pengembangan
entrepreneurship. Program Adiwiyata yang diterapkan
sekolah
dapat
menunjang
penguatan
pendidikan karakter dan kecakapan hidup. Nilai-nilai
religius,
nasionalis,
kemandirian, gotong royong dan integritas
dalam PPK manjadi nilai-nilai yang
terintegrasi
dalam
kecakapan
dasar
maupun intrumental. Keduanya menjadi
faktor penting dalam mengembangkan
entrepreneurship. Kecakapan Hidup Pengembangan
entrepreneurship
didasarkan pada nilai-nilai kecakapan
hidup berbasis potensi lokal serta mengacu
kepada penguatan pendidikan karakter
bangsa dalam berbagai kegiatan di sekolah
seperti
pada
pembelajaran
di
kelas,
kegiatan ekstrakurikuler, maupun dalam
kegiatan-kegiatan lainnya diharapkan dapat
menumbuhkembangkan
jiwa
pantang
menyerah, etos kerja yang tinggi, mandiri,
kreatif, inovatif dalam menciptakan karya-
karya di segala bidang, sehingga pada
akhirnya mampu menciptakan sumber
daya manusia yang mumpuni bagi negara
ini. Adapun
berbasis
potensi
lokal
dimaksudkan agar kelak generasi muda
dapat menjadi agen of change bagi
pembangunan di daerahnya sendiri. Pengembangan unsur-unsur utama
yang ada dalam entrepreneurship, yaitu:
penerapan
kreativitas
dan
inovasi,
pemanfaatan peluang, membuat perubahan,
dan memberikan nilai tambah bagi diri
sendiri dan orang lain dilaksanakan dalam RISTEK: Jurnal Riset, Inovasi dan Teknologi Page 88 Page 88 Gambar 1
Model Pengembangan Entrepreneurship melalui Kecakapan Hidup Berbasis Potensi
Lokal sebagai Model Penguatan Pendidikan Karakter di Sekolah Adiwiyata
Potensi Lokal
Adiwiyata
Entre-
preneurship
Kecakapan
Hidup
PPK Potensi Lokal Entre-
preneurship Adiwiyata Gambar 1
Model Pengembangan Entrepreneurship melalui Kecakapan Hidup Berbasis Potensi
Lokal sebagai Model Penguatan Pendidikan Karakter di Sekolah Adiwiyata daya manusia yang mumpuni bagi negara
ini. Adapun
berbasis
potensi
lokal
dimaksudkan agar kelak generasi muda
dapat menjadi agen of change bagi
pembangunan di daerahnya sendiri. METODOLOGI PENELITIAN Metode penelitian ini menggunakan
penelitian deskriptif dengan pendekatan
kualitatif. Data
dikumpulkan
dengan
menggunakan
metode
survey
dengan
langkah-langkah meliputi : 1) merumuskan
masalah penelitian dan menentukan tujuan
survei; 2) menentukan konsep dan hipotesa
serta
menggali
kepustakaan;
3)
pengambilan
sampel;
4)
pembuatan
instrumen
penelitian;
5)
pekerjaan
lapangan; 6) pengolahan data; 7) analisa
dan pelaporan. Langkah-langkah strategis
tersebutdipersiapkan untuk
mengurangi
pengangguran dengan mengembangkan
entrepreneurship didasarkan pada nilai-
nilai kecakapan hidup atau kecakapan
hidup berbasis potensi lokal serta mengacu
kepada penguatan pendidikan karakter
bangsa dalam berbagai kegiatan di sekolah
seperti
pada
pembelajaran
di
kelas,
kegiatan ekstrakurikuler, maupun dalam
kegiatan-kegiatan lainnya diharapkan dapat
menumbuhkembangkan
jiwa
pantang
menyerah, etos kerja yang tinggi, mandiri,
kreatif, inovatif dalam menciptakan karya-
karya di segala bidang, sehingga pada
akhirnya mampu menciptakan sumber Data
yang
digunakan
dalam
penelitian ini diperoleh dengan cara
melakukan observasi (lembar penilaian),
wawancara, dan dokumentasi di lokasi
penelitian meliputi: SMP N 3 Tersono,
SMP N 1 Batang, SMP N 2 Wonotunggal,
SD N Wonobodro 01, SD N Karangtengah
01 dan SMA N 1 Bandar. Observasi
dilaksanakan
dengan
menggunakan
instrumen pengukuran lembar penilaian
peer-assessment
untuk
mengukur
indikator-indikator
penerapan
entrepreneurship melalui kecakapan hidup
berbasis potensi lokal sebagai model
penguatan
pendidikan
karakter. Wawancara
dilaksanakan
untuk
mengungkap nilai-nilai karakter bangsa,
unsur
kecakapan
hidup
dan
entrepreneurship yang muncul selama
penelitian
berlangsung. Dokumentasi
dilaksanakan untuk mengabadikan proses
penerapan
entrepreneurship
melalui
kecakapan hidup berbasis potensi lokal RISTEK: Jurnal Riset, Inovasi dan Teknologi Page 89 HASIL DAN PEMBAHASAN sebagai
model
penguatan
pendidikan
karakter. Program
pengembangan
entrepreneurship melalui kecakapan hidup
berbasis potensi lokal sebagai model
penguatan pendidikan karakter di sekolah
adiwiyata
secara
umum
akan
menghasilkan: Validitas
instrumen
dalam
penelitian ini menggunakan face validity
dan
content
validity. Face
validity
(validitas rupa) adalah validitas yang
menunjukan
apakah
alat
pengukur/instrumen penelitian dari segi
rupanya nampak mengukur apa yang ingin
diukur, validitas ini lebih mengacu pada
bentuk
dan
penampilan
instrumen(Kenneth, 1999). 1. Kemampuan menerapkan kreativitas
dan inovasi, memanfaatkan peluang,
membuat perubahan dan kemampuan
mewujudkan nilai tambah bagi diri
sendiri dan orang lain. (
,
)
Selanjutnya,
untuk
keperluan
mengklasifikasikan
kualitas
entrepreneurship melalui kecakapan hidup
berbasis potensi lokal sebagai model
penguatan pendidikan karakter dengan
menggunakan
lembar
penilaian
peer
assessment, peneliti menentukan kriteria
pencapaian prosentase perolehan skor
secara kualitatif sendiri yaitu menurut
interval sebagai berikut:
Baik sekali > 90 %, 80 % < Baik < 90 %,
70 % < Cukup < 80 %, Kurang < 70 % 2. Kecakapan
dasar
dan
kecakapan
intrumental peserta didik yang sangat
dibutuhkan
dalam
menghadapi
tantangan
hidup
yang
semakin
kompleks. entrepreneurship melalui kecakapan hidup
berbasis potensi lokal sebagai model
penguatan pendidikan karakter dengan
menggunakan
lembar
penilaian
peer
assessment, peneliti menentukan kriteria
pencapaian prosentase perolehan skor
secara kualitatif sendiri yaitu menurut
interval sebagai berikut: 3. Karakter peserta didik yang religius,
nasionalis,
mandiri
dan
memiliki
semangat gotong royong tinggi. Program kegiatan pengembangan
yang
dimaksud
sebagaimana
tersebut
dalam tabel 1,2,3, dan 4. Tabel 4 Program Kegiatan Memberikan Nilai Tambah bagi Diri Sendiri dan Orang
Lain HASIL DAN PEMBAHASAN Baik sekali > 90 %, 80 % < Baik < 90 %,
70 % < Cukup < 80 %, Kurang < 70 % Tabel 1 Program Kegiatan Penerapan Kreativitas dan Inovasi Tabel 1 Program Kegiatan Penerapan Kreativitas dan Inovasi
Unsur
Entrepreneurship
Yang Dikembangkan
Life Skill Yang
Digunakan
Program Kegiatan
Penguatan
Pendidikan
Karakter Yang
Dicapai
Penerapan kreativitas
dan inovasi
Kecakapan
Dasar:
mandiri,
membaca,
menulis dan menghitung,
sistem, kreatif, mengelola
raga,
merumuskan
kepentingan dan upaya-
upaya untuk mencapainya
Kecakapan
Instrumental:
Memanfaatkan teknologi,
Bekerja
sama
dengan
orang
lain,
menjaga
harmoni
dengan
lingkungan
Berbasis Kelas:
Membuat
taman,
membuat
tempat
pemilahan
sampah
dari
bahan
bekas,
membuat pot bunga
dan
tanaman
dari
bahan-bahan bekas
Religius:
Bersih,
Cinta
Lingkungan
Nasionalis:
menghargai
kebhinekaan
Mandiri: kerja
keras,
kreatif,
disiplin, berani,
pembelajar
Gotong royong:
kerja sama
Integritas: cinta
pada kebenaran RISTEK: Jurnal Riset, Inovasi dan Teknologi
Page 90
Tabel 2 Program Kegiatan Pemanfaatan Peluang Tabel 2 Program Kegiatan Pemanfaatan Peluang RISTEK: Jurnal Riset, Inovasi dan Teknologi Page 90 Page 90 Unsur
Entrepreneurship
Yang Dikembangkan
Life Skill Yang
Digunakan
Program Kegiatan
Penguatan
Pendidikan
Karakter Yang
Dicapai
Pemanfaatan peluang
Kecakapan
Dasar:
mandiri, membaca, menulis
dan menghitung, sistem,
kreatif,
eksploratif,
pengambilan
keputusan,merumuskan
kepentingan
dan
upaya-
upaya untuk mencapainya
Kecakapan Instrumental:
Memanfaatkan
teknologi,
mengelelola
sumber
daya,Bekerja sama dengan
orang lain, memanfaatkan
informasi,menggunakan
sistem,
berwirausaha,
memilih,
menyiapkan,menjaga
harmoni
dengan
lingkungan,
menyatukan
bangsa
Berbasis
Masyarakat:
Olah
hasil
kebun
untuk komoditi dan
pemasarannya:
Minuman Jahe, bibit
tanaman,
jajan
tradisional
Religius: beriman,
cinta lingkungan
Mandiri:
kerja
keras dan kreatif
Gotong
royong:
kerja sama Tabel 3 Program Kegiatan Membuat Perubahan
Unsur
Entrepreneurship
Yang
Dikembangkan
Life Skill Yang Digunakan
Program Kegiatan
Penguatan
Pendidikan
Karakter Yang
Dicapai
Membuat perubahan
Kecakapan Dasar: mandiri,
membaca,
menulis
dan
menghitung, lateral, sistem,
kreatif,
mengelola
raga,
merumuskan
kepentingan
dan
upaya-upaya
untuk
mencapainya
Kecakapan Instrumental:
Memanfaatkan
teknologi,
mengelelola
sumber
daya,Bekerja sama dengan
orang lain, memanfaatkan
informasi,menggunakan
sistem,
berwirausaha,
memilih,
menyiapkan,menjaga
harmoni dengan lingkungan,
menyatukan bangsa
Berbasis
Budaya
Sekolah
dan
Masyarakat:
Gerakan
go
green
dan
go
clean,
gerakan
membaca
(literasi), berpacu
dalam
prestasi,
ekstrakurikuler,
pemberian
penghargaan,
publikasi
facebook,
blogspot, lewat koran
dan
media
cetak
lainnya
Religius: beriman,
bertaqwa, Bersih,
toleransi,
Cinta
Lingkungan
Nasionalis:
cinta
tanah air, semangat
kebangsaan
Mandiri:
kerja
keras,
kreatif,
disiplin,
berani,
pembelajar
Gotong
royong:
kerja sama, saling
menolong
Integritas:
keteladanan Tabel 3 Program Kegiatan Membuat Perubahan RISTEK: Jurnal Riset, Inovasi dan Teknologi
Page 91
Tabel 4 Program Kegiatan Memberikan Nilai Tambah bagi Diri Sendiri dan Orang
Lain RISTEK: Jurnal Riset, Inovasi dan Teknologi Page 91 Page 91 Unsur
Entrepreneurship
Yang Dikembangkan
Life Skill Yang
Digunakan
Program
Kegiatan
Penguatan
Pendidikan
Karakter Yang
Dicapai
Memberikan
nilai
tambah
Kecakapan
Dasar:
mandiri,
kreatif,
mengelola
raga,
merumuskan kepentingan
dan upaya-upaya untuk
mencapainya
Kecakapan
Instrumental:
Memanfaatkan teknologi,
mengelelola sumber daya,
Bekerja
sama
dengan
orang lain, memanfaatkan
informasi,
berwirausaha,
menyiapkan,menjaga
harmoni
dengan
lingkungan,
menyatukan
bangsa
Berbasis
Kelas,
Budaya
Sekolah
dan Masyarakat:
Pembenihan,
pembibitan,
pengembangbiakan
vegetatif (cangkok,
tempel, sisip, stek),
produksi
pupuk
bokhasi,
pupuk
organik cair dan
Minuman
Jahe
Merah
Religius:
beriman,
bertaqwa,
toleransi, Cinta
Lingkungan
Nasionalis:
cinta tanah air,
menghargai
kebhinekaan
Mandiri: kerja
keras,
kreatif,
disiplin, berani
Gotong
royong:
kerja
sama,
saling
menolong penguatan pendidikan karakter di sekolah
adiwiyata sebagaimana terlihat pada grafik
1,2,3,4 dan 5. HASIL DAN PEMBAHASAN Adapun
seberapa
besar
kualitas
pengembangan
pengembangan
entrepreneurship melalui kecakapan hidup
berbasis potensi lokal sebagai model entrepreneurship melalui kecakapan hidup
berbasis potensi lokal sebagai model Grafik 1 Pencapaian Unsur Penerapan Kreativitas dan Inovasi Grafik 1 Pencapaian Unsur Penerapan Kreativitas dan Inovasi
Berdasarkan grafik 1, pencapaian
skor unsur penerapan kreativitas dan
inovasi sebesar 78,75% yang merupakan
rerata dari capaian skor kecakapan dasar
81,25% dan capaian skor kecakapan
instrumental 76,25%. Dengan demikian,
Kualitas entrepreneurship ditinjau dari
pencapaian unsur penerapan kreativitas
dan inovasi masuk dalam kategori cukup. Grafik 1 Pencapaian Unsur Penerapan Kreativitas dan Inovasi instrumental 76,25%. Dengan demikian,
Kualitas entrepreneurship ditinjau dari
pencapaian unsur penerapan kreativitas
dan inovasi masuk dalam kategori cukup. Berdasarkan grafik 1, pencapaian
skor unsur penerapan kreativitas dan
inovasi sebesar 78,75% yang merupakan
rerata dari capaian skor kecakapan dasar
81,25% dan capaian skor kecakapan Grafik 2 Pencapaian Unsur Pemanfaatan Peluang Grafik 2 Pencapaian Unsur Pemanfaatan Peluang RISTEK: Jurnal Riset, Inovasi dan Teknologi Page 92 Page 92 Berdasarkan grafik 2, pencapaian
Dengan
demikian,
Kualitas Berdasarkan grafik 2, pencapaian
skor unsur pemanfaatan peluang sebesar
80% yang merupakan rerata dari capaian
Dengan
demikian,
Kualitas
entrepreneurship ditinjau dari pencapaian
unsur pemanfaatan peluang masuk dalam demikian, Berdasarkan grafik 2, pencapaian Berdasarkan grafik 2, pencapaian
skor unsur pemanfaatan peluang sebesar
80% yang merupakan rerata dari capaian
skor kecakapan dasar 81,25% dan capaian
skor kecakapan instrumental
78,75%. g
,
entrepreneurship ditinjau dari pencapaian
unsur pemanfaatan peluang masuk dalam
kategori cukup. Grafik 3 Pencapaian Unsur Membuat Perubahan
Berdasarkan grafik 3, pencapaian
skor unsur membuat perubahan sebesar
80,625% yang merupakan rerata dari
capaian skor kecakapan dasar 81,25% dan
Dengan
demikian,
Kualitas
entrepreneurship ditinjau dari pencapaian
unsur membuat perubahan masuk dalam
kategori baik. Grafik 3 Pencapaian Unsur Membuat Perubahan p demikian, Dengan Berdasarkan grafik 3, pencapaian
skor unsur membuat perubahan sebesar
80,625% yang merupakan rerata dari
capaian skor kecakapan dasar 81,25% dan
capaian skor kecakapan instrumental 80%. Dengan
demikian,
Kualitas
entrepreneurship ditinjau dari pencapaian
unsur membuat perubahan masuk dalam
kategori baik. G
fik 4 P
i
U
M
b
ik
Nil i T
b h Berdasarkan grafik 3, pencapaian
skor unsur membuat perubahan sebesar
80,625% yang merupakan rerata dari
capaian skor kecakapan dasar 81,25% dan
capaian skor kecakapan instrumental 80%. g
,
entrepreneurship ditinjau dari pencapaian
unsur membuat perubahan masuk dalam
kategori baik. g
,
entrepreneurship ditinjau dari pencapaian
unsur membuat perubahan masuk dalam
kategori baik. HASIL DAN PEMBAHASAN Perlu
dipikirkan
bagaimana
memasarkan
atau
memanfaatkan
produk kecakapan hidup seperti hasil
agrobisnis(hasil pengolahan jahe merah
dan hasil cangkok, sambung, sisip,
tempel tanaman ungggulan), hasta
karya bahan bekas, kreasi seni budaya,
pupuk bokhasi, pupuk organik cair dan
produk lainnya berbasis potensi lokal
yang ada di sekolah adiwiyata masing-
masing ke lingkungan yang lebih luas
sebagai tambahan pendapatan sekolah
untuk menunjang kegiatan sekolah. Berdasarkan grafik 5, pencapaian skor
entrepreneurship melalui kecakapan hidup
berbasis potensi lokal sebagai model
penguatan pendidikan karakter sebesar
80,15625% yang merupakan rerata dari
capaian skor kecakapan dasar 81,5625%
dan capaian skor kecakapan instrumental
78,75%. Dengan
demikian,
kualitas
entrepreneurship melalui kecakapan hidup
berbasis potensi lokal sebagai model
penguatan pendidikan karakter masuk
dalam kategori baik. HASIL DAN PEMBAHASAN Grafik 4 Pencapaian Unsur Memberikan Nilai Tambah RISTEK: Jurnal Riset, Inovasi dan Teknologi
Page 93
Grafik 4 Pencapaian Unsur Memberikan Nilai Tambah
Berdasarkan grafik 4, pencapaian
skor unsur memberikan nilai tambah
sebesar 80,625% yang merupakan rerata
dari capaian skor kecakapan dasar 81,25% Berdasarkan grafik 4, pencapaian
skor unsur memberikan nilai tambah sebesar 80,625% yang merupakan rerata
dari capaian skor kecakapan dasar 81,25% Berdasarkan grafik 4, pencapaian
skor unsur memberikan nilai tambah sebesar 80,625% yang merupakan rerata
dari capaian skor kecakapan dasar 81,25% RISTEK: Jurnal Riset, Inovasi dan Teknologi
Page 93
Berdasarkan grafik 4, pencapaian
skor unsur memberikan nilai tambah
sebesar 80,625% yang merupakan rerata
dari capaian skor kecakapan dasar 81,25% RISTEK: Jurnal Riset, Inovasi dan Teknologi RISTEK: Jurnal Riset, Inovasi dan Teknologi Page 93 dan capaian skor kecakapan instrumental
80%. Dengan
demikian,
Kualitas
entrepreneurship ditinjau dari pencapaian
unsur memberikan nilai tambah masuk
dalam kategori baik. Grafik 5 Pencapaian Skor entrepreneurship melalui kecakapan hidup berbasis potensi
lokal sebagai model penguatan pendidikan karakter unsur memberikan nilai tambah masuk
dalam kategori baik. Dengan
demikian,
Kualitas
eurship ditinjau dari pencapaian
dalam kategori baik. Pencapaian Skor entrepreneurship melalui kecakapan hidup berbasis potensi
lokal sebagai model penguatan pendidikan karakter neurship ditinjau dari pencapaian
Pencapaian Skor entrepreneurship melalui kecakapan hidup berbasis potensi
lokal sebagai model penguatan pendidikan karakter entrepreneurship ditinjau dari pencapaian
Grafik 5 Pencapaian Skor entrepreneurship melalui kecakapan hidup berbasis potensi
lokal sebagai model penguatan pendidikan karakter
Berdasarkan grafik 5, pencapaian skor
entrepreneurship melalui kecakapan hidup
berbasis potensi lokal sebagai model
penguatan pendidikan karakter sebesar
80,15625% yang merupakan rerata dari
capaian skor kecakapan dasar 81,5625%
dan capaian skor kecakapan instrumental
78,75%. Dengan
demikian,
kualitas
entrepreneurship melalui kecakapan hidup
berbasis potensi lokal sebagai model
penguatan pendidikan karakter masuk
dalam kategori baik. SARAN
2. Perlu
dipikirkan
bagaimana
memasarkan
atau
memanfaatkan
produk kecakapan hidup seperti hasil
agrobisnis(hasil pengolahan jahe merah
dan hasil cangkok, sambung, sisip,
tempel tanaman ungggulan), hasta
karya bahan bekas, kreasi seni budaya,
pupuk bokhasi, pupuk organik cair dan
produk lainnya berbasis potensi lokal
yang ada di sekolah adiwiyata masing-
masing ke lingkungan yang lebih luas
sebagai tambahan pendapatan sekolah
untuk menunjang kegiatan sekolah. 2. DAFTAR PUSTAKA Dari beberapa temuan dilapangan
selama pendidikan kecakapan hidup
dilaksanakan dan dari kesimpulan yang
diperoleh, maka penulis menyarankan
sebagai berikut: Ancok,
Djamaludin. 2012. Psikologi
Kepemimpinan & Inovasi. Jakarta:
Penerbit. Erlangga. Anonim. 2003. Undang-undang RI Nomor
20
Tahun
2003
tentang
Sistem
Pendidikan Nasional. Jakarta: Sinar
Grafika. 1. Kecakapan hidup berbasis potensi lokal
dalam pengembangan entrepreneurship
harapannya dapat dijadikan sebagai
alternatif pilihan model penguatan
pendidikan
karakter
karena
dapat
membangkitkan dan menguatkan sikap
wirausaha serta menguatkan karakter
bangsa siswa. Bailey Kenneth D.1999. Metods of Social
Research. New York: The Free Pers Bashori,
K. 2010. Menata
Ulang
Pendidikan
Karakter
Bangsa. http://www.mediaindonesia.co
m/read/2010/03/15/129378/68/11/Mena
ta-Ulang-Pendidikan-Karakter-Bangsa. Diakses 6 November 2017. RISTEK: Jurnal Riset, Inovasi dan Teknologi Page 94 Muhadjir, Noeng. 1996. Metode Penelitian
Kualitatif. Yogyakarta: Rake Sarasin. Budiman, Arie.2017. Gerakan Penguatan
Pendidikan
Karakter. Jakarta:
Kemendikbud
Depdiknas. (2002). Manajemen Mutu
Berbasis Sekolah. Jakarta: Departemen
Pendidikan
Nasional. Direktorat
Pendidikan Menengah Umum
Depdiknas. 2002. Pedoman Pelaksanaan
Pendidikan Kecakapan Hidup. Buku I,
II, dan III. Jakarta: Depdiknas. Depdiknas.2003. Pengembangan Model
Pendidikan Kecakapan Hidup. Jakarta:
Puskur, Balitbang Depdiknas
Depdiknas. 2003. Kurikulum Sekolah
Menengah Kejuruan. Jakarta : Bumi
Aksara Nasution.1988.metode Penelitian Kualitatif
Naturalistik. Jakarta: Sinar Grafika. Nuryanti
dan
Mahri.2010. Model
Pendidikan Pengembangan Kecakapan
Hidup
Berlandaskan
Jiwa
Kewirausahaan. Proceedings Of The 4th
International Conference On Teacher
Education; Join Conference UPI &
UPSI
Bandung,
Indonesia,
8-10
November 2010 II, dan III. Jakarta: Depdiknas. Depdiknas.2003. Pengembangan Model
Pendidikan Kecakapan Hidup. Jakarta:
Puskur, Balitbang Depdiknas Rutherford, Donald.2002. Dictionary of
Economics. London,: Routledge. Depdiknas. 2003. Kurikulum Sekolah
Menengah Kejuruan. Jakarta : Bumi
Aksara. Sugiyono. 2012. Metode
Penelitian
Kuantitatif. Bandung: Alfabeta. Sunyoto, Danang. 2013. Kewirausahaan
Untuk Kesehatan. Yogyakarta: Nuha
Medika Friday
O. Okpara,
“The
Value
of
Creativity
and
Innovation
in
Entrepreneurship”, Journal of Asia
Entrepreneurship
and
Sustainability,
Rossi Smith Academic Publishing,
Oxford, 2007, hal.1 Saryono,
Djoko. 2002. Pendidikan
Kecakapan
Hidup:
Konsepsi
dan
Implementasinya di Sekolah. Makalah
dalam Workshop Pengembangan Sistem
Pendidikan
Dasar
dan
Menengah
Berorientasi Kecakapan Hidup di Jawa
Timur, 11 November 2002, Universitas
Negeri Malang. Idiran, Zulkarnain. 2008. Pola Pelaksanaan
Pendidikan
Berorientasi
Kecakapan
Hidup
http://zulkarnainidiran.wordpress.com/2
008/11/28/pola-pelaksanaan-
pendidikan-berorientasi-kecakapan-
hidup-life-skill-education. diakses 28
November 2008 Suryana. 2013. Kewirausahaan Kiat dan
Proses
Menuju
Sukses. Jakarta:
Salemba Empat. Kasmir.2013. Kewirausahaan. Jakarta:
Raja Grafindo Persada Kasmir.2013. Kewirausahaan. Jakarta:
Raja Grafindo Persada Tim
Broad-Based
Education. 2002. Pendidikan
Berorientasi
Kecakapan
Hidup (LifeSkill) Melalui Pendekatan
Broad-Based
Education
(BBE). Departemen Pendidikan Nasional. Marlo,
Abu. 2013. Entrepreneurship
Hukum
Langit:
Jakarta:
Gramedia
Pustaka Utama Marlo,
Abu. 2013. Entrepreneurship
Hukum
Langit:
Jakarta:
Gramedia
Pustaka Utama Undang-undang Republik Indonesia N0. 20
Tahun
2003. 1Kepala SMP N 3 Tersono Batang Jawa Tengah, Guru Berprestasi I Batang Tahun 2007, Guru
Berprestasi Harapan II Jateng Tahun 2007, Juara II Simposium Guru Tk. Nasional Tahun 2005, Juara
III Jateng -Lomba Keberhasilan Pembelajaran Tahun 2006, dan Lulusan Terbaik PPG UNY Tahun 2009 DAFTAR PUSTAKA Tentang
Sistem
Pendidikan
Nasional. Departemen
Pendidikan
Nasional
Republik
Indonesia. 2003. Miles, MB & Huberman AM. 1984. Analisis Data Kualitatif. Terjemahan
oleh Tjetcep Rohendi Rohidi.1992. Jakarta:Universitas Indonesia Moleong,
J
Lexy.2007. Metodologi
Penelitian Kualitatif. Bandung: PT. Rosda Karya 1Kepala SMP N 3 Tersono Batang Jawa Tengah, Guru Berprestasi I Batang Tahun 2007, Guru
Berprestasi Harapan II Jateng Tahun 2007, Juara II Simposium Guru Tk. Nasional Tahun 2005, Juara
III Jateng -Lomba Keberhasilan Pembelajaran Tahun 2006, dan Lulusan Terbaik PPG UNY Tahun 2009 RISTEK: Jurnal Riset, Inovasi dan Teknologi Page 95 Page 95
|
https://openalex.org/W4221096917
|
https://jeo-esska.springeropen.com/counter/pdf/10.1186/s40634-022-00467-x
|
English
| null |
On the value and limitations of incorporating a “clean phase” into the surgical treatment of prosthetic joint infections – an illustrative cadaveric study using fluorescent powder
|
Journal of experimental orthopaedics
| 2,022
|
cc-by
| 4,779
|
Abstract The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. On the value and limitations
of incorporating a “clean phase” into the surgical
treatment of prosthetic joint infections –
an illustrative cadaveric study using fluorescent
powder eorges Vles1,2* , Jeroen Bossen2, Johannes Kloos2, Philippe Debeer1,2† and Stijn Ghijselings1, Journal of
Experimental Orthopaedics Journal of
Experimental Orthopaedics Journal of
Experimental Orthopaedics Vles et al. Journal of Experimental Orthopaedics (2022) 9:28
https://doi.org/10.1186/s40634-022-00467-x Open Access Abstract Purposes: A septic revision of an artificial joint is routinely split up in a so-called dirty phase and a clean phase. The
measures taken to initiate the start of the clean phase vary significantly between musculoskeletal infection centers. We performed simulations of one-step exchanges of infected THAs and sought to 1) determine the effect of differ-
ent clean phase protocols on the sterile field, and 2) determine whether or not it is possible to re-implant the new
prosthesis completely clean. Methods: Nine fresh frozen cadaveric hips were used and primary THA was undertaken via a direct anterior
approach. Before implantation of the components varying amounts of fluorescent powder (GloGerm) were depos-
ited, simulating bacterial infection. Second, a one-step exchange was performed via a posterolateral approach. After
implant removal, debridement, and lavage, randomization determined which clean phase protocol was followed, i.e. no, some or full additional measures. Finally, the new prosthesis was re-implanted. In order to determine the effect of different clean phase protocols on contamination of the sterile field standardized
UV light-enhanced photographs were obtained of 1) the gloves, 2) the instrument table, 3) the drapes, and 4) the
wound and these were ranked on cleanliness by a blinded panel of hip surgeons. In order to determine whether or not it is possible to re-implant the prosthesis completely clean, the implant was
taken out again at the end of the one-step exchange and inspected for contamination under UV light. Results: The gloves, the instrument table, the drapes and the wound were significantly cleaner after a clean phase
using full additional measures compared to partial or no additional measures (p < 0.000). Partial measures were able to
reduce some of the contamination of the gloves and the wound, but had no effect on the drapes and the instrument
table. All re-implanted implants were contaminated with some amount of fluorescent powder at the end of the one-
step exchange. © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://creativecommons.org/licenses/by/4.0/. *Correspondence: Georges.Vles@uzleuven.be
†Philippe Debeer and Stijn Ghijselings contributed equally to this work
and are therefore joint senior author.
1 Department of Development and Regeneration, Faculty of Medicine,
Institute for Orthopaedic Research and Training (IORT), Leuven, KU,
Belgium
Full list of author information is available at the end of the article Introduction gloves, the tip of the diathermy and suction, and place
one extra drape [10, 11]. h Prosthetic Joint Infection (PJI) remains a devastat-
ing complication following an otherwise highly suc-
cessful surgical intervention and generally requires
re-operation [1–4]. Regardless whether the next proce-
dure is a Debridement, Antibiotics and Implant Reten-
tion (DAIR), a one-step exchange or the 1st stage of a
two-step exchange, there routinely is a dirty phase and
a clean phase [5]. During the dirty phase the bacterial
load is reduced as much as possible by debridement of
infected and necrotic tissue and by extensive pulsatile
lavage with saline ± antiseptics. At one point, one needs
to re-implant the mobile parts, the entire implant or the
spacer, depending on the chosen strategy, and this marks
the beginning of the clean phase.h The question remains if we are not selling our patients
and ourselves short by this compromise. We therefore
performed simulations of one-step exchanges of infected
total hip arthroplasties on cadavers in order to 1) deter-
mine the effect of different clean phase protocols on the
sterile field, and 2) determine whether or not it is possible
to re-implant the prosthesis completely clean. Vles et al. Journal of Experimental Orthopaedics (2022) 9:28 Page 2 of 9 Conclusions: We advise to incorporate a clean phase with full additional measures into the surgical treatment of
prosthetic joint infections, as partial measures seem to be a poor compromise. Level of evidence: Not applicable (cadaveric study). Keywords: Prosthetic joint infection, Clean phase, Fluorescent powder, Cadaveric study, Revision arthroplasty,
Surgical technique, Contamination, Bacterial load reduction ns: We advise to incorporate a clean phase with full additional measures into the surgical treatment
joint infections, as partial measures seem to be a poor compromise. Materials and methods
Study design and settingh This cadaveric study was approved by our ethical review
board and registered at the Belgian National Council for
Bioethics (NH019 2021–6-01). The implant investigated
was an uncemented Total Hip Arthroplasty (THA, Pin-
nacle cup, Corail Stem, DePuy Synthes, Warsaw, Indi-
ana) and bacterial load was simulated using a fluorescent
powder (GloGerm, Hygienic Solutions, Lincoln, UK) fre-
quently used in studies on contamination [12, 13]. These
5 µm particles are the same size as bacteria and appear
orange under UV light (Fig. 1). All experiments were car-
ried out at the Vesalius institute of the Catholic Univer-
sity of Leuven, Belgium. There is considerable variation in practice between
well-established
musculoskeletal
infection
centers
regarding the steps just before the initiation of the clean
phase. Some will take no additional measures, i.e. after
debridement and lavage the clean phase begins, while
others will close the wound, get rescrubbed, re-prep and
re-drape the surgical field, and use new instruments to
re-implant the necessary parts and finish the operation
[6–9]. The latter is of course accompanied with extra
time, effort and costs and there are no studies to support
or negate its superiority. Therefore, most surgical teams
will only take some additional measures, e.g. change Cadaveric specimens Nine fresh frozen cadaveric hips (5 left, 4 female, mean
age 85 years SD 8.5, mean weight 73 kg SD 17) were used. Fig. 1 The appearance of orange fluorescent powder (GloGerm, Hygienic Solutions, Lincoln, UK) under normal daylight and under UV light. No
additional measures protocol (bacterial load of 3 × 2,5 mg GloGerm powder) Fig. 1 The appearance of orange fluorescent powder (GloGerm, Hygienic Solutions, Lincoln, UK) under normal daylight an
additional measures protocol (bacterial load of 3 × 2,5 mg GloGerm powder) Page 3 of 9 Vles et al. Journal of Experimental Orthopaedics (2022) 9:28 Vles et al. Journal of Experimental Orthopaedics Description of experiment followed by explantation of the prosthesis and subse-
quent pulsatile lavage with 6 L of saline. At this point ran-
domization determined which clean phase protocol was
followed, i.e. no, some or full additional measures (Fig. 2). Finally, a new prosthesis was implanted. First, a PJI of a THA had to be simulated (Fig. 2). The
cadaver was placed supine and a primary THA was
undertaken via a direct anterior approach. Right before
implantation of the components 1, 2.5, or 5 cc of orange
fluorescent powder was placed in the already prepared
acetabulum and an equal volume inside the femur. Next,
the components were implanted, the joint was relocated
and another 1, 2.5 or 5 cc of powder was distributed
around the implant in the soft tissues. After full closure,
the skin of the cadaveric specimen was checked for con-
tamination with fluorescent powder using a UV light
source. If this was the case, the specimen was washed
until clean. At the end of the experiment 9 one-step exchanges had
been performed, with 3 levels of bacterial load (1, 2.5 and
5 cc of fluorescent power behind the cup, along the stem,
and in the soft tissues) and using 3 different clean phase
protocols. Variables and outcome measuresf In order to determine the effect of different clean
phase protocols on contamination of the sterile field
the room was made pitch-dark and standardized UV
light-enhanced photographs (Fujifilm X-T30 camera,
focus 8, shutter speed 2, ISO 4000, diaphragm 3.5–4)
were obtained of 1) the gloves of the operating team, 2)
the instrument table, 3) the drapes, and 4) the wound. Blinded photographs were sent to the members of the
Belgian Hip Society who were asked to rank them from
most clean to most contaminated (www.survio.com Second, a one-step exchange was simulated and car-
ried out using all precautions one would normally take. The cadaver was placed in lateral decubitus using pelvic
supports. The operating team consisted of a lead surgeon
(SG), an assisting surgeon (JB), a scrub nurse (GV) and
a circulating nurse (PD). After routine prepping (chlo-
rhexidine/alcohol solution) and draping, a posterolateral
approach was performed (Fig. 3). Debridement of clearly
infected (orange) tissue was carried out from the start, Fig. 2 Simulating an infected Total Hip Arthroplasty Fig. 2 Simulating an infected Total Hip Arthroplasty Fig. 2 Vles et al. Journal of Experimental Orthopaedics (2022) 9:28 Page 4 of 9 Fig. 3 One-step exchange using a posterolateral approach. Retracting the external rotators and posterior capsule reveals the infected THA. S = superior, A = anterior, H = head of the implant, * = orange GloGerm powder mimicking the bacterial load Fig. 3 One-step exchange using a posterolateral approach. Retracting the external rotators and posterior capsule reveals the infected THA. S = superior, A = anterior, H = head of the implant, * = orange GloGerm powder mimicking the bacterial load – online survey tool). Thirty-eight out of 152 (28%) mem-
bers responded. clean phase protocols amongst themselves. Given the
presence of three groups, Bonferroni correction applied
and therefore significance was set at p ≤ 0.017. Effect
sizes (r = n cases / √Z) are furthermore reported. In order to determine whether or not it is possible to
re-implant the prosthesis completely clean, the implant
was taken out again at the end of the one-step exchange
and inspected for contamination under UV light. Statistical analysis Statistical analysis (Table 1) showed that the differences
in ranking of cleanliness (between the 9 experiments,
but also between the 3 clean phase protocols) were sig-
nificant (p < 0.000). Rater agreement ranged from moder-
ate (0.360) for drapes to very good (0.808) for gloves. The
gloves, the instrument table, the drapes and the wound
were all considered significantly cleaner after a clean
phase using full additional measures than after clean
phases using partial or no additional measures (p < 0.000,
medium to large effect sizes). Friedman’s test was used to explore differences in rank-
ing of cleanliness of the photographs of the nine experi-
ments. In order to assess agreement among raters
Kendall´s coefficient of concordance was used. If signifi-
cant differences were found, post hoc testing was per-
formed, again using Friedman´s test (Kendall´s W test),
to explore differences between the 3 clean phase proto-
cols. If significant differences were found, post hoc Wil-
coxon signed rank testing was performed, to compare Table 1 Statistical analysis and comparison of the results of the three different clean phase protocols
* indicates statistical significance
Friedman
test (9
groups)
Kendall´s W
test
Friedman
test (3
clean phase
protocols)
Kendall´s W
test
Post hoc Wilcoxon signed rank testing
Most clean
↔
(significance)
(effect size)
Intermediate
clean
↔
(significance)
(effect size)
Most
contaminated
Gloves
p < 0.000
0.837
p < 0.000
0.808
Full
p < 0.000*
r = 0.868
Partial
p < 0.000*
r = 0.509
None
Table
p < 0.000
0.762
p < 0.000
0.661
Full
p < 0.000*
r = 0.815
None
p = 0.935
Partial
Drapes
p < 0.000
0.422
p < 0.000
0.360
Full
p < 0.000*
r = 0.786
Partial
p = 0.093
None
Wound
p < 0.000
0.810
p < 0.000
0.474
Full
p < 0.000*
r = 0.632
Partial
p < 0.000*
r = 0.502
None Table 1 Statistical analysis and comparison of the results of the three different clean phase protocols Vles et al. Journal of Experimental Orthopaedics (2022) 9:28 Page 5 of 9 Vles et al. Journal of Experimental Orthopaedics (2022) 9:28 Vles et al. Journal of Experimental Orthopaedics (2022) 9:28 Vles et al. Journal of Experimental Orthopaedics (2022) 9:28 Vles et al. Journal of Experimental Orthopaedics Changing gloves during the one-step exchange (clean
phases with full and partial additional measures) led to
cleaner gloves at the end of the procedure (Fig. 4). Statistical analysis How-
ever, gloves remained significant cleaner (p < 0.000) dur-
ing a one-step exchange using a clean phase with full
additional measures, when compared to a clean phase
with partial additional measures. Stated in a different
way, fresh gloves will – for a degree – become contami-
nated again, if the sterile field is not completely renewed.h measures was rated as significantly cleaner than proce-
dures making use of clean phases with partial of no addi-
tional measures. No significant differences were observed
when comparing the latter two protocols amongst them-
selves. Stated in a different way, the contamination of the
instrument table occurs during the dirty phase and par-
tial clean phase measures do not seem to influence its
contamination at the end of the procedure. Similar observations were made for the drapes
(Fig. 6). Drapes were most clean after a clean phase
with full additional measures, while no significant i
The instrument table (Fig. 5) at the end of a one-
step exchange using a clean phase with full additional Fig. 4 Gloves of the surgical team at the end of the one-step exchange for three different clean phase protocols (bacterial load of 3 × 2,5 mg
GloGerm powder) Fig. 4 Gloves of the surgical team at the end of the one-step exchange for three different clean phase protocols (bacterial load of 3 × 2,5 mg
GloGerm powder) Fig. 5 Remaining instrument table at the end of the one-step exchange for three different clean phase protocols (bacterial load 3 × 2,5 mg of
GloGerm powder) Vles et al. Journal of Experimental Orthopaedics Vles et al. Journal of Experimental Orthopaedics (2022) 9:28 Page 6 of 9 Fig. 6 Drapes at the end of the one-step exchange for three different clean phase protocols (bacterial load 3 × 2,5 mg of GloGerm powder) not possible to re-implant the new prosthesis completely
clean, even from a macroscopic point of view. differences were found when comparing the other two
protocols. Stated in a different way, only changing the
U-drape does not influence the perception of contami-
nation of the drapes, while changing all drapes does. Of
note, the stockinette was often the most contaminated
drape.h Other relevant findings?h The extra costs associated with a partial and a full clean
phase for our hospital and for a surgical team of three are
€ 10.86 and € 120.30, respectively (supplement 1). The wound (Fig. 7) after a clean phase with full addi-
tional measures was significantly cleaner than the wound
after a clean phase with partial additional measures,
which was in turn significantly cleaner than the wound
after a clean phase with no additional measures. Is it possible to re‑implant the prosthesis completely
clean? Discussion Treatment of PJI remains challenging and it is essential
to get the basics right. This cadaveric study therefore
investigated the effect of different clean phase protocols
on contamination of the sterile field during one-step
exchanges of infected implants. Although it was shown
that a more extensive clean phase helps to significantly
reduce contamination of the sterile field before re-
implantation, some contamination of the new implant
appears inevitable.h Without exceptions, all re-implanted joint replace-
ments turned out to be contaminated with some orange
GloGerm powder at the end of the one-step exchange
regardless of the followed clean phase protocol or the
amount of bacterial load at the beginning of the surgical
procedure (Fig. 8). Therefore, in our experiments, it was This experimental cadaveric study has its limita-
tions. Firstly, the absence of bleeding tissue might have Fig. 7 Surgical wound at the end of the one-step exchange for three different clean phase protocols (bacterial load 3 × 2,5 mg of GloGerm
powder) Fig. 7 Surgical wound at the end of the one-step exchange for three different clean phase protocols (bacterial load 3 × 2,
powder) Vles et al. Journal of Experimental Orthopaedics (2022) 9:28 Vles et al. Journal of Experimental Orthopaedics Page 7 of 9 Fig. 8 The newly implanted THA at the end of the one-step exchange for three different clean phase protocols (bacterial load 3 × 2,5 mg of
GloGerm powder) of the contamination of the gloves and the wound, but
had no effect on the drapes and the instrument table. Although, the extra costs (€ 120.30) and extra operat-
ing time (approximately 10 min) spent on a full clean
phase seem manageable, perhaps the biggest concern
is the added layer of complexity to an already com-
plex procedure. Many additional actions are required,
which can cause disruption of sterility themselves and
can lead to increased turbulence in the operating room. The extent of the clean phase therefore is a trade-off
between theoretical bacterial load reduction and what
is feasible in daily clinical practice. It is our opinion
that in high-volume musculoskeletal infection centres
with well-defined protocols and well-trained staff, a full
clean phase should perhaps be preferred over a partial
clean phase, as the benefits of the latter are limited. In centres where septic surgeries are only occasionally
performed, some clean phase measures should still be
possible without disrupting the surgical flow too much,
e.g. Is it possible to re‑implant the prosthesis completely
clean? In our experiments, we were unable to re-implant the
new prosthesis entirely clean, even when minimal bac-
terial load was present from the beginning and a full
clean phase was performed. This observation under-
lines the importance of adequate tissue levels of appro-
priate antibiotics at the time of implantation (which
should therefore be given at induction and not with-
held until cultures have been obtained) and might also Discussion providing a new U-drape and stockinette, chang-
ing the tip of the diathermy and suction, and the entire
team changing gloves. influenced operating time, timing and number of sur-
gical actions, and distribution of fluorescent powder
around the wound and drapes. It is also possible that the
fluorescent powder used is more adherent than bacteria
and therefore more difficult to eradicate, which could
have led to a visual overrepresentation of contamina-
tion. Secondly, there was no laminar air flow on the site
of the experiment, although its relevance is questionable
[14]. Thirdly, we used visual interpretation of the degree
of contamination, rather than a quantitative evaluation. It must be stated that these ratings were done by experi-
enced hip surgeons with generally high inter rater agree-
ment and large effect sizes. Fourthly, one could argue that
antiseptics would kill bacteria, and therefore not all con-
tamination observed is still relevant. Lavage with antisep-
tics however is still controversial with opponents being
concerned about cytotoxicity and bacterial rebound. Finally, although it was illustrated that a full clean
phase leads to the highest level of bacterial load reduc-
tion, definitive recommendations for clinical practice
cannot be drawn from this experimental setup. Nev-
ertheless, considering the difficulties associated with
research in the field of septic surgery, performing a
clinical RCT on this topic will be extremely difficult and
in light of the results above, perhaps even unethical. What is the effect of different clean phase protocols
on the sterile field? It was illustrated clearly that the gloves, the instrument
table, the drapes and the wound after a clean phase with
full additional measures are significantly cleaner than
those after a clean phase with partial or no additional
measures. Partial measures were able to reduce some Vles et al. Journal of Experimental Orthopaedics (2022) 9:28 Vles et al. Journal of Experimental Orthopaedics Page 8 of 9 Availability of data and materials
Upon request. Availability of data and materials
Upon request. provide a rationale for the use of antibiotic-loaded
cemented implants during one-step exchange [15, 16]. g
g
The usage of fluorescent GloGerm powder allowed
us to get instant feedback on our actions during a one-
step exchange of the hip and therefore this set-up might
also be used as a training tool for surgeons treating
musculoskeletal infections. Several observations were
of particular interest to us. While it is common prac-
tice for the lead surgeon to change gloves at certain
stages during the operation, we noticed that the gloves
of the scrub nurse were often the most contaminated. Another custom in centers where they perform a par-
tial clean phase, is to provide a new U-drape. However,
we noticed that the stockinette surrounding the leg is
often equally or more contaminated. It would there-
fore make sense to change this as well since the leg is
manipulated repeatedly during the clean phase for
exposure and testing of stability and leg length. References 1. Izakovicova P, Borens O, Trampuz A (2019) Periprosthetic joint infection:
current concepts and outlook. EFORT Open Rev 2(4):482 1. Izakovicova P, Borens O, Trampuz A (2019) Periprosthetic joint infection:
current concepts and outlook. EFORT Open Rev 2(4):482 Finally, although inspection of the explanted prosthe-
sis might reveal useful information on sizes, wear, and
impingement, it should be regarded as the most con-
taminated object of the entire procedure and therefore
handled accordingly. 2. Kapadia BH, Berg RA, Daley JA, Fritz J, Bhave A, Mont MA (2016) Peripros-
thetic joint infection. Lancet 387:386–394 3. Parisi T, Konopka J, Bedair H (2017) What is the Long-term Economic
Societal Effect of Periprosthetic Infections After THA? A Markov Analysis. Clin Orthop Relat Res 475:1891–1900 4. Wildeman P, Rolfson O, Söderquist B, Wretenberg P, Lindgren V (2021)
What Are the Long-term Outcomes of Mortality, Quality of Life, and Hip
Function after Prosthetic Joint Infection of the Hip? A 10-year Follow-up
from Sweden. Clin Orthop Relat Res 479:2203–2213 Received: 1 February 2022 Accepted: 11 March 2022 Received: 1 February 2022 Accepted: 11 March 2022 Conclusionh This study investigated the effect of different clean phase
protocols on contamination of the sterile field during
one-step exchanges of infected THAs on cadavers. It was
found that a more extensive clean phase helps to signifi-
cantly reduce contamination before re-implantation. We
acknowledge that some contamination of the new implant
appears inevitable and that the extent of a clean phase
therefore is a trade-off between theoretical load reduction
and what is feasible in daily clinical practice. However,
we advise surgical teams in high volume musculoskel-
etal infection centres to try and incorporate a clean phase
with full additional measures into their work flow, as par-
tial measures seem to be a poor compromise. A compre-
hensive registry on septic surgeries could help to better
understand the clinical relevance of a clean phase. 5. Stragier B, Renard A, Vanlaer L, Verhaegen J, Neyt J (2018) The importance
of cleanliness in periprosthetic joint infection surgery: a comparitve reviw
of the introduction of the ’clean phase’ concept. Available via Orthopae-
dic Proceedings. https://online.boneandjoint.org.uk/doi/abs/ https://doi.
org/10.1302/1358-992X.2017.22.087. Accessed 30 Oct 2021 6. Baker RP, Tafin UF, Borens O (2015) Patient-Adapted Treatment for Pros-
thetic Hip Joint Infection. Hip Int 25:316–322 7. Choo K, Austin M, Parvizi J (2019) Irrigation and Debridement, Modular
Exchange, and Implant Retention for Acute Periprosthetic Infection After
Total Knee Arthroplasty. JBJS Essent Surg Tech 9:e38 8. George DA, Konan S, Haddad FS (2015) Single-Stage Hip and
Knee Exchange for Periprosthetic Joint Infection. J Arthroplasty
30:2264–2270 8. George DA, Konan S, Haddad FS (2015) Single-Stage Hip and
Knee Exchange for Periprosthetic Joint Infection. J Arthroplasty
30:2264–2270 9. Zahar A, Klaber I, Gerken A-M, Gehrke T, Gebauer M, Lausmann C, Citak
M (2019) Ten-Year Results Following One-Stage Septic Hip Exchange
in the Management of Periprosthetic Joint Infection. J Arthroplasty
34:1221–1226 9. Zahar A, Klaber I, Gerken A-M, Gehrke T, Gebauer M, Lausmann C, Citak
M (2019) Ten-Year Results Following One-Stage Septic Hip Exchange
in the Management of Periprosthetic Joint Infection. J Arthroplasty
34:1221–1226 10. Abdelaziz H, Zahar A, Lausmann C, Gehrke T, Fickenscher H, Suero
E, Gebauer M, Citak M (2018) High bacterial contamination rate of
electrocautery tips during total hip and knee arthroplasty. Int Orthop
42:755–760 10. Abdelaziz H, Zahar A, Lausmann C, Gehrke T, Fickenscher H, Suero
E, Gebauer M, Citak M (2018) High bacterial contamination rate of
electrocautery tips during total hip and knee arthroplasty. Int Orthop
42:755–760 Ethics approval and consent to participate Ethics approval and consent to participate
This cadaveric study was approved by our ethical review board and registered
at the Belgian National Council for Bioethics (NH019 2021–6-01). Supplementary Information 11. Robinson AH, Drew S, Anderson J, Bentley G, Ridgway GL (1993) Suction
tip contamination in the ultraclean-air operating theatre. Ann R Coll Surg
Engl 75:254 11. Robinson AH, Drew S, Anderson J, Bentley G, Ridgway GL (1993) Suction
tip contamination in the ultraclean-air operating theatre. Ann R Coll Surg
Engl 75:254 The online version contains supplementary material available at https://doi.
org/10.1186/s40634-022-00467-x. The online version contains supplementary material available at https://doi.
org/10.1186/s40634-022-00467-x. 12. Fraser JF, Young SW, Valentine KA, Probst NE, Spangehl MJ (2015) The
Gown-glove Interface Is a Source of Contamination: A Comparative
Study. Clin Orthop Relat Res 473:2291–2297 12. Fraser JF, Young SW, Valentine KA, Probst NE, Spangehl MJ (2015) The
Gown-glove Interface Is a Source of Contamination: A Comparative
Study. Clin Orthop Relat Res 473:2291–2297 Additional file 1. Supplement 1. y
13. Vermeiren A, Verheyden M, Verheyden F (2020) Do Double-fan Surgical
Helmet Systems Result in Less Gown-particle Contamination Than Single-
fan Designs? Clin Orthop Relat Res 478:1359–1365 13. Vermeiren A, Verheyden M, Verheyden F (2020) Do Double-fan Surgical
Helmet Systems Result in Less Gown-particle Contamination Than Single-
fan Designs? Clin Orthop Relat Res 478:1359–1365 Authors’ contributions Authors contributions
All authors contributed significantly to this study / paper and approved this
final version submitted to the Journal of Experimental Orthopaedics. All
author(s) read and approved the final manuscript. 14. Bischoff P, Kubilay NZ, Allegranzi B, Egger M, Gastmeier P (2017) Effect of
laminar airflow ventilation on surgical site infections: a systematic review
and meta-analysis. Lancet Infect Dis 17:553–561 14. Bischoff P, Kubilay NZ, Allegranzi B, Egger M, Gastmeier P (2017) Effect of
laminar airflow ventilation on surgical site infections: a systematic review
and meta-analysis. Lancet Infect Dis 17:553–561 15. Gehrke T, Zahar A, Kendoff D (2013) One-stage exchange: it all began
here. Bone Joint J 95(B):77–83 15. Gehrke T, Zahar A, Kendoff D (2013) One-stage exchange: it all began
here. Bone Joint J 95(B):77–83 Author details
1 1 Department of Development and Regeneration, Faculty of Medicine, Institute
for Orthopaedic Research and Training (IORT), Leuven, KU, Belgium. 2 Division
of Orthopaedics, University Hospitals Leuven, Leuven, Belgium. Received: 1 February 2022 Accepted: 11 March 2022 Funding
l Not applicable. Page 9 of 9 Vles et al. Journal of Experimental Orthopaedics (2022) 9:28 Vles et al. Journal of Experimental Orthopaedics (2022) 9:28 Vles et al. Journal of Experimental Orthopaedics (2022) 9:28 16. Wouthuyzen-Bakker M, Benito N, Soriano A (2017) The Effect of Pre-
operative Antimicrobial Prophylaxis on Intraoperative Culture Results
in Patients with a Suspected or Confirmed Prosthetic Joint Infection: a
Systematic Review. J Clin Microbiol 55:2765–2774 16. Wouthuyzen-Bakker M, Benito N, Soriano A (2017) The Effect of Pre-
operative Antimicrobial Prophylaxis on Intraoperative Culture Results
in Patients with a Suspected or Confirmed Prosthetic Joint Infection: a
Systematic Review. J Clin Microbiol 55:2765–2774 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations.
|
https://openalex.org/W3116604898
|
https://www.researchsquare.com/article/rs-42020/v1.pdf?c=1595467973000
|
English
| null |
Communication of Pharmacogenomic test results and treatment plans in pediatric oncology: deliberative stakeholder consultations with parents
|
BMC palliative care
| 2,021
|
cc-by
| 7,611
|
Research article License:
This work is licensed under a Creative Commons Attribution 4.0 International
License. Read Full License Version of Record: A version of this preprint was published on January 12th, 2021. See the published
version at https://doi.org/10.1186/s12904-021-00709-2. Page 1/16 Abstract Background: Effective communication in support of clinical decision making is central to the pediatric
cancer care experience for families. A new laboratory derived pharmacogenetic test (LDT) to diagnose
difficult-to-treat brain cancers has been developed to stratify children based on their ability to respond to
available treatment, but may hinder effective communication since it can potentially remove the option
for curative treatment in those children identified as non-responders, i.e. those with a catastrophic
diagnosis. Objective: We solicited the perspectives of parents of children with difficult-to-treat brain cancer on
communication preferences surrounding implementation of the LDT in standard care using deliberative
stakeholder consultations. Methods: Eight bereaved parents of children who succumbed to difficult-to-treat brain cancer, and four
parents of children currently undergoing treatment for similar cancers attended single deliberative
consultations. Parents discussed four questions about potential communication issues that may arise
with the LDT in practice. Of these, five parents attended a mixed consultation for the purpose of
developing concrete recommendations for implementation of the LDT. Results: Explaining the risks, benefits, and accuracy of the LDT were considered essential to parents. Once an LDT-based diagnosis/prognosis can be made, parents valued honesty, empathy, and clarity in
communication. Parents also requested that all results and treatment options be presented to them in
measured doses, and in an unbiased manner over the course of several meetings. This communication
strategy allowed sufficient time to understand and accept the diagnosis/prognosis, particularly if it was
catastrophic. Continuous access to the appropriate psychological and social support or counselling at
and post-diagnosis was also strongly recommended. Conclusions: Deliberants co-created family-centered recommendations surrounding communication
issues of the LDT, providing guidance to pediatric oncologists implementing the test in practice. Introduction Effective communication in pediatric oncology is a crucial component of the overall care experience and
fostering a supportive, honest, and trusting environment for shared decision making [1]. Nevertheless,
delivering bad news, whether it is the diagnosis of cancer itself or test results indicating poor prognosis, is
one of the most challenging aspects of practicing oncology, particularly since communication skills are
not systematically taught as part of most standard residency training or continuing medical education
programs [1, 2]. In the pediatric oncology setting, distinct issues arise in the communication between
physician, parent, and child. Despite fighting their own emotions and distress regarding the unnatural
threat to a child’s life [3], the clinical team must not only determine the best manner in which to deliver
bad news to the parents and/or caregivers as well as the child, but also navigate the complex tripartite Page 2/16 Page 2/16 decision-making process, where applicable [1]. Moreover, oncologists must achieve a delicate balance
when disclosing poor, yet honest and clear prognoses, while also taking care not to remove all hope [4]. With the growing implementation of pharmacogenomic tests that can accurately stratify children with
cancer by expected treatment outcome and risk, effective communication, and shared decision-making
are increasingly difficult to navigate for pediatric oncologists. For example, a laboratory-derived blood test
that can accurately identify fatal genetic mutations in ~ 20% children with difficult-to-treat brain cancers
(i.e. high-grade glioblastomas, HGGs) was being considered for implementation as a standard of care [5,
6]. For children carrying these fatal mutations, palliative care is introduced at diagnosis because these
tumors are treatment-refractory. Nevertheless, treatments known to be ineffective are frequently
administered to children despite serious side-effects that reduce quality of life without clinical benefit. The clinical decision to pursue comfort, rather than curative therapies for children with fatal HGG
diagnoses, herein referred to as a ‘catastrophic diagnosis,’ is distressing for both the pediatric oncologist
and the family. Emotional coping strategies, such as anger and disbelief at the limited treatment options
may hinder effective communication and shared decision-making in these devastating cases. In this study, we used deliberative stakeholder consultations to solicit the perspectives of parents of
children with HGG on the use of this potential LDT test. We explored what information parents deem
necessary to understand the nature of the diagnosis, how results should be communicated, and their
preferences with regard to the available therapy options. Research Design and Recruitment We conducted deliberative consultations with two stakeholder groups and future users of the LDT: 1)
bereaved parents of children diagnosed with HGG, and 2) parents of children currently undergoing
treatment for brain cancer, including but not limited to HGG. Parents were identified and purposefully
recruited via email with the help of Meagan’s Walk, a community-based organization supporting families
affected by childhood brain cancer located in Toronto, Ontario. This study was approved by McGill
University’s Institutional Review Board. Introduction Our goal was to develop recommendations for
communicating genetic test results and treatment options to parents, with an emphasis on breaking the
news of a catastrophic diagnosis, providing guidance to pediatric oncologists when these challenging
scenarios arise. Informing Parents: details regarding the genetic test for
HGA treatment decision-making At the beginning of each consultation, the facilitator (GB) explained the goals of the deliberative process,
i.e. to propose and raise competing arguments and solicit diverse opinions from interested stakeholders,
to parents in this regard. The expert also obtained informed consent at the start of each deliberation to
record the sessions. GB then provided a lay summary of the following regarding the non-invasive LDT : 1) Page 3/16 the accuracy by which the test identifies fatal mutations from biomarkers in blood (and is therefore less
invasive than a brain tumor biopsy); 2) the survival statistics of those children carrying the fatal
mutations (e.g. H3.3 K27M and H3.1 K27M) despite having received conventional treatment (i.e. chemotherapy) based on previous research; 3) the logistics of the genetic test (e.g. the test will be part of
the initial diagnostic workup and rapid turnaround time for results); 4) the proposed course of action
based on the test results; and 5) benefit-risk considerations relating to chemotherapy and/or radiation as
well as its potential impact on the child’s quality of life. Consultation Design and Data Collection Our consultation design has been previously described in detail elsewhere [7]. Briefly, bereaved parents
and those with children currently undergoing treatment attended a 2 to 3 hour small-group preliminary
session (Phase I), separately, where they first listened to a lay presentation of the policy issue by the
expert and were then given the opportunity to ask questions or clarifications. Following this initial
information session, deliberants were then asked to discuss their opinions and preferences relating to
implementation of the LDT in standard care. Four questions guided this discussion and are presented in
Table 1. The expert only intervened to answer clarifying questions about the LDT once the deliberative
process began so as to minimize undue influence on participant opinions and preferences. In Phase II, both bereaved parents and those of children undergoing treatment attended a mixed
consultation. The introductory process as preliminary small-group sessions were repeated, changing only
the topics for deliberation inspired by findings from the previous separate sessions. The goal of the
mixed consultation was to develop family-centered recommendations regarding the implementation and
communication of the LDT results in the pediatric oncology care context, based on consensus areas
corroborated by the deliberations. All consultations were audio recorded and transcribed verbatim. Authors CL and VR produced detailed reports on the interactions among participants as well as points of
deliberative agreement and disagreement for each phase. All authors discussed and refined key themes
that emerged from qualitative analyses of the transcripts. Data Analysis and Deliberative Outputs The goal of the deliberations was to provide family-centered recommendations to the pediatric oncology
multidisciplinary health team regarding how best to approach the communication of the genetic test
results in the scenario of a catastrophic diagnosis. After thematically analyzing the transcripts and notes
(CL and VR), a summary highlighting key issues and conclusive collective statements made by the
parents, outlining agreement, disagreement, or a recommendation for future course of action, were
considered to be the primary outputs in each deliberative session. Summary documents were submitted
to, and ratified by participants following each session. Deliberation was achieved in each session,
determined through the evaluation criteria described by De Vries and colleagues (Online Supplement)[8]. Page 4/16 Table 1
Consultations with parents and deliberation questions
Deliberation Questions Table 1
Consultations with parents and deliberation questions
Consultation
Group
Deliberation Questions
Phase I
Small-group
consultations
1) Bereaved
parents
2) Parents of
children
undergoing
treatment
1. What information do you think families would need to understand the results
of the genomic test and the implications for treatment? 2. What concerns/problems do you think families might have in accepting
treatment plans based on the genomic test? 3. What members of the health care team do you think need would need to be
involved in discussing the test results with children and families? i.e. the
oncologist, the palliative care physician, etc.? 4. Are there any other issues you can think of that may be a problem when this
test was part of regular clinical care for children with brain tumors? Phase II
Mixed
consultation
- Bereaved parents
and parents of
children
undergoing
treatment
1. What and how do you think test results and treatment options should be
communicated to the parents by the health care team in the initial and
subsequent follow-up meetings (if different)? 2. Who do you think should be involved in the initial diagnostic consultation
and subsequent meetings about potential treatment options, i.e. clinical and
psychosocial support? 3. Would your recommendations from Q1) and Q2) change if the test results
indicated a de-escalation of care/therapy? What margin of error do you think
would make you accept the test-based diagnosis as “fact"? In other words, how
certain is certain? 1. What and how do you think test results and treatment options should be
communicated to the parents by the health care team in the initial and
subsequent follow-up meetings (if different)? 2. Data Analysis and Deliberative Outputs Who do you think should be involved in the initial diagnostic consultation
and subsequent meetings about potential treatment options, i.e. clinical and
psychosocial support? 3. Would your recommendations from Q1) and Q2) change if the test results
indicated a de-escalation of care/therapy? What margin of error do you think
would make you accept the test-based diagnosis as “fact"? In other words, how
certain is certain? Overview of discussions and deliberants In the small-group consultations, 8 were parents who were bereaved and 4 were parents of children
currently undergoing treatment, while 5 parents participated in the mixed group consultation. In the
bereaved parents consultation, the discussions centered around important physician-parent
communication values, such as honesty and empathy, providing families with the appropriate
psychological and social support or counselling at and post-diagnosis, as well as providing parents with
unbiased information to enable shared decision-making of the child’s treatment trajectory. In addition to
these themes, the parents of children currently undergoing treatment deliberated on communication
strategy details, i.e. the structure and content of physician-family meetings, as well as the inclusion of the
child in such conversations. Lastly, the mixed consultation led to concrete family-centered
recommendations for LDT-based communication in practice. 16
Small-Group Consultation: Bereaved Parents Page 5/16
Small-Group Consultation: Bereaved Parents Page 5/16 Page 5/16 Parents first discussed how the LDT would be introduced at initial presentation. The accuracy, risks,
benefits (non-invasiveness compared to biopsy), turn-around time, and implications of the test
concerning response to treatment were seen as key communication needs. At initial presentation, parents
also expressed frustration with the health care team when asked questions that implied their parenting
skills or the environment in which they lived were the potential causes of the brain tumor: “Did [your child] ever hit his head?” (BP3) Once a diagnosis and prognosis can be communicated to the family, parents suggested that oncologists
“be human” when delivering the news, expecting a balance of honesty, compassion, and empathy. One
participant remembers: “I asked if [my child] was going to die, and [the surgeon] looked at me and said, well everyone is going to
die someday.” (BP1) “I asked if [my child] was going to die, and [the surgeon] looked at me and said, well everyone is going to
die someday.” (BP1) Participants unanimously agreed that honesty was appreciated when delivering the diagnosis, even if it
was extremely difficult news to cope with. This honesty, participants said, would allow parents to have an
open conversation about the potential next steps, i.e. treatment options, with their physician. Participants unanimously agreed that honesty was appreciated when delivering the diagnosis, even if it
was extremely difficult news to cope with. This honesty, participants said, would allow parents to have an
open conversation about the potential next steps, i.e. treatment options, with their physician. “It’s that human side of the doctors, and still that professional side of having the deliver the information.”
(BP1) (BP1) Participants also acknowledged how challenging it must be for the oncologists, since they are given the
difficult task of delivering the results and prognosis in a realistic and honest manner, without giving false
hope to the parents: “But thank goodness, most of them are honest, and that’s what we needed. And I think that’s important
for physicians and oncologists to know that.” (BP8) In the event that the genetic test suggests the tumor would not respond to any available therapies,
parents first discussed what type of information should be presented when disclosing a catastrophic
diagnosis and the support the care team should have at the initial diagnostic consultation, including from
the palliative care team. First, they highlighted that physicians should expect parents to request a second
and/or third medical opinion to reduce uncertainty of the prognosis, regardless of the expected accuracy
of the test. Second, participants agreed that physicians should provide supporting evidence for the
diagnosis from other tests such as an MRI such to help parents better understand and accept their child’s
prognosis. Finally, participants recommended presenting statistics of the tumor type and why this
specific tumor would not respond to chemotherapy to parents during the initial diagnostic consultation: “If you have more pieces [of information] that you can put together later when you’re thinking about the
[encounter] then it’s a little bit easier.” (BP5) Participants agreed that this may cause
parents to feel as though they were to blame for their child’s tumor and that their other children may be at
risk of acquiring this deadly form of cancer: “If you can test for this, then I want my other children tested… because if it’s in the DNA, they would see
that as a whole family thing as I can imagine […] I would just sit there and think… is this going to happen
to my other child?” (BP6) Participants agreed that it would also be important to provide psychological support to the siblings of the
diagnosed child, as they may worry about eventually developing the disease. Participants agreed that it would also be important to provide psychological support to the siblings of the
diagnosed child, as they may worry about eventually developing the disease. “I’m sure it’s in the back of [my other son’s] mind… what if this is genetic and I have it?” (BP4) “I’m sure it’s in the back of [my other son’s] mind… what if this is genetic and I have it?” (BP4) Whether and when the palliative care team should be introduced was a subject of deliberation amongst
bereaved parents. Initially, parents disagreed about the timing of the introduction; some felt that the
palliative care team should be present at the initial diagnostic encounter, while others felt overwhelmed
by the ‘mob of physicians’: “It’s almost like they were all there to protect themselves from us freaking out on them. I don’t know. I felt
a little up against the wall with so many people being there delivering that news.”(BP8) “It’s almost like they were all there to protect themselves from us freaking out on them. I don’t know. I felt
a little up against the wall with so many people being there delivering that news.”(BP8) After further discussion, participants agreed that the palliative care team and other health care
professionals should be introduced to the family early on, but not at the initial diagnostic consultation. After further discussion, participants agreed that the palliative care team and other health care
professionals should be introduced to the family early on, but not at the initial diagnostic consultation. “If you have more pieces [of information] that you can put together later when you’re thinking about the
[encounter] then it’s a little bit easier.” (BP5) “If you have more pieces [of information] that you can put together later when you’re thinking about the
[encounter] then it’s a little bit easier.” (BP5) “If you have more pieces [of information] that you can put together later when you’re thinking about the
[encounter] then it’s a little bit easier.” (BP5) Page 6/16 With regard to support, participants also recommended having an assigned social worker, bereaved
parent, or liaison person in the room at the initial diagnosis who can relate to what the parents are going
through as well as help them navigate the process: With regard to support, participants also recommended having an assigned social worker, bereaved
parent, or liaison person in the room at the initial diagnosis who can relate to what the parents are going
through as well as help them navigate the process: “You need someone who knows what has been happening […] almost like a bereaved parent, who can
guide you through it.” (BP8) Participants also suggested having someone who can take notes during the initial diagnostic encounter
with the treating physician, summarizing important points and outlining the potential treatment options: Participants also suggested having someone who can take notes during the initial diagnostic encounter
with the treating physician, summarizing important points and outlining the potential treatment options: “Maybe it’s the social worker or someone else who comes in to take notes […] because you’re listening but
you’re not really listening. As soon as the diagnosis comes in, it’s like bla bla bla, the teacher from Charlie
Brown.” (BP1) “Maybe it’s the social worker or someone else who comes in to take notes […] because you’re listening but
you’re not really listening. As soon as the diagnosis comes in, it’s like bla bla bla, the teacher from Charlie
Brown.” (BP1) Participants were also concerned about the anxiety and worry that may result from the association of the
word “DNA” or “genomic” to the inheritance of the brain tumor. “If you have more pieces [of information] that you can put together later when you’re thinking about the
[encounter] then it’s a little bit easier.” (BP5) Once the prognosis has been communicated, parents recommended that oncologists hold another
meeting with the family to discuss potential treatment options as well as the risks and benefits of each
option since it was widely felt “the more knowledge, the better.” Parents discussed how the options
should be presented, the burden of making the treatment decision, and how the LDT could improve the Page 7/16 quality of life of children with intractable brain tumors by not pursuing highly toxic therapies known to be
ineffective. Regardless of the test results, participants valued honesty from the health professional; it is this honesty
that parents perceived allows them to accept the dismal prognosis and eventually guide treatment
decisions that maximize their child’s quality of life. Some participants also highlighted how important it is
for the treating physician to be in the know, as many of the parents will base treatment decisions on their
physician’s recommendations and trust that they will provide them with accurate, honest, and unbiased
information to make this decision. “We needed someone to say to us that chances are, even if it does work, [your child] will probably not last
another year… So how do you want that year to go? A lot depends on what the physicians are telling you.”
(BP8) Participants agreed that the treating physicians should explain all possible treatment options, including
clinical trial enrolment, to the parents without persuading them toward one option or another. It should be
up to the parents and child, where appropriate, to make the final treatment decision: “It’s going to be tough for the physicians to give the right “answer” [regarding treatment options]. They
gave us both sides of the coin […] Ultimately, it’s a chance of a chance, but you as parents are the ones
that have to make that decision. And I think that’s tough for a lot of parents to hear. Even though they are
not the professional, they will still have to make that choice at some point. “If you have more pieces [of information] that you can put together later when you’re thinking about the
[encounter] then it’s a little bit easier.” (BP5) Our care was amazing […] it
was always that honest, up front, there isn’t hope but we’ll help you if you want to look at other things…
We’ll help you if you want to do these trials, but we’re also here to make [your child’s] life as comfortable
as possible for what we think is left of it.” (BP1) Lastly, parents also agreed the LDT has the potential to significantly improve the quality of life of children
with HGAs not responsive to chemotherapy. Non-invasive treatment, i.e. radiation, was viewed favorably
in terms of improving quality of life since the side effects were minimal and symptom management was
critical. In contrast, chemotherapy was seen as more invasive and many parents expressed their regret for
deciding to follow this treatment path having known retrospectively of its inefficacy in treating their
child’s specific tumor” “The worst part was chemotherapy. […] At that time you think you’re doing the right thing, because that’s
what they are telling you to do […] They’re sick, then they’re skinny, then they’re fat, then they’re skinny,
then they’re fat. Thinking back on it now, I wish I would have just taken him home and enjoyed my time
with him.” (BP3) Deliberative Consultation: Parents of Children Currently
Undergoing Treatment Page 8/16
As with the bereaved parents group, parents highlighted that the non-invasiveness and ability to confirm
the diagnosis in children with inoperable brain tumors with a quick turnaround time were key benefits of As with the bereaved parents group, parents highlighted that the non-invasiveness and ability to confirm
the diagnosis in children with inoperable brain tumors with a quick turnaround time were key benefits of Page 8/16 Page 8/16 the test. Parents then discussed how the information divulged in the first clinical encounter can be
overwhelming, and, similar to bereaved parents, suggested that physicians schedule multiple meetings
with the parents, where the information can be delivered in measured doses and important concepts can
be repeated at each meeting. The gap between successive meetings would allow the parents to process
the information and return to the next meeting with questions. One participant recalls: the test. Parents then discussed how the information divulged in the first clinical encounter can be
overwhelming, and, similar to bereaved parents, suggested that physicians schedule multiple meetings
with the parents, where the information can be delivered in measured doses and important concepts can
be repeated at each meeting. The gap between successive meetings would allow the parents to process
the information and return to the next meeting with questions. One participant recalls: “It took a while to register. It took at least 4 meetings with [our doctor] before it got into our heads that
there is nothing [they can do].” (PT4) “It took a while to register. It took at least 4 meetings with [our doctor] before it got into our heads that
there is nothing [they can do].” (PT4) Deliberants also suggested that treatment decisions should not be made at the initial diagnostic
encounter, given that parents typically have trouble understanding and/or accepting the diagnosis. Moreover, parents felt that oncologists use medical jargon during the initial diagnostic meeting, which
exacerbates the issue of understanding and/or accepting the diagnosis: “I had no idea what [the diagnosis] was […] I didn’t even know [my child] had cancer until they presented
us with this clinical trial and as my mom was reading the drugs, she’s like… well that’s chemotherapy. Deliberative Consultation: Parents of Children Currently
Undergoing Treatment Do you have any questions?” (PT1) With respect to delivering news about the failure of treatment, parents suggested that the care team
approach the delivery of such information in a coordinated way so as to avoid confusion: With respect to delivering news about the failure of treatment, parents suggested that the care team
approach the delivery of such information in a coordinated way so as to avoid confusion: “One doctor came in and said, we’ve got bad news, the drug didn’t work and you guys are going to go into
radiation. Then he left and another doctor comes in and says, oh we have wonderful news, we’re going to
start the radiation soon. One doctor is telling you one thing, and the other says wonderful news. Which
one is it?” (P4) Lastly, deliberants also valued trust building with their child’s treating physician as a key factor in the
acceptance of the LDT and an LDT-based diagnosis/prognosis. Deliberative Consultation: Parents of Children Currently
Undergoing Treatment And
then, I’m like… [my child] has cancer?” (PT3) “I had no idea what [the diagnosis] was […] I didn’t even know [my child] had cancer until they presented
us with this clinical trial and as my mom was reading the drugs, she’s like… well that’s chemotherapy. And
then, I’m like… [my child] has cancer?” (PT3) Parents agreed that physicians tend to ‘protect’ the parents form the truth by not saying the word ‘cancer’
or refraining from referring to the treatment as “chemotherapy.” Similar to the bereaved group, honesty
and empathy from the outset was recommended by parents: “I kept telling people, oh no, my son doesn’t have cancer. It’s just a low grade glioma. He doesn’t have
cancer […] and then the Canadian Cancer society sent us a letter saying [my child] is registered as a
cancer patient. Oh okay – maybe it’s cancer then […] but I haven’t heard [cancer] from a doctor. [...] The
[doctors] try to help us to cope, you know. But sometimes it’s better you give us some straight
information.” (P4) With respect to communicating the LDT-based diagnosis, as in the bereaved group, parents also
highlighted the likelihood of asking for a second opinion, particularly if the LDT revealed the tumor would
not respond to treatment and the parents had not built a trusting relationship with the treating oncologist. As with the bereaved group, deliberants highlighted how a negative results could take away hope for a
cure and this will be difficult for the parents to accept. Deliberants also suggested that physicians
acknowledge the uncertainty around the accuracy of the LDT, as no test is 100% accurate, and that many
parents will try new therapies, even if the child may not respond to them: “If your child is in the 1% that doesn’t respond to anything, you’re taking away their hope. And the human
spirit and hope is a huge factor. If you believe in something and have faith that something is going to
work, sometimes [it works].” (P1) Participants also discussed who should be present at the initial diagnostic meeting. Parents
recommended that the child not be present during the initial diagnostic meeting, as the parents would
need time to fully understand and accept the diagnosis before being able to explain it to their child. Mixed Consultation The mixed consultation ratified many of the themes highlighted in single dliberative consultations and
led to important family-centered communication recommendations for the implementation of the LDT in
clinical practice. Deliberative Consultation: Parents of Children Currently
Undergoing Treatment Social Page 9/16 workers, general practitioners or physicians who have built trust with the family, and other psychological
support personnel were deemed to be helpful as support during this initial encounter. However, parents
also mentioned that too many people in the room could also be intimidating: “The first time, we were just sitting in the waiting room, and this guy comes in… we’ve never seen him
before and says, your son has a brain tumor. That’s how we found out. I didn’t even know if he was a
doctor […] but then the second time, we had all these people and I thought… Oh no, it must be bad. Because the first time, it was the janitor who told me and now we have the social workers here so it must
be worse.” (PT2) In addition to providing psychosocial support during the initial encounter, parents also suggested that the
support should always be available regardless of whether the child is being diagnosed for the first time or
the parents are receiving news about the tumor not responding to treatment, or that the child relapsed
after initial treatment. In subsequent meetings following the initial diagnostic encounter, parents discussed when the child
should be included in treatment conversations with the oncologist. Deliberants recommended that the
oncologists initially speak to the parents first about the treatments and then have the parents choose
what information is relevant to the child, although still including the child in the decision-making, and to
avoid giving them information that may lead to unnecessary worry: “You have your discussion with the doctor and then bring [your child] in after and say,[…] we’ve been
discussing different [treatment] options and what we think is best for you is this. And what do you think? Do you have any questions?” (PT1) “You have your discussion with the doctor and then bring [your child] in after and say,[…] we’ve been
discussing different [treatment] options and what we think is best for you is this. And what do you think? Discussion Page 10/16 The single and mixed deliberative consultations with bereaved parents and those undergoing current
treatment for difficult-to-treat brain cancer highlighted communication needs between the clinical care
team and families, and led to six concrete recommendations for LDT-based communication to parents in
this context. First, the risks, benefits, and accuracy of this non-invasive blood test, highlighting that it
avoids brain biopsies to confirm diagnosis and can also inform treatment strategies, should be clearly
communicated to parents at the initial encounter. Second, to address issues with medical jargon,
disjointed communication in the medical team, as well as overwhelming parents with too much
information that they will likely not remember or understand, a lay written summary of the main points,
including a summary and interpretation of the LDT and other test results, from consultations with the
care team should be provided to families to further improve communication, understanding, and
acceptance of the diagnosis and prognosis of the child. Third, oncologists should pre-plan several
successive meetings with the families to ensure that they provide detailed and clear information to
parents in measured doses without overwhelming them during the initial encounter, which is particularly
crucial when the diagnosis is catastrophic. Fourth, all families should be provided with a standardized
protocol for psychosocial support, including access to social workers, a parent who has gone through a
similar situation that can help them navigate the system, or other counselling support programs that may
exist in the hospital. In the event that the diagnosis is catastrophic, the palliative care team should be
introduced early on in the process rather than later. Fifth, with respect to how the treating physician
should approach communication of the diagnosis and/or prognosis to the family, honesty, empathy, and
clarity were stressed as central values to avoid causing families unnecessary confusion and distress. Last, training in compassionate and empathetic communication of medical information, including
diagnostic, treatment, and prognostic information, should be provided to medical students, residents, and
physicians in pediatric oncology. Moreover, the salient themes in our deliberations align with five of six
core functions of family-centered communication, as outlined by a systematic review conducted by Sisk
et al.: fostering healing relationships, exchanging information, responding to emotions, making decisions
and managing uncertainty [9]. Our finding that, to foster healing relationships, prognostication should be communicated in an honest,
empathic, and compassionate way, catastrophic or not, was consistent with the literature [3, 10, 11]. Discussion Parents expressed how unclear or withholding information led to heightened distress and confusion,
which have both been shown to affect trust in the treating physician[12] and the clinical information
conveyed to them [13]. How and what diagnostic, prognostic, and treatment information should be exchanged was also
addressed in all deliberations. Consistent with other studies [3, 14, 15], parents stressed the importance of
providing detailed, clear, and unbiased information as well as supporting evidence for the diagnosis. Parents cautioned, however, about being overwhelmed at the initial diagnostic meeting, recommending
that information be delivered and repeated on an ongoing basis in subsequent meetings. This
recommendation is consistent with several previous studies highlighting the ongoing information needs
of parents of children with cancer [10, 13, 16–20]. Moreover, parents of children currently undergoing
treatment for difficult-to-treat brain tumors deliberated on the direct involvement of their children when Page 11/16 Page 11/16 information is exchanged, suggesting that discussions first take place with parents and then with their
children. While some previous studies showed variation in parents’ preferences for the direct involvement
of children in discussions with the physician [10, 21], others have also found that parents preferred to first
process the information separately from their children [18, 22, 23]. Avoiding the word ‘cancer’ and “chemotherapy” or using vague medical jargon were identified by some
parents as possible coping mechanisms used by oncologists to protect themselves when delivering bad
news. Honesty and empathy were re-iterated as crucial communication components of emotional
support, as in other studies [22, 24], allowing parents to react and eventually accept the diagnosis. Parents named emotional support personnel, including social workers, parent guides, counsellors, or the
palliative care team (for a catastrophic diagnosis) as especially key to helping them navigate their new
reality, and recommended such services be offered to all families regardless of the prognosis. In the event
of a catastrophic diagnosis, when the palliative care team should be introduced at the initial diagnostic
meeting was a point of disagreement between the bereaved and treatment parent groups, although both
groups acknowledged that they should be introduced early on in the process. While this introduction
rarely occurred early for many of the parents, it continues to demonstrate a potential divide between
palliative care and oncology teams [25]. Discussion Parents also expressed a need for emotionally supporting
siblings of affected children, due to the links between genetics and the type of brain cancer, which, to our
knowledge, has not yet been described in previous studies. When it comes to decision-making, parents preferred to be engaged and presented with all treatment
options, both in terms of benefits and risk, in a clear and unbiased way, i.e. without persuasion into a
clinical trial [3, 26]. In terms of the risk-benefit of each treatment option, parents also requested
information of the treatment’s impact on the child’s quality of life. Allowing time in between the first
diagnostic meeting and subsequent meetings where treatment options are explored was also viewed
favorably [14]. These communication strategies can foster trust, understanding, and peace-of-mind,
making parents feel supported during the decision-making process. Lastly, uncertainty around the diagnosis was a salient theme in all deliberations. Regardless of a test’s
demonstrated accuracy, parents expressed that they will always question a result suggesting poor or
catastrophic diagnosis. Acknowledging the margin of error around the information presented to parents
was deemed important in discussions and parents suggested that physicians be supportive of a request
for a second or third opinion to reduce even the slightest uncertainty. Our study corroborates many findings from the cancer communication literature, but is novel in at least
two primary ways. First, it is among the first to apply deliberative consultation methods to develop family-
centered communication recommendations, and second, to explore communication needs ahead of
clinical translation and implementation of a pharmacogenomic test in the pediatric cancer clinic. These
methods not only fostered a feeling of peer-support between members of each deliberation, but also led
to a trusting research partnership with Meagan’s Walk. Results from this study, for example, informed the
development of a communication intervention study targeting pediatric oncology physicians and Page 12/16 Page 12/16 Page 12/16 trainees. We did not recruit a representative random sample of bereaved parents or parents of children
currently undergoing treatment for brain cancer, which is a potential limitation of the study. The themes
highlighted in our deliberations are, however, consistent with those reported across other pediatric cancer
contexts and lends robustness to this method of engagement. trainees. We did not recruit a representative random sample of bereaved parents or parents of children
currently undergoing treatment for brain cancer, which is a potential limitation of the study. Consent for publication Not applicable. Discussion The themes
highlighted in our deliberations are, however, consistent with those reported across other pediatric cancer
contexts and lends robustness to this method of engagement. Our findings have direct implications for implementing the LDT in clinical practice. The agreements
reached by parents in the mixed deliberation provide key family-centered recommendations for
communicating results and informing clinical decision-making. Among the recommendations, lay written
summaries of the information disclosed in each meeting with the clinical care team as well as access to
emotional support personnel, including genetic counselling for the family, should be offered alongside the
implementation of the LDT. Standardized training in sensitive and honest communication should be
offered to medical students, residents, and physicians in pediatric oncology to better equip them when
faced with the challenge of disclosing poor diagnoses. Further work should focus on interventional
research, such as developing a standardized medical education curriculum in effective and sensitive
communication to families in pediatric oncology. Availability of data and materials The datasets generated and/or analysed during the current study are not available due to privacy and
confidentiality of the statements made by deliberants in documented transcripts. Ethics approval and consent to participate This study was approved by the McGill University’s Institutional Ethics Board (#A03-B13-16B). All parents
consented to the deliberations, the recording of the deliberations, and having the results of the
deliberations published. Competing interests The authors have no competing interests to declare. Acknowledgments This work was made possible through the funding support of Genome Canada and Genome Quebec. The
authors would like to thank Meagan’s Walk and the parents who took the time to participate in our
deliberative consultations. Without their support, this research would not have been possible. Authors’ contributions GB, VR, and CL contributed to the conception, design, data collection, interpretation of the results. CL and
VR analyzed the data and CL wrote the manuscript. GB and VR critically revised the manuscript. LDT: Laboratory Derived Test LDT: Laboratory Derived Test LDT: Laboratory Derived Test Funding This study was part of the “Biomarkers for Pediatric Glioblastoma through Genomics and Epigenomics”
project, which was funded by Genome Canada and Genome Quebec with Dr. Nada Jabado as principle
investigator. Dr. Vaso Rahimzadeh was funded by the David McCutcheon Pediatric Palliative Care
Fellowship and Dr. Cristina Longo was funded by the Fonds de Recherche Santé du Québec. Page 13/16 Page 13/16 Abbreviations HGG: High-Grade Glioblastoma References 1. Dobrozsi S, Trowbridge A, Mack JW, Rosenberg AR. Effective Communication for Newly Diagnosed
Pediatric Patients With Cancer: Considerations for the Patients, Family Members, Providers, and
Multidisciplinary Team. Am Soc Clin Oncol Educ Book. 2019;39:573–81. 1. Dobrozsi S, Trowbridge A, Mack JW, Rosenberg AR. Effective Communication for Newly Diagnosed
Pediatric Patients With Cancer: Considerations for the Patients, Family Members, Providers, and
Multidisciplinary Team. Am Soc Clin Oncol Educ Book. 2019;39:573–81. 2. Odeniyi F, Nathanson PG, Schall TE, Walter JK. Communication Challenges of Oncologists and
Intensivists Caring for Pediatric Oncology Patients: A Qualitative Study. J Pain Symptom Manage. 2017;54(6):909–15. 2. Odeniyi F, Nathanson PG, Schall TE, Walter JK. Communication Challenges of Oncologists and
Intensivists Caring for Pediatric Oncology Patients: A Qualitative Study. J Pain Symptom Manage. 2017;54(6):909–15. 3. Mack JW, Wolfe J, Grier HE, Cleary PD, Weeks JC. Communication about prognosis between parents
and physicians of children with cancer: Parent preferences and the impact of prognostic information. J Clin Oncol. 2006;24(33):5265–70. 4. Marron JM, Cronin AM, Kang TI, Mack JW. Intended and unintended consequences: Ethics,
communication, and prognostic disclosure in pediatric oncology. Cancer. 2018;124(6):1232–41. 5. Karremann M, Gielen GH, Hoffmann M, Wiese M, Colditz N, Warmuth-Metz M, Bison B, Claviez A, van
Vuurden DG, von Bueren AO, et al. Diffuse high-grade gliomas with H3 K27M mutations carry a
dismal prognosis independent of tumor location. Neuro Oncol. 2018;20(1):123–31. 6. Khuong-Quang DA, Buczkowicz P, Rakopoulos P, Liu XY, Fontebasso AM, Bouffet E, Bartels U,
Albrecht S, Schwartzentruber J, Letourneau L, et al. K27M mutation in histone H3.3 defines clinically
and biologically distinct subgroups of pediatric diffuse intrinsic pontine gliomas. Acta Neuropathol. 2012;124(3):439–47. 7. Longo C, Rahimzadeh V, O'Doherty K, Bartlett G. Addressing ethical challenges at the intersection of
pharmacogenomics and primary care using deliberative consultations. Pharmacogenomics. 2016;17(16):1795–805. 7. Longo C, Rahimzadeh V, O'Doherty K, Bartlett G. Addressing ethical challenges at the intersection of
pharmacogenomics and primary care using deliberative consultations. Pharmacogenomics. 2016;17(16):1795–805. Page 14/16 8. De Vries R, Stanczyk AE, Ryan KA, Kim SYH. A Framework for Assessing the Quality of Democratic
Deliberation: Enhancing Deliberation as a Tool for Bioethics. J Empir Res Hum Res. 2011;6(3):3–17. 9. Sisk BA, Mack JW, Ashworth R, DuBois J. Communication in pediatric oncology: State of the field
and research agenda. Pediatr Blood Cancer 2018, 65(1). 10. Zwaanswijk M, Tates K, van Dulmen S, Hoogerbrugge PM, Kamps WA, Bensing JM. References Young patients',
parents', and survivors' communication preferences in paediatric oncology: results of online focus
groups. BMC Pediatr. 2007;7:35. 11. Snaman JM, Torres C, Duffy B, Levine DR, Gibson DV, Baker JN. Parental Perspectives of
Communication at the End of Life at a Pediatric Oncology Institution. J Palliat Med. 2016;19(3):326–
32. 12. von Essen L, Enskar K, Skolin I. Important aspects of care and assistance for parents of children, 0–
18 years of age, on or off treatment for cancer. Parent and nurse perceptions. Eur J Oncol Nurs. 2001;5(4):254–64. 13. Clarke JN, Fletcher P. Communication Issues Faced by Parents who have a Child Diagnosed with
Cancer. J Pediatr Oncol Nurs. 2003;20(4):175–91. 13. Clarke JN, Fletcher P. Communication Issues Faced by Parents who have a Child Diagnosed with
Cancer. J Pediatr Oncol Nurs. 2003;20(4):175–91. 14. Baker JN, Leek AC, Salas HS, Drotar D, Noll R, Rheingold SR, Kodish ED. Suggestions from
adolescents, young adults, and parents for improving informed consent in phase 1 pediatric
oncology trials. Cancer. 2013;119(23):4154–61. 15. Hinds PS, Drew D, Oakes LL, Fouladi M, Spunt SL, Church C, Furman WL. End-of-life care preferences
of pediatric patients with cancer. J Clin Oncol. 2005;23(36):9146–54. 16. Jackson AC, Stewart H, O'Toole M, Tokatlian N, Enderby K, Miller J, Ashley D. Pediatric brain tumor
patients: Their parents' perceptions of the hospital experience. J Pediatr Oncol Nurs. 2007;24(2):95–
105. 17. Kastel A, Enskar K, Bjork O. Parents' views on information in childhood cancer care. European
Journal of Oncology Nursing. 2011;15(4):290–5. 18. Kessel RM, Roth M, Moody K, Levy A. Day One Talk: parent preferences when learning that their child
has cancer. Support Care Cancer. 2013;21(11):2977–82. 19. Kilicarslan-Toruner E, Akgun-Citak E. Information-seeking behaviours and decision-making process of
parents of children with cancer. European Journal of Oncology Nursing. 2013;17(2):176–83. 20. Parker TA, Johnston DL. Parental perceptions of being told their child has cancer. Pediatr Blood
Cancer. 2008;51(4):531–4. 20. Parker TA, Johnston DL. Parental perceptions of being told their child has cancer. Pediatr Blood
Cancer. 2008;51(4):531–4. 21. Young B, Dixon-Woods M, Windridge KC, Heney D. Managing communication with young people who
have a potentially life threatening chronic illness: qualitative study of patients and parents. Brit Med
J. 2003;326(7384):305–8b. 21. Young B, Dixon-Woods M, Windridge KC, Heney D. Managing communication with young people who
have a potentially life threatening chronic illness: qualitative study of patients and parents. Brit Med
J. 2003;326(7384):305–8b. 22. References Zwaanswijk M, Tates K, van Dulmen S, Hoogerbrugge PM, Kamps WA, Beishuizen A, Bensing JM. Communicating with child patients in pediatric oncology consultations: a vignette study on child
patients', parents', and survivors' communication preferences. Psycho-Oncology. 2011;20(3):269–77. 22. Zwaanswijk M, Tates K, van Dulmen S, Hoogerbrugge PM, Kamps WA, Beishuizen A, Bensing JM. Communicating with child patients in pediatric oncology consultations: a vignette study on child
patients', parents', and survivors' communication preferences. Psycho-Oncology. 2011;20(3):269–77. 22. Zwaanswijk M, Tates K, van Dulmen S, Hoogerbrugge PM, Kamps WA, Beishuizen A, Bensing JM. Communicating with child patients in pediatric oncology consultations: a vignette study on child
patients', parents', and survivors' communication preferences. Psycho-Oncology. 2011;20(3):269–77. Page 15/16 Page 15/16 23. Young B, Ward J, Salmon P, Gravenhorst K, Hill J, Eden T. Parents' Experiences of Their Children's
Presence in Discussions With Physicians About Leukemia. Pediatrics. 2011;127(5):E1230–8. 24. Orioles A, Miller VA, Kersun LS, Ingram M, Morrison WE. "To Be a Phenomenal Doctor You Have To Be
the Whole Package": Physicians' Interpersonal Behaviors during Difficult Conversations in Pediatrics. J Palliat Med. 2013;16(8):929–33. 25. Johnston DL, Nagel K, Friedman DL, Meza JL, Hurwitz CA, Friebert S. Availability and use of
palliative care and end-of-life services for pediatric oncology patients. J Clin Oncol. 2008;26(28):4646–50. 26. Hazen RA, Zyzanski S, Baker JN, Drotar D, Kodish E. Communication about the risks and benefits of
phase I pediatric oncology trials. Contemp Clin Trials. 2015;41:139–45. SupplementaryMaterials.docx Supplementary Files This is a list of supplementary files associated with this preprint. Click to download. This is a list of supplementary files associated with this preprint. Click to download. SupplementaryMaterials.docx Page 16/16
|
https://openalex.org/W4214590247
|
https://ruc.udc.es/dspace/bitstream/2183/30757/3/feart-10-838172.pdf
|
English
| null |
Editorial: Flood Susceptibility and Risk Maps as a Crucial Tool to Face the Hydrological Extremes in Developing Countries: Technical and Governance Aspects Linked by a Participatory Approach
|
Frontiers in earth science
| 2,022
|
cc-by
| 1,743
|
EDITORIAL EDITORIAL
published: 25 February 2022
doi: 10.3389/feart.2022.838172 Editorial on the Research Topic Flood Susceptibility and Risk Maps as a Crucial Tool to Face the Hydrological Extremes in
Developing Countries: Technical and Governance Aspects Linked by a Participatory Approach Editorial: Flood Susceptibility and Risk
Maps as a Crucial Tool to Face the
Hydrological Extremes in Developing
Countries: Technical and Governance
Aspects Linked by a Participatory
Approach Rosanna Bonasia 1*, Luis Cea 2 and Umberto Fratino 3 1CONACYT—Instituto Politécnico Nacional ESIA UZ, Mexico City, Mexico, 2Department of Civil Engineering, Universidade da
Coruña Environmental and Water Engineering Group, A Coruña, Spain, 3DICATECh, Politecnico di Bari, Bari, Italy 1CONACYT—Instituto Politécnico Nacional ESIA UZ, Mexico City, Mexico, 2Department of Civil Engineering, Universidade da
Coruña Environmental and Water Engineering Group, A Coruña, Spain, 3DICATECh, Politecnico di Bari, Bari, Italy Keywords: flood susceptibility, participatory approach, risk maps, hydrological extremes, developing countries Floods have been the subject of repeated alarming news in 2021, for their recurrence has
increased in frequency all over the world with very high-cost consequences in terms of
economic losses and human suffering. Floods represent one of the most common and
severe natural disasters, capable of disrupting people’s livelihoods around the world. Their
effects are particularly evident in low-income countries, where the lack of adequate
infrastructure, limited know-how and the scarcity of monitoring systems, make the
situation even more serious. Edited and reviewed by:
Carmine Galasso,
University College London,
United Kingdom *Correspondence:
Rosanna Bonasia
rbonasia@conacyt.mx Rentschler and Salhab (2020) in their recent study of 189 countries’ exposure to floods and
poverty, find that 19% of the world’s population is in conditions of direct exposure to risk during
flood events that occur once every 100 years. Of this percentage, the majority of the population at
risk is concentrated in developing countries, most located in sub-Saharan Africa and in South
and East Asia. The number of people exposed to flood risk has increased enormously in recent
decades, surpassing, already in 2020, the estimates made for 2050 (Jongman et al., 2012;
Rentschler and Salhab, 2020). This situation is due to various factors, among which climate
change plays a fundamental role. Emanuel (2013) estimates that the frequency of tropical
cyclones is increasing by 10–40% during the first three quarters of the twenty-first century, while
their intensity is estimated at an increase of about 15% as calculated by Hong et al. (2021). However, it is not only major and infrequent floods that pose a danger to exposed populations,
even less rare and minor events can jeopardize particularly vulnerable populations (Hoeppe,
2016). The increase in susceptibility to flooding is also determined by the increase in population
and urbanization of high-risk areas. The uncontrolled urbanization determines a change in land
use which causes a direct influence on the increase in surface runoff levels (Huang et al., 2014;
Liu et al., 2019). Specialty section:
This article was submitted to
Geohazards and Georisks,
a section of the journal
Frontiers in Earth Science
Received: 17 December 2021
Accepted: 31 January 2022
Published: 25 February 2022 Specialty section:
This article was submitted to
Geohazards and Georisks,
a section of the journal
Frontiers in Earth Science Specialty section:
This article was submitted to
Geohazards and Georisks,
a section of the journal
Frontiers in Earth Science Received: 17 December 2021
Accepted: 31 January 2022
Published: 25 February 2022 Received: 17 December 2021
Accepted: 31 January 2022 Citation: Bonasia R, Cea L and Fratino U (2022)
Editorial: Flood Susceptibility and Risk
Maps as a Crucial Tool to Face the
Hydrological Extremes in Developing
Countries: Technical and Governance
Aspects Linked by a
Participatory Approach. Front. Earth Sci. 10:838172. doi: 10.3389/feart.2022.838172 Bonasia R, Cea L and Fratino U (2022)
Editorial: Flood Susceptibility and Risk
Maps as a Crucial Tool to Face the
Hydrological Extremes in Developing
Countries: Technical and Governance
Aspects Linked by a In the light of these evidence, this Research Topic wanted to emphasize the importance of local
strategies in managing flood risk, in order to increase public awareness and preparedness. February 2022 | Volume 10 | Article 838172 1 Frontiers in Earth Science | www.frontiersin.org Editorial: Flood Suscepribility and Participatory Approach Bonasia et al. the risk of floods. The methodology presented can be applied
in other areas and represents a useful tool for establishing the
active involvement of stakeholders in the evaluation and
implementation of NBS. Scaini et al. focus attention on the importance of citizen
involvement
in
flood
risk
management
plans. Their
participation in the development of local inputs for the
measures to be taken to deal with the risk of flooding is still
very limited, given the difficulty of quantifying the perceived risk
because it is based on personal opinions. However, authors point
out that perceived risks may correspond to evidence-based data
and in their work, they present a project questionnaire which
aims to characterize the perception of flood risk and its spatial
distribution at the basin scale. The proposal was tested on the
Tagliamento river, in Italy, one of the last free-flowing rivers in
Western Europe and the last free-flowing Alpine river. The
questionnaire
was
tested
for
assess
risk
perception
and
compare it to available hazard/risk maps. Results show that
perceived risk is in good agreement with existing hazard and
risk assessments. Furthermore, the population has shown a strong
interest in being involved in handling the risk debate, expressing
the belief that risk reduction is compatible with river conservation
and both of these objectives should be achieved together. Finally, Mokkenstorm et al. present a methodology for flood
detection based on the use of Passive Microwave Remote Sensing
(PMRS). The authors stress the importance of the application of
remote sensing satellite data, particularly in ungauged river basins
of developing countries, due to their low cost and low
maintenance. Citation: Particularly, this study focuses on determining
whether PMRS images can actually be used for the prevention
of damage caused by floods and the cases of application are two
small-scale basins in Malawi. Two indices were analyzed: the m
index, expressed as the magnitude relative to the average flow,
and an index that takes into account an additional wet calibration
cell. Results, compared with the discharge estimates of the Global
Flood Awareness System, show that the detected flood levels are
very accurate, demonstrating that the methodology is valid and
particularly useful for monitoring rivers of data-scarce areas. j
g
The concepts of conservation of natural areas and a
participatory approach in support of decision makers are also
developed by Scrieciu et al. who analyze the potential of nature-
based solutions (NBS) to deal with the risk of flooding. The
authors start from the basic hypothesis that the barriers posed to
the wide diffusion of NBS are mainly due to the lack of
involvement and knowledge integration of stakeholders during
the design phases of NBS. For this reason, they propose a
combined
approach
based
on
the
integration
of
Fuzzy
Cognitive
Maps
(FCM),
Hydraulic
Modeling
(HM),
and
participatory
Bayesian
Belief
Networks
(BBN)
aiming
to
facilitate the stakeholders’ engagement and the knowledge
integration
process in
NBS
design
and
assessment. The
proposal has been applied to the Lower Danube demo site,
with the aim also of supporting the process of the Potelu
Wetland restoration. The proposed methodology aims to
demonstrate the suitability of the BBN, in support of FCM
and HM, to evaluate the effectiveness of NBS in addressing In summary, this Research Topic wants to draw attention to
two points of reflection: 1) The need to involve stakeholders at all
levels in the planning of actions aimed at reducing the risks
associated with floods is preponderant, given the considerable
increase in urban settlements in high-risk areas of developing
countries. 2) Likewise, we hope that new lines of research will
open up to explore innovative solutions that help in risk
management without having to resort to costly methods,
which involve manpower or require continuous maintenance. AUTHOR CONTRIBUTIONS The authors are co-editors which organized this Research Topic,
substantially, directly, and intellectually contributed to the work
on this Research Topic and the manuscript and approved it for
publication. All authors contributed to the article and approved
the submitted version. Rentschler, J., and Salhab, M. (2020). People in Harm’s Way : Flood Exposure and
Poverty in 189 Countries. Policy Research Working Paper: No. 9447. Washington,
DC:
World
Bank. Available
at:
https://openknowledge. worldbank.org/handle/10986/34655 License: CC BY 3.0 IGO. © World Bank. REFERENCES Emanuel, K. A. (2013). Downscaling CMIP5 Climate Models Shows Increased
Tropical Cyclone Activity over the 21st century. Proc. Natl. Acad. Sci. 110,
12219–12224. doi:10.1073/pnas.1301293110 Conflict of Interest: The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could be construed as a
potential conflict of interest. Hoeppe, P. (2016). Trends in Weather Related Disasters - Consequences for
Insurers and Society. Weather Clim. Extremes 11, 70–79. doi:10.1016/j.wace. 2015.10.002 Publisher’s Note: All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their affiliated organizations, or those of
the publisher, the editors and the reviewers. Any product that may be evaluated in
this article, or claim that may be made by its manufacturer, is not guaranteed or
endorsed by the publisher. Hong, C.-C., Tsou, C.-H., Hsu, P.-C., Chen, K.-C., Liang, H.-C., Hsu, H.-H., et al. (2021). Future Changes in Tropical Cyclone Intensity and Frequency over the
Western North Pacific Based on 20-km HiRAM and MRI Models. J. Clim. 34,
2235–2251. doi:10.1175/jcli-d-20-0417.1 2235–2251. doi:10.1175/jcli-d-20-0417.1 Huang, C.-J., Hsu, M.-H., Teng, W.-H., and Wang, Y.-H. (2014). The Impact of
Building Coverage in the Metropolitan Area on the Flow Calculation. Water 6,
2449–2466. doi:10.3390/w6082449 Copyright © 2022 Bonasia, Cea and Fratino. This is an open-access article
distributed under the terms of the Creative Commons Attribution License (CC
BY). The use, distribution or reproduction in other forums is permitted, provided the
original author(s) and the copyright owner(s) are credited and that the original
publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with
these terms. Jongman, B., Ward, P. J., and Aerts, J. C. J. H. (2012). Global Exposure to River and
Coastal Flooding: Long Term Trends and Changes. Glob. Environ. Change 22
(4), 823–835. doi:10.1016/j.gloenvcha.2012.07.004 Liu, S., Huang, S., Xie, Y., Wang, H., Leng, G., Huang, Q., et al. (2019). Identification of
the Non-stationarity of Floods: Changing Patterns, Causes, and Implications. Water Resour. Manage. 33, 939–953. doi:10.1007/s11269-018-2150-y February 2022 | Volume 10 | Article 838172 Frontiers in Earth Science | www.frontiersin.org 2
|
https://openalex.org/W4250933875
|
https://zenodo.org/records/1751693/files/article.pdf
|
English
| null |
Reviews
|
Analyst (London. 1877. Online)/Analyst
| 1,915
|
public-domain
| 2,235
|
View Article Online / Journal Homepage / Table of Contents for this issue View Article Online / Journal Homepage / Table of Contents for this issue 424 424 REPORT ALCOHOLOMETRIC TABLES. By SIR EDWARD
THORPE,
C.B., F.R.S. London : Long-
Price 3s. 6d. net. mans, Green and Go, 1915. ALCOHOLOMETRIC TABLES. By SIR EDWARD
THORPE,
C.B., F.R.S. London : Long-
Price 3s. 6d. net. mans, Green and Go, 1915. There has been a long felt want by chemists and technologists for a complete
and trustworthy set of Tables of Alcoholometry based on a critical examination
of the whole of the experimental data collected during the past hundred years or so. This want has been especially felt by those who require for their scientific researches
the utmost refinement of accuracy in the observed relations between density and
alcoholic concentrations. Up to recently the most complete and accurate Tables available have been those
compiled in 1880 by the late Sir Thomas Stevenson, who states that for spirits
ranging in specific gravity from 0-8250 to 0.9983 they were based on Blagden and View Article Online 425 REVIEWS Gilpin’s tables of 1794. The great accuracy of Blagden and Gilpin’s observations
have been confirmed by all subsequent workers in this field, and especially by
MendelBef. For spirits of less specific gravity than 0.8250 Stevenson made use
of the data of Fownes, but he appears to have done this with a certain amount
of distrust, since he was quite aware that the work of Fownes was much inferior
in accuracy to that of Blagden and Gilpin. For spirits of higher specific gravity
than 0-9983 Stevenson does not state his authority, but apparently his resnlts were
calculated by extrapolation. y
p
In the Tables of 1880 the specific gravity of absolute alcohol is given as 0.7938 at
60’ F., compared with water at the same temperature, all weighings beings made in
air ; but in the Preface to the second edition of his “Treatise on Alcohol,” pub-
lished in 1888, Stevenson recognises that this value is somewhat too high, and in
the body of the work he incorporates more recent investigations on this point,
notably those of Messrs. Squibb. He did not consider it necessary, however, to alter
the Tables published eight years previously, since he regarded them as ‘‘ sufficiently
accurate for all practical purposes ; and the percentages of proof spirit so necessary
to the public analyst, the brewer, the distiller, and the officers of Excise and Customs,
are of course unaffected by the new determinations of the specific gravity of absolute
alcohol.” During the quarter of a century which has elapsed since these words were
written no further attempt appears to have been made to introduce a higher degree
of accuracy into the Alcohol Tables until Sir Edward Thorpe undertook the laborious
task of re-examining and criticising all the existing data. The result is now given in
this little volume in the form of a, very complete set of Tables compiled under his
direction by Mr. T. J. Chester and Mr. John Holmes of the Government Laboratory. They are preceded by a short Introduction giving a highly interesting summary of
the progress of alcoholornetry since the year 1790, and an account of the standard
fiscal spirits of various countries, and of the hydrometers used for the determination
of their strength. g
The new Tables of Sir Edward Tborpe have been founded on the work of four
authorities -
Blagden and Gilpin, Drinkwater, Mendeldef, and the Kaiserlich
Normal-Eichungs Kommission. The labour expended on their compilation must
have been very great, entailing as it did the comparison and collation of a vast
amount of experimental data, and their reduction to a uniform system ; but it has
been well justified by the result, and from the experience of the writer of this notice
it may confidently be stated that the Tables may be safely used in researches
requiring a, high degree of scientific accuracy. q
g
g
g
y
The specific gravity of absolute alcohol at 15*6”/15*6”
C. is taken as 0.79359 in
air, a . value deduced from Mendeldef’s observations. The corresponding value for
British Fiscal Proof Spirit containing 49.28 per cent, of alcohol by weight is 0.91976,
against 0,9198 as given by Stevenson. Table I. gives for each interval of density of 0-0002 the percentage of alcohol
by weight and by volume, and also the percentage of fiscal proof spirit, the specific
gravities being determined at 60” 3’. or 1 5 6 O C. against water at the same tempera-
ture. A comparison of this Table with that of Stevenson does not show many points View Article Online View Article Online 4 26 REVIEWS of absolute agreement in the percentages of alcohol by weight. The differences as
far as they have been tested appear to vary from about 0.10 to 0.01 per cent., but
there is no consistent order in their distribution, and they are sometimes plus and
sometimes minus. There is, as might be expected, a very much closer approximation
to identity in the percentages of proof spirit as given by the old and the new Tables
respectively, since any variation in the assumed specific gravity of absolute alcohol
necessarily alters the standard value of proof spirit to a corresponding extent. Table 11. gives a synopsis of the standard fiscal spirits of Great Britain, America,
Germany, and France, as compared with the indications of the Sikes hydrometer,
which since 1816 has been the legal instrument for determining the strength of spirits
for revenue purposes in this country. Published on 01 January 1915. Downloaded on 28/10/2014 13:05:05. Table 111. gives a similar comparison of the readings of the Sikes hydrometer
with the hydrometers in use in Russia, Holland, Spain, and Switzerland. All the Tables are printed in bold type, and in such a form as to admit of
ready use. HORACE
T. BROWN. All the Tables are printed in bold type, and in such a form as to admit of
d
HORACE
T BROWN HORACE
T. BROWN ELEMENTARY
CHEMICAL MICROSCOPY. By EMILE MONNIN CHAMOT. 1915. New
York : John Wiley and Sons. London : Chapman and Hall, Ltd. Price
12s. 6d. net. The keynote of this work is to be found in the first paragraph of the preface, in
which the author says: “ H e (ie., the American chemist) has also failed to grasp
the fact that the modern microscope is in reality a more important adjunct to his
laboratory than the spectrometer, poiarimeter, or refractiometer ; in fact, it way be
said that the microscope is entitled to as important a place as the analytical balance. No one other instrument can perform so many functions and do them all well.” From this it will be inferred that the author is an enthusiast, and as such he
possesses the defects of his qualities. In food analysis, in biochemistry, in textile
chemistry, and in toxicology, to name only a few departments of chemistry, the
microscope has for long been recognised as an indispensable item of laboratory
equipment. Many textbooks of varying degrees of excellence have been written, and
nothing more is needed to emphasise the value to the chemist of microscopical
investigation. It is not, however, with these highly important and more obvious
uses of the microscope that the author concerns himself, but rather with the employ-
ment of that instrument for the purpose of carrying out well-known chemical
reactions on a very small scale, whilst he devotes some space to indicating how, by
the exercise of ingenuity, the microscope may be made to do work for which it is not
particularly well adapted. The limitations of the microscope as applied to chemical analysis are fairly well
recognised, and if ‘the professional chemist does not more frequently utilise so-called
micro-chemical reactions, it is because these are not as a rule sufficiently definite
and specific. The difficulties, when working on a minute quantity of material, of separating in
a state of purity any particular substance prior to applying microscopical tests, are
so great as to render the method useless in the great majority of cases. It does,
however, happen, notably in toxicological investigations, that micro-chemical indicrt- View Article Online 427 REVIEWS tions are sometimes very valuable indeed, and there can be no doubt that the author
is right in suggesting that the chemist might with advantage employ the microscope
more frequently than he does. Whether or not the writer is correct in thinking that
that portion of the book which deals with the micro-chemical reactions of the
common elements will not prove to be of much value to the chemical practitioner,
nothing but praise can be found for the remainder of the work, This is thoroughly
well written, and embraces a clear and admirable exposition of microscopical
technique. Even experienced microscopists could not fail to derive some useful
information and many valuable hints from a study of it, whilst the average chemist
will find that it constitutes a trustworthy and handy guide in his microscopical
work. Published on 01 January 1915. Downloaded on 28/10/2014 13:05:05. The print is clear, there ere very few typographical errors, there is a good index
and the spelling is not aggressively Transatlantic. A. CHASTON CHAPMAN. g
A. CHASTON CHAPMAN. THE CHEMISTS’
YEAR-BOOK, 1915. Edited by F. W. ATACK. Manchester : Sherratt
and Hughes. 1915. Price 10s. 6d. net. Mr. Atack has hammered a highly commendable little nail into the coffin of
German chemical indispensability by producing an excellent British substitute for
the more familiar Cherniker Kalender. The 900 closely-printed pages contain vast
stores of information upon all kinds of ohemical topics, and the concluding sections
dealing with essential oils, textile fibres, synthetic dyestuffs, alkaloids, pharma-
ceutical synonyms, trade names of drugs, indiarubber, tobacco, and photography,
are valuable features. The discussion of indicators will bring joy to the heart of the
ionophil, besides being serviceable to the mere chemist, and it is a useful idea
to assemble the volume numbers of the principal journals against the year in
which they are published. The lists of minerals, inorganic compounds and organic
compounds with their properties, are very comprehensive, the last-named group
including nearly 3,000 individuals ; in connection with the organic section, I would
suggest the inclusion of the empirical formulae in place of the “ crystalline form and
colour” column, which is not of any particular value, whereas in those cases where
the constitution is complex the empirical formula would be helpful in the consulta-
tion of Richter’s Lexikon, ,
In war-time one searches naturally for a, section on explosives, and in consequence
of its absence we are thrown back on the Daily Mail to inform us each morning why
cotton should be contraband, and who is the greatest living ‘‘ scientist.”
A more
serious drawback is the obstinate inclusion of a diary which occupies 116 pages, and
which, for at least two reasons, it is highly improbable that anyone will use : in the
first place, most people note their most interesting engagements in the treasured
intimacy of a waistcoat pocket, and then in war-time nobody should desecrate a
10s. 6d. volume with ephemeral jottings. It is difficult to understand, also, why
the editor clings so strenuously to the tongue-and-pouch fetish, with the inadequate
pencil, because although they were a part of the Chemiker Kalender, a comparison
unfavourable to Volume I. is imtituted by dividing the present work into two volumes
and leaving the second volume in ordinary book form. View Article Online 428 REVIEWS The care with which the proofs appear to have been corrected is highly praise-
worthy, because the tedium of checking so much tabular matter must be appalling,
and tabular matter is valueless unless it is carefully checked. It is to be hoped that,
Mr. Atack’s applied patriotism will meet with the support which it deserves, not
only from all British chemists who had found satisfaction in the Chemiker Kalender,
but from those who, on general principles, recognise the vital necessity of supporting
home industries from this time forth for evermore. M. 0. FORSTER. y
pp
M. 0. FORSTE OUTLINES
OF ORGANIC CHEMISTRY. By F. J. MOORE,
Ph.D. Professor of Organic
Chemistry in the Massachusetts Institute of Technology. Pp. xi + 325. New
York : John Wiley and Sons, Inc. ; London : Chapman and Hall, Ltd. 1914. Price 6s. 6d. net. Published on 01 January 1915. Downloaded on 28/10/2014 13:05:05. The author explains in the preface to the first edition that he annually delivers
a course of about thirty lectures on the underlying principles of organic chemistry to
a class made up of candidates taking their degree in Physics, Biology, and Sanitary
Engineeripg. As these students need a knowledge of organic chemistry, but do
not intend to become skilled organi,c chemists, the eubiect-matter is handled in a,
different manner from thaf usually adopted when the primary object i s to train
organic chemists. In dealing with a group of compounds, those substances are used
as examples which are of more practical importance. The book has been foundedon the course of lectures and is very readable. The
lectures themselves are based on a syllabus which comprises most of the subjects
with which a scientific man who is not a speoialist in organic chemistry might:
advantageously be acquainted. general knowledge of organic chemistry without specialising in the subject,
The book should prove useful to those who desire j
,
J. T. HEWITT
|
https://openalex.org/W3047066454
|
https://diglib.uibk.ac.at/ulbtirolfodok/content/titleinfo/5950833/full.pdf
|
English
| null |
Pros and cons of using a standard protocol to test germination of alpine species
|
Plant ecology
| 2,020
|
cc-by
| 15,335
|
Plant Ecol (2020) 221:1045–1067
https://doi.org/10.1007/s11258-020-01061-w (0123456789().,-volV)(0123456789().,-volV) Pros and cons of using a standard protocol to test
germination of alpine species Vera Margreiter
. Konrad Pagitz . Christian Berg . Patrick Schwager . Brigitta Erschbamer Received: 12 February 2020 / Accepted: 28 July 2020 / Published online: 3 August 2020
The Author(s) 2020 community, occurrence at elevation belts, bedrock
types, as well as CSR strategies, and further, seed mass
was examined. We could not find statistically signif-
icant differences of FGP within these classes, but
74.9% of all tested species germinated using the
standard protocol, and half of them had FGP
C 20.1–100%. A treatment with gibberellic acid
enhanced the germination in half of the species to
which this treatment was applied. Common families in
alpine regions, i.e. Asteraceae, Poaceae and Saxifra-
gaceae were highlighted in terms of their germination
behaviour. The results provide an evaluation of the
application of standard protocols to a broad Alpine
species pool on the one hand, and on the other hand,
provide ecological insights of the species tested. Germination is not only one of the most important
events of the reproductive cycle of plants but could
also be a key feature in species’ responses to changing
environmental conditions. Abstract
Storing seeds in seed banks is an effective
way to preserve plant diversity and conserve species. An essential step towards a valuable conservation is
the validation of germination. This study presents a
germination screening of seeds from 255 species of the
European Eastern Alps, which were to be stored at the
Millennium Seed Bank (Kew, UK). The final germi-
nation percentage (FGP) was determined using a
standard protocol in the laboratory. Species were
classified
according
to
species
rarity,
plant Communicated by Thomas Abeli. Electronic supplementary material
The online version of
this article (https://doi.org/10.1007/s11258-020-01061-w)
contains supplementary material, which is available to autho-
rized users. V. Margreiter (&) K. Pagitz B. Erschbamer
Department of Botany, University of Innsbruck,
Sternwartestr. 15, 6020 Innsbruck, Austria
e-mail: vera.margreiter4@gmail.com
K. Pagitz
e-mail: konrad.pagitz@uibk.ac.at
B. Erschbamer
e-mail: brigitta.erschbamer@uibk.ac.at
C. Berg P. Schwager
Institute of Biology, Department of Plant Science,
University of Graz, Holteigasse 6, 6020 Graz, Austria
e-mail: christian.berg@uni-graz.at
P. Schwager
e-mail: p.schwager@gmx.at V. Margreiter (&) K. Pagitz B. Erschbamer
Department of Botany, University of Innsbruck,
Sternwartestr. 15, 6020 Innsbruck, Austria
e-mail: vera.margreiter4@gmail.com Keywords
Seed bank Eastern alps Plant
community Rarity Seed mass Strategy Keywords
Seed bank Eastern alps Plant
community Rarity Seed mass Strategy Introduction To release this
type of dormancy, the seeds can be scarified so that
water can penetrate through the damaged seed coat. Morphological dormancy occurs in seeds with imma-
ture embryos. These seeds require a post-ripening
phase and often some additional treatments, such as a
series of cold and warm treatments or vice versa, to
break the dormancy (Baskin and Baskin 2014). A
generally simple and timesaving method is treatment
with gibberellic acid (GA3). This method is recom-
mended by ENSCONET (2009a; b) when a standard
protocol is used or when the type of dormancy is
unknown. Alpine plant species require relatively high tem-
peratures for germination (Bliss 1958; Amen 1966;
Billings and Mooney 1968; Cavieres and Arroyo
2000; Gime´nez-Benavides et al. 2005; Graae et al. 2008; Schwienbacher et al. 2011; Mondoni et al. 2012;
Baskin and Baskin 2014), whereas species below
treeline use a broader temperature range (Walder and
Erschbamer 2015; Ferna´ndez-Pascual et al. 2017). If
so, increasing temperatures in the course of climate
warming could be beneficial for both, alpine species
and upward-migrating lowland species, in contrast to
habitat specialists (Casazza et al. 2014) and species
with specific dormancy strategies (Mondoni et al. 2012, 2018). Field studies on that matter were hardly
performed and lab studies lack a comprehensive
screening of species from different elevations and
plant communities in the Alps (Mondoni et al. 2011, 2012; Isselin-Nondedeu and Be´de´carrats 2013;
Orsenigo et al. 2015; Walder and Erschbamer 2015;
Tudela-Isanta et al. 2018a, b). Here, a broad germi-
nation project can close this gap. The germination temperature can be regarded as a
proxy for the germination niche of a species (‘regen-
eration niche’ according to Grubb 1977), which
represents the ability of the species to adapt to
environmental conditions (Rosbakh and Poschlod
2015; Jime´nez-Alfaro et al. 2016). To investigate
optimal seed germination temperatures, several tem-
perature regimes (Mondoni et al. 2012) and seed pre-
treatments (Ferna´ndez-Pascual et al. 2017) are needed. In cases where seeds, time or working space is limited,
a standard protocol with only one temperature regime Among seed traits, seed mass seems to influence
germination (Schwienbacher and Erschbamer 2001;
Bu et al. 2007; Erschbamer et al. 2010; Liu et al. 2013), whereby large seeds are considered to have a
superior germination and survival. Introduction C. Berg P. Schwager
Institute of Biology, Department of Plant Science,
University of Graz, Holteigasse 6, 6020 Graz, Austria
e-mail: christian.berg@uni-graz.at
P. Schwager
e-mail: p.schwager@gmx.at Seed germination is one of the essential processes in a
plantslife cycle (Harper 1977; Fenner and Thompson
2005). It is highly influenced by individual species
requirements
(Baskin
and
Baskin
2014), 12 3 Plant Ecol (2020) 221:1045–1067 1046 characteristics of species’ niches (Schwienbacher
et al. 2012; Rosbakh and Poschlod 2015; Jime´nez-
Alfaro et al. 2016; Tudela-Isanta et al. 2018a, b), and
the environmental conditions during seed develop-
ment (Donohue et al. 2010; Bernareggi et al. 2016;
Ferna´ndez-Pascual et al. 2017). With increasing
temperatures in the Alps (Gobiet et al. 2014), knowl-
edge on germination traits have become highly
important. Generally, for common species, germina-
tion may be enhanced by increased temperatures
(Milbau et al. 2009; Mondoni et al. 2012; Rosbakh and
Poschlod 2015; Orsenigo et al. 2015). However,
specific germination requirements (Graae et al. 2008;
Shevtsova et al. 2009) and local adaptation (Kawecki
and Ebert 2004; Mondoni et al. 2009) are bottlenecks
for the growth of plant populations in a changing
environment. can be used. This procedure is practiced, for example,
when testing the germination ability of seed lots that
are to be stored in ex situ seed banks. It remains
questionable whether such standard protocols can
provide reasonable data on the germination ability of
species and whether the results allow an ecological
interpretation. One factor that hinders the success of germination
tests is seed dormancy (Baskin and Baskin 2004, 2014;
Finch-Savage and Leubner-Metzger 2006). It was
found that physiological dormancy occurs frequently
in alpine species (Schwienbacher et al. 2011). Here, a
distinction must be made between non-deep, interme-
diate and deep physiological dormancy (Baskin and
Baskin 2004). Non-deep physiological dormancy can
be broken by dry-cold storage (Wang et al. 2010). This
is done by storing the seeds in paper bags in a
refrigerator, which is a very simple method of
releasing dormancy. In contrast, breaking dormancy
in seeds with medium or low physiological dormancy
requires more effort, e.g. via stratification. In this
method, the seeds are stored under wet-cold conditions
for a certain period of time prior to a germination test
(Baskin and Baskin 2004, 2014). Furthermore, phys-
ical, or morphological dormancy and all combinations
can occur. Physical dormancy is a typical feature of
hard-coated seeds, such as of Fabaceae. 123 Seed collection Seeds were collected in the European Eastern Alps
with two centres in Austria, one in the western part
(Tyrol) and one in the south-eastern part (Styria). The
test species (Table 1) were selected in agreement with
the Millennium Seed Bank (MSB), representing 255
characteristic Alpine taxa from different plant com-
munities, endemic/rare species, and species that were
not already stored in the MSB. Nomenclature of the
species and affiliation to families follow Fischer et al. (2008). The seed collection was carried out in the
summers of 2016, 2017 and 2018. In all three years, an
above-average number of warm days combined with
above-average periods of sunshine were recorded in
Austria by the research institute ZAMG (https://www. zamg.ac.at). The summer of 2016 was a warm and wet
summer, the summer of 2017 was wet in mountain
areas, while the summer of 2018 was recorded as a
summer with above-average dry periods (ZAMG
2016, 2017, 2018). Collections were done in eleva-
tions from 600 m above sea level (a.s.l.) to 2700 m
a.s.l. with a central focus at the subalpine and alpine
elevation belt (1900–2400 m a.s.l.). According to
international standards (ENSCONET 2009a), collec-
tions were performed at the time of natural dispersal. The major fraction of the cleaned seeds was sent to the
MSB. A small fraction of each collection was stored
dry in paper bags in a refrigerator at ? 4 C (average
relative humidity of a refrigerator * 40%) until the
germination tests started. The storage period lasted
between 1 and 7 months. Fruits such as achenes,
legumes or caryopses were treated as seeds. Having access to various species from different
habitats and elevations, the study aimed to test the
germination success of these different species under a
standard protocol. Only one day/night temperature
regime was used to evaluate if such an approach
delivers appropriate germination data for an ex situ
conservation. We were interested in finding out
whether the results of our standard protocol provide
evidence for adaptive mechanisms of species distri-
bution along elevation gradients and habitats, and
whether conclusions can be drawn about functional
effects that influence the final germination percentage
(FGP). Based on this focus, we addressed the follow-
ing questions: (a) Does germination differ between
common,
scattered,
rare
and
endemic
species? (b) Does the FGP differ between species from
different plant communities and (c) from different
elevation belts (e.g. alpine, subalpine and mon-
tane/colline)? Introduction However, contra-
dicting results are found in the literature, noting only
weak or no influence of seed mass on germination
(Vera 1997; Eriksson 1999; Schwienbacher et al. 2012; Walder and Erschbamer 2015). 123 123 Plant Ecol (2020) 221:1045–1067 1047 The CSR strategy concept of Grime (1979) classi-
fies the life strategies of plants as stress tolerant (S),
competitive (C), ruderal (R) and intermediate strate-
gists (e.g. CSR). Ruderal strategists do not have
specific requirements for germination, while stress
tolerant species may be less flexible and more
demanding. In alpine environments, stress tolerant
species represent the most common strategy group
(Caccianiga et al. 2006). it possible to assign a germination pattern to the CSR
strategy types sensu Grime (1979)? (f) Is there a
correlation between FGP and seed mass? (g) Does
treatment with gibberellic acid (GA3) enhance FGP? (h) Do important families have a consistent germina-
tion pattern among their genera/species? Answers of
these questions improve the interpretation of germi-
nation data from standard protocols in terms of
ecological significance and can also be useful when
predicting species responses to global warming. g
This study presents the results of a broad-based
germination screening of 255 Alpine plant species
from the European Eastern Alps, including species
from lowlands (dry grasslands, montane meadows),
subalpine dwarf-shrub heaths, forest understorey
species, species of wet habitats, tall herbs, species of
alpine grasslands, rock crevices and scree sites. Seeds
were collected by two partners of the ‘‘Alpine Seed
Conservation and Research Network’’ (https://www. alpineseedconservation.eu/). The network was foun-
ded as a partner of the Millennium Seed Bank Part-
nership of the Royal Botanic Gardens, Kew (UK). One
goal of this institution is to ‘‘store seeds of 25% of the
world’s flora by 2020’’ (Mu¨ller et al. 2017, https://
www.kew.org/science/). Contemporaneously,
the
study of germination behaviour is one of the main
tasks of current seed ecology research, with the aim to
evaluate the needs of species, local adaptation and the
suitability of a species for ex situ conservation (Mu¨ller
et al. 2017). Seed collection (d) Are there differences between
species from siliceous and calcareous bedrock? (e) Is 123 Plant Ecol (2020) 221:1045–1067 1048 Table 1 Test species in alphabetical order, their family affiliation and rarity classification following Fischer et al. (2008)
Species
Family
Rar-
ity
Commun-
ity
T
R
CSR
Seed
mass
FGP
(%)
SE
FGP
GA
FGP (%)
SE
GA
Achillea atrata
Asteraceae
Sca
ROSC
1
5
css
0.242
92.80
3.44
Achillea clavennae*
Asteraceae
Sca
GRASS
1.5
5
css
0.447
70.40
3.71
Achillea clusiana*
Asteraceae
end
ROSC
1.5
5
css
0.326
81.60
6.76
Achillea moschata
Asteraceae
Com
ROSC
1
2
css
0.220
51.72
5.59
Aconitum napellus
Ranunculaceae
Com
TALL
3
3
ccs
3.100
0.00
0.00
1.80
1.11
Aconitum variegatum
Ranunculaceae
Com
TALL
3
5
ccs
0.843
0.00
0.00
0.00
0.00
Adenostyles alliariae*
Asteraceae
Com
TALL
2
3
ccs
1.258
45.60
5.60
Agrostis alpina
Poaceae
Sca
GRASS
1.5
4
crs
0.277
26.40
4.12
Agrostis rupestris*
Poaceae
Sca
GRASS
1
2
crs
0.108
60.80
1.50
Allium schoenoprasum*
Alliaceae
Sca
GRASS
2.5
4
crs
0.990
56.80
6.97
Androsace lactea
Primulaceae
Sca
GRASS
2.5
5
crs
1.013
0.00
0.00
15.58
4.66
Androsace obtusifolia
Primulaceae
Com
GRASS
1.5
2
css
0.730
0.00
0.00
0.80
0.80
Antennaria carpatica
Asteraceae
Sca
GRASS
1.5
3
crs
0.080
20.80
5.85
Antennaria dioica
Asteraceae
Sca
SHFO
2
2
crs
0.047
86.40
3.25
Anthericum ramosum
Anthericaceae
Sca
DRYM
3.5
4
crs
4.790
0.00
0.00
0.00
0.00
Anthyllis vulneraria ssp. Seed collection alpicola
Fabaceae
Com
GRASS
1.5
5
crs
2.366
12.80
1.50
0.00
0.00
Aposeris foetida
Asteraceae
Sca
SHFO
2.5
4
ccs
2.170
1.60
1.55
64.24
3.28
Arabidopsis halleri
Brassicaceae
Sca
MEAD
2.5
2
crs
0.160
77.60
7.96
Arabis ciliata*
Brassicaceae
Sca
ROSC
2.5
4
crs
0.111
92.80
5.43
Arabis stellulata
Brassicaceae
Sca
ROSC
1.5
5
rss
0.234
4.74
2.31
58.80
4.96
Arenaria marschlinsii*
Brassicaceae
Rar
WET
1
3
rss
n.a
95.20
1.50
Armeria alpina*
Plumbaginaceae
Sca
ROSC
1
2
css
2.206
18.40
3.71
Arnica montana
Asteraceae
Com
MEAD
2
3
crs
1.670
82.40
5.00
Artemisia mutellina*
Asteraceae
Sca
ROSC
1
3
rss
0.280
64.80
11.83
Aster alpinus
Asteraceae
Sca
GRASS
2
3
crs
1.350
80.80
3.44
Astragalus australis
Fabaceae
Rar
GRASS
1.5
4
ccs
3.110
15.20
2.33
Athamantha cretensis
Apiaceae
Sca
ROSC
2
5
css
2.150
1.60
0.98
Atocion rupestre*
Caryophyllaceae
Com
ROSC
2.5
2
css
0.074
43.20
6.50
Bartsia alpina
Orobanchaceae
Com
GRASS
2
3
rss
0.425
0.00
0.00
59.38
3.25
Blysmus compressus
Poaceae
Sca
WET
3
4
crs
0.905
10.40
3.49
0.00
0.00
Bothriochloa ischaemum
Poaceae
Sca
DRYM
4.5
3
crs
0.600
42.40
0.98
Buphthalmum salicifolium*
Asteraceae
Com
DRYM
3
4
crs
0.700
61.60
3.25
Calluna vulgaris
Ericaceae
Com
SHFO
2.5
1
css
0.030
52.80
5.43
Campanula alpina
Campanulaceae
Rar
ROSC
2
2
css
0.333
0.80
0.00
58.13
3.94
Campanula barbata*
Campanulaceae
Com
GRASS
2
2
crs
0.050
49.60
4.66
Campanula cochleariifolia*
Campanulaceae
Com
ROSC
2
5
css
0.045
7.00
2.41
34.73
7.62
Campanula pulla
Campanulaceae
end
GRASS
1.5
4
css
0.132
1.60
0.98
0.00
0.00
Campanula rapunculus
Campanulaceae
Rar
DRYM
4.5
3
crs
0.040
43.20
3.44
Campanula scheuchzeri
Campanulaceae
Com
GRASS
1.5
3
crs
0.090
7.20
2.65
90.26
3.77
Carduus personata*
Asteraceae
Sca
TALL
2.5
4
ccr
1.709
91.20
2.94
Carex brachystachys
Cyperaceae
Sca
GRASS
2.5
5
css
0.496
0.00
0.00
0.00
0.00 12 3 Plant Ecol (2020) 221:1045–1067 1049 Table 1 continued
Species
Family
Rar-
ity
Commun-
ity
T
R
CSR
Seed
mass
FGP
(%)
SE
FGP
GA
FGP (%)
SE
GA
Carex curvula
Cyperaceae
Com
GRASS
1
2
ccs
2.120
0.00
0.00
0.00
0.00
Carex davalliana
Cyperaceae
Com
WET
3
4
css
0.839
8.00
3.58
42.14
11.85
Carex echinata
Cyperaceae
Com
WET
3
2
css
0.840
29.60
3.49
Carex ferruginea*
Cyperaceae
Com
GRASS
2
4
ccs
1.210
28.00
2.83
Carex firma
Cyperaceae
Com
GRASS
1.5
5
css
0.770
53.60
3.25
Carex frigida
Cyperaceae
Sca
WET
1.5
4
css
0.320
2.50
2.50
5.00
2.24
Carex nigra
Cyperaceae
Com
WET
2.5
2
css
0.760
0.00
0.00
0.00
0.00
Carex panicea
Cyperaceae
Com
WET
3
4
css
2.473
0.00
0.00
0.00
0.00
Carex parviflora*
Cyperaceae
Sca
GRASS
1
5
crs
0.450
7.20
1.96
79.68
4.94
Carex rostrata
Cyperaceae
Sca
WET
3
2
css
1.500
17.60
6.01
Carex vesicaria
Cyperaceae
Sca
WET
3.5
3
css
1.800
0.00
0.00
0.00
0.00
Carlina acaulis
Asteraceae
Com
DRYM
3
2
crs
3.400
100.00
0.00
Centaurea pseudophrygia
Asteraceae
Sca
DRYM
2.5
3
ccs
n.a
38.40
3.49
Centaurea stoebe
Asteraceae
Sca
DRYM
4.5
4
crs
1.404
71.20
5.57
Centaurium erythraea
Gentianaceae
Sca
DRYM
3.5
3
rrs
0.010
92.80
2.33
Cerastium carinthiacum
Caryophyllaceae
Sca
ROSC
1.5
5
css
0.200
32.80
4.63
Cerastium eriophorum
Caryophyllaceae
Sca
GRASS
1.5
4
css
0.288
54.40
8.82
Cervaria rivini
Apiaceae
Sca
TALL
4
4
ccs
4.020
0.80
0.80
28.10
6.00
Chaerophyllum aureum
Apiaceae
Sca
TALL
3
4
ccc
8.100
0.00
0.00
0.00
0.00
Chaerophyllum villarsii
Apiaceae
Sca
TALL
2
3
ccs
7.666
0.00
0.00
0.00
0.00
Cirsium carniolicum
Asteraceae
Rar
TALL
2.5
4
ccs
5.400
40.80
0.80
Cirsium erisithales
Asteraceae
Rar
TALL
2.5
4
ccs
1.956
55.20
4.08
Cirsium spinosissimum
Asteraceae
Sca
TALL
1.5
3
ccs
2.168
49.60
3.71
Comarum palustre
Rosaceae
Rar
WET
3
2
ccs
0.320
1.60
1.60
6.50
1.57
Crepis alpestris
Asteraceae
Sca
MEAD
2
5
ccs
2.319
85.60
3.71
Crepis aurea
Asteraceae
Com
MEAD
2
3
ccs
0.856
26.40
8.73
Crepis conyzifolia*
Asteraceae
Rar
MEAD
2
2
ccs
2.403
87.20
3.44
Crepis jacquinii
Asteraceae
Sca
ROSC
1.5
5
css
0.897
92.00
2.83
Crepis kerneri
Asteraceae
Sca
GRASS
1.5
5
css
n.a
86.40
0.98
Crepis paludosa
Asteraceae
Sca
WET
3
3
ccs
0.652
59.20
4.27
Crepis pyrenaica*
Asteraceae
Sca
TALL
2.5
4
ccs
3.097
36.80
6.37
Dactylorhiza incarnata
Orchidaceae
Sca
WET
3.5
4
crs
0.001
0.00
0.00
0.00
0.00
Dianthus alpinus*
Caryophyllaceae
end
ROSC
1
5
css
0.580
47.20
6.12
Dianthus superbus*
Caryophyllaceae
Sca
WET
3.5
4
crs
0.583
68.00
5.06
Dianthus sylvestris
Caryophyllaceae
Sca
ROSC
3
3
crs
0.850
48.00
4.38
Drosera rotundifolia
Droseraceae
Com
WET
3
1
sss
0.020
0.00
0.00
12.00
3.00
Dryas octopetala
Rosaceae
Com
GRASS
1.5
5
crs
0.372
83.20
4.96
Epilobium alpestre*
Onagraceae
Com
TALL
2.5
4
ccs
0.267
96.80
1.50
Epilobium anagallidifolium
Onagraceae
Sca
WET
1.5
3
css
0.060
92.00
2.53
Epilobium hirsutum
Onagraceae
Sca
TALL
3.5
4
ccc
0.100
99.20
0.80
Epilobium nutans*
Onagraceae
Rar
WET
2
2
crs
0.075
48.00
7.38
Epilobium palustre
Onagraceae
Sca
WET
3
3
crs
0.090
4.80
2.94
72.72
5.01 Plant Ecol (2020) 221:1045–1067 1050 Table 1 continued
Species
Family
Rar-
ity
Commun-
ity
T
R
CSR
Seed
mass
FGP
(%)
SE
FGP
GA
FGP (%)
SE
GA
Epipactis palustris*
Orchidaceae
Rar
WET
3.5
4
crs
0.004
0.00
0.00
0.00
0.00
Erigeron atticus
Asteraceae
Rar
GRASS
2
2
crs
0.234
65.60
7.65
Erigeron glabratus*
Asteraceae
Com
GRASS
1.5
5
css
0.199
58.40
5.31
Erigeron uniflorus*
Asteraceae
Sca
GRASS
1
3
css
0.180
87.20
5.85
Eriophorum latifolium
Cyperaceae
com
WET
3
4
css
0.654
0.80
0.80
1.67
1.67
Eriophorum vaginatum
Cyperaceae
rar
WET
2.5
1
css
0.577
83.20
7.42
Erysimum rhaeticum
Brassicaceae
sca
GRASS
3
2
crs
n.a
76.00
6.07
Euphrasia officinalis
ssp. Seed collection picta
Orobanchaceae
sca
GRASS
2
3
rrs
n.a
0.80
0.80
11.52
3.91
Festuca picturata*
Poaceae
sca
GRASS
1.5
2
ccs
n.a
47.20
6.50
Festuca pseudodura
Poaceae
com
GRASS
1.5
2
ccs
0.334
29.60
9.26
Festuca varia
Poaceae
com
GRASS
1.5
2
ccs
1.150
10.40
4.12
27.14
12.33
Galium anisophyllon
Rubiaceae
com
GRASS
2
3
ccs
0.365
27.20
4.27
Gentiana acaulis
Gentianaceae
sca
GRASS
1.5
2
css
0.452
0.00
0.00
38.70
2.59
Gentiana bavarica
Gentianaceae
sca
WET
1.5
3
css
0.140
0.00
0.00
52.40
8.63
Gentiana clusii
Gentianaceae
sca
GRASS
1.5
5
css
0.410
0.00
0.00
47.50
2.13
Gentiana lutea
Gentianaceae
rar
GRASS
2.5
4
ccs
1.500
0.80
0.80
57.14
7.23
Gentiana nivalis
Gentianaceae
sca
GRASS
1.5
3
css
0.081
0.00
0.00
21.56
4.64
Gentiana pannonica
Gentianaceae
sca
GRASS
2
3
ccs
0.460
33.60
4.12
Gentiana pumila
Gentianaceae
sca
WET
1.5
5
css
0.167
0.00
0.00
70.66
2.23
Gentiana utriculosa
Gentianaceae
sca
WET
2.5
4
crs
0.060
0.00
0.00
77.60
6.01
Gentianella aspera*
Gentianaceae
com
GRASS
2.5
4
crs
0.219
0.00
0.00
12.38
2.24
Globularia cordifolia*
Globulariaceae
com
GRASS
2.5
5
css
0.530
32.80
3.44
Globularia nudicaulis
Globulariaceae
rar
GRASS
2
4
ccs
0.720
15.20
3.88
Gnaphalium sylvaticum
Asteraceae
sca
SHFO
3
2
crs
0.061
94.40
0.98
Gypsophila repens
Caryophyllaceae
sca
ROSC
2
5
crs
0.710
88.86
2.35
Hedysarum hedysaroides
Fabaceae
sca
SHFO
1.5
4
ccs
4.125
55.20
6.12
Helianthemum nummularium
Cistaceae
sca
GRASS
4.5
4
ccs
1.100
1.60
1.60
2.40
0.98
Helictotrichon parlatorei
Poaceae
sca
GRASS
1.5
4
css
1.789
0.80
0.80
1.60
0.98
Heracleum austriacum
Apiaceae
sca
SHFO
2.5
4
crs
5.429
0.00
0.00
9.72
2.62
Hieracium hoppeanum
Asteraceae
sca
SHFO
2.5
2
crs
0.276
57.60
6.27
Hieracium intybaceum
Asteraceae
sca
ROSC
2
2
css
0.647
7.20
2.33
68.80
3.88
Hieracium maculatum
Asteraceae
rar
DRYM
3.5
3
css
0.523
47.20
7.09
Hieracium racemosum
Asteraceae
rar
SHFO
4.5
2
css
0.511
56.80
6.62
Hieracium sphaerocephalum
Asteraceae
sca
GRASS
1.5
2
css
0.128
29.60
3.49
Hieracium umbellatum
Asteraceae
sca
SHFO
4
2
css
0.530
64.80
9.75
Hieracium villosum
Asteraceae
com
GRASS
2
5
css
0.445
0.80
0.80
63.36
4.85
Homogyne alpina
Asteraceae
com
SHFO
2
2
crs
0.946
30.40
7.86
Homogyne discolor*
Asteraceae
rar
GRASS
1.5
4
crs
0.891
13.89
4.00
66.00
2.00
Hornungia alpina
Brassicaceae
com
ROSC
1
4
rss
0.318
37.60
2.99
Hypochaeris uniflora*
Asteraceae
sca
MEAD
2
2
crs
3.800
60.80
8.14
Juncus trifidus
Juncaceae
sca
GRASS
1.5
1
css
0.176
0.00
0.00
14.56
2.23 12 3 Plant Ecol (2020) 221:1045–1067 1051 Table 1 continued
Species
Family
Rar-
ity
Commun-
ity
T
R
CSR
Seed
mass
FGP
(%)
SE
FGP
GA
FGP (%)
SE
GA
Knautia maxima
Dipsacaceae
com
TALL
3
3
ccs
3.225
12.00
3.58
96.36
2.23
Kobresia myosuroides
Cyperaceae
sca
GRASS
1.5
3
ccs
0.655
0.80
0.80
19.68
4.01
Kobresia simpliciuscula*
Cyperaceae
rar
WET
1.5
4
css
0.491
88.00
3.10
Koeleria hirsuta
Poaceae
rar
ROSC
1.5
2
css
n.a
85.33
6.67
Laserpitium halleri
Apiaceae
sca
GRASS
2
2
ccs
6.829
0.00
0.00
1.00
1.00
Laserpitium krapfii*
Apiaceae
rar
SHFO
3
3
ccs
n.a
0.00
0.00
0.00
0.00
Laserpitium latifolium
Apiaceae
sca
SHFO
3
4
ccs
9.680
0.00
0.00
0.00
0.00
Leontopodium alpinum
Asteraceae
rar
GRASS
1.5
4
ccs
0.127
74.40
3.25
Leucanthemopsis alpina
Asteraceae
com
GRASS
1
2
ccs
0.380
29.29
3.45
Leucanthemum atratum
Asteraceae
end
GRASS
1.5
5
css
0.578
95.20
1.50
Leucanthemum vulgare
Asteraceae
com
MEAD
3
3
crs
0.420
92.80
1.96
Linaria alpina
Antirrhinaceae
com
ROSC
1.5
4
rrs
0.278
0.00
0.00
89.38
1.69
Linum alpinum
Linaceae
sca
GRASS
2
4
crs
1.883
52.00
3.10
Luzula alpina
Juncaceae
sca
GRASS
1.5
2
crs
0.257
86.40
2.71
Luzula alpinopilosa*
Juncaceae
sca
GRASS
1
2
css
0.200
54.40
2.99
Luzula glabrata*
Juncaceae
sca
GRASS
1.5
4
css
0.281
6.40
3.49
Luzula spicata
Juncaceae
sca
GRASS
1.5
2
crs
0.340
0.00
0.00
0.80
0.80
Luzula sudetica*
Juncaceae
rar
WET
2
2
crs
0.373
67.20
4.63
Menyanthes trifoliata
Menyanthaceae
rar
WET
3
3
css
2.400
0.00
0.00
15.20
2.65
Minuartia austriaca*
Caryophyllaceae
sca
ROSC
1.5
5
css
0.290
0.80
0.80
0.76
0.80
Minuartia cherlerioides
Caryophyllaceae
rar
ROSC
1
5
css
0.061
10.40
6.76
82.24
5.66
Minuartia sedoides
Caryophyllaceae
sca
GRASS
1
3
css
0.258
7.20
2.65
Moehringia ciliata
Caryophyllaceae
sca
ROSC
1
4
css
0.387
0.00
0.00
1.60
0.98
Molinia arundinacea
Poaceae
sca
SHFO
3.5
4
css
n.a
4.00
1.79
12.75
2.66
Mutellina adonidifolia*
Apiaceae
com
SHFO
1.5
2
ccs
2.185
0.00
0.00
0.00
0.00
Noccaea crantzii
Brassicaceae
end
GRASS
2
5
crs
0.828
7.20
1.50
90.30
3.73
Oreochloa disticha*
Poaceae
sca
GRASS
1
1
crs
0.373
20.00
4.90
Orobanche flava
Orobanchaceae
sca
TALL
2.5
4
rss
0.006
0.00
0.00
0.00
0.00
Orobanche salviae
Orobanchaceae
rar
SHFO
3
4
rss
n.a
0.00
0.00
0.00
0.00
Oxytropis montana*
Fabaceae
sca
GRASS
1.5
5
ccs
3.253
6.40
0.98
0.00
0.00
Pachypleurum mutellinoides
Apiaceae
rar
GRASS
1
3
crs
1.020
0.80
0.80
7.26
2.19
Papaver alpinum*
Papaveraceae
sca
ROSC
1.5
4
sss
0.143
15.20
3.44
84.15
1.64
Parnassia palustris*
Parnassiaceae
sca
WET
2
4
css
0.030
60.00
6.69
Pedicularis palustris
Orobanchaceae
rar
WET
3
3
css
0.829
0.00
0.00
8.00
2.19
Pedicularis portenschlagii
Orobanchaceae
end
GRASS
1
3
css
0.646
1.60
0.98
81.28
3.76
Pedicularis recutita
Orobanchaceae
sca
TALL
2
4
css
0.820
0.00
0.00
57.76
4.22
Pedicularis rostratocapitata
Orobanchaceae
com
GRASS
1.5
5
css
1.055
0.00
0.00
53.50
3.32
Petasites paradoxus
Asteraceae
com
TALL
2
5
ccs
0.476
2.40
1.60
Peucedanum oreoselinum*
Apiaceae
sca
SHFO
4.5
3
ccs
4.520
56.80
6.37
Peucedanum ostruthium
Apiaceae
sca
TALL
2
3
ccc
1.208
26.40
4.12
Peucedanum verticillare
Apiaceae
rar
TALL
4
4
ccs
n.a
72.00
5.51
Phleum hirsutum
Poaceae
sca
TALL
2
4
ccs
0.169
89.60
2.99 3 Plant Ecol (2020) 221:1045–1067 1052 Table 1 continued
Species
Family
Rar-
ity
Commun-
ity
T
R
CSR
Seed
mass
FGP
(%)
SE
FGP
GA
FGP (%)
SE
GA
Phleum rhaeticum
Poaceae
com
TALL
2
3
ccs
0.464
67.13
5.34
Phyteuma betonicifolium
Campanulaceae
sca
TALL
2
2
ccs
0.045
3.20
1.50
67.82
14.11
Phyteuma confusum
Campanulaceae
sca
GRASS
1.5
2
css
0.043
8.00
2.53
89.78
4.10
Phyteuma globulariifolium*
Campanulaceae
sca
GRASS
1
2
css
0.052
3.20
1.50
81.23
4.61
Phyteuma hemisphaericum
Campanulaceae
com
GRASS
1.5
1
css
0.101
21.60
4.49
Phyteuma orbiculare*
Campanulaceae
sca
MEAD
2.5
4
css
0.215
0.00
0.00
74.97
1.64
Phyteuma ovatum
Campanulaceae
sca
TALL
2
3
ccs
0.180
0.00
0.00
84.68
4.56
Pinguicula leptoceras*
Lentibulariaceae
rar
WET
1.5
3
sss
0.020
4.00
2.19
66.15
5.79
Plantago alpina
Plantaginaceae
rar
GRASS
1.5
2
css
0.600
42.40
3.71
Poa alpina
Poaceae
com
GRASS
1.5
3
crs
0.354
48.80
7.74
Poa chaixii
Poaceae
rar
TALL
2.5
2
ccs
0.503
5.60
2.71
95.00
2.74
Poa minor
Poaceae
sca
GRASS
1
5
css
0.299
84.80
6.37
Potentilla aurea*
Rosaceae
com
MEAD
1.5
2
ccs
0.368
35.20
3.88
Potentilla brauneana
Rosaceae
rar
GRASS
1
5
css
0.615
0.80
0.80
46.96
4.43
Potentilla clusiana*
Rosaceae
end
ROSC
1.5
5
css
0.373
25.60
5.74
Potentilla frigida*
Rosaceae
rar
GRASS
1
2
css
0.200
3.20
0.80
6.93
2.26
Potentilla grandiflora*
Rosaceae
sca
GRASS
1.5
2
ccs
0.449
8.80
3.20
48.00
13.80
Primula auricula
Primulaceae
sca
GRASS
1.5
5
css
0.250
0.00
0.00
87.14
3.18
Primula clusiana*
Primulaceae
end
GRASS
1.5
5
css
0.315
32.00
8.10
Primula elatior
Primulaceae
com
MEAD
3
4
ccs
0.800
0.00
0.00
35.40
3.28
Primula farinosa
Primulaceae
rar
WET
2
4
css
0.060
5.60
2.71
97.60
1.60
Primula glutinosa
Primulaceae
sca
GRASS
1
2
css
0.121
5.60
1.60
16.00
8.00
Primula minima*
Primulaceae
com
GRASS
1.5
2
sss
0.123
25.38
2.96
Pulsatilla alpina*
Ranunculaceae
sca
SHFO
2
4
crs
6.680
0.00
0.00
69.10
8.29
Pulsatilla alpina ssp. Seed collection Apiifolia*
Ranunculaceae
sca
SHFO
2
2
crs
n.a
0.00
0.00
63.00
13.00
Ranunculus alpestris
Ranunculaceae
com
GRASS
1.5
4
css
0.278
18.40
6.40
Ranunculus montanus
Ranunculaceae
com
MEAD
2
4
ccs
1.547
0.00
0.00
54.76
2.72
Rhododendron ferrugineum*
Ericaceae
com
SHFO
2
2
ccs
0.027
38.40
3.71
Rhododendron hirsutum*
Ericaceae
com
SHFO
2
4
ccs
0.029
38.40
4.31
Rhodothamnus
chamaecistus*
Ericaceae
sca
SHFO
2
4
ccs
0.040
0.00
0.00
84.00
7.04
Salix helvetica*
Salicaceae
sca
SHFO
1.5
2
ccs
n.a
0.00
0.00
Salix pentandra*
Salicaceae
rar
SHFO
2.5
3
ccc
0.096
91.20
2.94
Saponaria pumila
Caryophyllaceae
com
GRASS
1
2
css
1.210
0.00
0.00
10.16
2.59
Saussurea pygmaea
Asteraceae
rar
GRASS
1
4
css
4.750
95.20
2.33
Saxifraga adscendens
Saxifragaceae
rar
ROSC
1.5
4
rss
0.011
39.00
2.52
Saxifraga aizoides
Saxifragaceae
com
WET
2
4
css
0.054
83.00
1.91
Saxifraga androsacea
Saxifragaceae
sca
ROSC
1
4
sss
0.036
0.00
0.00
25.00
4.50
Saxifraga aspera
Saxifragaceae
sca
ROSC
2
2
css
0.015
0.00
0.00
0.00
0.00
Saxifraga biflora
Saxifragaceae
sca
ROSC
1
4
css
0.016
11.33
7.33
63.53
6.70
Saxifraga bryoides
Saxifragaceae
com
ROSC
1
2
css
0.042
0.00
0.00
2.17
2.17
Saxifraga burseriana
Saxifragaceae
rar
ROSC
2
5
sss
0.052
0.00
0.00
33.03
2.97 12 3 Plant Ecol (2020) 221:1045–1067 1053 Table 1 continued
Species
Family
Rar-
ity
Commun-
ity
T
R
CSR
Seed
mass
FGP
(%)
SE
FGP
GA
FGP (%)
SE
GA
Saxifraga caesia
Saxifragaceae
com
GRASS
1.5
5
sss
0.024
14.00
2.58
55.60
6.58
Saxifraga exarata
Saxifragaceae
rar
ROSC
1
2
sss
0.055
0.00
0.00
4.00
1.63
Saxifraga hostii
Saxifragaceae
end
GRASS
2
5
sss
0.055
0.00
0.00
12.87
4.04
Saxifraga moschata
Saxifragaceae
com
ROSC
1
4
sss
0.040
22.00
4.16
Saxifraga mutata
Saxifragaceae
sca
ROSC
2.5
4
sss
0.030
38.00
7.75
Saxifraga oppositifolia
Saxifragaceae
com
ROSC
1
4
css
0.014
0.00
0.00
29.33
3.71
Saxifraga paniculata
Saxifragaceae
com
ROSC
2
4
sss
0.040
0.00
0.00
19.35
2.89
Saxifraga rotundifolia
Saxifragaceae
sca
TALL
2.5
3
ccs
0.039
96.00
2.83
Saxifraga rudolphiana
Saxifragaceae
end
ROSC
1
4
css
0.070
80.00
1.63
Saxifraga sedoides
Saxifragaceae
sca
ROSC
1.5
5
css
0.077
0.00
0.00
15.13
1.85
Saxifraga seguieri
Saxifragaceae
rar
ROSC
1
2
css
0.037
1.00
1.00
47.36
4.37
Saxifraga squarrosa
Saxifragaceae
end
ROSC
2
5
sss
0.029
29.00
8.70
Saxifraga stellaris
Saxifragaceae
com
WET
2
3
css
0.040
93.00
3.42
Saxifraga tridactylites
Saxifragaceae
rar
DRYM
4.5
4
rrs
0.011
0.00
0.00
54.48
10.11
Scabiosa lucida*
Dipsacaceae
sca
GRASS
1.5
4
crs
1.841
52.80
6.37
Scorzonera humilis
Asteraceae
sca
WET
3.5
4
ccs
4.068
3.20
1.50
Scorzoneroides crocea
Asteraceae
rar
MEAD
2
3
crs
2.198
51.20
4.27
Scorzoneroides helvetica
Asteraceae
com
MEAD
1.5
2
crs
1.570
73.60
2.04
Sedum album
Crassulaceae
com
ROSC
3
4
sss
0.040
72.00
5.51
Senecio abrotanifolius ssp. Seed collection If FGP was below 15%, a treatment with gibberellic acid (GA) was
applied Elevation values (T) 1–1.5 = nival-alpin, 2–2.5 = subalpine, 3–4.5 = montane-colline. Bedrock values (R) 1 = acidic, 2–3 = acidic-
neutral, 4–5 = neutral-alkaline. Stress tolerant (S) = SSS, CSS, RSS; competitive (C) = CCC, CCS, CCR; intermediate strategists
CSR; ruderal (R) = RRR, RRS, RRC Elevation values (T) 1–1.5 = nival-alpin, 2–2.5 = subalpine, 3–4.5 = montane-colline. Bedrock values (R) 1 = acidic, 2–3 = acidic-
neutral, 4–5 = neutral-alkaline. Stress tolerant (S) = SSS, CSS, RSS; competitive (C) = CCC, CCS, CCR; intermediate strategists
CSR; ruderal (R) = RRR, RRS, RRC com common, sca scattered, rar rare, end endemic, ROSC rocky habitats, scree sites, GRASS alpine grasslands above treeline, WET
bogs, mires, wetlands, SHFO scrub heaths, understorey of forests, subalpine meadows, TALL tall forbs, DRYM dry meadows of low
elevations, MEAD meadows of the colline and montane belt, n.a. not available, SE Standard error to the protocol of ENSCONET (2009b). Due to former
experiences with alpine species (Schwienbacher and
Erschbamer 2001; Erschbamer and Pfattner 2002) and
due to technical reasons, we did not stratify the seeds. The only exceptions were Saxifraga species; for them
we applied cold-wet stratification in the fridge
at ? 4 C for 12 weeks. Seed collection abrotanifolius*
Asteraceae
sca
SHFO
2
4
css
1.989
53.60
8.16
Senecio abrotanifolius ssp. tiroliensis
Asteraceae
sca
SHFO
1.5
2
css
n.a
20.80
2.65
Senecio doronicum
Asteraceae
sca
GRASS
1.5
4
css
1.800
13.60
4.49
65.34
6.80
Seseli libanotis
Apiaceae
sca
DRYM
3
4
ccs
1.309
8.80
3.20
28.80
2.94
Sesleria ovata
Poaceae
sca
ROSC
1
5
css
0.008
0.00
0.00
Silene acaulis ssp. exscapa
Caryophyllaceae
sca
GRASS
1
2
css
0.419
32.00
6.57
Silene acaulis ssp. Seed collection longiscapa*
Caryophyllaceae
com
GRASS
1
4
css
0.379
39.20
4.80
Silene nutans
Caryophyllaceae
com
SHFO
3
3
ccs
0.368
69.60
6.01
Silene vulgaris*
Caryophyllaceae
com
MEAD
3
3
crs
1.535
81.60
9.17
Soldanella alpina
Primulaceae
com
SHFO
2
3
css
0.240
83.20
1.50
Soldanella pusilla*
Primulaceae
com
GRASS
1.5
2
css
0.069
84.00
2.83
Solidago virgaurea*
Asteraceae
com
SHFO
3.5
3
crs
0.690
75.86
4.96
Tofieldia calyculata*
Tofieldiaceae
sca
WET
2.5
4
css
0.030
74.62
6.26
Tofieldia pusilla*
Tofieldiaceae
rar
WET
1.5
3
css
0.035
92.80
2.33
Trichophorum alpinum
Cyperaceae
rar
WET
3
2
css
0.182
0.00
0.00
1.67
3.73
Trichophorum cespitosum*
Cyperaceae
sca
WET
2.5
3
css
0.584
0.00
0.00
0.00
0.00
Trifolium badium
Fabaceae
sca
GRASS
2
4
crs
0.750
55.20
3.44
Triglochin palustris*
Juncaginaceae
sca
WET
2.5
4
sss
1.761
54.40
5.60
Trollius europaeus
Ranunculaceae
sca
MEAD
2.5
3
ccs
0.810
0.80
0.80
41.50
2.01
Valeriana celtica*
Valerianaceae
end
GRASS
1
2
css
0.675
0.00
0.00
Valeriana elongata*
Valerianaceae
end
ROSC
2
5
crs
n.a
9.60
2.40 123 3 Plant Ecol (2020) 221:1045–1067 1054 Table 1 continued
Species
Family
Rar-
ity
Commun-
ity
T
R
CSR
Seed
mass
FGP
(%)
SE
FGP
GA
FGP (%)
SE
GA
Valeriana montana
Valerianaceae
com
ROSC
2
5
css
1.103
4.80
2.94
35.80
6.53
Valeriana supina
Valerianaceae
end
ROSC
1
5
css
1.232
3.20
1.50
Valeriana tripteris
Valerianaceae
com
TALL
2.5
3
css
0.926
7.20
2.33
31.10
5.03
Veratrum album
Melanthiaceae
sca
TALL
2.5
4
ccs
7.060
0.00
0.00
3.08
1.28
Veronica alpina
Antirrhinaceae
com
GRASS
1.5
2
css
0.040
80.00
5.66
Veronica aphylla
Antirrhinaceae
com
GRASS
1.5
5
css
0.099
0.00
0.00
3.20
2.33
Veronica bellidioides*
Antirrhinaceae
sca
GRASS
1
1
css
0.110
0.80
0.80
68.46
4.63
Veronica fruticans
Antirrhinaceae
sca
GRASS
1.5
3
css
0.124
1.80
0.98
4.06
2.19 Table 1 continued
Species
Family
Rar-
ity
Commun-
ity
T
R
CSR
Seed
mass
FGP
(%)
SE
FGP
GA
FGP (%)
SE
GA
Valeriana montana
Valerianaceae
com
ROSC
2
5
css
1.103
4.80
2.94
35.80
6.53
Valeriana supina
Valerianaceae
end
ROSC
1
5
css
1.232
3.20
1.50
Valeriana tripteris
Valerianaceae
com
TALL
2.5
3
css
0.926
7.20
2.33
31.10
5.03
Veratrum album
Melanthiaceae
sca
TALL
2.5
4
ccs
7.060
0.00
0.00
3.08
1.28
Veronica alpina
Antirrhinaceae
com
GRASS
1.5
2
css
0.040
80.00
5.66
Veronica aphylla
Antirrhinaceae
com
GRASS
1.5
5
css
0.099
0.00
0.00
3.20
2.33
Veronica bellidioides*
Antirrhinaceae
sca
GRASS
1
1
css
0.110
0.80
0.80
68.46
4.63
Veronica fruticans
Antirrhinaceae
sca
GRASS
1.5
3
css
0.124
1.80
0.98
4.06
2.19
Community affiliation, temperature value (T), soil reaction value (R) and CSR strategy type were extracted from Landolt et al. Seed collection (2010)
and grouped as described in Material and Methods. Seed mass (mg/seed) was determined by seed collectors or extracted from data
bases. Germination was tested in a growth cabinet using a standard protocol with one alternating temperature 25/15 C or, if marked
with asterisks (*) with 20/10 C. Final germination percentage (FGP %) is calculated as germinated seeds from the initial number of
seeds laid per species (see Supplementary Material Table S1). If FGP was below 15%, a treatment with gibberellic acid (GA) was
applied
Elevation values (T) 1–1.5 = nival-alpin, 2–2.5 = subalpine, 3–4.5 = montane-colline. Bedrock values (R) 1 = acidic, 2–3 = acidic-
neutral, 4–5 = neutral-alkaline. Stress tolerant (S) = SSS, CSS, RSS; competitive (C) = CCC, CCS, CCR; intermediate strategists
CSR; ruderal (R) = RRR, RRS, RRC
com common, sca scattered, rar rare, end endemic, ROSC rocky habitats, scree sites, GRASS alpine grasslands above treeline, WET
bogs, mires, wetlands, SHFO scrub heaths, understorey of forests, subalpine meadows, TALL tall forbs, DRYM dry meadows of low
elevations, MEAD meadows of the colline and montane belt, n.a. not available, SE Standard error Community affiliation, temperature value (T), soil reaction value (R) and CSR strategy type were extracted from Landolt et al. (2010)
and grouped as described in Material and Methods. Seed mass (mg/seed) was determined by seed collectors or extracted from data
bases. Germination was tested in a growth cabinet using a standard protocol with one alternating temperature 25/15 C or, if marked
with asterisks (*) with 20/10 C. Final germination percentage (FGP %) is calculated as germinated seeds from the initial number of
seeds laid per species (see Supplementary Material Table S1). If FGP was below 15%, a treatment with gibberellic acid (GA) was
applied Community affiliation, temperature value (T), soil reaction value (R) and CSR strategy type were extracted from Landolt et al. (2010)
and grouped as described in Material and Methods. Seed mass (mg/seed) was determined by seed collectors or extracted from data
bases. Germination was tested in a growth cabinet using a standard protocol with one alternating temperature 25/15 C or, if marked
with asterisks (*) with 20/10 C. Final germination percentage (FGP %) is calculated as germinated seeds from the initial number of
seeds laid per species (see Supplementary Material Table S1). Germination protocol The germination tests were carried out at the labora-
tory in Innsbruck, following a standard protocol. Tests
were done on Petri dishes filled with three layers of
filter paper moistened with deionized water. In each
Petri dish, 25 seeds were laid and five Petri dishes per
species were prepared. Depending on seed availabil-
ity, either the number of seeds laid per Petri dish or the
number of Petri dishes was adjusted (total numbers of
seed laid are given in Supplementary Material
Table S1). The Petri dishes were then put in a growth
cabinet
(SANYO
MLR-350H,
Sanyo
Electric
Biomedical
Co
Ltd,
Japan)
under
16 h day
(20,000 l9) and 8 h night (0 l9) regime with 60%
relative air humidity. The maximum photosynthetic
photon flux density in the growth cabinet was
180 lmol m-2 s-1. Initially, a day/night temperature
regime of 20/10 C was used (see species with * in
Table 1) according to former germination protocols
for alpine species (Schwienbacher and Erschbamer
2001;
Schwienbacher et al. 2012;
Walder
and
Erschbamer 2015). Later, the setting was changed to
a day/night temperature regime of 25/15 C according The number of germinated seeds was counted every
3–4 days and germinated seeds (radicle emergence at
least half as long as the seed) were removed. Seeds
infested with fungi were subjected to a pressure test
using tweezers and, in the case of softness, removed
from the Petri dish and counted as non-viable. Using
the same method, non-germinated seeds were checked
for vitality at the end of the test period and were
counted as non-viable if not passing the test. One
germination period lasted for 30 days. The final
germination percentage (FGP) after 30 days was
calculated as germinated seeds from the initial number
of seeds laid, pooled from all Petri dishes, representing
a mean per species. Seeds of species with no or B 15% germination
were soaked in gibberellic acid (0.001 molar GA3,
10 ml) for 3 days in the growth cabinet to stimulate 123 Plant Ecol (2020) 221:1045–1067 1055 2–3 = aci-neu (weakly acidic, neutral to alkaline);
4–5 = neu-alk
(alkaline). CSR
strategies
were
grouped in S = stress tolerant (SSS, CSS, RSS);
C = competitive (CCC, CCS, CCR); CSR = interme-
diate strategists; R = ruderal (RRR, RRS, RRC). Germination protocol Further, seed mass (mg/seed) was either determined
by seed collectors, or extracted from the following
databases: KEW Seed Information Database (SID),
The LEDA Traitbase, BiolFlor Database, The VISTA
Plant Trait Database, Ecological Flora of the British
Isles. their germination, assuming non-deep physiological
dormancy (Baskin and Baskin 2014). After 3 days,
seeds were washed with deionized water and were
again laid out on Petri dishes with three layers of filter
paper in the growth cabinet. Final germination
percentage after GA3 treatment (GA FGP) was
determined after another 30 days. For species of the
families Cyperaceae and Apiaceae from the 2018
collection, parallel GA3 treatment to the standard tests
were performed due to the availability of enough
seeds. For 10 species with zero or B 15% FGP, GA3
treatment was not possible due to scarcity of seeds. Numbers of seeds laid, germinated seeds and non-
viable seeds are given in Supplementary Material
Table S1. Data analyses We defined FGP C 20.1% (i.e. species with interme-
diate and high germination) as successful proof of the
functioning of the standard protocol. For species with
FGP B 20.0%, we assumed that the standard protocol
did not work and therefore excluded these species
from further analyses. In addition, the strategy type R
was excluded from the analyses, because this group
contained only one species with FGP C 20.1% (Cen-
taurium erythraea). Using this reduced dataset, FGP
(species means) was analysed by a generalized linear
model (GLM) with a quasi-binomial setting and a
logit-link function. A binomial distribution is typically
used for proportion data such as germinated seeds
from seeds laid. The model showed an over-disper-
sion. Therefore, a quasi-binomial setting was used. This setting accounts for over-dispersion by means of
a dispersion parameter that allows additional variance
in the data and returns adjusted standard errors. Classifications Similarly to Alsos et al. (2013), species with FGP
B 10.0% were classified as very weak germinators,
FGP of 10.1–20.0% as low germinators, 20.1–50.0%
as intermediate ones and those with 50.1–100% as
species with high germination. Species rarity was
extracted from Fischer et al. (2008) and the following
classes were defined: com (common, species occur-
ring very frequently in the Eastern Alps); sca (scat-
tered, species not very frequent, but in some regions of
the Eastern Alps they may be locally common); rar
(rare); end (endemic for the North-Eastern or South
Alps). A classification of the test species addressing
their ecology was done using ecological indicator
values (i.e. temperature (T), soil reaction (R), CSR
strategy) and affiliation to plant communities provided
by (Landolt et al. 2010). The test species were grouped
in seven plant communities: ROSC = rocky habitats,
scree sites; GRASS = alpine grasslands above tree-
line; WET = bogs, mires, wetlands; SHFO = scrub
heaths, understorey of forests, subalpine meadows;
TALL = tall forbs; DRYM = dry meadows of low
elevations; MEAD = meadows of the colline and
montane belt. Temperature values (T) act as a proxy
for the elevation distribution of a species, they were
grouped as follows: 1–1.5 = niv-alp (nival, upper
alpine belt, lower alpine belt); 2–2.5 = subalpine
(subalpine belt, upper montane belt); 3–4.5 = mon-
coll (montane belt, colline belt). Ranges of soil
reaction (R) represent the bedrock on which the
species mostly occurs. These values were distin-
guished as: 1 = acidic (very acidic and acidic); The full model included all factors, i.e. rarity (sca,
com, rar, end), plant community (ROSC, GRASS,
WET, SHFO, TALL, DRYM, MEAD), elevation belt
(niv-alp, subalpine, mon-coll), bedrock type (acidic,
aci-neu, neu-alk), strategy type (S, C, CSR) and seed
mass. F-test was used to test the significance of the
factors in the model. Post hoc tests with Bonferroni-
Holm correction were applied to test for significances
between the levels of the factors. The significance
level was set to alpha 0.05; the model was verified via
diagnostic residual plots. Because only one factor
showed significance, the factors were ranked accord-
ing to least significance and were successively
excluded from the model to check the influence of
the factors on the model. Classifications This procedure was done to
prevent an over-fitting of the model, but as this did not
lead to any changes in factor level differences 123 12 3 3 Plant Ecol (2020) 221:1045–1067 1056 tested (Carlina acaulis) showed 100% FGP. Interme-
diate germination was observed in 17.6%, low germi-
nation was noted for 6.3%, and 19.6% had a weak
germination. Scattered species represented about half
of the species (48.6%, Fig. 1b), being the largest group
in this class. Species from alpine grasslands repre-
sented 35.7% of the species tested (Fig. 1c); 43.9%
were classified as species from the niv-alp elevation
belt (Fig. 1d). Of all species tested, only 2.8% were
classified as species on very acidic bedrock (Fig. 1e),
while the classes aci-neu and neu-alk contained 47.1%
and 50.2% of species, respectively. Half of the species
tested (50.6%) was classified as being stress tolerant
(S, Fig. 1f), and only 1.6% represented the ruderal
strategy type (R). Seed mass ranged from 0.001 to
9.68 mg/seed, with Dactylorhiza incarnata having the
lightest seeds and Laserpitium latifolium the heaviest. compared to the full model, the full model was chosen
as the final one. FGP for species from the families Asteraceae,
Poaceae, and Saxifragaceae are visually shown to
highlight the germination behaviour of these families. The analyses were performed in R (version 3.6.1, R
Core Team 2020) and RStudio (version 1.2.1568,
RStudio 2016) with the MASS package (Venables and
Ripley 2002). For data manipulation, we used the
packages tidyverse (Wickham 2017) and reshape
(Wickham 2018a), for descriptive statistics the pack-
age pastecs (Grosjean et al. 2018). For data visualiza-
tion, we used the packages ggplot2 (Wickham and
Winston 2019), gridExtra (Auguie and Antonov 2017)
and scales (Wickham 2018b). Results The mean FGP was analysed by a GLM using a
dataset of 124 species, comprising species with
FGP C 20.1%. Strategy type was the only significant
factor in the model (Table 2). However, post hoc tests
did not reveal any significances between the strategy
type factor levels. We could not find any statistically
significant
effects
of
rarity,
plant
community, In total, 255 species out of 39 families were tested. Regarding the germination classes, 74.9% germinated
under the standard protocol, and 25.1% failed to
germinate (Table 1, Fig. 1a). From the 255 species,
31.4% had a high FGP. Only one species among the Fig. 1 Species tested using the standard protocol (total 255
species) and classifications made in the study. The graphs
present the numbers of species per class: a Germination class
defined as zero: FGP = 0.0%, weak: FGP B 10.0%, low:
FGP C 10.1 B 20.0%,
intermediate:
FGP C 20.1 B 50.0%
and high: FGP C 50.1–100.0%. b Rarity (com common, scat
scattered, rar rare, end endemic). c Plant community (ROSC
rocky habitats, scree sites, GRASS alpine grasslands above
treeline, WET bogs, mires, wetlands, SHFO scrub heaths,
understorey of forests, subalpine meadows, TALL tall forbs,
DRYM dry meadows of low elevations, MEAD meadows of the
colline and montane belt). d Elevation belt (nival, subalpine,
montane-colline). e Bedrock type (acidic, acidic-neutral, acidic-
alkaline), and f Strategy type S stress tolerant, C competitive,
CSR intermediate, R ruderal strategy
1 treeline, WET bogs, mires, wetlands, SHFO scrub heaths,
understorey of forests, subalpine meadows, TALL tall forbs,
DRYM dry meadows of low elevations, MEAD meadows of the
colline and montane belt). d Elevation belt (nival, subalpine,
montane-colline). e Bedrock type (acidic, acidic-neutral, acidic-
alkaline), and f Strategy type S stress tolerant, C competitive,
CSR intermediate, R ruderal strategy Fig. 1 Species tested using the standard protocol (total 255
species) and classifications made in the study. The graphs
present the numbers of species per class: a Germination class
defined as zero: FGP = 0.0%, weak: FGP B 10.0%, low:
FGP C 10.1 B 20.0%,
intermediate:
FGP C 20.1 B 50.0%
and high: FGP C 50.1–100.0%. b Rarity (com common, scat
scattered, rar rare, end endemic). Results Species from the mon-coll elevation belt
had the highest mean FGP (67.4%) followed by
subalpine species (60.5%) and niv-alp species (58.5%;
Fig. 2c). Species classified to the acidic bedrock type
had the lowest FGP (52.5%), neu-alk the highest
(62.5%, Fig. 2d). Regarding strategy types (Fig. 1e),
C strategists had the lowest (56.7%) and intermediate
strategists the highest FGP (65.2%, Fig. 2e). Seed
mass (Fig. 2f) of species with FGP C 20.1% ranged
from
0.01 mg/seed
(Saxifraga
adscendens)
to
5.40 mg/seed (Cirsium carniolicum). elevation belt, bedrock type or seed mass on FGP. Consequently, post hoc tests on the single factors per
class resulted in non-significant differences. Common
and endemic species had the lowest mean FGP (55.4%
and 55.8%, respectively), and rare species the highest
(67.7%; Fig. 2a). Within plant communities (Fig. 2b),
the lowest FGP was found in GRASS species (54.7%),
the highest in the WET community class (72.5%). In
contrast to this high FGP, the WET class had the
highest proportion of non-germinating, weak and low
germinating species (i.e. 23 out of 38 species, Fig. 1c
and Fig. 2b). Species from the mon-coll elevation belt
had the highest mean FGP (67.4%) followed by
subalpine species (60.5%) and niv-alp species (58.5%;
Fig. 2c). Species classified to the acidic bedrock type
had the lowest FGP (52.5%), neu-alk the highest
(62.5%, Fig. 2d). Regarding strategy types (Fig. 1e),
C strategists had the lowest (56.7%) and intermediate
strategists the highest FGP (65.2%, Fig. 2e). Seed
mass (Fig. 2f) of species with FGP C 20.1% ranged
from
0.01 mg/seed
(Saxifraga
adscendens)
to
5.40 mg/seed (Cirsium carniolicum). Results c Plant community (ROSC
rocky habitats, scree sites, GRASS alpine grasslands above 12 123 Plant Ecol (2020) 221:1045–1067 1057 Table 2 Analyses of deviance-table listing the effects of tested factors on species mean FGP, comprising species with FGP C 20.1%
Factors
Df
Deviance
F value
p value
Label
Rarity
3
16.604
0.816
0.488
n.s
Plant community
6
56.765
1.395
0.224
n.s
Elevation belt
2
7.374
0.544
0.582
n.s
Bedrock type
2
15.817
1.166
0.316
n.s
Strategy type
3
42.312
3.119
0.049
*
Seed mass
1
3.201
0.472
0.494
n.s
A GLM with quasi-binomial error and logit-link function was used to analyse the data; F-test was used to test the significance of the
factors in the model
Dispersion parameter in the GLM was taken to be 6.8
Df degrees of freedom; n.s. not statistically significant
Significance levels are labelled as. p * 0.05, *p \ 0.05, **p \ 0.01, ***p \ 0.001 The most abundant species collected were from the
family Asteraceae (53 species, Fig. 4), whereby
86.8% had FGP C 20.1–100%, and 13.2% had low
and weak germination. None of the Asteraceae species
was noted to have zero germination using the standard
protocol. Poaceae species (17 species) germinated
with a similar pattern to Asteraceae species (Fig. 5). 58.8% had high to intermediate germination, 35.3%
had low and weak germination, and only one species
did not germinate under the standard protocol (Sesle-
ria ovata). Species among the family Saxifragaceae
(21 species) were cold stratified before the germina-
tion test. Here, 38.1% had FGP C 20.1%, and 47.6%
did not germinate under the standard protocol
(Fig. 6a). On 13 species a GA3 treatment was applied
(Fig. 6b), which enhanced the germination of 12
species, one did not germinate at all (Saxifraga
aspera). elevation belt, bedrock type or seed mass on FGP. Consequently, post hoc tests on the single factors per
class resulted in non-significant differences. Common
and endemic species had the lowest mean FGP (55.4%
and 55.8%, respectively), and rare species the highest
(67.7%; Fig. 2a). Within plant communities (Fig. 2b),
the lowest FGP was found in GRASS species (54.7%),
the highest in the WET community class (72.5%). In
contrast to this high FGP, the WET class had the
highest proportion of non-germinating, weak and low
germinating species (i.e. 23 out of 38 species, Fig. 1c
and Fig. 2b). Discussion In total, 74.9% of all investigated species germinated
using the standard protocol, and among them, 49% had
an intermediate to high FGP. If it is noted that half of
the species germinate, this is a fairly good result,
indicating that the collected seeds were in good
condition and able to germinate without pre-treatment
(except for the Saxifragaceae which were stratified). High-quality seeds are mandatory for ex situ seed
banking. Alsos et al. (2013) suggested to consider the
annual variation of seed production and to extend seed A treatment with gibberellic acid (GA3) was
applied to 44.7% out of 255 species (Fig. 3). Germi-
nation was stimulated in half of them (50.8% out of
114 species), resulting in GA FGPs of 20.1–97.6%. The other half of the GA3 tested species comprised
29.8%, classified as low and weak germinators, and
19.3% did not germinate (Fig. 3). Illustrations of GA
FGP of the species according to the classifications can
be found in the Supplementary Material Fig. S1. 12 3 3 Plant Ecol (2020) 221:1045–1067 1058 collections over several seasons to overcome possible
insufficient germination from one season. This would
further increase the genetic diversity of the stored
period that the necessary amount of seeds for seed
banking and germination trials cannot be obtained
within one year. A complete ex situ seed bank of rare
Fig. 2 Final germination percentage (FGP) of intermediate and
high germinating species (in total 124 species), tested using the
standard protocol and analysed via a GLM. Comparisons
between the levels of the classes a Rarity, b Plant community,
c Elevation belt, d Bedrock type and f Strategy type were made
with Post hoc tests and Bonferroni-Holm correction. For
abbreviations see Fig. 1 or Material and Methods. n number
of species per class. In e Seed mass was not available for 8
species. Error bars show confidence intervals (95%). p indicates
the p value label from the post hoc tests, n.s. not statistically
significant
(
) Fig. 2 Final germination percentage (FGP) of intermediate and
high germinating species (in total 124 species), tested using the
standard protocol and analysed via a GLM. Comparisons
between the levels of the classes a Rarity, b Plant community,
c Elevation belt, d Bedrock type and f Strategy type were made
with Post hoc tests and Bonferroni-Holm correction. For abbreviations see Fig. 1 or Material and Methods. n number
of species per class. Discussion In e Seed mass was not available for 8
species. Error bars show confidence intervals (95%). p indicates
the p value label from the post hoc tests, n.s. not statistically
significant period that the necessary amount of seeds for seed
banking and germination trials cannot be obtained
within one year. A complete ex situ seed bank of rare
and endemic species seems therefore hardly possible. Nonetheless, seed banks make an important contribu-
tion to the conservation of these species, but strict collections over several seasons to overcome possible
insufficient germination from one season. This would
further increase the genetic diversity of the stored
seeds. From our field collections, we have learned that
several rare and endemic species often produce seeds
in small quantities and over such a long ripening 12 3 Plant Ecol (2020) 221:1045–1067 1059 Fig. 3 Numbers of species tested with a gibberellic acid
treatment (GA) using the standard protocol (total 114 species),
and grouped in the germination classes zero: FGP = 0.0%,
weak: FGP B 10.0%, low: FGP C 10.1 B 20.0%, intermedi-
ate: FGP C 20.1 B 50.0% and high: FGP C 50.1–100.0% available. In any case, besides germination in
the laboratory, field experiments to analyse
recruitment success under different environ-
mental conditions are necessary, to test the
performance and population persistence of these
groups (Margreiter et al. in review). (b) (b)
Although
germination
traits
are
important
drivers of community composition and species
persistence (Donohue et al. 2010; Jime´nez-
Alfaro et al. 2016; Tudela-Isanta et al. 2018a, b),
no clear effect of habitat on germination was
found in our study. The highest germination was
recorded for species from WET environments
(although several Carex species did not germi-
nate, Table 1) and from MEAD. The conditions
for seed development might have been favour-
able in these habitats, whereas in other habitats
the growing seasons are often too short (ROSC,
GRASS) or unfavourable for a complete seed
maturation (DRYM, TALL, SHFO). This is in
line with the findings of Bu et al. (2007) who
germinated seeds of alpine species at the eastern
Tsinghai-Tibet plateau, China, and in other
habitats, when used germination temperatures
that did not reflect the distribution areas of the
tested species (Schu¨tz et al. 1997; Thompson
et al. 1999). Fig. Discussion 3 Numbers of species tested with a gibberellic acid
treatment (GA) using the standard protocol (total 114 species),
and grouped in the germination classes zero: FGP = 0.0%,
weak: FGP B 10.0%, low: FGP C 10.1 B 20.0%, intermedi-
ate: FGP C 20.1 B 50.0% and high: FGP C 50.1–100.0% conservation measures in the field are indispensable,
including the protection of entire plant communities. The standard protocol used in this study with only
one temperature regime, worked well in a high
percentage of species, but 25.1% did not germinate. This amount of non-germinating species weakens an
ecological interpretation of the results regarding
germination success. It seems to be mandatory to
perform germination trials at least under two (or
preferably more) different temperature regimes to
fully cover the germination niche of a species, or to
identify unknown germination needs. In addition, seed
pre-treatments, such as cold-wet stratification and
scarification, would have been necessary for at least
some of the species studied. For technical reasons,
many species have been tested after a relatively long
storage period (dry-cold). In some of these cases,
storage may have induced seed dormancy (Baskin and
Baskin 2014). Considering these limitations of the
study, we must interpret our results with caution. (c) (c)
Under the temperature regime used, there was
no significant pattern of FGP between species
from high and low elevations, but we may spot a
trend with nival-alpine-subalpine species hav-
ing lower FGPs compared to montane-colline
centred species. In situ, a low germination may
be the bottleneck for alpine seedling establish-
ment under climate change (Shevtsova et al. 2009; Graae et al. 2011), since a high mortality
of alpine seedlings was evidenced by several
authors (Marcante et al. 2009; Graae et al. 2011;
Winkler et al. 2015; Milbau et al. 2017). A high
germination percentage could mean an impor-
tant pre-requisite for a successful invasion of
species from lower to higher elevations. Thus,
novel communities can arise (Alexander et al. 2015, 2018), and it is unknown how coexistence
or competition will shape species turnovers and
community compositions in the future. (a)
A classification of the tested species into
common, scattered, rare and endemic species
did not reveal differences of FGP. Despite the
uneven quantity of species within these classes,
about 50% of the species in each class germi-
nated, and about 50% did not germinate. Plant Ecol (2020) 221:1045–1067 2017), because this simulates a
winter period which is an essential factor for
inducing germination in spring. Cavieres and
Sierra-Almeida (2018) recently found that this
stratification effect does not increase with
elevation. For some alpine species, even nega-
tive effects of stratification occur (Gime´nez-
Benavides and Milla 2013). Morphophysiolog-
ical dormancy might have also occurred in the
non-germinated
species
of
our
study,
for
instance in species from Apiaceae, Campanu-
laceae,
Gentianaceae,
and
Ranunculaceae
(Threadgill et al. 1981; Forbis and Diggle
2001; Baskin and Baskin 2005; Vandelook (g)
A treatment with GA3 was applied to 44.7% of
the investigated species. In half of these species,
germination was stimulated. They can be clas-
sified as having non-deep and intermediate
physiological dormancy (Baskin and Baskin
2004, 2014). For the other half, deep physio-
logical dormancy must be assumed. According
to Schwienbacher et al. (2011), breaking deep
physiological dormancy requires a cold-wet
stratification for more than four months and
additional preconditions such as temperature
fluctuations (Baskin and Baskin 2014). Several
authors suggested that cold-wet stratification is
necessary for alpine species to germinate
(Cavieres and Arroyo 2000; Shimono and Kudo
2003; Garcı´a Ferna´ndez et al 2015; Ferna´ndez mentioned that germination does hardly corre-
late with CSR or LHS traits. This can be
supported by our results, as no pattern was
found. Stress tolerant species built the biggest
group in alpine environments (Caccianiga et al. 2006) as also seen by our collections, and
ruderal species occur only in a small proportion. (f) (f)
No correlation was found between seed mass
and FGP of species with high and intermediate
germination. This is in line with results of a
glacier foreland study, where representative
species from different successional stages were
investigated (Schwienbacher et al. 2012). In
contrast, seed mass was highly related to FGP of
selected glacier foreland species, after a winter
burial trial in the field (Schwienbacher and
Erschbamer 2001). Liu et al. (2013) reported of
significant responses to temperature fluctuations
of small seeded species compared to larger
seeded ones. Generally, seed mass is thought to
be an indicator for seedling survival (Grubb
1977; Westoby 1998; Coomes and Grubb 2003),
and to correlate with several species traits
(Fenner
and
Thompson
2005;
Moles
and
Westoby 2006). Further, seed mass has also
been regarded as an indicator for the competi-
tive ability of a species (Tilman 1994). Plant Ecol (2020) 221:1045–1067 1061 mentioned that germination does hardly corre-
late with CSR or LHS traits. This can be
supported by our results, as no pattern was
found. Stress tolerant species built the biggest
group in alpine environments (Caccianiga et al. 2006) as also seen by our collections, and
ruderal species occur only in a small proportion. (f)
No correlation was found between seed mass
and FGP of species with high and intermediate
germination. This is in line with results of a
glacier foreland study, where representative
species from different successional stages were
investigated (Schwienbacher et al. 2012). In
contrast, seed mass was highly related to FGP of
selected glacier foreland species, after a winter
burial trial in the field (Schwienbacher and
Erschbamer 2001). Liu et al. (2013) reported of
significant responses to temperature fluctuations
of small seeded species compared to larger
seeded ones. Generally, seed mass is thought to
be an indicator for seedling survival (Grubb
1977; Westoby 1998; Coomes and Grubb 2003),
and to correlate with several species traits
(Fenner
and
Thompson
2005;
Moles
and
Westoby 2006). Further, seed mass has also
been regarded as an indicator for the competi-
tive ability of a species (Tilman 1994). Thus,
seed mass might be an essential trait for seedling
performance, but not for FGP per se. (g)
A treatment with GA3 was applied to 44.7% of
the investigated species. In half of these species,
germination was stimulated. They can be clas-
sified as having non-deep and intermediate
physiological dormancy (Baskin and Baskin
2004, 2014). For the other half, deep physio-
logical dormancy must be assumed. According
to Schwienbacher et al. (2011), breaking deep
physiological dormancy requires a cold-wet
stratification for more than four months and
additional preconditions such as temperature
fluctuations (Baskin and Baskin 2014). Several
authors suggested that cold-wet stratification is
necessary for alpine species to germinate
(Cavieres and Arroyo 2000; Shimono and Kudo
2003; Garcı´a-Ferna´ndez et al. 2015; Ferna´ndez-
Pascual et al. 2017), because this simulates a
winter period which is an essential factor for
inducing germination in spring. Cavieres and
Sierra-Almeida (2018) recently found that this
stratification effect does not increase with
elevation. For some alpine species, even nega-
tive effects of stratification occur (Gime´nez-
Benavides and Milla 2013). Morphophysiolog-
ical dormancy might have also occurred in the
non-germinated
species
of
our
study,
for
instance in species from Apiaceae, Campanu-
laceae,
Gentianaceae,
and
Ranunculaceae
(Threadgill et al. Discussion This
may indicate that widespread species and
species with a small distribution range do not
have different germination temperature require-
ments, or these results may indicate that there
are differences. For ex situ seed banking, these
results could mean a 50/50 chance of germina-
tion success for rare and endemic Alpine
species, when testing under a single standard
protocol. Further, this strongly suggests apply-
ing more temperature regimes if seeds are (d) (d)
The effect of bedrock (soil reaction) on germi-
nation has been notably investigated in the field
(Forbis 2003; Wenk and Dawson 2007; Isselin- (d)
The effect of bedrock (soil reaction) on germi-
nation has been notably investigated in the field
(Forbis 2003; Wenk and Dawson 2007; Isselin- 12 3 3 Plant Ecol (2020) 221:1045–1067 1060 Fig. 4 Final germination percentages (FGP) of 53 species from the family Asteraceae, tested under the standard protocol. Error bars
show standard errors Fig. 4 Final germination percentages (FGP) of 53 species from the family Asteraceae, tested under the standard protocol. Error bars
show standard errors species from very acidic bedrock having a lower
germination rate than species from slightly
acidic to neutral soil reaction sites. Nondedeu and Be´de´carrats 2013), with the
result, that soil chemistry was not related to
FGP. Recent studies however (Ferna´ndez-Pas-
cual et al. 2017; Tudela-Isanta et al. 2018a, b),
reported contrasting results stating a consistent
relation between germination strategy and bed-
rock type. Our results on FGP classified to three
bedrock types, did not reveal any significant
differences. However, we may note a trend of Nondedeu and Be´de´carrats 2013), with the
result, that soil chemistry was not related to
FGP. Recent studies however (Ferna´ndez-Pas-
cual et al. 2017; Tudela-Isanta et al. 2018a, b),
reported contrasting results stating a consistent
relation between germination strategy and bed-
rock type. Our results on FGP classified to three
bedrock types, did not reveal any significant
differences. However, we may note a trend of (e) (e)
Key parts of the CSR strategy concept (Grime
1979; Pierce et al. 2014) and the leaf-height-
seed concept (LHS, Westoby 1998) are seed
traits such as seed mass and volume, where seed
mass is thought to be an indicator for seedling
survival. Jime´nez-Alfaro
et
al. (2016) 12 3 Plant Ecol (2020) 221:1045–1067 Fig. 5 Final germination
percentage (FGP) of 17
species from the family
Poaceae, tested under the
standard protocol. Error bars
show standard errors Plant Ecol (2020) 221:1045–1067 1981; Forbis and Diggle
Fig. 5 Final germination
percentage (FGP) of 17
species from the family
Poaceae, tested under the
standard protocol. Error bars
show standard errors Fig. 5 Final germination
percentage (FGP) of 17
species from the family
Poaceae, tested under the
standard protocol. Error bars
show standard errors ermination
GP) of 17
he family
d under the
col. Error bars
errors Fig. 5 Final germination
percentage (FGP) of 17
species from the family
Poaceae, tested under the
standard protocol. Error bars
show standard errors A treatment with GA3 was applied to 44.7% of
the investigated species. In half of these species,
germination was stimulated. They can be clas-
sified as having non-deep and intermediate
physiological dormancy (Baskin and Baskin
2004, 2014). For the other half, deep physio-
logical dormancy must be assumed. According
to Schwienbacher et al. (2011), breaking deep
physiological dormancy requires a cold-wet
stratification for more than four months and
additional preconditions such as temperature
fluctuations (Baskin and Baskin 2014). Several
authors suggested that cold-wet stratification is
necessary for alpine species to germinate
(Cavieres and Arroyo 2000; Shimono and Kudo
2003; Garcı´a-Ferna´ndez et al. 2015; Ferna´ndez-
Pascual et al. 2017), because this simulates a
winter period which is an essential factor for
inducing germination in spring. Cavieres and
Sierra-Almeida (2018) recently found that this
stratification effect does not increase with
elevation. For some alpine species, even nega-
tive effects of stratification occur (Gime´nez-
Benavides and Milla 2013). Morphophysiolog-
ical dormancy might have also occurred in the
non-germinated
species
of
our
study,
for
instance in species from Apiaceae, Campanu-
laceae,
Gentianaceae,
and
Ranunculaceae
(Threadgill et al. 1981; Forbis and Diggle
2001; Baskin and Baskin 2005; Vandelook (g)
A treatment with GA3 was applied to 44.7% of
the investigated species. In half of these species,
germination was stimulated. They can be clas-
sified as having non-deep and intermediate
physiological dormancy (Baskin and Baskin
2004, 2014). For the other half, deep physio-
logical dormancy must be assumed. According
to Schwienbacher et al. (2011), breaking deep
physiological dormancy requires a cold-wet
stratification for more than four months and
additional preconditions such as temperature
fluctuations (Baskin and Baskin 2014). Several
authors suggested that cold-wet stratification is
necessary for alpine species to germinate
(Cavieres and Arroyo 2000; Shimono and Kudo
2003; Garcı´a-Ferna´ndez et al. 2015; Ferna´ndez-
Pascual et al. Plant Ecol (2020) 221:1045–1067 Thus,
seed mass might be an essential trait for seedling
performance, but not for FGP per se. 12 3 3 Plant Ecol (2020) 221:1045–1067 1062 Fig. 6 a Final germination percentage (FGP) of 21 species from the family Saxifragaceae, tested under the standard protocol and b FGP
after a treatment with gibberellic acid (GA). These species were cold-wet stratified before the germination test Fig. 6 a Final germination percentage (FGP) of 21 species from the family Saxifragaceae, tested under the standard protocol and b FGP
after a treatment with gibberellic acid (GA). These species were cold-wet stratified before the germination test (Andersson and Milberg 1998). The loss of
physiological dormancy may be closely related
to the environment of the maternal habitat
(Rosbakh and Poschlod 2015). Thus, it would
be advisable to test a variety of temperature
regimes for germination in combination with
seed pre-treatments. et al. 2009). In particular, Ranunculaceae have
rudimentary underdeveloped embryos (Finch-
Savage and Leubner-Metzger 2006). Noccaea
crantzii and Pedicularis portenschlagii, both
endemic species, were stimulated by GA3,
however, for other endemic species (i.e. Vale-
riana species) we did not have enough seeds to
perform the GA3 treatment and it remains
unknown if the dormancy of these species could
have been broken by a GA3 treatment. Further,
some endemic species require a cold-wet strat-
ification (i.e. 90 days for an endemic Telekia,
Brusa et al. 2007; at least nine months for an
endemic Androsace, Frattaroli et al. 2013),
whereas for an endemic Campanula, 24 h seed
imbibition and a specific germination medium is
required (Frattaroli et al. 2013). These examples
and our results illustrate that no generalisations
for GA3 treatment are possible, due to the
individual germination response of endemic
species. Moreover, even between individuals of
the same population, dormancy may vary et al. 2009). In particular, Ranunculaceae have
rudimentary underdeveloped embryos (Finch-
Savage and Leubner-Metzger 2006). Noccaea
crantzii and Pedicularis portenschlagii, both
endemic species, were stimulated by GA3,
however, for other endemic species (i.e. Vale-
riana species) we did not have enough seeds to
perform the GA3 treatment and it remains
unknown if the dormancy of these species could
have been broken by a GA3 treatment. Further,
some endemic species require a cold-wet strat-
ification (i.e. 90 days for an endemic Telekia,
Brusa et al. 2007; at least nine months for an
endemic Androsace, Frattaroli et al. Plant Ecol (2020) 221:1045–1067 However, after a stratifica-
tion period of three month, some species did still
not germinate, and several were stimulated only
weakly by a GA3 treatment. Gime´nez-Bena-
vides and Milla (2013) even noted a negative
effect of cold-wet stratification on germination
success on two Saxifraga species from northern
Spain. The only annual species among the
collected, S. tridactylites, did not germinate
under the standard protocol. It needs low
temperatures of 5 C and 10 C for germination
(Pemadasa and Lovell 1975) while high tem-
peratures enforce dormancy. This is a nice
example of germination temperature reflecting
adaptation to habitat conditions of winter
annuals. climate warming may be tolerated for germina-
tion without problems, which may lead to
increases in seedling numbers, and may turn
into
increased
seedling
recruitment. This
hypothesis is in line with several statements
made already for Poaceae (Venn et al. 2014;
Porro et al. 2019). In contrast, species from the
family Saxifragaceae seemed to have problems
with the conditions of the standard protocol,
although Schwienbacher et al. (2011) recorded
an
optimum
germination
temperature
of
20/10 C for alpine Saxifraga species. Their
germination might be restricted by different
types of dormancy. However, after a stratifica-
tion period of three month, some species did still
not germinate, and several were stimulated only
weakly by a GA3 treatment. Gime´nez-Bena-
vides and Milla (2013) even noted a negative
effect of cold-wet stratification on germination
success on two Saxifraga species from northern
Spain. The only annual species among the
collected, S. tridactylites, did not germinate
under the standard protocol. It needs low
temperatures of 5 C and 10 C for germination
(Pemadasa and Lovell 1975) while high tem-
peratures enforce dormancy. This is a nice
example of germination temperature reflecting
adaptation to habitat conditions of winter
annuals. Seedling recruitment is a highly risky developmen-
tal stage in nival-alpine environments (Sto¨cklin and
Ba¨umler 1996; Erschbamer et al. 2008; Marcante et al. 2009; Erschbamer and Caccianiga 2017). Therefore,
high germination rates are mandatory for alpine plants
to persist or migrate. The same holds for upward-
migrating species along elevation gradients. Climate
warming may enhance germination as shown for arctic
species (Alsos et al. 2013). However, according to
niche models (Schwager and Berg 2019) habitat
suitability will change remarkably, and several Alpine
species might be suffering from habitat loss. Importance of findings for ex situ seed
conservation and seedling recruitment in the field In any classification in this study, about half of the
species germinated up to intermediate and high extents
using the standard protocol. All the species tested
originated from the European Eastern Alps. It seems
that most Alpine species require relatively high
temperatures for germination, but that germination is
restricted by different types of dormancy. For ex situ
seed banking, an application of different dormancy
release treatments is advisable when testing germina-
tion of Alpine species. However, for rare and endemic
species, seeds for several parallel trials are often
limited or not available. In these cases, standard
protocols must be applied, and for Alpine species, a
high temperature regime might be recommended. Data availability
The data are currently not on a repository. The datasets are available from the corresponding author on
reasonable request. Plant Ecol (2020) 221:1045–1067 2013),
whereas for an endemic Campanula, 24 h seed
imbibition and a specific germination medium is
required (Frattaroli et al. 2013). These examples
and our results illustrate that no generalisations
for GA3 treatment are possible, due to the
individual germination response of endemic
species. Moreover, even between individuals of
the same population, dormancy may vary (h) (h)
Among the studied species with weak or zero
germination, only few might have had physical
dormancy. This mechanism is typical for
Fabaceae and for sure, scarification might have
been an efficient dormancy release for species
from this family (Schu¨tz 1988; Flu¨eler 2011;
Schwienbacher et al. 2011). However, we did
not consider such treatments in our screening
project. Asteraceae and Poaceae are among the
most widespread families in the Alps (Ko¨rner
2003). According to our results, most species of
these two families germinated to a high extent
using the standard protocol, i.e. at warm tem-
peratures (20/10 C or 25/15 C). This means,
higher temperatures in the field that occur due to (h)
Among the studied species with weak or zero
germination, only few might have had physical
dormancy. This mechanism is typical for
Fabaceae and for sure, scarification might have
been an efficient dormancy release for species
from this family (Schu¨tz 1988; Flu¨eler 2011;
Schwienbacher et al. 2011). However, we did
not consider such treatments in our screening
project. Asteraceae and Poaceae are among the
most widespread families in the Alps (Ko¨rner
2003). According to our results, most species of
these two families germinated to a high extent
using the standard protocol, i.e. at warm tem-
peratures (20/10 C or 25/15 C). This means,
higher temperatures in the field that occur due to 12 3 Plant Ecol (2020) 221:1045–1067 1063 climate warming may be tolerated for germina-
tion without problems, which may lead to
increases in seedling numbers, and may turn
into
increased
seedling
recruitment. This
hypothesis is in line with several statements
made already for Poaceae (Venn et al. 2014;
Porro et al. 2019). In contrast, species from the
family Saxifragaceae seemed to have problems
with the conditions of the standard protocol,
although Schwienbacher et al. (2011) recorded
an
optimum
germination
temperature
of
20/10 C for alpine Saxifraga species. Their
germination might be restricted by different
types of dormancy. Plant Ecol (2020) 221:1045–1067 Germi-
nation experiments in the laboratory are valuable,
however, they do not substitute for observations in the
field under real environmental conditions. Investiga-
tions of the reproductive output, seedling survival and
recruitment, as well as the demographic behaviour of
species should be focussed by future projects. Acknowledgements
Open
access
funding
provided
by
University of Innsbruck and Medical University of Innsbruck. We are thankful to all seed collectors (Monika Hamacher,
Andreas Ba¨r, Bettina Mittendrein, Ca¨cilia Lechner-Pagitz, Lena
Nicklas, Moritz Falch, Martina Po¨ltl, Matthias Kaltenbo¨ck),
helping hands in the lab (Barbara Pernfuß, Charlotte Permann,
Katharina
Ramskogler)
and
to
Matthias
Hepp
for
the
germination tests of the seed collection 2017. Further, we are
grateful to two anonymous reviewers for their valuable
comments on the manuscript. Funding
This study was done as part of the ‘‘Alpine Seed
Conservation
and
Research
Network’’
(https://www. alpineseedconservation.eu/), a project managed by the Royal
Botanic Gardens, Kew, funded by the David and Claudia
Harding Foundation. VM was additionally supported by a PhD
scholarship grant from the Vice rector for research at the
University of Innsbruck. Importance of findings for ex situ seed
conservation and seedling recruitment in the field References Alexander JM, Diez JM, Levine JM (2015) Novel competitors
shape species’ responses to climate change. Nature
525:515–518. https://doi.org/10.1038/nature14952 Cavieres LA, Arroyo MTK (2000) Seed germination response to
cold stratification period and thermal regime in Phacelia
secunda (Hydrophyllaceae): altitudinal variation in the
mediterranean andes of central Chile. Plant Ecol 149:1–8. https://doi.org/10.1023/A:1009802806674 Alexander JM, Chalmandrier L, Lenoir J, Burgess TI, Essl F,
Haider S, Kueffer C, McDougall K, Milbau A, Nun˜ez MA,
Pauchard A, Rabitsch W, Rew LJ, Sanders NJ, Pellissier L
(2018) Lags in the response of mountain plant communities
to climate change. Glob Chang Biol 24:563–579. https://
doi.org/10.1111/gcb.13976 Cavieres LA, Sierra-Almeida A (2018) Assessing the impor-
tance of cold-stratification for seed germination in alpine
plant species of the High-Andes of central Chile. Perspect
Plant Ecol Evol Syst 30:125–131. https://doi.org/10.1016/
j.ppees.2017.09.005 Alsos IG, Mu¨ller E, Eidesen PB (2013) Germinating seeds or
bulbils in 87 of 113 tested arctic species indicate potential
for ex situ seed bank storage. Polar Biol 36:819–830. https://doi.org/10.1007/s00300-013-1307-7 Coomes DA, Grubb PJ (2003) Colonization, tolerance, com-
petition and seed-size variation within functional groups. Trends Ecol Evol 18:283–291. https://doi.org/10.1016/
S0169-5347(03)00072-7 Amen RD (1966) The extent and role of seed dormancy in alpine
plants. Q Rev Biol 41:271–281. https://doi.org/10.1086/
405055 Andersson L, Milberg P (1998) Variation in seed dormancy
among mother plants, populations and years of seed col-
lection. Seed Sci Res 8:29–38. https://doi.org/10.1017/
s0960258500003883 Donohue K, Rubio de Casas R, Burghardt L, Kovach K, Willis
CG (2010) Germination, postgermination adaptation, and
species ecological ranges. Annu Rev Ecol Evol Syst
41:293–319. https://doi.org/10.1146/annurev-ecolsys-
102209 144715 Auguie B, Antonov A (2017) Package gridExtra. https://cran.r-
project.org/web/packages/gridExtra/gridExtra.pdf. Acces-
sed 14 Nov 2019 102209 144715
ENSCONET (2009a) ensconet seed collecting manual for wild
species. https://www.publicgardens.org/resources/
ensconet-seed-collecting-manual-wild-species. Accessed
4 Feb 2020 Baskin JM, Baskin CC (2004) A classification system for seed
dormancy. Seed Sci Res 14:1–16. https://doi.org/10.1079/
ssr2003150 ENSCONET (2009b) ENSCONET Curation protocols & Rec-
ommendations. https://ensconet.maich.gr/PDF/Curation_
protocol_English.pdf. Accessed 4 Feb 2020 Baskin CC, Baskin JM (2005) Underdeveloped embryos in
dwarf seeds and implications for assignment to dormancy
class. Seed Sci Res 15:357–360. https://doi.org/10.1079/
ssr2005224 Eriksson O (1999) Seed size variation and its effect on germi-
nation and seedling performance in the clonal herb Con-
vallaria majalis. Acta Oecol 20:61–66. https://doi.org/10. 1016/S1146-609X(99)80016-2 Baskin CC, Baskin JM (2014) Seeds: ecology, biogeography,
and evolution of dormancy and germination, 2nd edn. Compliance with ethical standards Conflict of interest
The authors declare that they have no
conflicts of interest. Ethical approval
The study did not cause any harm to the
environment, animals, or plants. The study did not involve
human participants and/or animals. Informed Consent
The publication in Plant Ecology has been
approved by all co-authors. 123 123 Plant Ecol (2020) 221:1045–1067 1064 Billings WD, Mooney HA (1968) The ecology of arctic and
alpine plants. Biol Rev 43:481–529. https://doi.org/10. 1111/j.1469-185X.1968.tb00968.x Open Access
This article is licensed under a Creative Com-
mons Attribution 4.0 International License, which permits use,
sharing, adaptation, distribution and reproduction in any med-
ium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative
Commons licence, and indicate if changes were made. The
images or other third party material in this article are included in
the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not
included in the article’s Creative Commons licence and your
intended use is not permitted by statutory regulation or exceeds
the permitted use, you will need to obtain permission directly
from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. Bliss LC (1958) Seed germination in arctic and alpine species. Arctic 11:180. https://doi.org/10.14430/arctic3743 Brusa G, Ceriani R, Cerabolini B (2007) Seed germination in a
narrow endemic species (Telekia speciosissima, Aster-
aceae): implications for ex situ conservation. Plant Biosyst
141:56–61. https://doi.org/10.1080/11263500601153495 Bu H, Chen X, Xu X, Liu K, Jia P, Du G (2007) Seed mass and
germination in an alpine meadow on the eastern Tsinghai-
Tibet plateau. Plant Ecol 191:127–149. https://doi.org/10. 1007/s11258-006-9221-5 Caccianiga M, Luzzaro A, Pierce S, Ceriani RM, Cerabolini B
(2006) The functional basis of a primary succession
resolved by CSR classification. Oikos 112:10–20. https://
doi.org/10.1111/j.0030-1299.2006.14107.x Casazza G, Giordani P, Benesperi R, Foggi B, Viciani D,
Filigheddu R, Farris E, Bagella S, Pisanu S, Mariotti MG
(2014) Climate change hastens the urgency of conservation
for range-restricted plant species in the central-northern
Mediterranean region. Biol Conserv 179:129–138. https://
doi.org/10.1016/j.biocon.2014.09.015 References Elsevier, London Erschbamer B, Pfattner M (2002) Das Keimverhalten von
alpinen Arten in der Klimakammer und im Gela¨nde Ger-
mination characteristics of alpine species in the growth
chamber and in the field site. Ber Nat Med Verein Inns-
bruck 89:87–97 Bernareggi G, Carbognani M, Mondoni A, Petraglia A (2016)
Seed dormancy and germination changes of snowbed
species under climate warming: the role of pre-and post-
dispersal temperatures. Ann Bot 118:529–539. https://doi. org/10.1093/aob/mcw125 Erschbamer B, Caccianiga MS (2017) Glacier forelands: lessons
of plant population and community development. In: 12 3 Plant Ecol (2020) 221:1045–1067 1065 Ca´novas FM, Lu¨ttge U, Matyssek R (eds) Progress in
botany, vol 78. Springer, Cham, pp 259–284 Ca´novas FM, Lu¨ttge U, Matyssek R (eds) Progress in
botany, vol 78. Springer, Cham, pp 259–284 germination of dwarf shrub seeds from arctic, alpine and
boreal sites. Plant Ecol 198:275–284. https://doi.org/10. 1007/s11258-008-9403-4 Erschbamer B, Niederfriniger Schlag R, Winkler E (2008)
Colonization processes on a central alpine glacier foreland. J Veg Sci 19:855–862. https://doi.org/10.3170/2008-8-
18464 Graae BJ, Ejrnæs R, Lang SI, Meineri E, Ibarra PT, Bruun HH
(2011) Strong microsite control of seedling recruitment in
tundra. Oecologia 166:565–576. https://doi.org/10.1007/
s00442-010-1878-8 Erschbamer B, Anich C, Benirschke M, Ganthaler A, Grassmair
R, Hasibeder R, Huter V, Konzett D, Lechleitner M,
Magauer M, Miller R, Newerkla S, Schneider J, Zeisler B,
Schwienbacher E (2010) Das Keimverhalten von 13 alpi-
nen Arten der Familie Asteraceae im Licht und im Dun-
keln. Ber Nat-Med Verein Innsbruck 96:73–78 Grime JP (1979) Plant strategies and vegetation processes. Wiley, Chichester Grosjean P, Ibanez Frederic, Etienne M (2018) Package ‘pas-
tecs’: package for analysis of space-time ecological series. https://cran.r-project.org/web/packages/pastecs/pastecs. ttps://cran.r-project.org/web/packages/pastecs/pastecs keln. Ber Nat-Med Verein Innsbruck 96:73–78 Fenner M, Thompson K (2005) The ecology of seeds, 1st edn. Cambridge University Press, Cambridge pdf. Accessed 14 Nov 2019 Grubb PJ (1977) The maintenance of species-richness in plant
communities: the importance of the regeneration niche. Biol
Rev
52:107–145. https://doi.org/10.1111/j.1469-
185x.1977.tb01347.x Ferna´ndez-Pascual E, Jime´nez-Alfaro B, Bueno A (2017)
Comparative seed germination traits in alpine and sub-
alpine grasslands: higher elevations are associated with
warmer germination temperatures. Plant Biol 19:32–40. https://doi.org/10.1111/plb.12472 Harper JL (1977) Population biology of plants. Academic Press,
London Finch-Savage WE, Leubner-Metzger G (2006) Seed dormancy
and the control of germination. New Phytol 171:501–523. https://doi.org/10.1111/j.1469-8137.2006.01787.x Isselin-Nondedeu F, Be´de´carrats A (2013) Germination and
seedling responses of subalpine plants to different soil
substrates. Folia Geobot 48:39–53. https://doi.org/10. References 1007/s12224-012-9136-y Fischer MA, Oswald K, Adler W (2008) Exkursionsflora von
O¨ sterreich, Liechtenstein und Su¨dtirol. Biologiezentrum
der Obero¨sterreichischen Landesmuseen, 3rd edn. Land
Obero¨sterreich, Linz Jime´nez-Alfaro B, Silveira FAO, Fidelis A, Poschlod P, Com-
mander LE (2016) Seed germination traits can contribute
better to plant community ecology. J Veg Sci 27:637–645. https://doi.org/10.1111/jvs.12375 Flu¨eler RP (2011) Experimentelle Untersuchungen u¨ber Kei-
mung und Etablierung von alpinen Leguminosen. Diss
Vero¨ff Geobt Inst ETH Stift Ru¨bel Zu¨rich. https://doi.org/
10.3929/ethz-a-000643314 Kawecki TJ, Ebert D (2004) Conceptual issues in local adap-
tation. Ecol Lett 7:1225–1241. https://doi.org/10.1111/j. 1461-0248.2004.00684.x Forbis TA (2003) Seedling demography in an Alpine ecosystem. Am J Bot 90:1197–1206. https://doi.org/10.3732/ajb.90.8. 1197 Ko¨rner C (2003) Alpine plant life. Functional plant ecology of
high mountain ecosystems. Springer, Berlin Landolt E, Ba¨umler B, Erhardt A, Hegg O, Klo¨tzli F, La¨mmler
W, Nobis M, Rudmann-Maurer K, Schweingruber FH,
Theurillat J-P, Urmi E, Vust M, Wohlgemuth T (2010)
Flora indicativa. Ecological indicator values and biological
attributes of the Flora of Switzerland and the Alps, 2nd edn. Haupt, Bern Forbis TA, Diggle PK (2001) Subnivean embryo development
in the alpine herb Caltha leptosepala (Ranunculaceae). Can
J Bot 79:635–642. https://doi.org/10.1139/b01-037 Frattaroli AR, Di Martino L, Di Cecco V, Catoni R, Varone L,
Di Santo M, Gratani L (2013) Seed germination capability
of four endemic species in the Central Apennines (Italy):
relationships with seed size. Lazaroa 34:43–53. https://doi. org/10.5209/rev_LAZA.2013.v34.n1.42253 Liu K, Baskin JM, Baskin CC, Bu H, Du G, Ma M (2013) Effect
of diurnal fluctuating versus constant temperatures on
germination of 445 species from the Eastern Tibet Plateau. PLoS ONE. https://doi.org/10.1371/journal.pone.0069364 Garcı´a-Ferna´ndez A, Escudero A, Lara-Romero C, Iriondo JM
(2015) Effects of the duration of cold stratification on early
life stages of the Mediterranean alpine plant Silene ciliata. Plant Biol 17:344–350. https://doi.org/10.1111/plb.12226 Marcante S, Winkler E, Erschbamer B (2009) Population
dynamics along a primary succession gradient: do alpine
species fit into demographic succession theory? Ann Bot
103:1129–1143. https://doi.org/10.1093/aob/mcp047 g
Gime´nez-Benavides L, Milla R (2013) Comparative germina-
tion ecology of two altitudinal vicariant Saxifraga species
endemic to the north of Spain. Plant Biol 15:593–600. https://doi.org/10.1111/j.1438-8677.2012.00663.x Margreiter V, Walde J, Erschbamer B (in review) Germination
and seedling recruitment of alpine species in a two-year
field experiment along an elevation gradient Gime´nez-Benavides L, Escudero A, Pe´rez-Garcı´a F (2005) Seed
germination of high mountain Mediterranean species:
altitudinal, interpopulation and interannual variability. Ecol Res 20:433–444. doi.org/10.1007/s10113-019-01554-z Mondoni A, Probert RJ, Rossi G, Vegini E, Hay FR (2011)
Seeds of alpine plants are short lived: Implications for
long-term conservation. Ann Bot 107:171–179. https://doi. org/10.1093/aob/mcq222 Schwienbacher E, Erschbamer B (2001) Longevity of seeds in a
glacier foreland of the Central Alps-a burial experiment. Bull Geobot Inst ETH 68:63–71. https://doi.org/10.5169/
seals-377848 Mondoni A, Orsenigo S, Mu¨ller JV, Carlsson-Graner U, Jime´-
nez-Alfaro B, Abeli T (2018) Seed dormancy and longevity
in subarctic and alpine populations of Silene suecica. Alp
Bot 128:71–81. https://doi.org/10.1007/s00035-017-0194-
x Schwienbacher E, Navarro-Cano JA, Neuner G, Erschbamer B
(2011) Seed dormancy in alpine species. Flora Morphol
Distrib Funct Ecol Plants 206:845–856. https://doi.org/10. 1016/j.flora.2011.05.001 Mondoni A, Rossi G, Orsenigo S, Probert RJ (2012) Climate
warming could shift the timing of seed germination in
alpine plants. Ann Bot 110:155–164. https://doi.org/10. 1093/aob/mcs097 Schwienbacher E, Navarro-Cano JA, Neuner G, Erschbamer B
(2012) Correspondence of seed traits with niche position in
glacier foreland succession. Plant Ecol 213:371–382. https://doi.org/10.1007/s11258-011-9981-4 Shevtsova A, Graae BJ, Jochum T, Milbau A, Kockelbergh F,
Beyens L, Nijs I (2009) Critical periods for impact of cli-
mate warming on early seedling establishment in subarctic
tundra. Glob Chang Biol 15:2662–2680. https://doi.org/10. 1111/j.1365-2486.2009.01947.x Mu¨ller JV, Berg C, De´traz-Me´roz J, Erschbamer B, Fort N,
Lambelet-Haueter C, Margreiter V, Mombrial F, Mondoni
A, Pagitz K, Porro F, Rossi G, Schwager P, Breman E
(2017) The alpine seed conservation and research net-
work—a new initiative to conserve valuable plant species
in the European Alps. J Mt Sci 14:806–810. https://doi.org/
10.1007/s11629-016-4313-8 Shimono Y, Kudo G (2003) Intraspecific variations in seedling
emergence and survival of Potentilla matsumurae (Rosa-
ceae) between alpine fellfield and snowbed habitats. Ann
Bot 91:21–29. https://doi.org/10.1093/aob/mcg002 Orsenigo S, Abeli T, Rossi G, Bonasoni P, Pasquaretta C,
Gandini M, Mondoni A (2015) Effects of autumn and
spring heat waves on seed germination of high mountain
plants. PLoS ONE. https://doi.org/10.1371/journal.pone. 0133626 Sto¨cklin J, Ba¨umler E (1996) Seed rain, seedling establishment
and clonal growth strategies on a glacier foreland. J Veg
Sci 7:45–56. https://doi.org/10.2307/3236415 Pemadasa MA, Lovell PH (1975) Factors controlling germina-
tion of some dune annuals. J Ecol 63:41–59 Thompson K, Gastonn KJ, Band SR (1999) Range size, dis-
persal and niche breadth in the herbaceous flora of central
England. J Ecol 87:150–155. https://doi.org/10.1046/j. 1365-2745.1999.00334.x Pierce S, Bottinelli A, Bassani I, Ceriani RM, Cerabolini BEL
(2014) How well do seed production traits correlate with
leaf traits, whole-plant traits and plant ecological strate-
gies? References https://doi.org/10.1007/s11284-
005-0059-4 Milbau A, Graae BJ, Shevtsova A, Nijs I (2009) Effects of a
warmer climate on seed germination in the subarctic. Ann
Bot 104:287–296. https://doi.org/10.1093/aob/mcp117 Milbau A, Vandeplas N, Kockelbergh F, Nijs I (2017) Both seed
germination and seedling mortality increase with experi-
mental warming and fertilization in a subarctic tundra. AoB Plants 9:1–13. https://doi.org/10.1093/aobpla/plx040 Gobiet A, Kotlarski S, Beniston M, Heinrich G, Rajczak J,
Stoffel M (2014) 21st century climate change in the
European
Alps-a
review. Sci
Total
Environ
493:1138–1151. https://doi.org/10.1016/j.scitotenv.2013. 07.050 Moles AT, Westoby M (2006) Seed size and plant strategy
across the whole life cycle. Oikos 113:91–105. https://doi. org/10.1111/j.0030-1299.2006.14194.x Mondoni A, Daws MI, Belotti J, Rossi G (2009) Germination
requirements of the alpine endemic Silene elisabethae Jan: Mondoni A, Daws MI, Belotti J, Rossi G (2009) Germination
requirements of the alpine endemic Silene elisabethae Jan: Graae BJ, Alsos IG, Ejrnaes R (2008) The impact of temperature
regimes
on
development,
dormancy
breaking
and 12 3 3 1066 Plant Ecol (2020) 221:1045–1067 effects of cold stratification, light and GA3. Seed Sci
Technol 37:79–87. https://doi.org/10.15258/sst.2009.37.1. 10 Schwager P, Berg C (2019) Global warming threatens conser-
vation status of alpine EU habitat types in the European
Eastern Alps. Reg Environ Chang 19:2411–2421. https://
doi.org/10.1007/s10113-019-01554-z doi.org/10.1007/s10113-019-01554-z Plant Ecol 215:1351–1359. https://doi.org/10.1007/
s11258-014-0392-1 Threadgill PF, Baskin JM, Baskin CC (1981) Dormancy in seeds
of
Frasera
caroliniensis
(Gentianaceae). Am
J
Bot
68:80–86. https://doi.org/10.1002/j.1537-2197.1981. tb06358.x Porro F, Tomaselli M, Abeli T, Gandini M, Gualmini M,
Orsenigo S, Petraglia A, Rossi G, Carbognani M (2019)
Could plant diversity metrics explain climate-driven veg-
etation changes on mountain summits of the GLORIA
network? Biodivers Conserv 28:3575–3596. https://doi. org/10.1007/s10531-019-01837-1 Tilman D (1994) Competition and biodiversity in spatially
structured habitats. Ecology 75:2–16. https://doi.org/10. 2307/1939377 Tudela-Isanta M, Ferna´ndez-Pascual E, Wijayasinghe M,
Orsenigo S, Rossi G, Pritchard HW, Mondoni A (2018)
Habitat-related seed germination traits in alpine habitats. Ecol Evol 8:150–161. https://doi.org/10.1002/ece3.3539 R Core Team (2020) A language and environment for statistical
computing. In: R Found. Stat. Comput. https://www.r-
project.org. Accessed 15 May 2020 Tudela-Isanta M, Ladouceur E, Wijayasinghe M, Pritchard HW,
Mondoni A (2018) The seed germination niche limits the
distribution of some plant species in calcareous or siliceous
alpine bedrocks. Alp Bot 128:83–95. https://doi.org/10. 1007/s00035-018-0199-0 Rosbakh S, Poschlod P (2015) Initial temperature of seed ger-
mination as related to species occurrence along a temper-
ature gradient. Funct Ecol 29:5–14. https://doi.org/10. 1111/1365-2435.12304 Vandelook F, Bolle N, Van Assche JA (2009) Morphological
and physiological dormancy in seeds of aegopodium
podagraria (apiaceae) broken successively during cold
stratification. Seed Sci Res 19:115–123. https://doi.org/10. 1017/S0960258509301075 RStudio RsT (2016) Integrated development for R. RStudio. In:
RStudio, Inc., Boston, MA. https://www.rstudio.com/. Accessed 15 May 2020 Schu¨tz M (1988) Genetisch-o¨kologische Untersuchungen an
alpinen Pflanzenarten auf verschiedenen Gesteinsunterla-
gen : Keimungs- und Aussaatversuche = Genetic-ecologi-
cal studies in alpine plants from various substrata :
germination
trials
and
sowing
experiments. Vero¨ffentlichungen des Geobot Institutes der Eidg Tech
Hochschule, Stift Ru¨bel Zu¨rich, pp 963027–963032 Venables WN, Ripley BD (2002) Modern applied statistics with
S, 4th edn. Springer, Berlin Venn S, Pickering C, Green K (2014) Spatial and temporal
functional changes in alpine summit vegetation are driven
by increases in shrubs and graminoids. AoB Plants. https://
doi.org/10.1093/aobpla/plu008 Schu¨tz W, Milberg P, Schutz W (1997) Seed dormancy in carex
canescens: regional differences and ecological conse-
quences. Oikos 78:420. https://doi.org/10.2307/3545604 Vera ML (1997) Effects of altitude and seed size on germination
and seedling survival of heathland plants in north spain. Vera ML (1997) Effects of altitude and seed size on germination
and seedling survival of heathland plants in north spain. 12 3 Plant Ecol (2020) 221:1045–1067 1067 Plant
Ecol
133:101–106. doi.org/10.1007/s10113-019-01554-z https://doi.org/10.1023/A:
1009729201384 Wickham H (2018b) Package ‘scales’—scale functions for
visualization. In: CRAN Repos. https://cran.r-project.org/
web/packages/scales/scales.pdf. Accessed 15 May 2020 Walder T, Erschbamer B (2015) Temperature and drought drive
differences in germination responses between congeneric
species
along
altitudinal
gradients. Plant
Ecol
216:1297–1309. https://doi.org/10.1007/s11258-015-
0509-1 Wickham H, Winston C (2019) create elegant data visualisa-
tions using the grammar of graphics. In: Packag. ‘ggplot2’. https://cran.r-project.org/web/packages/ggplot2/ggplot2. pdf. Accessed 15 May 2020 pdf. Accessed 15 May 2020 Wang JH, Baskin CC, Chen W, Du GZ (2010) Variation in seed
germination between populations of five sub-alpine woody
species from eastern Qinghai-Tibet Plateau following dry
storage at low temperatures. Ecol Res 25:195–203. https://
doi.org/10.1007/s11284-009-0643-0 Winkler E, Marcante S, Erschbamer B (2015) Demography of
the alpine pioneer species Saxifraga aizoides in different
successional stages at the glacier fore-land of the Rot-
moosferner
(Obergurgl,
o¨tztal,
Austria). Tuexenia
35:267–283. https://doi.org/10.14471/2015.35.002 Wenk EH, Dawson TE (2007) Interspecific differences in seed
germination, establishment, and early growth in relation to
preferred soil type in an alpine community. Arctic Antarct
Alp
Res
39:165–176. https://doi.org/10.1657/1523-
0430(2007)39[165:IDISGE]2.0.CO;2 ZAMG (2016) Jahresbericht/Annual report. https://www.zamg. ac.at/cms/de/topmenu/ueber-uns/jahresberichte. Accessed
29 June 2020 ZAMG (2017) Jahresbericht/Annual report. https://www.zamg. ac.at/cms/de/topmenu/ueber-uns/jahresberichte. Accessed
29 June 2020 Westoby M (1998) A leaf-height-seed (LHS) plant ecology
strategy scheme. Plant Soil 199:213–227. https://doi.org/
10.1023/A:1004327224729 ZAMG (2018) Jahresbericht/Annual report. https://www.zamg. ac.at/cms/de/topmenu/ueber-uns/jahresberichte. Accessed
29 June 2020 Wickham H (2017) Package ‘tidyverse’. In: CRAN repos. https://tidyverse.org. Accessed 15 May 2020 Wickham H (2018a) Package ‘reshape’ flexibly reshape data. https://had.co.nz/reshape/. Accessed 15 May 2020 Publisher’s Note
Springer Nature remains neutral with
regard
to
jurisdictional
claims
in
published
maps
and
institutional affiliations. 123 123
|
https://openalex.org/W4393280178
|
https://www.mdpi.com/2673-8449/4/2/8/pdf?version=1711640250
|
English
| null |
Classification and Molecular Functions of Heparan Sulfate Proteoglycans and Their Molecular Mechanisms with the Receptor
|
Biologics
| 2,024
|
cc-by
| 20,080
|
Citation: Matsuzaka, Y.; Yashiro, R. Classification and Molecular
Functions of Heparan Sulfate
Proteoglycans and Their Molecular
Mechanisms with the Receptor. Biologics 2024, 4, 105–129. https://
doi.org/10.3390/biologics4020008 Yasunari Matsuzaka 1,2,* and Ryu Yashiro 2,3 Yasunari Matsuzaka 1,2,* and Ryu Yashiro 2,3 Yasunari Matsuzaka 1,2,* and Ryu Yashiro 2,3 y
p
*
Correspondence: yasunari80808@ims.u-tokyo.ac.jp; Tel.: +81-3-5449-5372 Abstract: Heparan sulfate proteoglycans are highly glycosylated proteins in which heparan sulfate,
a glycosaminoglycan sugar chain, is an acidic sugar chain consisting of a repeating disaccharide
structure of glucuronic acid and N-acetylglucosamine is locally sulfated. Syndecan, one of the
transmembrane HSPGs, functions as a receptor that transmits signals from the extracellular microen-
vironment to the inside of the cell. In the vascular system, heparan sulfate proteoglycans, a major
component of the glycocalyx, enable the binding of various plasma-derived molecules due to their
diversity, epimerization of glycosaminoglycans chains, long chains, and sulfation. Heparan sulfate
proteoglycans present in the extracellular matrix serve as a reservoir for bioactive molecules such as
chemokines, cytokines, and growth factors. Aberrant expression of heparan sulfate proteoglycans,
heparanase, and sulfatase is observed in many pathological conditions. Therefore, it can be applied
to therapeutic strategies for a wide range of fields including Alzheimer’s disease, heart failure, cancer,
organ transplants, diabetes, chronic inflammation, aging, and autoimmune diseases. Keywords: glycosaminoglycans; growth factor; heparan sulfate proteoglycans; matrix metallopeptidase;
syndecan Review
Classification and Molecular Functions of Heparan Sulfate
Proteoglycans and Their Molecular Mechanisms with the Receptor Yasunari Matsuzaka 1,2,* and Ryu Yashiro 2,3 1. Introduction The heparan sulfate proteoglycans (HSPGs) are highly glycosylated proteins in which
heparan sulfate (HS), a glycosaminoglycan sugar chain, is an acidic sugar chain consist-
ing of a repeating disaccharide structure of glucuronic acid and N-acetylglucosamine is
locally sulfated (Figure 1) [1,2]. Chondroitin sulfate (CS) and dermatan sulfate also be-
long to sulfated glycosaminoglycans (GAGs), but they differ from HS in that they contain
galactosamine as an amino sugar. Furthermore, heparin localized within mast cells is
homologous to HS, which exists on the cell surface and basement membrane as HSPG
covalently bound to core protein and plays important physiological roles while interacting
with various substances. The number of sugar chains of HSPGs on the cell surface varies
from one to three or four [3]. Received: 12 January 2024
Revised: 26 February 2024
Accepted: 25 March 2024
Published: 28 March 2024 Received: 12 January 2024
Revised: 26 February 2024
Accepted: 25 March 2024
Published: 28 March 2024 In addition, in vivo, it is covalently bound to a protein, core protein, and expressed
as HSPG, which are broadly classified into four types: Syndecan, glypican, perlecan, and
agrin [4]. Among them, Syndecan and glypican are proteoglycans expressed on the cell
surface. Three to five sugar chains and the core protein of glypican are bound to the cell
membrane via glycated phosphatidylinositol (GPI) [5]. HSPGs with many sugar chains,
including Syndecan, are recognized as a “full-time type”, in which the sugar chains are
constantly modified [3]. Syndecan in mammals consists of four genetically distinct core
proteins. The expression of Syndecan-1/Syndecan is primarily in the epithelium and some Copyright: © 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article
distributed
under
the
terms
and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/). https://www.mdpi.com/journal/biologics Biologics 2024, 4, 105–129. https://doi.org/10.3390/biologics4020008 Biologics 2024, 4 106 leukocytes, Syndecan-2/fibroglycan is expressed mostly in mesenchymal cells, Syndecan-
3/N-syndecan is expressed in the nervous system, endothelium, and muscle tissue, and
Syndecan-4/ryudocan is expressed almost ubiquitously [6,7]. Syndecan mRNAs encode
a family of diverse extracellular domains consisting of an N-terminal signal peptide and
an HS or CS addition site, a single transmembrane domain, and an intracellular domain
of a short C-terminal consisting of conserved C1 and C2 domains separated by a diverse
domain unique to each Syndecan [8]. Although the cytoplasmic domain of Syndecan
is only a small portion of the molecules, this domain has important functions, such as
regulating adhesion to the extracellular matrix, which contributes significantly to the
progression of tumors. Syndecan-4 interacts with protein kinase C, alpha (PKCa), and
Phosphatidylinositol 4,5-bisphosphate (PtdIns4,5P2) in fibroblasts cultured on fibronectin,
resulting in enhanced activation of PKCa. This interaction is mediated by the LGKKPIYKK
sequence, which is absent in other Syndecans and therefore cannot activate PKC, at the
center of the short intracellular domain of Syndecan-4. y
HSPGs, as a molecular group, have been involved in developmental and homeostatic
processes but also many pathological conditions, particularly cancer [9–11]. There are six
types of glypicans, including Glypican-1/Glypican, Glypican-2/Cerebroglycan, Glypican-
3/OCI-5, Glypican-4/L-Glypican, Glypican-5, and Glypican-6 [12]. The glypican family
has 14 cysteine residues necessary for forming a compact three-dimensional structure and
a COOH-terminal structure that is covalently bounded to GPI and anchored to the cell
membrane. The glypican family is mainly distributed on the cell membrane at the tip of
the cell, but it is also expressed in a cell-type-specific manner, with only Glypican-1 present
in vascular endothelial cells. On the other hand, perlecan and agrin are proteoglycans that
are secreted outside the cell and form the extracellular matrix [13,14]. The core protein
of perlecan has a molecular weight of 467 kDa in humans and consists of five functional
domains like other extracellular matrix (ECM) molecules. The first domain at the N-
terminus is a binding site for three HS chains, the second domain is a repeating structure
like the low-density lipoprotein (LDL) receptor, the third domain is a repeating structure
like a laminin short arm, the fourth domain contains an Ig-type NCAM (neural cell adhesion
molecules)-like repeat, and the fifth domain contains a laminin A chain-like repeat and an
epidermal growth factors (EGF) motif [14]. The core protein portion alone shows extreme
multifunctionality [14,15]. There is also a hybrid type with added dermatan sulfate chains. In addition to HS chains, a total of about 20kDa of N-linked and O-linked oligosaccharide
chains are added, suggesting their involvement in secretion. On the other hand, agrin
was discovered as a protein that induces the aggregation of acetylcholine receptors and
acetylcholinesterase in the postsynaptic membrane of the neuromuscular junction [16]. It is also present in basement membranes other than synapses and is more abundant
than perlecan in renal glomeruli. The core protein consists of three globular domains,
including a laminin-binding site at the head at N-terminus, a follistatin-like rod, and a
laminin G-like globular domain at the tail at C-terminus. Four HS chains and four N-linked
oligosaccharide chains are predicted to be modified [17]. Agrin exists in the basement
membranes of nerve tissue and other tissues, but those that have undergone unique splicing
are localized at synapses. In this review, we summarize the structure, biological activity,
diversity, modification, and therapeutic applications of HSPGs. Glypicans consist of six members in mammals, are anchored in the cell membrane via
a GPI anchor, and carry two to three HS chains [18]. Figure 1. Cell surface and extracellular HSPGs. Syndecan core proteins are transmembrane proteins
that contain a highly conserved C-terminal cytoplasmic domain. HS chains attach to serine residues
distal from the plasma membrane. Some Syndecans also contain a CS chain(s) that attaches to a
serine residue(s) near the membrane. The glypican core proteins are disulfide-stabilized globular
core proteins that are linked to the plasma membrane by a glycosylphosphatidylinositol (GPI) link-
age. HS chains link to serine residues adjacent to the plasma membrane. Perlecans are secreted
HSPGs that carry HS chains [20]. Figure 1. Cell surface and extracellular HSPGs. Syndecan core proteins are transmembrane proteins
that contain a highly conserved C-terminal cytoplasmic domain. HS chains attach to serine residues
distal from the plasma membrane. Some Syndecans also contain a CS chain(s) that attaches to a
serine residue(s) near the membrane. The glypican core proteins are disulfide-stabilized globular core
proteins that are linked to the plasma membrane by a glycosylphosphatidylinositol (GPI) linkage. HS
chains link to serine residues adjacent to the plasma membrane. Perlecans are secreted HSPGs that
carry HS chains [20]. HS chains attach to serine residues
distal from the plasma membrane. Some Syndecans also contain a CS chain(s) that attaches to a
serine residue(s) near the membrane. The glypican core proteins are disulfide-stabilized globular core
proteins that are linked to the plasma membrane by a glycosylphosphatidylinositol (GPI) linkage. HS
chains link to serine residues adjacent to the plasma membrane. Perlecans are secreted HSPGs that
carry HS chains [20]. 2. Modification and Biosynthesis of HS
2.1. Functions, Structures, and Biosynthesis of HS Chain
Due to its great structure diversity, HS chains of Syndecan bind with a wide range of
ligands that strongly influence adhesion, migration, proliferation, and survival [21]. The HS
chain is a strongly anionic, structurally heterogeneous, linear polysaccharide and retains
a common overall pattern, characterized by highly, intermediately, and lightly sulfated
regions. The expression of HS is in almost all cell types, and its synthesis occurs mainly
in the Golgi apparatus [22,23]. Although HS chains are synthesized by the same set of
enzymes, the length and amount of sulfation of HS chains on the same core protein differ
in different tissues. For example, when calculating the ratio of saturated/unsaturated
disaccharides, the average HS chain length in rats is found in muscle (to 60 disaccharides
(dpc) per chain), liver (to 40 dpc), brain (to 50 dpc), spleen (to 62 dpc), lung (to 42 dpc),
kidney (to 49 dpc), and heart (to 33 dpc). These differences are due to the expression of
biosynthetic enzymes in the Golgi apparatus and differences in nucleotide sugar donors
during synthesis. Figure 1. Cell surface and extracellular HSPGs. Syndecan core proteins are transmembrane proteins
h
h
hl
d C
l
l
d
S h
h
d 2. Modification and Biosynthesis of HS
2. Modification and Biosynthesis of HS 2. Modification and Biosynthesis of HS
2 1 Functions Structures and Biosynthesis of HS Chain
2.1. Functions, Structures, and Biosynthesis of HS Chain Thus, multiple types of core proteins
bind to HS, which is a sulfated sugar chain and has a structure, a sulfate group is added
to a repeating disaccharide skeleton of uronic acid and glucosamine, and the local sulfa-
tion of their sugar chains caused by various sulfotransferases, such as N-deacetylase/N-
sulfotransferase, C5 epi transferase, 6-O sulfotransferase, 2-O sulfotransferase, and 3-O
sulfotransferase, etc., resulted in the construction of a diverse group of glycoproteins [19]. Biologics 2024, 4
Biologics 2024, 4, F 107
3 Biologics 2024, 4
107
g
, ,
Figure 1. Cell surface and extracellular HSPGs. Syndecan core proteins are transmembrane proteins
that contain a highly conserved C-terminal cytoplasmic domain. HS chains attach to serine residues
distal from the plasma membrane. Some Syndecans also contain a CS chain(s) that attaches to a
serine residue(s) near the membrane. The glypican core proteins are disulfide-stabilized globular
core proteins that are linked to the plasma membrane by a glycosylphosphatidylinositol (GPI) link-
age. HS chains link to serine residues adjacent to the plasma membrane. Perlecans are secreted
HSPGs that carry HS chains [20]. 2. Modification and Biosynthesis of HS
2.1. Functions, Structures, and Biosynthesis of HS Chain
Due to its great structure diversity, HS chains of Syndecan bind with a wide range of
ligands that strongly influence adhesion, migration, proliferation, and survival [21]. The
HS chain is a strongly anionic, structurally heterogeneous, linear polysaccharide and re-
tains a common overall pattern, characterized by highly, intermediately, and lightly sul-
fated regions. The expression of HS is in almost all cell types, and its synthesis occurs
mainly in the Golgi apparatus [22,23]. Although HS chains are synthesized by the same
set of enzymes, the length and amount of sulfation of HS chains on the same core protein
differ in different tissues. For example, when calculating the ratio of saturated/unsatu-
rated disaccharides, the average HS chain length in rats is found in muscle (to 60 disac-
charides (dpc) per chain), liver (to 40 dpc), brain (to 50 dpc), spleen (to 62 dpc), lung (to
42 dpc), kidney (to 49 dpc), and heart (to 33 dpc). These differences are due to the expres-
sion of biosynthetic enzymes in the Golgi apparatus and differences in nucleotide sugar
donors during synthesis. Figure 1. Cell surface and extracellular HSPGs. Syndecan core proteins are transmembrane proteins
that contain a highly conserved C-terminal cytoplasmic domain. 2. Modification and Biosynthesis of HS
2 1 Functions Structures and Biosynthesis of HS Chain
2.1. Functions, Structures, and Biosynthesis of HS Chain 2.1. Functions, Structures, and Biosynthesis of HS Chain
Due to its great structure diversity, HS chains of Syndecan bind with a wide range of
ligands that strongly influence adhesion, migration, proliferation, and survival [21]. The
HS chain is a strongly anionic, structurally heterogeneous, linear polysaccharide and re-
tains a common overall pattern, characterized by highly, intermediately, and lightly sul-
fated regions. The expression of HS is in almost all cell types, and its synthesis occurs
mainly in the Golgi apparatus [22,23]. Although HS chains are synthesized by the same
set of enzymes, the length and amount of sulfation of HS chains on the same core protein
differ in different tissues. For example, when calculating the ratio of saturated/unsatu-
rated disaccharides, the average HS chain length in rats is found in muscle (to 60 disac-
charides (dpc) per chain), liver (to 40 dpc), brain (to 50 dpc), spleen (to 62 dpc), lung (to
42 dpc), kidney (to 49 dpc), and heart (to 33 dpc). These differences are due to the expres-
sion of biosynthetic enzymes in the Golgi apparatus and differences in nucleotide sugar
d
d
i
h
i
Due to its great structure diversity, HS chains of Syndecan bind with a wide range of
ligands that strongly influence adhesion, migration, proliferation, and survival [21]. The HS
chain is a strongly anionic, structurally heterogeneous, linear polysaccharide and retains
a common overall pattern, characterized by highly, intermediately, and lightly sulfated
regions. The expression of HS is in almost all cell types, and its synthesis occurs mainly
in the Golgi apparatus [22,23]. Although HS chains are synthesized by the same set of
enzymes, the length and amount of sulfation of HS chains on the same core protein differ
in different tissues. For example, when calculating the ratio of saturated/unsaturated
disaccharides, the average HS chain length in rats is found in muscle (to 60 disaccharides
(dpc) per chain), liver (to 40 dpc), brain (to 50 dpc), spleen (to 62 dpc), lung (to 42 dpc),
kidney (to 49 dpc), and heart (to 33 dpc). These differences are due to the expression of
biosynthetic enzymes in the Golgi apparatus and differences in nucleotide sugar donors
during synthesis. 2. Modification and Biosynthesis of HS
2 1 Functions Structures and Biosynthesis of HS Chain
2.1. Functions, Structures, and Biosynthesis of HS Chain Biologics 2024, 4 108 Biosynthesis of the HS chain initiates with the generation of the core tetrasaccharide,
in which stepwise addition of a tetrasaccharide chain (GlcA-Gal-Gal-Xyl-O-Ser) consisting
of xylose (Xyl), galactose (Gal), and glucuronic acid (GlcA) to serine residues of core pro-
teins [23]. This occurs through the translocation of xylose to a specific serine residue in the
HSPG core protein by xylose transferase, followed by the addition of two galactose residues
by galactosyltransferases I and II and glucuronic acid (GlucA) by glucuronidyltransferase. After the core tetrasaccharide is formed, elongation of the disaccharide repeating structure
(GlcA-GlcNAc) of acetylglucosamine (GlcNAc) and GlcA. Elongation of the HS chain
begins with the addition of N-acetylglucosamine (GlcNAc) by N-acetylglucosamine trans-
ferase I (GlcNAcT-I), followed by alternating addition of GlcA and GlcNAc by exostosis
family enzymes [24]. Subsequently, sulfation and epimerization modify the extending
[GlcA-GlcNAc]n polymer. The first modification that occurs and is required in advance
of all other modifications is GlcNAc N-deacetylation/N-sulfation of GlcNAc residues by
N-deacetylases/N-sulfotransferases (NDSTs). This reaction is followed by the isomeriza-
tion of GlcA and the resulting 2-O-sulfation of the iduronic acid (IdoA) residue. Finally,
O-sulfation at position 6 and/or position 3 of the glucosamine residue occurs. After the ini-
tial N-deacetylation/N-sulfation, all modification reactions are incompletely accomplished,
resulting in various sulfation states of HS, in which HS chain composition determines
HSPG-mediated cellular responses [23]. Due to the exostosis glycosyltransferase 1 (Ext1)
gene mutation, primary fibroblasts express a smaller amount and shorter chain of HS than
the wild-type (Wt) fibroblasts and are unable to form vinculin-containing focal adhesions
compared to the Wt, which is consistent with the known role of Syndecan-4 in promoting
focal adhesion formation through interaction with the fibronectin Hep II domain [25]. N-sulfation is essential for focal adhesion formation in primary rat fibroblasts plated on
fibronectin in the presence of competitive heparin oligomers or variably desulfated heparin
oligomers or in the absence of oligomers, and 6-O-sulfate is the next most important. 2.3. Sugar Skeleton of HS The sugar skeleton of HS is a repeat of GlcA/IdoA-GlcNAc, but there are theoretically
1036 different molecules depending on the position and number of sulfate groups attached. Therefore, it is currently not possible to determine whether the “ligand binding redun-
dancy” of HS is truly redundant or binding via each specific sugar chain structure. Signal
pathways mediated by cell surface receptors usually involve a dimerization process upon
ligand binding. Syndecan can self-associate through a transmembrane domain containing
a specific conserved motif GXXXG [26]. The delicate balance between anabolism and
catabolism is critical to tissue homeostasis, structure, and function. For example, matrix
metalloproteinases (MMPs) cleave many extracellular matrix proteins, shedding Syndecan
from the cell membrane and affecting signaling [29,30]. g
g
g
Furthermore, in some cancer types, HSPG expression is different from healthy tis-
sue. Thus, enzymes that modify, cleave, shed, and degrade HSPG are important factors
in catabolic reactions, while glycosaminoglycan biosynthetic enzymes control anabolic
reactions [22,30–32]. Syndecan functions as a receptor that transmits signals from the extra-
cellular microenvironment to the inside of the cell [8,33]. Fibroblast growth factor-2 (FGF-2)
bound to HS significantly increases affinity for FGFR [34,35]. Cells lacking cell surface
HSPGs have reduced proliferation rates that can be restored by exogenous heparin [36]. In
addition, interaction with HS protects growth factors from proteolytic degradation, thereby
increasing their half-life in the cellular environment. Antithrombin III specifically inter-
acts with heparin, while HGF can bind both HS and dermatan sulfate to exert biological
activity and requires a unique pentasaccharide structure containing a rare 3-O-sulfated
N-unsubstituted glucosamine (GlcN3S), which induces structural changes in the protein
and increases the degree of thrombin inhibition. Glucosamine 3-O-sulfation of the HS chain
is a rare modification but critical for several specific HS–protein interactions [37,38]. i
pi
p
Because heparin has a high IdoA2S-GlcNS6S content, in the absence of HSPGs, core
protein papillomavirus (HPV-16) can infect epidermal cells [39]. Therefore, structural
changes in the sulfation pattern of HS can change ligand binding and HS-mediated cellular
signaling [23]. The ligand interacts with heparan sulfate on the cell surface in cooperation
with high-affinity receptors [36]. The fact that the ligand binding site is the HS sugar side
chain means that Syndecan could theoretically serve as a receptor for all heparin-binding
molecules [40]. Many extracellular matrix constituent molecules and cell growth factors
bind to the HS sugar side chain of Syndecan. 2.2. Functions of the Syndecan Family The sulfation pattern of HS also influences the mode of ligand binding to HSPGs [23]. Syndecan-1, which are expressed at approximately the same level on the surface of myeloma
cells, share a common core protein and have HS chain sugars of approximately similar
size. However, only Syndecan-1 on MPC-11 cells, which is a B lymphocyte cell line of a
mouse with plasmacytoma, mediates cell adhesion by binding to type I collagen. Syndecan-
1 HS from MPC-11 exhibits high N-sulfation 2-0-sulfation and low 6-O sulfation. This
indicates that HS sulfation and the specific sulfation pattern, rather than the true charge, are
important for differences in binding to collagen. Each of these HSPGs has a cell-type-specific
distribution, and in vascular endothelial cells, expression of Syndecan-1, -2, and -4 is mainly
observed between cells and on the basement membrane side [1]. The Syndecan family
belongs to type I membrane proteins and forms a family with homology to each other in
their transmembrane and cytoplasmic domains. The cytoplasmic amino acid sequences of
Syndecan proteins show high homology within the family. In particular, the submembrane
and C-terminal regions are highly conserved and are therefore responsible for triggering
signal transduction cascades common to the family. Also, since the central arrangement
is unique to each member, it triggers a cascade unique to each member. In addition,
this transmembrane–intracellular domain contains four tyrosine residues with conserved
positions, and phosphorylation of any of the tyrosine residues participates in intracellular
signal transduction [26]. These intracellular domain ends are involved in association with
intracellular microfilaments in Syndecan-1 and focal adhesion in Syndecan-4. Furthermore, in Syndecan-1, which was cloned from mouse mammary epithelial cells,
the sugar chain contains CS in addition to heparan sulfate, and the ratio and size of these
differ depending on the tissue in which they are expressed. Additionally, Syndecan-1 is
expressed as an HSPG with anticoagulant activity in rat microvascular endothelial cells. It
was isolated from golden hamsters as a fibroblast growth factor 2 (FGF-2) receptor, and the Biologics 2024, 4 109 sugar chain has been modified to have different functions depending on the cell in which it
is expressed [27,28]. 3. Structure and Function of Vascular Endothelial Glycocalyx
3.1. Structure and Function of Proteoglycan Proteoglycan, the main component of the vascular endothelial glycocalyx, which is a
structure composed of glycoproteins and polysaccharides that covers the cell surface of the
vascular lumen, functions as the most important skeletal molecule of the glycocalyx, which
is usually composed of proteoglycan and glycoproteins that are bound to the cell membrane
and surface layer, as well as polysaccharides that are not bound to the cell membrane surface
in the outer layer [49,50]. The main components of the glycocalyx are glycosaminoglycans
such as HS, CS, and hyaluronic acid (HA), and core proteins such as Syndecan that support
them, which are also made up of complex bonds of sugar chains. Furthermore, since
the presence of glycocalyx in the vascular endothelium has the function of preventing
substances from flowing out of the blood vessel and adhering to the vascular endothelium,
it is attracting attention in intensive care fields such as infusions and anesthesia [51,52]. Glycocalyx is a jelly-like, band-like structure of glycoprotein that covers the surface of
cells such as vascular endothelium and has various physiological functions [53,54]. The
glycocalyx has a variety of functions, including maintaining blood flow, protecting the
endothelium, antithrombotic effects, binding and signal transmission of hormones and
growth factors, and serving as a nutritional source consisting of sugar chains [54,55]. Vascular endothelial glycocalyx is a carbohydrate-rich layer located on the luminal surface
of the vascular endothelium [56]. It binds to the endothelium primarily through skeletal
molecules such as proteoglycans and glycoproteins and incorporates soluble molecules
derived from plasma or endothelium to form a mesh structure [49,54]. The lumen side
of the glycocalyx is formed by soluble plasma components bound directly to membrane-
bound proteoglycans and glycoproteins or indirectly through sugar chains such as soluble
proteoglycans [57]. There is a dynamic equilibrium between the soluble components on the
lumen side of the glycocalyx and the flowing blood, and this intermittently influences the
composition and thickness of the glycocalyx [57,58]. 2.3. Sugar Skeleton of HS Adsorption of Herpes simplex virus type 1
(HSV-1) to the target cell surface via HS occurs when viral glycoproteins gB and gC bind
to HS. This binding of HS by gB and gC is involved in the binding of virus particles to
the cell surface, but not in the entry of virus particles into cells. For HSV-1 to enter cells,
the glycoprotein gD must bind to entry receptors, resulting in membrane fusion between
the viral envelope and the host cell membrane [41,42]. In addition to nectin cell adhesion
molecule 1 (NECTIN-1), nectin cell adhesion molecule 2 (NECTIN-2), and herpesvirus
entry mediator (HVEM), 3-O-sulfated HS has been identified as one of these entry receptors. Interestingly, among the six human 3-O-sulfotransferases (3-OSTs), 3-O-sulfated HS, which
is sulfated by six 3-OSTs (3-OST-2, 3-OST-3A, 3-OST-3B, 3-OST-4, 3-OST-5, and 3-OST-6),
except 3-OST-1, interacts with gD and functions as an entry receptor [37,43,44]. Multiple
sulfotransferases share activity and form sugar chains with more complex sulfation patterns,
thereby acting as separate functional domains. In addition to HSV-1, a wide variety of
virus particles interact with HS, including human immunodeficiency virus, hepatitis B
virus, dengue virus, human papillomavirus, cold coronavirus, vaccinia virus, and Lassa
virus [39,45,46]. The electrostatic interaction between the positive charge of the viral protein
that makes up the surface of the virus particle and the negative charge of HS is involved in
the adsorption of the virus particle to the cell surface. SARS-CoV-2 (severe acute respiratory
syndrome coronavirus 2), the virus that causes COVID-19 (CoronaVirus Infectious Disease Biologics 2024, 4 110 in 2019), also uses HS to bind to target cell surfaces [46–48]. SARS-CoV-2 is caused by the
Spike protein, an S protein present in the envelope binding to the Angiotensin-converting
enzyme 2 (ACE2) receptor present in the cell membrane, entering the cell, when the S
protein also binds to HS, increasing its binding to the ACE2 receptor [37–39]. 3.2. Vascular Endothelial Glycocalyx As a result of performing acute normovolemic hemodilution and volume loading (VL)
using the same amount of HES (20) mL on surgical patients, only VL increases blood Enzymes such as MMP, heparinase, and hyaluronidase are involved in the decomposi-
tion process of glycocalyx [54]. These enzymes are activated in the presence of inflammatory
cytokines and reactive oxygen species. Since ANP-induced glycocalyx damage was reduced
under the use of MMP inhibitors, activation of MMP by ANP contributes to glycocalyx
damage [68]. The vascular endothelial glycocalyx is a complex, self-assembled three-dimensional
mesh of various polysaccharide molecules. This is sheared and released by enzyme reac-
tions associated with inflammation and shear stress [69]. In the vascular system, HSPGs,
a major component of the glycocalyx, enable the binding of various plasma-derived
molecules. Because of their diversity, epimerization of sugar chains, long chains, and
sulfation account for approximately 50 to 90% of the total amount of proteoglycans in
the glycocalyx [32,70]. The binding of enzymes, ligands, and their receptors to sugar
chains locally increases their concentrations, facilitating necessary signaling and enzyme
modification reactions, and functions in blood vessel protection [71]. Several important
anticoagulant mediators are present in the endothelial glycocalyx, including antithrombin,
heparin cofactor II, thrombomodulin, and tissue factor pathway inhibitor (TFPI), which is
produced in vascular endothelial cells, is a strong inhibitor of FVIIa and FXa, and exists
bound to the glycocalyx via heparan sulfate. Syndecan-4 derived from human vascular en-
dothelial cells binds to heparin-binding proteins FGF-2, midkine, and TFPI via its heparan
sulfate sugar chains. Antithrombin, produced and secreted by hepatocytes, is a powerful
inhibitor of thrombin and procoagulant proteases such as FXa and FIXa [72]. Their binding
to a specific region of heparan sulfate enhances anticoagulant ability hundreds of times [73].i pi
g
p
g
y
[
]
Heparin cofactor II, which is also produced and secreted by hepatocytes, is a specific
inhibitor of thrombin and exhibits antithrombin activity by binding to dermatan sulfate
in the glycocalyx. Thrombomodulin is a CS-containing glycoprotein expressed on the
vascular endothelial cell membrane, and when it binds to thrombin, it abolishes its proco-
agulant effect [74,75]. On the other hand, it exhibits an anticoagulant effect by activating
protein C, a coagulation inhibitor produced by liver cells. However, proteoglycan ex-
pression in endothelial cells changes due to various stimuli. 3.2. Vascular Endothelial Glycocalyx Glycocalyx is a structure with a thickness of several hundred nm that exists on the
vascular endothelium [56,59,60]. It is composed of glycoproteins such as proteoglycans and
adhesion molecules that are bound to the cell membrane, and HA that is not directly bound
to the cell membrane. Proteoglycan is a core protein with GCG bound to it. Plasma proteins
such as albumin and antithrombin bind to the vascular lumen side of glycocalyx, forming
an endothelial surface layer (ESL). The physiological functions of glycocalyx and ESL are
diverse, including regulation of vascular permeability, mechanoreceptors for shear stress,
regulation of leukocyte adhesion and migration, and control of coagulation and fibrinolysis
within blood vessels [61]. Glycocalyx is a very fragile structure. Due to various stimuli and
pathological conditions, it peels off, releases into the blood, and then regenerates, a process
that repeats repeatedly. As one way to recognize glycocalyx damage, blood concentrations
of these released glycocalyx components, including Syndecan-1, HS, and HA, etc., are
measured for research purposes [62,63]. On the other hand, visualizing glycocalyx is an
extremely difficult task, and even with current technology, it is difficult to directly evaluate
glycocalyx structure in vivo. Injury factors for glycocalyx include sepsis, ischemia–reperfusion, hypervolemia due
to the use of cardiopulmonary bypass, and hyperglycemia [63]. Among these, hyper-
volemia damages glycocalyx via atrial natriuretic peptide (ANP) secretion [64]. ANP
administration reduces the glycocalyx structure observed by electron microscopy and
promotes hydroxyethyl starch (HES) extravasation, even at physiological concentrations. Biologics 2024, 4 111 As a result of performing acute normovolemic hemodilution and volume loading (VL)
using the same amount of HES (20) mL on surgical patients, only VL increases blood
ANP concentration and increases the blood concentration of Syndecan-1 and HA [65]. In
addition, in coronary artery bypass surgery, regardless of on-pump or off-pump surgery,
blood concentrations of ANP and glycocalyx degradation products increased by several to
several dozen times [66]. The increase in blood concentration of decomposition products
precedes the increases in inflammatory cytokines and is triggered by the increase in blood
ANP concentration [67]. Furthermore, multivariate analysis has shown that there is a
relationship between cumulative fluid volume during initial resuscitation and glycocalyx
injury in patients with sepsis. 3.2. Vascular Endothelial Glycocalyx In
classical theory, the movement of substances from the capillary side to the interstitial side,
or from the interstitial side to the capillary side, is caused by the difference in osmotic
pressure between the two sides and moves through the small gap between endothelial cells. When many substances, which move due to the difference in osmotic pressure between the
upper and lower sides of the glycocalyx covering the capillaries, pass through, the gap in
the endothelium opens [54,92,93].i p
The microstructure of the lumen of capillaries is classified into three types: the continu-
ous type with almost no gaps in the lumen, the fenestrated type with small holes distributed,
and the sinusoidal type, with holes of various sizes. The capillary endothelium in the brain,
heart, lungs, etc., is continuous, and its surface is densely covered with a glycocalyx [94,95]. The digestive tract, kidneys, and endocrine organs are fenestrated and have a glycocalyx
that covers the holes [55,65,96]. The liver and bone marrow are sinusoidal, with large holes,
and are not covered by a glycocalyx [97]. Thus, substances enter and exit the capillaries of
the gastrointestinal tract, kidneys, liver, etc., freely, to a certain extent, through holes that
are originally opened and are regulated by the function of the glycocalyx [94,98,99]. The
permeability of capillaries in each organ can be explained by a combination of the three
classifications of their microstructure and the glycocalyx [94,100,101]. Even though the
capillary structure is the same, the thickness of the glycocalyx differs depending on the
organ [51,59,102]. Comparing their ultrafine three-dimensional structures by organ, the
fine structure of the vascular endothelium in the lungs, heart, and brain is classified into
the same continuous types, and the same structure was confirmed in electron microscopy
images [103]. However, the glycocalyx is thin in the lungs, intermediate in the heart, and very thick
in the brain [98,102]. Even if the blood vessels have the same structure, the glycocalyx
differs depending on the organ, in which the brain is thick to protect brain cells from
cytokines and toxins, and the lungs are thin to carry out the gas exchange of oxygen and
carbon dioxide. The vascular endothelial glycocalyx of mouse models of diabetes, aging,
and dietary restrictions is thinner than that of normal mice [104]. 3.2. Vascular Endothelial Glycocalyx For example, Syndecan is
changed and regulated by the activation of endothelial cells and stimulated by different
chemokines [49,76]. The second most common sugar chain in the vascular endothelial
glycocalyx is CS/dermatan sulfate, and the abundance ratio of heparan sulfate and CS in
the vascular endothelium is approximately 4:1 [55,60,77]. Another important sugar chain in
the vascular endothelial glycocalyx is hyaluronan [60,78,79]. These long polymer molecules
with up to 10 kDa differ from other sugar chains in that they are not bound to a core protein,
forming surprisingly viscous solutions [80]. Sugar chains have specific binding sites for
many plasma proteins, and even slight differences in chain modification can significantly
change their function, such as loss of binding ability [81]. Modifications at individual sugar
units within sugar chains confer unique functions to proteoglycans, but there are typically
16 to 48 different sulfation patterns per disaccharide chain [82]. The length of the functional
domain is typically five to ten glycans. At least 163 different sulfation patterns are theoreti-
cally possible, and the various structures correspond to the diverse biological functions of
the sugar chains [83]. Indeed, modification patterns change over time and under different
pathological and physiological stimuli [81,84–86]. The diversity of sugar chain sulfation Biologics 2024, 4 112 patterns and their effects on specific protein binding and functional regulation change the
thickness of the glycocalyx, sugar chain sulfation patterns, and their charges, regulating
vascular permeability, and specific binding proteins and their activities [82,85,87].l p
y
pi
g p
In addition, the vascular endothelial glycocalyx sensitively reflects the health status
of vascular endothelial cells, and under normal conditions, it remains thick and slippery,
but when the vascular endothelium itself is damaged due to diabetes, etc., the glycocalyx
becomes thinner and its function decreases [78,88,89]. In particular, the glycocalyx in
peripheral capillaries is more important than in large blood vessels because the thickness
of the glycocalyx is relatively larger than the vessel diameter [55]. Capillaries, which
extend throughout the body’s organs, account for 99% of all blood vessels. This exchanges
information substances such as nutrients, oxygen, and hormones with six billion cells, and
is not surrounded by smooth muscle, instead consisting of endothelial cells that form the
vascular lumen and pericytes that are sparsely present around the periphery [90,91]. 3.2. Vascular Endothelial Glycocalyx When inflammation is
induced in model mice if the glycocalyx is thick, the endothelium is likely to be protected,
but if it is thin, the inflammation will be prolonged [56,105]. Additionally, if infected with
COVID-19, people with underlying diseases such as diabetes or cancer are more likely
to develop blood clots and become more seriously ill, which may also be due to the thin
glycocalyx [106,107]. In addition, when the glycocalyx becomes thinner, damage caused by
stimulation and an increase in abnormal signal transmission occurs, leading to the onset or
worsening of various diseases such as prolonged inflammation and cancer growth [105]. Elucidation of the mechanisms leading to the onset and deterioration of the disease and
the development of treatments targeting the glycocalyx still require further research, but
this is an area that is expected to continue to develop [51,55,106]. In addition, blood vessels
are essential organs that control circulation throughout the body, as well as inflammatory
cytokines such as Tumor Necrosis Factor (TNF) and Interleukin-6 (IL-6) [108,109]. Further,
interleukin-1 beta (IL-1beta) and growth factors such as hepatocyte growth factor (HGF), Biologics 2024, 4 113 fibroblast growth factor (FGF), and vascular endothelial growth factor (VEGF) are also
transported to organs through blood vessels and are involved in the development of various
diseases and tissue repair [110,111]. fibroblast growth factor (FGF), and vascular endothelial growth factor (VEGF) are also
transported to organs through blood vessels and are involved in the development of various
diseases and tissue repair [110,111]. 4. Molecular Mechanisms of the Syndecan Family
4.1. Syndecan-4 Functions Syndecan-4 plays an important role in biological defense against inflammation and
tissue damage [6,76,112]. In addition, the expression of Syndecan-4 is increased in the
vascular wall after vascular injury. Syndecan-4 functions in cell proliferation and migration
of vascular smooth muscle cells induced by thrombin. For HS–growth factor interactions,
it is still unclear whether distinct HS sequences are required. The degree of global sulfation
defines HS-FGF interactions in a nonspecific manner. Using FGF-1 and FGF-2, 2-O-sulfation
and 6-O-sulfation can be interconverted without loss of physiologically significant interac-
tions. Therefore, for many proteins, what is important is the composition of the HS domain,
rather than the precisely determined HS sugar chain sequence. The expression of 3-O-
sulfotransferase was regulated through RAS/mitogen-activated protein kinase (MAPK),
fibroblast growth factor receptor 2 (FGFR2), and Kit signals [113]. Biotechnically produced
3-O-sulfate-rich HS interacts with FGFR2b, thereby promoting the formation of a ternary
complex between fibroblast growth factor receptor 10 (FGF-10) and FGFR2b [114]. Ad-
ditionally, microRNAs can control HS chain modification, which changes the binding of
growth factors to sugar chains [3,22]. Sydecan-4 always exists as a dimer on the cell surface, which becomes multimerized
by binding to ligands such as the heparin-binding domain of fibronectin [115]. Furthermore,
in focal adhesions, high concentrations of Syndecan-4 increase the probability of oligomer
formation that promotes PKC signaling. Membrane-bound HSPGs, including Syndecan,
glypican, beta-glycan, neuropilin-1, and CD44v3, an isoform of CD44, are found in the
plasma membrane of nearly all eukaryotic tissues and can act as co-receptors that regulate
cell signals and behavior, or as receptors for endocytosis and adhesion [3]. 4.2. HS Chains and Inflammatory Responses Syndecan HS chains are also important in inflammatory responses by binding to
chemokines [60]. In rheumatoid arthritis (RA), the chemokine CXC chemokine ligand
8 (CXCL8) binds to the HS chain of Syndecan-3, and the excision of the HS chain from
Syndecan-3 by bacterial heparanases I and III abolished the binding of CXCL8 to rheuma-
toid synovial endothelium [116,117]. In addition, the administration of chemokines engi-
neered to have reduced affinity for HS abolishes leukocyte recruitment to RA joints [118]. Chemokine CXC chemokine ligand 1 (CXCL1), a functional rodent homolog of CXCL8,
binds to HS on the cell surface of mice [119]. Syndecan-3 knockout mice exhibit reduced
neutrophil accumulation in the knee joints due to CXCL1 administration and attenuated in-
flammatory responses and RA compared to Wt mice [116]. Thus, the role of Syndecan-3 on
leukocyte activation and migration via the HS chain is shown [116,120]. Specific deficiency
of HS 2-O-sulfotransferase in endothelial and bone marrow cells reveals a specific HS func-
tion in inflammatory cell recruitment. Inactivation of this gene in endothelial cells markedly
altered the sulfation status of disaccharides, showing a decrease in 2-O-sulfate content
as expected, but 6-O-sulfation and N-sulfation increased. This altered sulfation pattern
promoted increased binding of cytokines [121,122]. Furthermore, neutrophil recruitment
was enhanced in these animals after inflammatory stimuli, leading to intensified neutrophil
rolling [123]. The latter is because of increased interaction between neutrophil L-selectin
and HS on endothelial cells. In addition to chemokine binding, the pleiotropic roles of
HSPGs in immune system signaling are clear. In vitro experiments show that Syndecan-4
HS in CTCL cells acts as a cell surface reservoir for TGFb1 [112]. Additionally, the same HS
site was shown to bind dendritic cell HSPG–integrin ligand on activated T cells, in which
interactions produce immunosuppressive effects. Biologics 2024, 4 114 5.1. Sulfatases Function and Group Sulfatases are enzymes that remove sulfate groups from a variety of molecules, in-
cluding steroids, sulfated lipids, glycoproteins, and proteoglycans. Removal of sulfate
groups from HS alters interactions with growth factors and affects downstream signals [10]. Extracellular endosulfatases Sulfatase 1 (Sulf-1) and Sulfatase 2 (sulf-2) selectively re-
move 6-O-sulfate from glucosamine, thereby changing the binding properties of sugar
chains [129]. These sulfatases need to be processed by furin-type proteases to become
active enzymes. 6-O-sulfate of the HS chain is important for signals from FGF. Sulf-1 in-
hibits FGF-2-promoted thymoma viral proto-oncogene 1 (AKT/PKB) and ERK signaling in
hepatoma cell lines. Sulf-1 decreases the expression of cyclin D1 and survives, leading to
cell cycle arrest and apoptosis. Ultimately, Sulf-1 attenuates hepatoma cell proliferation
in vivo. Similarly, Sulf-1 expression is decreased in several types of cancer, including breast
cancer. Furthermore, expression of Sulf-1 in MDA-MB-468 breast cancer cells overexpress-
ing epidermal growth factor receptor (EGFR) reduces cell proliferation in vitro and cancer
size in vivo. The molecular mechanism underlying these Sulf-1-mediated effects involves
the inhibition of autocrine ERK signaling induced by amphiregulin and heparin-binding
EGF [130]. Apart from their impact on cancer biology, sulfatases have a role in maintaining
tissue homeostasis. Sulf-1 and Sulf-2 knockout mice survived, and no obvious physiological
defects were observed [131,132]. On the other hand, double-knockout mice for both sulfa-
tases died around birth. Sulf-1 knockout mice exhibit more severe cartilage degeneration
in surgically induced osteoarthritis than Wt mice. In addition, this mouse exhibits high
expression levels of matrix metallopeptidase 13 (MMP-13) and a disintegrin and metal-
loproteinase with thrombospondin motifs 5 (ADAMTS-5) and an abnormal extracellular
matrix, in which there are low mRNA levels of collagen type 2a1 (col2a1) and aggrecan. This phenotype is associated with increased phosphorylation of Erk1/2, while decreased
Smad1 protein expression and Smad1/5 phosphorylation. Sulf-2 knockout mice exhibit a
similar phenotype. The degree of sulfation differs between various organs in Sulf-1 and
Sulf-2 knockout mice, and each of these sulfatases controls organ-specific sulfation patterns
of HS chains [133]. 4.3. Syndecan and Cell Signals Syndecan can control cell signals in vitro and in vivo experiments [120]. Several
phenotypic changes in FGF-2-mediated signaling induced overexpression of Syndecan-
4 in endothelial cells [115,124]. Syndecan-1 null mice were viable and fertile but had
reduced mammary branching morphogenesis, in which a reduced number of collaterals
characterized the glands of lower morphology [125]. Additionally, by crossing Syndecan-1 null mice with mice expressing constitutively
active intracellular beta-catenin, Syndecan-1 was suggested to be required for catenin-
induced mammary gland hyperplasia and cancer development in vivo [126]. Furthermore,
the expression of Syndecan is enhanced in myocardial infarction models. In patients with
myocardial infarction, a marked increase in blood Syndecan-4 was detected, peaking two
weeks after onset, and clear Syndecan-4 staining was observed in the tissue repair area in
autopsy tissue images, indicating that Syndecan-4 is rapidly expressed during ischemic
myocardial injury and function in tissue repair [127,128]. 5.2. Sulfatases and Cell Signaling In addition, alternative splicing of the Sulf-1 gene provides further enzymatic regula-
tion of cellular signals [130]. Short-splicing variant of Sulf-1B suppresses Wnt signaling in
osteoblasts, while full-length Sulf-1 induces Wnt signaling [134–136]. Furthermore, Sulf-1B
induces angiogenesis, whereas Sulf-1 causes the opposite effect. Expression of Sulf-2 is
elevated in glioblastoma patients and correlates with cell proliferation and tumorigene-
sis [137]. The molecular mechanism of Sulf-2-mediated effects involves the regulation of
several tyrosine kinase-type receptors, such as Platelet-derived Growth Factor Receptor a Biologics 2024, 4 115 (PDGDRa), insulin-like growth factor 1 receptor 1 beta (IGF1Rbeta), and Erythropoietin-
producing hepatoma receptor-A2 (EPHA2). Sulf-2 promotes cell proliferation, which was
associated with increased expression of glypican-3 [138,139]. This HSPG induces higher
binding of FGF-2 and activation of ERK and AKT signaling pathways in hepatoma cell
lines [45]. Sulf-2 enhances tumor growth in nude mice and is associated with poor patients’
prognosis [140,141]. Sulf-2 is elevated in breast cancer compared to healthy controls and
promotes angiogenesis [142]. Furthermore, the knockdown of Sulf-2 reduces breast cancer
proliferation and increases luminal cell apoptosis. Furthermore, tumors lacking Sulf-2
retain lumen formation and basement membrane integrity, which is concomitant with
decreased matrix metalloproteinase-9 (MMP-9) expression and activity. On the other hand,
when MDA-MB-231 cells, basal-like breast cancer were treated with Sulf-2, phosphory-
lation of ERK induced by FGF-2 or HB-EGF was reduced, resulting in a decrease in cell
invasion and proliferation. Furthermore, heparanase is secreted as an inactive zymogen,
taken up through HSPGs such as Syndecan-1, and, finally, transported to lysosomes, where
heparanase is converted into an active form by cathepsin-L [143,144]. Heparanase rec-
ognizes a specific HS sequence and directly cleaves between GlcA and GlcNS with 3-O-
or 6-O-sulfate [145]. On the other hand, polysaccharides containing IdoA2S-GlcNS (2-O-
sulfo-α-L-iduronic acid- glucosamine-N-sulfate) inhibit heparanase enzyme activity. The
released HS fragments serve as competitors for HSPG–protein interactions but can also bind
to Toll-like receptor 4 (TLR-4), eliciting innate immune responses [146,147]. Heparanase
induces malignant transformation and is associated with poor prognosis in several cancer
types [118]. Therefore, several heparanase inhibitors have been developed, but most of the
inhibitors are sulfated sugars or HS analogs and exhibit anti-angiogenic, anti-metastatic,
and anticancer activities [148,149]. 6. Function of Syndecan via Extracellular Vesicle (EV) and miRNAs
6.1. EVs and Heparanase
6. Function of Syndecan via Extracellular Vesicle (EV) and miRNAs
6.1. EVs and Heparanase EVs are extracellular granules that represent important communication pathways be-
tween cells, transporting molecules that affect cellular signals, such as proteins, lipids,
microRNAs, and mRNAs [156,157]. Most cells produce EVs, but under pathological con-
ditions, EV secretion is increased and correlated with cancer progression [158–160]. Mye-
loma cells that highly express heparanase secrete significantly more EVs than control cells
[161]. In addition, exogenously applied heparanase enhances EV secretion in other types
of cancer cells [162,163]. These effects depend on the enzymatic activity of catalytically
inactivated heparanase, as it did not induce EV release [164]. EVs derived from cells with
high heparanase expression contain increased amounts of Syndecan-1, VEGF, and HGF,
which regulate cancer and host cell behavior [128,165,166]. Syndecan core protein is key
to EV formation through interaction with the syntenin/ALIX (ALG-2-interacting protein
X) complex [167,168]. Besides controlling EV biogenesis, HSPGs are important for EV up-
take and biological activities, e.g., EV-induced ERK1/2 signal activation [169,170]. Poor
prognosis in myeloma patients correlates with high HGF levels [171,172]. In this context,
heparanase influences cellular signals by regulating growth factor expression and activity
[147,173]. Analysis of bone marrow biopsies from myeloma patients shows a positive cor-
relation between heparanase and HGF levels [161]. In vitro experiments showed that hep-
aranase induces the expression and secretion of active HGF in myeloma cells, and that
enzyme activity is not required for this effect. Furthermore, heparanase increases the for-
mation of active complexes between HGF and Syndecan-1 released from the membrane
[165]. EVs are extracellular granules that represent important communication pathways
between cells, transporting molecules that affect cellular signals, such as proteins, lipids,
microRNAs, and mRNAs [156,157]. Most cells produce EVs, but under pathological
conditions, EV secretion is increased and correlated with cancer progression [158–160]. Myeloma cells that highly express heparanase secrete significantly more EVs than control
cells [161]. In addition, exogenously applied heparanase enhances EV secretion in other
types of cancer cells [162,163]. These effects depend on the enzymatic activity of catalytically
inactivated heparanase, as it did not induce EV release [164]. EVs derived from cells
with high heparanase expression contain increased amounts of Syndecan-1, VEGF, and
HGF, which regulate cancer and host cell behavior [128,165,166]. Syndecan core protein
is key to EV formation through interaction with the syntenin/ALIX (ALG-2-interacting
protein X) complex [167,168]. Besides controlling EV biogenesis, HSPGs are important for
EV uptake and biological activities, e.g., EV-induced ERK1/2 signal activation [169,170]. Figure 2. Schematic representation of the interaction between HSPGs, growth factors, receptors, and
downstream signaling pathways. HSPGs bind several other growth factors such as hepatocyte
growth factor (HGF), fibroblast growth factor (FGF), heparin-binding epidermal growth factor-like
growth factor (HB-EGF), transforming growth factor (TGF) beta, and vascular endothelial growth
factor (VEGF), and modulate their signaling, including SMAD, PI3K/AKT, MAPK/ERK, JAK/STAT,
and Rho/PKC [155]. Figure 2. Schematic representation of the interaction between HSPGs, growth factors, receptors,
and downstream signaling pathways. HSPGs bind several other growth factors such as hepatocyte
growth factor (HGF), fibroblast growth factor (FGF), heparin-binding epidermal growth factor-like
growth factor (HB-EGF), transforming growth factor (TGF) beta, and vascular endothelial growth
factor (VEGF), and modulate their signaling, including SMAD, PI3K/AKT, MAPK/ERK, JAK/STAT,
and Rho/PKC [155]. 6. Function of Syndecan via Extracellular Vesicle (EV) and miRNAs
6.1. EVs and Heparanase
6. Function of Syndecan via Extracellular Vesicle (EV) and miRNAs
6.1. EVs and Heparanase Poor prognosis in myeloma patients correlates with high HGF levels [171,172]. In this
context, heparanase influences cellular signals by regulating growth factor expression
and activity [147,173]. Analysis of bone marrow biopsies from myeloma patients shows
a positive correlation between heparanase and HGF levels [161]. In vitro experiments
showed that heparanase induces the expression and secretion of active HGF in myeloma
cells, and that enzyme activity is not required for this effect. Furthermore, heparanase
increases the formation of active complexes between HGF and Syndecan-1 released from
the membrane [165]. 5.3. HSPGs and Reservoir for Bioactive Molecules HSPGs present in the extracellular matrix serve as a reservoir for bioactive molecules
(Figure 2) [150]. In a physiological situation using a model of mammary duct branching
morphogenesis, heparanase activity promotes the release of fibroblast growth factor-10
(FGF-10) bound to the HS of perlecan. This allows growth factors to bind to fibroblast
growth factor receptors (FGFRs), increasing mitogen-activated protein kinase (MAPK)
activation, epithelial invasion, and lateral branch formation [151]. Similarly, fibroblast
growth factor-2 (FGF-2) and VEGF are released from the extracellular matrix by the action
of heparanase and control the angiogenic response [152], while heparanase can regulate
cellular signals independently of its enzymatic activity. For example, mutated and inactive
heparanase promotes VEGF activation mediated by proto-oncogene tyrosine-protein kinase
Src and p38 mitogen-activated protein kinase activation. Besides hydrolyzing glycosidic
bounds, heparanase promotes changes in the sulfation pattern of HS, which, in turn,
modulates cellular signals [23,153]. Heparanase overexpression increases HS turnover
as well as N- and O-sulfate content in the liver [154]. Additionally, HS6-O-sulfation is
higher in other organs of mice overexpressing heparanase. Similarly, cancers that highly
express heparanase correlate with increased HS 6-O-sulfation [22]. This change in sulfation
promotes the formation of a tripartite complex between fibroblast growth factor (FGF) and
FGFR1 and modulates signal transduction. Biologics 2024, 4
Biologics 2024, 4, 116
12 6.2. Sydecans Control Cellular Functions
6.2. Sydecans Control Cellular Functions 6
Sy e a s Co
o Ce u a
u
io s
Transmembrane Sydecans control cell adhesion and migration through interactions
with signal molecules and the cytoskeleton. On the other hand, it interacts with
Transmembrane Sydecans control cell adhesion and migration through interactions with
signal molecules and the cytoskeleton. On the other hand, it interacts with extracellular Biologics 2024, 4 117 matrix molecules, integrins, growth factors, and other ligands [66]. For example, Syndecan-
2, as a receptor for the cell adhesion substrate fibronectin, controls the fibronectin-binding
information of integrin a5b1 and induces different cytoskeleton formation. Thus, Syndecan
mediates cell adhesion and migration via cytoskeleton architecture as an extracellular matrix
receptor and cell proliferation as a growth factor low-affinity receptor. Syndecan shedding is
an important mechanism that controls the amount of Syndecan on the cell surface. Furthermore, the shedding of Syndecan is an important phenomenon because shed
Syndecan can serve as a soluble ligand [174]. Although the Syndecan extracellular domain
may be constitutively shed into the extracellular space, the rate at which this occurs in vivo
is poorly understood. However, pathological conditions and some stimuli such as growth
factors, chemokines, microbial toxins, insulin, heparanase, and cellular stress, increase the
rate of shedding. Matrix metallopeptidase (MMPs) primarily catalyze Syndecan shedding,
usually cleaving the extracellular domain of Syndecan close to the transmembrane do-
main. Shredded Syndecan-1 acts as a paracrine factor during cancer progression [127,165]. Syndecan-1 shed from fibroblasts induces cell division response in human breast cancer
cells using a three-dimensional co-culture model. HS chains, FGF-1, and stromal-derived
factor 1 are key factors in this paracrine phenomenon [1,114]. Furthermore, in MCF-7 breast cancer cells, overexpression of Syndecan-1 ectodomain
rapidly promotes the acquisition of invasiveness in a normally poorly invasive cell line [175]. Apart from metalloproteases, heparanase promotes the shedding of Syndecan-1 from
myeloma cells by several mechanisms [166]. First, heparanase cleaves HS chains and
enhances the sensitivity of core proteins to matrix metalloproteinases (MMPs). Increased
heparanase maintains ERK signaling [176,177]. This increases the expression level of MMP-
9, potentially increasing the shedding rate of Syndecan-1, which is shed from myeloma cells
and acts in a paracrine manner [166]. For example, culture supernatants from myeloma
cells that highly express heparanase contain shed Syndecan-1 and VEGF. g
y
p
p
y
Chemotherapy and radiotherapy, the most common treatments for cancer patients, in-
duce Syndecan shedding [127,165]. 6.2. Sydecans Control Cellular Functions
6.2. Sydecans Control Cellular Functions However, this can potentially be considered a negative
side effect as it creates a favorable microenvironment for cancer progression [178,179]. In
addition, treatment of myeloma mice with chemotherapeutic drugs increases the amount
of shed Syndecan-1 compared to control animals. The molecular mechanism involved in
chemotherapy-induced shedding of Syndecan-1 involves the activation of caspases and
ADAM metallopeptidase domain (ADAMs) consisting of disintegrin and metal proteinase
domains [127,166]. MMP was the metalloprotease that mediated Syndecan-1 shedding in
this case [22]. Furthermore, miR-494 was identified as a microRNA involved in Syndecan-1
shedding [127]. 7.2. Structure, Expression, and Functions of CD44 CD44 on the surface of lymphocytes and fibroblasts binds to HA, transmits signals
by phosphorylating a kinase-like protein kinase C and phosphorylating serine/threonine
residues within the CD44 molecules, and controls cell adhesion, proliferation, connective
tissue component synthesis, etc. [191,192]. CD44 is a family with multiple isotopes and is a
type 1 transmembrane glycoprotein that is expressed in a wide range of tissues and cells in
normal living organisms [193,194]. These molecules bind to the extracellular matrix and
are involved in cell movement and aggregation. Abnormal expression of these molecules
has been reported to play an important role in tumor invasion and metastasis. The human CD44 gene locus is in the p13 region of chromosome 11, and its genetic
structure consists of at least 20 exons. Some exons undergo alternative splicing, resulting in
multiple isotopes and isoforms, and, to date, more than 20 combinations of isotopes have
been reported. In the upstream region of the CD44 gene, a typical TATA box sequence is not
observed, but an AP-1 binding site, TPA (12-O-tetradecanoylphorbol 13-acetate) response
element (TRE), and epidermal growth factor (EGF) response element (ERE) is present,
leading to activation by intracellular signals. In addition to sugar chains, CS and HS are
added to this gene product, and it is finally localized in the cell membrane. These molecules
are classified into two major forms and variants. One of the major forms has a molecular
weight of 80 to 90 kDa and is also called the standard form (CD44s) or the hematopoietic
form (CD44H) since it was first discovered in blood cells. In particular, the role of binding to
HA and its association with ankyrin in the intracellular domain in cell movement has been
pointed out. The second major form has a molecular weight of 110 to 160 kDa and is called
the epithelial form (CD44E) because it was first discovered in epithelial cells. Isotopes
other than these two are collectively called variant forms (CD44v). Five well-conserved
N-glycosylation sites exist in the N-terminal 120 amino acid region. The intracellular
region of the CD44 molecules is usually selected from exon 20 and is called the long form
(CD44-L) [195]. Exon 19 is rarely selected, and when it is selected, the molecule is called
short form CD44 (CD44s), and this isotope exists at a ratio of about 1/100 to 1/200 times
less. 7.1. Cell Surface Type and Extracellular Matrix Type of HSPGs 7.1. Cell Surface Type and Extracellular Matrix Type of HSPGs With very few exceptions, HS is covalently bound to core proteins, and together they
form HSPG, which are classified into two types: cell surface type and extracellular matrix
type. The two main cell surface HSPGs are glypicans and Syndecans. As mentioned above,
Syndecan has an extracellular region where HS chains and chondroitin sulfate chains bind,
as well as a transmembrane region and an intracellular region. Glypicans are bound to
membranes via glycosylphosphatidylinositol. Other cell surface HSPGs include CD44 and
beta glycan [190]. On the other hand, the main HSPG in the extracellular matrix is perlecan. Cell surface HSPGs may also be secreted into the matrix because of enzymatic cleavage of
core proteins. 6.3. Diversity in HS Chain Length and Modification The great diversity in HS chain length and modifications contributes to a high degree
of complexity, but important aspects remain unclear, especially in vivo [180,181]. This sug-
gests a mechanism by which heparanase expression is linked to subsequent HS sulfation
and ultrastructural changes [1]. It involves sensitivity to changes in the pericellular envi-
ronment, such as from interactions between growth factors and HS [182]. In cancer tissues,
aberrant expression of HSPGs, heparanase, and sulfatase is observed and influences disease
progression and outcome [10,131,183]. There is increasing interest in developing HS and
HS-modifying enzymes as useful targets for novel cancer therapeutics, such as inhibitors
of heparanase [10,184,185]. HSPGs are expressed on accessible cell surfaces, but it remains
unclear how abnormalities in HSPG expression affect changes in cell physiology and how
different cell types in the cancer microenvironment are affected [10,55,186]. To achieve an
understanding of the function of HSPGs in cancer progression, it is necessary to clarify
how the expression of HSPG is regulated and maintained during normal development
and how these molecules are deregulated during disease [9,10,55,187]. In addition, since
HS is abundant on the cell surface, viruses use it as a scaffold for adsorption to target
cells [29,38,188]. On the other hand, the structure of HS changes depending on the cell Biologics 2024, 4 118 type, individual, age, etc., and there is various structural diversity of HS, which determines
the infection tropism of viruses and the infection susceptibility of cells [4,29,39,189]. More
detailed analysis of the interaction between virus particles and HS may lead to a better
understanding of the various HS [37,38]. 7.2. Structure, Expression, and Functions of CD44 The intracytoplasmic region contains ezrin, radixin, moesin (ERM family) binding
sites, ankyrin binding sites, and two phosphorylation sites (Ser325 and Ser327).i CD44 is expressed in many cell types, including blood cells, fibroblasts, epithelial cells,
vascular endothelial cells, muscle cells, and neuroglial cells. It also appears and disappears
during the differentiation and proliferation processes of each cell lineage, and changes
dynamically [191,195]. In intestinal mucosa, gastric mucosa, and squamous epithelial
mucosa, CD44 expression is enhanced in areas with strong basal proliferation, but the
expression is weak or absent in superficial areas. When T cells are activated by antigen Biologics 2024, 4 119 stimulation, etc., they begin to express isotopes containing v6, which are not observed in
the resting phase and are involved in T cell migration to tissues. stimulation, etc., they begin to express isotopes containing v6, which are not observed in
the resting phase and are involved in T cell migration to tissues. HA is a long-chain glycosaminoglycan consisting of repeating disaccharides and is
involved as an extracellular matrix in cell migration during morphogenesis, wound healing,
and tumor progression. Cell surface molecules such as CD44 and RHAMM mediate cell
movement by binding to HA as a ligand [196]. The binding of CD44 to HA is influenced not
only by the amount of CD44 expressed but also by its distribution density and activation
state [197]. Covalent homodimerization of CD44 via cysteine 286 in the transmembrane
region is important for high-level binding of HA [198,199]. Furthermore, since it has
been suggested that five well-conserved N-glycosylation sites present in the N-terminal
120 amino acid region play an important role in this region, the HA-binding ability is
attenuated by these N-glycosylation and sialic acid additions. This also provides the
basis for explaining the diversity of CD44 function across different tissues and cell types. Furthermore, the presence of a motif (B(7x)B) consisting of two basic amino acids into which
seven non-acidic amino acids are inserted is important for binding to HA. In wildtype
CD44, this motif exists repeatedly, once in the cartilage attachment region and twice in the
center of the extracellular domain. Serglycin, a new ligand for CD44s, is a proteoglycan
secreted by immune hematopoietic cells and is involved in the activation of immune cells. Osteopontin is an extracellular phosphorylated protein that is secreted by activated T cells,
osteoblasts, histiocytes, etc. 7.2. Structure, Expression, and Functions of CD44 In addition to interacting with integrins, it also binds to CD44,
an isotope containing v7 to v10, and the pathological conditions caused by this interaction
are chemotaxis rather than motility. 7.5. CD44 and Angiogenesis The growth of tumors larger than one to two mm requires blood vessel induction
and maintenance, and tumor metastasis is dependent on angiogenesis. Angiogenesis is
controlled by the balance between promoting and suppressing factors secreted by tumors. Inside a tumor, microvessels supply the oxygen and nutrients necessary for tumor cell
growth and maintenance and serve as passageways for tumor cells to migrate away from
their primary tumor [212]. Factors that promote this include bFGF, VEGF, and IL-8, and
some of these growth factors can bind to the HS chain of CD44, and cytokines have an
affinity for heparin and can bind to HS. Vascular endothelial cells displaying growth factors
on the CD44 molecule promote the survival, proliferation, and migration of tumor cells
out of the blood vessels. The expression of CD44 is enhanced in vascular endothelial cells
in tumors compared to endothelial cells in normal tissues, in which enhanced expression
of CD44 in endothelial cells is induced by stimulation with bFGF or VEGF [213]. In ad-
dition, by blocking CD44 expressed on endothelial cells with an anti-CD44 monoclonal
antibody, proliferation, migration, and lumen formation of endothelial cells are inhibited. In fibroblasts and alveolar macrophages, CD44 binds to HA, takes it into the cells, and is
degraded by hyaluronidase. These degraded products of HA act on tyrosine phosphoryla-
tion in endothelial cells, promoting endothelial cell proliferation. On the other hand, when
CD44 isotopes containing v10 were introduced into human mammary gland epithelial cells
expressing CD44s, the expression of bFGF and IL-8 was enhanced. In addition, introduc-
ing CD44E into CD44-negative human lung cancer cells enhances VEGF production and
induces angiogenesis in vitro and in vivo. 7.4. Molecular Mechanism of CD44 and Tumor Invasion and Metastasis In metastatic breast cancer cell lines, CD44v3m v8 to V10 and MMP-9 combine and
localize on the cell surface, contributing to invadopodia formation and cell migration [208]. In mouse breast cancer cell lines and human melanoma cell lines, CD44 binds to active
MMP-9, localizes on the cell membrane, and contributes to tumor invasion by promoting
the degradation of ECM collagen IV. This binding between CD44 and MMP-9 is due to
the ability of CD44 to bind to HA. The migration ability of cancer cells via HA, a ligand
for CD44, is related to the cleavage of CD44 by membrane-type metalloproteases [209]. In
lung cancer cell lines, CD44 expression induces and enhances the expression of membrane
type 2-matrix metalloproteinase (MT2-MMP) is induced and enhanced, promoting the
activation of proMMP-2, contributing to the decomposition of the surrounding ECM, and
being involved in tumor invasion [210,211]. Thus, in elucidating the molecular mechanism
of cancer invasion and metastasis involving CD44, an adhesion molecule, not only its ligand
binding ability but also MMPs, which are ECM degrading enzymes, are involved [191]. 7.3. Therapeutic Application of HSPGs HA, whose functions are being further elucidated, is expected to be applied to a wide
range of fields beyond the areas in which it has been primarily used, such as musculoskeletal
diseases and ophthalmology. In addition, sugar chains such as heparin and CS, which
belong to the same proteoglycan region as HA, have also been widely used as medicines. Heparin has a unique function of expressing anticoagulant activity through antithrombin
III in the blood and has long held an unwavering position as a drug that maintains blood
circulation. Low-molecular-weight heparin, an improved version of heparin with reduced
bleeding effects, is an antithrombotic drug. Furthermore, as the oligosaccharide structure of
the active moiety is elucidated, progress is being made in the development of antithrombotic
agents using this active oligosaccharide chain. CS has been used as a medical drug for
more than half a century. Its efficacy has been recognized for joint pain, neuralgia, frozen
shoulders, etc. It is also used to prevent corneal dryness and is included in medical and
non-prescription eye drops. In addition, polysaccharides such as chitin and chitosan, which
are proteoglycan analogs, are being developed for use in medical devices [200,201]. In
addition, polysaccharides such as dextran produced by microorganisms have long been
used in infusions as plasma expanders, and polysaccharides derived from plants are used
in anticancer drugs and healthy foods [202–204]. g
y
There are still not many examples of applying knowledge from glycoscience to pharma-
ceuticals. However, carbohydrates play unique roles in developmental and aging processes
and diseases such as infection, immunity, inflammation, and cancer, and, as the elucidation
of their mechanisms is rapidly progressing, there is no doubt that new opportunities for
clinical application will open [205]. For example, influenza treatments, neuraminidase
inhibitors, and other drugs are developed based on the knowledge and technology of gly-
coscience. The various cell surface sugar chain molecules involved in viral and microbial
infections and the mechanisms involved in these infections have been elucidated, which
is expected to further accelerate the development of infection prevention and treatment
agents. Elucidation of the changes in sugar chains on the cell surface associated with can-
cerization and their mechanisms have become increasingly useful as an indicator for cancer
diagnosis and treatment [206,207]. Furthermore, progress has been made in elucidating the
involvement of HS and HA in cancer metastasis, implantation, and proliferation. Biologics 2024, 4 120 7.6. CD44 and Cell Motility via Cell Skeletal Protein Ezrin/radixin/moesin (ERM) protein family are localized directly beneath protoplasmic
membranes such as microvilli, ruffling membranes, cleavage furrows, and adherens junctions,
and are involved in their formation and function. These regions are where the actin filament
is tightly bound to the membrane, and the ERM family is directly involved in the molecular
mechanisms that control the interaction of the actin filament with the plasma membrane. One
of the major membrane proteins that this ERM protein binds to is CD44. In addition, CD44 and
ERM proteins show strong affinity and form a complex in the presence of phosphoinositides. When the CD44/ERM complex was immunoprecipitated, this complex contained a Rho-GDP
dissociation inhibitor (GDI). Furthermore, C3 toxin, a specific inhibitor of Rho, inhibits the
binding of recombinant ERM proteins to CD44, reducing motor performance. Therefore, Rho
and phosphorylation play a critical role in the formation of the CD44/ERM complex and are
also involved in the regulation of locomotor performance. Biologics 2024, 4 121 7.7. Soluble CD44 CD44 can be rapidly downregulated in response to signals inside and outside the cell,
but this is mainly due to the dissociation of soluble CD44 from the extracellular region. The
CD44 released in this way can be found in the serum, and soluble CD44 is elevated in the
serum of patients with malignant tumors. Serum CD44 and CD44v6 levels in patients with
malignant lymphoma are both increased compared to healthy individuals and, as these
values decreased in response to treatment, they can be used to monitor the effectiveness
of treatment. Furthermore, soluble CD44s-Ig fusion protein suppresses the metastasis of
lymphoid malignant tumors that highly express CD44s and can be applied to therapy. 8. Conclusions HSPGs have been involved in developmental and homeostatic processes, as well as
in many pathological conditions. Therefore, it can be applied to a wide range of fields for
therapeutic strategies, including Alzheimer’s disease, heart failure, cancer, organ transplants,
diabetes, chronic inflammation, aging, and autoimmune diseases. HSPGs, a major component
of the glycocalyx, enable the binding of various plasma-derived molecules due to their diver-
sity, epimerization of sugar chains, long chains, and sulfation. The glycocalyx has a variety
of functions, including maintaining blood flow, protecting the endothelium, antithrombotic
effects, binding and signal transmission of hormones and growth factors, and serving as a nu-
tritional source consisting of sugar chains. The various structural diversity of HS determines
the infection tropism of viruses and the infection susceptibility of cells. Author Contributions: Writing—review and editing, Y.M.; supervision, R.Y.; funding acquisition,
Y.M. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Funding: This research received no external funding. Data Availability Statement: Not appliable. Data Availability Statement: Not appliable. Data Availability Statement: Not appliable. Conflicts of Interest: The authors declare that this research was conducted in the absence of any
commercial or financial relationships that could be construed as potential conflicts of interest. Conflicts of Interest: The authors declare that this research was conducted in the absence of any
commercial or financial relationships that could be construed as potential conflicts of interest. References 1. Ravikumar, M.; Smith, R.A.A.; Nurcombe, V.; Cool, S.M. Heparan Sulfate Proteoglycans: Key Mediators of Stem Cell Function. Front. Cell Dev. Biol. 2020, 8, 581213. [CrossRef] 1. Ravikumar, M.; Smith, R.A.A.; Nurcombe, V.; Cool, S.M. Heparan Sulfate Proteoglycans: Key Mediators of Stem Cell Function. Front. Cell Dev. Biol. 2020, 8, 581213. [CrossRef] 2. Moon, S.; Zhao, Y.T. Spatial, temporal and cell-type-specific expression profiles of genes encoding heparan sulfate biosynthesis
enzymes and proteoglycan core proteins. Glycobiology 2021, 31, 1308–1318. [CrossRef] 2. Moon, S.; Zhao, Y.T. Spatial, temporal and cell-type-specific expression profiles of genes encoding heparan sulfate biosynthesis
enzymes and proteoglycan core proteins. Glycobiology 2021, 31, 1308–1318. [CrossRef] 3. Condomitti, G.; de Wit, J. Heparan Sulfate Proteoglycans as Emerging Players in Synaptic Specificity. Front. Mol. Neurosci. 2018,
11, 14. [CrossRef] 3. Condomitti, G.; de Wit, J. Heparan Sulfate Proteoglycans as Emerging Players in Synaptic Specificity. Front. Mol. Neurosci. 2018,
11, 14. [CrossRef] 4. De Pasquale, V.; Quiccione, M.S.; Tafuri, S.; Avallone, L.; Pavone, L.M. Heparan Sulfate Proteoglyca
Treatment: A Special Focus on SARS-CoV-2. Int. J. Mol. Sci. 2021, 22, 6574. [CrossRef] p
5. Sparn, C.; Dimou, E.; Meyer, A.; Saleppico, R.; Wegehingel, S.; Gerstner, M.; Klaus, S.; Ewers, H.; Nickel, W. Glypican-1 drives
unconventional secretion of fibroblast growth factor 2. eLife 2022, 11, e75545. [CrossRef] 5. Sparn, C.; Dimou, E.; Meyer, A.; Saleppico, R.; Wegehingel, S.; Gerstner, M.; Klaus, S.; Ewers, H.; Nickel, W. Glypican-1 drives
unconventional secretion of fibroblast growth factor 2. eLife 2022, 11, e75545. [CrossRef] i
g
f
6. Gopal, S.; Arokiasamy, S.; Pataki, C.; Whiteford, J.R.; Couchman, J.R. Syndecan receptors: Pericellular regulators in development
and inflammatory disease. Open Biol. 2021, 11, 200377. [CrossRef] 6. Gopal, S.; Arokiasamy, S.; Pataki, C.; Whiteford, J.R.; Couchman, J.R
and inflammatory disease. Open Biol. 2021, 11, 200377. [CrossRef] 6. Gopal, S.; Arokiasamy, S.; Pataki, C.; Whiteford, J.R.; Couchman, J.R. Syndecan receptors: Pericellular regulators in development
and inflammatory disease. Open Biol. 2021, 11, 200377. [CrossRef] l
y
p
7. Gondelaud, F.; Ricard-Blum, S. Structures and interactions of syndecans. FEBS J. 2019, 286, 2994–3007. p
rd-Blum, S. Structures and interactions of syndecans. FEBS J. 2019, 286, 2994–3007. [CrossRef] l
y
p
7. Gondelaud, F.; Ricard-Blum, S. Structures and interactions of synd 8. Jang, B.; Kim, A.; Hwang, J.; Song, H.K.; Kim, Y.; Oh, E.S. Emerging Role of Syndecans in Extracellular Matrix Remodeling in
Cancer. J. Histochem. Cytochem. 2020, 68, 863–870. [CrossRef] y
9. References Onyeisi, J.O.S.; Ferreira, B.Z.F.; Nader, H.B.; Lopes, C.C. Heparan sulfate proteoglycans as targets for cancer therapy: A review. Cancer Biol. Ther. 2020, 21, 1087–1094. [CrossRef] 10. Nagarajan, A.; Malvi, P.; Wajapeyee, N. Heparan Sulfate and Heparan Sulfate Proteoglycans in Cancer Initiation and Progression. Front. Endocrinol. 2018, 9, 483. [CrossRef] 11. Crespo, A.; García-Suárez, O.; Fernández-Vega, I.; Solis-Hernandez, M.P.; García, B.; Castañón, S.; Quirós, L.M. Heparan sulfate
proteoglycans undergo differential expression alterations in left sided colorectal cancer, depending on their metastatic character. BMC Cancer 2018, 18, 687. [CrossRef] ,
,
[
]
12. Thota, L.N.R.; Chignalia, A.Z. The role of the glypican and syndecan families of heparan sulfate proteoglycans in cardiovascular
function and disease. Am. J. Physiol. Cell Physiol. 2022, 323, C1052–C1060. [CrossRef] Biologics 2024, 4 122 13. Guss, E.J.; Akbergenova, Y.; Cunningham, K.L.; Littleton, J.T. Loss of the extracellular matrix protein Perlecan disrupts axonal
and synaptic stability during Drosophila development. eLife 2023, 12, RP88273. [CrossRef] 14. Hayes, A.J.; Farrugia, B.L.; Biose, I.J.; Bix, G.J.; Melrose, J. Perlecan, A Multi-Functional, Cell-Instructive, Matrix-Stabilizing
Proteoglycan with Roles in Tissue Development Has Relevance to Connective Tissue Repair and Regeneration. Front. Cell Dev. Biol. 2022, 10, 856261. [CrossRef] 15. Yamashita, Y.; Nakada, S.; Yoshihara, T.; Nara, T.; Furuya, N.; Miida, T.; Hattori, N.; Arikawa-Hirasawa, E. Perlecan, a heparan
sulfate proteoglycan, regulates systemic metabolism with dynamic changes in adipose tissue and skeletal muscle. Sci. Rep. 2018,
8, 7766. [CrossRef] 16. Bai, G.; Zhang, M. Clustering acetylcholine receptors in neuromuscular junction by phase-separated Rapsn condensates. Neuron
2021, 109, 1907–1909. [CrossRef] 17. Guarino, S.R.; Canciani, A.; Forneris, F. Dissecting the Extracellular Complexity of Neuromuscular Junction Organizers. Front. Mol. Biosci. 2020, 6, 156. [CrossRef] 18. Dong, C.; Choi, Y.K.; Lee, J.; Zhang, X.F.; Honerkamp-Smith, A.; Widmalm, G.; Lowe-Krentz, L.J.; Im, W. Structure, Dynamics,
and Interactions of GPI-Anchored Human Glypican-1 with Heparan Sulfates in a Membrane. Glycobiology 2021, 31, 593–602. [CrossRef] [
]
19. Yu, P.; Pearson, C.S.; Geller, H.M. Flexible Roles for Proteoglycan Sulfation and Receptor Signaling. Trends Neurosci. 2018, 41, 47–61. [CrossRef] [
]
20. Lin, X. Functions of heparan sulfate proteoglycans in cell signaling during development. Development 2004, 131, 6009–6021. [CrossRef] 21. Park, P.W. Isolation and functional analysis of syndecans. Methods Cell Biol. 2018, 143, 317–333. [Cr 22. Annaval, T.; Wild, R.; Crétinon, Y.; Sadir, R.; Vivès, R.R.; Lortat-Jacob, H. References Heparan Sulfate Proteoglycans Biosynthesis and Post
Synthesis Mechanisms Combine Few Enzymes and Few Core Proteins to Generate Extensive Structural and Functional Diversity. Molecules 2020, 25, 4215. [CrossRef] s, C.A.; Vivès, R.R.; Magalhães, A. Heparan Sulfate Biosynthesis and Sulfation Profiles as Modulators of Can
rogression. Front. Oncol. 2021, 11, 778752. [CrossRef] 23. Marques, C.; Reis, C.A.; Vivès, R.R.; Magalhães, A. Heparan Sulfate Biosynthesis and Sulfation Prof
Signalling and Progression. Front. Oncol. 2021, 11, 778752. [CrossRef] 24. Leisico, F.; Omeiri, J.; Le Narvor, C.; Beaudouin, J.; Hons, M.; Fenel, D.; Schoehn, G.; Couté, Y.; Bonnaffé, D.; Sadir, R.; et al. Structure of the human heparan sulfate polymerase complex EXT1-EXT2. Nat. Commun. 2022, 13, 7110. [CrossRef] 25. Niwa, A.; Taniguchi, T.; Tomita, H.; Okada, H.; Kinoshita, T.; Mizutani, C.; Matsuo, M.; Imaizumi, Y.; Kuroda, T.; Ichihashi, K.;
et al. Conditional ablation of heparan sulfate expression in stromal fibroblasts promotes tumor growth in vivo. PLoS ONE 2023,
18, e0281820. [CrossRef] 26. Hwang, J.; Jang, B.; Kim, A.; Lee, Y.; Lee, J.; Kim, C.; Kim, J.; Moon, K.M.; Kim, K.; Wagle, R.; et al. Syndecan Transmembrane
Domain Specifically Regulates Downstream Signaling Events of the Transmembrane Receptor Cytoplasmic Domain. Int. J. Mol. Sci. 2021, 22, 7918. [CrossRef] 27. Hara, T.; Sato, A.; Yamamoto, C.; Kaji, T. Syndecan-1 downregulates syndecan-4 expression by suppressing the ERK1/2 and p38
MAPK signaling pathways in cultured vascular endothelial cells. Biochem. Biophys. Rep. 2021, 26, 101001. [CrossRef] 28. Ren, Z.; Spaargaren, M.; Pals, S.T. Syndecan-1 and stromal heparan sulfate proteoglycans: Key moderat
and myeloma pathogenesis. Blood 2021, 137, 1713–1718. [CrossRef] 29. Zhang, D.; Zhang, J.T.; Pan, Y.; Liu, X.F.; Xu, J.W.; Cui, W.J.; Qiao, X.R.; Dong, L. Syndecan-1 Shedding by Matrix Metalloproteinase-
9 Signaling Regulates Alveolar Epithelial Tight Junction in Lipopolysaccharide-Induced Early Acute Lung Injury. J. Inflamm. Res. 2021, 14, 5801–5816. [CrossRef] 30. Jang, B.; Yun, J.H.; Choi, S.; Park, J.; Shin, D.H.; Lee, S.T.; Lee, W.; Oh, E.S. Tyrosine 51 residue of
domain is involved in the interaction with and activation of pro-matrix metalloproteinase-7. Sci. R 30. Jang, B.; Yun, J.H.; Choi, S.; Park, J.; Shin, D.H.; Lee, S.T.; Lee, W.; Oh, E.S. Tyrosine 51 residue of the syndecan-2 extracellular
domain is involved in the interaction with and activation of pro-matrix metalloproteinase-7. Sci. Rep. 2019, 9, 10625. [CrossRef]
31. De Pasquale, V.; Sarogni, P.; Pistorio, V.; Cerulo, G.; Paladino, S.; Pavone, L.M. References Targeting Heparan Sulfate Proteoglycans as a
N
l Th
i S
f
M
l
h
id
M l Th
M h d Cli
D
2018 10 8 16 [C
R f] domain is involved in the interaction with and activation of pro matrix metalloproteinase 7. Sci. Rep. 2019, 9, 10625. [CrossRef]
31. De Pasquale, V.; Sarogni, P.; Pistorio, V.; Cerulo, G.; Paladino, S.; Pavone, L.M. Targeting Heparan Sulfate Proteoglycans as a
Novel Therapeutic Strategy for Mucopolysaccharidoses. Mol. Ther. Methods Clin. Dev. 2018, 10, 8–16. [CrossRef] Novel Therapeutic Strategy for Mucopolysaccharidoses. Mol. Ther. Methods Clin. Dev. 2018, 10, 8–16. [CrossRef]
32. Pretorius, D.; Richter, R.P.; Anand, T.; Cardenas, J.C.; Richter, J.R. Alterations in heparan sulfate proteoglycan synthesis and
sulfation and the impact on vascular endothelial function. Matrix Biol. Plus 2022, 16, 100121. [CrossRef] 32. Pretorius, D.; Richter, R.P.; Anand, T.; Cardenas, J.C.; Richter, J.R. Alterations in heparan sulfate proteoglycan synthesis and
sulfation and the impact on vascular endothelial function. Matrix Biol. Plus 2022, 16, 100121. [CrossRef] 33. Yang, Z.; Chen, S.; Ying, H.; Yao, W. Targeting syndecan-1: New opportunities in cancer therapy. Am. J. Physiol. Cell Physiol. 2022,
323, C29–C45. [CrossRef] 34. Garcia, J.; Patel, N.; Basehore, S.; Clyne, A.M. Fibroblast Growth Factor-2 Binding to Heparan Sulfate Proteoglycans Varies with
Shear Stress in Flow-Adapted Cells. Ann. Biomed. Eng. 2019, 47, 1078–1093. [CrossRef] 35. Koledova, Z.; Sumbal, J.; Rabata, A.; de La Bourdonnaye, G.; Chaloupkova, R.; Hrdlickova, B.; Damborsky, J.; Stepankova, V. Fibroblast Growth Factor 2 Protein Stability Provides Decreased Dependence on Heparin for Induction of FGFR Signaling and
Alters ERK Signaling Dynamics. Front. Cell Dev. Biol. 2019, 7, 331. [CrossRef] 36. Hayashida, K.; Aquino, R.S.; Park, P.W. Coreceptor functions of cell surface heparan sulfate proteoglycans. Am. J. Physiol. Cell
Physiol. 2022, 322, C896–C912. [CrossRef] 37. Chopra, P.; Joshi, A.; Wu, J.; Lu, W.; Yadavalli, T.; Wolfert, M.A.; Shukla, D.; Zaia, J.; Boons, G.J. The
sulfate modulates protein binding and lyase degradation. Proc. Natl. Acad. Sci. USA 2021, 118, e201293 37. Chopra, P.; Joshi, A.; Wu, J.; Lu, W.; Yadavalli, T.; Wolfert, M.A.; Shukla, D.; Zaia, J.; Boons, G.J. The 3-O-sulfation of heparan
sulfate modulates protein binding and lyase degradation. Proc. Natl. Acad. Sci. USA 2021, 118, e2012935118. [CrossRef]
f
f 38. Ferreira, A.; Timmerman, E.; Staes, A.; Vuylsteke, M.; De Muynck, L.; Gevaert, K. Protein interactors
sulfates in human MCI and age-matched control cerebrospinal fluid. Sci. Data 2023, 10, 121. References Update on the Role of the Endothelial Glycocalyx in Angiogenesis and Vascular Inflammation. Front. Cell Dev. Biol. 2021, 9, 734276. [CrossRef] 51. Uchimido, R.; Schmidt, E.P.; Shapiro, N.I. The glycocalyx: A novel diagnostic and therapeutic target in sepsis. Crit. Care 2019,
23, 16. [CrossRef] 52. Aldecoa, C.; Llau, J.V.; Nuvials, X.; Artigas, A. Role of albumin in the preservation of endothelial gly
microcirculation: A review. Ann. Intensive Care 2020, 10, 85. [CrossRef] 53. Möckl, L. The Emerging Role of the Mammalian Glycocalyx in Functional Membrane Organization and Immune System
Regulation. Front. Cell Dev. Biol. 2020, 8, 253. [CrossRef] g
54. Jin, J.; Fang, F.; Gao, W.; Chen, H.; Wen, J.; Wen, X.; Chen, J. The Structure and Function of the Glycocalyx and Its Connection
With Blood-Brain Barrier. Front. Cell. Neurosci. 2021, 15, 739699. [CrossRef] 54. Jin, J.; Fang, F.; Gao, W.; Chen, H.; Wen, J.; Wen, X.; Chen, J. The Structure and Function of the Glyco
With Blood-Brain Barrier. Front. Cell. Neurosci. 2021, 15, 739699. [CrossRef] 55. Moore, K.H.; Murphy, H.A.; George, E.M. The glycocalyx: A central regulator of vascular function. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2021, 320, R508–R518. [CrossRef] 56. Milusev, A.; Rieben, R.; Sorvillo, N. The Endothelial Glycocalyx: A Possible Therapeutic Target in Cardiovascular Disorders. Front. Cardiovasc. Med. 2022, 9, 897087. [CrossRef] 57. Schött, U.; Solomon, C.; Fries, D.; Bentzer, P. The endothelial glycocalyx and its disruption, protection and regeneration: A
narrative review. Scand. J. Trauma. Resusc. Emerg. Med. 2016, 24, 48. [CrossRef]ll 58. Kundra, P.; Goswami, S. Endothelial glycocalyx: Role in body fluid homeostasis and fluid managemen
63, 6–14. [CrossRef] 59. Haymet, A.B.; Bartnikowski, N.; Wood, E.S.; Vallely, M.P.; McBride, A.; Yacoub, S.; Biering, S.B.; Harris, E.; Suen, J.Y.; Fraser, J.F. Studying the Endothelial Glycocalyx in vitro: What Is Missing? Front. Cardiovasc. Med. 2021, 8, 647086. [CrossRef] y
g
y
y
g
60. Foote, C.A.; Soares, R.N.; Ramirez-Perez, F.I.; Ghiarone, T.; Aroor, A.; Manrique-Acevedo, C.; Padilla, J.; Martinez-Lemus, L. Endothelial Glycocalyx. Compr. Physiol. 2022, 12, 3781–3811. [CrossRef] 61. Jiang, X.Z.; Goligorsky, M.S. Biomechanical properties of endothelial glycocalyx: An imperfect pendulum. Matrix Biol. Plus 2021,
12, 100087. [CrossRef] 62. Bol, M.E.; Huckriede, J.B.; van de Pas, K.G.H.; Delhaas, T.; Lorusso, R.; Nicolaes, G.A.F.; Sels, J.E.M.; van de Poll, M.C.G. Multimodal measurement of glycocalyx degradation during coronary artery bypass grafting. Front. Med. 2022, 9, 1045728. [CrossRef] [
]
63. Patterson, E.K.; Cepinskas, G.; Fraser, D.D. References [CrossRef Biologics 2024, 4 123 39. Cagno, V.; Tseligka, E.D.; Jones, S.T.; Tapparel, C. Heparan Sulfate Proteoglycans and Viral Attachment: True Receptors or
Adaptation Bias? Viruses 2019, 11, 596. [CrossRef] 40. Gómez Toledo, A.; Sorrentino, J.T.; Sandoval, D.R.; Malmström, J.; Lewis, N.E.; Esko, J.D. A Systems View of the Heparan Sulfate
Interactome. J. Histochem. Cytochem. 2021, 69, 105–119. [CrossRef] 41. Pataki, Z.; Rebolledo Viveros, A.; Heldwein, E.E. Herpes Simplex Virus 1 Entry Glycoproteins Form Complexes before and during
Membrane Fusion. mBio 2022, 13, e0203922. [CrossRef] 42. Weed, D.J.; Nicola, A.V. Herpes simplex virus Membrane Fusion. Adv. Anat. Embryol. Cell Biol. 201 la, A.V. Herpes simplex virus Membrane Fusion. Adv. Anat. Embryol. Cell Biol. 2017, 223, 29–47. [CrossRef] 43. Elste, J.; Chan, A.; Patil, C.; Tripathi, V.; Shadrack, D.M.; Jaishankar, D.; Hawkey, A.; Mungerson, M.S.; Shukla, D.; Tiwari, V. Archaic connectivity between the sulfated heparan sulfate and the herpesviruses—An evolutionary potential for cross-species
interactions. Comput. Struct. Biotechnol. J. 2023, 21, 1030–1040. [CrossRef] p
J
[
]
44. Zhao, J.; Zhu, Y.; Song, X.; Xiao, Y.; Su, G.; Liu, X.; Wang, Z.; Xu, Y.; Liu, J.; Eliezer, D.; et al. 3-O-Sulfation of Heparan Sulfate
Enhances Tau Interaction and Cellular Uptake. Angew. Chem. Int. Ed. Engl. 2020, 59, 1818–1827. [CrossRef] p
g
g
45. Kearns, F.L.; Sandoval, D.R.; Casalino, L.; Clausen, T.M.; Rosenfeld, M.A.; Spliid, C.B.; Amaro, R.E.; Esko, J.D. Spike-heparan
sulfate interactions in SARS-CoV-2 infection. Curr. Opin. Struct. Biol. 2022, 76, 102439. [CrossRef] 46. Zhang, Q.; Chen, C.Z.; Swaroop, M.; Xu, M.; Wang, L.; Lee, J.; Wang, A.Q.; Pradhan, M.; Hagen, N.; Chen, L.; et al. Heparan
sulfate assists SARS-CoV-2 in cell entry and can be targeted by approved drugs in vitro. Cell Discov. 2020, 6, 80. [CrossRef] 47. Clausen, T.M.; Sandoval, D.R.; Spliid, C.B.; Pihl, J.; Perrett, H.R.; Painter, C.D.; Narayanan, A.; Majowicz, S.A.; Kwong, E.M.;
McVicar, R.N.; et al. SARS-CoV-2 Infection Depends on Cellular Heparan Sulfate and ACE2. Cell 2020, 183, 1043–1057.e15. [CrossRef] [
]
48. Zhang, Q.; Tang, W.; Stancanelli, E.; Jung, E.; Syed, Z.; Pagadala, V.; Saidi, L.; Chen, C.Z.; Gao, P.; Xu, M.; et al. Host heparan
sulfate promotes ACE2 super-cluster assembly and enhances SARS-CoV-2-associated syncytium formation. Nat. Commun. 2023,
14, 5777. [CrossRef] 49. Villalba, N.; Baby, S.; Yuan, S.Y. The Endothelial Glycocalyx as a Double-Edged Sword in Microvascular Homeostasis and
Pathogenesis. Front. Cell Dev. Biol. 2021, 9, 711003. [CrossRef]l 50. Hu, Z.; Cano, I.; D’Amore, P.A. References [CrossRef] , T.; Thachil, J.; Asakura, H.; Levy, J.H.; Iba, T. Thrombomodulin in disseminated intravascular coagulation
nditions-a multi-faceted anticoagulant protein with therapeutic potential. Crit. Care 2019, 23, 280. [CrossRe 75. Watanabe-Kusunoki, K.; Nakazawa, D.; Ishizu, A.; Atsumi, T. Thrombomodulin as a Physiological Modulator of Intravascular
Injury. Front. Immunol. 2020, 11, 575890. [CrossRef] j
y
[
]
76. Gopal, S. Syndecans in Inflammation at a Glance. Front. Immunol. 2020, 11, 227. [CrossRef] 77. Ahn, S.J.; Le Master, E.; Granados, S.T.; Levitan, I. Impairment of endothelial glycocalyx in atherosclerosis and obesity. Curr. Top. Membr. 2023, 91, 1–19. [CrossRef] 78. Wang, G.; Tiemeier, G.L.; van den Berg, B.M.; Rabelink, T.J. Endothelial Glycocalyx Hyaluronan: Regulation and Role in
Prevention of Diabetic Complications. Am. J. Pathol. 2020, 190, 781–790. [CrossRef] 79. Dogné, S.; Flamion, B. Endothelial Glycocalyx Impairment in Disease: Focus on Hyaluronan Shedding. Am. J. Pathol. 2020,
190, 768–780. [CrossRef]l va, K.; Morozkina, S.; Olekhnovich, R.; Uspenskaya, M. Hyaluronic Acid: The Influence of Molecular Weight 80. Snetkov, P.; Zakharova, K.; Morozkina, S.; Olekhnovich, R.; Uspenskaya, M. Hyaluronic Acid: The Influ
on Structural, Physical, Physico-Chemical, and Degradable Properties of Biopolymer. Polymers 2020, 1 on Structural, Physical, Physico-Chemical, and Degradable Properties of Biopolymer. Polymers 2020, 12, 1800. [CrossRef]
81. Shi, D.; Sheng, A.; Chi, L. Glycosaminoglycan-Protein Interactions and Their Roles in Human Disease. Front. Mol. Biosci. 2021,
8, 639666. [CrossRef] y
y
g
p
p
y
y
81. Shi, D.; Sheng, A.; Chi, L. Glycosaminoglycan-Protein Interactions and Their Roles in Human Disease. Front. Mol. Biosci. 2021,
8, 639666. [CrossRef] ,
[
]
82. Soares da Costa, D.; Reis, R.L.; Pashkuleva, I. Sulfation of Glycosaminoglycans and Its Implications in Human Health and
Disorders Annu Rev Biomed Eng 2017 19 1 26 [CrossRef] [
]
82. Soares da Costa, D.; Reis, R.L.; Pashkuleva, I. Sulfation of Glycosaminoglycans and Its Implications in Human Health and
Disorders. Annu. Rev. Biomed. Eng. 2017, 19, 1–26. [CrossRef]
k
l
l
l
f
l
l
b l
[
f] 82. Soares da Costa, D.; Reis, R.L.; Pashkuleva, I. Sulfation of Glycosaminoglycans and Its Implication
Disorders. Annu. Rev. Biomed. Eng. 2017, 19, 1–26. [CrossRef] g
83. Varki, A. Biological roles of glycans. Glycobiology 2017, 27, 3–49. [CrossRef] g
g y
y
gy
[
]
84. Morla, S. Glycosaminoglycans and Glycosaminoglycan Mimetics in Cancer and Inflammation. Int. J. Mol. Sci. 2019, 20, 1963. [CrossRef] 84. Morla, S. Glycosaminoglycans and Glycosaminoglycan Mimetics in Cancer and Inflammation. References Endothelial Glycocalyx Degradation in Critical Illness and Injury. Front. Med. 2022,
9, 898592. [CrossRef] 64. Damén, T.; Kolsrud, O.; Dellgren, G.; Hesse, C.; Ricksten, S.E.; Nygren, A. Atrial natriuretic peptide does not degrade the
endothelial glycocalyx: A secondary analysis of a randomized porcine model. Acta. Anaesthesiol. Scand. 2021, 65, 1305–1312. [CrossRef] 65. Hahn, R.G.; Patel, V.; Dull, R.O. Human glycocalyx shedding: Systematic review and critical appraisal. Acta. Anaesthesiol. Scand. 2021, 65, 590–606. [CrossRef] Biologics 2024, 4 124 66. Passov, A.; Schramko, A.; Salminen, U.S.; Aittomäki, J.; Andersson, S.; Pesonen, E. Endothelial glycocalyx during early reperfusion
in patients undergoing cardiac surgery. PLoS ONE 2021, 16, e0251747. [CrossRef] 67. Zhang, J.; Luo, W.; Miao, C.; Zhong, J. Hypercatabolism and Anti-catabolic Therapies in the Persistent Inflammation, Immuno-
suppression, and Catabolism Syndrome. Front. Nutr. 2022, 9, 941097. [CrossRef] 68. Zha, D.; Fu, M.; Qian, Y. Vascular Endothelial Glycocalyx Damage and Potential Targeted Therapy in COVID-19. Cells 2022,
11, 1972. [CrossRef] 69. Wang, G.; Kostidis, S.; Tiemeier, G.L.; Sol, W.M.P.J.; de Vries, M.R.; Giera, M.; Carmeliet, P.; van den Berg, B.M.; Ton, J.; Rabelink,
T.J. Shear Stress Regulation of Endothelial Glycocalyx Structure Is Determined by Glucobiosynthesis. Arterioscler. Thromb. Vasc
Biol. 2020, 40, 350–364. [CrossRef] 70. Oshima, K.; King, S.I.; McMurtry, S.A.; Schmidt, E.P. Endothelial Heparan Sulfate Proteoglycans in Sepsis: The Role of the
Glycocalyx. Semin. Thromb. Hemost. 2021, 47, 274–282. [CrossRef] y
y
71. Wight, T.N. A role for proteoglycans in vascular disease. Matrix Biol. 2018, 71–72, 396–420. [CrossRef] g
p
g y
72. Zhou, X.; Chen, X.; Du, H.; Ye, Y.; Miu, Y.; Su, T.; Guo, X.; Wang, S.; Qiu, Y.; Wang, J.; et al. Antithrombin III activity is associated
with prognosis, infection, and inflammation in patients with hepatitis B virus-related acute-on-chronic liver failure. Eur. J. Gastroenterol. Hepatol. 2023, 35, 914–920. [CrossRef] p
[
]
73. Arnold, K.; Xu, Y.; Liao, Y.E.; Cooley, B.C.; Pawlinski, R.; Liu, J. Synthetic anticoagulant heparan sulfate attenuates liver ischemia
reperfusion injury. Sci. Rep. 2020, 10, 17187. [CrossRef] p
j
y
p
74. Ito, T.; Thachil, J.; Asakura, H.; Levy, J.H.; Iba, T. Thrombomodulin in disseminated intravascular coagulation and other critical
conditions-a multi-faceted anticoagulant protein with therapeutic potential. Crit. Care 2019, 23, 280. [CrossRef] sakura, H.; Levy, J.H.; Iba, T. Thrombomodulin in disseminated intravascular coagulation and other critical
aceted anticoagulant protein with therapeutic potential. Crit. Care 2019, 23, 280. References Villar-Fincheira, P.; Sanhueza-Olivares, F.; Norambuena-Soto, I.; Cancino-Arenas, N.; Hernandez-Vargas, 108. Villar-Fincheira, P.; Sanhueza-Olivares, F.; Norambuena-Soto, I.; Cancino-Arenas, N.; Hernandez-Vargas, F.; Troncoso, R.; Gabrielli,
L.; Chiong, M. Role of Interleukin-6 in Vascular Health and Disease. Front. Mol. Biosci. 2021, 8, 641734. [CrossRef] g
L.; Chiong, M. Role of Interleukin-6 in Vascular Health and Disease. Front. Mol. Biosci. 2021, 8, 641734. 109. Kang, S.; Kishimoto, T. Interplay between interleukin-6 signaling and the vascular endothelium in cy
Med. 2021, 53, 1116–1123. [CrossRef] 110. Fahey, E.; Doyle, S.L. IL-1 Family Cytokine Regulation of Vascular Permeability and Angiogenesis. Front. Immunol. 2019, 10, 1426. [CrossRef] 111. Wilson, S.E. Interleukin-1 and Transforming Growth Factor Beta: Commonly Opposing, but Sometimes Supporting, Master
Regulators of the Corneal Wound Healing Response to Injury. Investig. Ophthalmol. Vis. Sci. 2021, 62, 8. [CrossRef] 112. Tanino, Y.; Wang, X.; Nikaido, T.; Misa, K.; Sato, Y.; Togawa, R.; Kawamata, T.; Kikuchi, M.; Frevert, C.W.; Tanino, M.; et al. Syndecan-4 Inhibits the Development of Pulmonary Fibrosis by Attenuating TGF-β Signaling. Int. J. Mol. Sci. 2019, 20, 4989. [CrossRef] 113. Farooq, M.; Khan, A.W.; Kim, M.S.; Choi, S. The Role of Fibroblast Growth Factor (FGF) Signaling in Tissue Repair and
Regeneration. Cells 2021, 10, 3242. [CrossRef] g
114. Chen, L.; Fu, L.; Sun, J.; Huang, Z.; Fang, M.; Zinkle, A.; Liu, X.; Lu, J.; Pan, Z.; Wang, Y.; et al. Structural basis for FGF hormone
signalling. Nature 2023, 618, 862–870. [CrossRef] 115. Onyeisi, J.O.S.; Lopes, C.C.; Götte, M. Syndecan-4 as a Pathogenesis Factor and Therapeutic Target in Cancer. Biomolecules 2021,
11, 503. [CrossRef] 116. Eustace, A.D.; McNaughton, E.F.; King, S.; Kehoe, O.; Kungl, A.; Mattey, D.; Nobbs, A.H.; Williams, N.; Middleton, J. Soluble
syndecan-3 binds chemokines, reduces leukocyte migration in vitro and ameliorates disease severity in models of rheumatoid
arthritis. Arthritis Res. Ther. 2019, 21, 172. [CrossRef] 117. Gerlza, T.; Nagele, M.; Mihalic, Z.; Trojacher, C.; Kungl, A. Glycosaminoglycans located on neutrophils and monocytes impact on
CXCL8- and CCL2-induced cell migration. Cytokine 2021, 142, 155503. [CrossRef] 118. Crijns, H.; Vanheule, V.; Proost, P. Targeting Chemokine-Glycosaminoglycan Interactions to Inhibit Inflammation. Front. Immunol. 2020, 11, 483. [CrossRef] thnam, K. Structural basis of chemokine interactions with heparan sulfate, chondroitin sulfate, and dermatan
. 2019, 294, 15650–15661. [CrossRef] 119. Sepuru, K.M.; Rajarathnam, K. Structural basis of chemokine interactions with heparan sulfate, chon
sulfate. J. Biol. Chem. 2019, 294, 15650–15661. [CrossRef] 120. References Int. J
[CrossRef] 85. Lepedda, A.J.; Nieddu, G.; Formato, M.; Baker, M.B.; Fernández-Pérez, J.; Moroni, L. Glycosaminoglycans: From Vascular
Physiology to Tissue Engineering Applications. Front. Chem. 2021, 9, 680836. [CrossRef] 86. Zapp, C.; Mundinger, P.; Boehm, H. Natural Presentation of Glycosaminoglycans in Synthetic Matrices for 3D Angiogenesis
Models. Front. Cell Dev. Biol. 2021, 9, 729670. [CrossRef] 87. Wang, Q.; Chi, L. The Alterations and Roles of Glycosaminoglycans in Human Diseases. Polymers 2 87. Wang, Q.; Chi, L. The Alterations and Roles of Glycosaminoglycans in Human Diseases. Polymers 2022, 14, 5014. [CrossRef]
88. Li, Z.; Wu, N.; Wang, J.; Zhang, Q. Roles of Endovascular Calyx Related Enzymes in Endothelial Dysfunction and Diabetic
Vascular Complications. Front. Pharmacol. 2020, 1, 590614. [CrossRef] g Q
y
g y
y
[
]
88. Li, Z.; Wu, N.; Wang, J.; Zhang, Q. Roles of Endovascular Calyx Related Enzymes in Endothelial Dysfunction and Diabetic
Vascular Complications. Front. Pharmacol. 2020, 1, 590614. [CrossRef] p
89. Dogné, S.; Flamion, B.; Caron, N. Endothelial Glycocalyx as a Shield Against Diabetic Vascular Complications: Involvement of
Hyaluronan and Hyaluronidases. Arterioscler. Thromb. Vasc Biol. 2018, 38, 1427–1439. [CrossRef] 90. van Splunder, H.; Villacampa, P.; Martínez-Romero, A.; Graupera, M. Pericytes in the disease spotlight. Trends Cell Biol. 2023,
34, 58–71. [CrossRef] 91. Brown, L.S.; Foster, C.G.; Courtney, J.M.; King, N.E.; Howells, D.W.; Sutherland, B.A. Pericytes and Neurovascular Function in
the Healthy and Diseased Brain. Front. Cell. Neurosci. 2019, 13, 282. [CrossRef] 92. Escribano, J.; Chen, M.B.; Moeendarbary, E.; Cao, X.; Shenoy, V.; Garcia-Aznar, J.M.; Kamm, R.D.; Spill, F. Balance of mechanical
forces drives endothelial gap formation and may facilitate cancer and immune-cell extravasation. PLoS Comput. Biol. 2019,
15, e1006395. [CrossRef] 93. Claesson-Welsh, L.; Dejana, E.; McDonald, D.M. Permeability of the Endothelial Barrier: Identifying and Reconciling Controver-
sies. Trends Mol. Med. 2021, 27, 314–331. [CrossRef] 94. Okamoto, T.; Kawamoto, E.; Takagi, Y.; Akita, N.; Hayashi, T.; Park, E.J.; Suzuki, K.; Shimaoka, M. Gap junction-mediated
regulation of endothelial cellular stiffness. Sci. Rep. 2017, 7, 6134. [CrossRef] 95. Zhao, F.; Zhong, L.; Luo, Y. Endothelial glycocalyx as an important factor in composition of blood-bra
Ther. 2021, 27, 26–35. [CrossRef] 95. Zhao, F.; Zhong, L.; Luo, Y. Endothelial glycocalyx as an important factor in composition of blood-brain barrier. CNS Neurosci. Ther. 2021, 27, 26–35. [CrossRef] Biologics 2024, 4 125 96. Yu, H.; Song, Y.Y.; Li, X.H. Early diabetic kidney disease: Focus on the glycocalyx. World J. Diabetes. 96.
Yu, H.; Song, Y.Y.; Li, X.H. Early diabetic kidney disease: Focus on the glycocalyx. World J. Diabetes. 2023, 14, 460–480. [CrossRef]
97
Suzuki K ; Miura T ; Okada H The endothelial glycocalyx All the same? No it is not Acute Med Surg 2023 10 e896 [CrossRef] g
y
y
g y
y
97.
Suzuki, K.; Miura, T.; Okada, H. The endothelial glycocalyx-All the same? No, it is not. Acute Med. Surg. References 2023, 14, 460–480. [CrossRef] 96. Yu, H.; Song, Y.Y.; Li, X.H. Early diabetic kidney disease: Focus on the glycocalyx. World J. Diabetes. 2023, 14, 460–480. [CrossRef]
97. Suzuki, K.; Miura, T.; Okada, H. The endothelial glycocalyx-All the same? No, it is not. Acute Med. Surg. 2023, 10, e896. [CrossRef] 98. Kong, C.; Elderman, M.; Cheng, L.; de Haan, B.J.; Nauta, A.; de Vos, P. Modulation of Intestinal Epithelial Glycocalyx Development
by Human Milk Oligosaccharides and Non-Digestible Carbohydrates. Mol. Nutr. Food Res. 2019, 63, e1900303. [CrossRef] y
g
g
y
99. Sun, W.W.; Krystofiak, E.S.; Leo-Macias, A.; Cui, R.; Sesso, A.; Weigert, R.; Ebrahim, S.; Kachar, B. Nanoarchitecture and dynamics of
the mouse enteric glycocalyx examined by freeze-etching electron tomography and intravital microscopy. Commun. Biol. 2020, 3, 5. [CrossRef] [
]
100. Fernández-Sarmiento, J.; Salazar-Peláez, L.M.; Carcillo, J.A. The Endothelial Glycocalyx: A Fundamental Determinant of Vascular
Permeability in Sepsis. Pediatr. Crit. Care Med. 2020, 21, e291–e300. [CrossRef] 101. Dull, R.O.; Hahn, R.G. The glycocalyx as a permeability barrier: Basic science and clinical evidence. Crit. Care 2022, 26, 273. [CrossRef] A.; Tomita, H.; Okada, H. Form follows function: The endothelial glycocalyx. Transl. Res. 2022, 247, 158–167. 102. Suzuki, A.; Tomita, H.; Okada, H. Form follows function: The endothelial glycocalyx. Transl. Res. 2022, 102. Suzuki, A.; Tomita, H.; Okada, H. Form follows function: The endothelial glycocalyx. Transl. Res. 2022, 247, 158–167. [CrossRef]
103. Trimm, E.; Red-Horse, K. Vascular endothelial cell development and diversity. Nat. Rev. Cardiol. 2023, 20, 197–210. [CrossRef]
104. Lim, J.; Machin, D.R.; Donato, A.J. The role of hyaluronan in endothelial glycocalyx and potential preventative lifestyle strategy
with advancing age. Curr. Top. Membr. 2023, 91, 139–156. [CrossRef] 103. Trimm, E.; Red-Horse, K. Vascular endothelial cell development and diversity. Nat. Rev. Cardiol. 2023, p
y
104. Lim, J.; Machin, D.R.; Donato, A.J. The role of hyaluronan in endothelial glycocalyx and potential preventative lifestyle strategy
with advancing age. Curr. Top. Membr. 2023, 91, 139–156. [CrossRef] g g
p
105. Qu, J.; Cheng, Y.; Wu, W.; Yuan, L.; Liu, X. Glycocalyx Impairment in Vascular Disease: Focus on Inflammation. Front. Cell Dev. Biol. 2021, 9, 730621. [CrossRef] [
]
106. Zavras, P.D.; Mehta, V.; Goel, S.; Billett, H.H. Increased incidence of thrombosis in a cohort of cancer patients with COVID-19. medRxiv 2021. [CrossRef] 107. Levi, M.; van Es, N. COVID-19 associated coagulopathy and thrombosis in cancer. Thromb. Res. 2022, 213 (Suppl. S1), S72–S76. [CrossRef] 108. u, H.; Song, Y.Y.; Li, X.H. Early diabetic kidney disease: Focus on the glycocalyx. World J. Diabetes. 2023, 14, 46
uzuki, K.; Miura, T.; Okada, H. The endothelial glycocalyx-All the same? No, it is not. Acute Med. Surg. 2023, 1 References Jones, F.K.; Stefan, A.; Kay, A.G.; Hyland, M.; Morgan, R.; Forsyth, N.R.; Pisconti, A.; Kehoe, O. Syndecan-3 regulates MSC
adhesion, ERK and AKT signalling in vitro and its deletion enhances MSC efficacy in a model of inflammatory arthritis in vivo. Sci. Rep. 2020, 10, 20487. [CrossRef] ker, T.E.; Kim, H.; Stevens, M.M. Glycosaminoglycan-based biomaterials for growth factor and cytokine
i ht h i
J Control Release 2019 313 131 147 [C
R f] p
121. Hachim, D.; Whittaker, T.E.; Kim, H.; Stevens, M.M. Glycosaminoglycan-based biomaterials for
delivery: Making the right choices. J. Control. Release 2019, 313, 131–147. [CrossRef] ker, T.E.; Kim, H.; Stevens, M.M. Glycosaminoglycan-based biomaterials for growth factor and cytokine
e right choices. J. Control. Release 2019, 313, 131–147. [CrossRef] y
g
g
122. Shamdani, S.; Chantepie, S.; Flageollet, C.; Henni-Chebra, N.; Jouan, Y.; Eymard, F.; Hay, E.; Cohen-Solal, M.; Papy-Garcia, D.;
Chevalier, X.; et al. Heparan sulfate functions are altered in the osteoarthritic cartilage. Arthritis Res. Ther. 2020, 22, 283. [CrossRef] 122. Shamdani, S.; Chantepie, S.; Flageollet, C.; Henni-Chebra, N.; Jouan, Y.; Eymard, F.; Hay, E.; Cohen-S
Chevalier, X.; et al. Heparan sulfate functions are altered in the osteoarthritic cartilage. Arthritis Res. Ther 122. Shamdani, S.; Chantepie, S.; Flageollet, C.; Henni-Chebra, N.; Jouan, Y.; Eymard, F.; Hay, E.; Cohen-Solal, M.; Papy-Garcia, D.;
Chevalier, X.; et al. Heparan sulfate functions are altered in the osteoarthritic cartilage. Arthritis Res. Ther. 2020, 22, 283. [CrossRef]
123. Margraf, A.; Lowell, C.A.; Zarbock, A. Neutrophils in acute inflammation: Current concepts and translational implications. Blood
2022, 139, 2130–2144. [CrossRef] p
g
123. Margraf, A.; Lowell, C.A.; Zarbock, A. Neutrophils in acute inflammation: Current concepts and translational implications. Blood
2022, 139, 2130–2144. [CrossRef] [
]
124. Hara, T.; Yabushita, S.; Yamamoto, C.; Kaji, T. Cell Density-Dependent Fibroblast Growth Factor-2 Signaling Regulates Syndecan-4
Expression in Cultured Vascular Endothelial Cells. Int. J. Mol. Sci. 2020, 21, 3698. [CrossRef] Biologics 2024, 4 126 125. Huang, X.; Reye, G.; Momot, K.I.; Blick, T.; Lloyd, T.; Tilley, W.D.; Hickey, T.E.; Snell, C.E.; Okolicsanyi, R.K.; Haupt, L.M.; et al. Heparanase Promotes Syndecan-1 Expression to Mediate Fibrillar Collagen and Mammographic Density in Human Breast Tissue
Cultured ex vivo. Front. Cell Dev. Biol. 2020, 8, 599. [CrossRef] 126. Guo, S.; Wu, X.; Lei, T.; Zhong, R.; Wang, Y.; Zhang, L.; Zhao, Q.; Huang, Y.; Shi, Y.; Wu, L. The Role and Therapeutic Value of
Syndecan-1 in Cancer Metastasis and Drug Resistance. Front. References Cell Dev. Biol. 2022, 9, 784983. [CrossRef] 127. Strand, M.E.; Vanhaverbeke, M.; Henkens, M.T.H.M.; Sikking, M.A.; Rypdal, K.B.; Braathen, B.; Almaas, V.M.; Tønnessen, T.;
Christensen, G.; Heymans, S.; et al. Inflammation and Syndecan-4 Shedding from Cardiac Cells in Ischemic and Non-Ischemic
Heart Disease. Biomedicines 2023, 11, 1066. [CrossRef] 128. Shaik, F.; Balderstone, M.J.M.; Arokiasamy, S.; Whiteford, J.R. Roles of Syndecan-4 in cardiac injury and repair. Int. J. Biochem. Cell
Biol. 2022, 146, 106196. [CrossRef] 129. El Masri, R.; Seffouh, A.; Roelants, C.; Seffouh, I.; Gout, E.; Pérard, J.; Dalonneau, F.; Nishitsuji, K.; Noborn, F.; Nikpour, M.; et al. Extracellular endosulfatase Sulf-2 harbors a chondroitin/dermatan sulfate chain that modulates its enzyme activity. Cell Rep. 2022, 38, 110516. [CrossRef] ,
,
[
]
130. Dao, D.T.; Anez-Bustillos, L.; Adam, R.M.; Puder, M.; Bielenberg, D.R. Heparin-Binding Epidermal Growth Factor-Like Growth
Factor as a Critical Mediator of Tissue Repair and Regeneration. Am. J. Pathol. 2018, 188, 2446–2456. [CrossRef] . Jayatilleke, K.M.; Hulett, M.D. Heparanase and the 131. Jayatilleke, K.M.; Hulett, M.D. Heparanase and the hallmarks of cancer. J. Transl. Med. 2020, 18, 453. [CrossRef] 132. Wang, J.; Feng, Y.; Liu, B.; Xie, W. Estrogen sulfotransferase and sulfatase in steroid homeostasis, metabolic disease, and cancer. Steroids 2023, 201, 109335. [CrossRef] 133. Okada, T.; Keino-Masu, K.; Nagamine, S.; Kametani, F.; Ohto, T.; Hasegawa, M.; van Kuppevelt, T.H.; Kunita, S.; Takahashi,
S.; Masu, M. Desulfation of Heparan Sulfate by Sulf1 and Sulf2 Is Required for Corticospinal Tract Formation. Sci. Rep. 2017,
7, 13847. [CrossRef] ,
[
]
134. Gao, Y.; Chen, N.; Fu, Z.; Zhang, Q. Progress of Wnt Signaling Pathway in Osteoporosis. Biomolecules 2 g
g
g
g
y
p
135. Yang, C.; Wang, C.; Zhou, J.; Liang, Q.; He, F.; Li, F.; Li, Y.; Chen, J.; Zhang, F.; Han, C.; et al. Fibronectin 1 activates WNT/β-catenin
i
li
t i d
t
i diff
ti ti
i i t
i
β1 i t
ti
L b I
ti
2020 100 1494 1502 [C
R f] 135. Yang, C.; Wang, C.; Zhou, J.; Liang, Q.; He, F.; Li, F.; Li, Y.; Chen, J.; Zhang, F.; Han, C.; et al. Fibronectin 1 activates WNT/β-catenin
signaling to induce osteogenic differentiation via integrin β1 interaction. Lab. Investig. 2020, 100, 1494–1502. [CrossRef] 136. Liu, J.; Xiao, Q.; Xiao, J.; Niu, C.; Li, Y.; Zhang, X.; Zhou, Z.; Shu, G.; Yin, G. Wnt/β-catenin signalling: Function, biological
mechanisms, and therapeutic opportunities. References Signal Transduct. Target. Ther. 2022, 7, 3. [CrossRef] Liu, J.; Xiao, Q.; Xiao, J.; Niu, C.; Li, Y.; Zhang, X.; Zhou, Z.; Shu, G.; Yin, G. Wnt/β-catenin signalling
mechanisms and therapeutic opportunities Signal Transduct Target Ther 2022 7 3 [CrossRef] 136. Liu, J.; Xiao, Q.; Xiao, J.; Niu, C.; Li, Y.; Zhang, X.; Zhou, Z.; Shu, G.; Yin, G. Wnt/β-catenin signal
mechanisms, and therapeutic opportunities. Signal Transduct. Target. Ther. 2022, 7, 3. [CrossRef] , J ;
, Q ;
, J ;
,
;
,
;
g,
;
,
;
,
;
,
/β
g
g
,
g
mechanisms, and therapeutic opportunities. Signal Transduct. Target. Ther. 2022, 7, 3. [CrossRef] ,
p
pp
g
g
, ,
[
]
137. Kinoshita, T.; Tomita, H.; Okada, H.; Niwa, A.; Hyodo, F.; Kanayama, T.; Matsuo, M.; Imaizumi, Y.; Kuroda, T.; Hatano, Y.; et al. Endothelial cell-specific reduction of heparan sulfate suppresses glioma growth in mice. Discov. Oncol. 2021, 12, 50. [CrossRef]
138. Du, Z.; Lovly, C.M. Mechanisms of receptor tyrosine kinase activation in cancer. Mol. Cancer. 2018, 17, 58. [CrossRef] p
pp
g
g
137. Kinoshita, T.; Tomita, H.; Okada, H.; Niwa, A.; Hyodo, F.; Kanayama, T.; Matsuo, M.; Imaizumi, Y.; Kuroda, T.; Hatano, Y.; et al. Endothelial cell-specific reduction of heparan sulfate suppresses glioma growth in mice. Discov. Oncol. 2021, 12, 50. [CrossRef] y
p
y
[
]
139. Lanzi, C.; Cassinelli, G. Receptor tyrosine kinases and heparan sulfate proteoglycans: Interplay providing anticancer targeting
strategies and new therapeutic opportunities. Biochem. Pharmacol. 2020, 178, 114084. [CrossRef] 140. Jiang, T.; Chen, Z.H.; Chen, Z.; Tan, D. SULF2 promotes tumorigenesis and inhibits apoptosis of cervical cancer cells through the
ERK/AKT signaling pathway. Braz. J. Med. Biol. Res. 2020, 53, e8901. [CrossRef] 141. Wang, C.; Shang, C.; Gai, X.; Song, T.; Han, S.; Liu, Q.; Zheng, X. Sulfatase 2-Induced Cancer-Associated Fibroblasts Promote
Hepatocellular Carcinoma Progression via Inhibition of Apoptosis and Induction of Epithelial-to-Mesenchymal Transition. Front. Cell Dev. Biol. 2021, 9, 631931. [CrossRef] 142. Ayoub, N.M.; Jaradat, S.K.; Al-Shami, K.M.; Alkhalifa, A.E. Targeting Angiogenesis in Breast Cancer: Current Evidence and
Future Perspectives of Novel Anti-Angiogenic Approaches. Front. Pharmacol. 2022, 13, 838133. [CrossRef] p
g
g
pp
143. Scarcella, M.; d’Angelo, D.; Ciampa, M.; Tafuri, S.; Avallone, L.; Pavone, L.M.; De Pasquale, V. The Key Role of Lysosomal
Protease Cathepsins in Viral Infections. Int. J. Mol. Sci. 2022, 23, 9089. [CrossRef] 144. 151. Latko, M.; Czyrek, A.; Por˛ebska, N.; Kuci´nska, M.; Otlewski, J.; Zakrzewska, M.; Opali´nski, Ł. Cross-Talk between Fibroblast
Growth Factor Receptors and Other Cell Surface Proteins. Cells 2019, 8, 455. [CrossRef] References [CrossRef] g
y y
g
g
164. Coombe, D.R.; Gandhi, N.S. Heparanase: A Challenging Cancer Drug Target. Front. Oncol. 2019, 9, 1316. [CrossRef] 164. Coombe, D.R.; Gandhi, N.S. Heparanase: A Challenging Cancer Drug Target. Front. Oncol. 2019, 9, 1316. [CrossRef]
165. Rangarajan, S.; Richter, J.R.; Richter, R.P.; Bandari, S.K.; Tripathi, K.; Vlodavsky, I.; Sanderson, R.D. Heparanase-enhanced
Shedding of Syndecan 1 and Its Role in Driving Disease Pathogenesis and Progression J Histochem Cytochem 2020 68 823 840 164. Coombe, D.R.; Gandhi, N.S. Heparanase: A Challenging Cancer Drug Target. Front. Oncol. 2019, 9, 1316. [CrossRef]
165. Rangarajan, S.; Richter, J.R.; Richter, R.P.; Bandari, S.K.; Tripathi, K.; Vlodavsky, I.; Sanderson, R.D. Heparanase-enhanced
Shedding of Syndecan-1 and Its Role in Driving Disease Pathogenesis and Progression. J. Histochem. Cytochem. 2020, 68, 823–840. [CrossRef] 166. Maacha, S.; Bhat, A.A.; Jimenez, L.; Raza, A.; Haris, M.; Uddin, S.; Grivel, J.C. Extracellular vesicles-mediated intercellular
communication: Roles in the tumor microenvironment and anti-cancer drug resistance. Mol. Cancer. 2019, 18, 55. [CrossRef] 167. Lee, K.M.; Seo, E.C.; Lee, J.H.; Kim, H.J.; Hwangbo, C. The Multifunctional Protein Syntenin-1: Regulato
Cellular Function, and Tumor Progression. Int. J. Mol. Sci. 2023, 24, 9418. [CrossRef] Cellular Function, and Tumor Progression. Int. J. Mol. Sci. 2023, 24, 9418. [CrossRef]
168. Gurung, S.; Perocheau, D.; Touramanidou, L.; Baruteau, J. The exosome journey: From biogenesis to uptake and intracellular
signalling Cell Commun Signal 2021 19 47 [CrossRef] 168. Gurung, S.; Perocheau, D.; Touramanidou, L.; Baruteau, J. The exosome journey: From biogenesis to
signalling. Cell Commun. Signal. 2021, 19, 47. [CrossRef] 169. Cerezo-Magaña, M.; Bång-Rudenstam, A.; Belting, M. The pleiotropic role of proteoglycans in extracellular vesicle mediated
communication in the tumor microenvironment. Semin. Cancer Biol. 2020, 62, 99–107. [CrossRef] 170. Castro-Cruz, M.; Hyka, L.; Daaboul, G.; Leblanc, R.; Meeussen, S.; Lembo, F.; Oris, A.; Van Herck, L.; Granjeaud, S.; David,
G.; et al. PDZ scaffolds regulate extracellular vesicle production, composition, and uptake. Proc. Natl. Acad. Sci. USA 2023,
120, e2310914120. [CrossRef] [
]
171. Klotz, D.M.; Link, T.; Wimberger, P.; Kuhlmann, J.D. Prognostic relevance of longitudinal HGF levels in serum of patients with
ovarian cancer. Mol. Oncol. 2021, 15, 3626–3638. [CrossRef] 172. Li, F.; Liu, P.; Huang, Y.; Li, L.; Zhang, S.; Yang, Z.; Wang, R.; Tao, Z.; Han, Z.; Fan, J.; et al. The Incremental Prognostic Value of
Hepatocyte Growth Factor in First-Ever Acute Ischemic Stroke: An Early Link Between Growth Factor and Interleukins. Front. Neurol. References Jia, T.; Jacquet, T.; Dalonneau, F.; Coudert, P.; Vaganay, E.; Exbrayat-Héritier, C.; Vollaire, J.; Josserand, V.; Ruggiero, F.; Coll, J.L.;
et al. FGF-2 promotes angiogenesis through a SRSF1/SRSF3/SRPK1-dependent axis that controls VEGFR1 splicing in endothelial
cells. BMC Biol. 2021, 19, 173. [CrossRef] [
]
153. Ferreira, A.; Royaux, I.; Liu, J.; Wang, Z.; Su, G.; Moechars, D.; Callewaert, N.; De Muynck, L. The 3-O sulfation of heparan sulfate
proteoglycans contributes to the cellular internalization of tau aggregates. BMC Mol. Cell Biol. 2022, 23, 61. [CrossRef] au aggregates. BMC Mol. Cell Biol. 2022, 23, 61. [Cross 154. Poli, M.; Anower-E-Khuda, F.; Asperti, M.; Ruzzenenti, P.; Gryzik, M.; Denardo, A.; Gordts, P.L.S.M.; Arosio, P.; Esko, J.D. Hepatic
heparan sulfate is a master regulator of hepcidin expression and iron homeostasis in human hepatocytes and mice. J. Biol. Chem. 2019, 294, 13292–13303. [CrossRef] 155. De Pasquale, V.; Pavone, L.M. Heparan Sulfate Proteoglycan Signaling in Tumor Microenvironment. Int. J. Mol. Sci. 2020, 21, 6588. [CrossRef] 156. Liu, Y.J.; Wang, C. A review of the regulatory mechanisms of extracellular vesicles-mediated intercellular communication. Cell
Commun. Signal. 2023, 21, 77. [CrossRef] g
157. Ginini, L.; Billan, S.; Fridman, E.; Gil, Z. Insight into Extracellular Vesicle-Cell Communication: From Cell Recognition to
Intracellular Fate. Cells 2022, 11, 1375. [CrossRef] 158. Chang, W.H.; Cerione, R.A.; Antonyak, M.A. Extracellular Vesicles and Their Roles in Cancer Progression. Methods Mol. Biol. 2021, 2174, 143–170. [CrossRef] 159. Hu, W.; Liu, C.; Bi, Z.Y.; Zhou, Q.; Zhang, H.; Li, L.L.; Zhang, J.; Zhu, W.; Song, Y.Y.; Zhang, F.; et al. Comprehensive landscape of
extracellular vesicle-derived RNAs in cancer initiation, progression, metastasis and cancer immunology. Mol. Cancer 2020, 19, 102. [CrossRef] [
]
160. Wu, Q.; Zhang, H.; Sun, S.; Wang, L.; Sun, S. Extracellular vesicles and immunogenic stress in cancer. Cell Death Dis. 2021, 12, 894. [CrossRef] 161. Purushothaman, A.; Sanderson, R.D. Heparanase: A Dynamic Promoter of Myeloma Progression. Adv. Exp. Med. Biol. 2020,
1221, 331–349. [CrossRef] u, K.S.; Gao, J.; Zhai, W.L.; Yang, J.H.; Zhang, S.J. Exosome biogenesis: Machinery, regulation, and therapeutic
er. Mol. Cancer 2022, 21, 207. [CrossRef] 162. Han, Q.F.; Li, W.J.; Hu, K.S.; Gao, J.; Zhai, W.L.; Yang, J.H.; Zhang, S.J. Exosome biogenesis: Machinery, re
implications in cancer. Mol. Cancer 2022, 21, 207. [CrossRef] 163. Kooijmans, S.A.A.; de Jong, O.G.; Schiffelers, R.M. Exploring interactions between extracellular vesicles and cells for innovative
drug delivery system design. Adv. Drug Deliv. Rev. 2021, 173, 252–278. References Gomes, C.P.; Fernandes, D.E.; Casimiro, F.; da Mata, G.F.; Passos, M.T.; Varela, P.; Mastroianni-Kir 144. Gomes, C.P.; Fernandes, D.E.; Casimiro, F.; da Mata, G.F.; Passos, M.T.; Varela, P.; Mastroianni
Cathepsin L in COVID-19: From Pharmacological Evidences to Genetics. Front. Cell. Infect. Microbiol 144. Gomes, C.P.; Fernandes, D.E.; Casimiro, F.; da Mata, G.F.; Passos, M.T.; Varela, P.; Mastroianni Kirsztajn, G.; Pesquero, J.B. Cathepsin L in COVID-19: From Pharmacological Evidences to Genetics. Front. Cell. Infect. Microbiol. 2020, 10, 589505. [CrossRef]
145. Yu, Y.; Williams, A.; Zhang, X.; Fu, L.; Xia, K.; Xu, Y.; Zhang, F.; Liu, J.; Koffas, M.; Linhardt, R.J. Specificity and action pattern of
heparanase Bp, a β-glucuronidase from Burkholderia pseudomallei. Glycobiology 2019, 29, 572–581. [CrossRef] Cathepsin L in COVID-19: From Pharmacological Evidences to Genetics. Front. Cell. Infect. Microbiol. 20 p
g
f
145. Yu, Y.; Williams, A.; Zhang, X.; Fu, L.; Xia, K.; Xu, Y.; Zhang, F.; Liu, J.; Koffas, M.; Linhardt, R.J. Specificity and action pattern of
heparanase Bp, a β-glucuronidase from Burkholderia pseudomallei. Glycobiology 2019, 29, 572–581. [CrossRef] p
p
g
p
y
gy
146. Collins, L.E.; Troeberg, L. Heparan sulfate as a regulator of inflammation and immunity. J
[CrossRef] 146. Collins, L.E.; Troeberg, L. Heparan sulfate as a regulator of inflammation and immunity. J. Leukoc. Biol. 2019, 105, 81–92. [CrossRef] 147. Farrugia, B.L.; Lord, M.S.; Melrose, J.; Whitelock, J.M. The Role of Heparan Sulfate in Inflammation, and the Development of
Biomimetics as Anti-Inflammatory Strategies. J. Histochem. Cytochem. 2018, 66, 321–336. [CrossRef] 148. de Boer, C.; Armstrong, Z.; Lit, V.A.J.; Barash, U.; Ruijgrok, G.; Boyango, I.; Weitzenberg, M.M.; Schröder, S.P.; Sarris, A.J.C.;
Meeuwenoord, N.J.; et al. Mechanism-based heparanase inhibitors reduce cancer metastasis in vivo. Proc. Natl. Acad. Sci. USA
2022, 119, e2203167119. [CrossRef] 149. He, P.; Zhang, X.; Xia, K.; Green, D.E.; Baytas, S.; Xu, Y.; Pham, T.; Liu, J.; Zhang, F.; Almond, A.; et al. Chemoenzymatic synthesis
of sulfur-linked sugar polymers as heparanase inhibitors. Nat. Commun. 2022, 13, 7438. [CrossRef] 150. Hassan, N.; Greve, B.; Espinoza-Sánchez, N.A.; Götte, M. Cell-surface heparan sulfate proteoglycans as multifunctional integrators
of signaling in cancer. Cell Signal. 2021, 77, 109822. [CrossRef] 151. Latko, M.; Czyrek, A.; Por˛ebska, N.; Kuci´nska, M.; Otlewski, J.; Zakrzewska, M.; Opali´nski, Ł. Cross-Talk between Fibroblast
Growth Factor Receptors and Other Cell Surface Proteins. Cells 2019, 8, 455. [CrossRef] Biologics 2024, 4 127 152. References Putting the brakes on autophagy: The role of heparan sulfate modified proteins in the
balance of anabolic and catabolic pathways and intracellular quality control. Matrix Biol. 2021, 100–101, 173–181. [CrossRef]
183
M k
N L kk
M G
h
d
K S k l
I M
h
i
l W
S
d
M l
l
S
f P
i
ll l
L p
p
g
182. Schultheis, N.; Jiang, M.; Selleck, S.B. Putting the brakes on autophagy: The role of heparan sulfate modified proteins in the
balance of anabolic and catabolic pathways and intracellular quality control. Matrix Biol. 2021, 100–101, 173–181. [CrossRef]
183. Makarova, N.; Lekka, M.; Gnanachandran, K.; Sokolov, I. Mechanical Way to Study Molecular Structure of Pericellular Layer. ACS Appl. Mater. Interfaces 2023, 15, 35962–35972. [CrossRef] 183. Makarova, N.; Lekka, M.; Gnanachandran, K.; Sokolov, I. Mechanical Way to Study Molecular St
ACS Appl. Mater. Interfaces 2023, 15, 35962–35972. [CrossRef] pp
f
184. Alshehri, M.A.; Alshehri, M.M.; Albalawi, N.N.; Al-Ghamdi, M.A.; Al-Gayyar, M.M.H. Heparan sulfate proteoglycans and their
modification as promising anticancer targets in hepatocellular carcinoma. Oncol. Lett. 2021, 21, 173. [CrossRef] i
p
g
g
p
185. Lanzi, C.; Yates, E.A.; Cassinelli, G. Editorial: Heparan Sulfate Proteoglycans and Their Endogenous Modifying Enzymes: Cancer
Players, Biomarkers and Therapeutic Targets. Front. Oncol. 2020, 10, 195. [CrossRef] y
p
g
186. Furini, S.; Falciani, C. Expression and Role of Heparan Sulfated Proteoglycans in Pancreatic Cancer. Front. Oncol. 2021, 11, 695858. [CrossRef] 187. Hoffmann, M.; Snyder, N.L.; Hartmann, L. Polymers Inspired by Heparin and Heparan Sulfate for Viral Targeting. Macromolecules
2022, 55, 7957–7973. [CrossRef] 188. Volland, A.; Lohmüller, M.; Heilmann, E.; Kimpel, J.; Herzog, S.; von Laer, D. Heparan sulfate proteoglycans serve as alternative
receptors for low affinity LCMV variants. PLoS Pathog. 2021, 17, e1009996. [CrossRef] 189. Yu, M.; Zhang, T.; Zhang, W.; Sun, Q.; Li, H.; Li, J.P. Elucidating the Interactions Between Heparin/Heparan Sulfate and
SARS-CoV-2-Related Proteins-An Important Strategy for Developing Novel Therapeutics for the COVID-19 Pandemic. Front. Mol. Biosci. 2021, 7, 628551. [CrossRef] 190. Kamimura, K.; Maeda, N. Glypicans and Heparan Sulfate in Synaptic Development, Neural Plasticity, and Neurological Disorders. Front. Neural Circuits 2021, 15, 595596. [CrossRef] Syafruddin, S.E.; Mohtar, M.A.; Syahir, A. CD44: A Multifunctional Mediator of Cancer Progression
, 1850. [CrossRef] 191. Hassn Mesrati, M.; Syafruddin, S.E.; Mohtar, M.A.; Syahir, A. CD44: A Multifunctional Med
Biomolecules 2021, 11, 1850. [CrossRef] 192. References 2021, 12, 691886. [CrossRef] 173. Belvedere, R.; Novizio, N.; Palazzo, M.; Pessolano, E.; Petrella, A. The pro-healing effects of heparan sulfate and growth factors
are enhanced by the heparinase enzyme: New association for skin wound healing treatment. Eur. J. Pharmacol. 2023, 960, 176138. [CrossRef] nn, M. Soluble syndecans: Biomarkers for diseases and therapeutic options. Br. J. Pharmacol. 2019, 176, 67–81 174. Bertrand, J.; Bollmann, M. Soluble syndecans: Biomarkers for diseases and therapeutic options. Br. [CrossRef] 175. Hassan, N.; Bückreiß, N.; Efing, J.; Schulz-Fincke, M.; König, P.; Greve, B.; Bendas, G.; Götte, M. The Heparan Sulfate Proteoglycan
Syndecan-1 Triggers Breast Cancer Cell-Induced Coagulability by Induced Expression of Tissue Factor. Cells 2023, 12, 910. [CrossRef] 176. Nadanaka, S.; Bai, Y.; Kitagawa, H. Cleavage of Syndecan-1 Promotes the Proliferation of the Basal-Like Breast Cancer Cell Line
BT-549 Via Akt SUMOylation. Front. Cell Dev. Biol. 2021, 9, 659428. [CrossRef] 177. Koganti, R.; Suryawanshi, R.; Shukla, D. Heparanase, cell signaling, and viral infections. Cell. Mol. Life Sci. 2020, 77, 5059–5077. [CrossRef] Biologics 2024, 4 128 178. Xue, S.; Zhou, F.; Zhao, T.; Zhao, H.; Wang, X.; Chen, L.; Li, J.P.; Luo, S.Z. Phase separation on cell surface facilitates bFGF signal
transduction with heparan sulphate. Nat. Commun. 2022, 13, 1112. [CrossRef] p
p
179. Jang, B.; Song, H.K.; Hwang, J.; Lee, S.; Park, E.; Oh, A.; Hwang, E.S.; Sung, J.Y.; Kim, Y.N.; Park, K.; et al. Shed syndecan-2
enhances colon cancer progression by increasing cooperative angiogenesis in the tumor microenvironment. Matrix Biol. 2022,
107, 40–58. [CrossRef] [
]
180. Czarnowski, D. Syndecans in cancer: A review of function, expression, prognostic value, and therapeutic significance. Cancer
Treat. Res. Commun. 2021, 27, 100312. [CrossRef] i, R.; Yadavalli, T.; Agelidis, A.; Suryawanshi, R.; Shukla, D. Heparanase-Regulated Syndecan-1 Shedding Hadigal, S.; Koganti, R.; Yadavalli, T.; Agelidis, A.; Suryawanshi, R.; Shukla, D. Heparanase-Regulated S
Facilitates Herpes Simplex Virus 1 Egress J Virol 2020 94 e01672-19 [CrossRef] 181. Hadigal, S.; Koganti, R.; Yadavalli, T.; Agelidis, A.; Suryawanshi, R.; Shukla, D. Heparanase-Regulated Syndecan-1 Shedding
Facilitates Herpes Simplex Virus 1 Egress. J. Virol. 2020, 94, e01672-19. [CrossRef] 181. Hadigal, S.; Koganti, R.; Yadavalli, T.; Agelidis, A.; Suryawanshi, R.; Shukla, D. Heparana
Facilitates Herpes Simplex Virus 1 Egress. J. Virol. 2020, 94, e01672-19. [CrossRef] ates Herpes Simplex Virus 1 Egress. J. Virol. 2020, 94, e01672-19. [CrossRef] p
p
g
182. Schultheis, N.; Jiang, M.; Selleck, S.B. 204. Guo, R.; Chen, M.; Ding, Y.; Yang, P.; Wang, M.; Zhang, H.; He, Y.; Ma, H. Polysaccharides as Potential Anti-tumor
Biomacromolecules—A Review. Front. Nutr. 2022, 9, 838179. [CrossRef] References Weng, X.; Maxwell-Warburton, S.; Hasib, A.; Ma, L.; Kang, L. The membrane receptor CD44: Novel insights into metabolism. Trends Endocrinol. Metab. 2022, 33, 318–332. [CrossRef] 193. Gaiteiro, C.; Soares, J.; Relvas-Santos, M.; Peixoto, A.; Ferreira, D.; Paulo, P.; Brandão, A.; Fernandes, E.; Azevedo, R.; Palmeira,
C.; et al. Glycoproteogenomics characterizes the CD44 splicing code associated with bladder cancer invasion. Theranostics 2022,
12, 3150–3177. [CrossRef] [
]
194. Liao, C.; Wang, Q.; An, J.; Chen, J.; Li, X.; Long, Q.; Xiao, L.; Guan, X.; Liu, J. CD44 Glycosylation as a Therapeutic Target in
Oncology. Front. Oncol. 2022, 12, 883831. [CrossRef] 195. Patil, S. CD44 Sorted Cells Have an Augmented Potential for Proliferation, Epithelial-Mesenchymal Transition, Stemness, and a
Predominantly Inflammatory Cytokine and Angiogenic Secretome. Curr. Issues Mol. Biol. 2021, 43, 423–433. [CrossRef] 196. Hinneh, J.A.; Gillis, J.L.; Moore, N.L.; Butler, L.M.; Centenera, M.M. The role of RHAMM in cancer: Exposing novel therapeutic
vulnerabilities. Front. Oncol. 2022, 12, 982231. [CrossRef] 197. Chaudhry, G.E.; Akim, A.; Naveed Zafar, M.; Safdar, N.; Sung, Y.Y.; Muhammad, T.S.T. Understanding Hyaluronan Receptor
(CD44) Interaction, HA-CD44 Activated Potential Targets in Cancer Therapeutics. Adv. Pharm. Bull. 2021, 11, 426–438. [CrossRef] 198. Ma, Z.; Shi, S.; Ren, M.; Pang, C.; Zhan, Y.; An, H.; Sun, F. Molecular mechanism of CD44 homodimerization modulated by
palmitoylation and membrane environments. Biophys. J. 2022, 121, 2671–2683. [CrossRef] 199. Guo, Q.; Yang, C.; Gao, F. The state of CD44 activation in cancer progression and therapeutic targeting. F
[CrossRef] ao, F. The state of CD44 activation in cancer progression and therapeutic targeting. FEBS J. 2022, 89, 7970–7986 200. Sanjanwala, D.; Londhe, V.; Trivedi, R.; Bonde, S.; Sawarkar, S.; Kale, V.; Patravale, V. Polysaccharide-based hydrogels for medical
devices, implants and tissue engineering: A review. Int. J. Biol. Macromol. 2024, 256, 128488. [CrossRef] p
g
g
J
[
]
201. Farasati Far, B.; Naimi-Jamal, M.R.; Safaei, M.; Zarei, K.; Moradi, M.; Yazdani Nezhad, H. A Review on Biomedical Application of
Polysaccharide-Based Hydrogels with a Focus on Drug Delivery Systems. Polymers 2022, 14, 5432. [CrossRef] 202. Wang, A.; Liu, Y.; Zeng, S.; Liu, Y.; Li, W.; Wu, D.; Wu, X.; Zou, L.; Chen, H. Dietary Plant Polysaccharides for Cancer Prevention:
Role of Immune Cells and Gut Microbiota, Challenges and Perspectives. Nutrients 2023, 15, 3019. [CrossRef] 203. Khan, R.; Shah, M.D.; Shah, L.; Lee, P.C.; Khan, I. Bacterial polysaccharides-A big source for prebiotics and therapeutics. Front. Nutr. 2022, 9, 1031935. 203. Khan, R.; Shah, M.D.; Shah, L.; Lee, P.C.; Khan, I. Bacterial polysaccharides-A big source for prebiotics and therapeutics. Front.
Nutr. 2022, 9, 1031935. [CrossRef] References [CrossRef] 204. Guo, R.; Chen, M.; Ding, Y.; Yang, P.; Wang, M.; Zhang, H.; He, Y.; Ma, H. Polysaccharides as Potential Anti-tumor
Biomacromolecules—A Review. Front. Nutr. 2022, 9, 838179. [CrossRef] Biologics 2024, 4 129 205. Zhao, S.; Han, T.; Pei, X.; Song, Y.; Zhang, Y.; Liu, L.; Wang, X.; Hou, W.; Sun, C. The association of diet carbohydrates consumption
with cognitive function among US older adults modification by daily fasting duration. Front. Aging Neurosci. 2022, 4, 991007. [CrossRef] [
]
206. ˇCaval, T.; Alisson-Silva, F.; Schwarz, F. Roles of glycosylation at the cancer cell surface: Opportunities for large scale glycopro-
teomics. Theranostics 2023, 13, 2605–2615. [CrossRef] ,
,
[
]
207. Choi, Y.; Kim, J.; Chae, J.; Hong, J.; Park, J.; Jeong, E.; Kim, H.; Tanaka, M.; Okochi, M.; Choi, J. Surface glycan targeting for cancer
nano-immunotherapy. Control. Release 2022, 342, 321–336. [CrossRef] py
208. Perrin, L.; Belova, E.; Bayarmagnai, B.; Tüzel, E.; Gligorijevic, B. Invadopodia enable cooperative invasion and metastasis of
breast cancer cells. Commun. Biol. 2022, 5, 758. [CrossRef] 208. Perrin, L.; Belova, E.; Bayarmagnai, B.; Tüzel, E.; Gligorijevic, B. Invadopodia enable cooperative invasion and metastasis of
breast cancer cells. Commun. Biol. 2022, 5, 758. [CrossRef] 209. Medrano-González, P.A.; Rivera-Ramírez, O.; Montaño, L.F.; Rendón-Huerta, E.P. Proteolytic Processing of CD44 and Its
Implications in Cancer. Stem Cells Int. 2021, 2021, 6667735. [CrossRef] 209. Medrano-González, P.A.; Rivera-Ramírez, O.; Montaño, L.F.; Rendón-Huerta, E.P. Proteolytic Processing of CD44 and Its
Implications in Cancer. Stem Cells Int. 2021, 2021, 6667735. [CrossRef] p
210. Wei, C. The multifaceted roles of matrix metalloproteinases in lung cancer. Front. Oncol. 2023, 13, 1 p
210. Wei, C. The multifaceted roles of matrix metalloproteinases in lung can p
g
211. Wang, Y.Y.; Vadhan, A.; Chen, P.H.; Lee, Y.L.; Chao, C.Y.; Cheng, K.H.; Chang, Y.C.; Hu, S.C.; Yuan, S.F. CD44 Promotes Lung
Cancer Cell Metastasis through ERK-ZEB1 Signaling. Cancers 2021, 13, 4057. [CrossRef] p
g
211. Wang, Y.Y.; Vadhan, A.; Chen, P.H.; Lee, Y.L.; Chao, C.Y.; Cheng, K.H.; Chang, Y.C.; Hu, S.C.; Yuan, S.F. CD44 Promotes Lung
Cancer Cell Metastasis through ERK-ZEB1 Signaling. Cancers 2021, 13, 4057. [CrossRef] g
g
g
212. Emami Nejad, A.; Najafgholian, S.; Rostami, A.; Sistani, A.; Shojaeifar, S.; Esparvarinha, M.; Nedaeinia, R.; Haghjooy Javanmard,
S.; Taherian, M.; Ahmadlou, M.; et al. The role of hypoxia in the tumor microenvironment and development of cancer stem cell:
A novel approach to developing treatment. Cancer Cell Int. References 2021, 21, 62. [CrossRef] g
g
g
212. Emami Nejad, A.; Najafgholian, S.; Rostami, A.; Sistani, A.; Shojaeifar, S.; Esparvarinha, M.; Nedaeinia, R.; Haghjooy Javanmard,
S.; Taherian, M.; Ahmadlou, M.; et al. The role of hypoxia in the tumor microenvironment and development of cancer stem cell:
A novel approach to developing treatment. Cancer Cell Int. 2021, 21, 62. [CrossRef] pp
p
g
,
,
[
]
213. Essa, A.A.M.; Deraz, E.M. Expression of CD44 (NKI-P1) in oral squamous cell carcinoma associated vascular endothelial cells: A
relationship to tumor angiogenesis. Saudi Dent. J. 2022, 34, 21–26. [CrossRef] pp
p
g
[
]
213. Essa, A.A.M.; Deraz, E.M. Expression of CD44 (NKI-P1) in oral squamous cell carcinoma associated vascular endothelial cells: A
relationship to tumor angiogenesis. Saudi Dent. J. 2022, 34, 21–26. [CrossRef] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.
|
https://openalex.org/W3183365833
|
https://www.scielo.br/j/reben/a/qmvyc4QwHcmh4ZC8Df3vtYd/?lang=en&format=pdf
|
English
| null |
Dialogic educational practices in the context of child intoxication: an approach based on Paulo Freire
|
Revista Brasileira de Enfermagem
| 2,021
|
cc-by
| 7,845
|
RESUMO Objetivo: Analisar os saberes e práticas educativas dos profissionais da Estratégia Saúde da
Família em intoxicação infantil para autonomia e empoderamento das famílias. Métodos:
Pesquisa qualitativa, realizada por meio de entrevistas semiestruturadas com 50 profissionais
das equipes da Estratégia Saúde da Família de um município do noroeste do Paraná. Os dados
foram problematizados por meio da concepção dialógica e da autonomia da abordagem
sociocultural de Paulo Freire. Resultados: Os depoimentos revelaram várias formas de
perceber o cuidado em intoxicação infantil e sua relação com a ação educativa, mas estas
eram práticas educativas tradicionais, com elementos centrados em difundir informações
sobre saúde-doença, provavelmente porque a abordagem dialógica ainda é abstrata e sem
mediação com a prática concreta. Considerações finais: Os saberes e as práticas educativas
dos profissionais para autonomia e empoderamento das famílias sugeriram uma fase de
transição entre o modelo curativista/biomédico e o dialógico. Magda Lúcia Félix de OliveiraI
ORCID: 0000-0003-4095-9382 Magda Lúcia Félix de OliveiraI
ORCID: 0000-0003-4095-9382 I Universidade Estadual de Maringá. Maringá, Paraná, Brazil. II Universidade Estadual de Londrina. Londrina, Paraná, Brazil. I Universidade Estadual de Maringá. Maringá, Paraná, Brazil. II Universidade Estadual de Londrina. Londrina, Paraná, Brazil. How to cite this article:
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ,
Tacla MTGM, Baldissera VDA, et al. Dialogic educational
practices in the context of child intoxication: an approach
based on Paulo Freire. Rev Bras Enferm. 2021;74(5):e20201196. https://doi.org/10.1590/0034-7167-2020-1196
Corresponding author:
Camila Cristiane Formaggi Sales Ribeiro
E-mail: camila_cfs14@hotmail.com
EDITOR IN CHIEF: Antonio José de Almeida Filho
ASSOCIATE EDITOR: Hugo Fernandes
Submission: 11-26-2020 Approval: 03-05-2021 How to cite this article:
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ,
Tacla MTGM, Baldissera VDA, et al. Dialogic educational
practices in the context of child intoxication: an approach
based on Paulo Freire. Rev Bras Enferm. 2021;74(5):e20201196. https://doi.org/10.1590/0034-7167-2020-1196 Descritores: Educação em Saúde; Atenção Primária à Saúde; Envenenamento; Saúde da
Criança; Abordagem Freireana. ORIGINAL ARTICLE ORIGINAL ARTICLE Camila Cristiane Formaggi Sales RibeiroI
ORCID: 0000-0002-6645-1299 Martina Mesquita TononI
ORCID: 0000-0001-9113-9750 Práticas educativas dialógicas no contexto das intoxicações infantis: uma abordagem freireana
Prácticas educativas dialógicas en el contexto de las intoxicaciones infantiles: un abordaje basado en Paulo Freire Práticas educativas dialógicas no contexto das intoxicações infantis: uma abordagem freireana
Prácticas educativas dialógicas en el contexto de las intoxicaciones infantiles: un abordaje basado en Paulo Freire ABSTRACT Camila Cristiane Formaggi Sales RibeiroI
ORCID: 0000-0002-6645-1299
Martina Mesquita TononI
ORCID: 0000-0001-9113-9750
Jéssica Yumi de OliveiraI
ORCID: 0000-0002-6434-0579
Márcia Regina Jupi GuedesI
ORCID: 0000-0002-2480-0438
Mauren Teresa Grubisich Mendes TaclaII
ORCID: 0000-0001-8928-3366
Vanessa Denardi Antoniassi BaldisseraI
ORCID: 0000-0003-1680-9165
Magda Lúcia Félix de OliveiraI
ORCID 0000 0003 4095 9382 Objective: To analyze the knowledge and practices of education of Family Health Strategy
professionals regarding child intoxication for the autonomy and empowering of the families. Methods: Qualitative research, carried out through semistructured interviews with 50
professionals from the Family Health Strategy of a city in the Northwest of Paraná. Data was
problematized using Paulo Freire’s sociocultural approach and its concepts of dialogism
and autonomy. Results: The statements showed the many ways in which the care for child
intoxication is perceived and their relations with educational action, but these were traditional
educational practices, with elements focused on the dissemination of information about
health-disease, probably because the dialogic approach is still abstract and has no interface
with the concrete practices. Final considerations: The education knowledge and practices
of the professionals for the autonomy and empowering of families suggested that there is a
transition stage between the biomedical model, focused on a cure, and the dialogical one. Descriptors: Health Education; Primary Health Care; Poisoning; Child Health; Freirean Approach. 1
Rev Bras Enferm. 2021;74(5): e20201196
8
of METHODS Understanding the home environment of the family requires the
mobilization of the health professionals, especially those from the
Family Health Strategy (ESF), since the intoxication can manifest
differently depending on the many environments where it takes
place. Therefore, the proposals for integral attention to children
and family’s health should intervene through educational prac-
tices that include listening, the satisfaction of needs, and allow
for autonomy, extrapolating simple assistance and connecting
these people in social networks based on perspectives that lead
to the confrontation and minimization of vulnerability factors(5-6). Study setting This study was carried out in the PHC network of the city of
Maringá, located in the Northwest of the State of Paraná, in Brazil. This includes 35 Primary Health Care Units (UBS), associated to 9
Centers for Support to Family Health (NASF) and 74 ESF teams
(ESFt). To determine the places where to carry out the study, the
PHC units were separated according to the NASFs to which they
were associated. One PHC Unit was selected in the scope of each
NASF, adding up to 9 ESFts. As the perspective of the ESF workers is broadened with
regard to their professional activities, it can be noted that an
authentic dialog makes the professional-family relations easier,
helping subjects to feel valued and aware of how important
their social role is in the transformation of reality. This study is
relevant because it aims to encourage ESF workers to reflect on RESUMEN
Obj
i
A Objetivo: Analizar los saberes y prácticas educativas de profesionales de la Estrategia
Salud de la Familia en intoxicación infantil para autonomía y empoderamiento de familias. Métodos: Investigación cualitativa, realizada por medio de entrevistas semiestructuradas
con 50 profesionales de equipos de la Estrategia Salud de la Familia de un municipio del
noroeste de Paraná. Datos fueron problematizados por medio de la concepción dialógica y de
la autonomía del abordaje sociocultural de Paulo Freire. Resultados: Deposiciones revelaron
varias maneras de percibir el cuidado en intoxicación infantil y su relación con la acción
educativa, pero estas eran prácticas educativas tradicionales, con elementos centrados en
difundir informaciones sobre salud-enfermedad probablemente porque el abordaje dialógico
aún es abstracto y sin mediación con la práctica concreta. Consideraciones finales: Los saberes
y las prácticas educativas de los profesionales para autonomía y empoderamiento de las
familias sugirieron una fase de transición entre el modelo curativa/biomédico y el dialógico. Descriptores: Educación en Salud; Atención Primaria de Salud; Envenenamiento; Salud del
Niño; Enfoque Freireano. Corresponding author:
Camila Cristiane Formaggi Sales Ribeiro
E-mail: camila_cfs14@hotmail.com EDITOR IN CHIEF: Antonio José de Almeida Filho
ASSOCIATE EDITOR: Hugo Fernandes Submission: 11-26-2020 Approval: 03-05-2021 1
Rev Bras Enferm. 2021;74(5): e20201196
8
of https://doi.org/10.1590/0034-7167-2020-1196 Dialogic educational practices in the context of child intoxication: an approach based on Paulo Freire
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ, Tacla MTGM, Baldissera VDA, et al. Dialogic educational practices in the context of child intoxication: an approach based on Paulo Freire
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ, Tacla MTGM, Baldissera VDA, et al. Ethical aspects This research was approved by the Permanent Research Eth-
ics Committee for Researches Involving Human Beings from the
Universidade Estadual de Maringá. The study was approved by the
Municipal Secretariat of Health from Maringá, and participants
gave their spontaneous consent to participate in the study by
signing the Free and Informed Consent Form. However, the review of literature carried out for this study
showed the scarcity of research that analyze the way in which
care should be offered by the PHC in order to prevent child intoxi-
cation, from a dialogic perspective that aims to develop critical
awareness, based on the understanding that the community is a
place for social and cultural changes, as opposed to a place where
to simply carry out investigations. Authors(5,7-8) have highlighted
that educational actions based on dialogic, pedagogical, and
liberating referents are capable of promoting the autonomy of
families and encourage them to assume the main role in the
performance of collective action and in the exercise of citizenship. OBJECTIVE To analyze the educational knowledge and practices of Fam-
ily Health Strategy professionals about child intoxication for the
autonomy and empowering of the families. INTRODUCTION dialogic educational practices based on effective communicational
processes, to encourage autonomy, joint responsibility, and the
valuing of families, in addition to aiding them to make decisions
and change behaviors that could put their health at risk(8,12). Intoxication is an important public health problem and a sig-
nificant cause of morbidity and mortality in children, representing
approximately 3% of all patients hospitalized in emergency ser-
vices(1). Every year, nearly 45 thousand children and adolescents
below 20 years old die due to intoxication, and it has been reported
that, among people in this age group, the rate of death due to
poisoning is of 1.8 per 100 thousand throughout the world(1-2). Therefore, considering child accidents, especially intoxications,
as preventable health issues that can be prevented and impacted
by PHC actions, and considering the references of this work, that
is, the educational practices in the perspective of Paulo Freire’s
dialogism, this study contributes for the knowledge in the field of
child and family health, improving the actions of health workers
as mediators of educational actions, attuned to the idea that the
role of the family is strategic and central for policies that promote
health and empowering. To Disfani et al.(2), the rates of incidence and mortality due to
child intoxication can be diminished by implementing educational
intervention programs, since intoxications are avoidable events
that can be prevented and are conditions that can be impacted
by actions from the Primary Health Care (PHC). As a result, home,
the family environment, is a privileged setting for the develop-
ment of actions that promote health. The home can be especially
unsafe for children, since it contains objects and materials that
offer risk in all of its rooms. This is especially true with regard to
the diversity of chemical products, which favor intoxication and
are mostly stored in places that the child can easily access(3-4). Type of study This study was designed following a qualitative, descriptive,
and exploratory approach(13). It results from a report that origi-
nated from the Master’s Degree dissertation Práticas educativas no
cuidado à intoxicação infantil na atenção básica: uma perspectiva
dialógica (Educational Practices in the Care for Child Intoxication
in Primary Care: a Dialogic Perspective), and followed all Con-
solidated Criteria for Reporting Qualitative Research (COREQ)(14),
which guide the writing of qualitative research articles. The dialogic perspective proposed by Paulo Freire(9) in his socio-
cultural approach is important as a reference on which to base the
autonomy and empower families. This perspective is conducted by
reflection-action, aiming to reach liberty not only in the cognitive
field, but, essentially, in the social and political ones. In the scope of
health, the epistemological constructs of Paulo Freire have pushed
forward(5,10-11) researches that propose to offer health education for
the groups in the community, which is seen as a fruitful path to
make viable new actors in a production of care that is committed
to autonomy and to a critical awareness about reality. RESULTS Regarding the characterization of the professionals who were
interviewed, 46 were female, and the age group varied from 19
to 67 years old, most participants being from 30 to 49 years old. Educational level: most CHAs had complete high school; 42.9% of
nursing auxiliaries had complete higher education; and 77.8% of the
licensed nurses had lato sensu post-graduation courses (Table 1). The data to characterize the professionals were compiled in
an electronic spreadsheet in the software Microsoft Office Excel
10.0 and went through descriptive analysis. The answers to the
triggering questions were transcribed in full, being typed into a
text document in the software Microsoft Office Word 10.0. Table 1 - Socioeconomic characteristics of professionals interviewed,
Maringá, Paraná, Brazil, July-August, 2016 Participants of the study The study counted on the participation of 50 professionals who
worked in the 9 ESFts: 9 nurses, 7 nursing auxiliaries, and 34 community
health agents (CHAs). The participants were chosen intentionally, 2
Rev Bras Enferm. 2021;74(5): e20201196
8
of 2
8
of Dialogic educational practices in the context of child intoxication: an approach based on Paulo Freire
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ, Tacla MTGM, Baldissera VDA, et al. Dialogic educational practices in the context of child intoxication: an approach based on Paulo Freire
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ, Tacla MTGM, Baldissera VDA, et al. since they were the main responsible for health education processes
in the ESFts(15). The study included those who had been working in the
ESF for more than 12 months, regardless of their time working in the
particular ESFt where they worked at the moment of the interview. Professionals who were on leave at the time of data collection were
excluded — be it due to vacations, health leave, maternity leave,
especial leave, or any other type of legal leave. process in the professional practice for the empowering and
autonomy of the family; and Identifying educational activities
in the context of preventing child intoxication. The narratives of
professionals were identified using codes with the initial letters of
the professional categories in Portuguese, followed by numbers
referring to the transcription of the interviews: “Enf1”, “Aux1”, “ACS1”. 1. Emphasis on critical dialog and on activities
that involve the practice of reflection, listening
to the other;
2. The use of the situations experienced by
groups, in the form of dialog;
3. A horizontal participation of professionals
and family;
4. Active subject participation. 1. Valuing the knowledge of family, believing
that their knowledge is based on their
concrete conditions of existence;
2. Interactions that establish bonds and
respect to particularities;
3. Emphasis in the experience and the
building of a knowledge between the
technical and the popular;
4. Individual and/or group empowering
through relations between social subjects. Health education
practices
Promotion of
autonomy in the family
Dialogic relation
Empowering
Criteria
Parameters
Figure 1 - Criteria and parameters for the approach of the practices, Maringá,
Paraná, Brazil, July-August, 2016 Parameters Collection and organization of data 1. Emphasis on critical dialog and on activities
that involve the practice of reflection, listening
to the other; The semistructured narrative interviews were chosen as the
strategy to obtain data(16), and aimed to understand dialogic
educational practices in the context of child intoxication. The
ESFts professionals were invited in person, by the main researcher,
to participate in this study. The interviews were scheduled ac-
cording to the time and place of work of each interviewee and
carried out by the researcher herself, in the PHC units of each
ESFt. They were conducted in private rooms, with as little external
influence as possible, and recorded using digital media, lasting
for a mean of 30 minutes. Data collection took place from July
to August 2016, and the professionals were approached twice. In the second occasion, they were presented with the transcrip-
tion of their narratives, so they could change and validate them. The data collection instrument was a semistructured script,
made up by two moments, each including: the sociodemographic
and professional profile of the interviewees; and the description
of educational practices for the empowering of the families. The
latter contained the following triggering questions: How do you
carry out the teaching-learning process in your health education
practices in the ESF? For you, how should the relation between
educator and learner (professional and patient) be in the process
of health education? What educational activities do you carry out
regarding accidents involving children? Figure 1 - Criteria and parameters for the approach of the practices, Maringá,
Paraná, Brazil, July-August, 2016 Data analysis Variables
Results
Female (%)
92.0
Age (years), mean
42.6
Nursing qualification, lato sensu specialization (%)
77.8
Nursing auxiliary educational level, complete higher
education (%)
42.9
Community health agent educational level, complete
high school (%)
70.6 Strengths and weaknesses for the dialogic relation in the
practice of health education This category shows the strengths and weaknesses of the dialogic
relation in the practice of health education. As the professionals
responded to questions about how the relations between educator
and learner (professional and patient) should be, according to the
parameter of situations experienced by the families, they valued
the knowledge of the family through the exchange of experiences
and, especially, through dialog. Also, the professionals are inserted
in the territory and have a bond with their families, which was seen
as a facilitating factor for a dialogic relation. This process of teaching the patient is very difficult, because they
just learn and listen to what they want; if it isn’t connected to their
consultation, they won’t come. (Enf4) That story of promotion and prevention is still very distant from
the community. They only believe in medical treatments. (ACS4) There are patients who, sometimes, don’t follow treatment or
something like that, so you have to be stern with them. I’m more
the type to blackmail them, to get them afraid. (Aux7) [...] on the same footing as them, not letting them be ashamed of
their questions even when they’re pointless. To exchange experiences,
sometimes. First, he tells me what he needs and what is happening,
and then, we give them some guidance. (Enf3) In these circumstances, it should be highlighted that valuing
the knowledge of the families in the context of their territory,
creating bonds, and having dialog are strengths of the dialogic
process in the practice of health education. However, the lack of
involvement of families in this process is seen as a weakness in
the dialogic relation, transforming the educational practice of
the professionals in a top-to-bottom affair. First and foremost, to respect the people and listen, because our
priority, of us community agents, is to listen to ideas, ways of
thinking, respecting their beliefs and advancing conversations
based on these premises. (ACS28) Variables The narratives were analyzed using the content analysis technique,
thematic modality, following the stages: Pre-analysis, Exploration
of material, and Treatment of results/inference/interpretation(17). The answers were codified and analyzed in a critical-reflective
process(9-10), regarding the educational practices developed and
problematized through the concepts of dialogism and autonomy
that stem from the sociocultural approach of the Brazilian educa-
tor Paulo Freire(9). Directives were established for the discussion of
the findings, based on the theoretical referents and on principals
that should guide educational practices as expressions of care in
public health(9,18), divided in evaluative criteria and parameters of
the approach, as presented in Figure 1. Although the mean professional experience was 10.5 years
old in the field of health and 9.2 years old in family health teams,
only 13 (26%) professionals stated to have participated in train-
ing sessions to start working in the ESFt. However, all of them
stated to have participated in continued education courses every
four months, as they exercise their activities. Only 16% received
training for activities to prevent intoxication. After the data collected were codified, they were organized, and
the narratives in the interviews were discussed in three thematic
categories: Strengths and weaknesses for the dialogic relation in
the practice of health education; Developing the teaching-learning 3
Rev Bras Enferm. 2021;74(5): e20201196
8
of Dialogic educational practices in the context of child intoxication: an approach based on Paulo Freire
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ, Tacla MTGM, Baldissera VDA, et al. did not mention the active participation of the subjects in the
teaching-learning process, and, therefore, show their rigidity in
the education process. Developing the teaching-learning process in the profes-
sional practices for the empowering and autonomy of
the family By exchanging experiences, because patients know very well and
have a lot to teach too. So we always sit in a circle and talk. It’s
not like we’re on foot and they are siting, everyone is sitting down
and on the same level. (Aux7) When the CHAs were asked about how they develop their
teaching-learning process to empower and promote the autonomy
of the family in the context of the ESFt, it was found that they
consider, especially, knowledge, experiences, and the valoriza-
tion of popular knowledge. The interactions in which bonds of
confidence are established and the respect to particularities were
also mentioned by many professionals. The horizontal nature of this educational practice was not
manifested in the statements of all professionals, but was nar-
rated by professionals from all categories. I think that we must be equal, act as if it wasn’t from top to bottom,
but between equals [...] starting from their reality so they can
clarify any doubts. I explain, some subject is raised, we talk, this
is nice. (Enf2) [...] we end up giving recommendations, always considering the
reality of the person we’re advising, because it’s pointless for me
to say “you must buy this” if the person can’t afford it. (ACS12) Clear language, a language they can understand. Sometimes you
talk, and they pretend they understood, but they didn’t understand
anything. So you try a clearer and more popular language, so they
can understand and don’t get frightened. (Aux6) [...] if I treat a patient aggressively and I want to impose myself,
it won’t work, it will break the bond I was trying to keep. (ACS6) It was possible to infer that the use of scientific terms by
professionals is seen as a barrier for the dialog with the families. Therefore, they seek to use a clearer language that is more ac-
cessible by the population. First I need to have a bond with the patient so he can trust me. The rest flows from this bond. It has to be like a relation of real
trust. (ACS30) Mutual bonds and respect. I think that I have to gain the trust of
this patient. He needs to trust me as a professional, to trust me
as a person. (Enf9) Sometimes, we see the physician speaking with a little more theory,
and the patients don’t understand. Developing the teaching-learning process in the profes-
sional practices for the empowering and autonomy of
the family We come back later, and they
say they don’t understand. Sometimes, they don’t tell the physician,
they get embarrassed, but to us, the CHAs, they say it. (ACS14) There are some patients we adopted around here. You create such a
strong bond [...] we even have the number of their relatives. (Aux3) The exchange of knowledge must be as natural as possible, transmit
it as accessibly as possible to the population, because the patient
frequently can’t understand the technical terms, scientific terms,
in a conversation and they don’t know what to do for their treat-
ment, for instance. (Enf6) However, some practices mentioned by the professionals during
the teaching-learning process did not establish the construction
of knowledge between the technical and popular knowledge
and were punitive towards the families. Although the statements indicate a valuing of the knowl-
edge for the process of health education, most professionals I have to scare some of them into adhering. I don’t know if it’s
right, but it’s what I do, by explaining the risks “if you don’t do 4
Rev Bras Enferm. 2021;74(5): e20201196
8
of 4
8
of Dialogic educational practices in the context of child intoxication: an approach based on Paulo Freire
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ, Tacla MTGM, Baldissera VDA, et al. this, that will happen”, in the clearest, most direct and upfront
way possible. (Enf2) and nursing homes, can be used as spaces to train the families in
the prevention of intoxications and safety at home. this, that will happen”, in the clearest, most direct and upfront
way possible. (Enf2) So we kind of go to their weakest spot: their fear of dying. (ACS7) I think the schools and nursing homes should organize for the
teachers to call families, or we should organize this with the
schools. This would be a very good way, because the approach
would address bot the kids and their parents. (Enf8) I’ll give you an example. Sometimes we even make up examples we
don’t have, just to make the person afraid so they won’t do it. (ACS4) If you do this at the school, the public will be greater and both
mother and father will be present, or grandmother and grandfa-
ther, and even an older brother who brings the kid home. (ACS12) Some professionals stated not to promote the strengthening
of individuals and/or groups through the relations between social
subjects. DISCUSSION The sociodemographic characteristics of the professionals
interviewed influence the variables of the teaching-learning
process in the context of child intoxication, considering the
references used for this study(9), since they integrate a team of
professionals inserted in the community, who have established
a bond with it throughout the years. To this end, it was found
that these characteristics corroborate data on the composition
of the Brazilian health workforce: most are female, young, with
a higher educational level than required for their work, not to
mention a high number of professionals with lato sensu post-
graduation courses(3,19). Developing the teaching-learning process in the profes-
sional practices for the empowering and autonomy of
the family However, these professionals seem to be in a transition
stage of their educational practices, as shown in some statements,
especially by nurses. Paradoxically, the workers mentioned the families themselves
as a barrier for the education activities to promote health and
prevent child intoxication. We’ve entered the houses of these people and we know how they’re
inside. We do our part, which is giving advice. If they’ll follow it, if
they’ll do it, that’s their problem. (ACS9) Normally, the population comes here to the unit and wants a cure,
they don’t care much about preventing, about a healthy life. (Enf7) We’re working and trying to change this point of view focused on
curing and aiming towards promotion and prevention; actually,
especially towards health promotion. (Enf1) It’s really hard for them to participate, because they don’t adhere
to this type of thing; unless they get some benefit, they gain some-
thing, otherwise no one really shows interest. (ACS25) We work by building a work plan and demystifying everything
that is going on in the community. The community is where
the health education process takes place, and we carry out our
activities according to the openings it shows for the educational
process, because you can’t stop the work dynamic, it’s a continuous
process of learning. (Enf8) Prevention, for them, doesn’t exist, they always just want the
medication [...] it would be much more interesting if it was really
just family health. (ACS32) The professionals show that, although families have the poten-
tial to prevent child intoxications, they sometimes limit it when
they avoid active participation and do not become involved in
education activities, seeking only a type of assistance based on
the biomedical model. Considering this setting, it stands out that the interaction of
the ESFt for promoting family autonomy is developed through
bonds of trust, respect, and valorization of the community
knowledge. However, it is still necessary to rethink the domain
of technical knowledge, and to reconsider the strengthening of
the relations that exist between social subjects and the commu-
nication between them, including their opening and availability
for dialogic relations, without any punishment for the families. Identifying educational activities in the context of prevent-
ing child intoxication This category represents the conditions that cause the phenom-
enon under study. When asked about the activities they develop
regarding child intoxication, most professionals responded that
they develop “no activity” or few activities, but they believe the
families could be trained to prevent intoxications in their homes. It was necessary to organize the understanding of educa-
tional practices developed and problematized by the concepts
of dialogism and autonomy, as divided in evaluative criteria and
parameters(9) (practices in health education and promotion of fam-
ily autonomy), to comprehend the multiple factors that affected
and determined the relations of empowerment and dialog of
the families, required for child intoxications could be dealt with. I’ll only give advice if I see something, but no, I don’t address
something specific. (ACS34) It has to be addressed in their homes, in the home visits [...], in their
home, we see what is happening, how things are. Sometimes it’s
not even the mother who takes care of the child. Sometimes it’s
the grandmother, so I think this has to be at their home. (Enf1) Regarding the conceptual framework for the dialogic relations
in the practice of health education, which was presented in the first
thematic category, the participants and the users share, according
to some statements, their knowledge. This is done through active
listening, dialog, the exchange of experiences, through teaching
and learning, and showing the change in perspective that avoids a I think the easiest way is the home visit, really, because to convince them
to get out of their homes and come participate is very hard. (ACS25) In addition to the space of their homes, many narrated that
other environments that favor social interaction, such as schools 5
Rev Bras Enferm. 2021;74(5): e20201196
8
of Dialogic educational practices in the context of child intoxication: an approach based on Paulo Freire
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ, Tacla MTGM, Baldissera VDA, et al. technical knowledge, popular knowledge, and personal and/or
group improvement through the relations of trust between pro-
fessionals and family, the interviewees were seen to understand
Paulo Freire’s presuppositions(9), but do not seem to use them
in practice. Identifying educational activities in the context of prevent-
ing child intoxication There is an observable trend towards the effective
employment of the dialogic model, but the community, as well as
the workers in the field of health, must change the predominant
point of view according to which health is the absence of disease,
also changing actions based on the transmission of information
that relies on emphasizing the possibility of becoming ill(21-22). reductionist view of the health-disease binomial. Health education,
carried out according to Paulo Freire’s guidance, makes it possible
to strengthen the bonds and the trust between health services and
users with regard to spaces of dialog and construction of knowledge
and practices. In these practices, users and professionals discuss
strategies to confront their problems in group, to propose alterna-
tives and realize effective health practices that can be integral and
capable of solving the problems(5,20-21). It stands out that the knowledge
and practices must be understood by the health professionals and
respected, valued, and treated horizontally. When one considers the
words and the knowledge of their users, there is a higher chance of
educating, of improving behavior. One must respect and consider
the culture of people and promote the empowering of the action
and reflection, so there is a change in behavior(9,20-21). In the same vein, the activities developed by the workers in this
study are, still, focused on disease and on the use of traditional
methodologies, such as educational practices focused on the
prevention of diseases and health problems. Although activities
focused on the prevention of diseases and health problems are
important, considering the epidemiological changes in health,
educational practices targeted at health promotion, which include
the family, express the opportunity of providing an integral care
to health, considering the complexity and interaction between
users, families, professionals, and society(25). These educational
practices are directed at the transformation of the behavior of
people, so they can act in the improvement of their quality of
life and health. This includes a stronger participation in society,
so they can have control over this process(26). However, narratives were weak with regard to a dialogic
relationship with the families in practice and the use of method-
ologies to empower and give autonomy to them. This could be
observed in the statements about the fact that subjects did not
participate in the process, and in those about the authoritarian-
ism of the professionals. Identifying educational activities in the context of prevent-
ing child intoxication It can be understood, as a result, that
the educational practices that are dialogic and, consequently,
emancipatory, can be less frequent among the professionals due
to their own personal experiences and professional formation,
which can be distant from reality(22). Regarding the process of dialogic relations with the families,
the fact that some professionals live in the territory of the ESFt was
pointed out as an instrument that facilitates the establishment of
bonds. When the workers take action regarding health promotion
and the prevention of child intoxications, they use different technolo-
gies, and belonging to the same community as the family who is
being cared for is an element that connects them, a “link” between
the population and of the territory and the team as a whole(23). Professionals mentioned little involvement in activities to
prevent child intoxication and were unaware of the reality of the
families concerning intoxication, as the third thematic category
showed. Considering these premises, child intoxication has been
discussed as an event where the values and beliefs of a family are
important causal factors — the existence of toxic plants around
the houses to “protect the family”, the use of highly toxic cleaning
products to diminish the personal effort spent in cleaning, the
self-medication and storage of drugs at home(4,27). Considering the active listening and critical dialog parameters, the
horizontality of professional and family, and the active participation
of people and groups in the teaching-learning process, the group
interviewed presented multiple and occasionally contradictory
narratives, with regard to how they believed the relation between
educator and learner should take place. The use of devices from
the traditional model of education in health repeated itself in the
narratives, especially as the workers were rigid in the process of
educating the families and did not bring them into this process
active members of it. Bringing into effect educational practices
that are actually committed to the emancipation of the subjects
is a challenge that must be faced in the process of work of the
interviewees, since this type of education is still rare. Traditional
methodologies which compromise the creation of a bond between
the professionals and the population continue to be used, especially
with regard to blaming the family for toxicological events(21-22). FINAL CONSIDERATIONS The strategies used in the educational practices to prevent
against child intoxication were focused on traditional health
models of education, without the active participation of the
families, although professionals stated that they carried out
collective practices, targeted at increasing the autonomy and
wellbeing of the same families. This suggests that they are in a
transition stage between the curative/biomedical model and
the dialogic one. Identifying educational activities in the context of prevent-
ing child intoxication Dialogic educational practices in the context of child intoxication: an approach based on Paulo Freire
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ, Tacla MTGM, Baldissera VDA, et al. against the curative model, emphasizing the principles of inte-
gral and equal health, based on the presuppositions of Paulo
Freire (dialog, participation, and autonomy) as a possible way
to reorganize education and care-related practices of health(9). In this context, the narratives that mentioned the families them-
selves as barriers for educational activities were opposed to the
discourse for the transition of educational practices in general. In
other words, there is a theoretical and general approximation to the
perspective of Paulo Freire, while the distance from them is greater
in the specificities of daily practice. Educational interventions should
be constructed in society (family, school, health units, churches) in
a participative and co-responsible manner, with multisectoral and
universal interventions. The communities and the participation of
the community must be the goal, involving environmental changes,
as well as the implantation of public policies and legislation(27,29). against the curative model, emphasizing the principles of inte-
gral and equal health, based on the presuppositions of Paulo
Freire (dialog, participation, and autonomy) as a possible way
to reorganize education and care-related practices of health(9). practice of nursing, as it targets the attention of the health needs
and is based on a teaching-learning process which includes strate-
gies that allow for the active participation of subjects capable of
forming competent ESFts professionals, with the necessary skills
to generate changes in the practices of education in health to
deal with child intoxication. In this context, the narratives that mentioned the families them-
selves as barriers for educational activities were opposed to the
discourse for the transition of educational practices in general. In
other words, there is a theoretical and general approximation to the
perspective of Paulo Freire, while the distance from them is greater
in the specificities of daily practice. Educational interventions should
be constructed in society (family, school, health units, churches) in
a participative and co-responsible manner, with multisectoral and
universal interventions. The communities and the participation of
the community must be the goal, involving environmental changes,
as well as the implantation of public policies and legislation(27,29). Contributions to the fields of Nursing, Health or Public Policy This work was carried out with the support of the Coordina-
tion for the Improvement of Higher Education Personnel - Brazil
(CAPES) - Financing Code 001, from March/2015 to February/2017. The dialogic action, as a strategy to empower the families,
stands out as a contribution of this research for the professional Study limitations Dialogic educational practices about child intoxication, which
is a complex and multifaceted phenomenon, must recognize
homes and social spaces as priorities for their practice. However,
the discourse that blames the families for intoxications must be
pushed away from the educational process, and dialog must be
more frequent for the promotion of health and the prevention of
intoxication, in a way that targets the social reality of these families. The context in which the research was carried out can be seen
as its limitation, since this is a study involving the ESFts of a city
with its own specific characteristics, although the interviewees
shared professional and social characteristics that are true for the
whole country. For many reasons, the professionals presented
care for child intoxication as something dispensable, which is not
their responsibility. REFERENCES 1. Wynn PM, Zou K, Young B, Majsak-Newman G, Hawkins A, Kay B, et al. Prevention of childhood poisoning in the home: overview of systematic
reviews and a systematic review of primary studies. Int J Inj Contr Saf Promot. 2016;23:3-28. https://doi.org/10.1080/17457300.2015.1032978 2. Disfani HF, Kamandi M, Mousavi SM, Sadrzadeh SM, Farzaneh R, Doolabi N, Rahmani K. Risk factors contributing to the incidence and
mortality of acute childhood poisoning in emergency department patients in Iran: a hospital-based case-control study. Epidemiol Health. 2019;41:e2019016. https://doi.org/10.4178/epih.e2019016 2. Disfani HF, Kamandi M, Mousavi SM, Sadrzadeh SM, Farzaneh R, Doolabi N, Rahmani K. Risk factors contributing to the incidence and
mortality of acute childhood poisoning in emergency department patients in Iran: a hospital-based case-control study. Epidemiol Health. 2019;41:e2019016. https://doi.org/10.4178/epih.e2019016 3. Sales CCFS, Oliveira MLF. Health education practices of poisoning prevention for child in Family Health Strategy. Esc Anna Nery. 2019;23(1):e20180140. https://doi.org/10.1590/2177-9465-EAN-2018-0140 4. Vilaça L, Volpe FM, Ladeira RM. Accidental poisoning in children and adolescents admitted to a referral toxicology department of a Brazilian
emergency hospital. Rev Paul Pediatr. 2020;38:e2018096. https://doi.org/10.1590/1984-0462/2020/38/2018096 5. Brandão Neto W, Silva MAI, Aquino JM, Lima LS, Monteiro EMLM. Violence in the eye of adolescents: education intervention with Culture
Circles. Rev Bras Enferm. 2015;68(4):617-25. https://doi.org/10.1590/0034-7167.2015680407i 6. Roma KMS, Gomes MFP, Reticena K, Capellini VK, Fracolli LA. Prevenção de acidentes na primeira infância na estratégia saúde da família:
perspectiva dos pais. Rev Enferm UFPI [Internet]. 2018 [cited 2020 Oct 23];7(2):28-34. Available from: https://revistas.ufpi.br/index.php/
reufpi/article/view/6670/pdf 7. Amestoy SC, Oliveira AFL, Thofehrn MB, Trindade LL, Santos BP, Bao ACP. Contributions of Paulo Freire to understanding the dialogic leadership
exercise of nurses in the hospital setting. Rev Gaúcha Enferm. 2017;38(1):e64764. https://doi.org/10.1590/1983- 1447.2017.01.64764 7. Amestoy SC, Oliveira AFL, Thofehrn MB, Trindade LL, Santos BP, Bao ACP. Contributions of Paulo Freire to understanding the dialogic leadership
exercise of nurses in the hospital setting. Rev Gaúcha Enferm. 2017;38(1):e64764. https://doi.org/10.1590/1983- 1447.2017.01.64764 g
g
8. Lee BS, Eo YS, Lee MA. Leadership experience of clinical nurses: applying focus group interviews. J Korean Acad Nurs. 2015;45(5):671-83. https://doi.org/10.4040/jkan.2015.45.5.671 8. Lee BS, Eo YS, Lee MA. Leadership experience of clinical nurses: applying focus group interviews. J Korean Acad Nurs. 2015;45(5):671-83. https://doi.org/10.4040/jkan.2015.45.5.671 9. Freire P. Pedagogia da autonomia: saberes necessários à prática educativa. 53. ed. São Paulo: Paz e Terra; 2016. 10. Labegalini CMG, Nogueira IS, Hammerschmidt KSA, Jaques AE, Carreira L, Baldissera VDA. Identifying educational activities in the context of prevent-
ing child intoxication The educational practices for this event should be based on
an approach that not only demands more attention from adult
caregivers when children are at risk of accidents, but also discusses
the socioenvironmental determinants and the vulnerability of the
families(28). Considering Paulo Freire’s presuppositions, they should
include the “3 Es” — Education, Engineering, and Enforcement,
to which Economy and Environment can be added(29). The professionals interviewed considered that families can
be addressed to prevent intoxication in their houses, that is,
according to their social reality, and recognize other spaces, in
addition to the home spaces, to interact with them in the pre-
vention of intoxication. If these activities are only articulated as
norms, they become general conducts and procedures to solve
health problems, but do not guarantee a type of care that can
respond to the needs of children and to the ways of life of their
families. If an educational practice is to be, simultaneously, a care
practice, with the power to transform health, it must be reason-
able, participative, non-coercive or punitive, and focused on the
family, as opposed to revolving around aspects of its inability to
promote safety measures in the home environment(9,28). Regarding the use of methodologies to guide the teaching-
learning process that empowers and promotes the autonomy
of the families within ESF practice, there has been a theoretical
transition of educational practices, as the second thematic cat-
egory shows. The change is from a reductive perspective about
the health-disease binomial into a perspective that involves
the exchange of knowledge as the main condition for a trans-
formative education(24). Considering how the knowledge and
experiences of the family are valued, the integration between To develop educational activities in the context of child intoxica-
tion, the social, economic, and cultural contexts of the individuals
involved in the educational process must be considered(30). One
must seek a humane practice, that privileges health promotion 6
Rev Bras Enferm. 2021;74(5): e20201196
8
of Dialogic educational practices in the context of child intoxication: an approach based on Paulo Freire
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ, Tacla MTGM, Baldissera VDA, et al. Dialogic educational practices in the context of child intoxication: an approach based on Paulo Freire
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ, Tacla MTGM, Baldissera VDA, et al. REFERENCES Educational and care-related dialogical
pathway on active aging with family health strategy professionals. Texto Contexto Enferm. 2020;29:e20180235. https://doi. org/10.1590/1980-265X-TCE2-2018-0235 7
Rev Bras Enferm. 2021;74(5): e20201196
8
of Dialogic educational practices in the context of child intoxication: an approach based on Paulo Freire
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ, Tacla MTGM, Baldissera VDA, et al. Dialogic educational practices in the context of child intoxication: an approach based on Paulo Freire
Ribeiro CCFS, Tonon MM, Oliveira JY, Guedes MRJ, Tacla MTGM, Baldissera VDA, et al. 11. Rodrigues DMMR, Nogueira IS, Higarashi IH, Heidemann ITSB, Baldissera VDA. Critical disclosures about sexuality among elderly women as a
dialogical educational assessment device. Rev Baiana Enferm. 2019;33:e27754. https://doi.org/10.18471/rbe.v33.27754 11. Rodrigues DMMR, Nogueira IS, Higarashi IH, Heidemann ITSB, Baldissera VDA. Critical disclosures about sexuality among elderly women as a
dialogical educational assessment device. Rev Baiana Enferm. 2019;33:e27754. https://doi.org/10.18471/rbe.v33.27754 12. Dignam D, Duffield C, Stasa H, Gray J, Jackson D, Daly J. Management and leadership in nursing: an Australian educational perspective. J
Nurs Manag. 2012;20(1):65-71. https://doi.org/10.1111/j.1365-2834.2011.01340.x 12. Dignam D, Duffield C, Stasa H, Gray J, Jackson D, Daly J. Management and leadership in nursing: an Australian educational perspective. J
Nurs Manag. 2012;20(1):65-71. https://doi.org/10.1111/j.1365-2834.2011.01340.x 13. Gil AC. Como elaborar projetos de pesquisa.6ª ed. São Paulo: Atlas; 2017. 14. Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for
groups. Int J Qual Health Care. 2007;19(6):349–57. https://doi.org/10.1093/intqhc/mzm042 15. Ramos CFV, Silva MSB, Rosa AS, Santana CLA, Tanaka LH. Educational actions: an action research with Family Health Strategy professionals
and users. Rev Bras Enferm. 2020;73(5):e20180969. https://doi.org/10.1590/0034-7167-2018-0969 15. Ramos CFV, Silva MSB, Rosa AS, Santana CLA, Tanaka LH. Educational actions: an action research with Family Health
and users. Rev Bras Enferm. 2020;73(5):e20180969. https://doi.org/10.1590/0034-7167-2018-0969 16. Castellanos METP. The narrative in qualitative research in health. Ciênc Saúde Colet. 2014;19(4):1065-76. https://doi
org/10.1590/1413-81232014194.12052013 17. Minayo MCS. O desafio do conhecimento: pesquisa qualitativa em saúde. 14ª ed. São Paulo: Editora Hucitec; 2015. 18. Ministério da Saúde (BR). Secretaria de Vigilância em Saúde. Secretaria de Atenção à Saúde. Política Nacional de Promoção da Saúde. Brasília
(DF): Ministério da Saúde; 2015. 19. Machado MH, Aguiar Filho W, Lacerda WF, Oliveira E, Lemos W, Wermelinger M, et al. Características gerais da enfermagem: o perfil sócio
demográfico. Enfermagem em foco [Internet]. 2016 [cited 2020 Oct 23];3(Esp):9-14. Available from: http://revista.portalcofen.gov.br/index. php/enfermagem/article/view/686/296 20. Rodrigues DMMR, Labegalini CMG, Higarashi IH, Heidemann ITSB, Baldissera VDA. REFERENCES The dialogic educational pathway as a strategy of care
with elderly women in sexuality. Esc Anna Nery 2018;22(3):e20170388. https://doi.org/10.1590/2177-9465-EAN-2017-0388 21. Soares NA, Souza V, Santos FBO, Carneiro ACL, Gazzinelli MF. Dispositivo educação em saúde: reflexões sobre práticas educativas na Atenção
Primária e formação em enfermagem. Texto Contexto Enferm. 2017;26(3):e0260016. https://doi.org/10.1590/0104-07072017000260016 22. Porto LA, Valente RN, Esteves CO, Escalda J. Práticas de aleitamento materno: concepções das trabalhadoras de saúde de um hospital
público de Salvador. Rev Baiana Saúde Pública [Internet]. 2018 [cited 2020 Oct 23];42(4):653-70. Available from: http://www.rbsp.sesab. ba.gov.br/index.php/rbsp/article/view/2828/2568 23. Assis AS, Castro-Silva CR. Agente comunitário de saúde e o idoso: visita domiciliar e práticas de cuidado. Physis: Rev Saúde Coletiva
[Internet]. 2018 [cited 2020 Oct 23];28(3):e280308. Available from: https://www.scielosp.org/pdf/physis/2018.v28n3/e280308/pt 24. Viaro VD, Linhares FMP, Coriolano-Marinus MWL, Guedes TG, Vanderley LSL. Limits and possibilities for teaching and learning about
breastfeeding. Rev Bras Enferm. 2019;72(1):3-8. https://doi.org/10.1590/0034-7167-2018-0017 24. Viaro VD, Linhares FMP, Coriolano-Marinus MWL, Guedes TG, Vanderley LSL. Limits and possibilities for teaching and learning about
breastfeeding. Rev Bras Enferm. 2019;72(1):3-8. https://doi.org/10.1590/0034-7167-2018-0017
25. Fagundes DQ, Oliveira AE. Educação em saúde no pré-natal a partir do referencial teórico de Paulo Freire. Trab Educ Saúde. 2017;15(1):223-
43 https //doi org/10 1590/1981 7746 sol00047 25. Fagundes DQ, Oliveira AE. Educação em saúde no pré-natal a partir do referencial teórico de Paulo Freire. Trab Educ
43. https://doi.org/10.1590/1981-7746-sol00047 26. Santos FPA, Acioli S, Rodrigues VP, Machado JC, Souza MS, Couto TA. Nurse care practices in the Family Health Strate
2016;69(6):1060-7. https://doi.org/10.1590/0034-7167-2016-0273 27. Uthkarsh PS, Suryanarayana SP, Gautham MS, Shivraj NS, Murthy NS, Pruthvish S. Profile of injury cases admitted to a tertiary level hospital
in south India. Int J Inj Contr Saf Promot. 2012;19(1):47-51. https://doi.org/10.1080/17457300.2011.603149 28. Nascimento EN, Gimeniz-Paschoal SR, Sebastião LT. Prevention of domestic child accidents: an educational intervention conducted by
Speech Therapy trainees in a Family Health Care Unit. Rev CEFAC. 2019;21(5):e17018. https://doi.org/10.1590/1982-0216/201921517018 Speech Therapy trainees in a Family Health Care Unit. Rev CEFAC. 2019;21(5):e17018. https://doi.org/10.1590/1982-0216/201921517018
29. Patel D, Sandell JM. Prevention of unintentional injury in children. Paediatrics Child Health. 2013;23(9):402-8. https://doi.org/10.1016/j. paed.2013.06.001 29. Patel D, Sandell JM. Prevention of unintentional injury in children. Paediatrics Child Health. 2013;23(9):402-8. https://doi.org/10.1016/j. paed.2013.06.001 30. Branquinho ID, Lanza FM. Child health in primary care: evolution of brazilian policies and nurses’ performance. Rev Enferm C-Oeste Min. 2018;8:e2753. https://doi.org/10.19175/recom.v8i0.2753 30. Branquinho ID, Lanza FM. Child health in primary care: evolution of brazilian policies and nurses’ performance. Rev Enferm C-Oeste Min. 2018;8:e2753. REFERENCES https://doi.org/10.19175/recom.v8i0.2753 8
Rev Bras Enferm. 2021;74(5): e20201196
8
of 8
Rev Bras Enferm. 2021;74(5): e20201196
8
of
|
https://openalex.org/W2350766159
|
https://jelectrochem.xmu.edu.cn/cgi/viewcontent.cgi?article=1437&context=journal
|
Chinese
| null |
Study of Hydrogen Diffusion Behavior in Hydrogen Storage Alloy MlNi_(3.75)Co_(0.65) Mn_(0.4) Al_(0.2) with Potential Step Method
|
Deleted Journal
| 2,001
|
cc-by
| 2,841
|
Journal of Electrochemistry
Journal of Electrochemistry Journal of Electrochemistry
Journal of Electrochemistry Volume 7
Issue 4 Recommended Citation
Recommended Citation
Xian xia YUAN, Nai xin XU. Study of Hydrogen Diffusion Behavior in Hydrogen Storage Alloy
MlNi_(3.75)Co_(0.65) Mn_(0.4) Al_(0.2) with Potential Step Method[J]. Journal of Electrochemistry, 2001 ,
7(4): 445 451 Study of Hydrogen Diffusion Behavior in Hydrogen Storage Alloy
Study of Hydrogen Diffusion Behavior in Hydrogen Storage Alloy
MlNi_(3.75)Co_(0.65) Mn_(0.4) Al_(0.2) with Potential Step
MlNi_(3.75)Co_(0.65) Mn_(0.4) Al_(0.2) with Potential Step
Method
Method Recommended Citation
Recommended Citation
Xian xia YUAN, Nai xin XU. Study of Hydrogen Diffusion Behavior in Hydrogen Storage Alloy
MlNi_(3.75)Co_(0.65) Mn_(0.4) Al_(0.2) with Potential Step Method[J]. Journal of Electrochemistry, 2001 ,
7(4): 445-451. DOI: 10.61558/2993-074X.1437
Available at: https://jelectrochem.xmu.edu.cn/journal/vol7/iss4/5 Recommended Citation
Recommended Citation
Xian xia YUAN, Nai xin XU. Study of Hydrogen Diffusion Behavior in Hydrogen Storage Alloy
MlNi_(3.75)Co_(0.65) Mn_(0.4) Al_(0.2) with Potential Step Method[J]. Journal of Electrochemistry, 2001 ,
7(4): 445-451. Journal of Electrochemistry
Journal of Electrochemistry Journal of Electrochemistry
Journal of Electrochemistry Recommended Citation
Recommended Citation Recommended Citation
Recommended Citation Available at: https://jelectrochem.xmu.edu.cn/journal/vol7/iss4/5 Available at: https://jelectrochem.xmu.edu.cn/journal/vol7/iss4/5 This Article is brought to you for free and open access by Journal of Electrochemistry. It has been accepted for
inclusion in Journal of Electrochemistry by an authorized editor of Journal of Electrochemistry. 电化学
EL ECTROCHEMISTRY 电化学 第7 卷 第4 期
2001 年11 月 Vol. 7 No. 4
Nov. 2001 Vol. 7 No. 4
Nov. 2001 EL ECTROCHEMISTRY 电极,辅助电极为容量远大于贮氢合金电极的烧结镍电极,Hg/ HgO (6 mol/ L KOH) 电极作参
比电极,6 mol/ L KOH 溶液作电解液. 实验所用仪器为美国EG&G公司的M273A 恒电位/ 恒
电流仪,所用软件为该公司的M270. 电极,辅助电极为容量远大于贮氢合金电极的烧结镍电极,Hg/ HgO (6 mol/ L KOH) 电极作参
比电极,6 mol/ L KOH 溶液作电解液. 实验所用仪器为美国EG&G公司的M273A 恒电位/ 恒
电流仪,所用软件为该公司的M270. 经过活化的贮氢合金电极在室温(～25 ℃) 下以60 mA/ g 的电流充电7. 5 h ,搁置10 min
后以相同的电流放电,经一定的时间达到相应的放电深度DOD (定义见式(1) ) ,稳定2 h 后再
将电位阶跃一定的幅值并记录阶跃后的响应电流随时间的变化. 或者,活化后的贮氢合金电极
先在室温下以60 mA/ g 的电流充放电并达到50 %DOD ,然后再转移到相应的测试温度,待其
电位稳定后再阶跃一定的幅值并记录响应电流随时间的变化. DOD = id td
Q × 100 %
式
中
为贮氢合金 (1) 式(1) 中, Q 为贮氢合金电极活化后达到的稳定容量(C) , id 和td 分别为放电电流(A) 和
放电时间(s) . 文章编号
恒电位阶跃法研究贮氢合金
3. 75Co0. 65Mn0. 4Al0. 2中氢的扩散行为
原鲜霞3 徐乃欣 原鲜霞3 ,徐乃欣
(中国科学院上海冶金研究所, 上海200050) 摘要: 本文用恒电位阶跃法研究了不同放电深度(DOD) 和不同温度下贮氢合金MlNi3. 75Co0. 65
Mn0. 4Al0. 2中氢的扩散行为. 结果表明:室温下该合金中氢的扩散系数随DOD 的增大而增大,在
50 %DOD 的该合金中,氢的扩散系数随温度的升高而增大,扩散活化能为19. 87 kJ/ mol. 关键词: 贮氢合金;氢的扩散系数;恒电位阶跃法
中图分类号: O 646 文献标识码: A 文献标识码: A MH/ Ni 电池由于具有高比能量、可高倍率充放电、耐过充放电能力强、循环寿命长以及无
记忆效应、无污染、免维护、使用安全等特点而引起了人们的极大兴趣. 国内外学者在有关方面
已经做了大量的研究工作,普遍认为,MH/ Ni 电池中贮氢合金电极的性能不仅与电极/ 溶液界
面电化学动力学有关,而且在很大程度上受贮氢合金中氢扩散行为的影响[1 ]. 表征扩散行为
的一个重要参数是扩散系数[2 ] ,要深入研究贮氢合金电极以及MH/ Ni 电池免不了研究贮氢
合金中氢的扩散系数. 本文应用恒电位阶跃法研究了贮氢合金
中氢的扩散系数随其放 合金中氢的扩散系数
本文应用恒电位阶跃法研究了贮氢合金MlNi3. 75Co0. 65Mn0. 4Al0. 2中氢的扩散系数随其放
电深度及环境温度的变化,并用测得的数据估算了在50 %DOD 的该合金中,氢扩散的活化能. 1 实 验
MlNi3. 75Co0. 65Mn0. 4Al0. 2贮氢合金用纯度不低于99. 9 %的富镧混合稀土和Ni、Co 、Mn、Al
等金属按比例混合后于中频真空感应炉中熔炼制得. 为了保证合金的均匀性,熔炼过程中翻身
重熔5 次. 铸锭经机械破碎后用目数相近的两个筛子筛分,所得贮氢合金粉末,用英国Malvern
公司的激光粒度仪Mastersizer2000 测得其平均粒径(直径) 为125. 2μ m. 取上述合金粉0. 3 g
与纯镍粉( T255 ,加拿大INCO 公司) 1. 2 g 混匀后在1. 8 × 108 Pa 下压到泡沫镍的两侧,制得
直径15 mm、厚1. 68 mm 的贮氢合金电极. 1 实 验
MlNi3. 75Co0. 65Mn0. 4Al0. 2贮氢合金用纯度不低于99. 9 %的富镧混合稀土和Ni、Co 、Mn、Al
等金属按比例混合后于中频真空感应炉中熔炼制得. 为了保证合金的均匀性,熔炼过程中翻身
重熔5 次. 铸锭经机械破碎后用目数相近的两个筛子筛分,所得贮氢合金粉末,用英国Malvern
公司的激光粒度仪Mastersizer2000 测得其平均粒径(直径) 为125. 2μ m. 取上述合金粉0. 3 g
与纯镍粉( T255 ,加拿大INCO 公司) 1. 2 g 混匀后在1. 8 × 108 Pa 下压到泡沫镍的两侧,制得
直径15 mm、厚1. 68 mm 的贮氢合金电极. 电化学实验在三电极电解池中进行. 其中,MlNi3. 75Co0. 65Mn0. 4Al0. 2贮氢合金电极为工作 收稿日期:2000-12-25 , 修订日期: 2001-03-30 电 化 学 2001 年 ·
6
4
4
· 2. 1 电位阶跃幅值的选择 文献中用于测定贮氢合金中氢扩散系数的恒电位阶跃法,其电位阶跃的幅值选择各不相
同[2～7 ] ,Yang2Li 等[8 ]特别强调了阶跃幅值选择的重要性. 他们认为,在保证电极表面的电化
学反应足够快以至于扩散成为整个电极过程的控制步骤前提下,较小的电位阶跃幅值有利于
扩散系数的准确求解,过大的阶跃幅值不仅使被测系统的状态发生较大的偏移而且有可能导
致副反应发生,但如果阶跃幅值过小则可
能使响应电流很快降低到噪音水平. 所有
这些都会导致实验结果不准确性. 为此,本
文首先比较了不同的电位阶跃幅值对放电
态(100 %DOD) 贮氢合金电极响应电流的
影响. 文献中用于测定贮氢合金中氢扩散系数的恒电位阶跃法,其电位阶跃的幅值选择各不相
同[2～7 ] ,Yang2Li 等[8 ]特别强调了阶跃幅值选择的重要性. 他们认为,在保证电极表面的电化
学反应足够快以至于扩散成为整个电极过程的控制步骤前提下,较小的电位阶跃幅值有利于
扩散系数的准确求解,过大的阶跃幅值不仅使被测系统的状态发生较大的偏移而且有可能导 图1 分别示出,放电态贮氢合金电极
的电位分别阶跃+ 10 mV 、+ 20 mV 、+ 50
mV 、+ 80 mV 和+ 100 mV 后的电流2时间
响应曲线. 可以看出,当阶跃幅值为+ 10
mV 或+ 20 mV 时,阶跃后的响应电流很快
降低到噪音的水平;当阶跃幅值为+ 80 mV
或+ 100 mV 时,响应电流几乎相等,说明
此时扩散到电极表面的氢全部被氧化,且
电极表面氢的浓度为0 ;阶跃幅值为+ 50
mV ,响应电流比较适中,此际,从电极内部
扩散过来的氢虽在电极表面全被氧化,但 图1 100 %DOD 的MlNi3. 75Co0. 65Mn0. 4Al0. 2贮氢合金
电极在电位阶跃不同幅值后的电流2时间响应
曲线 图1 100 %DOD 的MlNi3. 75Co0. 65Mn0. 4Al0. 2贮氢合金
电极在电位阶跃不同幅值后的电流2时间响应
曲线 Fig. 1 Current2time responses of MlNi3. 75 Co0. 65 Mn0. 4
Al0. 2
hydrogen
storage
alloy
electrode
with
100 %DOD after application of potential step
with various amplitudes Fig. 1 Current2time responses of MlNi3. 75 Co0. 65 Mn0. 4
Al0. 2
hydrogen
storage
alloy
electrode
with
100 %DOD after application of potential step
with various amplitudes ·
7
4
4
·
第4 期 原鲜霞等:恒电位阶跃法研究贮氢合金MlNi3. 75Co0. 65Mn0. 4Al0. 2中氢的扩散行为 图2 100 % DOD 的MlNi3. 75 Co0. 65 Mn0. 4Al0. 2贮氢合
金电极在恒电位阶跃初期的电流2时间响应线
电极表面氢的浓度并不为0. 本文研究即以
+ 50 mV 作为电位阶跃幅值. 2. 2 扩散系数计算公式的比较
假设贮氢合金颗粒为大小均一的圆
球,则其电极在恒电位阶跃初期和阶跃后
期的响应电流可以分别表示为[2 ] :
阶跃初期: i = - nFSD ( Co - Cs)
× [
1
π Dt
+ 1
r ] (2)
阶跃后期:log i = log[ ± 6 FD
dr2 ( Co -
Cs) ] -
π 2 D
2. 303 r2 t
(3) 电极表面氢的浓度并不为0. 本文研究即以
+ 50 mV 作为电位阶跃幅值. 2. 2 扩散系数计算公式的比较
假设贮氢合金颗粒为大小均一的圆
球,则其电极在恒电位阶跃初期和阶跃后
期的响应电流可以分别表示为[2 ] :
阶跃初期: i = - nFSD ( Co - Cs)
× [
1
π Dt
+ 1
r ] (2)
阶跃后期:log i = log[ ± 6 FD
dr2 ( Co -
Cs) ] -
π 2 D
2. 303 r2 t
(3) 期的响应电流可以分别表示为
:
阶跃初期: i = - nFSD ( Co - Cs)
× [
1
π Dt
+ 1
r ] (2)
阶跃后期:log i = log[ ± 6 FD
dr2 ( Co -
Cs) ] -
π 2 D
2. 303 r2 t
(3) 图2 100 % DOD 的MlNi3. 75 Co0. 65 Mn0. 4Al0. 2贮氢合 图2 100 % DOD 的MlNi3. 75 Co0. 65 Mn0. 4Al0. 2贮氢合
金电极在恒电位阶跃初期的电流2时间响应线 图2 100 % DOD 的MlNi3. 75 Co0. 65 Mn0. 4Al0. 2贮氢合
金电极在恒电位阶跃初期的电流2时间响应线
Fig. 2 Short2time plot of current2time response of Ml2
Ni3. 75 Co0. 65 Mn0. 4 Al0. 2. 1 电位阶跃幅值的选择 2 hydrogen storage alloy
electrode with 100 % DOD after application of
potential step with various amplitudes 式(2) 、(3) 中, i 为响应电流(A) , n 为
反应过程中的电子转移数, F 为法拉第常
数(96 485 C/ mol) , S 为电化学活性表面积
(cm2) , D 为氢的扩散系数(cm2/ s) , Co 和
Cs 分别为电极中氢的初始浓度和电位阶跃
后恒电位下电极表面氢的浓度(mol/ cm3) ,
可由库仑滴定曲线(开路电位随氢含量的
变化关系曲线) 测得, r 为合金颗粒的半径
(cm) , d 为扩散层厚度(cm) , t 为时间(s) ,
± 号中+ 号表示放电, - 号表示充电. Fig. 2 Short2time plot of current2time response of Ml2
Ni3. 75 Co0. 65 Mn0. 4 Al0. 2 hydrogen storage alloy
electrode with 100 % DOD after application of
potential step with various amplitudes 式(2) 给出,响应电流在阶跃初期与时
间平方根的倒数成线性关系,在合金颗粒
粒径已知的情况下,由该直线的斜率和截
距即可求出扩散系数D ;式(3) 表明,阶跃
后期响应电流对数随时间成直线变化,在
合金颗粒半径已知情况下,由直线斜率即
可求出扩散系数D. 式(2) 给出,响应电流在阶跃初期与时
间平方根的倒数成线性关系,在合金颗粒
粒径已知的情况下,由该直线的斜率和截
距即可求出扩散系数D ;式(3) 表明,阶跃
后期响应电流对数随时间成直线变化,在
合金颗粒半径已知情况下,由直线斜率即
可求出扩散系数D. 电 化 学 28 × 10 - 8
+ 80
9. 30 × 10 - 8
1. 30 × 10 - 8
+ 100
2. 16 × 10 - 7
1. 36 × 10 - 8
图4 室温下不同放电深度的MlNi3. 75 Co0. 65 Mn0. 4
Al0. 2贮氢合金电极在电位阶跃+ 50 mV 后的
电流2时间响应曲线
Fig. 4 Current2time responses of MlNi3. 75 Co0. 65 Mn0. 4
Al0. 2hydrogen storage alloy electrode with vari2
ous DOD after application of a potential step of
+ 50 mV at room temperature
从表1 可以看出,当用阶跃初期的响
应电流公式(式(2) ) 求解D 时,阶跃幅值的
大小对测试结果影响很大,如幅值为+ 100
mV 时的测试值比幅值为+ 10 mV 时的大
4 个数量级;而当利用阶跃后期的式(3) 求
解时,则阶跃幅值对测试结果的影响甚小,
如表所见,各不同阶跃幅值下所测得的D
大体接近,况且,跃幅值为+ 50 mV 时,由
以上2 个公式求得的结果相差不大;阶跃
幅值小于或大于+ 50 mV 时,两者相差分
别达到2～3 个数量级和7～15 倍. Tatau
Nishina 等[2 ,7 ]在用恒电位阶跃法研究Pd
和LaNi5 颗粒中氢的扩散行为时,也得到
过类似结果,并认为在阶跃初期较短的时
间内氢的扩散层主要建立在合金颗粒的表
面,由于表面粗糙程度的影响,实测的响应
电流远大于光滑表面电极的响应电流,而
且阶跃幅值越大这种影响越大. 而在阶跃后期,氢的扩散层已经扩展到合金颗粒的内部,响应
电流受合金表面粗糙程度的影响减小. Mingming Geng 等[4 ]曾讨论了用恒电位阶跃法研究贮
氢合金中氢的扩散系数时,其阶跃初期和阶跃后期响应电流两个关系式的适用性. 指出阶跃初
期的响应电流主要是界面双层电容的充电电流,阶跃后期的响应电流才是扩散过程的体现. 但
是,以上两种说法都只能解释阶跃幅值大于+ 50 mV 时,应用式(2) 产生较大的偏差,而不能
解释阶跃幅值小于+ 50 mV 时,应用式(3) 出现结果偏大的现象. 这种差别的真正原因还有待 图4 室温下不同放电深度的MlNi3. 75 Co0. 65 Mn0. 4
Al0. 2贮氢合金电极在电位阶跃+ 50 mV 后的
电流2时间响应曲线
Fig. 4 Current2time responses of MlNi3. 75 Co0. 65 Mn0. 4
Al0. 2hydrogen storage alloy electrode with vari2
ous DOD after application of a potential step of
+ 50 mV at room temperature
从表1 可以看出,当用阶跃初期的响
应电流公式(式(2) ) 求解D 时,阶跃幅值的
大小对测试结果影响很大,如幅值为+ 100
mV 时的测试值比幅值为+ 10 mV 时的大
4 个数量级;而当利用阶跃后期的式(3) 求
解时,则阶跃幅值对测试结果的影响甚小,
如表所见,各不同阶跃幅值下所测得的D
大体接近,况且,跃幅值为+ 50 mV 时,由
以上2 个公式求得的结果相差不大;阶跃
幅值小于或大于+ 50 mV 时,两者相差分
别达到2～3 个数量级和7～15 倍. Tatau
Nishina 等[2 ,7 ]在用恒电位阶跃法研究Pd
和LaNi5 颗粒中氢的扩散行为时,也得到
过类似结果,并认为在阶跃初期较短的时
间内氢的扩散层主要建立在合金颗粒的表
面,由于表面粗糙程度的影响,实测的响应
电流远大于光滑表面电极的响应电流,而
且阶跃幅值越大这种影响越大. 而在阶跃后期,氢的扩散层已经扩展到合金颗粒的内部,响应
电流受合金表面粗糙程度的影响减小. Mingming Geng 等[4 ]曾讨论了用恒电位阶跃法研究贮
氢合金中氢的扩散系数时,其阶跃初期和阶跃后期响应电流两个关系式的适用性. 指出阶跃初
期的响应电流主要是界面双层电容的充电电流,阶跃后期的响应电流才是扩散过程的体现. 但
是,以上两种说法都只能解释阶跃幅值大于+ 50 mV 时,应用式(2) 产生较大的偏差,而不能
解释阶跃幅值小于+ 50 mV 时,应用式(3) 出现结果偏大的现象. 这种差别的真正原因还有待 图3 100 %DOD 的MlNi3. 75Co0. 65Mn0. 4Al0. 2贮氢合金
电极在恒电位阶跃后期的电流2时间响应曲线 图3 100 %DOD 的MlNi3. 75Co0. 65Mn0. 4Al0. 2贮氢合金
电极在恒电位阶跃后期的电流2时间响应曲线 Fig. 3 Long2time plot of the current2time response of
MlNi3. 75Co0. 65Mn0. 4Al0. 2hydrogen storage alloy
electrode with 100 %DOD after application of
potential step with various amplitudes 电 化 学 电 化 学 2001 年 ·
8
4
4
· 此求得的氢的扩散系数列于表1. 表1 100 %DOD 的贮氢合金MlNi3. 75Co0. 65Mn0. 4Al0. 2中氢的扩散系数
Tab. 1 Hydrogen diffusion coefficient in hydrogen storage alloy MlNi3. 75Co0. 65Mn0. 4Al0. 2with 100 %DOD
Potential step amplitude/ mV
D/ cm2· s - 1
By equation(2)
By equation(3)
+ 10
6. 80 × 10 - 11
1. 16 × 10 - 8
+ 20
3. 63 × 10 - 10
1. 19 × 10 - 8
+ 50
1. 05 × 10 - 8
1. 28 × 10 - 8
+ 80
9. 30 × 10 - 8
1. 30 × 10 - 8
+ 100
2. 16 × 10 - 7
1. 36 × 10 - 8
图4 室温下不同放电深度的MlNi3. 75 Co0. 65 Mn0. 4
Al0. 2贮氢合金电极在电位阶跃+ 50 mV 后的
电流2时间响应曲线
Fig. 4 Current2time responses of MlNi3. 75 Co0. 65 Mn0. 4
Al0. 2hydrogen storage alloy electrode with vari2
ous DOD after application of a potential step of
+ 50 mV at room temperature
从表1 可以看出,当用阶跃初期的响
应电流公式(式(2) ) 求解D 时,阶跃幅值的
大小对测试结果影响很大,如幅值为+ 100
mV 时的测试值比幅值为+ 10 mV 时的大
4 个数量级;而当利用阶跃后期的式(3) 求
解时,则阶跃幅值对测试结果的影响甚小,
如表所见,各不同阶跃幅值下所测得的D
大体接近,况且,跃幅值为+ 50 mV 时,由
以上2 个公式求得的结果相差不大;阶跃
幅值小于或大于+ 50 mV 时,两者相差分
别达到2～3 个数量级和7～15 倍. Tatau
Nishina 等[2 ,7 ]在用恒电位阶跃法研究Pd
和LaNi5 颗粒中氢的扩散行为时,也得到
过类似结果,并认为在阶跃初期较短的时
间内氢的扩散层主要建立在合金颗粒的表
面,由于表面粗糙程度的影响,实测的响应
电流远大于光滑表面电极的响应电流,而
且阶跃幅值越大这种影响越大. 而在阶跃后期,氢的扩散层已经扩展到合金颗粒的内部,响应
电流受合金表面粗糙程度的影响减小. Mingming Geng 等[4 ]曾讨论了用恒电位阶跃法研究贮
氢合金中氢的扩散系数时,其阶跃初期和阶跃后期响应电流两个关系式的适用性. 指出阶跃初
期的响应电流主要是界面双层电容的充电电流,阶跃后期的响应电流才是扩散过程的体现. 但 表1 100 %DOD 的贮氢合金MlNi3. 75Co0. 65Mn0. 4Al0. 2中氢的扩散系数 表1 100 %DOD 的贮氢合金MlNi3. 75Co0. 65Mn0. 4Al0. 2中氢的扩散系数
Tab. 1 Hydrogen diffusion coefficient in hydrogen storage alloy MlNi3. 75Co0. 65Mn0. 4Al0. 2with 100 %DOD
Potential step amplitude/ mV
D/ cm2· s - 1
By equation(2)
By equation(3)
+ 10
6. 80 × 10 - 11
1. 16 × 10 - 8
+ 20
3. 63 × 10 - 10
1. 19 × 10 - 8
+ 50
1. 05 × 10 - 8
1. 28 × 10 - 8
+ 80
9. 30 × 10 - 8
1. 30 × 10 - 8
+ 100
2. 16 × 10 - 7
1. 36 × 10 - 8 en diffusion coefficient in hydrogen storage alloy MlNi3. 75Co0. 65Mn0. 4Al0. 2with 100 %DOD Potential step amplitude/ mV
D/ cm2· s - 1
By equation(2)
By equation(3)
+ 10
6. 80 × 10 - 11
1. 16 × 10 - 8
+ 20
3. 63 × 10 - 10
1. 19 × 10 - 8
+ 50
1. 05 × 10 - 8
1. 图4 室温下不同放电深度的MlNi3. 75 Co0. 65 Mn0. 4
Al0. 2贮氢合金电极在电位阶跃+ 50 mV 后的
电流2时间响应曲线 Fig. 4 Current2time responses of MlNi3. 75 Co0. 65 Mn0. 4
Al0. 2hydrogen storage alloy electrode with vari2
ous DOD after application of a potential step of
+ 50 mV at room temperature + 50 mV at room temperature
电流远大于光滑表面电极的响应电流,而
且阶跃幅值越大这种影响越大. 而在阶跃后期,氢的扩散层已经扩展到合金颗粒的内部,响应
电流受合金表面粗糙程度的影响减小. Mingming Geng 等[4 ]曾讨论了用恒电位阶跃法研究贮
氢合金中氢的扩散系数时,其阶跃初期和阶跃后期响应电流两个关系式的适用性. 指出阶跃初
期的响应电流主要是界面双层电容的充电电流,阶跃后期的响应电流才是扩散过程的体现. 但
是,以上两种说法都只能解释阶跃幅值大于+ 50 mV 时,应用式(2) 产生较大的偏差,而不能
解释阶跃幅值小于+ 50 mV 时,应用式(3) 出现结果偏大的现象. 这种差别的真正原因还有待 第4 期 原鲜霞等:恒电位阶跃法研究贮氢合金MlNi3. 75Co0. 65Mn0. 4Al0. 2中氢的扩散行为 于进一步探讨. 考虑到特定状态下贮氢合
金中氢的扩散速度应该是大体一定的,扩
散系数的测试结果不应随阶跃幅值的变化
而有数量级的变化,我们认为利用阶跃后
期的式(3) 求贮氢合金中氢的扩散系数较
为合理. 于进一步探讨. 考虑到特定状态下贮氢合
金中氢的扩散速度应该是大体一定的,扩
散系数的测试结果不应随阶跃幅值的变化
而有数量级的变化,我们认为利用阶跃后
期的式(3) 求贮氢合金中氢的扩散系数较
为合理. 2. 3 贮氢合金MlNi3. 75Co0. 65Mn0. 4 上述结果与Chiaki Iwakura[6 ]所测得
的MmNi4. 2Al0. 5M0. 3 (M = Cr、Mn、Fe、Co 、
Ni) 合金中氢的扩散系数随氢含量的增加
而减小及氢的扩散活化能为19. 9 kJ/ mol
相当吻合. 3 结 论
(1) 恒电位阶跃法求贮氢合金中氢
的扩散系数时,宜采用阶跃后期的log i～t
变化公式. (2) 室温下,贮氢合金MlNi3. 75Co0. 65
Mn0. 4Al0. 2中氢的扩散系数随其放电深度
的增大而增大. (3) 50 %DOD 的贮氢合金MlNi3. 75
Co0. 65Mn0. 4Al0. 2中氢的扩散系数随温度的
升高而增大,其中氢扩散的活化能为19. 87 kJ/ mol. 图7 50 %DOD MlNi3. 75 Co0. 65 Mn0. 4Al0. 2的贮氢合金
中氢的扩系数与温度的关系
Fig. 7 Temperature dependence of hydrogen diffusion
coefficient in hydrogen storage alloy MlNi3. 75
Co0. 65Mn0. 4Al0. 2with 50 %DOD
的活化能为19. 87 kJ/ mol. 上述结果与Chiaki Iwakura[6 ]所测得
的MmNi4. 2Al0. 5M0. 3 (M = Cr、Mn、Fe、Co 、
Ni) 合金中氢的扩散系数随氢含量的增加
而减小及氢的扩散活化能为19. 9 kJ/ mol
相当吻合. 3 结 论
(1) 恒电位阶跃法求贮氢合金中氢
的扩散系数时,宜采用阶跃后期的log i～t
变化公式. (2) 室温下,贮氢合金MlNi3. 75Co0. 65
Mn0. 4Al0. 2中氢的扩散系数随其放电深度
的增大而增大. (3) 50 %DOD 的贮氢合金MlNi3. 75
Co0. 65Mn0. 4Al0. 2中氢的扩散系数随温度的
升高而增大,其中氢扩散的活化能为19. 87 kJ/ mol. (1) 恒电位阶跃法求贮氢合金中氢
的扩散系数时,宜采用阶跃后期的log i～t
变化公式. (2) 室温下,贮氢合金MlNi3. 75Co0. 65
Mn0. 4Al0. 2中氢的扩散系数随其放电深度
的增大而增大. 图7 50 %DOD MlNi3. 75 Co0. 65 Mn0. 4Al0. 2的贮氢合金
中氢的扩系数与温度的关系 Fig. 7 Temperature dependence of hydrogen diffusion
coefficient in hydrogen storage alloy MlNi3. 75
Co0. 65Mn0. 4Al0. 2with 50 %DOD YUAN Xian xia
, XU Nai xin
( S hanghai Institute of Meatallurgy , Chinese Academy of Sciencs , S hanghai 200050 , China) S hanghai Institute of Meatallurgy , Chinese Academy of Sciencs , S hanghai 20 Abstract : Potential step method was employed to study the hydrogen diffusion behavior in hy2
drogen storage alloy MlNi3. 75Co0. 65Mn0. 4Al0. 2 with various depth of discharge (DOD) at ambient
temperature and with 50 %DOD at various temperatures. It was found that hydrogen diffusion
cofficient in the hydrogen storage alloy increaes with the increase in DOD at ambient temperature. and for the alloy with 50 %DOD ,the hydrogen diffusion coefficient increases with the increase in
temperature and the activation energy for hydrogen diffusion in the alloy is 19. 87 kJ/ mol. Key words : Hydrogen storage alloy ,Hydrogen diffusion coefficient ,Potential step method [1 ] Van Rijiswick M H J . Hydride for Energy Storage[ M] Oxford : Pergamon Press , 1978. 261. 2. 3 贮氢合金MlNi3. 75Co0. 65Mn0. 4 2. 3 贮氢合金MlNi3. 75Co0. 65Mn0. 4 2. 3 贮氢合金MlNi3. 75Co0. 65Mn0. 4 图5 室温下贮氢合金MlNi3. 75 Co0. 65 Mn0. 4Al0. 2中氢
的扩散系数与其放电深度的关系
Fig. 5 DOD dependence of hydrogen diffusion coeffi2
cient in hydrogen storage alloy MlNi3. 75 Co0. 65
Mn0. 4Al0. 2at room temperature
2. 3 贮氢合金MlNi3. 75Co0. 65Mn0. 4
Al0. 2中氢扩散系数的测定
图4 示出室温下不同放电深度的贮氢
合金电极在电位阶跃+ 50 mV 后的电流2
时间响应曲线. 如图可见,当时间大于250 s
时,响应电流对数随时间的变化成一直线. 由该直线的斜率求得的贮氢合金中氢的扩
散系与其放电深度的关系见图5 ,可以看
出,MlNi3. 75Co0. 65Mn0. 4Al0. 2中氢的扩散系
图6 50 %DOD MlNi3. 75Co0. 65Mn0. 4Al0. 2贮氢合金
电极于各不同温度下电位阶跃+ 50 mV 后
的电流2时间响应曲线
Fig. 6 Current2time responses of MlNi3. 75 Co0. 65
Mn0. 4Al0. 2 hydrogen storage alloy electrode
with 50 %DOD after application of a poten2
tial step of 50 mV at various temperatures
数随着DOD 的增大而增大. 这是因为放电深
度较低时,电极中氢的浓度较大,而且多数以
金属氢化物的形式存在,氢在扩散前必须从吸
收态转变为吸附态(即电极中的活性物质从β
相的金属氢化物转变为α 相的固溶体) . 此外,
合金中要有足够多的可存放氢的金属原子间
空位以保证氢在其中顺利扩散. 但当放电深度
较低时,电极中较高的氢浓度却使得可存放氢
的空位相对减少,此时,氢的扩散一方面受到
相变过程的影响,另一方面还受到较少的可存
氢的空位的制约. 当放电深度较高时,电极中
氢的浓度很小,其中可存放氢的空位很多,而
且此时电极中的氢主要以固溶体的形式存在,
氢不必经历相变就可以顺利扩散,所以合金中
氢的扩散受上述两个因素的影响很小甚至根
本不受其影响,从而扩散系数也就较大. 50 %DOD 的贮氢合金电极于各个温度下
恒电位阶跃+ 50 mV 后的电流2时间响应曲线
即如图6 所示. 如图,当t > 250 s 时,log i 与t 成直线关系. 由该直线的斜率求得的氢的扩散系
数与温度的关系见图7. 可以看出,该合金中氢的扩散系数D 随温度的升高而增大,而且D 的
自然对数与温度的倒数基本上也成直线关系. 根据Arrhenius 公式求得,该贮氢合金中氢扩散 Al0. 2中氢扩散系数的测定
图
示出室温下不同放电深度的贮氢 图
F
Al0. 2中氢扩散系数的测定
图4 示出室温下不同放电深度的贮氢
合金电极在电位阶跃+ 50 mV 后的电流2
时间响应曲线. 如图可见,当时间大于250 s
时,响应电流对数随时间的变化成一直线. 由该直线的斜率求得的贮氢合金中氢的扩
散系与其放电深度的关系见图5 ,可以看
出,MlNi3. 75Co0. 65Mn0. 4Al0. 2中氢的扩散系
数随着DOD 的增大而增大. 这是因为放电深
度较低时,电极中氢的浓度较大,而且多数以
金属氢化物的形式存在,氢在扩散前必须从吸
收态转变为吸附态(即电极中的活性物质从β
相的金属氢化物转变为α 相的固溶体) . 此外,
合金中要有足够多的可存放氢的金属原子间
空位以保证氢在其中顺利扩散. 但当放电深度
较低时,电极中较高的氢浓度却使得可存放氢
的空位相对减少,此时,氢的扩散一方面受到
相变过程的影响,另一方面还受到较少的可存
氢的空位的制约. 当放电深度较高时,电极中
氢的浓度很小,其中可存放氢的空位很多,而
且此时电极中的氢主要以固溶体的形式存在,
氢不必经历相变就可以顺利扩散,所以合金中
氢的扩散受上述两个因素的影响很小甚至根
本不受其影响,从而扩散系数也就较大. 的贮氢合金电极于各个温度下 图5 室温下贮氢合金MlNi3. 75 Co0. 65 Mn0. 4Al0. 2中氢
的扩散系数与其放电深度的关系 Fig. 5 DOD dependence of hydrogen diffusion coeffi2
cient in hydrogen storage alloy MlNi3. 75 Co0. 65
Mn0. 4Al0. 2at room temperature Mn0. 4Al0. 2at room temperature 图6 50 %DOD MlNi3. 75Co0. 65Mn0. 4Al0. 2贮氢合金
电极于各不同温度下电位阶跃+ 50 mV 后
的电流2时间响应曲线 图6 50 %DOD MlNi3. 75Co0. 65Mn0. 4Al0. 2贮氢合金
电极于各不同温度下电位阶跃+ 50 mV 后
的电流2时间响应曲线 Fig. 6 Current2time responses of MlNi3. 75 Co0. 65
Mn0. 4Al0. 2 hydrogen storage alloy electrode
with 50 %DOD after application of a poten2
tial step of 50 mV at various temperatures w t
50 % O
a te app cat o o a pote
tial step of 50 mV at various temperatures
50 %DOD 的贮氢合金电极于各个温度下
恒电位阶跃+ 50 mV 后的电流2时间响应曲线
即如图6 所示. 如图,当t > 250 s 时,log i 与t 成直线关系. 由该直线的斜率求得的氢的扩散系
数与温度的关系见图7. 可以看出,该合金中氢的扩散系数D 随温度的升高而增大,而且D 的
自然对数与温度的倒数基本上也成直线关系. 根据Arrhenius 公式求得,该贮氢合金中氢扩散 电 化 学 2001 年 ·
0
5
4
· 图7 50 %DOD MlNi3. 75 Co0. 65 Mn0. 4Al0. 2的贮氢合金
中氢的扩系数与温度的关系
Fig. 7 Temperature dependence of hydrogen diffusion
coefficient in hydrogen storage alloy MlNi3. 75
Co0. 65Mn0. 4Al0. 2with 50 %DOD
的活化能为19. 87 kJ/ mol. Study of Hydrogen Diffusion Behavior in Hydrogen Storage Alloy
MlNi3. 75Co0. 65Mn0. 4Al0. 2 with Potential Step Method YUAN Xian2xia 3 , XU Nai2xin
( S hanghai Institute of Meatallurgy , Chinese Academy of Sciencs , S hanghai 200050 , China) YUAN Xian2xia 3 , XU Nai2xin YUAN Xian2xia
, XU Nai2xin
( S hanghai Institute of Meatallurgy , Chinese Academy of Sciencs , S hanghai 200050 , China) References :
[1 ] Van Rijiswick M H J . Hydride for Energy Storage[ M] Oxford : Pergamon Press , 1978. 261. ·
1
5
4
·
第4 期 原鲜霞等:恒电位阶跃法研究贮氢合金MlNi3. 75Co0. 65Mn0. 4Al0. 2中氢的扩散行为 [2 ] Tatsuo Nishna , Hironori Ura , Isamu Uchida. Determination of the chemical diffusion coefficients in metal
hydride particles with a microelectrode technique[J ].J . electrochem. Soc ,1997 ,144 :1 273. [3 ] Miyamura H , Kuriyama N ,Sakai T , et al. Hydrogen diffusion in amorphous LaNi2 Hx thin film[J ]. J . alloys
and compounds , 1993 ,192 :188. [4 ] Mingming2Geng ,Jianmen Han ,Fengfeng ,et al. Electrochemical measurements of a metal hydride for the Ni/
MH battery[J ]. Int.J . Hydrogen Energy , 1999 ,25 :203. [5 ] Cai Chenxin , Zhao Dongjiang , Wang Baochen. Study of the diffusion behavior of hydrogen in hydrogen stor2
age alloy[J ]. Chinese J . Power Sources ,1993 ,17(5) :9. [6 ] Chiaki Iwakura , Takafumi Oura , Hiroshi Inoue , et al. Effect of alloy composition on hydrogen diffusion in
the AB52type hydrogen storage alloys[J ].J . Electroanal. Chem. ,1995 ,398 :37. [7 ] Hironori Ura , Tasuo Nishina , Isamu Uchida. Electrochemical measurements of single particles of Pb and
LaNi5 with a microelectrode technique[J ]. J . Electroanal. Chem. , 1995 ,396 :169. [8 ] Li Yang ,Cheng Yang2Tse. Hydrogen diffusion and solubility in Palladium thin films[J ]. Int.J . Hydrogen En2
ergy , 1996 ,21(4) :281.
|
https://openalex.org/W1967415975
|
https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0064729&type=printable
|
English
| null |
Serum Resistin, Cardiovascular Disease and All-Cause Mortality in Patients with Type 2 Diabetes
|
PloS one
| 2,013
|
cc-by
| 8,046
|
Abstract Background: High serum resistin has been associated with increased risk of cardiovascular disease in the general
population, Only sparse and conflicting results, limited to Asian individuals, have been reported, so far, in type 2 diabetes. We studied the role of serum resistin on coronary artery disease, major cardiovascular events and all-cause mortality in type
2 diabetes. Methods: We tested the association of circulating resistin concentrations with coronary artery disease, major cardiovascular
events (cardiovascular death, non-fatal myocardial infarction and non-fatal stroke) and all-cause mortality in 2,313 diabetic
patients of European ancestry from two cross-sectional and two prospective studies. In addition, the expression of resistin
gene (RETN) was measured in blood cells of 68 diabetic patients and correlated with their serum resistin levels. Results: In a model comprising age, sex, smoking habits, BMI, HbA1c, and insulin, antihypertensive and antidyslipidemic
therapies, serum resistin was associated with coronary artery disease in both cross-sectional studies: OR (95%CI) per SD
increment = 1.35 (1.10–1.64) and 1.99 (1.55–2.55). Additionally, serum resistin predicted incident major cardiovascular events
(HR per SD increment = 1.31; 1.10–1.56) and all-cause mortality (HR per SD increment = 1.16; 1.06–1.26). Adjusting also for
fibrinogen levels affected the association with coronary artery disease and incident cardiovascular events, but not that with
all cause-mortality. Finally, serum resistin was positively correlated with RETN mRNA expression (rho = 0.343). Conclusions: This is the first study showing that high serum resistin (a likely consequence, at least partly, of increased RETN
expression) is a risk factor for cardiovascular disease and all-cause mortality in diabetic patients of European ancestry. Citation: Menzaghi C, Bacci S, Salvemini L, Mendonca C, Palladino G, et al. (2013) Serum Resistin, Cardiovascular Disease and All-Cause Mortality in Patients with
Type 2 Diabetes. PLoS ONE 8(6): e64729. doi:10.1371/journal.pone.0064729
Editor: Francesco Dotta, University of Siena, Italy Editor: Francesco Dotta, University of Siena, Italy Received January 16, 2013; Accepted April 18, 2013; Published June 3, 2013 Received January 16, 2013; Accepted April 18, 2013; Published June 3, 2013 pyright: 2013 Menzaghi et al. This is an open-access article distributed under the terms of the Creative Commons Attribution L
restricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Abstract Funding: This research was supported by Accordo Programma Quadro in Materia di Ricerca Scientifica nella Regione Puglia-PST 2006 and PO Puglia FESR 2007–
2013, Italian Ministry of Health grants RC2011, RC2012, European Foundation for the Study of Diabetes/Pfizer grant (CM) and National Institutes of Health grant
HL073168 (AD). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: CM received funding from the EFSD/Pfizer. MF is a PLOS ONE Editorial Board member. There are no patents, products in development or
marketed products to declare. This does not alter the authors’ adherence to all the PLOS ONE policies on sharing data and materials. * E-mail: c.menzaghi@operapadrepio.it (CM); v.trischitta@operapadrepio.it (VT) Serum Resistin, Cardiovascular Disease and All-Cause
Mortality in Patients with Type 2 Diabetes Claudia Menzaghi1*, Simonetta Bacci2, Lucia Salvemini1, Christine Mendonca3, Giuseppe Palladino1,
Andrea Fontana4, Concetta De Bonis1, Antonella Marucci1, Elizabeth Goheen3, Sabrina Prudente5,
Eleonora Morini1, Stefano Rizza6, Alyssa Kanagaki3, Grazia Fini1, Davide Mangiacotti1,
Massimo Federici6, Salvatore De Cosmo2, Fabio Pellegrini4,7, Alessandro Doria3,8,
Vincenzo Trischitta1,5,9* 1 Research Unit of Diabetes and Endocrine Diseases, IRCCS Casa Sollievo della Sofferenza, San Giovanni Rotondo, Italy, 2 Unit of Endocrinology, IRCCS Casa Sollievo della
Sofferenza, San Giovanni Rotondo, Italy, 3 Research Division, Joslin Diabetes Center, Boston, Massachusetts, United States of America, 4 Unit of Biostatistics, IRCCS Casa
Sollievo della Sofferenza, San Giovanni Rotondo, Italy, 5 IRCSS Casa Sollievo della Sofferenza-Mendel Laboratory, Rome, Italy, 6 Department of Internal Medicine, University
of Rome Tor Vergata, Rome, Italy, 7 Unit of Biostatistics, Consorzio Mario Negri Sud, Santa Maria Imbaro, Italy, 8 Department of Medicine, Harvard Medical School, Boston,
Massachusetts, United States of America, 9 Department of Experimental Medicine, Sapienza University of Rome, Italy Data Collection and Definitions Clinical data were obtained from a standardized interview and
examination. Body mass index (BMI) was calculated by dividing
the weight (in kilograms) by the square of height (in meters). Smoking habits and history of hypertension (as indicated by the
presence of anti-hypertensive therapy), dyslipidemia (as indicated
by the presence of anti-dyslipidemic therapy), and MI as well as
glucose-lowering
treatment
were
also
recorded
at
time
of
examination. Data regarding medications were confirmed by
review of medical records. Those who reported smoking cigarettes
regularly during the year before the examination were considered
current smokers. Joslin Heart Study (JHS). This study consists of a series of
868 CAD cases and controls, all with type 2 diabetes (ADA 2003
criteria), who lived in the greater Boston area and received
treatment at the Joslin Clinic and/or the Beth Israel Deaconess
Medical Center (BIDMC) at the time of their recruitment [21]. Joslin Heart Study (JHS). This study consists of a series of
868 CAD cases and controls, all with type 2 diabetes (ADA 2003
criteria), who lived in the greater Boston area and received
treatment at the Joslin Clinic and/or the Beth Israel Deaconess
Medical Center (BIDMC) at the time of their recruitment [21]. All participants were self-reported non-Hispanic Whites. Case
participants with CAD were a random sample of patients with
type 2 diabetes who had a stenosis greater than 50% in a major
coronary artery or a main branch thereof that was documented by
cardiac catheterization at the BIDMC between 2001 and 2008. Sixty percent of the case patients received diabetes management
care at the Joslin Clinic. Control participants without CAD were
randomly selected from among Joslin patients who were identified
between 2001 and 2008 as fulfilling the following criteria: (1)
current age between 55 and 74 years; (2) type 2 diabetes for 5 years
or more; (3) negative cardiovascular history (i.e., normal resting
electrocardiogram, absence of cardiac symptoms, and no hospi-
talization for cardiovascular events); and (4) non inducible
ischemia to an exercise treadmill test performed for screening
purposes. In the GHS-cross sectional and-prospective designs and GMS,
blood samples were collected between 8:00 and 9:00 AM after an
overnight fast. In the JHS, blood samples were obtained between
7:00 AM and 6:00 PM without the requirement of fasting. Serum
aliquots were stored at 280uC. Methods Gargano
Mortality
Study
(GMS). One thousand and
twenty-eight patients with type 2 diabetes (ADA 2003 criteria)
were consecutively recruited from November 1th 2000 to
September 30th 2005 at the Endocrine Unit of IRCCS ‘‘Casa
Sollievo della Sofferenza’’ in San Giovanni Rotondo, for a study
having all-cause mortality as the end-point (Figure S1). The only
exclusion criterion was the presence of poor life expectancy due to
non diabetes-related disorders. This cohort was followed until
2010 by obtaining information on the participants’ vital status by
direct contact with patients and/or their relatives or by queries to
the registry offices of the cities of residence. Such information was
available in 838 individuals whose data were therefore analyzed in
the present study. One hundred and three of the GMS
participants are also participants of the GHS-prospective design
(Figure S1). Introduction [4,5]. Several cross-sectional studies based on resistin serum levels
and/or tissue expression have pointed to this molecule as a pro-
inflammatory adipokine contributing to atherosclerosis and the
clinical phenotypes resulting from it [5,6,7,8,9,10,11]. High serum
resistin
levels
have also been
found, although
with
some
inconsistencies, to predict incident cardiovascular events in
prospective studies [12,13,14,15,16,17]. This evidence, however,
mostly concerns the general population since the few published
studies of resistin as a CVD marker in diabetic subjects are small,
limited to Asian individuals, and contradictory in their findings
[9,11,18]. Given that cardiovascular risk may be differently shaped Cardiovascular disease (CVD) is a major cause of morbidity and
mortality among patients with type 2 diabetes [1]. Although
several components of the diabetic milieu contribute to the
increased risk of CVD associated with diabetes, insulin resistance
and inflammation have been recognized as particularly important
pathogenic factors [2]. Both conditions have been linked to
cytokines released by the adipose tissue and collectively known as
adipokines [3]. Among these is resistin, a 12.5 kDa cysteine-rich
protein, which, in humans, is primarily secreted by macrophages June 2013 | Volume 8 | Issue 6 | e64729 1 PLOS ONE | www.plosone.org June 2013 | Volume 8 | Issue 6 | e64729 Resistin and RETN Levels in Type 2 Diabetes in non diabetic as compared to diabetic individuals answering the
question of whether or not resistin plays a role in the development
of CVD also among the latter group is definitely needed. To
address this question, we analyzed data from over 2,300 European
subjects with type 2 from four different studies: two case-control
collections of such patients with and without evidence of coronary
artery disease (CAD), a prospective cohort of patients with type 2
diabetes followed over time for incident major cardiovascular
events and another prospective cohort of patients with type 2
diabetes followed over time with regard to all-cause mortality. point was a combination of major cardiovascular events including
cardiovascular death (i.e. according to the international classifica-
tion of diseases’ codes: 428.1- ninth edition - and I21.0–I21.9,
I25.9, I46.9–I50.9, I63.0, I63.9, I70,2– tenth edition), non-fatal
MI and non-fatal stroke [22]. For all non-fatal MI and strokes,
confirmation of the events was obtained from the hospital medical
records. Introduction In the case of patients who did not show up at the
scheduled clinical control, information on the incident cardiovas-
cular events was obtained through telephone interviews with the
patients or their primary care physicians or from death certificates. Serum resistin was measured in 359 (98%) participants. Serum resistin was measured in 359 (98%) participants. Data Collection and Definitions Peripheral whole blood cells
(PWBC) RNA was obtained from 68 fasting patients with type 2
diabetes, with no clinical evidence of CVD (38 males/30 females,
age 65.167.0 years, BMI 30.965.0 kg/m2, HbA1c 7.961.7%) by
PAXgene Blood RNA collection tubes (PreAnalytiX, GmbH,
Germany). These patients, not belonging to any of the previous
samples, were consecutively recruited at the Endocrine Unit of
IRCCS ‘‘Casa Sollievo della Sofferenza’’ in San Giovanni
Rotondo, with the specific purpose of correlating gene expression
levels on PWBC with clinical features and/or biomarkers levels. Serum resistin was measured in 861 (99%) participants. Case-Control Studies Gargano heart study (GHS)-cross sectional design. This
study includes 798 European subjects from Italy with type 2
diabetes (ADA 2003 criteria) who were consecutively recruited at
the Endocrine Unit of IRCCS ‘‘Casa Sollievo della Sofferenza’’ in
San Giovanni Rotondo (Gargano, Center East Coast of Italy) from
2001 to 2008, as part of an ongoing investigation on the genetics of
CAD in type 2 diabetes [19,20] (Figure S1). Cases are patients who underwent coronary angiography and
had a stenosis .50% in at least one coronary major vessel or with
previous myocardial infarction (MI). Controls include asymptom-
atic patients without signs of myocardial ischemia at resting and
maximal symptom limited stress ECG. The latter was conducted
on a treadmill according to a Bruce protocol after cardiovascular
drugs as b-blockers and Ca-channel blockers were stopped for 48
hours. The test was defined as maximal if 85% of the predicted
heart rate for the participant’s age was reached. Ischemia was
defined as a horizontal or downsloping ST segment depression of
1 mm or more calculated at 0.08 s after the J point (i.e. the
junction between QRS complex and ST segment) or the
development of typical angina pectoris. (
g
)
Serum resistin was measured in 779 (93%) participants. Prospective Studies Each study protocol and the informed consent procedures were
approved by the local Institutional Ethic Committee IRCCS
(Istituto di Ricovero e Cura a Carattere Scientifico) ‘‘Casa Sollievo
della Sofferenza’’ for GHS and GMS and by the Joslin Committee
on Human Studies and the Beth Israel Deaconess Medical Center
Committee on Clinical Investigations for JHS. All participants
gave written consent. Statistical Methods Patients’
baseline
characteristics
were
reported
as
mean
6standard deviation (SD) and percentages for continuous and
categorical variables, respectively. Correlations between contin-
uous variables were assessed by Pearson coefficient. In case-control studies, the association between resistin circu-
lating
levels
and
CAD
was
assessed
with
univariate
and
multivariate logistic regression models with CAD status as the
dependent and resistin as the independent variable. Separate
analyses were performed for continuous resistin values and tertiles
of its distribution. The strength of the associations was estimated
by means of odds ratios (ORs), along with their 95% Confidence
Intervals (95% CI), per SD increase in baseline resistin level and
for tertiles of its distribution. In addition, a test for linear trend in
OR estimates over resistin tertiles was performed by including
resistin tertiles (coded as 1, 2, and 3) as a continuous variable into
the logistic model. A p-value ,0.05 was considered as significant. All analyses were
performed using SAS Release 9.1.3 (SAS Institute, Cary, NC,
USA). Measurement of RETN mRNA Levels Total RNA from PWBC was extracted using PAXgene Blood
RNA kit (PreAnalytiX, GmbH, Germany). RNA was eluted in
RNAse free-water and stored at 280uC until used. Total RNA
yield and purity were determined spectrophotometrically using the
NanoDrop
ND-1000
(Wilmington,
DE,
USA). Integrity
of
resuspended total RNA was determined by electrophoretic
separation and subsequent laser induced florescence detection
using the RNA 6000 Nano Assay Chip Kit on the Bioanalyzer
2100 (Agilent Technologies, Waldbronn, Germany). Five hundred nanograms of RNA were reverse transcripted by
AMV Reverse Transcription System (Promega Corp., Wis, USA)
and
used
as
template
in
subsequent
analyses. RETN
(Hs00220767_m1) and GAPDH (Hs99999905_m1) gene expres-
sion assays on demand kit reagents Applied Biosystems (Foster
City, CA) were used to quantify in triplicates relative gene
expression on ABI-PRISM 7500 Applied Biosystems (Foster City,
CA). RETN transcription levels were normalized using the GAPDH
housekeeping gene. RETN/GAPDH mRNA ratios were obtained
from the equation 22DCt, where DCt is the difference in threshold
cycles between RETN and GAPDH. Reclassification improvement offered by resistin was quantified
using the survival-based net reclassification index (NRI) following
the Kaplan-Meier approach with one-sided bootstrapped p-values
based on 1000 re-samplings with replacement [30,31] and by
Integrated Discrimination Improvement (IDI) (28). Since no
established risk cut-offs were available for our high risk patients
as those affected by diabetes, we computed the categories-free
version of NRI (i.e. cNRI) [31]. The main difference consists in the
definition of a reclassified subject: for the NRI a subject has to
move from one risk-category to another one; the cNRI requires
that the subject’s risk probability changes, without any limit, to
define an upward or downward reclassification. cNRI is a more
objective measure of improvement in risk prediction while NRI
has a more attractive interpretation for clinicians. The time
horizon of risk prediction was set to 7 years. (i.e. upper cut-off of
the 3th quartile). Resistin and RETN Levels in Type 2 Diabetes Measurement of Circulating Resistin Levels
Serum resistin concentrations were measured by a commercial
ELISA (Bio Vendor, Brno Czech Republic) at the Research Unit
of Diabetes and Endocrine Diseases in San Giovanni Rotondo, as
previously described [23]. Inter- and intra-assay coefficients of
variation were 3.2–4% and 6.3–7.2% respectively. resistin tertiles was performed by including resistin tertiles as a
continuous variable into the Cox proportional hazard models. Adjusted survival curves were derived from the Cox proportional
hazard models, using the direct approach [24]. g
Serum resistin concentrations were measured by a commercial
ELISA (Bio Vendor, Brno Czech Republic) at the Research Unit
of Diabetes and Endocrine Diseases in San Giovanni Rotondo, as
previously described [23]. Inter- and intra-assay coefficients of
variation were 3.2–4% and 6.3–7.2% respectively. g
pp
[
]
Predicted risk probabilities were derived from the Framingham
Risk Score (FRS), which is an established risk model for
cardiovascular event in the general population [25] and from
the UKPDS risk engine, which is a model for the risk of coronary
heart disease in type 2 diabetes [26]. Models’ calibration, i.e. the
agreement between observed outcomes and predictions, was
assessed using the survival-based Hosmer-Lemeshow (HL) good-
ness-of-fit
test
[27],
a
chi-square
test
based
on
grouping
observations into deciles of predicted risk and testing associations
with observed outcomes. Models’ discrimination, i.e. the ability to
distinguish subjects who will develop an event from those who will
not, was assessed by computing the modified C statistic for
censored survival data [28,29]. Comparison between C-indices
was carried out following Pencina and D’Agostino’s approach
[29]. GHS-prospective design. GHS-prospective design. This study comprises 368 patients
with type 2 diabetes and CAD (as previously defined), who were all
case participants of the GHS-cross sectional design (Figure S1). Follow-up information on outcomes was collected yearly from
2002 to 2011. The only exclusion criterion was the presence of
poor life expectancy for non diabetes-related diseases. The end- June 2013 | Volume 8 | Issue 6 | e64729 PLOS ONE | www.plosone.org 2 Resistin and RETN Levels in Type 2 Diabetes Case-Control Studies Clinical features of participants in the GHS-cross sectional
design and the JHS are summarized in Table 1. In both studies,
serum resistin concentrations were significantly higher in CAD-
positive cases than in CAD-negative controls (Table 1). For each
SD increment in resistin levels, the odds of CAD increased by
,30% in the GHS cross-sectional design (OR = 1.29, 95% CI:
1.10–1.51, p = 0.002) and by ,80% in the JHS (OR = 1.83, 95%
CI: 1.49–2.24; p = 3.53610212) (Table 2). This association was
unaffected by adjustments for age, sex, smoking habits, BMI,
HbA1c, and insulin, antihypertensive and antidyslipidemic ther-
apies (Table 2). The linear relationship between resistin levels and
CAD risk was confirmed in an analysis by resistin tertiles (Figure 1). In both studies, individuals in the second tertile had an OR of
CAD that was intermediate between the first and third tertiles,
with p-values for linear trend of 0.012 for the GHS-cross sectional
design and 3.76610211 for the JHS (Figure 1). In both prospective studies, the time variable was defined as the
time between the baseline examination and date of the event
(namely, major cardiovascular events for GHS-prospective, and
all-cause mortality for GMS), or, for subjects who did not
experience any event, the date of the last available clinical
follow-up. Incidence rates for the endpoint of interest were
expressed as the number of new events per total number of person-
years (py) and were compared between baseline serum resistin
levels tertiles using a Poisson regression model. In addition, a test
for linear trend in incidence rates over resistin tertiles was
performed by including resistin tertiles as a continuous variable
into the Poisson model. Univariate and multivariate Cox proportional hazards regres-
sions analyses were performed to assess the association between
resistin values or tertiles of its distribution and the event
occurrence. Risks were reported as Hazard Ratios (HR) along
with their 95% CI per SD increase in resistin levels and for tertiles
of its distribution. Test for linear trend in HR estimates over Given the role of resistin in low-grade inflammation, we also
tested plasma fibrinogen as a covariate. After adjustment for this
variable, the association between resistin and CAD in the GHS-
cross sectional design was no longer significant (Table 2). Data on
fibrinogen levels were not available for the JHS. Continuous variables were reported as mean6SD whereas categorical variables were reported as total frequency and percentages. GHS: Gargano Heart Study; JHS:
Joslin Heart Study; CAD: Coronary Artery Disease; BMI: Body Mass Index; HbA1c: glycated haemoglobin. N.A: Not Available.
doi:10 1371/journal pone 0064729 t001 Prospective Studies The GHS-prospective design. The clinical features of study
participants are summarized in Table 3. During follow-up
(5.462.5 years), 58 cardiovascular deaths, 6 non-fatal MIs and 9
non-fatal strokes occurred, corresponding to an overall annual
incidence rate of 3.8% (73 events/1,934 py). Given that this cohort
comprises only very high risk individuals (i.e. diabetic patients who
already suffered by coronary stenosis and/or previous MI), it is not
surprising that most of major cardiovascular events are represent- After stratification by tertiles of baseline resistin levels, the
incidence rate of major cardiovascular events was 2.4% (17
events/711 py) in the first, 3.5% (23 events/665py) in the second,
and 5.9% (33 events/558 py) in the third tertile (p for
trend = 0.001). Accordingly, the HR of major cardiovascular
events progressively increased across resistin tertiles, and persisted
after adjustment for age, sex, smoking habits, BMI, HbA1c and
insulin,
antihypertensive
and
antidyslipidemic
therapies
(HR = 1.68, 95% CI: 0.80 to 3.53 and HR = 2.79, 95% CI: 1.37
to 5.69 in the second and third tertile, respectively; p for
trend = 0.004) (Figure 2). Table 3. Clinical characteristics of patients from prospective
studies. GHS-prospective
design
GMS
n = 359
n = 779
Males (%)
242 (67.4)
397 (50.7)
Age (yrs)
64.568.1
62.169.5
Smokers (%)
160 (44.5)
177 (22.6)
Diabetes duration (yrs)
13.869.2
10.969.1
BMI (kg/m2)
30.264.8
31.065.6
HbA1C(%)
8.661.9
8.761.9
Glucose-lowering therapy
Diet only (%)
23 (6.6)
107 (13.7)
Oral agents (%)
127 (35.4)
326 (41.6)
Insulin w/wo oral agents (%)
194 (54.0)
326 (41.6)
Antihypertensive therapy (%)
305 (85.0)
405 (51.7)
Antidyslipidemic therapy (%)
233 (64.9)
261 (33.3)
Fibrinogen (mg/dl)
385.96124.1
365.36109.6
Resistin (ng/ml)
10.766.6
10.168.1
Continuous variables were reported as mean6SD whereas categorical variables
as total frequency and percentages. GHS: Gargano Heart Study; GMS: Gargano
Mortality Study; BMI: body mass index; HbA1c: glycated haemoglobin. doi:10.1371/journal.pone.0064729.t003 Table 3. Clinical characteristics of patients from prospective
studies. Table 3. Clinical characteristics of patients from prospective
studies. Survival C statistic, IDI and cNRI indices were used to evaluate
the incremental prognostic information of serum resistin for major
cardiovascular events as obtained by the FRS [25] and the
UKPDS risk engine [26]. Time horizon prediction was set to 7
years, replacing the baseline survival probability accordingly. Patients whose information of some clinical variable used in FRS
and/or UKPDS risk engine was not available (n = 61) were
excluded. Case-Control Studies June 2013 | Volume 8 | Issue 6 | e64729 PLOS ONE | www.plosone.org 3 Resistin and RETN Levels in Type 2 Diabetes Figure 1. Odds ratios (95% CI) of CAD in cross sectional studies, according to baseline tertile (T1–T3, range in parentheses) of
resistin levels. ORs were estimated by logistic regression after adjusting for age, sex, smoking habits, BMI, HbA1c and insulin, antihypertensive and
antidyslipidemic therapies. doi:10.1371/journal.pone.0064729.g001 Figure 1. Odds ratios (95% CI) of CAD in cross sectional studies, according to baseline tertile (T1–T3, range in parentheses) of
resistin levels. ORs were estimated by logistic regression after adjusting for age, sex, smoking habits, BMI, HbA1c and insulin, antihypertensive and
antidyslipidemic therapies. doi:10.1371/journal.pone.0064729.g001 Table 1. Clinical Characteristics of patients from case-control studies. Table 1. Clinical Characteristics of patients from case-control studies. p
GHS-cross sectional design
JHS
CAD Negative
CAD Positive
CAD Negative
CAD Positive
n = 416
n = 360
n = 443
n = 418
Males (%)
184 (44.2)
246 (68.3)
250 (56.4)
304 (72.7)
Age (yrs)
59.968.6
64.468.1
64.366.3
64.367.5
Smokers (%)
122 (29.3)
160 (44.4)
169 (38.1)
274 (65.6)
Diabetes duration (yrs)
11.168.3
13.869.2
12.666.7
12.768.8
BMI (kg/m2)
31.565.3
30.264.8
32.365.6
32.266.0
HbA1c (%)
8.561.9
8.761.9
7.361.2
7.461.4
Glucose-lowering therapy
Diet only (%)
41 (9.9)
22 (6.1)
30 (6.8)
24 (5.8)
Oral agents (%)
215 (51.7)
128 (35.6)
225 (51.3)
188 (45.4)
Insulin w/wo oral agents (%)
142 (34.1)
195 (54.2)
184 (41.9)
202 (48.8)
Antihypertensive therapy (%)
279 (67.1)
306 (85.0)
323 (72.9)
365 (87.3)
Antidyslipidemic therapy (%)
141 (33.9)
235 (65.3)
312 (70.4)
343 (82.0)
Fibrinogen (mg/dl)
348.6687.8
389.36128.7
N.A. N.A. Resistin (ng/ml)
9.3365.03
10.7266.68
6.4664.25
8.6465.99
Continuous variables were reported as mean6SD whereas categorical variables were reported as total frequency and percentages. GHS: Gargano Heart Study; JHS:
Joslin Heart Study; CAD: Coronary Artery Disease; BMI: Body Mass Index; HbA1c: glycated haemoglobin. N.A: Not Available. doi:10.1371/journal.pone.0064729.t001 June 2013 | Volume 8 | Issue 6 | e64729 PLOS ONE | www.plosone.org 4 Resistin and RETN Levels in Type 2 Diabetes Table 2. Association between serum resistin levels and CAD occurrence in case-control studies. Case-Control Studies GHS-cross sectional design
JHS
CAD Negative n = 416
CAD Positive n = 360
CAD Negative n = 443
CAD Positive n = 418
OR (95% CI)
p-value
OR (95% CI)
p-value
Model 1
1.29 (1.10–1.51)
0.002
1.83 (1.49–2.24)
3.53610212
Model 2
1.35 (1.10–1.64)
0.002
1.99 (1.55–2.55)
2.95610213
Model 3
1.10 (0.93–1.30)
0.260
N.A. N.A. GHS: Gargano Heart Study; JHS: Joslin Heart Study; CAD: Coronary Artery Disease. OR (95% CI) are given for SD increase of resistin levels. Model 1: unadjusted. Model 2: adjusted for age, sex, smoking habits, BMI, HbA1c, insulin therapy, hypertension and dyslipidemia. Model 3: adjusted for plasma fibrinogen. N.A: Not Available. doi:10.1371/journal.pone.0064729.t002 ed by death. Each SD increment of serum resistin levels was
associated with a 31% increase in the risk of major cardiovascular
events (HR = 1.31, 95% CI: 1.13–1.53; p = 3.38610 –4). As in the
case-control studies, the association was not affected by adjustment
for
age,
sex,
smoking
habits,
BMI,
HbA1c,
and
insulin,
antihypertensive and antidyslipidemic therapies (HR = 1.31, 95%
CI: 1.10–1.56; p = 0.003), but was attenuated and lost significance
after adjusting for fibrinogen levels (HR = 1.18, 95% CI: 0.97–
1.45; p = 0.099). It is of note that serum resistin levels in CAD negative controls
from both GHS-prospective design and JHS were significantly
higher than those previously measured in our laboratory [23] in
non diabetic controls (p,0,001 and ,0,01, respectively; data not
shown). Prospective Studies FRS did not perform well in our sample with survival C statistic
being equal to 0.584 (95% CI: 0.510–0.657). The addition of
serum resistin produced a significant (p = 0.028) improvement,
with survival C statistic becoming 0.640 (95% CI: 0.568–0.713). Both models resulted well calibrated (HL p-values being 0.532 and
0.256, respectively). Moreover, a significant improvement in
discrimination was also detected by IDI: 0.022, 95% CI: 0.004–
0.048, p = 0.003. Finally, the addition of serum resistin to the FRS
allowed to reclassify correctly 95/298 patients (cNRI = 0.433,
p = 0.006): 92/244 (37.7%) and 3/54 (5.6%) in those without and
with incident events, respectively. The UKPDS risk engine too performed poorly (survival C
statistic = 0.674; 95% CI: 0.607–0.741). The addition of serum
resistin produced a significant (p = 0.025) improvement (survival C
statistic = 0.704; 95% CI: 0.641–0.766). Both models resulted well
calibrated (HL p-values being 0.169 and 0.176, respectively). In
contrast, resistin addition did not result in a significant IDI (0.005; June 2013 | Volume 8 | Issue 6 | e64729 June 2013 | Volume 8 | Issue 6 | e64729 PLOS ONE | www.plosone.org PLOS ONE | www.plosone.org 5 Figure 2. Survival curves for major cardiovascular events in the GHS-prospective design, according to baseline tertile (T1–T3, range
in parentheses) of resistin levels. Curves are estimated by Cox regression after adjusting for age, sex, smoking habits, BMI, HbA1c and insulin,
antihypertensive and antidyslipidemic therapies. doi:10.1371/journal.pone.0064729.g002
Resistin and RETN Levels in Type 2 Diabetes Resistin and RETN Levels in Type 2 Diabetes yp Figure 2. Survival curves for major cardiovascular events in the GHS-prospective design, according to baseline tertile (T1–T3, range
in parentheses) of resistin levels. Curves are estimated by Cox regression after adjusting for age, sex, smoking habits, BMI, HbA1c and insulin,
antihypertensive and antidyslipidemic therapies. doi:10.1371/journal.pone.0064729.g002 95% CI: 20.01–0.024, p = 0.286). Finally, the addition of serum
resistin to the UKPDS risk engine allowed to reclassify correctly
98/298 patients (cNRI = 0.459, p = 0.004): 94/244 (38.5%) and
4/54 (7.4%) in those without and with incident events, respec-
tively. adjusted HRs for all-cause mortality were similar in the second
and third tertile (HRs = 1.72, 95% CI: 1.11 to 2.68, and 1.81, 95%
CI: 1.16 to 2.82, respectively, p for trend = 0.019) (Figure 3). Correlation between Serum and mRNA Resistin Levels Since the data above indicate that serum resistin is a marker of
cardiovascular risk in patients with type 2 diabetes, we measured
RETN mRNA levels in circulating PWBC and serum resistin levels
in 68 diabetic patients in order to obtain mechanistic insights on
the biology of serum resistin variability. The two variables were
positively correlated (rho = 0.343, p = 0.006) (Figure S2). Serum resistin predicted the risk for all-cause death with an HR
of 1.18 (95% CI: 1.12–1.27; p = 1.561026) per SD increment. One hundred and three GMS participants overlapped with those
of GHS-prospective design. Exclusion of these subjects did not
substantially alter the results (HR = 1.17, 95% CI: 1.07–1.27;
p = 0.00047). Similar results were obtained in the whole cohort
after adjusting for age, sex, smoking habits, BMI, HbA1c and
insulin,
antihypertensive
and
antidyslipidemic
therapies
(HR = 1.16, 95% CI: 1.06–1.26; p = 0.001). Prospective Studies Also
at odds with the other studies, the association between resistin and
mortality remained significant, though less robustly, after adjusting
for fibrinogen levels (HR per resistin SD = 1.15, 95% CI: 1.05–
1.27; p = 0.03). The GMS. Given that CVD is the main cause of death among
patients with type 2 diabetes, we assessed the role of serum resistin
in predicting all-cause mortality in a cohort of such patients. The
clinical features of the GMS participants are summarized in
Table 3. During follow-up (7.662.1 years), 150 deaths occurred,
corresponding to an annual incidence rate of 2.7% (150 events/
5,659 py). The GMS. Given that CVD is the main cause of death among
patients with type 2 diabetes, we assessed the role of serum resistin
in predicting all-cause mortality in a cohort of such patients. The
clinical features of the GMS participants are summarized in
Table 3. During follow-up (7.662.1 years), 150 deaths occurred,
corresponding to an annual incidence rate of 2.7% (150 events/
5,659 py). June 2013 | Volume 8 | Issue 6 | e64729 Discussion Despite improvements in the treatment of CVD and in the
control of risk factors, individuals with type 2 diabetes remain at
increased
cardiovascular
risk
as
compared
to
the
general
population [32] with cardiovascular events being the most
important cause of death in these patients [1]. Discovering novel
biomarkers able to predict CVD in diabetic patients is therefore
urgently needed to decrease the burden of this devastating
complication. The all-cause mortality rates in the first, second, and third tertile
of resistin levels were 1.8% (36 events/2,022 py), 2.8% (52 events/
1,862 py), and 3.5% (62 events/1,775 py), respectively (p for
trend = 0.001). At variance with the linear trend observed in both
cross-sectional studies and the GHS-prospective design, the June 2013 | Volume 8 | Issue 6 | e64729 PLOS ONE | www.plosone.org 6 Figure 3. Survival curves for all-cause mortality in the GMS, according to baseline tertile (T1–T3, range in parentheses) of resisti
levels. Curves are estimated by Cox regression after adjusting for age, sex, smoking habits, BMI, HbA1c and insulin, antihypertensive an
tid
li id
i th
i
Resistin and RETN Levels in Type 2 Diabete Figure 3. Survival curves for all-cause mortality in the GMS, according to baseline tertile (T1–T3, range in parentheses) of resistin
levels. Curves are estimated by Cox regression after adjusting for age, sex, smoking habits, BMI, HbA1c and insulin, antihypertensive and
antidyslipidemic therapies. doi:10.1371/journal.pone.0064729.g003
Resistin and RETN Levels in Type 2 Diabetes Resistin and RETN Levels in Type 2 Diabetes Figure 3. Survival curves for all-cause mortality in the GMS, according to baseline tertile (T1–T3, range in parentheses) of resistin
levels. Curves are estimated by Cox regression after adjusting for age, sex, smoking habits, BMI, HbA1c and insulin, antihypertensive and
antidyslipidemic therapies. doi:10.1371/journal.pone.0064729.g003 To the best of our knowledge, this is the first study to show that
elevated serum resistin concentration is a risk factor for CVD in
patients with type 2 diabetes of European ancestry. While a rich
literature exists on serum resistin as a cardiovascular risk factor in
the general population [13,14,15,16,17], data concerning the type
2 diabetes population have been thus far sparse, contradictory,
and limited to Asian individuals [9,11,18]. Discussion Two small cross-
sectional studies of patients with type 2 diabetes from Japan and
Korea described an association of serum resistin with CAD and
stroke, respectively, but a third study on 343 diabetic Korean
patients failed to confirm such findings in a prospective setting. By
contrast, we have obtained strong and consistent evidence of
association between serum resistin and CVD from both cross-
sectional and prospective studies on patients with type 2 diabetes
of European ancestry. This effect of resistin is independent of the
most established cardiovascular risk factor including sex, smoking
habits, BMI, HbA1c and insulin, antihypertensive and antidyslipi-
demic therapies. However, if the analysis is adjusted for fibrinogen,
the association is no longer significant, suggesting it is mediated at
least in part by low-grade inflammation – an established
cardiovascular risk factor. Whether resistin is a true risk factor
which may causally contribute to CVD or, in contrast, a simple
biomarker of pro-inflammatory status, cannot be addressed by our
study. as obtained by well established models such as the FRS and the
UKPDS risk engine [25,26] which, similarly to other models, are
known not to perform well in the subset of patients with type 2
diabetes [33,34]. Thus, it was not unexpected that both models
performed poorly also in our sample in predicting major
cardiovascular events. Of note, the addition of serum resistin
improved the two models in terms of both discriminatory and
reclassification performance. Such improvement was not only
statistically significant, but also of clinical relevance. Further larger
studies, are needed to deeper address the relative importance of
resistin as an additional marker of clinical utility. for predicting
CVD in patients with type 2 diabetes. Our study also shows for the first time that serum resistin is an
independent predictor of all-cause mortality in a study comprising
779 patients with type 2 diabetes. In contrast to what was observed
with CVD risk, the association with mortality is only modestly
affected by adjustment for fibrinogen levels. Therefore, though
fibrinogen is not the best marker of low-grade inflammation, it
may be hypothesized that the effects of resistin on all-cause
mortality are mediated by mechanisms that are independent of
this pathway. June 2013 | Volume 8 | Issue 6 | e64729 Resistin and RETN Levels in Type 2 Diabetes Resistin and RETN Levels in Type 2 Diabetes Resistin and RETN Levels in Type 2 Diabetes mortality observed in the general population [35,36,37] seems to
be independent from cardiovascular mortality [35,36]. Unfortu-
nately, data on cause of death that could confirm this finding
among type 2 diabetes patients were not available in the GMS. Finally, whether our finding can be generalized to other
populations of different ethnicity having different environmental
and/or genetic background remains to be established. This issue
deserves particular attention given that a different genetic
regulation in different ethnic groups has been hypothesized for
serum resistin [40]. Therefore, additional studies are certainly
needed to confirm our present finding in a broader context. An additional finding of our study is that RETN mRNA in
PWBC is correlated to serum resistin. This result, which is
consistent with the observation that human resistin is mainly
produced by macrophages [5], strongly suggest that resistin
circulating levels are modulated by gene expression levels. The
mechanism(s) underlying such modulation are not yet known and
need further studies to be unraveled. In conclusion, our study is the first to show that high serum
resistin (a likely consequence, at least in part, of increased resistin
mRNA expression) is a risk factor for CVD and all-cause mortality
in patients with type 2 diabetes of European ancestry. Further
studies are warranted to determine whether this biomarker can be
used in a clinical setting to improve the stratification of diabetic
patients with regard to their risk for CVD and death. One strength of our study is the overall sample size, consisting of
a total of 2,313 diabetic patients from two cross-sectional and two
prospective
investigations,
and
the
completeness
of
clinical
information, including standardized clinical evaluations and hard
end-points validated by medical records or death certificates. Another strength is the fact that the resistin measurements were
centralized and all the samples were handled identically. In this
context, the observed difference in serum resistin concentration
between the two cross-sectional studies, with JHS participants
having 20–30% lower mean levels as compared to GHS
individuals, is somewhat surprising. One possibility is that such
difference was due to the different proportion of patients treated
with lipid lowering agents in the two studies (77% in the JHS as
compared to 48% in the GHS). Such agents are mainly statins,
which are known to decrease serum resistin levels [38]. 12. Lubos E, Messow CM, Schnabel R, Rupprecht HJ, Espinola-Klein C, et al.
(2007) Resistin, acute coronary syndrome and prognosis results from the
AtheroGene study. Atherosclerosis 193: 121–128. 11. Tsukahara T, Nakashima E, Watarai A, Hamada Y, Naruse K, et al. (2009)
Polymorphism in resistin promoter region at 2420 determines the serum resistin
levels and may be a risk marker of stroke in Japanese type 2 diabetic patients.
Diabetes Res Clin Pract 84: 179–186. Acknowledgments Of note, resistin levels in CAD-negative controls from both
GHS-cross sectional and JHS were clearly higher than those from
non diabetic controls [23]. We are indebted to the staffs and participants of the GHS, JHS and GMS
for their dedication and contributions. The major limitation of our study is represented by the lack of
C-reactive protein measurements which surely would have help
clarify the link between resistin and chronic inflammation state. In
fact, the role of fibrinogen on inflammation remains largely
speculative, thus making not possible to draw firm conclusions
about the biology underlying the association we observed in our
present study. (TIF) Figure S2
Correlation between serum and mRNA resistin levels. Correlation between serum and mRNA resistin levels in 68
patients with type 2 diabetes. RETN mRNA levels in PWBC are
expressed as arbitrary units (AU) of RETN/GAPDH ratios. (TIF) Despite these differences, the fact that the association between
serum resistin and CAD, that was found in the GHS-cross
sectional was fully replicated in the JHS makes our finding
especially convincing. Author Contributions Conceived and designed the experiments: C. Menzaghi SB MF SDC AD
VT. Performed the experiments: LS C. Mendonca GP CDB AM SP EM
SR GF DM. Analyzed the data: C. Menzaghi SB GP AF FP AD. Contributed reagents/materials/analysis tools: SB EG AK MF SDC FP
AD. Wrote the paper: C. Menzaghi AD VT. Resistin and RETN Levels in Type 2 Diabetes This
hypothesis is supported by the observation that participants in the
GHS-cross sectional design who were on statins had serum resistin
levels 10–15% lower than patients who were not on statins (data
not shown). At variance, given that no substantial effects of fasting
on resistin levels have been described [39], we can exclude that the
observed difference in serum resistin levels between GHS and JHS
is due to the different fasting status of the two studies. 7. Langheim S, Dreas L, Veschini L, Maisano F, Foglieni C, et al. (2010) Increased
expression and secretion of resistin in epicardial adipose tissue of patients with
acute coronary syndrome. Am J Physiol Heart Circ Physiol 298: H746–753. 10. Ohmori R, Momiyama Y, Kato R, Taniguchi H, Ogura M, et al. (2005)
Associations between serum resistin levels and insulin resistance, inflammation,
and coronary artery disease. J Am Coll Cardiol 46: 379–380. 3. Scherer PE (2006) Adipose tissue: from lipid storage compartment to endocrine
organ. Diabetes 55: 1537–1545. 4. Steppan CM, Bailey ST, Bhat S, Brown EJ, Banerjee RR, et al. (2001) The
hormone resistin links obesity to diabetes. Nature 409: 307–312. 2. Haffner SM (2006) The metabolic syndrome: inflammation, diabetes mellitus,
and cardiovascular disease. Am J Cardiol 97: 3A–11A. Discussion The possibility that different mechanisms underlie
the effects of resistin on CVD and all-cause mortality is also
suggested by the fact that the relationship between resistin levels
and all-cause mortality does not appear to be linear as that
between resistin and CAD or major CVD events. Consistent with
such hypothesis, the association between resistin and all-cause We wanted to investigate whether serum resistin provides
incremental information in predicting major cardiovascular events June 2013 | Volume 8 | Issue 6 | e64729 PLOS ONE | www.plosone.org June 2013 | Volume 8 | Issue 6 | e64729 PLOS ONE | www.plosone.org 7 Supporting Information Figure S1
Diagrams and participants of Gargano Heart Study
and Gargano Mortality Study in whom serum resistin levels were
available. Gargano Heart Study (GHS)-cross sectional design includes 776
European subjects with type 2 diabetes mellitus (T2DM), 360
CAD positive and 416 CAD negative as defined in methods. GHS-
prospective design comprises 359 patients with T2D and CAD who
were all case participants of the GHS-cross sectional design as
described in methods. Gargano Mortality Study (GMS) comprises 779
patients with T2D as describes in methods. One hundred and
three of the GMS participants are also participants of the GHS-
prospective design (gray box). Raw data on resistin levels and
association with related variables can be provided upon request for
collaborative purposes. 6. Burnett MS, Lee CW, Kinnaird TD, Stabile E, Durrani S, et al. (2005) The
potential role of resistin in atherogenesis. Atherosclerosis 182: 241–248. 1. Seshasai SR, Kaptoge S, Thompson A, Di Angelantonio E, Gao P, et al. (2011)
Diabetes mellitus, fasting glucose, and risk of cause-specific death. N Engl J Med
364: 829–841. References Uno H, Cai T, Pencina MJ, D’Agostino RB, Wei LJ (2011) On the C-statistics
for evaluating overall adequacy of risk prediction procedures with censored
survival data. Stat Med 30: 1105–1117. 15. Frankel DS, Vasan RS, D’Agostino RB Sr, Benjamin EJ, Levy D, et al. (2009)
Resistin, adiponectin, and risk of heart failure the Framingham offspring study. J Am Coll Cardiol 53: 754–762. 29. Pencina MJ, D’Agostino RB (2004) Overall C as a measure of discrimination in
survival analysis: model specific population value and confidence interval
estimation. Stat Med 23: 2109–2123. 16. Butler J, Kalogeropoulos A, Georgiopoulou V, de Rekeneire N, Rodondi N, et
al. (2009) Serum resistin concentrations and risk of new onset heart failure in
older persons: the health, aging, and body composition (Health ABC) study. Arterioscler Thromb Vasc Biol 29: 1144–1149. 30. Pencina MJ, D’Agostino RB Sr, D’Agostino RB Jr, Vasan RS (2008) Evaluating
the added predictive ability of a new marker: from area under the ROC curve to
reclassification and beyond. Stat Med 27: 157–172; discussion 207–112. 17. Luc G, Empana JP, Morange P, Juhan-Vague I, Arveiler D, et al. (2010)
Adipocytokines and the risk of coronary heart disease in healthy middle aged
men: the PRIME Study. Int J Obes (Lond) 34: 118–126. 31. Pencina MJ, D’Agostino RB, Steyerberg EW (2011) Extensions of net
reclassification improvement calculations to measure usefulness of new
biomarkers. Stat Med 30: 11–21. 18. Lim S, Koo BK, Cho SW, Kihara S, Funahashi T, et al. (2008) Association of
adiponectin and resistin with cardiovascular events in Korean patients with type
2 diabetes: the Korean atherosclerosis study (KAS): a 42-month prospective
study. Atherosclerosis 196: 398–404. 32. Preis SR, Pencina MJ, Hwang SJ, D’Agostino RB Sr, Savage PJ, et al. (2009)
Trends in cardiovascular disease risk factors in individuals with and without
diabetes mellitus in the Framingham Heart Study. Circulation 120: 212–220. diabetes mellitus in the Framingham Heart Study. Circulation 120: 2 19. Bacci S, Menzaghi C, Ercolino T, Ma X, Rauseo A, et al. (2004) The +276 G/T
single nucleotide polymorphism of the adiponectin gene is associated with
coronary artery disease in type 2 diabetic patients. Diabetes Care 27: 2015–
2020. 33. Coleman RL, Stevens RJ, Retnakaran R, Holman RR (2007) Framingham,
SCORE, and DECODE risk equations do not provide reliable cardiovascular
risk estimates in type 2 diabetes. Diabetes Care 30: 1292–1293. yp
34. References 8. Pischon T, Bamberger CM, Kratzsch J, Zyriax BC, Algenstaedt P, et al. (2005)
Association of plasma resistin levels with coronary heart disease in women. Obes
Res 13: 1764–1771. 1. Seshasai SR, Kaptoge S, Thompson A, Di Angelantonio E, Gao P, et al. (2011)
Diabetes mellitus, fasting glucose, and risk of cause-specific death. N Engl J Med
364: 829–841. 9. On YK, Park HK, Hyon MS, Jeon ES (2007) Serum resistin as a biological
marker for coronary artery disease and restenosis in type 2 diabetic patients. Circ J 71: 868–873. 2. Haffner SM (2006) The metabolic syndrome: inflammation, diabetes mellitus,
and cardiovascular disease. Am J Cardiol 97: 3A–11A. 3. Scherer PE (2006) Adipose tissue: from lipid storage compartment to endocrine
organ. Diabetes 55: 1537–1545. 10. Ohmori R, Momiyama Y, Kato R, Taniguchi H, Ogura M, et al. (2005)
Associations between serum resistin levels and insulin resistance, inflammation,
and coronary artery disease. J Am Coll Cardiol 46: 379–380. 4. Steppan CM, Bailey ST, Bhat S, Brown EJ, Banerjee RR, et al. (2001) The
hormone resistin links obesity to diabetes. Nature 409: 307–312. 5. Reilly MP, Lehrke M, Wolfe ML, Rohatgi A, Lazar MA, et al. (2005) Resistin is
an inflammatory marker of atherosclerosis in humans. Circulation 111: 932–
939. 6. Burnett MS, Lee CW, Kinnaird TD, Stabile E, Durrani S, et al. (2005) The
potential role of resistin in atherogenesis. Atherosclerosis 182: 241–248. June 2013 | Volume 8 | Issue 6 | e64729 PLOS ONE | www.plosone.org 8 Resistin and RETN Levels in Type 2 Diabetes 13. Hoefle G, Saely CH, Risch L, Koch L, Schmid F, et al. (2007) Relationship
between the adipose-tissue hormone resistin and coronary artery disease. Clin
Chim Acta 386: 1–6. 26. Stevens RJ, Kothari V, Adler AI, Stratton IM, Group UKPDS (2001) The
UKPDS risk engine: a model for the risk of coronary heart disease in Type II
diabetes (UKPDS 56). Clin Sci (Lond) 101: 671–679. (
)
(
)
27. D’Agostino R, Nam BH (2004) Evaluation of the performance of survival
analysis models: discrimination and calibration measures. Handbook of
Statistics, vol 23 Elsevier Science BV. 14. Weikert C, Westphal S, Berger K, Dierkes J, Mohlig M, et al. (2008) Plasma
resistin levels and risk of myocardial infarction and ischemic stroke. J Clin
Endocrinol Metab 93: 2647–2653. 28. References Chamnan P, Simmons RK, Sharp SJ, Griffin SJ, Wareham NJ (2009)
Cardiovascular risk assessment scores for people with diabetes: a systematic
review. Diabetologia 52: 2001–2014. 20. Prudente S, Morini E, Larmon J, Andreozzi F, Di Pietro N, et al. (2011) The
SH2B1 obesity locus is associated with myocardial infarction in diabetic patients
and with NO synthase activity in endothelial cells. Atherosclerosis 219: 667–672. g
35. Lee SH, Ha JW, Kim JS, Choi EY, Park S, et al. (2009) Plasma adiponectin and
resistin levels as predictors of mortality in patients with acute myocardial
infarction: data from infarction prognosis study registry. Coron Artery Dis 20:
33–39. 21. Qi L, Parast L, Cai T, Powers C, Gervino EV, et al. (2011) Genetic susceptibility
to coronary heart disease in type 2 diabetes: 3 independent studies. J Am Coll
Cardiol 58: 2675–2682. 22. Bacci S, Rizza S, Prudente S, Spoto B, Powers C, et al. (2011) The ENPP1 Q121
variant predicts major cardiovascular events in high-risk individuals: evidence
for interaction with obesity in diabetic patients. Diabetes 60: 1000–1007. 36. Pilz S, Weihrauch G, Seelhorst U, Wellnitz B, Winkelmann BR, et al. (2007)
Implications of resistin plasma levels in subjects undergoing coronary
angiography. Clin Endocrinol (Oxf) 66: 380–386. 37. Zhang MH, Na B, Schiller NB, Whooley MA (2011) Association of resistin with
heart failure and mortality in patients with stable coronary heart disease: data
from the heart and soul study. J Card Fail 17: 24–30. 23. Menzaghi C, Coco A, Salvemini L, Thompson R, De Cosmo S, et al. (2006)
Heritability of serum resistin and its genetic correlation with insulin resistance-
related features in nondiabetic Caucasians. J Clin Endocrinol Metab 91: 2792–
2795. 38. von Eynatten M, Schneider JG, Hadziselimovic S, Hamann A, Bierhaus A, et al. (2005) Adipocytokines as a novel target for the anti-inflammatory effect of
atorvastatin in patients with type 2 diabetes. Diabetes Care 28: 754–755. 24. Ghali WA, Quan H, Brant R, van Melle G, Norris CM, et al. (2001)
Comparison of 2 methods for calculating adjusted survival curves from
proportional hazards models. JAMA 286: 1494–1497. 39. Weikert C, Westphal S, Luley C, Willich SN, Boeing H, et al. (2007) Within-
subject variation of plasma resistin levels over a 1-year period. Clin Chem Lab
Med 45: 899–902. p
p
25. D’Agostino RB, Vasan RS, Pencina MJ, Wolf PA, Cobain M, et al. References (2008)
General cardiovascular risk profile for use in primary care: the Framingham
Heart Study. Circulation 117: 743–753. 40. Menzaghi C, Trischitta V (2010) Genetics of serum resistin: a paradigm of
population-specific regulation? Diabetologia 53: 226–228. PLOS ONE | www.plosone.org June 2013 | Volume 8 | Issue 6 | e64729 9
|
https://openalex.org/W2883528982
|
https://heritagesciencejournal.springeropen.com/track/pdf/10.1186/s40494-018-0209-y
|
English
| null |
A novel LC–MS method using collagen marker peptides for species identification of glue applicable to samples with multiple animal origins
|
Heritage science
| 2,018
|
cc-by
| 6,358
|
© The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License
(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium,
provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license,
and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/
publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Abstract Animal glues derived from collagen-rich tissue are widely used as adhesives and binders in artworks. Identification of
the animal source of glue is important not only to understand the artistic technique and historical background, but
also to conduct appropriate treatment for restoration and conservation of artworks. Recently, mass spectrometric
approaches have been used for species identification of glue. However, it appears that the existing methods were
not applied to samples containing multiple animal sources, and the range of target species has been limited because
of incomplete sequence information of collagen in public databases. In the present study, we established a novel
method for discrimination of glues from eight animals applicable to samples with multiple animal origins. Trypsin-
digested glue samples were analyzed by LC–MS in multiple reaction monitoring mode, and the animal source was
determined based on the detection patterns of 12 type I collagen-derived marker peptides. Using the database-
independent method, we successfully identified the animal source of commercial products and found the declared
species for some glues to be incorrect. In addition, some products were identified to contain two different animal
origins. Relative abundance of the animal origins in several impure glues was estimated using the marker peptides,
which helped to speculate the reason for the detection of multiple species. We analyzed a painting (The Harvest by
Camille Pissarro, 1882) and clarified that glues used in the ground and size layers of the canvas were derived from cat-
tle (~ 65%) and sheep (~ 35%). Keywords: Glue, Collagen, LC–MS, Species identification, Marker peptide A novel LC–MS method using collagen
marker peptides for species identification
of glue applicable to samples with multiple
animal origins Yuki Kumazawa1, Yuki Taga2* , Miho Takashima3 and Shunji Hattori1,2 Kumazawa et al. Herit Sci (2018) 6:43
https://doi.org/10.1186/s40494-018-0209-y Kumazawa et al. Herit Sci (2018) 6:43
https://doi.org/10.1186/s40494-018-0209-y Open Access *Correspondence: y‑taga@nippi‑inc.co.jp
2 Nippi Research Institute of Biomatrix, 520‑11 Kuwabara, Toride, Ibaraki
302‑0017, Japan
Full list of author information is available at the end of the article Introduction and craftsmen use animal glue for various purposes; for
example, as adhesives for wooden products and paper, as
binding media for pigments in paint and ground layers,
and as sizings for paper, panel, and canvas. Animal glue
exhibits varied physical, chemical, and mechanical prop-
erties depending on its biological origin and the method
used to prepare it [2]. Therefore, in addition to determin-
ing the use of glue, identification of its animal source
provides an understanding of artistic material/technique
and context of the era. Such information is also helpful
in identifying appropriate treatments for restoration and
conservation of artworks. Skin and bone of mammalians and fish are mainly com-
posed of type I collagen, consisting of Gly–Xaa–Yaa
repeats, where Xaa and Yaa are frequently Pro and
4-hydroxyproline (4-Hyp), respectively. Through heat-
ing in boiling water, collagen is extracted from these tis-
sues as a partially hydrolyzed form, referred to as gelatin
or glue. Animal glue was used by Egyptians more than
4000 years ago and has been commercially manufactured
since the 1700s [1]. In the preparation of artworks, artists *Correspondence: y‑taga@nippi‑inc.co.jp
2 Nippi Research Institute of Biomatrix, 520‑11 Kuwabara, Toride, Ibaraki
302‑0017, Japan
Full list of author information is available at the end of the article Detection of collagen-based glue in artworks with dif-
ferentiation from other types of proteinaceous binders,
such as milk casein and egg, can be performed based Kumazawa et al. Herit Sci (2018) 6:43 Page 2 of 9 Page 2 of 9 to unambiguously determine the animal source of glue
even when multiple-species glues are used. on the high content of Gly and Hyp estimated by HPLC,
GC–MS, or capillary zone electrophoresis after acid
hydrolysis [3, 4]. However, these methods are unsuitable
for identification of the animal source of glue because of
the loss of species specificity in the amino acid sequence. Some studies have reported an immunology-based
approach using ELISA and a DNA-based approach using
PCR for species identification of glue in artworks [5, 6]. ELISA can selectively and sensitively detect the use of
glue [5], but cross-reaction of antibodies is a concern
because of similarities in the amino acid sequences of
collagen among mammals. In contrast, PCR amplifica-
tion of mitochondrial DNA has high taxonomic speci-
ficity for tracing the animal origin of glue [6], but DNA
molecules are relatively unstable and can be degraded
during glue manufacture and aging after use. Introduction p
p
g
Collagen-based materials, such as leather, gelatin, and
glue, are prepared from various animals. However, col-
lagen sequence information of these domestic animals
is scarce in public databases, which makes species iden-
tification difficult [16, 20]. For example, only partial
sequences of rabbit type I collagen (53 and 526 amino
acids for α1(I) and α2(I) chains, respectively) are reg-
istered in the UniProt database as reviewed sequence
data, and deer α1(I) and α2(I) chains are completely
unknown. We recently established a novel approach for
species identification of leather using LC–MS [21]. Using
authentic leathers, we first performed MS/MS identifica-
tion of type I collagen-derived peptides using currently
available sequence information after trypsin digestion. Among many identified peptides, six tryptic marker pep-
tides were selected with no regard for the animal species
of type I collagen to give different and species-specific
detection patterns between cattle, horse, pig, sheep, goat,
and deer. Detection of the marker peptides by LC–MS
in multiple reaction monitoring (MRM) mode enabled
simple certification of the animal source of leather. MRM
analysis based on MS/MS fragmentation can highly
selectively detect target analytes by using a predefined
precursor ion and its fragments [22]. Of note, this data-
base-independent methodology can be applied to various
animals regardless of whether their collagen sequences
are publicly available. In the present study, we applied
this approach to species identification of glues from eight
animals, including cattle, horse, pig, sheep, goat, deer,
rabbit, and sturgeon. In contrast to leather, there is the
potential for a mixture of multiple sources for glue. Thus,
we further designed the method to give different detec-
tion patterns even between samples containing two spe-
cies. Commercial animal glues and The Harvest painted
by Camille Pissarro were analyzed using the newly estab-
lished method. g g
g
g
Recently, mass spectrometric approaches have been
increasingly used for species identification for various
kinds of samples, including bone [7, 8], meat [9, 10], milk
[11, 12], and gelatin [13, 14]. Target proteins are enzy-
matically digested into peptide fragments, typically using
trypsin. MS can simultaneously detect the resulting pep-
tides with high sensitivity, which allows discrimination
of the animal species based on interspecies differences in
the amino acid sequence. As for glue used in artworks,
Kuckova et al. successfully identified the animal origin
using peptide mass fingerprinting with MALDI–TOF–
MS detection [15]. Introduction While this analysis can be straight-
forward using the detected peak pattern in a single
mass spectrum, unambiguous species determination
is difficult, especially for complex samples. Proteomic
approaches with database search of LC–MS/MS spectra
also have been reported for species identification of glue
[16–19]. This strategy enables more reliable determina-
tion of the animal species because of the exact sequence
information of the identified peptides determined based
on their MS/MS fragmentation patterns. Dallongeville
et al. used species-specific peptides identified for type I
and type III collagens to discriminate between bovine,
rabbit, and sturgeon glues [17]. Using this method, they
identified bovine glue in an 18th century gilt sample and
found that bovine-derived glue was supplied as rabbit
glue for some products. However, identification of spe-
cies-specific peptides becomes challenging when extend-
ing the target species, and all the above existing methods
were assumed not to be applied to samples containing
multiple glues derived from different animals. It is known
that some suppliers provide rabbit glue as a blend with
bovine glue to alter its properties [2]. In addition, con-
tamination with other glues is possible during and after
the manufacturing process. Furthermore, blending tech-
niques or multiple use of glues derived from different ori-
gins may be used in artworks. Therefore, it is important Glue samples defined to be 95%. We have deposited the MS datasets to
the ProteomeXchange consortium via the jPOST partner
repository with the dataset identifier PXD009059 [25,
26]. Commercial animal glues were purchased from several
suppliers. The glue samples were dissolved in distilled
water at 60 °C for 4 h and stored at 4 °C until analysis. Standard samples Cattle, horse, and pig collagens were purified from the
skin of respective animals by pepsin digestion and salt
precipitation as described previously [23]. Sheep colla-
gen and collagen from rabbit skin were purchased from
Sigma-Aldrich (St. Louis, MO). Deer glue prepared from
deer skin (Cervus nippon) was obtained from Amanosan
Cultural Heritages Research Institute (Osaka, Japan), and
sturgeon glue made from bladder was kindly gifted from
Prof. Izuru Seki (Tokyo University of the Arts). A certi-
fied chrome-tanned goat leather was used as a standard
sample after dechroming as described previously [21]. Kumazawa et al. Herit Sci (2018) 6:43 Page 3 of 9 Analysis of The Harvest Samples from two separate locations (approximately
1 mg each) were taken from the tacking margin of
the canvas of The Harvest (Camille Pissarro, 1882,
70.3 × 126 cm, glue tempera on canvas), which is part of
the collection of The National Museum of Western Art
(Tokyo, Japan; inventory number P. 1984-0003). The sam-
ples included ground layers, size layers, and fibers of the
canvas. The samples were digested with 2.5 µg of trypsin
as described above after heating at 60 °C for 30 min in
100 µL of 100 mM Tris–HCl/1 mM CaCl2 (pH 7.6). After
acidification of the reactant with formic acid, the super-
natants were subjected to MRM analysis. The relative abundance of the animal origins in glue
samples or The Harvest was estimated using peak areas
of MRM chromatograms of marker peptides detected for
either but not both of identified two animal species (P11
for cattle, P9 for rabbit, and P5 for sheep). Type I collagen
concentrations in standard samples of cattle, rabbit, and
sheep were predetermined using stable isotope-labeled
collagen (SI-collagen) as reported previously [27]. Using
the standard samples, external calibration curves of cat-
tle, rabbit, and sheep type I collagen were prepared by
MRM analysis of the marker peptides (P11, P9, and
P5) following trypsin digestion for quantification of the
respective animal’s type I collagen concentrations in the
samples. MRM analysis Collagen or glue samples (protein amount of 50 µg) were
heated at 60 °C for 30 min and digested with 2.5 µg of
trypsin (Sigma-Aldrich) at 37 °C for 16 h in 100 µL of
100 mM Tris–HCl/1 mM CaCl2 (pH 7.6). The dechromed
leather was similarly digested with trypsin as reported
previously [21]. The tryptic digests were acidified with
formic acid (final 1%) and subjected to LC–MS/MS or
MRM analysis. Type I collagen-derived marker peptides were monitored
by LC–MS in MRM mode. The tryptic peptide solutions
were separated using an Ascentis Express C18 HPLC
column (5 μm particle size, L × I.D. 150 mm × 2.1 mm;
Supelco) at a flow rate of 500 μL/min with a binary gradi-
ent as follows: 98% solvent A (0.1% formic acid in water)
for 2 min, linear gradient of 2–60% solvent B (100% ace-
tonitrile) for 4 min, 90% solvent B for 2 min, and 98%
solvent A for 2 min. The MRM transitions and collision
energy settings for the marker peptides are shown in
Additional file 1: Table S1. The detection threshold for
marker peptides was set to a peak intensity of 100 with a
signal-to-noise ratio of 5.h LC–MS/MS analysisi Identification of collagen marker peptides were per-
formed by LC–MS/MS using a 3200 QTRAP mass
spectrometer (AB Sciex, Foster City, CA) coupled to an
Agilent 1200 Series HPLC system (Agilent Technolo-
gies, Palo Alto, CA). Tryptic digests of standard samples
were loaded onto an Ascentis Express C18 HPLC column
(2.7 μm particle size, L × I.D. 150 mm × 2.1 mm; Supelco,
Bellefonte, PA) at a flow rate of 200 μL/min with a binary
gradient as follows: 98% solvent A (0.1% formic acid
in water) for 5 min, linear gradient of 2–50% solvent B
(100% acetonitrile) for 15 min, 90% solvent B for 5 min,
and 98% solvent A for 5 min. Three injections of 10 µL
were performed for each sample. The MS scan and MS/
MS acquisition were performed over m/z ranges of 400–
1300 and 100–1700, respectively. The collision energy
was automatically determined based on the mass and
charge state of the precursor ions using rolling collision
energy. Acquired MS/MS spectra were searched against a
local type I collagen database, which included type I col-
lagen sequences of various animals [8], using ProteinPilot
software 4.0 (AB Sciex) with the Paragon algorithm [24]. The confidence threshold of the identified peptides was Kumazawa et al. Herit Sci (2018) 6:43 Page 4 of 9 is because of the high sequence homology of collagen
among mammals. In contrast, almost all peptides identi-
fied for sturgeon glue were from fish species.i Identification of the animal source of commercial glues Identification of the animal source of commercial glues
We next investigated commercial animal glues using the
established method (Table 2). Among 22 glue samples,
the animal source was unambiguously determined for 18
samples, including glues for which the biological origin
was not provided. In addition, several samples (C, K, and
S) were identified to have two different animal origins. In sample S identified as sheep and goat, two marker
peptides, P4 and P5, sharing same sequence position
(733–756 and 741–756) were detected, which strongly
supports the co-existence of different origins in the one
product. We could not determine the animal species for
some samples (B, F, L, and P) because the detection pat-
terns of the marker peptides did not match any of those
established for glue containing one or two animal origins. We assumed that these glue samples contained more
than three species or were derived from unexpected ani-
mals. In seven samples identified as cattle (A, D, E, G,
and H) or sturgeon (Q and R), the identification results
matched the animal source given on the label. However,
some commercial glues showed detection patterns that
were different from those predicted for the declared ani-
mal species. For example, the animal source of sample I
was declared as deer, but was identified as cattle based on
the detection of P1, P4, and P11. Surprisingly, almost all
of the glue products supplied as rabbit were identified to
be completely derived from a different animal source (pig
for sample J and cattle for samples M, N, and O). In addi-
tion, sample K was shown to contain cattle as an addi-
tional animal source other than rabbit. There was no pure
rabbit glue among the commercial products we analyzed. This is consistent with a previous observation that some ii
From the identified peptide list, we ruled out peptides
in which not all Pro residues at the Yaa position were
hydroxylated to 4-Hyp and a missed cleavage except at
the Arg- or Lys-Hyp bond was observed. MRM transi-
tions were then set for the remaining marker peptide
candidates to monitor the presence or absence of the
peptides for each standard sample (data not shown). c The presence and absence of marker peptides are denoted by + and −, respectively Selection of collagen marker peptides Selection of collagen marker peptides To establish an LC–MS method for animal source iden-
tification of glue, we first selected marker peptides using
authentic standards of cattle, horse, pig, sheep, and rabbit
(collagen), deer and sturgeon (glue), and goat (leather). The authentic standards were digested with trypsin, and
the generated peptides were subjected to LC–MS/MS
analysis. Many type I collagen-derived peptides from
various animal species were identified for the respec-
tive standard samples by searching the acquired MS/
MS spectra against an in-house type I collagen database
(data have been deposited to the ProteomeXchange con-
sortium with the identifier PXD009059). For example,
although the sequence of deer type I collagen is not in
public databases, analysis of deer glue identified a total
of 396 peptides for type I collagen α1 and α2 chains
from various animals, including 55 peptides for bovine
(Bos taurus) and 37 peptides for sheep (Ovis aries). This Kumazawa et al. Herit Sci (2018) 6:43 a The numbering of residues begins with the triple-helical region of the chains
b O indicates 4-Hyp Kumazawa et al. Herit Sci (2018) 6:43 Herit Sci (2018) 6:43 Page 5 of 9 Fig. 1 MRM chromatograms of type I collagen-derived marker peptides for eight animal species. Single intense peaks out of two MRM transitions
are depicted as representatives for respective marker peptides. Detected marker peptides for each authentic animal standard are indicated in
boldface Fig. 1 MRM chromatograms of type I collagen-derived marker peptides for eight animal species. Single intense peaks out of two MRM transitions
are depicted as representatives for respective marker peptides. Detected marker peptides for each authentic animal standard are indicated in
boldface modern skin glues provided as rabbit were in fact cattle
glues [17]. we speculate that cattle glue was accidentally contami-
nated with rabbit glue, such as when these glues are pre-
pared through the same processes. In contrast, sample K
declared as rabbit was composed of nearly equal parts of
the two animal origins (55.0% cattle and 45.0% rabbit),
suggesting that cattle glue was purposefully blended into
rabbit glue. We also estimated the relative abundance of the animal
origin for the glue samples revealed to have both cattle
and rabbit origins (samples C and K) (Table 3). Peak areas
of MRM chromatograms of marker peptides detected
for either but not both animals (P11 for cattle and P9 for
rabbit) were used for quantification of respective origins. Concentrations of these animal origins were calculated
by external calibration curves prepared using authentic
standards of cattle and rabbit for which type I collagen
concentrations were predetermined using a previously
developed internal standard, named SI-collagen [27]. The proportion of the undeclared origin in sample C was
estimated to be low (9.9% for rabbit). From this result, Kumazawa et al. Herit Sci (2018) 6:43 From the results, collagen marker peptides were selected
to give at least two differences in the patterns of the pres-
ence or absence of marker peptides between each animal,
as previously reported for leather [21]. Furthermore, we
added a rule that there is at least one difference in the
detection patterns of marker peptides when glue samples
have two animal origins. To satisfy these requirements,
we selected seven marker peptides (P1–P7) from α1(I)
chain and five marker peptides (P8–P12) from α2(I) chain
(Table 1). Two MRM transitions were set for the respec-
tive marker peptides to definitively judge the presence of
the peptides (Additional file 1: Table S1). From the detec-
tion patterns of the MRM peaks, we were able to clearly
discriminate the eight animals as shown in Fig. 1 (repre-
senting single intense transition for each of the marker
peptides). In addition, use of the 12 marker peptides
enabled determination of the animal species even when
two different origins are present in one sample (Addi-
tional file 2: Table S2). For example, marker peptides P1,
P4, and P11 are detected for cattle, and P1, P4, and P9
are detected for rabbit, which means P1, P4, P9, and P11
should be detected for a glue sample with cattle and rab-
bit origins. Table 1 List of marker peptides for animal species identification of glue
a The numbering of residues begins with the triple-helical region of the chains
Chain
Positiona
Marker peptide
Sequenceb
Cattlec
Horsec
Pigc
Sheepc
Goatc
Deerc
Rabbitc
Sturgeonc
α1(I)
316–327
P1
GFOGADGVAGPK
+
+
+
−
−
+
+
−
316–327
P2
GFOGSDGVAGPK
−
−
−
+
+
−
−
−
688–704
P3
GAAGPOGATGFOGAAGR
−
−
−
−
−
−
−
+
733–756
P4
GETGPAGROGEVGPOGPOGPAGEK
+
−
−
−
+
+
+
−
741–756
P5
AGEVGPOGPOGPAGEK
−
−
−
+
−
−
−
−
889–906
P6
GEAGPAGPAGPIGPVGAR
−
+
−
−
−
−
−
−
889–906
P7
GETGPAGPAGPAGPAGAR
−
−
−
−
−
−
−
+
α2(I)
238–252
P8
GIOGPAGAAGATGAR
−
+
+
−
−
−
−
−
253–264
P9
GLVGEOGPAGTK
−
−
−
−
−
−
+
−
361–374
P10
GFOGSOGNVGPAGK
−
−
+
−
−
−
−
−
978–990
P11
IGQOGAVGPAGIR
+
−
−
−
−
−
−
−
978–990
P12
TGQOGAVGPAGIR
−
−
+
+
+
+
−
− Table 1 List of marker peptides for animal species identification of glue Kumazawa et al. Harvest We applied our method to investigate the glue used in
the ground and size layers of The Harvest painted in
1882 by Camille Pissarro (Fig. 2a). Samples were taken
from two separate locations (#1 and #2) of the tack-
ing margin of the canvas (Fig. 2b). The samples were Kumazawa et al. Herit Sci (2018) 6:43 Page 6 of 9 Table 2 Species identified for commercial glue products
N.D. not determined
Sample
Species declared
P1
P2
P3
P4
P5
P6
P7
P8
P9
P10
P11
P12
Identified species
A
Cattle
+
−
−
+
−
−
−
−
−
−
+
−
Cattle
B
Cattle
+
+
−
+
−
−
−
+
−
+
+
+
N.D. C
Cattle
+
−
−
+
−
−
−
−
+
−
+
−
Cattle, rabbit
D
Cattle
+
−
−
+
−
−
−
−
−
−
+
−
Cattle
E
Cattle
+
−
−
+
−
−
−
−
−
−
+
−
Cattle
F
Cattle
+
−
−
+
−
+
−
+
−
+
+
+
N.D. G
Cattle
+
−
−
+
−
−
−
−
−
−
+
−
Cattle
H
Cattle
+
−
−
+
−
−
−
−
−
−
+
−
Cattle
I
Deer
+
−
−
+
−
−
−
−
−
−
+
−
Cattle
J
Rabbit
+
−
−
−
−
−
−
+
−
+
−
+
Pig
K
Rabbit
+
−
−
+
−
−
−
−
+
−
+
−
Cattle, rabbit
L
Rabbit
+
+
−
+
−
−
−
−
−
−
+
−
N.D. M
Rabbit
+
−
−
+
−
−
−
−
−
−
+
−
Cattle
N
Rabbit
+
−
−
+
−
−
−
−
−
−
+
−
Cattle
O
Rabbit
+
−
−
+
−
−
−
−
−
−
+
−
Cattle
P
Rabbit
+
−
−
+
−
−
−
+
+
−
+
−
N.D. Harvest Q
Sturgeon
−
−
+
−
−
−
+
−
−
−
−
−
Sturgeon
R
Sturgeon
−
−
+
−
−
−
+
−
−
−
−
−
Sturgeon
S
None
−
+
−
+
+
−
−
−
−
−
−
+
Sheep, goat
T
None
+
−
−
−
−
−
−
+
−
+
−
+
Pig
U
None
+
−
−
+
−
−
−
−
−
−
+
−
Cattle
V
None
+
−
−
+
−
−
−
−
−
−
+
−
Cattle Table 2 Species identified for commercial glue products Table 3 Relative
abundance
of
the
animal
origin
in commercial glues containing cattle and rabbit origins
The data represent the mean ± SD of three separate measurements
Sample
Species declared
Identified species
Content (%)
C
Cattle
Cattle
90.1 ± 1.4
Rabbit
9.9 ± 1.4
K
Rabbit
Cattle
55.0 ± 2.5
Rabbit
45.0 ± 2.5 Table 3 Relative
abundance
of
the
animal
origin
in commercial glues containing cattle and rabbit origins dispersive X-ray spectroscope revealed that the ground
layer of The Harvest had a two-layer structure (Addi-
tional file 4: Figure S2). Our results indicate that cattle
and sheep glues were used for these two layers and also
for the size layer(s) located under the ground layers. y
g
y
Cleavage of peptide bonds of the collagen molecule
may occur during the heat extraction process of glue
from animal tissues [2]. In addition, during storage, glue
present in artworks possibly undergoes various chemi-
cal reactions, such as peptide bond cleavage and deami-
dation of asparagine and glutamine residues [28, 29]. Modified marker peptides cannot be detected by MRM
analysis because of the mass shifts caused by these
changes. Despite the low sample amount taken from
The Harvest (approximately 1 mg including fibers of the
canvas), identification of the animal source of glue was
successfully achieved using collagen-derived marker pep-
tides. This indicates that although protein damage might
have occurred during aging, the remaining intact pep-
tides were sufficient to allow species identification of glue
for the historical artwork painted more than 100 years
ago. Harvest Although we cannot rule out the possibility that par-
tial degradation of glue affects the accuracy of quantifi-
cation of the relative abundance of the animal origin, the
similar quantitative values for the different two locations The data represent the mean ± SD of three separate measurements directly subjected to trypsin digestion for species iden-
tification by MRM analysis. As shown in Fig. 3, collagen
marker peptides were clearly detected for location #1. Based on the detection pattern of the marker peptides
(P1, P2, P4, P5, P11, and P12), the glue was determined
to be derived from two animal species, cattle and sheep
(Table 4). Location #2 also showed the same detec-
tion pattern (Additional file 3: Figure S1). By estimat-
ing the relative abundance of cattle and sheep origins
using specific marker peptides P5 for sheep and P11
for cattle, both samples from the two different loca-
tions showed similar results; the glues were composed
of approximately 65% cattle and 35% sheep (Table 4). Scanning electron microscopy equipped with an energy Kumazawa et al. Herit Sci (2018) 6:43 Page 7 of 9 Fig. 2 The Harvest. a Small samples from the tacking margin of the canvas were analyzed for species identification of glue. The sampling locations
are indicated by dotted boxes (#1 and #2). b The tacking margin of the canvas is enclosed by white dotted lines Fig. 2 The Harvest. a Small samples from the tacking margin of the canvas were analyzed for species identification of glue. The sampling locations
are indicated by dotted boxes (#1 and #2). b The tacking margin of the canvas is enclosed by white dotted lines Fig. 2 The Harvest. a Small samples from the tacking margin of the canvas were analyzed for species identification of glue. The sampling locations
are indicated by dotted boxes (#1 and #2). b The tacking margin of the canvas is enclosed by white dotted lines Fig. 3 MRM chromatograms of marker peptides detected for
location #1 of The Harvest. Detected marker peptides are indicated in
boldface In addition, our method is applicable not only to simple
glue samples but also to samples containing two animal
origins. We demonstrated its application in glue certifica-
tion by detecting undeclared animal species and blends
or contamination of different animal origins in commer-
cial glue products. Harvest Furthermore, we clarified that both
cattle- and sheep-originated animal glues were used in
the ground and size layers of The Harvest. Many studies
have investigated glues present in artworks and identi-
fied their animal origins using MS [15–19]. However,
only one glue species was determined for each sample in
almost all of the investigations, except for the simultane-
ous detection of bovine- and fish-specific peptides in a
polychromy sample [18]. Our novel method may uncover
hidden facts about glue used for artworks, which will be
helpful to understand the use of artist’s materials in his-
toric objects. We could not determine the animal source
of four commercial products. These were assumed to be
derived from more than three animals or other unex-
pected animals. Species identification under these situ-
ations will be an area of future work. We consider that
reselection of marker peptides in response to the target
animals should lead to successful determination of the
animal source for such complicated glue samples. Fig. 3 MRM chromatograms of marker peptides detected for
location #1 of The Harvest. Detected marker peptides are indicated in
boldface Fig. 3 MRM chromatograms of marker peptides detected for
location #1 of The Harvest. Detected marker peptides are indicated in
boldface from the painting enhanced the reliability of the analysis
(Table 4). Additional files 7. Buckley M, Collins M, Thomas-Oates J, Wilson JC. Species identification by
analysis of bone collagen using matrix-assisted laser desorption/ionisa-
tion time-of-flight mass spectrometry. Rapid Commun Mass Spectrom. 2009;23(23):3843–54. Additional file 1: Table S1. MRM transitions of marker peptides. Additional file 2: Table S2. Detection patterns of marker peptides for
glue samples with two different animal origins. Additional file 3: Figure S1. MRM chromatograms of marker peptides
detected for location #2 of The Harvest. Detected marker peptides are
indicated in boldface. Additional file 4: Figure S2. Cross-section of The Harvest. (a) The scan-
ning electron microscopy (SEM) image shows the ground layer consisting
of two layers under a paint layer. The upper ground layer contained lead
white and calcium carbonate, and the bottom ground layer contained
calcium sulfate only, or calcium sulfate and calcium carbonate. The sample
was examined by SEM (JSM-5800LV; JEOL, Tokyo, Japan) equipped with
an energy dispersive X-ray spectroscope (DX4; EDAX, Mahwah, NJ). (b)
Schematic illustration of the cross section Additional file 1: Table S1. MRM transitions of marker peptides. Additional file 2: Table S2. Detection patterns of marker peptides for
glue samples with two different animal origins. Additional file 3: Figure S1. MRM chromatograms of marker peptides
detected for location #2 of The Harvest. Detected marker peptides are
indicated in boldface. Additional file 4: Figure S2. Cross-section of The Harvest. (a) The scan-
ning electron microscopy (SEM) image shows the ground layer consisting Additional file 1: Table S1. MRM transitions of marker peptides. Additional file 2: Table S2. Detection patterns of marker peptides for
glue samples with two different animal origins. 8. Buckley M, Larkin N, Collins M. Mammoth and Mastodon col-
lagen sequences; survival and utility. Geochim Cosmochim Acta
2011;75(7):2007–16. 8. Buckley M, Larkin N, Collins M. Mammoth and Mastodon col-
lagen sequences; survival and utility. Geochim Cosmochim Acta. 2011;75(7):2007–16. Additional file 3: Figure S1. MRM chromatograms of marker peptides
detected for location #2 of The Harvest. Detected marker peptides are
indicated in boldface. 9. von Bargen C, Dojahn J, Waidelich D, Humpf HU, Brockmeyer J. New
sensitive high-performance liquid chromatography-tandem mass spec-
trometry method for the detection of horse and pork in halal beef. J Agric
Food Chem. 2013;61(49):11986–94. Additional file 4: Figure S2. Cross-section of The Harvest. (a) The scan-
ning electron microscopy (SEM) image shows the ground layer consisting
of two layers under a paint layer. Conclusion In the present study, we established a novel LC–MS
method for animal source identification of glue using
detection patterns of 12 type I collagen-derived marker
peptides. This database-independent approach ena-
bles discrimination between eight animals regardless
of whether their type I collagen sequences are known. Table 4 Glue species identification and determination of relative abundance of the animal origin in The Harvest
P1
P2
P3
P4
P5
P6
P7
P8
P9
P10
P11
P12
Identified species
Content (%)
#1
+
+
−
+
+
−
−
−
−
−
+
+
Cattle
65.3
Sheep
34.7
#2
+
+
−
+
+
−
−
−
−
−
+
+
Cattle
62.4
Sheep
37.6 es identification and determination of relative abundance of the animal origin in The Harvest Kumazawa et al. Herit Sci (2018) 6:43 Page 8 of 9 Author details 16. Fremout W, Dhaenens M, Saverwyns S, Sanyova J, Vandenabeele P,
Deforce D, et al. Tryptic peptide analysis of protein binders in works of art
by liquid chromatography-tandem mass spectrometry. Anal Chim Acta. 2010;658(2):156–62. 1 Japan Institute of Leather Research, 520‑11 Kuwabara, Toride, Ibaraki
302‑0017, Japan. 2 Nippi Research Institute of Biomatrix, 520‑11 Kuwabara,
Toride, Ibaraki 302‑0017, Japan. 3 The National Museum of Western Art, 7‑7
Ueno‑Koen, Taito‑ku, Tokyo 110‑0007, Japan. 17. Dallongeville S, Koperska M, Garnier N, Reille-Taillefert G, Rolando C,
Tokarski C. Identification of animal glue species in artworks using
proteomics: application to a 18th century gilt sample. Anal Chem. 2011;83(24):9431–7. Additional files The upper ground layer contained lead
white and calcium carbonate, and the bottom ground layer contained
calcium sulfate only, or calcium sulfate and calcium carbonate. The sample
was examined by SEM (JSM-5800LV; JEOL, Tokyo, Japan) equipped with
an energy dispersive X-ray spectroscope (DX4; EDAX, Mahwah, NJ). (b)
Schematic illustration of the cross-section. 10. Watson AD, Gunning Y, Rigby NM, Philo M, Kemsley EK. Meat authentica-
tion via multiple reaction monitoring mass spectrometry of myoglobin
peptides. Anal Chem. 2015;87(20):10315–22. 11. Cuollo M, Caira S, Fierro O, Pinto G, Picariello G, Addeo F. Toward milk
speciation through the monitoring of casein proteotypic peptides. Rapid
Commun Mass Spectrom. 2010;24(11):1687–96. 12. Bernardi N, Benetti G, Haouet NM, Sergi M, Grotta L, Marchetti S, et al. A
rapid high-performance liquid chromatography–tandem mass spectrom-
etry assay for unambiguous detection of different milk species employed
in cheese manufacturing. J Dairy Sci. 2015;98(12):8405–13. 13. Zhang G, Liu T, Wang Q, Chen L, Lei J, Luo J, et al. Mass spectrometric
detection of marker peptides in tryptic digests of gelatin: a new method
to differentiate between bovine and porcine gelatin. Food Hydrocolloids. 2009;23(7):2001–7. Abbreviations
H
h d Hyp: hydroxyproline; MRM: multiple reaction monitoring; SI-collagen: stable
isotope-labeled collagen. Hyp: hydroxyproline; MRM: multiple reaction monitoring; SI-collagen: stable
isotope-labeled collagen. 14. Cheng XL, Wei F, Xiao XY, Zhao YY, Shi Y, Liu W, et al. Identification of five
gelatins by ultra performance liquid chromatography/time-of-flight mass
spectrometry (UPLC/Q-TOF-MS) using principal component analysis. J
Pharm Biomed Anal. 2012;62:191–5. The authors declare that they have no competing interests. 18. Dallongeville S, Richter M, Schafer S, Kuhlenthal M, Garnier N, Rolando C,
et al. Proteomics applied to the authentication of fish glue: application to
a 17th century artwork sample. Analyst. 2013;138(18):5357–64. Funding This work was supported by JSPS KAKENHI Grant Number JP16K01187. 19. Mazurek J, Svoboda M, Maish J, Kawahara K, Fukakusa S, Nakazawa T, et al. Characterization of binding media in Egyptian Romano portraits using
enzyme-linked immunosorbant assay and mass spectrometry. e-Preserv
Sci. 2014;11:76–83. Authors’ contributions YK and YT designed and developed the method. YK performed the experi-
ments. MT provided the glue/painting samples and historical background. YK,
YT, and MT wrote the manuscript. All authors analyzed the data. All authors
read and approved the final manuscript. 15. Kuckova S, Nemec I, Hynek R, Hradilova J, Grygar T. Analysis of organic
colouring and binding components in colour layer of art works. Anal Bio-
anal Chem. 2005;382(2):275–82. Received: 30 April 2018 Accepted: 5 July 2018 21. Kumazawa Y, Taga Y, Iwai K, Koyama Y. A rapid and simple LC–MS method
using collagen marker peptides for identification of the animal source of
leather. J Agric Food Chem. 2016;64(30):6051–7. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in pub-
lished maps and institutional affiliations. 20. Tripkovic T, Charvy C, Alves S, Lolic AD, Baosic RM, Nikolic-Mandic SD,
et al. Identification of protein binders in artworks by MALDI-TOF/TOF
tandem mass spectrometry. Talanta. 2013;113:49–61. Received: 30 April 2018 Accepted: 5 July 2018 Competing interests The authors declare that they have no competing interests. 29. Leo G, Bonaduce I, Andreotti A, Marino G, Pucci P, Colombini MP, et al.
Deamidation at asparagine and glutamine as a major modification upon
deterioration/aging of proteinaceous binders in mural paintings. Anal
Chem. 2011;83(6):2056–64. 28. Manzano E, Romero-Pastor J, Navas N, Rodríguez-Simón LR, Cardell C.
A study of the interaction between rabbit glue binder and blue copper
pigment under UV radiation: a spectroscopic and PCA approach. Vib
Spectrosc. 2010;53(2):260–8. Kumazawa et al. Herit Sci (2018) 6:43 References g
22. Lange V, Picotti P, Domon B, Aebersold R. Selected reaction monitoring
for quantitative proteomics: a tutorial. Mol Syst Biol. 2008;4:222. 1. Ockerman HW, Hansen CL. Animal by-product processing & utilization. Lancaster: Technomic publishing; 2000. 23. Taga Y, Kusubata M, Ogawa-Goto K, Hattori S. Developmental stage-
dependent regulation of prolyl 3-hydroxylation in tendon type I collagen
J Biol Chem. 2016;291(2):837–47. 2. Schellmann NC. Animal glues: a review of their key properties relevant to
conservation. Stud Consev. 2007;8:55–66. 3. Kaml I, Vcelakova K, Kenndler E. Characterisation and identification of
proteinaceous binding media (animal glues) from their amino acid profile
by capillary zone electrophoresis. J Sep Sci. 2004;27(3):161–6. 24. Shilov IV, Seymour SL, Patel AA, Loboda A, Tang WH, Keating SP, et al. The
Paragon Algorithm, a next generation search engine that uses sequence
temperature values and feature probabilities to identify peptides from
tandem mass spectra. Mol Cell Proteomics. 2007;6(9):1638–55. y
y
4. Prikryl P, Havlickova L, Pacakova V, Hradilova J, Stulik K, Hofta P. An evalu-
ation of GC-MS and HPLC-FD methods for analysis of protein binders in
paintings. J Sep Sci. 2006;29(17):2653–63. 25. Vizcaino JA, Deutsch EW, Wang R, Csordas A, Reisinger F, Rios D, et al. ProteomeXchange provides globally coordinated proteomics data sub-
mission and dissemination. Nat Biotechnol. 2014;32(3):223–6. 5. Cartechini L, Vagnini M, Palmieri M, Pitzurra L, Mello T, Mazurek J, et al. Immunodetection of proteins in ancient paint media. Acc Chem Res. 2010;43(6):867–76. 26. Okuda S, Watanabe Y, Moriya Y, Kawano S, Yamamoto T, Matsumoto M,
et al. jPOSTrepo: an international standard data repository for proteomes. Nucleic Acids Res. 2017;45(D1):D1107–11. 6. Albertini E, Raggi L, Vagnini M, Sassolini A, Achilli A, Marconi G, et al. Tracing the biological origin of animal glues used in paintings through
mitochondrial DNA analysis. Anal Bioanal Chem. 2011;399(9):2987–95. 27. Taga Y, Kusubata M, Ogawa-Goto K, Hattori S. Stable isotope-labeled col-
lagen: a novel and versatile tool for quantitative collagen analyses using
mass spectrometry. J Proteome Res. 2014;13(8):3671–8. Page 9 of 9
|
https://openalex.org/W3133591262
|
https://aprp.msal.ru/jour/article/download/2193/1636
|
Russian
| null |
Protection of Scientifi c Research Results and Scientifi c Information in Elite Sports ("Formula 1" Case Study)
|
Aktualʹnye problemy rossijskogo prava
| 2,021
|
cc-by
| 5,999
|
DOI: 10.17803/1994-1471.2021.123.2.175-182 DOI: 10.17803/1994-1471.2021.123.2.175-182 А. М. Камалян* © Камалян А. М., 2021
* Камалян Артур Михайлович, кандидат юридических наук, преподаватель кафедры интеграционного и
европейского права Московского государственного юридического университета имени О.Е. Кутафина
(МГЮА)
Садовая-Кудринская ул., д. 9, г. Москва, Россия, 125993
arthur.kamalyan@gmail.com Камалян А. М. Камалян А. М. Камалян А. М. Охрана результатов научных исследований и научной информации в спорте высших достижений на примере «Формулы-1» Охрана результатов научных исследований и научной информации в спорте высших достижений на примере «Формулы-1» Актуальные проблемы российского права. 2021. Т. 16. № 2 (123) февраль Охрана результатов научных исследований
и научной информации в спорте высших
достижений на примере «Формулы-1»1 Аннотация. В статье исследуется специфика охраны результатов научных исследований в профессиональ-
ном спорте на примере «Формулы-1». Подчеркивается, что обладателем прав на результаты научных иссле-
дований может быть не только сама гоночная команда, но и одно из тесно связанных юридических лиц. Приводятся конкретные примеры, которые демонстрируют, что́ может охраняться патентом и кто может
обладать правами на объекты интеллектуальной собственности. Отмечается, что, несмотря на большое
количество высокотехнологичных решений, команды «Формулы-1» нередко сознательно отказываются
от патентования. Анализ ситуации в этом виде спорта показывает, что в условиях постоянного внесения
изменений и доработок в гоночные автомобили получение патента неоправданно из-за продолжитель-
ности этой процедуры. В то же время результаты научных исследований и научная информация охраня-
ются режимом коммерческой тайны. Оговаривается, что такой режим сопровождается риском разглаше-
ния конфиденциальной информации лицами, прежде всего действующими и бывшими сотрудниками. В качестве иллюстрации приводится один из крупнейших шпионских скандалов в истории «Формулы-1». Особое внимание уделяется проблеме перехода членов персонала из одного коллектива «Формулы-1» в
другой, в том числе разграничению собственных навыков сотрудника и полученных им на предыдущей
работе охраняемых сведений. Дополнительно подчеркивается, что такие переходы часто сопровождаются
обязательным отпуском без права выхода на новую работу, чтобы имеющиеся знания о наработках уже
бывшего работодателя потеряли свою актуальность. Отмечается также, что режим коммерческой тайны
не препятствует командам «Формулы-1» знакомиться с результатами научных исследований своих конку-
рентов ввиду правил о максимальной открытости автомобилей на время проведения Гран-при. Приведены
конкретные примеры заимствования гоночными командами успешных инженерных решений соперников
путем создания собственных аналогов. Ключевые слова: научные исследования; результаты научных исследований; научная информация; инно-
вации; технологии; правовое регулирование; интеллектуальная собственность; охрана интеллектуальной
собственности; патент; коммерческая тайна; спорт; Формула-1; автогонки; Международная федерация
автоспорта (FIA). Для цитирования: Камалян А. М. Охрана результатов научных исследований и научной информации в
спорте высших достижений на примере «Формулы-1» // Актуальные проблемы российского права. —
2021. — Т. 16. — № 2. — С. 175–182. — DOI: 10.17803/1994-1471.2021.123.2.175-182. 1
Исследование выполнено при финансовой поддержке РФФИ в рамках научного проекта № 18-29-15022. Актуальные проблемы российского права. 2021. Т. 16. № 2 (123) февраль 175 ИНТЕГРАЦИОННОЕ ПРАВО p
y
y
g
p
j
3
Например: патент 182952 на охрану рулевого колеса Ferrari, который использовался в болиде Ferrari SF90
для выступлений в 2019 г. (см.: официальный сайт Всемирной организации интеллектуальной собственности
(ВОИС). URL: https://www3.wipo.int/designdb/en/showData.jsp?ID=CAID.182952 (дата обращения: 17.05.2020)).
4
Например: патент 007693569-0001 на охрану болида McLaren MCL35 для выступления в сезоне-2020 (см.:
официальный сайт ВОИС. URL: https://www3.wipo.int/designdb/en/showData.jsp?ID=EMID.007693569-0001
(дата обращения: 17.05.2020)). Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration and European Law Department, Kutafin
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration and European Law Department, Kutafin
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com Abstract. The paper examines the legal regulation of scientific research in professional sports as in the case of
Formula 1. It is emphasized that the owner of the rights to the results of scientific research can be not only the
racing team itself, but also one of the closely related legal entities. Specific examples are provided demonstrating
what can be protected by a patent and who can own intellectual property rights. It is noted that, despite a large
number of high-tech solutions, Formula 1 teams often deliberately refuse to patent. An analysis of the situation
in this sport shows that in the conditions of constant changes and improvements in racing cars, obtaining a patent
is unjustified due to the length of this procedure. At the same time, the results of scientific research and scientific
information are protected by the trade secret regime. It is stipulated that such a regime is accompanied by the risk
of disclosure of confidential information by persons, primarily current and former employees. One of the biggest
spy scandals in the history of Formula 1 is cited as an illustration. Particular attention is given to the problem of
the transfer of staff members from one Formula 1 team to another, including the delineation of the employee’s
own skills and the protected information obtained by him in previous work. In addition, it is emphasized that
such transitions are often accompanied by compulsory leave without the right to go to a new job, so that the
existing knowledge about the work of the former employer loses its relevance. It is also noted that the trade
secret regime does not prevent Formula 1 teams from getting acquainted with the results of scientific research of
their competitors due to the rules on the maximum openness of cars during the Grand Prix. Specific examples of
borrowing by racing teams of successful engineering solutions of rivals by creating their own analogues are given. Keywords: scientific research; scientific research results; scientific information; innovation; technologies; legal
regulation; intellectual property; protection of intellectual property; patent; trade secret; sport; Formula 1; auto
racing; International Motor Sport Federation (FIA). 2
The reported study was funded by RFBR according to the research project № 18-29-15022. Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration and European Law Department, Kutafin
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com Cite as: Kamalyan AM. Okhrana rezultatov nauchnykh issledovaniy i nauchnoy informatsii v sporte vysshikh
dostizheniy na primere «Formuly-1» [Protection of Scientific Research Results and Scientific Information in Elite
Sports ("Formula 1" Case Study)]. Aktualnye problemy rossiyskogo prava. 2021;16(2):175-182. DOI: 10.17803/1994-
1471.2021.123.2.175-182. (In Russ., abstract in Eng.). Н Н
аучные исследования в спорте, в частно-
сти в «Формуле-1», имеют определенную
специфику. Как правило, регламенты со-
ревнований содержат требования к конечному
результату исследований, хотя «серые зоны» все
равно возникают, предоставляя конструкторам
и инженерам определенную свободу. Подоб-
ное положение дел не могло не сказаться и на охране таких результатов научных исследова-
ний, которая тоже имеет свои особенности. Как известно, «Формула-1» является высоко-
технологичным спортом, поэтому объектов охра-
ны субъективных прав довольно много: отдель-
ные детали автомобиля3, сам дизайн болида4
и даже его раскраска5 и т.д. При этом сто́ит
иметь в виду, что правообладателем может охране таких результатов научных исследова-
ний, которая тоже имеет свои особенности. Как известно, «Формула-1» является высоко-
технологичным спортом, поэтому объектов охра-
ны субъективных прав довольно много: отдель-
ные детали автомобиля3, сам дизайн болида4
и даже его раскраска5 и т.д. При этом сто́ит
иметь в виду, что правообладателем может p
y
y
g
p
j
3
Например: патент 182952 на охрану рулевого колеса Ferrari, который использовался в болиде Ferrari SF90
для выступлений в 2019 г. (см.: официальный сайт Всемирной организации интеллектуальной собственности
(ВОИС). URL: https://www3.wipo.int/designdb/en/showData.jsp?ID=CAID.182952 (дата обращения: 17.05.2020)). 4
Например: патент 007693569-0001 на охрану болида McLaren MCL35 для выступления в сезоне-2020 (см.:
официальный сайт ВОИС. URL: https://www3.wipo.int/designdb/en/showData.jsp?ID=EMID.007693569-0001
(дата обращения: 17.05.2020)). Актуальные проблемы российского права. 2021. Т. 16. № 2 (123) февраль 176 Камалян А. М. Охрана результатов научных исследований и научной информации в спорте высших достижений на примере «Формулы-1» Камалян А. М. а ал
. . Охрана результатов научных исследований и научной информации в спорте высших достижений на примере «Формулы-1» выступать как гоночный коллектив, так и тесно
связанное с ним иное юридическое лицо. На-
пример, британские команды McLaren (McLaren
Racing Limited) и Williams (Williams Grand Prix
Engineering Limited) сами являются правообла-
дателями своих изобретений, а правами на
результаты научных исследований гоночной
команды Ferrari обладает головное акционер-
ное общество Ferrari S.p.A. затраты. Немаловажным обстоятельством явля-
ется также тот факт, что система KERS активно
внедряется в систему общественного транспорта
Великобритании: уже в июле 2014 г. 5
Например: патенты 006813234-0002 и 006813234-0003 на охрану раскраски болида McLaren в це-
лом и отдельного элемента орнамента (см.: официальный сайт ВОИС. URL: https://www3.wipo.int/
designdb/en/showData.jsp?ID=EMID.006813234-0002; https://www3.wipo.int/designdb/en/showData.
jsp?ID=EMID.006813234-0003 (дата обращения: 17.05.2020)). Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration and European Law Department, Kutafin
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com (спустя три
месяца после покупки подразделения Williams)
GKN заключила контракт с одним из крупней-
ших британских перевозчиков Go-Ahead Group
на поставку системы KERS для 500 автобусов в
течение двух лет, прежде всего для Лондона и
Оксфорда8, а в 2015 г. объем заказов увеличился
до 750 автобусов9. выступать как гоночный коллектив, так и тесно
связанное с ним иное юридическое лицо. На-
пример, британские команды McLaren (McLaren
Racing Limited) и Williams (Williams Grand Prix
Engineering Limited) сами являются правообла-
дателями своих изобретений, а правами на
результаты научных исследований гоночной
команды Ferrari обладает головное акционер-
ное общество Ferrari S.p.A. Необходимо также понимать, что результаты
научных исследований коллективов «Форму-
лы-1» могут применяться не только в спорте,
но и в обычных автомобилях. В таком контек-
сте ценность и значение патентов значительно
возрастает. Например, в 2009 г. в «Формуле-1»
появилась система рекуперации кинетической
энергии KERS (Kinetic Energy Recovery System) и
британский коллектив Williams одним из первых
зарегистрировал патент на свою разработку в
этой области6. Через пять лет, в апреле 2014 г.,
предприятие Williams Hybrid Power, которое и
занималось исследованием системы KERS для
команды «Формулы-1», было продано со всеми
правами на результаты научных исследований
за 8 млн фунтов стерлингов британской авто-
мобильной и аэрокосмической компании GKN с
дополнительными выплатами от продаж в раз-
мере 3,5 % в первые пять лет и 1,5 % в после-
дующие годы7. И хотя стоимость сделки сравни-
тельно невысока, необходимо отметить, что в
свое время команда Williams вложила в данное
предприятие 1,5 млн фунтов стерлингов, то есть
цена продажи более чем в пять раз превысила В то же время, несмотря на большое коли-
чество патентов и высокий уровень результатов
научных исследований, в мире «Формулы-1»
крайне редко можно встретить судебные раз-
бирательства или «патентные войны», кото-
рые характерны для областей, связанных с
технологиями (например, патентные споры
Apple v. Samsung, Motorola v. Microsoft, Oracle
v. Google). Однако это вовсе не свидетельствует
о добросовестном поведении участников чем-
пионата мира. Отсутствие «патентных» войн в
«королеве автоспорта» объясняется тем, что
команды зачастую отказываются от патента
как формы оформления и защиты своих прав
на результат научных исследований (то есть в
идеале патентов в «Формуле-1» должно было
быть еще больше). Причин такого неожидан-
ного выбора несколько. Во-первых, процедура
патентования занимает довольно много време-
ни. Актуальные проблемы российского права. 2021. Т. 16. № 2 (123) февраль 6
Патент EP2422110 (B1) на «маховик в сборе» (Flywheel Assembly — альтернативное и более общее
название для KERS), выданный Европейским патентным ведомством (European Patent Office — EPO) (см.:
официальный сайт EPO. URL: https://data.epo.org/gpi/EP2422110B1-FLYWHEEL-ASSEMBLY (дата обращения:
17.05.2020)). Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration and European Law Department, Kutafin
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com Исключительно гоноч-
ные технологии принципиально не патентуются,
чтобы вместо «патентных войн» и дуэлей ад-
вокатов велись «инженерные войны» и битвы
конструкторов — в полном соответствии с ду-
хом соревнований. смена Технического регламента), оформленный
патент может уже не представлять актуальность
и интерес для правообладателя. Во-вторых, за-
крепление результата научного исследования в
форме патента предоставляет команде-право-
обладателю конкурентное преимущество, по-
скольку соперники не могут им воспользовать-
ся, а попытки создать собственный аналог могут
привести к нарушению прав. Казалось бы, это
обычное дело в любой другой отрасли, однако
важно помнить, что «Формула-1» — это прежде
всего спорт, правила которого вырабатываются
в том числе при участии самих команд. Иными
словами, если будет запатентована технология
или деталь, которая предоставляет одному
из коллективов значительное преимущество,
остальные команды на следующем обсужде-
нии правил (как правило, проходит в конце чем-
пионата) могут принять решение о запрете ее
использования в будущем10. Ценность такого
патента резко падает. В итоге команды «Фор-
мулы-1» патентуют лишь те детали и техно-
логии, которые можно будет использовать в
коммерческих целях при переносе на обычные
автомобили или прочую технику (как в случае
с системой KERS), и в том виде, в котором они
будут реализовываться. Исключительно гоноч-
ные технологии принципиально не патентуются,
чтобы вместо «патентных войн» и дуэлей ад-
вокатов велись «инженерные войны» и битвы
конструкторов — в полном соответствии с ду-
хом соревнований. используют режим коммерческой тайны, кото-
рый также позволяет обладателю конфиденци-
альной информации сохранять и использовать
преимущество в борьбе с конкурентами. Согла-
шение по торговым аспектам прав интеллек-
туальной собственности (ТРИПС)11 определяет
информацию, составляющую коммерческую
тайну, как любую секретную информацию в
том смысле, что она в целом или в определен-
ной конфигурации и подборе ее компонентов
не является общеизвестной и легко доступной
лицам в тех кругах, которые обычно имеют дело
с подобной информацией; имеет коммерческую
ценность ввиду своей секретности и является
предметом надлежащих в данных обстоятель-
ствах мер, направленных на сохранение ее
секретности, со стороны лица, правомерно
контролирующего эту информацию (п. 2 ст. 39). Похожее определение дается в Директиве ЕС
2016/943 о защите конфиденциальных ноу-хау
и деловой информации (коммерческой тайны)
от незаконного приобретения, использования
и раскрытия12 и в российском Федеральном за-
коне от 29.07.2004 № 98-ФЗ «О коммерческой
тайне»13. Подобный режим позволяет всем командам
использовать новейшие технологии, сохраняя
при этом в тайне процесс и детали их разра-
ботки. В истории «Формулы-1» было немало
случаев, когда результат научных исследований
одного коллектива быстро набирал популяр-
ность благодаря своей эффективности, застав-
ляя конкурентов создавать свои аналоги. Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration and European Law Department, Kutafin
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com С учетом того, как часто и сильно меняется
болид в течение одного года (не говоря уже о
нескольких годах, особенно если происходит 5
Например: патенты 006813234-0002 и 006813234-0003 на охрану раскраски болида McLaren в це-
лом и отдельного элемента орнамента (см.: официальный сайт ВОИС. URL: https://www3.wipo.int/
designdb/en/showData.jsp?ID=EMID.006813234-0002; https://www3.wipo.int/designdb/en/showData. jsp?ID=EMID.006813234-0003 (дата обращения: 17.05.2020)). 6
Патент EP2422110 (B1) на «маховик в сборе» (Flywheel Assembly — альтернативное и более общее
название для KERS), выданный Европейским патентным ведомством (European Patent Office — EPO) (см.:
официальный сайт EPO. URL: https://data.epo.org/gpi/EP2422110B1-FLYWHEEL-ASSEMBLY (дата обращения:
17.05.2020)). 6
Патент EP2422110 (B1) на «маховик в сборе» (Flywheel Assembly — альтернативное и более общее
название для KERS), выданный Европейским патентным ведомством (European Patent Office — EPO) (см.:
официальный сайт EPO. URL: https://data.epo.org/gpi/EP2422110B1-FLYWHEEL-ASSEMBLY (дата обращения:
17.05.2020)). 7
The Telegraph. URL: https://www.telegraph.co.uk/finance/newsbysector/industry/10737329/Williams-sells-
hybrid-power-unit-to-GKN.html (дата обращения: 17.05.2020). 8
Официальный сайт компании Go-Ahead Group. URL: https://www.go-ahead.com/media/press-relea
and-go-ahead-group-use-f1-technology-improve-fuel-efficiency-london-buses (дата обращения: 17.0 and-go-ahead-group-use-f1-technology-improve-fuel-efficiency-london-buses (дата обращения: 17.05.2020). 9
Официальный сайт Общества производителей и продавцов автомобилей (Society of Motor Manufacturers
and Traders, SMMT), одной из старейших в своей отрасли (образована в 1902 г.) и наиболее влиятельных
торговых ассоциаций Великобритании. URL: https://www.smmt.co.uk/2015/06/gkn-hybrid-power-flywheels-
in-750-uk-buses/ (дата обращения: 17.05.2020). 9
Официальный сайт Общества производителей и продавцов автомобилей (Society of Motor Manufacturers
and Traders, SMMT), одной из старейших в своей отрасли (образована в 1902 г.) и наиболее влиятельных
торговых ассоциаций Великобритании. URL: https://www.smmt.co.uk/2015/06/gkn-hybrid-power-flywheels-
in-750-uk-buses/ (дата обращения: 17.05.2020). Актуальные проблемы российского права. 2021. Т. 16. № 2 (123) февраль 177 ИНТЕГРАЦИОННОЕ ПРАВО смена Технического регламента), оформленный
патент может уже не представлять актуальность
и интерес для правообладателя. Во-вторых, за-
крепление результата научного исследования в
форме патента предоставляет команде-право-
обладателю конкурентное преимущество, по-
скольку соперники не могут им воспользовать-
ся, а попытки создать собственный аналог могут
привести к нарушению прав. Казалось бы, это
обычное дело в любой другой отрасли, однако
важно помнить, что «Формула-1» — это прежде
всего спорт, правила которого вырабатываются
в том числе при участии самих команд. Иными
словами, если будет запатентована технология
или деталь, которая предоставляет одному
из коллективов значительное преимущество,
остальные команды на следующем обсужде-
нии правил (как правило, проходит в конце чем-
пионата) могут принять решение о запрете ее
использования в будущем10. Ценность такого
патента резко падает. В итоге команды «Фор-
мулы-1» патентуют лишь те детали и техно-
логии, которые можно будет использовать в
коммерческих целях при переносе на обычные
автомобили или прочую технику (как в случае
с системой KERS), и в том виде, в котором они
будут реализовываться. 10 Слова анонимного высокопоставленного инженера «Формулы-1» на сайте Motorsport.com. URL: https://
www.motorsport.com/f1/news/patents-in-f1-explained/3218781/?utm_source=jaonf1.com&utm_
medium=acquisition-redirect&utm_content=patents-in-f1-explained (дата обращения: 17.05.2020). 14 Motorsport.com. URL: https://www.motorsport.com/f1/news/why-x-wings-were-banned/4776778/
#gal-4776778-m0-johnny-herbert-sauber-petronas-c17 (дата обращения: 15.05.2020). 11 Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement) // Официальный сайт
Всемирной торговой организации (ВТО). URL: https://www.wto.org/english/docs_e/legal_e/31bis_trips_01_e.
htm (дата обращения: 17.05.2020). Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration and European Law Department, Kutafin
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com При
этом разрешается использование чехлов для
покрытия поврежденных автомобилей или
компонентов; прозрачного лотка для инструм-
ентов глубиной не более 50 мм, расположен-
ного сверху на заднем крыле; утепляющих или
теплозащитных чехлов для двигателя и коробки
передач на стартовой решетке; чехла на заднее
крыло, разработанного специально для защиты
запускающего автомобиль механика от огня23;
термочехлов для шин24; чехла для закрытия мы исключительно по механическим причинам
(например, в противопожарных целях). Запре-
щается также покрывать двигатели, коробки
передач или радиаторы во время замены или
перемещения двигателей по гаражу и запасные
крылья, когда они не используются и находятся
на стойке на пит-лейне. Запрещается исполь-
зовать в качестве препятствий и заграждений
такие детали, как запасные днища, топливные
установки или тележки для инструментов. При
этом разрешается использование чехлов для
покрытия поврежденных автомобилей или
компонентов; прозрачного лотка для инструм-
ентов глубиной не более 50 мм, расположен-
ного сверху на заднем крыле; утепляющих или
теплозащитных чехлов для двигателя и коробки
передач на стартовой решетке; чехла на заднее
крыло, разработанного специально для защиты
запускающего автомобиль механика от огня23;
термочехлов для шин24; чехла для закрытия совместно с Кембриджским университетом16),
а также воздухозаборник McLaren F-duct17. а также воздухозаборник McLaren F duct . Необходимо отметить, что режим коммер-
ческой тайны не означает, что конкуренты не
могут подойти к болиду и изучить его, в том
числе на предмет нестандартных инженерных
решений. Как отмечал Р. Браун, если команда
сможет сохранить в тайне от конкурентов свое
инновационное решение на протяжении трех —
четырех месяцев, особенно с учетом участия в
гонках, ей крупно повезло18. Как правило, конку-
ренты изучают автомобили друг друга во время
гоночных уикендов19, поскольку в соответствии
со ст. 21.4 Спортивного регламента20 в эти дни
не допускаются в любое время и в любом месте
(в паддоке21, в гаражах, на пит-лейне22 или на
стартовой решетке) никакие экраны, крышки
или другие препятствия, которые каким-либо
образом заслоняют какую-либо часть автомо-
биля, за исключением тех, которые необходи- 15 Официальный сайт Moore Good Ink. URL: https://mooregoodink.com/how-renault-f1-found-faster-lap-times-
and-won-a-world-championshIP-by-creating-the-tuned-mass-damper/ (дата обращения: 15.05.2020). 15 Официальный сайт Moore Good Ink. URL: https://mooregoodink.com/how-renault-f1-found-faster-lap-times-
and-won-a-world-championshIP-by-creating-the-tuned-mass-damper/ (дата обращения: 15.05.2020). 16 См.: Smith M. C. Synthesis of mechanical networks: the inerter // IEEE Transactions on Automatic Control. 2002. Vol. 47. No 10. P. 1648–1662. 17 Autosport.com. URL: https://www.autosport.com/f1/news/149090/banned-the-f1-2010-season-fduct (дата
обращения: 17.05.2020). 18 См.: Jenkins M., Pasternak K., West R. Performance at the Limit. Cambridge University Press, 2016. Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration and European Law Department, Kutafin
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com К чис-
лу таких примеров можно отнести Х-крылья
(X-wings) британской команды Tyrell14, демпфер
масс Renault15 и инертер McLaren (разработан Отказ от патента как способа защиты субъ-
ективных прав на результаты научных иссле-
дований не означает, что они вообще никак не
защищаются. Чаще всего команды «Формулы-1» 10 Слова анонимного высокопоставленного инженера «Формулы-1» на сайте Motorsport.com. URL: https://
www.motorsport.com/f1/news/patents-in-f1-explained/3218781/?utm_source=jaonf1.com&utm_
medium=acquisition-redirect&utm_content=patents-in-f1-explained (дата обращения: 17.05.2020). 10 Слова анонимного высокопоставленного инженера «Формулы-1» на сайте Motorsport.com. URL: https://
www.motorsport.com/f1/news/patents-in-f1-explained/3218781/?utm_source=jaonf1.com&utm_
medium=acquisition-redirect&utm_content=patents-in-f1-explained (дата обращения: 17.05.2020). 11 Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement) // Официальный сайт
Всемирной торговой организации (ВТО). URL: https://www.wto.org/english/docs_e/legal_e/31bis_trips_01_e. htm (дата обращения: 17.05.2020). 11 Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement) // Официальный сайт
Всемирной торговой организации (ВТО). URL: https://www.wto.org/english/docs_e/legal_e/31bis_trips_01_e. htm (дата обращения: 17.05.2020). 12 Directive (EU) 2016/943 of the European Parliament and of the Council of 8 June 2016 on the protection of
undisclosed know-how and business information (trade secrets) against their unlawful acquisition, use and
disclosure (Text with EEA relevance) // OJ L 157. 15.06.2016. P. 1–18. 13 СЗ РФ 2004 №32 Ст 3283 12 Directive (EU) 2016/943 of the European Parliament and of the Council of 8 June 2016 on the protection of
undisclosed know-how and business information (trade secrets) against their unlawful acquisition, use and
disclosure (Text with EEA relevance) // OJ L 157. 15.06.2016. P. 1–18. 13 СЗ РФ 2004 №32 Ст 3283 14 Motorsport.com. URL: https://www.motorsport.com/f1/news/why-x-wings-were-banned/4776778/
#gal-4776778-m0-johnny-herbert-sauber-petronas-c17 (дата обращения: 15.05.2020). Актуальные проблемы российского права. 2021. Т. 16. № 2 (123) февраль 178 Камалян А. М. Камалян А. М. Охрана результатов научных исследований и научной информации в спорте высших достижений на примере «Формулы-1» Камалян А. М. Охрана результатов научных исследований и научной информации в спорте высших достижений на примере «Формулы-1» мы исключительно по механическим причинам
(например, в противопожарных целях). Запре-
щается также покрывать двигатели, коробки
передач или радиаторы во время замены или
перемещения двигателей по гаражу и запасные
крылья, когда они не используются и находятся
на стойке на пит-лейне. Запрещается исполь-
зовать в качестве препятствий и заграждений
такие детали, как запасные днища, топливные
установки или тележки для инструментов. Актуальные проблемы российского права. 2021. Т. 16. № 2 (123) февраль 15 Официальный сайт Moore Good Ink. URL: https://mooregoodink.com/how-renault-f1-found-faster-lap-times-
and-won-a-world-championshIP-by-creating-the-tuned-mass-damper/ (дата обращения: 15.05.2020). Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration and European Law Department, Kutafin
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com P 19 Гран-при «Формулы-1» проводится в течение трех дней, как правило, с пятницы по воскресенье (исклю-
чение составляет Гран-при Монако, где традиционно гоночный уикенд начинается в четверг и продол-
жается в субботу и воскресенье). В пятницу (для Монако — четверг) проводятся две сессии свободных
заездов, в субботу — третья сессия свободных заездов и квалификация, в воскресенье — гонка. 20 Спортивный регламент «Формулы-1» (Formula One Sporting Regulations) в ред. от 07.04.2020 // Офи-
циальный сайт Международной федерации автоспорта (FIA). URL: https://www.fia.com/sites/default/
files/2020_formula_1_sporting_regulations_-_iss_6_-_2020-04-07.pdf (дата обращения: 17.05.2020). 21 Как и ряд других понятий, термин «паддок» (англ. paddock) пришел в «Формулу-1» из скачек, где озна-
чает расположенный рядом со стойлом загон. В автоспорте под ним понимают специальное место для
развертывания команд недалеко от гаражей (боксов). 22 Пит-лейн представляет собой часть гоночной трассы, где располагаются гаражи (боксы) команд, участ-
вующих в гонке, и производятся пит-стопы — остановки для замены колес и/или поврежденных деталей. При заезде на пит-лейн гонщик обязан снизить скорость и соблюдать скоростной режим вплоть до выезда
из него. 23 Несмотря на применение огромного количества высоких технологий в «Формуле-1», двигатель болида
запускается не гонщиком, а группой механиков. Один из механиков, который непосредственно запускает
двигатель, располагается позади автомобиля. 23 Несмотря на применение огромного количества высоких технологий в «Формуле-1», двигатель болида
запускается не гонщиком, а группой механиков. Один из механиков, который непосредственно запускает
двигатель, располагается позади автомобиля. 24 Термочехлы планировалось запретить с переходом на новый Технический регламент в 2021 (2022) г.,
однако, учитывая и так довольно большое число изменений, их решили оставить с последующим запре-
том // Autosport.com. URL: https://www.autosport.com/f1/news/146522/f1-2021-tyre-blanket-ban-decision-
reversed (дата обращения: 17.05.2020). 24 Термочехлы планировалось запретить с переходом на новый Технический регламент в 2021 (2022) г.,
однако, учитывая и так довольно большое число изменений, их решили оставить с последующим запре-
том // Autosport.com. URL: https://www.autosport.com/f1/news/146522/f1-2021-tyre-blanket-ban-decision-
reversed (дата обращения: 17.05.2020). Актуальные проблемы российского права. 2021. Т. 16. № 2 (123) февраль 179 ИНТЕГРАЦИОННОЕ ПРАВО болида на ночь в режиме закрытого парка25,
а также любых приспособлений для укрытия
автомобиля от дождя на пит-лейне или на стар-
товой решетке. Иными словами, в ходе Гран-при
болиды максимально открыты для посторонних
глаз, прежде всего для зрителя. получился настолько похожим, что соперники
заподозрили Racing Point в нарушении правил,
но доказательства представить не смогли26. И тем не менее в «Формуле-1» все же возни-
кают конфликтные ситуации, когда именно сами
технологии (а не просто их внешнее выражение)
оказываются в руках конкурентов. Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration and European Law Department, Kutafin
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com Режим ком-
мерческой тайны, безусловно, более гибкий для
спорта, нежели патент, но у него есть один суще-
ственный недостаток — не все люди умеют оди-
наково хорошо хранить секреты. В этом контек-
сте одной из ключевых проблем является уход
членов персонала из команды «Формулы-1», в
том числе переход к конкурентам. До недавних пор такая открытость была толь-
ко на этапах Гран-при. На предсезонных тестах,
где новые болиды проезжают свои первые
километры на трассах перед началом сезона,
команды пользовались многочисленными при-
способлениями для закрытия автомобиля от по-
сторонних глаз (чаще всего использовали огром-
ные плакаты или экраны, которые полностью
загораживали боксы). Однако перед началом
сезона-2020 в ст. 10.5 Спортивного регламента
внесли изменения, запрещающие любые огра-
ждения с 09:00 до 18:00 на время проведения
предсезонных тестов с рядом дополнительных
изъятий. В частности, заграждения могут при-
меняться при ремонте и восстановлении авто-
мобиля, поврежденного в ходе тестов, а также
если днище болида не установлено. Переход сотрудников из одного коллектива
в другой довольно часто (и практически всегда,
если речь идет о высокопоставленном сотруд-
нике) сопровождается довольно продолжитель-
ным отпуском с сохранением заработной платы
и без права выхода на работу с тем, чтобы не
допустить утечку важной информации (коммер-
ческой тайны) в стан конкурентов. Например, в
июле 2018 г. команда McLaren подтвердила на-
значение на должность технического директора
Дж. Ки27, который занимал аналогичную долж-
ность в Toro Rosso, однако приступить к своим
новым обязанностям британский инженер смог
лишь 25 марта 2019 г.28, то есть спустя почти
восемь месяцев. К пилотам подобное правило
почти не применяется, поскольку они сами не
участвуют в разработке болида, а высказывают
свои пожелания. Однако если в середине сезона
уже известно, что по окончании чемпионата гон-
щик покинет коллектив и перейдет в команду
соперника, то по очевидным причинам его
перестают привлекать к участию в разработке Важно понимать, что если одна команда
подсмотрела у своих конкурентов инженерное
решение, она может воспользоваться лишь его
внешним выражением. Какой бы то ни был
доступ к чертежам, технической документации
и прочим документам, являющимся интел-
лектуальной собственностью, не может быть
предоставлен, это строго запрещено. Именно
таким образом и был сконструирован болид
британского коллектива Racing Point для уча-
стия в сезоне-2020, инженеры которого руко-
водствовались фотографиями с различных гонок
машины Mercedes образца 2019 г. Автомобиль 25 Закрытый парк (фр. parc fermé) — специально отведенное место на автодроме, где действует особый
режим работы с болидами, препятствующий их обслуживанию, за исключением особо оговоренных слу-
чаев и процедур. Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration and European Law Department, Kutafin
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com Машины переводятся в закрытый парк непосредственно после квалификаций и гонок
для проведения инспекции с целью убедиться в их соответствии требованиям технического регламента. Такой же режим устанавливается с момента выезда из боксов на стартовую решетку, когда разрешается
вносить лишь незначительные изменения в настройки болида. 25 Закрытый парк (фр. parc fermé) — специально отведенное место на автодроме, где действует особый
режим работы с болидами, препятствующий их обслуживанию, за исключением особо оговоренных слу-
чаев и процедур. Машины переводятся в закрытый парк непосредственно после квалификаций и гонок
для проведения инспекции с целью убедиться в их соответствии требованиям технического регламента. Такой же режим устанавливается с момента выезда из боксов на стартовую решетку, когда разрешается
вносить лишь незначительные изменения в настройки болида. 26 Официальный сайт «Формулы-1». URL: https://www.formula1.com/en/latest/article.we-absolutely-designed-
it-ourselves-racing-point-defend-rp20-after-mercedes.41SgWpnq6ytiHOAMvTCr8x.html (дата обращения:
17.05.2020). 26 Официальный сайт «Формулы-1». URL: https://www.formula1.com/en/latest/article.we-absolutely-designed-
it-ourselves-racing-point-defend-rp20-after-mercedes.41SgWpnq6ytiHOAMvTCr8x.html (дата обращения:
17.05.2020). 27 Autosport.com. URL: https://www.autosport.com/f1/news/137623/toro-rosso-tech-chief-key-to-move-to-
mclaren (дата обращения: 17.05.2020). 28 Autosport.com. URL: https://www.autosport.com/f1/news/141698/key-to-join-mclaren-after-2019-melbourne-
race (дата обращения: 17.05.2020). Актуальные проблемы российского права. 2021. Т. 16. № 2 (123) февраль 180 Камалян А. М. Камалян А. М. Охрана результатов научных исследований и научной информации в спорте высших достижений на примере «Формулы-1» Охрана результатов научных исследований и научной информации в спорте высших достижений на примере «Формулы-1» болида следующего года, и всё реже он участ-
вует в командных собраниях (чем ближе конец
сезона, тем больше брифингов посвящаются
подготовке к следующему чемпионату). Renault, результатом рассмотрения которого ста-
ло исключение обоих гонщиков французского
коллектива из итогового протокола Гран-при
Японии31. Один из самых громких скандалов32, связан-
ных с раскрытием коммерческой тайны, про-
изошел в 2007 г. и получил название «спай-
гейт» (Spygate), или Степнигейт (Stepneygate,
по фамилии инженера, ставшего центральной
фигурой). Н. Степни являлся частью «команды
мечты»33 Ferrari, которая привела команду из
Маранелло к шести чемпионским титулам под-
ряд (ранее эта же группа единомышленников
завоевала еще два титула чемпионов мира в
составе команды Benneton). После заверше-
ния карьеры М. Шумахера в 2006 г. в высших
эшелонах итальянского коллектива пошли
перестановки (в частности, команду покинул
технический директор Р. Браун), которые при-
шлись не по душе Н. Степни34. В результате
перед началом сезона 2007 г. он занял долж-
ность руководителя отдела развития и перестал
принимать непосредственное участие в работе
команды на Гран-при. Недовольный положе-
нием вещей, Н. Степни саботировал работу
команды (на Гран-при Монако в баках Ferrari
был обнаружен белый порошок, который, по
словам очевидцев, пронес именно Н. 29 См.: Kolasa M. Trade Secrets and Employee Mobility: In Search of an Equilibrium. Cambridge University Press,
2018. P. 131. 30 Force India Formula One Team Limited v. Aerolab srl [2013] EWCA Civ 780. Court of Appeal // Reports of Patent,
Design and Trade Mark Cases. 2013. Vol. 130. Iss. 12. P. 947–984.
31 Официальный сайт «Формулы-1». URL: https://www.formula1.com/en/latest/article.renault-disqualified-
from-results-of-the-japanese-grand-prix.5lDaMU1SWKaXnpEP93CLVB.html (дата обращения: 17.05.2020).
32 См.: Jenkins M., Pasternak K., West R. Op. cit. P. 78.
33 Помимо главного механика Н. Степни, в нее входили гонщик М. Шумахер, руководитель команды Ж. Тодт,
технический директор Р. Браун и главный конструктор Р. Бирн.
34 Autosport com URL: https://www autosport com/f1/news/56539/unhappy stepney wants a sabbatical (дата Актуальные проблемы российского права. 2021. Т. 16. № 2 (123) февраль 29 См.: Kolasa M. Trade Secrets and Employee Mobility: In Search of an Equilibrium. Cambridge University Press,
2018. P. 131.
30 Force India Formula One Team Limited v. Aerolab srl [2013] EWCA Civ 780. Court of Appeal // Reports of Patent,
Design and Trade Mark Cases. 2013. Vol. 130. Iss. 12. P. 947–984.
31 Официальный сайт «Формулы-1». URL: https://www.formula1.com/en/latest/article.renault-disqualified-
from-results-of-the-japanese-grand-prix.5lDaMU1SWKaXnpEP93CLVB.html (дата обращения: 17.05.2020).
32 См.: Jenkins M., Pasternak K., West R. Op. cit. P. 78.
33 Помимо главного механика Н. Степни, в нее входили гонщик М. Шумахер, руководитель команды Ж. Тодт,
технический директор Р. Браун и главный конструктор Р. Бирн.
34 Autosport.com. URL: https://www.autosport.com/f1/news/56539/unhappy-stepney-wants-a-sabbatical (дата
обращения: 17.05.2020).
35 Официальный сайт The Guardian. URL: https://www.theguardian.com/sport/2007/jul/05/motorsports.sport3
(дата обращения: 17.05.2020). БИБЛИОГРАФИЯ kins M., Pasternak K., West R. Performance at the Limit. — Cambridge University Press, 2016. — 266 p. 1. Jenkins M., Pasternak K., West R. Performance at the Limit. — Cambridge University Press, 2016. — 266 p. 2. Kolasa M. Trade Secrets and Employee Mobility: In Search of an Equilibrium. — Cambridge University Press,
2018. — 408 p. 1. Jenkins M., Pasternak K., West R. Performance at the Limit. — Cambridge University Press, 2016. — 266 p. 2. Kolasa M. Trade Secrets and Employee Mobility: In Search of an Equilibrium. — Cambridge University Press,
2018. — 408 p. ,
,
g
y
,
p
2. Kolasa M. Trade Secrets and Employee Mobility: In Search of an Equilibrium. — Cambridge University Press,
2018. — 408 p. 3. Smith M. C. Synthesis of mechanical networks: the inerter // IEEE Transactions on Automatic Control. —
2002. — Vol. 47. — № 10. 3. Smith M. C. Synthesis of mechanical networks: the inerter // IEEE Transactions on Automatic Control. —
2002. — Vol. 47. — № 10. 3. Smith M. C. Synthesis of mechanical networks: the inerter // IEEE Transactions on Automatic Control. —
2002. — Vol. 47. — № 10. Материал поступил в редакцию 17 мая 2020 г. Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration and European Law Department, Kutafin
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com тате дорожно-транспортного происшествия в
2014 г.) и денежному штрафу, факт промышлен-
ного шпионажа был установлен и подтвержден,
команда McLaren лишилась всех набранных
очков и была исключена из Кубка конструкто-
ров (при этом пилоты команды сохранили свои
баллы в личном зачете), а также была оштра-
фована на 100 млн долл. США с обязательством
предоставить болид перед началом нового се-
зона в FIA для более детального изучения на
предмет заимствования технологий Ferrari36. Требование о допуске болида McLaren к уча-
стию в чемпионате 2008 г. только по итогам его
тщательной проверки позже было отменено в
связи с принесением британским коллективом
публичных извинений37, хотя ряд мероприятий
к тому моменту уже успели состояться38. тате дорожно-транспортного происшествия в
2014 г.) и денежному штрафу, факт промышлен-
ного шпионажа был установлен и подтвержден,
команда McLaren лишилась всех набранных
очков и была исключена из Кубка конструкто-
ров (при этом пилоты команды сохранили свои
баллы в личном зачете), а также была оштра-
фована на 100 млн долл. США с обязательством
предоставить болид перед началом нового се-
зона в FIA для более детального изучения на
предмет заимствования технологий Ferrari36. Требование о допуске болида McLaren к уча-
стию в чемпионате 2008 г. только по итогам его
тщательной проверки позже было отменено в
связи с принесением британским коллективом
публичных извинений37, хотя ряд мероприятий
к тому моменту уже успели состояться38. Таким образом, охрана результатов научных
исследований в «Формуле-1» обладает опре- 36 Autosport.com. URL: https://www.autosport.com/f1/news/87071/stepney-sentenced-over-spy-affair (дата
обращения: 17.05.2020). 38 Autosport.com. URL: https://www.autosport.com/f1/news/63844/fia-begins-probe-of-mclaren-2008-car (дата
обращения: 17.05.2020). 37 Официальный сайт The Guardian. URL: https://www.theguardian.com/sport/2007/dec/14/motorsports.sport
(дата обращения: 17.05.2020). Artur M. Kamalyan, Cand. Sci. (Law), Lecturer, Integration and European Law Department, Kutafin
Moscow State Law University (MSAL)
ul. Sadovaya-Kudrinskaya, d. 9, Moscow, Russia, 125993
arthur.kamalyan@gmail.com Степ-
ни35) и передавал конфиденциальную инфор-
мацию (речь шла о документе объемом около
800 страниц) своему другу, высокопоставленно-
му инженеру из McLaren М. Кохлену. По итогам
судебного разбирательства и расследования
FIA Н. Степни был осужден итальянским судом
к 20 месяцам лишения свободы (наказание не
отбывал, в том числе ввиду гибели в резуль- В ситуации с переходом конструкторов и
инженеров нередко возникает вопрос: как
отграничить навыки и квалификацию от ком-
мерческой тайны, которую удалось запомнить
(неважно, намеренно или нет)? Своеобразным
водоразделом может служить контекст, при
котором сотрудник приобрел знания. Когда
такие знания были получены в ходе выполне-
ния деятельности у предыдущего работодателя
и неразрывно связаны с опытом, скорее всего,
речь следует вести о навыках, в отличие от ситуа-
ции, когда лицо получило информацию путем
ознакомления с документами или из разговоров
с коллегами, то есть за рамками выполнения
должностных обязанностей. Не менее важно,
сообщил ли подобные сведения работодатель
или работник сам получил эти знания29. В любом
случае, тот факт, что работник смог запомнить
какие-то сведения, не означает сам по себе, что
полученная информация перестала относиться к
коммерческой тайне и стала навыком30. В «Формуле-1» нередко случались ситуации,
когда сотрудники после перехода раскрывали
некоторые секреты своих предыдущих рабо-
тодателей. Один из недавних примеров свя-
зан с переходом в 2019 г. инженера Renault в
команду Racing Point, который раскрыл новому
работодателю принцип работы системы смеще-
ния баланса тормозов, которую использовала
французская «конюшня». Когда FIA запретила
британскому коллективу использовать подоб-
ную систему, Racing Point подала протест против 29 См.: Kolasa M. Trade Secrets and Employee Mobility: In Search of an Equilibrium. Cambridge University Press,
2018 P 131 35 Официальный сайт The Guardian. URL: https://www.theguardian.com/sport/2007/jul/05/motorsports.sport3
(дата обращения: 17.05.2020). Актуальные проблемы российского права. 2021. Т. 16. № 2 (123) февраль 181 ИНТЕГРАЦИОННОЕ ПРАВО деленной спецификой. Чаще всего команды
добровольно отказываются от патентования
своих изобретений, поскольку процедура зани-
мает слишком много времени для такого дина-
мично развивающегося спорта, а также есть
высокая вероятность, что конкуренты могут
запретить использование запатентованной
технологии. Там, где патенты все же были полу-
чены, правообладателями не всегда являются
гоночные коллективы, ими могут быть иные
юридические лица, которые непосредствен-
но связаны с командой, что может затруднить
процесс поиска охраняемых патентами изо-
бретений. В основном же результаты научных
исследований в «Формуле-1» охраняются ре-
жимом коммерческой тайны, который может
нарушаться недовольными сотрудниками и/
или членами персонала, которые перешли к
конкурентам. Актуальные проблемы российского права. 2021. Т. 16. № 2 (123) февраль REFERENCES (TRANSLITERATION) 1. Jenkins M., Pasternak K., West R. Performance at the Limit. — Cambridge University Press, 2016. — 266 p. 2. Kolasa M. Trade Secrets and Employee Mobility: In Search of an Equilibrium. — Cambridge University Press,
2018. — 408 p. 2. Kolasa M. Trade Secrets and Employee Mobility: In Search of an Equilibrium. — Cambridge University Press,
2018. — 408 p. 3. Smith M. C. Synthesis of mechanical networks: the inerter // IEEE Transactions on Automatic Control. —
2002. — Vol. 47. — № 10. 3. Smith M. C. Synthesis of mechanical networks: the inerter // IEEE Transactions on Automatic Control. —
2002. — Vol. 47. — № 10. 182
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.